Page 1

Format: 215 × 280 mm

1 Einheit: 7 mm

2/19 ACTA ORTHOPAEDICA

Preventing and reducing periprosthetic joint infections

For anaerobes

Bone cement with 2 antibiotics Broad spectrum of activity For MRSA/MRSE

Infection prevention Implant fixation High stability

High initial antibiotic release

Revision

www.heraeus-medical.com 1810_15791_AZ_COPAL_Produktfamilie_Acta_Orthop_Europe_215x280_EN.indd 1

Vol. 90, No. 2, 2019 (pp. 97–190)

PRODUCTS & SOLUTIONS YOU CAN TRUST.

Volume 90, Number 2, April 2019

ISSN 1745-3674

13.03.19 12:31


Acta Orthopaedica is owned by the Nordic Orthopaedic Federation and is the official publication of the Nordic Orthopaedic Federation

E DITORIAL O F FICE

Acta Orthopaedica Department of Orthopedics Lund University Hospital SE–221 85 Lund, Sweden E-mail: acta.ort@med.lu.se Homepage: http://www.actaorthop.org

EDITOR

THE FOUNDATION BOARD OF

Anders Rydholm Lund, Sweden

THE NORDIC O RTHOPAEDIC F EDERATION AND A CTA O RTHOPAEDICA

DEPUTY EDITOR

Peter A Frandsen Odense, Denmark CO-EDITORS

Li Felländer-Tsai Stockholm, Sweden Nils Hailer Uppsala, Sweden Ivan Hvid Oslo, Norway Urban Rydholm Lund, Sweden Bart A Swierstra Wageningen, The Netherlands Eivind Witsø Trondheim, Norway Rolf Önnerfält Lund, Sweden

Peter Frandsen Denmark Ragnar Jonsson Iceland Heikki Kröger Finland Anders Rydholm Sweden Kees Verheyen the Netherlands

WEB EDITOR

Magnus Tägil Lund, Sweden S TATISTICAL EDITOR

Jonas Ranstam Lund, Sweden P RODUCTION MANAGER

Kaj Knutson Lund, Sweden

Vol. 90, No. 2, 2019


SUBSCRIPTION INFORMATION Acta Orthopaedica [print 1745-3674, online 1745-3682] is a peerreviewed journal, published six times a year plus supplements by Taylor & Francis on behalf of Nordic Orthopaedic Federation. Annual Institutional Subscription, Volume 89, 2018 $1,173 £725 €939 The subscription fee purchases an online subscription. The price includes access to current content and back issues to January 1997 (if available). Printed copies of the journal are provided on request as a free supplementary service accompanying an online subscription. Supplements to the journal are also included in the subscription price. For more information, visit the journal’s website: http://www.tandfonline.com/IORT Manuscripts should be uploaded at http://www.manuscriptmanager.com/ao/ for further handling at: Acta Orthopaedica Editorial Office, Department of Orthopaedics, Lund University Hospital, SE-221 85 Lund, Sweden Correspondance concerning copyright and permissions should be sent to: Håkan Pårup P.O. Box 3255, SE-103 65 Stockholm, Sweden, Tel: +46 (0)8 440 80 40. Fax: +46 (0)8 440 80 50. E-mail: hakan.parup@informa.com. Ordering information: Please contact your local Customer Service Department to take out a subscription to the Journal: USA, Canada: Taylor & Francis, Inc., 530 Walnut Street, Suite 850, Philadelphia, PA 19106, USA. Tel: +1 800 354 1420; Fax: +1 215 207 0050. UK/ Europe/Rest of World: T&F Customer Services, Informa UK Ltd, Sheepen Place, Colchester, Essex, CO3 3LP, United Kingdom. Tel: +44 (0) 20 7017 5544; Fax: +44 (0) 20 7017 5198; Email: subscriptions@tandf.co.uk Dollar rates apply to all subscribers outside of Europe. Euro rates apply to all subscribers in Europe except the UK and Republic of Ireland. If you are unsure which applies, contact Customer Services. All subscriptions are payable in advance and all rates include postage. Journals are sent by air to the USA, Canada, Mexico, India, Japan and Australasia. Subscriptions are entered on an annual basis, i.e., January to December. Payment may be made by sterling check, US dollar check, euro check, international money order, National Giro, or credit card (Amex, Visa and Mastercard). Back issues: Taylor & Francis retains a two-year back issue stock of journals. Older volumes are held by our official stockists to whom all orders and enquiries should be addressed: Periodicals Service Company, 351 Fairview Ave., Suite 300, Hudson, New York 12534, USA. Tel: +1 518 537 4700; fax: +1 518 537 5899; e-mail: psc@ periodicals.com.

Airfreight and mailing in the USA by agent named Air Business Ltd, c/o Worldnet Shipping Inc., 156-15, 146th Avenue, 2nd Floor, Jamaica, NY 11434, USA. Periodicals postage paid at Jamaica NY 11431. US Postmaster: Send address changes to Acta Orthopaedica, Air Business Ltd, c/o Worldnet Shipping Inc., 156-15, 146th Avenue, 2nd Floor, Jamaica, NY 11434, USA. Subscription records are maintained at Taylor & Francis Group, 4 Park Square, Milton Park, Abingdon, OX14 4RN, United Kingdom.

Copyright © 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by-nc/3.0 . Informa UK Limited, trading as Taylor & Francis Group makes every effort to ensure the accuracy of all the information (the “Content”) contained in its publications. However, Informa UK Limited, trading as Taylor & Francis Group, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Informa UK Limited, trading as Taylor & Francis Group. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Informa UK Limited, trading as Taylor & Francis Group shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. Terms & Conditions of access and use can be found at http://www.tandfonline. com/page/terms-and-conditions Indexed/abstracted in: Allied and Complementary Medicine Library (Amed); ASCA (Automatic Subject Citation Alert); Biological Abstracts; Chemical Abstracts; Cumulative Index to Nursing and Allied Health Literature(CINAHL); Current Advances in Ecological and Environmental Sciences; Current Contents/Clinical Medicine; Current Contents/Life Sciences; Developmental Medicine and Child Neurology; Energy Research Abstracts; EMBASE/ Excerpta Medica; Faxon Finder; Focus On: Sports Science & Medicine; Health Planning and Administration; Index Medicus/MEDLINE; Index to Dental Literature; Index Veterinarius; INIS Atomindex; Medical Documentation Service; Nuclear Science Abstracts (Ceased); Periodicals Scanned and Abstracted. Life Sciences Collection; Research Alert; Science Citation Index; SciSearch; SportSearch; Uncover Veterinary Bulletin. Printed in England by Henry Ling


Acta Orthopaedica

ISSN 1745-3674

Vol. 90, No. 2, April 2019 Infection Target site antibiotic concentrations in orthopedic/trauma extremity surgery: is prophylactic cefazolin adequately dosed? A systematic review and meta-analysis A comparative study of intraoperative frozen section and alpha defensin lateral flow test in the diagnosis of periprosthetic joint infection Upper extremity Clinical significance of cervical MRI in brachial plexus birth injury Poor patient-reported outcome after shoulder replacement in young patients with cuff-tear arthropathy: a matched-pair analysis from the Danish Shoulder Arthroplasty Registry Early palmar plate fixation of distal radius fractures may benefit patients aged 50 years or older: a randomized trial comparing 2 different treatment protocols Patient-reported outcomes after a distal radius fracture in adults: a 3–4 years follow-up Hip The design of the cemented stem influences the risk of Vancouver type B fractures, but not of type C: an analysis of 82,837 Lubinus SPII and Exeter Polished stems Varying but reduced use of postoperative mobilization restrictions after primary total hip arthroplasty in Nordic countries: a questionnaire-based study An international comparison of THA patients, implants, techniques, and survivorship in Sweden, Australia, and the United States High annual surgeon volume reduces the risk of adverse events following primary total hip arthroplasty: a registry-based study of 12,100 cases in Western Sweden Estonian hip fracture data from 2009 to 2017: high rates of nonoperative management and high 1-year mortality

97

F R K Sanders, J C Goslings, R A A Mathôt, and T Schepers

105

I K Sigmund, J Holinka, S Lang, S Stenicka, K Staats, G Hobusch, B Kubista, and R Windhager

111 119

P Grahn, T Pöyhiä, A Sommarhem, and Y Nietosvaara M Ammitzboell, A Baram, S Brorson, B S Olsen, and J V Rasmussen

123

K Sirniö, J Leppilahti, P Ohtonen, and T Flinkkilä

129

R H Van Leerdam, F Huizing, F Termaat, S Kleinveld, S J Rhemrev, P Krijnen, and I B Schipper

135

G Chatziagorou, H Lindahl, and J Kärrholm

143

K Gromov, A Troelsen, M Modaddes, O Rolfson, O Furnes, G Hallan, A Eskelinen, P Neuvonen, and H Husted

148

E W Paxton, G Cafri, S Nemes, M Lorimer, J Kärrholm, H Malchau, S E Graves, R S Namba, and O Rolfson P Jolbäck, O Rolfson, P Cnudde, D Odin, H Malchau, H Lindahl, and M Mohaddes

153 159

P Prommik, H Kolk, P Sarap, E Puuorg, E Harak, A Kukner, M Pääsuke, and A Märtson

165

M Stilling, I Mechlenburg, C F Jepsen, L Rømer, O Rahbek, K Søballe, and F Madsen

172

E K Laende, J L Astephen Wilson, J Mills Flemming, E R Valstar, C G Richardson, and M J Dunbar

179

Y-H Pua, C L-L Poon, F J-T Seah, J Thumboo, R A Clark, M-H Tan, H-C Chong, J W-M Tan, E S-X Chew, and S-J Yeo

Case report Custom-made asymmetric polyethylene liner to correct tibial component malposition in total knee arthroplasty — a case report

187

A Kappel, C S Blom, and A El-Galaly

Correspondence Early recovery trajectories after fast-track primary total hip arthroplasty: the role of patient characteristics

190

T Bandholm versus J Porsius

Knee Superior fixation and less periprosthetic stress-shielding of tibial components with a finned stem versus an I-beam block stem: a randomized RSA and DXA study with minimum 5 years’ followup Equivalent 2-year stabilization of uncemented tibial component migration despite higher early migration compared with cemented fixation: an RSA study on 360 total knee arthroplasties Predicting individual knee range of motion, knee pain, and walking limitation outcomes following total knee arthroplasty

Information to authors (see http://www.actaorthop.org/)


Acta Orthopaedica 2019; 90 (2): 97–104

97

Target site antibiotic concentrations in orthopedic/trauma extremity surgery: is prophylactic cefazolin adequately dosed? A systematic review and meta-analysis Fay R K SANDERS 1, J Carel GOSLINGS 2, Ron A A MATHÔT 3, and Tim SCHEPERS 1 1 Trauma

Unit, Department of Surgery, Amsterdam UMC, University of Amsterdam; 2 Department of Surgery, OLVG, Amsterdam; 3 Department of Hospital Pharmacy, Amsterdam UMC, University of Amsterdam, the Netherlands Correspondence: t.schepers@amc.nl Submitted 2018-09-21. Accepted 2019-01-11.

Background and purpose — The incidence of surgical site infections (SSIs) in trauma/orthopedic surgery varies between different body parts. Antibiotic prophylaxis (e.g., with cefazolin) lowers infection rates in closed fracture surgery and in primary arthroplasty. For prophylactic antibiotics to prevent infections, sufficient concentrations at the target site (location of surgery) are required. However, dosage recommendations and the corresponding efficacy are unclear. This review assesses target site cefazolin concentrations and the effect of variation in dose and location of target site during orthopedic extremity surgery. Methods — For this meta-analysis and systematic review, the literature was searched using the following keywords: “cephalosporins,” “orthopedic,” “extremity,” “surgical procedures,” and “pharmacokinetics”. Trials measuring target site antibiotic concentrations (bone, soft tissue, synovia) during orthopedic surgery after a single dose of cefazolin were included. Results — The search identified 14 studies reporting on concentrations in the shoulder (n = 1), hip (n = 8), knee (n = 8), or foot (n = 1). A large variation was seen between studies, but the pooled results of 4 studies showed higher concentrations in hip than in knee (mean difference: 4 ug/g, 95% CI 0.8–7). Articles comparing different doses of cefazolin reported higher bone concentrations after 2 g than before, but pooling results did not lead to a statistically significant difference. Interpretation — Although not all results could be pooled, this study shows that cefazolin concentrations are higher in the hip than in the knee. These findings suggest that the dose of prophylactic cefazolin might not be sufficient in distal parts of the extremity. Further research should investigate whether a higher dose of cefazolin can lead to higher concentrations and fewer SSIs.

A surgical site infection (SSI) is one of the most common complications of extremity surgery, especially when implants are involved. The infection rate ranges from 1.3% to 10% in hip and knee procedures (Agodi et al. 2017, De Jong et al. 2017) to 12% to 25% (Backes et al. 2014, Feilmeier et al. 2014, Wiewiorski et al. 2015) in foot and ankle surgery. Antibiotic prophylaxis is widely used and has been shown to lower infection rates in closed fracture surgery (Burnett et al. 1980, Boxma et al. 1996), as well as in primary arthroplasty (AlBuhairan et al. 2008, Voigt et al. 2015). Because of their broad-spectrum effect on methicillin-sensitive staphylococci and streptococci and relatively low costs, first-generation cephalosporins (e.g., cefazolin, cephradine, or cephalexin) are the recommended prophylactics in orthopedic/trauma surgery (Mangram et al. 1999, Bauer et al. 2017). However, there is limited evidence to support dosage recommendations in this field. The studies that form the foundation for the dosage, as mentioned in several international guidelines on surgical prophylaxis, do not include patients undergoing fracture/implant surgery (Bratzler et al. 2013, Moine and Fish 2013, Brill et al. 2014). For prophylactic antibiotics to prevent infections it is necessary to achieve concentrations that exceed the minimum inhibitory concentration (MIC) of the targeted pathogen for at least the time between incision and closure of the wound (Burke 1961). The MIC is the serum concentration that an antibiotic should exceed to inhibit a certain pathogen (e.g., MIC of cefazolin for S. aureus is 0.5–2 µg/L, meaning that a cefazolin concentration in serum of 0.5–2 ug/L is necessary for adequate inhibition of S. aureus). Because drugs are not evenly distributed through the body, it is important to know that an antibiotic achieves sufficient concentrations not only in serum but also at the target site (location of surgery) (Müller et al. 2004). Data on target-site concentrations of cefazolin during orthopedic surgery of the extremities could provide us

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2019.1577014


98

with the necessary information to assess and improve the efficacy of prophylactic cefazolin. The aim of this systematic review was to answer the following questions: 1. What are the target-site concentrations of cefazolin in the extremities during orthopedic surgery? 2. What is the influence of location of the target site and dose of cefazolin on the target site concentration?

Methods Search strategy and criteria This meta-analysis and systematic review was performed according to the PRISMA statement (Moher et al. 2009) and registered in PROSPERO (no. CRD42018093697). A search was performed in MedLine (PubMed), EMBASE (Ovid), and the Cochrane Library. A clinical librarian was consulted on the search strategy. The search included the following keywords: “cephalosporins,” “orthopedic,” “extremity,” “surgical procedures,” and “pharmacokinetics” (see Supplementary data). The last search was run on January 15, 2018. In addition to the databases, bibliographies were checked for additional articles. Eligible for inclusion were randomized controlled trials or prospective cohort studies investigating “target site” antibiotic concentrations in human, adult subjects who received prophylactic cefazolin in a single, intravenously administered dose before orthopedic/trauma surgery of the extremity. “Target site” concentrations were defined as concentrations measured in soft tissue, bone, synovia, or wound/drain fluid at or near the site of surgery. To be able to compare dosages, we chose to limit the type of administered prophylaxis to cefazolin only, the most widely studied first-generation cephalosporin. No publication date or language restrictions were imposed. Study selection All identified studies were screened for relevance based on title and abstract by 2 reviewers (TS and FS). The remaining studies were independently screened for eligibility based on full-text reading by the same reviewers and were included if none of the exclusion criteria were met. Studies were excluded based on intervention (cephalosporin that was not cefazolin), study design (reviews or articles only available as abstract), population (when included patients received therapeutic antibiotics up to a week before surgery or had peripheral vascular disease), and outcome (solely serum concentrations measured). Conflicts were discussed until consensus was reached. The literature search identified 825 articles, of which 626 were assessed for eligibility and a final number of 14 studies were included in the systematic review after full text screening (Figure 1, Table 1). 5 studies were also included in the metaanalysis (3 in the comparison of different target sites, 1 in the comparison of different cefazolin dosages, 1 in both).

Acta Orthopaedica 2019; 90 (2): 97–104

Records identified through database searching n = 825 Records after duplicates removed n = 626 Records excluded after screening n = 499 Full-text articles assessed for eligibility n = 116 Full-text articles excluded (n = 102): – intervention, 90 – study design, 4 – duplicates, 4 – patient population, 2 – outcomes, 2 Studies included in qualitative synthesis n = 14 Studies included in quantitative synthesis (meta-analysis) n=5

Figure 1. Flow diagram of included studies.

Data collection Data were extracted using a customized extraction sheet (based on the Cochrane data extraction template), which was pilot-tested on 5 articles randomly selected from the included articles and adjusted accordingly. One reviewer (FS) extracted the data and the other reviewer (TS) verified it. Duplicate publications were filtered out by juxtaposing author names and carefully reviewing study designs and treatment combinations. In the case of multiple publications on one trial, all published information was combined to ensure comprehensiveness of data. Collected information included (1) study characteristics (study design, number of patients included, in/exclusion criteria, and year published), (2) patient characteristics (sex, age, weight, type of procedure, given demographic or disease specifics), (3) type of intervention(s) (dose, timing of administration, comparative), (4) outcome (site of measurement, timing of measurement, unit presented as, type of analysis, results). Study quality Studies were screened for quality and risk of bias using the Newcastle–Ottawa Scale (NOS), designed to assess the quality of nonrandomized cohort studies (Wells et al. 2013). Using this scale studies were judged on 9 items within 3 domains (selection, comparability, and outcome) as either good, poor, or unclear, a “good” score counting as one point, with a maximum of 9 points. The rating sheet was adjusted to this review using topic-specific rating criteria (shown in Appendix). Quality screening was performed by one reviewer (FS) and subsequently checked by another reviewer (TS). Due to heterogeneity of location of measurements, cefazolin dosages and measurement methods, assessment of publication bias—using for example a funnel plot—was not possible.


Acta Orthopaedica 2019; 90 (2): 97–104

99

Table 1. Study characteristics. Characteristics are reported for intervention groups (patients receiving cefazolin i.v. only) unless otherwise reported A B C D E

F

G H I J K L M

1 Angthong (2015) PC 18 C 1 g iv (12) – 70 (5) 25 61 (8) TKA (18) Yes < TQ / incision C 2 g iv (6) 68 (3) 17 62 (8) 2 Bryan (1988) DB 97 C 1 g iv (48) Cefamandole 2 g iv 59 (12) 46 THA (37)/ NR 30–60 min RCT (49) TKA (11) < anesthesia 3 Cunha (1977) PC 71 C 1 g iv (?) Cephradine 1 g iv (?) NR NR NR THA No Cephalothin 1 g iv (?) 4 Cunha (1984) PC 35 C 1 g iv (13) Cephradine 1 g iv (?) [61–88] NR NR TKA Yes 10–225 min < bone removal 5 Deacon (1996) PC 25 C 1 g iv (25) – < 55 NR NR Bunion- Yes 60 min < TQ ectomy 6 Friedman (1990) RCT 24 C 1 g – 52 TKA Yes Until 1, 2, 1 min < TQ (8) 67 (8) 94 (18) or 5 min < TQ 2 min < TQ (8) 63 (10) 98 (22) 5 min < TQ (8) 60 (8) 90 (25) 7 Miller (2004) PC 15 C 1 g iv (7) Cefazolin 1 g iv 49 57 NR Shoulder No 20 min + regional (8) [29–77] surgery < incision 51 (14) 8 Parsons (1978) PC 7 C 4 g iv (7) – 61 (1.3) 57 65 (5) THA No Immediately < anesthesia 9 Polk (1983) PC 20 C 10 mg/kg Cefazolin 10 mg/kg [27–82] 65 [51–83] THA No Shortly after SR bolus (9) infusion (11) anesthesia 10 Sharareh (2016) PC 34 C 1 g/2 g – 67 44 83 THA (12)/ Yes d < 60 min to < 70 kg/ [38–86] [50–125] TKA (22) incision / < TQ > 70 kg (31) a 64 (12) 85 (19) 11 Sørensen (1978) PC 20 C 1 g iv (1) b Erythromycin (8) 75 40 c NR Fixation No < surgery Methicillin (6) [48–92] c femur 73 (11) fracture 12 Williams (1983) PC 125 C 1 g iv (17) Cephalothin 0.5 g (38) 65 NR NR THA (13)/ Yes d 30 min < TQ C 2 g iv (6) Cefamandole 2 g (13) [13–91] c TKA (10) Cefoxitin 0.5 g (37) 59 (13) Ceforanide 0.5 g (14) 13 Yamada (2011) PC 43 C 2 g iv (43) – 75 (8) 16 55 (8) THA (16)/ Yes d < TQ /incision TKA (27) 14 Young (2013) RCT 22 C 1 g iv (11) Cefazolin 1 g io (11) 65 36 BMI 29 TKA Yes 10–30 min < TQ [48–83] [23–35] 65 (10)

30 NR NR 0.5 0.5–1.0 1

NR 0.25–4 NR 2 0.5–3 0.5–7

1–100 0.5–100

A Reference number B First author (year) C Study design DB = Double-blind PC = Prospective cohort RCT = Randomized controlled trial SR = Semi-randomized D Total number E Intervention (n) C = cefazolin a 3 patients did not receive cefazolin because of allergy, 4 received 1 g, 27 received 2g. b 6 patients received cefazolin, 3 excluded because of receiving > 1 dose, 2 no detectable levels. F Comparison (n) io = intra-osseously administered (ref 14) G Mean age (SD) When only median [range] are reported, values were transformed into estimated mean and standard deviation by using the calculations from Hozo et al. (2005), estimations are given in italics. c For all patients, not just intervention group. H Male sex (%) I Weight See G for values in italics. J Type of surgery (n) K Tourniquet use d in TKA L Timing of antibiotic administration < = before TQ = tourniquet M MIC reported (µg/mL)


100

Acta Orthopaedica 2019; 90 (2): 97–104

Statistics All statistical analyses were performed using Review Manager (RevMan, Version 5.3, the Nordic Cochrane Centre, Copenhagen, the Cochrane Collaboration, 2014). Mean target site antibiotic concentrations were compared between different locations of the target site and between different dosages of cefazolin. To compare these groups, while taking the heterogeneity between studies into account, only studies describing 2 different doses/measuring sites were included in the meta-analysis. Only concentrations measured in bone were included for meta-analysis, since soft tissue samples were not comparable (different locations). Antibiotic concentrations were expressed as weighted mean difference and the corresponding 95% confidence interval (CI). Statistical analyses were performed using a random-effects model, considering the heterogeneity of included trials (Borenstein et al. 2010). The I2 was used as a measure for consistency of data. To incorporate the heterogeneity in the estimation of difference in concentrations between hip and knee, a 95% prediction interval (PI) was also computed. Where the CI only represents the average of study effects, the PI presents the range of expected results for 95% of similar studies that might be conducted in the future (Higgins et al. 2009). However, as Partlett and Riley (2017) recently concluded, the PI performs best when heterogeneity is high but when there is also a minimum of 5 included studies. With a smaller amount of studies, the PI tends to be too wide, and should not be interpreted as true effect. Statistical significance was defined as p < 0.05.

Table 2. Cefazolin dose and site of measurement of included studies

Results

Target-site concentrations Bone concentrations Overall, target-site antibiotic concentrations in the bone ranged from 0.64 µg/g to 88 µg/g, but the variation of concentrations between different administered dosages of cefazolin and location of measurement was quite large. All of the 10 studies reporting a MIC (Table 1) reported mean target site concentrations higher than the minimum concentration for S. aureus (0.5–2 µg/mL). However, most studies also described MICs for other organisms, usually requiring higher concentrations of cefazolin. In 5 studies, mean cefazolin concentrations were higher than all of the MICs they reported for different pathogens/resistance patterns, ranging from 0.5 ug/mL for S. aureus, to 3–4 µg/mL for E. coli and Klebsiella species (4, 5, 8, 10, 11). 2 studies specifically reported the percentage of patients achieving bone concentrations higher than the MIC. Friedman et al. (1990) reported that 10 of 24 patients achieved bone concentrations higher than a MIC of 4 µg/mL at 30 minutes after administering 1g of cefazolin. Sharareh et al. (2016), using a MIC of 2 µg/mL, found that in the group receiving 1 g of cefazolin, 3 of 4 reached the MIC compared with 25 of 27 in the group receiving 2 g.

Included trials were mostly prospective cohorts (1, 3–5, 7–13, number of refs refers to numbers given in Table 1), except for 3 randomized controlled trials (2, 6, 14). Most studies included patients undergoing elective total hip replacement (n = 3) (3, 8, 9), total knee replacement (n = 4) (1, 4, 6, 14) or both (n = 5) (2, 4, 10, 12, 13) (of which ref no. 4 compared their results in the knee with results in the hip from a previously reported trial (3) in their most recent article) and 1 each included patients with a femur fracture (11), shoulder surgery (7), or bunionectomy (5). Table 2 gives an overview of the different target sites and given dose of cefazolin described by each study. All studies reported target site antibiotic concentrations in bone and some also measured concentrations in soft tissue (n = 4) (6–8, 14) or synovial fluid (n = 1) (4). Target site concentrations were reported as a mean value in 10 studies (1, 2, 5, 7–10, 12–14), as mean peak value in 2 studies (3, 4), as separate values per patient in one study (11), and as percentage of patients with values above the MIC in 2 studies (6, 10). Due to the heterogeneity of included studies regarding cefazolin doses, sampling, and analysis methods, the majority of data could not be pooled.

Site Dose Reference Hip 1 g 2 g Other Knee 1 g 2 g Foot 1 g Shoulder 1 g

Bryan et al. (1988) Cunha et al. (1977) Sharareh et al. (2016) Sorensen et al. (1978) Williams et al. (1983) Sharareh et al. (2016) Williams et al. (1983) Yamada et al. (2011) Parsons et al. (1978), 4 g cefazolin Polk et al. (1983), 10 mg/kg (mean: 684 (108) mg) Angthong et al. (2015) Bryan et al. (1988) Cunha et al. (1984) Friedman et al. (1990) Sharareh et al. (2016) Williams et al. (1983) Young et al. (2013) Angthong et al. (2015 Sharareh et al. (2016) Williams et al. (1983) Yamada et al. (2011) Deacon et al. (1996) Miller et al. (2004)

Quality assessment Although a minimal score on the NOS suggesting good quality has not been established, the overall risk of bias in the included studies seemed high, with only 5 studies scoring 5 or more out of 9 points on the NOS (Wells et al. 2013). The studies used in meta-analysis scored relatively high with 2 studies scoring seven stars and 1 each scoring 6, 5, and 4 points (Figure 2, see Supplementary data).


Acta Orthopaedica 2019; 90 (2): 97–104

101

Bone cefazolin concentration (µg/g)

Bone cefazolin concentration (µg/g)

1,000

1,000 foot knee hip shoulder

100

100

10

10

1

1

0.1

0

20

40

60

80

100

120

140

4g 2g 1g 10 mg/kg

160

Minutes after administration

0.1

0

20

40

60

80

100

120

140

160

Minutes after administration

Figure 3. Mean target site concentrations organized according to location of measurements. Mean or maximum target site concentrations of all included studies. When results were reported separately for individual patients or results were given for multiple time-points, these are depicted separately. The bar with the dotted line represents the reported MIC90 of Staphylococcus aureus (0.5–2.0 µg/L).

Figure 4. Mean target site concentrations organized according to dose of cefazolin. Mean or maximum target site concentrations of all included studies. When results were reported separately for individual patients or results were given for multiple time points, these are depicted separately. The bar with the dotted line represents the reported MIC90 of Staphylococcus aureus (0.5–2.0 µg/L).

Soft tissue concentrations Cunha et al. (1984) took samples of the synovia of the knee and found mean peak levels of 8 mg/L, which cannot be compared with measurements in bone due to the unit (mg/L vs. ug/g) and absence of standard deviations (Table 3, see Supplementary data). Miller et al. (2004) found a mean cefazolin concentration of 11 µg/g in the soft tissue of the shoulder (SD not reported), which was lower than the 36 µg/g they found in bone at the same time-point. Young et al. (2013) found values between 7 µg/g and 13 µg/g in fat around the knee joint at different time-points, comparable to the values measured simultaneously in bone. Friedman et al. (1990), conversely, reported a higher percentage of patients with concentrations above the MIC (4 µg/g) in soft tissue than in bone of the knee at each time-point. Parsons et al. (1978) also found higher levels in the hip capsule than in bone with mean concentrations of 35 µg/g (SD 7.2) and 14 (2.3) respectively.

other 3 studies either did not compare knee and hip directly or did not perform statistical testing for this comparison. Of the 5 studies measuring antibiotic concentrations at different target sites, 4 (2, 10, 12, 13) could be pooled. The results from Cunha et al. (1984) could not be included in the meta-analysis due to the fact that no standard deviations were provided. When pooled, target site cefazolin concentrations were significantly higher in the hip (acetabulum, femoral head, or proximal femur) than in the knee (distal femur or proximal tibia) with a mean difference of 4 µg/g (CI 0.8–7) (Figure 5, see Supplementary data). Although the time-points at which concentrations were measured differed between studies, the time-points of measurements in hip and knee within each study were similar. Heterogeneity between the studies is high, with an I2 of 83%.

Location of target site Mean (peak) antibiotic concentrations for different measuring sites were ranging from 1.6 µg/g to 88 µg/g in the hip, from 0.64 µg/g to 40 µg/g in the knee, 2.4 µg/g in the foot, and 36 µg/g in the shoulder, measured at varying time-points. The results of each individual study are displayed in Table 4–5 and Figure 5 (see Supplementary data). Figure 3 shows that although concentrations in the knee were lower than in the hip, nearly all measured concentrations were higher than the MIC of S. aureus. The concentrations that were lower or only just above the MIC were measured either more than 100 minutes after administration or in the foot. In 5 studies concentrations in the hip and knee were compared (2, 4, 10, 12, 13). All reported higher concentrations measured in the hip than in the knee, which were statistically significant in 2 (10, 12). The

Dose of cefazolin Bone concentrations were ranging from 0.64 µg/g to 36 µg/g when 1 g of cefazolin was given, 8.3 µg/g to 40 µg/g in 2 g, 10 µg/g to 88 µg/g in 4 g, and 7.7 µg/g in the study administering 10 mg/kg (Figure 4, Table 5 in Supplementary data). 3 studies compared target-site concentrations according to the given dose of cefazolin (either 1 g or 2 g) (1, 10, 12). All of these studies report higher levels for the group receiving 2 g of cefazolin, but only in 1 study statistical significance was achieved (1) (Figure 6, see Supplementary data). Figure 4 shows that the concentrations lower or just above the MIC were all measured after administration of 1 g of cefazolin. Only 2 studies (1, 10) could be used for meta-analysis, because of missing standard deviations in the study by Williams et al. (1983). As shown in Figure 6, pooling the results did not lead to a statistically significant difference in target site concentration (mean difference –9, CI –23 to 4). Time-points


102

of the measurements were similar both between and within the different studies. Nonetheless, heterogeneity between the studies was high with an I2 of 78%.

Discussion Because of the large clinical implications of SSIs in orthopedic trauma surgery, prophylactic antibiotics are widely initiated. However, dosage recommendations and corresponding efficacy remain unclear, partly because of the unequal distribution of drugs throughout the body. Sufficient concentrations of antibiotics at the target site (location of surgery) are required for optimal infection prevention. Therefore this meta-analysis focused on the target-site concentrations of cefazolin, the most commonly used prophylactic in orthopedic/trauma surgery. A few limitations can be pointed out for this review and the available literature. First of all, the quality of the individual studies included in this systematic review varied. Most studies presented data that were collected before the year 2000, which often resulted in incomplete reporting and possibly outdated analysis methods. Also, given the fact that most studies included only elective orthopedic surgery, selection bias, by including only relatively healthy patients, may have occurred. Second, given the large heterogeneity in methods used for sampling, timing of the samples, processing, and analyzing, less than half of the results could be pooled. Third, targetsite concentrations of an antibiotic may also be influenced by other patient or surgical characteristics such as renal function, obesity, bleeding, or tourniquet use. Although most of these characteristics do not seem to differ within study groups, tourniquet use could have an influence on the comparison between samples of hip and knee. Another limitation is the fact that all of the individual studies measured concentrations in samples of bone/soft tissue, the so called “whole tissue concentration.” As described by Mouton et al. (2008), these concentrations are only an estimate of the “unbound” or active part of the drug and cannot be credibly compared with the MIC, which is the total (unbound plus plasma protein bound) concentration in serum. Moreover, homogenizing or grinding up whole-tissue samples leads to dilution of the drug by mixing intracellular and extracellular fluids, resulting in, depending on the type of drug, under- or overestimation of its concentrations (Mouton et al. 2008). Nevertheless, simply using the serum concentration as a surrogate might not be sufficient for estimating effect, since the distribution of drugs throughout the body is not homogeneous (Müller et al. 2004). Finally, regarding the clinical implications of antibiotic prophylaxis, more than the concentration itself, the time that the concentration is higher than the MIC (T>MIC) is important. The T>MIC should at least overlap with the “decisive period.” This is the period that starts at incision and ends after 3 hours, during which antibiotics can effectively suppress the development of a wound infection (Burke 1961). Instead of reporting this T>MIC, all

Acta Orthopaedica 2019; 90 (2): 97–104

included studies solely measured antibiotic concentrations at individual moments in time. When samples are taken at multiple time-points, they can be used to predict levels of antibiotics in tissue over a course of time, using population kinetics, like Gergs et al. (2014). With population kinetics, model predicted time-concentration curves for each patient can be fabricated, which allows the evaluation of the T>MIC, and therefore a more precise estimate of clinical efficacy. We found a large variation in target site cefazolin concentrations of the extremity between different studies. In general, the achieved concentrations in bone surpassed the minimal inhibitory concentration (MIC) for S. aureus, the most common pathogen of an SSI. However, when stratifying the results on the location of the target site and dose of cefazolin, some measurements did not reach this MIC. Regarding the association between location of the target site and antibiotic concentrations, our study showed that the same dosage of cefazolin resulted in statistically significantly lower concentrations in the knee than in the hip. Although, with just a single study in the foot (Deacon et al. 1996), no definite conclusions can be drawn, this could mean that for surgery of the more distal parts of the extremity (e.g., foot/ankle), 1 g of cefazolin is not sufficient as prophylaxis. Whether or not a higher dose is beneficial in this area is yet to be determined. As for the relationship between cefazolin dose and concentration, all 3 articles that compared different dosages found higher concentrations when 2 g was given instead of 1 g, although only 1 with statistical significance. A visualization of these concentrations, including also the studies investigating only one dose, suggests that the higher the dose, the higher the concentration (Figure 4). However, even though concentrations seemed increasingly high, the clinical implications of this phenomenon have not been investigated. A ceiling effect, where higher concentrations would not lead to fewer SSIs, is likely to occur at a certain time and could pave the way for antimicrobial resistance. This first systematic review shows that there is a large variation in target site cefazolin concentrations of the extremity between different studies. In general, the achieved concentrations in bone surpassed the minimal inhibitory concentration (MIC) for S. aureus, the most common pathogen of an SSI. Most importantly, we found that the local concentration of cefazolin is associated with the location of the target site. Although no definite conclusions can be drawn based on this study, a higher dose of cefazolin seems to produce higher whole-tissue concentrations. These insights could be helpful on the path towards more efficient use of antibiotics. In particular, this study gives rise to the question whether the dose of prophylactic cefazolin needs to be adjusted to the location of the target site. To make any recommendations for the dose of prophylactic cefazolin in orthopedic/trauma surgery of the extremity however, additional prospective research is needed. We believe that a preclinical trial, comparing multiples dosages and locations of measurement in the extremity, is neces-


Acta Orthopaedica 2019; 90 (2): 97–104

sary before further investigating the efficacy of prophylactic cefazolin in preventing SSIs. Supplementary data Tables 3–5 and Figures 2 and 5–6 are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/17453674.2019.1577014 TS conceived of the presented idea; FS and TS performed the search, screened and analyzed the articles; FS drafted the manuscript; TS, JCG, and RM revised the manuscript critically for intellectual content. The authors would like to thank Faridi S. van Etten-Jamaludin, clinical librarian, for her help with the literature search. Acta thanks Anna Stefánsdóttir for help with peer review of this study.

Agodi A, Auxilia F, Barchitta M, Cristina M L, D’Alessandro D, Mura I, Nobile M, Pasquarella C, GISIO-SItI. Risk of surgical site infections following hip and knee arthroplasty: results of the ISChIA-GISIO study. Ann DI Ig Med Prev E DI Comunita 2017; 29(5): 422-30. AlBuhairan B, Hind D, Hutchinson A. Antibiotic prophylaxis for wound infections in total joint arthroplasty: a systematic review. J Bone Joint Surg Br 2008; 90(7): 915-19. Angthong C, Krajubngern P, Tiyapongpattana W, Pongcharoen B, Pinsornsak P, Tammachote N, Kittisupaluck W. Intraosseous concentration and inhibitory effect of different intravenous cefazolin doses used in preoperative prophylaxis of total knee arthroplasty. J Orthop Traumatol 2015; 16(4): 331-4. Backes M, Schepers T, Beerekamp M S H, Luitse J S K, Goslings J C, Schep N W L. Wound infections following open reduction and internal fixation of calcaneal fractures with an extended lateral approach. Int Orthop 2014; 38(4): 767-73. Bauer M P, van de Garde E M W, van Kasteren M E E, Prins J, Vos M. SWAB Richtlijn: [peri-operatieve profylaxe. [Dutch] 2017; (January): 1-10. Borenstein M, Hedges L V, Higgins J P T, Rothstein H R. A basic introduction to fixed-effect and random-effects models for meta-analysis. Res Synth Methods 2010; 1(2): 97-111. Boxma H, Broekhuizen T, Patka P, Oosting H. Randomised controlled trial of single-dose antibiotic prophylaxis in surgical treatment of closed fractures: the Dutch Trauma Trial. Lancet 1996; 347(9009): 1133-7. Bratzler D W, Dellinger E P, Olsen K M, Perl T M, Auwaerter P G, Bolon M K, Fish D N, Napolitano L M, Sawyer R G, Slain D, Steinberg J P, Weinstein R A. Clinical practice guidelines for antimicrobial prophylaxis in surgery. Am J Heal Pharm 2013; 70(3): 195-283. Brill M J E, Houwink A P I, Schmidt S, Van Dongen E P A, Hazebroek E J, Van Ramshorst B, Deneer V H, Mouton J W, Knibbe C A J. Reduced subcutaneous tissue distribution of cefazolin in morbidly obese versus nonobese patients determined using clinical microdialysis. J Antimicrob Chemother 2014; 69(3): 715-23. Bryan C S, Morgan S L, Caton R J, Lunceford E M. Cefazolin versus cefamandole for prophylaxis during total joint arthroplasty. Clin Orthop Relat Res 1988; (228): 117-22. Burke J F. The effective period of preventive antibiotic action in experimental incisions and dermal lesions. Surgery 1961; 50: 161-8. Burnett J W, Gustilo R B, Williams D N, Kind A C. Prophylactic antibiotics in hip fractures: a double-blind, prospective study. J Bone Joint Surg Ser A 1980; 62(3): 457-62.

103

Cunha B A, Gossling H R, Pasternak H S, Nightingale C H, Quintiliani R. The penetration characteristics of cefazolin, cephalothin, and cephradine into bone in patients undergoing total hip replacement. J Bone Joint Surg Am 1977; 59(7): 856-9. Cunha B A, Gossling H R, Pasternak H S, Nightingale C H, Quintiliani R. Penetration of cephalosporins into bone. Infection 1984; 12(2): 80-4. Deacon J S, Wertheimer S J, Washington J A. Antibiotic prophylaxis and tourniquet application in podiatric surgery. J Foot Ankle Surg 1996; 35(4): 344-9. De Jong L, Klem T M A L, Kuijper T M, Roukema G R. Factors affecting the rate of surgical site infection in patients after hemiarthroplasty of the hip following a fracture of the neck of the femur. Bone Joint J 2017; 99B(8): 1088-94. Feilmeier M, Dayton P, Sedberry S, Reimer R A. Incidence of surgical site infection in the foot and ankle with early exposure and showering of surgical sites: a prospective observation. J Foot Ankle Surg 2014; 53(2): 173-5. Friedman R J, Friedrich L, White R L, Kays M B, Brundage D M, Graham J. Antibiotic prophylaxis and tourniquet inflation in total knee arthroplasty. Clin Orthop Relat Res 1990; (260): 17-23. Gergs U, Clauss T, Ihlefeld D, Weiss M, Pönicke K, Hofmann G O, Neumann J. Pharmacokinetics of ceftriaxone in plasma and bone of patients undergoing hip or knee surgery. J Pharm Pharmacol 2014; 66(11): 1552-8. Higgins J P T, Thompson S G, Spiegelhalter D J. A re-evaluation of randomeffects meta-analysis. J R Stat Soc Ser A Stat Soc 2009; 172(1): 137–59. Hozo S P, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol 2005; 5: 13. Mangram A J, Horan T C, Pearson M L, Silver L C, Jarvis W R. Guideline for prevention of surgical site infection, 1999. Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol 1999; 20(4): 250-78; quiz 279-80. Miller B S, Harper W P, Hughes J S, Sonnabend D H, Walsh W R. Regional antibiotic prophylaxis in elbow surgery. J Shoulder Elb Surg 2004; 13(1): 57-9. Moher D, Liberati A, Tetzlaff J, Altman D G, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009; 339(17): b2535. Moine P, Fish D N. Pharmacodynamic modelling of intravenous antibiotic prophylaxis in elective colorectal surgery. Int J Antimicrob Agents 2013; 41(2): 167-73. Mouton J W, Theuretzbacher U, Craig W A, Tulkens P M, Derendorf H, Cars O. Tissue concentrations: do we ever learn? J Antimicrob Chemother 2008; 61(2): 235-7. Müller M, Dela Peña A, Derendorf H. Issues in pharmacokinetics and pharmacodynamics of anti-infective agents: distribution in tissue. Antimicrob. Agents Chemother 2004; 48(5): 1441-53. Parsons R L, Beavis J P, David J A, Paddock G M, Trounce J R. Plasma, bone, hip capsule, and drain fluid concentrations of cephazolin during total hip replacement. Br J Clin Pharmac 1978; 5: 331-6. Partlett C, Riley R D. Random effects meta-analysis: coverage performance of 95% confidence and prediction intervals following REML estimation. Stat Med 2017; 36(2): 301-17. Polk R, Hume A, Kline B J, Cardea J. Penetration of moxalactam and cefazolin into bone following simultaneous bolus or infusion. Clin Orthop Relat Res 1983; 177(7): 216-21. Sharareh B, Sutherland C, Pourmand D, Molina N, Nicolau D P, Schwarzkopf R. Effect of body weight on cefazolin and vancomycin trabecular bone concentrations in patients undergoing total joint arthroplasty. Surg Infect (Larchmt) 2016; 17(1): 71-7. Sørensen T S, Colding H, Schroeder E, Rosdahl V T. The penetration of cefazolin, erythromycin and methicillin into human bone tissue. Acta Orthop 1978; 49(6): 549-53.


104

Voigt J, Mosier M, Darouiche R. Systematic review and meta-analysis of randomized controlled trials of antibiotics and antiseptics for preventing infection in people receiving primary total hip and knee prostheses. Antimicrob Agents Chemother 2015; 59(11): 6696-707. Wells G A, Shea B, O’Connell D, Peterson J, Welch V, Losos M, Tugwell P. The Newcastle–Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in meta-analyses. Ottawa Hosp Res Inst 2013; (3): 1-4. Wiewiorski M, Barg A, Hoerterer H, Voellmy T, Henninger H B, Valderrabano V. Risk factors for wound complications in patients after elective orthopedic foot and ankle surgery. Foot Ankle Int 2015; 36(5): 479-87.

Acta Orthopaedica 2019; 90 (2): 97–104

Williams D N, Gustilo R B, Beverly R, Kind A C. Bone and serum concentrations of five cephalosporin drugs: relevance to prophylaxis and treatment in orthopedic surgery. Clin Orthop Relat Res 1983; (179): 253-65. Yamada K, Matsumoto K, Tokimura F, Okazaki H, Tanaka S. Are bone and serum cefazolin concentrations adequate for antimicrobial prophylaxis? Clin Orthop Relat Res 2011; 469(12): 3486-94. Young S W, Zhang M, Freeman J T, Vince K G, Coleman B. Higher cefazolin concentrations with intraosseous regional prophylaxis in TKA knee. Clin Orthop Relat Res 2013; 471(1): 244-9.


Acta Orthopaedica 2019; 90 (2): 105–110

105

A comparative study of intraoperative frozen section and alpha defensin lateral flow test in the diagnosis of periprosthetic joint infection Irene K SIGMUND 1, Johannes HOLINKA 1, Susanna LANG 2, Sandra STENICKA 1, Kevin STAATS 1, Gerhard HOBUSCH 1, Bernd KUBISTA 1, and Reinhard WINDHAGER 1 1 Medical

University of Vienna, Department of Orthopaedics and Trauma Surgery, Vienna; 2 Medical University of Vienna, Department of Pathology, Vienna, Austria Correspondence: reinhard.windhager@meduniwien.ac.at Submitted 2018-08-29. Accepted 2018-11-17.

Background and purpose — For decision-making (aseptic vs. septic), surgeons rely on intraoperatively available tests when a periprosthetic joint infection (PJI) cannot be confirmed or excluded preoperatively. We compared and evaluated the intraoperative performances of the frozen section and the alpha defensin lateral flow test in the diagnosis of PJI. Patients and methods — In this prospective study, consecutive patients with indicated revision surgery after arthroplasty were included. Patients were classified as having PJI using the MusculoSkeletal Infection criteria. The presence of alpha defensin was determined using the lateral flow test intraoperatively. During revision surgery, tissue samples were harvested for frozen and permanent section. Analysis of diagnostic accuracy was based on receiver-operating characteristics. Results — 101 patients (53 hips, 48 knees) were eligible for inclusion. Postoperatively, 29/101 patients were diagnosed with PJI, of which 8/29 cases were definitely classified as septic preoperatively. Of the remainder 21 septic cases, the intraoperative alpha defensin test and frozen section were positive in 13 and 17 patients, respectively. Sensitivities of the alpha defensin test and frozen section were 69% and 86%, respectively. The area under the curves of both tests showed a statistically significant difference (p = 0.006). Interpretation — The frozen section showed a significantly higher performance compared with the alpha defensin test and a near perfect concordance with the definitive histology, and therefore remains an appropriate intraoperative screening test in diagnosing PJI. Although the sensitivity of the alpha defensin test was lower compared with that of frozen section, this test is highly specific for confirming the diagnosis of PJI.

Despite various available diagnostic test methods, diagnosing periprosthetic joint infection (PJI) remains very challenging. When results are not available or cannot be interpreted preoperatively, an intraoperative screening test is required to confirm or exclude PJI. Some studies have shown good concordance between the intraoperative evaluation of frozen tissue samples and paraffin-embedded permanent sections ranging from 95% to 100% (Della Valle et al. 1999, Musso et al. 2003, Bori et al. 2007, Stroh et al. 2012, Kwiecien et al. 2017). In the recent study by Kwiecien et al. (2017), the sensitivity and specificity of intraoperative frozen section was 74% and 99%, respectively. This is in line with the reported 73% and 95% of permanent sections by Morawietz et al. (2009). Therefore, frozen sections show comparable results to the definitive histopathology and could be useful for confirming the presence or absence of PJI intraoperatively. On the other hand, in recent years attention was also focused on the biomarker alpha defensin, which is released by neutrophils into the synovial fluid and induces rapid microorganism death due to depolarization of the cell membrane (Ganz et al. 1985, Chalifour et al. 2004). The alpha defensin lateral flow test showed sensitivities ranging from 69% to 92% and high specificities close to 100% (Sigmund et al. 2017, Gehrke et al. 2018, Renz et al. 2018). While results of the quantitative alpha defensin test (ELISA) are available within one day, the lateral flow test is characterized by ease of use and quick results within 10 minutes, thus rendering its intraoperative use frequent. We evaluated the intraoperative performance of the alpha defensin lateral flow test and the histopathological analysis of frozen tissue samples using the Musculoskeletal Infection Society (MSIS) criteria (Parvizi and Gehrke 2014). In addition, the results of both diagnostic methods were compared.

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2019.1567153


106

Patients and methods Study design This prospective cohort study was conducted in a tertiary healthcare center providing advanced specialty care to patients with PJI with multidisciplinary collaborations. This study includes some data (31 patients) from a previously published study analyzing the sensitivity and specificity of qualitative alpha defensin in all kinds of revision surgery (including the second stage of a 2-stage revision and spacer exchanges, which were excluded in this study) (Sigmund et al. 2017). Study population Patients with an indicated revision surgery between January 2016 and February 2018 were eligible for inclusion. Inclusion criteria comprised an indicated revision surgery after TJA, sufficient synovial fluid derived from the affected joint, a conclusive alpha defensin lateral flow test, a frozen section, and permanent histopathology. Samples with obvious contaminated joint fluid and/or failed periprosthetic tissue samples for pathohistological analysis were excluded. Further exclusion criteria (and the difference from the previously published study) were surgery within the last 6 weeks, a joint aspiration with a cement spacer in place, the second stage of 2-stage revision, or a resection arthroplasty. Definition of infection In accordance with the MSIS criteria (Parvizi and Gehrke 2014), PJI was diagnosed when a sinus tract communicating with the joint was present, 2 or more periprosthetic cultures grew phenotypically identical organisms, or at least 3 of the following 5 minor criteria were present: (i) elevated serum C-reactive protein (CRP) (acute: > 100 mg/L or chronic: > 10 mg/L), (ii) elevated synovial fluid white blood cell (acute: WBC > 10,000/µL or chronic: WBC > 3,000/µL), (iii) elevated synovial fluid polymorphonuclear neutrophil percentage (acute: PMN% > 90% or chronic: PMN% > 80%), (iv) positive histological analysis of periprosthetic tissue, and/or (v) a single positive culture. The serum erythrocyte sedimentation rate (ESR), which has low sensitivity and specificity for PJI, was not routinely determined in our institution (Piper et al. 2010). We did not evaluate leucocyte esterase colorimetric test strips in our study as Deirmengian et al. (2015) have shown that alpha defensin is a better test. Determination of diagnostic tests For all patients, a standardized diagnostic workup was performed. First, blood samples were collected to assess the serum CRP levels. In line with proceedings of the International Consensus Meeting (Parvizi et al. 2013), a cut-off of 100 mg/L (acute) or 10 mg/L (chronic) was chosen as positive with suspicion of infection (systemic or local).

Acta Orthopaedica 2019; 90 (2): 105–110

Preoperatively, synovial fluid was aspirated under sterile conditions. 1 mL was placed into a vial containing ethylenediaminetetraacetic acid (EDTA) for quantification of the erythrocyte and leukocyte count as well as granulocyte percentage. Another 1 mL was sent for microbiological investigations and processed per standard laboratory protocol with cultures held for 14 days (Butler-Wu et al. 2011, Parvizi et al. 2013). For alpha defensin measurements, synovial joint fluid was aspirated in the operation room before arthrotomy by direct needle aspiration (Diagnostics 2013). For qualitative alpha defensin testing, the Synovasure™ test (Zimmer Inc., Warsaw, IN, USA) was used according to the manufacturer’s instructions. The synovial fluid was processed and the qualitative result (i.e., infection present yes or no) was read after 10 min. The control line (“C”) had to appear; otherwise the test was considered inconclusive. Intraoperatively, at least 3 periprosthetic tissue samples were sent for microbiological investigations and processed per standard laboratory protocol with cultures held for 14 days (Schafer et al. 2008, Butler-Wu et al. 2011, Parvizi et al. 2013). All the explanted prosthetic components were also sent for sonication culture analysis. In all cases, at least 1 tissue sample (median: 3, range 1–8) for frozen section was taken from 1 of several sites at the time of revision surgery. For frozen section, multiple sections or 1 representative area from each sample were prepared at –20° C in the Microm HM 550 cryocut (Thermo Fisher Scientific, Walldorf, Germany) for about 3 min. Sections of 3 µm thickness were cut and stained with hematoxylin and eosin. The samples were analyzed and evaluated under high power (×400 magnification). The diameter of the visual field was 0.625 mm; hence the visual field was 0.307 mm2. At least 40 high-powered fields (HPFs) were evaluated for each slice. The samples were interpreted by 1 of 3 senior pathologists specialized in musculoskeletal infections. After about 15–25 min, the pathologist communicated the results to the orthopedic surgeon using the intercom system. For definitive histopathological analysis, samples for frozen section and additional periprosthetic tissue specimens (median 5, range 2–12) were obtained intraoperatively in all patients. The samples were sent to histopathological analysis and fixed in 4.5% formaldehyde for 12 hours, paraffin embedded and also stained with hematoxylin and eosin. The samples were analyzed and classified according to the Krenn criteria by default (Morawietz et al. 2009) by 1 of the 3 pathologists. If the number of neutrophil granulocytes was > 23 in 10 HPFs the sample was classified as positive. Statistics Analysis of diagnostic accuracy is based on the receiveroperating characteristic (ROC). Sensitivity, specificity, area under the ROC curve (AUC), positive and negative likelihood ratios (LR+ and LR–, respectively), accuracy (calculated as the number of correct classifications/number of total classifi-


Acta Orthopaedica 2019; 90 (2): 105–110

107

Table 1. Distribution of microorganisms Conventional Isolated pathogens culture (n = 29) Staphylococcus aureus 5 Coag. negative Staphylococci 4 Escherichia coli 3 Cutibacterium acnes 1 Streptococcus agalactiae 1 Enterococcus faecalis 1 Finegoldia magna 1 Actinomyces neuii 1 Parvimonas micra 1 Moraxella osloensis 1 Candida albicans 1 Negative culture 9

Sensitivity

Specificity

Alpha defensin (n = 101) Frozen section (n = 101) C-reactive protein (n = 100) Synovial fluid white blood count (n = 62) Synovial fluid culture (n = 88) Tissue culture (n = 90) Sonication culture (n = 94) Histopathology (n = 101)

cations), positive (PPV) and negative predictive value (NPV), and their 95% confidence intervals (CI) were calculated. For comparison between the two tests, the AUC values were compared using the z-test. Statistical analyses were performed in XLSTATPM (version 2017; XLSTAT; Addinsoft, New York, USA). Ethics, funding, and potential conflicts of interest. Approval of the institutional review board was obtained (EK 1156/2016). The study was done in accordance with the Declaration of Helsinki. This research did not receive any funding. The authors have no competing interests to declare.

Results Patient demographic data and infection characteristics A total of 101 patients (63 women), fulfilled the inclusion criteria. The median age was 71 years (22–91). Previously performed surgeries were 53 total hip arthroplasties, and 48 total knee arthroplasties. Finally, 29 patients were diagnosed with PJI and 72 cases were classified as aseptic failure according to the MSIS criteria. In 1 of the 29 septic cases, a PJI was diagnosed due to a sinus tract, 2 positive cultures with a phenotypically identical Staphylococcus aureus, and positive minor criteria (permanent section: Type 2, elevated CRP). In 10 of the septic cases, at least 3 minor criteria were fulfilled but no major criterion. In the remaining 18 septic cases, 2 or more periprosthetic positive cultures with the phenotypically identical organisms plus positive minor criteria were present. None of the criteria was positive in 53 of the 72 aseptic cases. In 17 patients, only 1 criterion of the MSIS criteria was fulfilled. In the remaining 2 cases, 2 minor criteria were positive. Performance of diagnostic tests The median CRP (n = 100) was 5.7 mg/L (0.03–252). The sensitivity and specificity of serum CRP was 79% (CI 61–90)

0

10

20

30

40

50

60

70

80

90 100%

Figure 1. Sensitivity (■) and specificity (■) of different tests in diagnosing PJI when using the MSIS criteria.

and 82% (CI 71–89), respectively (Figure 1). The median WBC count (n = 62) in the synovial fluid was 5,167 cells per µL (< 1.0–55,140). The sensitivity and specificity of synovial fluid WBC count was 79% (CI 59–91) and 94% (C: 79–99), respectively. The sensitivity of synovial fluid culture (n = 88), tissue culture (n = 90), and sonication (n = 94) was 44% (CI 28–63), 59% (CI 41–74), and 63% (CI 44–78), respectively. The specificity was, in all 3 different culture methods, 100% (CI 93–100). The most commonly isolated microorganism was Staphylococcus aureus (n = 5), followed by coagulase negative Staphylococci (n = 4) and Escherichia coli (n = 3) (Table 1). 9 septic cases were culture negative. Overall, the sensitivity, specificity, accuracy, positive predictive value, negative predictive value, positive likelihood ratio, and negative likelihood ratio of the permanent section were 90% (CI 73–97), 92% (CI 83–96), 91% (CI 86–97), 81% (CI 68–95), 96% (CI 91–100), 11 (CI 5–23), and 0.11 (0.04– 0.33), respectively. Performance of frozen sections and alpha defensin Of the 284 samples from the 101 cases, 68 were positive for infection (n = 68 of 284 [24%]). In 30 cases (n = 30 of 101), at least 1 frozen section sample was positive. Overall percentage agreement between frozen section and definitive histology was 98.9% (CI 97.8–100%). A Cohen’s kappa of 0.97 (CI 0.94–1.0) indicated a near perfect agreement between frozen and permanent section. The positive and negative percentage agreement was 96% (CI 88–99) and 100% (CI 98–100), respectively. 3 samples showed a discrepancy between the results of frozen and permanent sections. All 3 tissue samples showed a negative frozen section, while the permanent section revealed an infection (all Type 3 according to Krenn classification). There was no statistically significant difference between the AUCs of the frozen section and the permanent section (p = 0.7). The frozen sections showed 25 true positive, 4 false positive, 67 true negative, and 5 false negative test results (Table 2).


108

Acta Orthopaedica 2019; 90 (2): 105–110

Table 2. Performance of the frozen section and alpha defensin using MSIS criteria for diagnosis of PJI Performance

Frozen section

Alpha defensin

Sensitivity (%) Specificity (%) Accuracy (%) Positive predictive value (%) Negative predictive value (%) Positive likelihood ratio Negative likelihood ratio Area under the curve

86.2 (68.7–95.0) 93.1 (84.3–97.3) 91.1 (85.5–96.6) 83.3 (70.0–96.7) 94.4 (89.0–99.7) 12.4 (5.3–29.3) 0.15 (0.06–0.37) 0.90 (0.83–0.97)

69.0 (50.6–82.8) 94.4 (86.0–98.2) 87.1 (80.6–93.7) 83.3 (68.4–98.2) 88.3 (81.1–95.5) 12.4 (4.6–33.2) 0.33 (0.19–0.57) 0.82 (0.73–0.91)

The 95% confidence interval is presented in parentheses

False negative and false positive results 10 cases with false negative results (frozen section [n = 4], alpha defensin [n = 9]) were found. In 3 patients (Table 3: Patients 2, 3, and 10), both diagnostic methods were negative while a PJI with at least 2 phenotypically identical microorganisms was present. In 6 patients with PJI (Patients 1, 4, 5, 7, 8, and 9), the alpha defensin test was negative, while the frozen section showed an infection. Moreover, in the frozen or permanent section of another septic case (Patient 6), no infection was found, whereas alpha defensin was positive and Cutibacterium acnes was identified in culture analysis (2/4 positive tissue cultures, positive sonication) (Table 3). On the other hand, using the MSIS criteria, 7 cases were classified as false positive (frozen section [n = 5], alpha defensin [n = 4]). In 2 cases (Patients 15 and 17), only the frozen

and permanent sections were positive, while no other MSIS criterion was fulfilled. In another patient (Patient 16) only the alpha defensin test was positive. These 3 cases were categorized as true false positive cases. In 2 patients (Patients 11 and 12), the positive results of frozen section and the alpha defensin test were regarded as potentially true positive. Both cases revealed a positive histopathology (Krenn classification: Type 2 [Patient 11], Type 3 [Patient 12]). In another patient with positive frozen section (Patient 14), only 2 minor criteria were fulfilled. Therefore, no PJI was present according to the MSIS criteria. Nevertheless, the synovial fluid WBC count was elevated (16,420 cells/µL) and the histopathology showed an infection as well (Type 2). The 3 latter cases could therefore be undetected infections when using the MSIS criteria (Ochsner et al. 2016, Renz et al. 2018). Moreover, the positive result of 1 alpha defensin test (Patient 13) is unclear. In this patient, a borderline leukocyte count (2,212 cells/µL) and elevated percentage of polymorphonuclear neutrophils (88%) was observed, while all other criteria were negative. However, both results (alpha defensin and leukocyte count) were categorized as false positive. Comparison between frozen sections and alpha defensin The AUCs of frozen section and the alpha defensin test were 0.90 (CI 0.83–0.97) and 0.82 (CI 0.73–0.91), respectively (Figure 2). The difference of both AUCs was 0.023. Of note, the AUCs of the frozen section and the alpha defensin test showed a statistically significant difference (p = 0.006).

Table 3. False negative and false positive alpha defensin test or frozen section based on Musculoskeletal Infection Society (MSIS) criteria Alpha Patient Sex Age Joint MSIS defensin

Frozen Permanent Sinus CRP WBC PMN section section a tract (mg/L) (cells/µL) (%)

Culture b

Antibiotics

False negatives 1 M 59 Knee PJI Negative Positive Type 3 No 11.2 1,456 67 Staph. epidermidis No 2 F 78 Hip PJI Negative Negative Type 3 No 6.6 645 54 Enterococcus faecalis No 3 M 81 Hip PJI Negative Negative Type 1 No 23.0 ND ND MRSA No 4 M 44 Hip PJI Negative Positive Type 2 No 65.1 91,343 92 No growth No 5 M 22 Hip PJI Negative Positive Type 3 No 45.0 44,853 91 No growth No 6 F 86 Hip PJI Positive Negative Type 1 No 17.5 12,987 86 Cutibacterium acnes No 7 F 82 Hip PJI Negative Positive Type 2 No 40.2 26,640 97 Staph. sacchrolyticus No 8 F 71 Knee PJI Negative Positive Type 3 No 46.6 14,638 82 No growth No 9 F 83 Knee PJI Negative Positive Type 2 No 160.0 5,068 87 No growth Yes 10 F 61 Hip PJI Negative Negative Type 1 No 8.9 ND ND Moraxella osloensis No False positives 11 F 71 Hip AF Positive Positive Type 3 No 4.0 ND ND No growth No 12 F 60 Hip AF Positive Positive Type 2 No 2.9 < 1.0 – No growth No 13 F 69 Knee AF Positive Negative Type 1 No 5.2 2,212 88 No growth No 14 F 71 Hip AF Negative Positive Type 2 No 2.6 16,420 71 No growth No 15 M 74 Hip AF Negative Positive Type 2 No 1.6 < 1.0 – No growth No 16 F 69 Hip AF Positive Negative Type 1 No 11.8 < 1.0 – No growth No 17 M 45 Knee AF Negative Positive Type 3 No 3.4 < 1.0 – No growth Yes a According b 2 or more

to Krenn and Morawietz classification (Morawietz et al. 2009). periprosthetic cultures grew phenotypically identical organisms. None of these patients showed bacterial growth in only 1 sample (synovial fluid, tissue, sonication fluid). PJI = periprosthetic joint infection; AF = aseptic failure; CRP = serum C-reactive protein; WBC = synovial fluid leukocyte count, PMN = polymorphonuclear neutrophils.


Acta Orthopaedica 2019; 90 (2): 105–110

109

Table 4. Preoperative and postoperative diagnosis of PJI according to the MSIS criteria and intraoperative diagnosis by using alpha defensin lateral flow test and frozen section. Values are frequency (percentage)

Sensitivity 1

0.8

Intraoperative positive Patients Preop. Postop. alpha frozen (n) PJI PJI defensin test section

0.6

0.4

0.2

0

frozen section alpha defensin 0

0.2

0.4

0.6

0.8

1

1 – specificity Figure 2. Receiver operating characteristic curves for diagnostic accuracy of periprosthetic joint infection based on the alpha defensin lateral flow test and the frozen section when using the MSIS criteria. There is a statistically significant difference between the two ROC curves (p = 0.006).

Intraoperative performance Preoperatively, only 8 of the 29 PJI cases were classified as septic (Table 4). Among these septic cases, 21 did not meet the criteria for PJI preoperatively. 4 did not fulfill any criterion before revision surgery, 4 were positive for only 1 minor criterion, and 13 cases had 2 positive minor criteria. Among the 21 septic cases, the intraoperative alpha defensin test was positive in 13 and frozen section was positive in 17 cases. There was no statistically significant difference between the two tests in intraoperative diagnosis (Fisher’s exact test, p = 0.3). Of the 57 cases with postoperative aseptic failures not fulfilling a single positive preoperative criterion, 4 had a positive frozen section (Table 3: Patients 11, 12, 15, 17) and 2 had a positive alpha defensin test (Patients 11 and 12). 15 cases diagnosed with aseptic failure met 1 preoperative positive criterion for PJI, of whom 1 patient had a positive frozen section (Patient 14) and 2 were positive for alpha defensin test (Patients 13 and 16).

Discussion The preoperative test results showed no definitive evidence of infection in three-quarters of the cases that actually qualified for infection based on postoperative MSIS criteria. When such a diagnosis (PJI, aseptic failure) cannot be made preoperatively, a rapid and accurate screening test is required to exclude PJI intraoperatively. According to the clinical practice guidelines supported by the American Academy of Orthopedic Surgeons, frozen sections can be useful in ruling in PJI intraoperatively (Della Valle et al. 2011, Parvizi et al. 2013). In addition, attention has also been paid to the alpha defensin lateral flow test to distinguish between infection and aseptic failure after a total joint replacement. Therefore, we evaluated the performance of the alpha defensin test and frozen section

Minor criteria 0 positive 61 1 positive 19 2 positive 13 3 positive 7 Major criterion 1

0 0 0 7 1

(0) 4 (7) (0) 4 (21) (0) 13 (100) (100) 7 (100) (100) 1 (100)

3 1 9 6 1

(75) (25) (69) (86) (100)

3 2 12 7 1

(75) (50) (92) (100) (100)

Minor criterion (serum CRP, synovial fluid leukocyte count, or percentage of polymorphonuclear neutrophils, synovial fluid culture). Major criterion (sinus tract). PJI = periprosthetic joint infection.

for ruling in PJI intraoperatively. To date, no comparative study of the performance of these diagnostic tests has been reported. Our results show that among preoperative cases with ambiguous diagnosis of infection (n = 21), the intraoperative frozen section and the alpha defensin test were able to confirm PJI in 17/21 and 13/21 of the cases respectively, suggesting that both these methods are reliable in the diagnosis of infection intraoperatively. However, frozen sections yielded very good diagnostic accuracy with a high sensitivity and specificity in diagnosing PJI (when analyzed by an experienced pathologist). There was no statistically significant difference in the ROC curves between frozen and permanent sections. Additionally, a near perfect agreement (Cohen’s kappa = 0.97) between the two histopathological analyses was shown. Overall, a 99% concordance of all cases could be illustrated, which is in line with the reported rates ranging from 95% to 98% in the literature (Wong et al. 2005, Stroh et al. 2012, Kwiecien et al. 2017). The very low discrepancy (1%) usually occurs due to differences in the quality of the sections and samplings. In the hands of experienced pathologists, frozen sections are as reliable as definitive histology and the technique scores in terms of cost-effectiveness, simplicity, and timely results. In addition, the frozen sections outperformed the alpha defensin test in our study. The alpha defensin test showed a lower sensitivity (69%) when compared with frozen section (86%), and a statistically significant difference in the ROC curves between the two tests. As described by Renz et al. (2018), the alpha defensin test therefore does not appear as an appropriate intraoperative PJI screening test. However, in situations in which the frozen section is unavailable (for example unavailable experienced pathologist) and/or a PJI cannot be confirmed or excluded preoperatively, the alpha defensin test may be a useful adjunct, especially as a confirmatory test, due to its high specificity.


110

Limitations of this study include, first, that the interpretation of the frozen and permanent sections is pathologist dependent and may be influenced by the results of other diagnostic tests. To limit this potential bias, analysis and interpretation is blinded and independent at our institution. Second, the tissue sampling is surgeon dependent and can result in selection bias by choosing samples with obvious presence/absence of infection, thus possibly altering the performance of frozen sections. Third, the number of collected tissue samples for frozen section varied between 1 and 8 per patient, which might have affected the final performance (higher numbers increase the sensitivity at the cost of specificity (Athanasou et al. 1995). In addition, in few cases, not all required test results were available for the infection evaluation when using the MSIS criteria (Figure 1), which is the reality in clinical routine (Bonanzinga et al. 2017). Finally, although MSIS criteria are considered the gold standard in diagnosing PJI, these criteria may miss some patients with PJI, especially infections caused by lowvirulence organisms (Kwiecien et al. 2017, Renz et al. 2018). In conclusion, the frozen section technique showed high performance and a near perfect concordance with the definitive histology, thus strengthening its position as an appropriate intraoperative PJI screening test in diagnosing PJI, especially when the results of preoperative tests are not interpretable. Nevertheless, some institutions do not have the opportunity to analyze tissue samples intraoperatively. In such cases, the alpha defensin test shows great advantage in confirming the diagnosis of PJI. Although sensitivity was lower compared with frozen sections, this test is highly specific and delivers timely results, thus allowing quick decision-making during surgery. SIK, HJ, LS, SS, SK, HG, KB, and WR: Substantial contributions to research design and data acquisition. SIK, HJ, WR: statistical analysis and interpretation. SIK, LS, WR: drafting and revising the paper. Acta thanks Christof Wagner for help with peer review of this study.

Athanasou N A, Pandey R, de Steiger R, Crook D, Smith P M. Diagnosis of infection by frozen section during revision arthroplasty. J Bone Joint Surg Br 1995; 77(1): 28-33. Bonanzinga T, Zahar A, Dutsch M, Lausmann C, Kendoff D, Gehrke T. How reliable is the alpha-defensin immunoassay test for diagnosing periprosthetic joint infection? A prospective study. Clin Orthop Relat Res 2017; 475(2): 408-15. doi: 10.1007/s11999-016-4906-0. Bori G, Soriano A, Garcia S, Mallofre C, Riba J, Mensa J. Usefulness of histological analysis for predicting the presence of microorganisms at the time of reimplantation after hip resection arthroplasty for the treatment of infection. J Bone Joint Surg Am 2007; 89(6): 1232-7. doi: 10.2106/jbjs.f.00741. Butler-Wu S M, Burns E M, Pottinger P S, Magaret A S, Rakeman J L, Matsen F A 3rd, Cookson B T. Optimization of periprosthetic culture for diagnosis of Propionibacterium acnes prosthetic joint infection. J Clin Microbiol 2011; 49(7): 2490-5. doi: 10.1128/jcm.00450-11. Chalifour A, Jeannin P, Gauchat J F, Blaecke A, Malissard M, N’Guyen T, Thieblemont N, Delneste Y. Direct bacterial protein PAMP recognition by human NK cells involves TLRs and triggers alpha-defensin production. Blood 2004; 104(6): 1778-83. doi: 10.1182/blood-2003-08-2820.

Acta Orthopaedica 2019; 90 (2): 105–110

Deirmengian C, Kardos K, Kilmartin P, Cameron A, Schiller K, Booth R E Jr, Parvizi J. The alpha-defensin test for periprosthetic joint infection outperforms the leukocyte esterase test strip. Clin Orthop Rel Res 2015; 473(1): 198-203. doi: 10.1007/s11999-014-3722-7. Della Valle C J, Bogner E, Desai P, Lonner J H, Adler E, Zuckerman J D, Di Cesare P E. Analysis of frozen sections of intraoperative specimens obtained at the time of reoperation after hip or knee resection arthroplasty for the treatment of infection. J Bone Joint Surg Am 1999; 81(5): 684-9. Della Valle C, Parvizi J, Bauer T W, DiCesare P E, Evans R P, Segreti J, Spangehl M, Watters W C 3rd, Keith M, Turkelson C M, Wies J L, Sluka P, Hitchcock K. American Academy of Orthopaedic Surgeons clinical practice guideline on: the diagnosis of periprosthetic joint infections of the hip and knee. J Bone Joint Surg Am 2011; 93(14): 1355-7. doi: 10.2106/ JBJS.9314ebo. Ganz T, Selsted M E, Szklarek D, Harwig S S, Daher K, Bainton D F, Lehrer R I. Defensins: natural peptide antibiotics of human neutrophils. J Clin Invest 1985; 76(4): 1427-35. doi: 10.1172/jci112120. Gehrke T, Lausmann C, Citak M, Bonanzinga T, Frommelt L, Zahar A. The accuracy of the alpha defensin lateral flow device for diagnosis of periprosthetic joint infection: comparison with a gold standard. J Bone Joint Surg Am 2018; 100(1): 42-8. doi: 10.2106/jbjs.16.01522. Kwiecien G, George J, Klika A K, Zhang Y, Bauer T W, Rueda C A. Intraoperative frozen section histology: matched for Musculoskeletal Infection Society criteria. J Arthroplasty 2017; 32(1): 223-7. doi: 10.1016/j. arth.2016.06.019. Morawietz L, Tiddens O, Mueller M, Tohtz S, Gansukh T, Schroeder J H, Perka C, Krenn V. Twenty-three neutrophil granulocytes in 10 high-power fields is the best histopathological threshold to differentiate between aseptic and septic endoprosthesis loosening. Histopathology 2009; 54(7): 84753. doi: 10.1111/j.1365-2559.2009.03313.x. Musso A D, Mohanty K, Spencer-Jones R. Role of frozen section histology in diagnosis of infection during revision arthroplasty. Postgrad Med J 2003; 79(936): 590-3. Ochsner P, Borens O, Bodler P, Schweizerische Gesellschaft für Orthopädie und Traumatologie. Infections of the musculoskeletal system: basic principles, prevention, diagnosis and treatment. Grandvaux, Switzerland: Swiss Orthopaedics and the Swiss Society for Infectious Diseases expert group “Infections of the musculoskeletal system”; 2016. Parvizi J, Gehrke T. Definition of periprosthetic joint infection. J Arthroplasty 2014; 29(7): 1331. doi: 10.1016/j.arth.2014.03.009. Parvizi J, Gehrke T, Chen A F. Proceedings of the International Consensus on Periprosthetic Joint Infection. Bone Joint J 2013; 95-b(11): 1450-2. doi: 10.1302/0301-620x.95b11.33135. Piper K E, Fernandez-Sampedro M, Steckelberg K E, Mandrekar J N, Karau M J, Steckelberg J M, Berbari E F, Osmon D R, Hanssen A D, Lewallen D G, Cofield R H, Sperling J W, Sanchez-Sotelo J, Huddleston P M, Dekutoski M B, Yaszemski M, Currier B, Patel R. C-reactive protein, erythrocyte sedimentation rate and orthopedic implant infection. PloS One 2010; 5(2): e9358. doi: 10.1371/journal.pone.0009358. Renz N, Yermak K, Perka C, Trampuz A. Alpha defensin lateral flow test for diagnosis of periprosthetic joint infection: not a screening but a confirmatory test. J Bone Joint Surg Am 2018; 100(9): 742-50. doi: 10.2106/ jbjs.17.01005. Schafer P, Fink B, Sandow D, Margull A, Berger I, Frommelt L. Prolonged bacterial culture to identify late periprosthetic joint infection: a promising strategy. Clin Infect Dis 2008; 47(11): 1403-9. doi: 10.1086/592973. Sigmund I K, Holinka J, Gamper J, Staats K, Bohler C, Kubista B, Windhager R. Qualitative alpha-defensin test (Synovasure) for the diagnosis of periprosthetic infection in revision total joint arthroplasty. Bone Joint J 2017; 99-b(1): 66-72. doi: 10.1302/0301-620x.99b1.bjj-2016-0295.r1. Stroh D A, Johnson A J, Naziri Q, Mont M A. How do frozen and permanent histopathologic diagnoses compare for staged revision after periprosthetic hip infections? J Arthroplasty 2012; 27(9): 1663-8.e1. doi: 10.1016/j. arth.2012.03.035. Wong Y C, Lee Q J, Wai Y L, Ng W F. Intraoperative frozen section for detecting active infection in failed hip and knee arthroplasties. J Arthroplasty 2005; 20(8): 1015-20. doi: 10.1016/j.arth.2004.08.003.


Acta Orthopaedica 2019; 90 (2): 111–118

111

Clinical significance of cervical MRI in brachial plexus birth injury Petra GRAHN ¹, Tiina PÖYHIÄ ², Antti SOMMARHEM ¹, and Yrjänä NIETOSVAARA ¹ 1 New

Children’s Hospital, HUS Helsinki University Hospital, Department of Children’s Orthopedics and Traumatology, Helsinki; 2 HUS Medical Imaging Center, HUS Helsinki University Hospital, Department of Radiology, Helsinki, Finland Correspondence: petra.grahn@hus.fi Submitted 2018-01-31. Accepted 2018-11-16.

Background and purpose — Patient selection for nerve surgery in brachial plexus birth injury (BPBI) is difficult. Decision to operate is mostly based on clinical findings. We assessed whether MRI improves patient selection. Patients and methods — 157 BPBI patients were enrolled for a prospective study during 2007–2015. BPBI was classified at birth as global plexus injury (GP) or upper plexus injury (UP). The global plexus injury was subdivided into flail upper extremity (FUE) and complete plexus involvement (CP). Patients were seen at set intervals. MRI was scheduled for patients that had either GP at 1 month of age or UP with no antigravity biceps function by 3 months of age. Type (total or partial avulsion, thinned root), number and location of root injuries and pseudomeningoceles (PMC) were registered. Position of humeral head (normal, subluxated, dislocated) and glenoid shape (normal, posteriorly rounded, pseudoglenoid) were recorded. Outcome was assessed at median 4.5 years (1.6–8.6) of age. Results — Cervical MRI was performed on 34/157 patients at median 3.9 months (0.3–14). Total root avulsions (n = 1–3) were detected on MRI in 12 patients (8 FUE, 4 CP). Reconstructive surgery was performed on 10/12 with total avulsions on MRI, and on all 10 with FUE at birth. Sensitivity and specificity of MRI in detecting total root avulsions was 0.88 and 1 respectively. Posterior shoulder subluxation/dislocation was seen in 15/34 patients (3.2–7.7 months of age). Interpretation — Root avulsion(s) on MRI and flail upper extremity at birth are both good indicators for nerve surgery in brachial plexus birth injury. Shoulder pathology develops very early in permanent BPBI.

Neonatal brachial plexus injury (BPBI) occurs in 0.5–4/1,000 vaginal births (Hoeksma et al. 2000, Foad et al. 2008, Pöyhiä et al. 2010). Approximately 80% of these patients recover spontaneously during their first year of life (Pöyhiä et al. 2010). In patients with a permanent injury BPBI causes muscle changes in infants often leading to shoulder and elbow contractures (Eisman et al. 2015, Gharbaoui et al. 2015). Without treatment at least one-third of these develop posterior instability, subluxation, and deformity of the glenohumeral joint (Hoeksma et al. 2000, 2003, Moukoko et al. 2004). Patients with root avulsions have a poorer prognosis compared with patients without (Kirjavainen et al. 2007). There is evidence that some of the patients with a permanent palsy benefit from surgical repair of the lesion (Waters 2005, Hale et al. 2010). The extent and the type of root injury can be evaluated by clinical, neurophysiological, and different radiographic methods. Different grading systems (Narakas classification, Gilbert shoulder and Gilbert–Raimondi classification, Active Movement Scale and 3-month Toronto Test Score, 9-month Cookie Test) (Narakas 1986, Curtis et al. 2002, Haerle and Gilbert 2004, Borschel and Clarke 2009, Bade et al. 2014) have been developed for prognostic purposes in an attempt to help in surgical decision-making. At present, however, there is no consensus regarding the indications and timing of surgery in BPBI. We have prospectively studied the clinical significance of high-resolution cervical MRI in BPBI patients who were considered for brachial plexus surgery during a 9-year period from 2007 to 2015. Our hypothesis was that evidence of total root avulsion injury on MRI is a good indicator for surgical repair.

Patients and method During the study period between 2007 and 2015 altogether 157 BPBI patients were referred to our brachial plexus clinic, which serves as a tertiary treatment center for a population of © 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2018.1562621


112

2 million people. All children were examined at birth by the referral centers’ pediatrician at a median age of 1 day (0–2) and at a median of 2 days (0–7) by a physiotherapist. Severity of the injury was classified at birth to global plexus injury or upper plexus injury (UP), UP meaning shoulder and elbow and in some patients wrist extension affected. Global plexus injury was further subclassified as flail upper extremity (FUE); no movement at all in the affected limb, complete plexus involvement (CP); shoulder, elbow, wrist, and hand affected. Presence of Horner’s sign was documented. Once referred to our clinic all patients were examined by a BPBI specialized team consisting of a hand surgeon, occupational therapist, and physiotherapist. Patients with persisting palsy were scheduled to be seen on regular basis by the same team at set time intervals from 1 month of age (at 3, 6, and 12 months, 2, 4, 7, 10, and 14 years of age). Active and passive range of motion (ROM) of all upper extremity joints were measured at each appointment using a goniometer. Muscle strength for shoulder abduction and flexion, elbow and wrist extension, and flexion, thumb and finger extension, and flexion was evaluated using the Medical Research Council’s scale for muscle strength. The 3-month Toronto Test Score was retrospectively calculated, since it was not used routinely in our institution during the study period. MRI was scheduled for patients who had either GP at 1 month of age or UP with no antigravity biceps function by 3 months of age. High-resolution cervical MRI (1.5T Philips Medical Systems, Achieva; Philips Healthcare, Best, Netherlands) was performed under general anesthesia. After localizer sequences, T1-weighted (T1-W) spin-echo images in sagittal plane were obtained. T2-weighted (T2-W) spin echo images were obtained in axial, sagittal, and coronal planes. Slice thickness was 2 mm for T2 coronal sequence and 3 mm for all other sequences. Heavily T2 weighted BFFE sequence in coronal and axial planes with 0.5 mm slice thickness allowed MR myelography view of the roots in each study. Type and number of root injuries (no avulsion, thinned roots, partial avulsion, and total avulsion) as well as location of pseudomeningoceles (PMC) were registered. Total root avulsion was defined as both anterior and posterior roots avulsed from the spinal cord. Partial avulsion was defined as either anterior or posterior root avulsed from the spinal cord. Thinned roots are seen on MRI when some of the rootlets emerging from the spinal cord, forming the anterior or posterior root, are ruptured (Silbermann-Hoffmann and Teboul 2013, Tse et al. 2014). T2-weighted axial sequences also covered both shoulders and therefore position (normal, posteriorly subluxated, posteriorly dislocated) of both humeral heads and the shape (normal, posteriorly rounded, pseudoglenoid) of both glenoids was recorded as well as the glenoscapular angle (GSA). A pediatric radiologist (TP), with more than 15 years of experience in musculoskeletal MRI, evaluated all images. Brachial plexus exploration was recommended to all patients with total root avulsion(s) on MRI. If no total avulsion(s) were

Acta Orthopaedica 2019; 90 (2): 111–118

detected observation was continued for another 3 months. Surgery was then again recommended if no improvement was clinically observed. The length of time in days from MRI referral to MRI examination and to brachial plexus surgery was recorded. Both sensitivity and specificity for total avulsions and PMC on MRI was calculated in relation to the intraoperative findings. Findings in MRI and surgery were also compared with clinical outcome at a mean follow-up of 4.5 years (1.6–8.6) to assess given treatment. None of the patients were lost during follow-up. Statistics Sensitivity and specificity for the MRI findings in comparison with the intraoperative findings as well as PMC in relation to root avulsion injury on MRI were calculated. The 95% confidence intervals (CI) were calculated using Wilson score intervals. Linear regression models were fitted for GSA difference and model assumptions were visually assessed. Statistical analysis was done using R program for statistical computing (R Foundation for Statistical Computing, Vienna, Austria. R Core Team 2017). The significance level p < 0.05 was used. Ethics, funding, and potential conflicts of interest The Ethics Committee of our hospital approved this study (registration number 79/E7/2001). Parental consent was gathered. No funding was received. No conflicts of interest were declared.

Results Inclusion criteria were met by 34/157 patients. Cervical MRI was performed on all 34 patients (18 boys) at median 3.9 months of age (0.3–14.3). Mean birth weight was 4,276 g (3,480–5,400). 22 of the injuries were on the right side. 1 of the 34 patients who had cervical MRI had a bilateral injury after breech delivery. 4 patients with an FUE at birth had a positive Horner’s sign. Our diagnostic and treatment protocol could not be followed exactly as planned for patient, parent, and hospital related reasons. Children with persisting FUE or CP (n = 18) had the referral for MRI at median 2.1 month of age (0.2–3.5) and the MRI was performed at median 3.5 months of age (0.3–14.3) respectively. The respective ages of children with UP and no antigravity biceps function by 3 months of age (n = 16) were 3.0 months (0.9–6.2) and 4.0 months (1.7–7.7). Median time to MRI from referral was 28 days (1–70), excluding patient number 13 whose MRI was postponed twice (up to 328 days) for miscellaneous reasons. Based on the preliminary findings of this study 1 child (patient 32) with an FUE at birth was immediately referred for MRI. Altogether 170 root levels were examined. 18 total root avulsions were detected in 12/34 patients (Figure 1).


Acta Orthopaedica 2019; 90 (2): 111–118

113

B

A

Figure 1. Coronal (A) and axial (B) BFFE MR images (0.5 mm) in a 3-month-old boy (patient 18) with brachial plexus birth injury on the right side. Total (both ventral and dorsal roots) avulsion of right C6 root with a PMC (asterisk): ventral root is avulsed from the cord (upper arrow), where a short stump of dorsal root is seen (lower arrow). Left ventral and dorsal C6 roots (arrowheads) are normal.

Figure 2. Axial BFFE MR image (0.5 mm) in a 4-month-old girl (patient 26) with brachial plexus birth injury on the right side. Partial avulsion of C8 root: ventral root is avulsed (red arrow), dorsal C8 root is intact (arrowhead). Left C8 nerve roots are normal (arrowheads).

B A Figure 3. Coronal BFFE image (0.5 mm) in a 3-month-old girl (patient 16) with brachial plexus birth injury on the right side. Right ventral C6 root is thinned (arrow) compared to the normal left ventral C6 root (arrowhead).

Figure 4. Coronal (A) and axial (B) BFFE image (0.5 mm) in 3-month-old boy (patient 31) with brachial plexus birth injury on both sides. (b) Intact ventral and dorsal nerve roots (red arrows) at C7 level despite a PMC clearly visible on both sides.

6 patients had partial avulsions only (dorsal root 2, ventral root 5) (Figure 2). 4 patients with total or partial root avulsions also had thinning of additional roots (Figure 3, Table 1). The most extensive injury was in patient number 8, who had total avulsions of C6–8 with thinning of both the ventral and dorsal C5 rootlets. The number of totally avulsed roots per patient varied from 1 to 3. The most commonly totally avulsed root was C8. Sensitivity and specificity of MRI in detecting total root avulsions was 0.88 (CI 0.5–1) and 1 (CI 1–0.9). PMC was seen in association with all 18 total root avulsions, and in 6 of the 8 partial avulsions at the level of the avulsion. 2 patients had PMC without evidence of root injuries (Figure 4). Specificity and sensitivity of PMC for total nerve root avulsion was 0.44 and 1. Asymmetry (> 5°) in GSA was recorded in 22 patients with a mean difference of 17° (6–35) (Table 2). GSA difference was modelled using linear regression with findings at birth and age at MRI as the covariates. Patient 13 was excluded due to significant delay until MRI. Both univariable and multivari-

able models were fitted. The findings at birth did not statistically significantly associate with the GSA difference in either the univariable or the multivariable models (p > 0.05 for both FUE and CP when compared with UP in both models). The age at MRI was associated significantly with GSA difference in both models; 4.7 (CI 3–6.5) per year in univariable and 5 (CI 3–7) in the multivariable model, p < 0.001 in both cases (Figure 5). Glenoid shape was normal in 20 patients, with a trend towards more severe incongruence in the patients with an older age at MRI. Reconstructive nerve surgery was recommended to all 12 patients with total avulsions on MRI and to 7 of the 22 patients without total avulsions. Parents consented to surgery in 10 patients with total avulsions on MRI (reconstruction with autologous nerve grafts 7, spinal accessory nerve (SAN) pro suprascapular nerve (SSN) transfer 2, contralateral C7 transfer 1), and 6 patients without total avulsions respectively (reconstruction with autologous nerve grafts 4, SAN pro SSN transfer 2). Median age at primary reconstruc-


114

Acta Orthopaedica 2019; 90 (2): 111–118

Table 1. Patient demographics. Patients are arranged in descending order of abnormal findings on MRI MRI findings b

Findings at birth

Plexus reconstruction

Extent 3-month Age Total root Partial root Thinning Patient of injury a test score (months) avulsion avulsion of roots PMC 8 25 15 34 22 3 32 9 20 24 13 18 17 11 21 1 26 10 14 28 16 31 19 27 29 23 6 33 30 2 7 4 5 12

FUE c FUE FUE c CP CP FUE FUE c FUE FUE CP FUE CP FUE CP FUE c CP CP CP UP UP CP UP UP CP CP CP UP UP CP CP UP UP UP UP

0 0 0.3 2.1 2.8 0 d 0 1.3 1.8 2.6 3.2 0.6 1.2 1.3 2.1 2.4 3.8 4.8 4.5 4.8 3.8 4.8 2.5 4.2 4.5 4.8 5.2 5.2 5.5 5.8 5.8 5.8 5.8

0.9 2.7 2.6 2.5 3.2 4.0 0.3 3.4 4.0 3.7 14.3 3.0 1.9 4.6 3.9 6.4 4.5 3.3 7.1 3.9 3.9 3.5 7.7 3.9 3.2 4.4 3.2 4.3 3.9 4.6 3.4 4.7 4.1 1.7

C6–8 C8–T1 C8–T1 C7–8 C7–8 C8 C8 C8 C8 C8 C7 C6

C7D C6V, C8V C8V C8V C6V C6D C6D

C5vd C7vd C6v C6v

C5–8 C8–T1 C8–T1 C7–8 C7–T1 C8 C7–T1 C8 C8–T1 C8 C7 C6 C6 8 C8 C8 C6 C6 C5 6 7 C8

Age Plexus (months) reconstruction 6.6 4.3 3.4 7.4 0.8 4.4 5.6 5.7 27.9 7.3 3.9 6.6 9.0 5.6 8.0 4.8

CC7 yes yes refused refused yes yes yes yes yes yes yes yes refused yes

a FUE = Flail upper extremity, CP = Complete plexus involvement, UP = Upper plexus involvement, b V =Ventral root, D = Dorsal root, v =Ventral root thinning, d = Dorsal root thinning, SAN = Spinal accessory

scapular nerve. c Horner sign. d Primary surgery before 3 months of age

tive nerve surgery was 5.6 months (0.8–9) excluding patient number 13 who was operated at 28 months of age. According to the retrospective calculation of the 3-month Toronto Test Score, 18 of the 19 patients for whom we recommended plexus surgery had a Test Score less than 3.5, which is an indication for plexus reconstruction (Borschel and Clarke 2009) (Table 1). Intraoperative findings concerning total avulsions were compared with corresponding findings on MRI (Table 3). Sensitivity and specificity of MRI in detecting total nerve root avulsions was 0.88 and 1. Three total C8 avulsions, one accompanied by a total C7 avulsion, were left unexplored due to good hand and wrist function at the time of surgery (Table 3). Median time from MRI to primary surgery was 49 days (13–173) excluding patient number 13 whose MRI and operation were delayed.

Converted to SAN pro SSN

yes

yes

yes

yes

nerve, SSN = Supra-

None of the 34 patients recovered completely during follow-up. Additional surgery was performed in 4 patients with total root avulsions, 1 patient with partial root avulsions, and 9 patients without root injuries (Table 4). Retrospectively, assessed by final outcome (Table 4) expressed by ratios (injured vs. uninjured side) of active antigravity shoulder, elbow, wrist, and finger ROM, all patients who had total root avulsions on MRI and all children born with FUE would have benefited from plexus surgery. On the other hand, based on the final outcome (Table 4), 2 (patients 7 and 19) of the 10 patients with upper plexus palsy at birth might have benefited from plexus reconstruction. Retrospectively analyzed, both of these patients would have failed the Cookie Test at 9 months. When looking at the patient final outcome, partial root avulsion alone, or in combination with thinned rootlets (6 patients), had no clinical significance (Table 4).


Acta Orthopaedica 2019; 90 (2): 111–118

115

Table 2. Glenohumeral joint (GHJ) MRI findings. Patients are arranged primarily in descending order based on incongruency of their affected shoulder, secondarily in descending order based on the difference between GSA of both shoulders Findings MRI age Glenoid GSA (°) GSA (°) GSA Patient at birth a (months) GHJ shape affected normal difference 19 UP 7.7 D PG –40 –5 35 11 CP 4.6 D PG –40 –6 34 14 UP 7.1 D PG –40 –6 34 5 UP 4.1 D PG –30 –3 27 3 FUE 4.0 D PG –25 3 22 10 CP 3.3 D PG –25 –8 17 16 CP 3.9 D PG –25 –20 5 26 CP 4.5 SL PR –40 –15 25 1 CP 6.4 SL PR –30 –9 21 21 FUE 3.9 SL PR –25 –5 20 4 UP 4.7 SL PR –40 –22 18 24 CP 3.7 SL PR –25 –10 15 9 FUE 3.4 SL PR –30 –20 10 6 UP 3.2 SL PR –25 –15 10 30 CP 3.9 SL N –20 –10 10 27 CP 3.9 N N –15 –3 12 23 CP 4.4 N N –20 –8 12 17 FUE 1.9 N N –20 –10 10 33 UP 4.3 N N –20 –11 9 7 UP 3.4 N N –20 –12 8 28 UP 3.9 N N –15 –8 7 15 FUE 2.6 N N –13 –7 6 28 CP 3.0 N N –13 –7 6 32 FUE 0.3 N N –15 –10 5 34 CP 2.5 N N –15 –10 5 22 CP 3.2 N N –10 –6 4 29 CP 3.2 N N –5 –1 4 8 FUE 0.9 N N –9 –6 3 13 FUE 14.3 N N –7 –10 3 2 CP 4.6 N N –8 –5 3 25 FUE 2.7 N N –5 –7 2 20 FUE 4.0 N N –5 –4 1 12 UP 1.7 N N –5 –5 0 31 UP/UP 3.5 N/N N/N –15/–20 a See

Table 1. D = Dislocated, SL = Subluxed, N = Normal, PG = Pseudoglenoid, PR = Posteriorely rounded

Discussion Clinical evaluation of the extent and type of root injuries in BPBI forms the basis for indication, timing, and planning of surgical repair. Distinction of BPBI patients with axonotmesis type of root injuries, with potential for spontaneous recovery, from infants with root ruptures and/or avulsions that usually benefit from surgical treatment, however, remains a challenge for brachial plexus surgeons. Our aim was to find out whether cervical MRI could be helpful in surgical decision-making in patients with permanent BPBI. CT myelography has long been the gold standard in BPBI diagnostic imaging, but during recent years there has been a clear trend towards MRI, possibly due to the fact that MRI does not involve ionizing radiation or the need for intrathecal

Glenoscapular angle difference 40 Upper plexus injury Flail upper extremity Complete plexus involvement 30

20

10

0

0

2

4

6

8

Age at MRI Figure 5. Multivariable model expressing GSA difference in relation to age at time of MRI and clinical findings at birth.

contrast injection. Earlier MRI studies with evaluation of the presence of PMC only (Tse et al. 2014) or of nerve root integrity with 1.5 mm MRI slice thickness (Medina et al. 2006) have demonstrated only moderate sensitivity or specificity levels for root avulsions. In contradiction to these earlier reports we found an excellent correlation between complete root avulsions and surgical findings using 1.5 T MRI with 0.5 mm slice thickness in axial and coronal views. Sensitivity and specificity for complete root avulsion on MRI in our study are in line with the more recent studies of Somashekar et al. (2014) and Menashe et al. (2015). Our study further confirmed that PMC has a high sensitivity but low specificity for total nerve root avulsions on MRI (Yilmaz et al. 1999, Medina et al. 2006). We did not explore all levels that showed root avulsion on MRI and thus left these unconfirmed findings outside the sensitivity and specificity calculation. We also found that evidence of total root avulsion(s) on MRI in itself was a good indicator for brachial plexus exploration and reconstruction. The clinical findings did not always correlate with a complete C8 avulsion on MRI. Hand recovery could therefore not be reliably predicted by MRI in our patients since functional recovery of the hand and wrist were good in some without exploration and surgical repair of C8, despite it appearing completely avulsed on the MRI. These patients underwent surgical reconstruction of the upper plexus, which appeared to be beneficial assessed by the final outcome. Our MRI protocol also enabled imaging of thinned rootlets and partial root avulsions. Their existence did not appear to influence the outcome negatively, which is why it is probably better to leave roots that show evidence of partial avulsion or thinning on MRI unexplored.


116

Acta Orthopaedica 2019; 90 (2): 111–118

that fail the Cookie Test at 9 months is controversial. Untreated posterior shoulder sub C5 C6 C7 C8 T1 PMC Patient MRI Surgery MRI Surgery MRI Surgery MRI Surgery MRI Surgery MRI luxation in BPBI leads to permanent reduction of shoulder ROM and gle9 N N N N N N A A N — Yes nohumeral deformity (Hoeksma et al. 11 N N N N N — N — N — No 15 N N N N N N A A A A Yes 2003, Pöyhiä et al. 2007). Maintenance 17 N N N N N N N N N — No of good passive shoulder ROM, treat18 N N A A N N N — N — Yes ment of posterior subluxation with 20 N N N N D — A — N — Yes 21 N N N N N N N — N — No Botulinum Toxin A injections and 24 N N N N N N A — N — Yes early surgical reduction of the shoul25 N N N N N N A A A A Yes der may prevent these adverse shoul26 N N N N v+d N V N N — Yes 32 N N N N N A A A N — Yes der sequelae in BPBI (El-Gammal et al. 2006, Ezaki et al. 2010, Pöyhiä et N = Normal, A = Avulsion, V = Ventral root avulsion, D = Dorsal root avulsion, al. 2011). Shoulder subluxation prov = Ventral root thinning, d = Dorsal root thinning, — = Not examined True positive was defined as a total root avulsion seen on MRI and detected during exploration. ceeds gradually to dislocation, which True negative was defined as no avulsion seen on MRI nor detected during surgery. Partial was evident also in our study where the and thinned roots seen on MRI as well as roots not explored during surgery were left out of the severity of changes in shoulder concalculation. gruency correlated to patient age. First signs of glenohumeral joint incongruAs of today, there have been no studies published regarding ence were recognized already under 2 months of age in some the use of 3T MRI for root diagnostics in infants with BPBI. of our patients. This is in accordance with the earlier study of Further studies are needed to establish the role of 3T MRI Pöyhiä et al. (2010) where half of the patients with permanent compared with 1.5 T MRI concerning root avulsion diagnos- BPBI and shoulder pathology had already developed posterior tics in BPBI. shoulder subluxation at 3 months of age. FUE with or without a positive Horner’s sign at birth was Our study has limitations despite its prospective nature. a good indicator for a difficult permanent injury and plexus First, the sample size is relatively small; second, our patient reconstruction. All 10 patients in this series with an FUE at population is heterogeneous and third, all totally avulsed roots birth had surgery. Root avulsions were evident in 8/10, and on MRI were not surgically explored. Therefore further studneuromas in continuity in all 10 of these children. This is in ies with more patients are needed to verify our main findings: line with previous findings by Grossman et al. (2004), Hale FUE at birth, total root avulsions on MRI, and/or a 3-month et al. (2010) and Abid et al. (2016), and is the reason why Test Score < 3.5 are good indicators for brachial plexus exploin our practice we have turned towards very early surgical ration and reconstruction in BPBI. exploration in patients with FUE. Some surgeons prefer to use the 3-month Toronto Test Score when evaluating the need for surgery (Borschel and Clarke 2009) and in fact this method PG: Main author, hand surgeon. Part of brachial plexus birth injury team. Clinical work and development of study protocol. TP: Second author, pedigave concurrent recommendations for surgery in 33 of the 34 atric radiologist, and developer of the imaging protocol. Part of brachial patients in our study. 1 patient with a test score of 4.8 was plexus injury team. AS: Co-author, hand surgeon. Part of brachial plexus scheduled for plexus reconstruction, but the operation was birth injury team. Clinical work and development of study protocol. YN: Senior author. Lead of brachial plexus birth injury team. Clinical work and converted to nerve transfer at the time of surgery due to better development of study protocol. than expected recovery (patient 14). In addition, preoperative clinical re-evaluation converted plexus reconstruction to nerve Acta thanks Lars Eldar Myrseth for help with peer review of this study. transfer for 3 more patients. Later analysis of these patients revealed that the decision for conversion was appropriate for 1 patient (patient 27) but 2 patients would most likely have benefited from plexus reconstruction (patients 13 and 22). 2/10 of Abid A. Brachial plexus birth palsy: management during the first year of life. our patients with UP at birth might have benefited from upper Orthop Traumatol Surg Res 2016; 102(Suppl. 1): S125-32. plexus reconstruction or extraplexal neurotization procedures Bade S A, Lin J C, Curtis C G, Clarke H M. Extending the indications for primary nerve surgery in obstetrical brachial plexus palsy. Biomed Res Int since they did not reach above horizontal shoulder abduction, 2014; 2014;2014:627067. doi: 10.1155/2014/627067. Epub 2014 Jan 12. with elbow and wrist movement also clearly compromised. Borschel G H, Clarke H M. Obsterical brachial plexus palsy. Plast Reconstr Retrospectively we found that both of these patients had failed Surg 2009; 124(Suppl.1): 144e-55e. the Cookie Test at 9 months of age. Thus, postponing the deciCurtis C, Stephens D, Clarke H M, Andrews D. The active movement scale: sion to do surgery in patients with no avulsions on MRI, but an evaluative tool for infants with obstetrical brachial plexus palsy. J Hand Surg Am 2002; 27: 470-8. a Toronto Test Score at 3 months below 3.5, or in patients

Table 3. MRI findings compared to intra-operative findings by root level


Acta Orthopaedica 2019; 90 (2): 111–118

117

Table 4. Patient outcome. Outcome expressed as ratio (%) of active antigravity range of motion of the affected side in comparison with the unaffected side. Patients are arranged primarily by the extent of injury at birth and secondarily by their 3-month Test Score Findings 3-month Patient at birth a test score Plexus surgery d

Nerve GHJ Other Outcome transfer relocation surgery A B C D E F G

8 FUE b 0 CC7 8.0 3 FUE 0 refused SAN > AN yes EIP > EPB, 8.6 BR > EDC c 32 FUE b 5 > 8, 6 > UT SAN > ISN 2.6 9 FUE 0 5–6 > UT, SAN > SSN 7.3 25 FUE 0 5 > 6, 6 > 1, 7 > 8, SAN > SSN yes forearm 3.7 osteotomy 15 FUE b 0.3 5 > UT, 6 > MT, 7 > 81, 4.5 SAN > SSN 17 FUE 0.6 5–6 > UT, 7 > MT, SAN > SSN 5.5 11 CP 1.2 5–6 > UT, SAN >SSN Oberlin yes 6.2 21 FUE b 1.3 5–6 > UT, 7 > MT, SAN > SSN FCU > ECR 5.0 20 FUE 1.3 5–6 > UT, SAN > SSN Oberlin TM > IS 4.5 24 CP 1.8 5 > UT, 6 > UT, 7 > MT, 4.2 SAN > SSN 1 CP 2.1 refused yes 5.2 34 CP 2.1 refused 1.6 26 CP 2.4 6 > UT, SAN > SSN 3.2 27 CP 2.5 converted to nerve transfer SAN > SSN 3.0 13 FUE 2.6 converted to nerve transfer SAN > SNN, 5.6 pRN > pAN 22 CP 2.8 converted to nerve transfer SAN > SSN 4.0 18 CP 3.2 5 > 6, SAN > SSN 5.0 10 CP 3.8 SAN > ISN 6.2 31 UP 3.8 SAN > ISN 2.9 29 CP 4,2 SAN > ISN 3.5 28 UP 4.5 2.2 23 CP 4.5 2.9 14 UP 4.8 converted to nerve transfer SAN > SSN 6.1 6 UP 4.8 5.2 19 UP 4.8 yes 4.5 16 CP 4,8 4.0 33 UP 5.2 SAN > ISN 2.4 30 CP 5.2 SAN > ISN 3.2 2 CP 5.5 TM > IS 7.4 7 UP 5.8 yes TM > IS 6.9 4 UP 5.8 yes 2.1 5 UP 5.8 8.4 12 UP 5.8 4.8

CP CP

91 28

0 73

33 17

0 50

0 0

CP CP CP

50 33 36

38 92 50

25 10 10

25 10 40

0 10 20

CP

50

63

72

0

10

CP UP CP UP UP

50 44 50 78 50

75 28 84 88 80

11 61 17 100 100

20 100 50 100 100

10 100 50 100 100

UP 40 100 100 100 100 UP 38 38 25 100 100 UP 44 62 100 100 100 UP 78 81 100 100 100 UP 39 63 56 100 100 UP UP UP UP UP UP UP UP UP UP UP UP UP UP UP UP UP UP

38 83 69 89 72 89 89 66 94 40 88 81 92 60 50 72 94 89

75 88 81 81 81 97 91 90 100 56 91 81 100 87 71 69 81 100

61 100 100 100 100 100 100 100 100 56 100 100 100 100 36 100 100 100

100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100

100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100

a See Table 1. b positive Horner sign c primary surgery before 3 months of age d AN = Axillary nerve, pAN = partial axillary

nerve, BR = Brachioradialis muscle, CC7 = contralateral C7 transfer, ECR = Extensor carpi radialis longus and brevis muscle, EDC = Extensor digitorum communis muscle, EIP = Extensor indicis proprius muscle, EPB = Extensor pollicis brevis muscle, FCU = Flexor carpi ulnaris muscle, IS = Infraspinatus muscle, ISN = Infraspinatus branch of suprascapular nerve, M = Middle trunk, pRN = partial Radial nerve, SAN = Spinal accessory nerve, SSN = Suprascapular nerve, TM = Teres major muscle, UT = Upper trunk A. Final follow-up age (years) B. Findings at final follow-up: CP = Complete plexus involvement, UP = Upper plexus involvement C. Shoulder abduction (%) D. Elbow flexion (%) E. Wrist ex-tension-flexion (%) F. Finger movement (%) G. Intrinsic (%)

Eisman E A, Little K J, Laor T, Cornwall R. Glenohumeral abduction contracture in children with unresolved neonatal brachial plexus palsy. J Bone Joint Surg Am 2015; 97(2): 112-18. El-Gammal T A, Saleh W R, El-Sayed A, Kotb M M, Imam H M, Fathi N A. Tendon transfer around the shoulder in obstetric brachial plexus paralysis: clinical and computed tomographic study. J Pediatr Orthop 2006; 26(5): 641-6.

Ezaki M, Malungpaishrope K, Harrison R J, Brown R H. Onabotulinum toxin A injection as an adjunct in the treatment of posterior shoulder subluxation in neonatal brachial plexus palsy. J Bone Joint Surg Am 2010; 92(12): 2171-7. Foad S L, Mehlman C T, Ying J. The epidemiology of neonatal brachial plexus palsy in the United States. J Bone Joint Surg Am 2008; 90(6): 1258-64.


118

Gharbaoui I S, Gogola G R, Aaron D H, Kozin S H. Perspectives on glenohumeral joint contractures and shoulder dysfunction in children with perinatal brachial plexus palsy. J Hand Ther 2015; 28(2): 176-84. Grossman J, Di Taranto P, Price A, Israel A. Multidisciplinary management of brachial plexus birth injuries: the Miami experience. Semin Plast Surg 2004; 18(4): 319-26. Haerle M, Gilbert A. Management of complete obstetric brachial plexus lesions. J Pediatr Orthop 2004; 24(2): 194-200. Hale H B, Bae D S, Waters P M. Current concepts in the management of brachial plexus birth palsy. J Hand Surg Am 2010; 35(2): 322-31. Hoeksma A F, Wolf H, Oei S L. Obstetrical brachial plexus injuries: incidence, natural course and shoulder contracture. Clin Rehabil 2000; 14(5): 523-6. Hoeksma A F, Ter Steeg A M, De Jong B A. Shoulder contracture and osseous deformity in obstetrical brachial plexus injuries. J Bone Joint Surg Am 2003; 85(2): 316-22. Kirjavainen O M, Remes V, Peltonen J, Nietosvaara Y. Long-term results of surgery for brachial plexus birth palsy. J Bone Joint Surg Am 2007; 89(1): 18-26. Medina L S, Yaylali I, Zurakowski D, Ruiz J, Altman N R, Grossman J A. Diagnostic performance of MRI and MR myelography in infants with a brachial plexus injury. Pediatr Radiol 2006; 36(12): 1295-9. Menashe S J, Tse R, Nixon J N, Ishak G E, Thapa M M, Mc Broom J A, Iyers R S. Brachial plexus birth palsy: multimodality imaging of spine and shoulder abnormalities in children. AJR Am J Roentgenol 2015; 204(2): W199-206. Moukoko D, Ezaki M, Wilkes D, Carter P. Posterior shoulder dislocations in infants with neonatal brachial plexus palsy. J Bone Joint Surg Am 2004; 86(4): 787-93.

Acta Orthopaedica 2019; 90 (2): 111–118

Narakas A O. Injuries to the brachial plexus. In: Bora F W Jr, editor. The pediatric upper extremity: diagnosis and management. Philadelphia: Saunders; 1986, p. 247-58. Pöyhiä T H, Koivikko M P, Peltonen J I, Kirjavainen M O, Lamminen A E, Nietosvaara A Y. Muscle changes in brachial plexus birth injury with elbow flexion contracture: an MRI study. Pediatr Radiol 2007; 37(2): 173-9. Pöyhiä T H, Lamminen A E, Nietosvaara Y. Brachial plexus birth injury: US screening for glenohumeral join instability. Radiology 2010; 254(1): 253-60. Pöyhiä T, Lamminen A, Peltonen J, Willamo P, Nietosvaara Y. Treatment of shoulder sequelae in brachial plexus birth injury. Acta Ortop 2011; 82(4): 482-8. Silbermann-Hoffman O, Teboul E. Post-traumatic brachial plexus MRI in practice. Diagn Interv Imaging 2013; 94(10): 925-43. Somashekar D, Yang L J, Ibrahim M, Parmar H A. High-resolution MRI evaluation of neonatal brachial plexus palsy: a promising alternative to traditional CT myelography. AJNR Am J Neuroradiol 2014; 35(6): 1 209-13. Tse R, Nixon J N, Iyer R S, Kuhlman-Wood K A, Ishak G E. The diagnostic value of CT myelography, MR myelography, and both in neonatal brachial plexus palsy. AJNR Am J Neuroradiol 2014; 35(7): 1425-32. Waters P M. Update on management of pediatric brachial plexus palsy. J Pediatr Orthop 2005; 25(1): 116-26. Yilmaz K, Caliskan M, Oge E, Aydinli N, Tunaci M, Ozmen M. Clinical assessment, MRI, and EMG in congenital brachial plexus palsy. Pediatr Neurol 1999; 21(4): 705-10.


Acta Orthopaedica 2019; 90 (2): 119–122

119

Poor patient-reported outcome after shoulder replacement in young patients with cuff-tear arthropathy: a matched-pair analysis from the Danish Shoulder Arthroplasty Registry Mette AMMITZBOELL 1, Amin BARAM 1, Stig BRORSON 2, Bo Sanderhoff OLSEN 1, and Jeppe V RASMUSSEN 1 1 Department

of Orthopaedic Surgery, Herlev and Gentofte Hospital, University of Copenhagen; 2 Department of Orthopaedic Surgery, Zealand University Hospital, University of Copenhagen, Denmark Correspondence: metteammitzboell@hotmail.com Submitted 2018-09-04. Accepted 2018-11-29.

Background and purpose — Reverse shoulder arthroplasty (RSA) has become the treatment of choice for cufftear arthropathy. There are, however, concerns about the longevity and the outcome of an eventual revision procedure. Thus, resurfacing hemiarthroplasty (RHA) with extended articular surface has been suggested for younger patients. We compared the patient-reported outcome of these arthroplasty designs for cuff-tear arthropathy. Patients and methods — We included patients operated on because of cuff-tear arthropathy and reported to the Danish Shoulder Arthroplasty Registry (DSR) from January 1, 2006 to December 31, 2013. 117 RHA cases were matched by age and sex with 233 RSA controls. 34 of the RHAs were conventional and 67 were RHAs with extended articular surface. The Western Ontario Osteoarthritis of the Shoulder (WOOS) Index at 1 year was used as primary outcome. The score was converted to a percentage of a maximum score. Revision, defined as removal or exchange of any component or the addition of a glenoid component, was used as secondary outcome. Results — Median WOOS was 49 (30–81) for RHA and 77 (50–92) for RSA (p < 0.001). For patients younger than 65 years, median WOOS was 58 (44–80) after RHA, similar to the 54 after RSA (37–85). For patients older than 65 years, median WOOS was 48 (28–82) after RHA and 79 (55–92) after RSA (p < 0.001). Interpretation — In all patients RSA had a clinically and statistically better patient-reported outcome than RHA. However, in patients younger than 65 years the functional outcome was similar and poor for either arthroplasty type. The optimal treatment of CTA in young patients remains a challenge.

According to Neer et al. (1983), cuff-tear arthropathy (CTA) is described as a large rotator cuff tear followed by inactivity, disuse of the shoulder, leaking of synovial fluid, and instability of the humeral head. Radiographic characteristics are superior migration, collapse of the proximal aspect of the humeral articular surface, and erosion or acetabularization of the acromion (Neer et al. 1983, Feeley et al. 2009). The clinical symptoms are pain, arthritis, muscle atrophy, decreased range of motion, and in a subset of late-stage patients a pseudoparalytic shoulder (Neer et al. 1983, Ramirez et al. 2012, Smith et al. 2012). In these cases, the result of reverse shoulder arthroplasty (RSA) is better than that of resurfacing hemiarthroplasty (RHA) (Young et al. 2013). There are, however, concerns not only regarding the longevity of RSA but also the outcome of an eventual revision procedure. Thus, RHA with posterior and superior extended articular surface has been suggested as an option in the treatment of younger patients with long life expectancy. We compared patient-reported outcome and revision rates of RHA and RSA in the treatment of CTA in average patients and in patients younger than 65 years.

Patients and methods Data were obtained from the Danish Shoulder Arthroplasty Registry (DSR). All Danish public hospitals and private clinics report to this registry. More than 90% of the procedures have been captured each year since 2007 when compared with the national patient registry (DSR 2016). Surgeons report information concerning the patient (name, date of birth, and sex) and the procedure (hospital, date of surgery, diagnosis, previous surgery, and arthroplasty type) electronically at the time of

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2018.1563855


120

Acta Orthopaedica 2019; 90 (2): 119–122

Frequency reverse arthroplasty the operation. 10–14 months after the procedure, Frequency resurfacing arthroplasty the patient-reported outcome is collected using 10 40 the Western Ontario Osteoarthritis of the Shoulder (WOOS) questionnaire index. WOOS consists 8 30 of 19 questions focused on shoulder-related quality of life. Each question is answered on a scale 6 from 0 to 100; thus the total score ranges from 0 to 20 1,900. In this study, the raw WOOS score was con4 verted to a percentage of a maximum score where a score of 100 was regarded as a healthy shoulder 10 with no functional impairment (Lo et al. 2001). A 2 validated Danish version of WOOS is used in the 0 DSR (Rasmussen et al. 2013). To our knowledge, 0 0 20 40 60 80 100 0 20 40 60 80 100 a minimal clinically important difference WOOS WOOS WOOS has not been established and validated. We used WOOS distribution after resurfacing arthroplasties (left panel) and reverse arthroplasa value of 10% of a maximum score, which is an ties (right panel). extrapolation from other shoulder-specific questionnaires (e.g., the Oxford Shoulder Score [OSS)]) (Rasmus- 81) for RHA and 77 (50–92) for RSA, (p < 0.001). There were sen et al. 2013). We defined a clinical failure as a WOOS below 44 (52%) RHAs and 42 (25%) RSAs with a WOOS below 50. Revision is defined as removal or exchange of any compo- 50 (Figure). 7 (6%) RHAs and 13 (6%) RSAs were revised. nent or the addition of a glenoid component. The reasons for revision after RHA were glenoid attrition (n All primary shoulder arthroplasties for CTA reported to the = 2), infection (n = 1), luxation (n = 1), other (n = 1), or no registry from January 1, 2006 to December 31, 2013 were reason reported (n = 2). The reasons for revision after RSA identified (n = 891). We matched each RHA (n = 119) with 2 were infection (n = 5), luxation (n = 3), loosening (n = 3), RSA controls based on age, sex, and response of the WOOS other (n = 1), or no reason reported (n = 1). at 1 year. For patients younger than 60 years we accepted Of the 117 RHAs, 34 were conventional RHAs and 67 had matches in intervals of 5 years. 2 patients did not have match- posterior and superior extended articular surface. Informaing controls available (they were both men and 43 years old). tion regarding the subtype was missing in 16 cases. In the 1 had a WOOS of 73 and 1 was recorded as a non-responder group with posterior and superior extended articular surface regarding WOOS. A 49-year-old man with a WOOS of 47 was 35 (52%) patients were women, mean age was 73 years, and matched with only 1 control. Thus, 117 RHAs and 233 RSAs the response rate of WOOS was 73%. In the group with conwere included in the study. 73% had a complete WOOS, 24% ventional RHAs 22 (65%) patients were women, mean age did not respond to WOOS, and 3% were revised or died within was 73 years, and the response rate of WOOS was 71%. The the first year after surgery. median WOOS was 48 (33–80) for RHA with posterior and superior extended articular surface and 57 (38–92) for conStatistics ventional RHA (p = 0.4). 5 (8%) RHAs with posterior and Data on WOOS were not normally distributed. Thus, the superior extended articular surface and 1 (3%) conventional results are presented as median and interquartile range (IQR) RHA were revised. and the Mann–Whitney U-test was used when groups were For patients younger than 65 years the median WOOS was compared. P-value < 0.05 was considered significant. The similar for RHA (n = 19) and RSA (n = 32) at 58 (44–80) analysis was performed using IBM SPSS statistics version and 54 (37–85). For patients older than 65 years the median 24.0 (IBM Corp, Armonk, NY, USA) WOOS for RHA (n = 66) and RSA (n = 137) was 48 (28–82) and 79 (55–92) (p < 0.001). Ethics, funding, and potential conflicts of interest Ethics approval was from the Danish Patient Safety Department (Study number 3-3013-1862/1/ Reference MOAD from March 28, 2017). No funding was recieved. No conflicts of Discussion interest. We found the patient-reported outcome after RSA to be better than after RHA. For patients younger than 65 years the results were disappointing for both arthroplasty types. The revision rates were the same. 52% of RHAs and 25% of RSAs had a Results WOOS below 50 but only 6% were revised for both arthroThe mean age of all patients was 73 years (50–92) and 60% plasty designs. A reason for this difference might be restraint were women. The median WOOS for all patients was 49 (30– in revision arthroplasties because of poor outcome.


Acta Orthopaedica 2019; 90 (2): 119–122

The results of RHA or RSA for CTA have frequently been reported in small case series, but to our knowledge there are only 2 studies comparing hemiarthroplasty with RSA for CTA. From the New Zealand Joint Registry Young et al. (2013) compared 102 hemiarthroplasties with 102 RSAs. 77 of 102 hemiarthroplasties were RHAs with posterior and superior extended articular surface. The patients were matched for age, sex, and ASA scores, and the functional outcome was evaluated using the OSS at 6 months and 5 years postoperatively (Young et al. 2013). They found no differences in mortality, revision rates, or OSS for patients younger than 65 years when comparing 24 hemiarthroplasties and 20 RSA. As in our study there was a statistically significant difference in functional outcome overall and in patients older than 65 years. The authors did not report separate results of the 77 RHAs with posterior and superior extended articular surface (Young et al. 2013). A retrospective study by Leung et al. (2012) from a 10-year period compared 20 hemiarthroplasties and 36 RSAs for CTA at minimum 2 years’ follow-up. They found a better Shoulder Pain and Disability Index (SPADI) for RSA. They found high complication rates in 14 patients from both arthroplasty types, with infections and pain in the hemiarthroplasty group and infection, loosening, and fracture of acromion or humerus in the RSA group (Leung et al. 2012). A radiographic study by Leung et al. (2017) with 97 arthroplasties reported 26 radiographic complications, e.g., acromion remodeling, fracture, subluxation, or loosening after RHA with posterior and superior extended articular surface for CTA. Most of these complications were seen within the first 3 months postoperatively. Occurrence of postoperative radiographic complications was associated with revision. 8 arthroplasties were revised (Leung et al. 2017). No functional outcome or results, particularly for younger patients, were reported. Alizadehkhaiyat et al. (2013) compared RHA with posterior and superior extended articular surface with conventional RHA for different diagnoses including 9 patients with CTA of whom 8 received RHA with posterior and superior extended articular surface. 5 of these 8 patients were revised. The CTA group had a mean OSS of 26 postoperatively at 4 years’ follow up. The authors concluded that cuff deficiency was a major reason for failure and revision (Alizadehkhaiyat et al. 2013). Our results of conventional RHA and RHA with posterior and superior extended articular surface show poor patient- reported outcome for both types. In a single-center retrospective study Ernstbrunner et al. (2017) looked at RSA for irreparable rotator cuff tears in 23 cases younger than 60 years. The functional and radiographic results at 10 years were good, but 6 patients were reoperated, mainly because of instability. The high revision rate suggests that RSA might not be the best choice for all younger patients. In the absence of alternative treatments a selected group of younger patients with irreparable rotator cuff tears can obtain good results with RSA (Ernstbrunner et al. 2017). To our knowledge there are no studies reporting the results of

121

revision arthroplasty after failed RHA in young patients with CTA, but data from the DSR showed poor outcome of the revision arthroplasty when the RHA was used for osteoarthritis (Rasmussen et al. 2016). Consequently the use of shoulder arthroplasty for younger patients should be considered carefully. The strength of our study is the systematic nationwide collection of data through the DSR with high external validity. There is no data selection, and our results reflect the actual outcome in the Danish population. The study has the limitations of observational studies including the possibility of differences in baseline characteristics such as in comorbidities and preoperative functional status. Another limitation of our study and other studies reporting the results of shoulder arthroplasty for CTA is the lack of a common definition and classification of CTA. There is a possible bias in distribution of surgeries among the participating centers and surgeons. We do not know whether indications for surgery or revisions are the same as in the studies used for comparison. Finally, we have no preoperative WOOS. To conclude, RSA had a clinically and statistically significant better patient-reported outcome compared with RHA. However, for patients younger than 65 years the functional outcome was disappointing for both arthroplasty types. The optimal treatment of CTA in young patients remains a challenge.

MA: collecting data from the DSR, writing the manuscript, and incorporation of input from the other authors. AB, SB, BSO: reviewing and correction of the manuscript. JVR: statistics, reviewing, and correction of the manuscript. The authors would like to thank the Danish surgeons for reporting information to the DSR. Special thanks are offered to Steen Lund Jensen (Farsø Sygehus), Lone Bjørklund (Aleris-Hamlet Hospitaler), and Lars Henrik Frich (Odense Universitetshospital) for collecting information on RHA subtypes. Acta thanks Björn Salomonsson for help with peer review of this study.

Alizadehkhaiyat O, Kyriakos A, Singer M S, Frostick S P. Outcome of Copeland shoulder resurfacing arthroplasty with a 4-year mean follow-up. J Shoulder Elbow Surg 2013; 22 (10): 1352-8. DSR. Annual Report. In: Danish Shoulder Arthroplasty Registry Annual Report. https://www.sundhed.dk/content/cms/3/4703_dsr_%C3%A5rsrap port2016.pdf; 2016. Ernstbrunner L, Suter A, Catanzaro S, Rahm S, Gerber C. Reverse total shoulder arthroplasty for massive, irreparable rotator cuff tears before the age of 60 years: long-term results. J Bone Joint Surg Am 2017; 99 (20): 1721-9. Feeley B T, Gallo R A, Craig E V. Cuff tear arthropathy: current trends in diagnosis and surgical management. J Shoulder Elbow Surg 2009; 18 (3): 484-94. Leung B, Horodyski M, Struk A M, Wright T W. Functional outcome of hemiarthroplasty compared with reverse total shoulder arthroplasty in the treatment of rotator cuff tear arthropathy. J Shoulder Elbow Surg 2012; 21 (3): 319-23.


122

Leung A S, Hippe D S, Ha A S. Cuff tear arthropathy shoulder hemiarthroplasty: a radiographic outcome study. Skeletal Radiol 2017; 46 (7): 909-18. Lo I K, Griffin S, Kirkley A. The development of a disease-specific quality of life measurement tool for osteoarthritis of the shoulder: the Western Ontario Osteoarthritis of the Shoulder (WOOS) index. Osteoarthritis Cartilage 2001; 9 (8): 771-8. Neer C S 2nd, Craig E V, Fukuda H. Cuff-tear arthropathy. J Bone Joint Surg Am 1983; 65(9): 1232-44. Ramirez M A, Ramirez J, Murthi A M. Reverse total shoulder arthroplasty for irreparable rotator cuff tears and cuff tear arthropathy. Clin Sports Med 2012; 3 (4): 749-59.

Acta Orthopaedica 2019; 90 (2): 119â&#x20AC;&#x201C;122

Rasmussen J V, Jakobsen J, Olsen B S, Brorson S. Translation and validation of the Western Ontario Osteoarthritis of the Shoulder (WOOS) index: the Danish version. Patient related outcome measures 2013; 4: 49-54. Rasmussen J V, Olsen B S, Al-Hamdani A, Brorson S. Outcome of revision shoulder arthroplasty after resurfacing hemiarthroplasty in patients with glenohumeral osteoarthritis. J Bone Joint Surg Am 2016; 98 (19): 1631-7. Smith C D, Guyver P, Bunker T D. Indications for reverse shoulder replacement: a systematic review. J Bone Joint Surg Br 2012; 94 (5): 577-83. Young S W, Zhu M, Walker C G, Poon P C. Comparison of functional outcomes of reverse shoulder arthroplasty with those of hemiarthroplasty in the treatment of cuff-tear arthropathy: a matched-pair analysis. J Bone Joint Surg Am 2013; 95 (10): 910-5.


Acta Orthopaedica 2019; 90 (2): 123–128

123

Early palmar plate fixation of distal radius fractures may benefit patients aged 50 years or older: a randomized trial comparing 2 different treatment protocols Kai SIRNIÖ 1, Juhana LEPPILAHTI 1, Pasi OHTONEN 2, and Tapio FLINKKILÄ 1 1 Department

of Surgery, Division of Orthopaedic and Trauma Surgery, Oulu University Hospital, Oulu; 2 Department of Anesthesiology, Surgery, and Intensive Care, Oulu University Hospital, Oulu, Finland Correspondence: Kai.Sirnio@ppshp.fi Submitted 2018-04-27. Accepted 2018-12-06.

Background and purpose — There is no consensus regarding optimal treatment of displaced distal radius fractures (DRFs). We compared the results of 2 treatment protocols: early palmar plating vs. primary nonoperative treatment of displaced DRFs. Patients and methods — We performed a prospective randomized controlled study including 80 patients aged ≥ 50 years with dorsally displaced DRFs, excluding AO type C3 fractures. Patients were randomized to undergo either immediate surgery with palmar plating (n = 38), or initial nonoperative treatment (n = 42) after successful closed reduction in both groups. Delayed surgery was performed in nonoperatively treated patients showing early loss of alignment (n = 16). The primary outcome measure was Disabilities of the Arm, Shoulder, and Hand (DASH) score. Results — Mean DASH scores at 24 months in the early surgery group were 7.9 vs. 14 in the initial nonoperative group (difference between means 6, 95% CI 0.1–11, p = 0.05). Delayed operation was performed on 16/42 of patients due to secondary displacement in the initial nonoperative group. In “as treated” analysis, DASH scores were 7 in the early surgery group, 13 in the nonoperative group, and 17 after delayed surgery (p = 0.02). The difference in DASH scores between early and delayed surgery was 9 points (CI 0.3–19, p = 0.02) Interpretation — Treatment of DRFs with early palmar plating resulted in better 2-year functional outcomes for ≥ 50-year-old patients compared with a primary nonoperative treatment protocol. Delayed surgery in case of secondary displacement was not beneficial in terms of function.

As palmar plating of distal radius fractures (DRFs) became more popular, the incidence of operative treatment increased, particularly among older women (Chung et al. 2009, Mattila et al. 2011, Mellstrand-Navarro et al. 2014). Percutaneous Kirschner wire (K-wire) fixation and ORIF with palmar plating are the most used fixation methods of displaced DRFs. The previous DRAFFT study showed no difference in functional results at 12 months between percutaneous K-wire fixation and volar plating in adults (Costa et al. 2014). Palmar fixed-angle plates enable near-anatomic reduction and stable fixation, creating optimal conditions for healing even in osteoporotic bone (Orbay and Fernandez 2004, Figl et al. 2010). However, while palmar plate fixation of DRF achieves stability and good radiographical results (Orbay and Fernandez 2004, Rozental and Blazar 2006), the relationship between radiographic reduction and outcome is not yet confirmed, particularly in elderly patients (Grewal and MacDermid 2007, Diaz-Garcia et al. 2011). Considering that nonoperative treatment has been in general the preferred treatment method, it is of interest to compare results of palmar plating and nonoperative treatment of DRFs. Only a few randomized controlled trials have compared mid-term results of palmar plating and nonoperative treatment of DRFs in populations predominantly including elderly patients, with no significant benefit of plating observed (Arora et al. 2011, Bartl et al. 2014). In contrast, a recent study of Martinez-Mendez et al. (2018) showed significantly better functional results in patients older than 60 years treated with palmar plating compared with cast treatment of DRF. After closed reduction of the displaced DRF in our department, fracture alignment is routinely evaluated at 1-week and 2-week follow-up visits. When malalignment reaches a specific threshold (> 10° of dorsal angulation, < 15° of radial incli-

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2018.1561614


124

Acta Orthopaedica 2019; 90 (2): 123–128

Table 1. Inclusion and exclusion criteria Inclusion criteria Displaced DRF (AO/OTA 23 type A2/A3 and C1/C2) Duration of < 1 week from primary injury Acceptable closed reduction achieved: – dorsal angulation ≤ 10° – radial inclination ≥ 15° – ulnar variance < –3 mm – articular step-off ≤ 2 mm Exclusion criteria Patients < 50 years old Acceptable closed reduction not achieved (see inclusion criteria) Bilateral/open fractures Fractures with neurovascular compromise Previous ipsilateral DRF Inflammatory joint disease Radiocarpal joint degeneration Limited cooperation or major comorbidity not allowing to operate Major concomitant fracture necessitating any operation

Enrollment

Assessed for eligibility (n = 110) Excluded (n = 30): – declined to participate, 17 – comorbidity and/or age related, 8 – senility, 1 – bilateral wrist fracture, 1 – rheumatoid arthritis, 1 – local dermal infection, 1 – previous ipsilateral fracture, 1

Allocation

Randomized (n = 80)

Allocated to early surgery (n = 38)

nation, or > 2° millimeters ulnar positive variance), patients are offered surgery. Loss of alignment within 2 weeks after closed reduction of DRF is common (Mackenney et al. 2006), and it is challenging to decide between operative and nonoperative treatment at this early stage. Early operative treatment may lead to more predictable radiographic and patient-rated outcomes, with fewer outpatient visits. In the present study, our main goal was to compare functional and radiographic results in patients of ≥ 50 years of age with dorsally displaced DRFs who underwent treatment with 2 different protocols: early palmar plating vs. primary nonoperative (control group) treatment. Our primary hypothesis was that the functional results at 24 months would be superior in the early surgery group compared with the control group. Our secondary aim was to compare the patient satisfaction, radiographic results, complications, and secondary operation rates between study groups.

Patients and methods We conducted a prospective randomized controlled singlecenter trial at Oulu University Hospital (Department of Surgery, Division of Orthopaedic and Trauma Surgery, Oulu, Finland). 80 patients of ≥ 50 years of age were recruited from the catchment area of our hospital district between November 2008 and January 2014. Patients Table 1 presents the criteria for inclusion and exclusion. All fractures were in both groups initially treated via closed reduction under hematoma block and acceptable radiographic reduction was achieved in all cases. The wrist was immobi-

Allocated to control (n = 42): – non-operative treatment, 26 – delayed operation, 16

Follow-up Lost to follow-up (n = 5): – lost interest, 2 – moved far, 1 – died unrelated to treatment, 1 – lost contact, 1

Lost to follow-up (n = 7): – lost interest, 3 – hip fractures, 2 – lost contact, 1

Analysis Intention-to-treat analysis at 2 years (n = 33/38)

Intention-to-treat analysis at 2 years (n = 35/42)

Figure 1. Flow diagram.

lized with a short-arm cast at the emergency unit. 110 patients were assessed for eligibility, of whom 30 were excluded (Figure 1). The remaining 80 patients were randomized into 1 of 2 study protocol groups: early surgery with palmar plating (n = 38) or primary nonoperative treatment (n = 42). Intervention Within 1 week after injury, patients underwent palmar plating of DRF applying a standard surgical technique. The distal part of the radius was exposed using a modified Henry’s approach. Fracture reduction was achieved by open manipulation, and the fracture was stabilized using a palmar fixed-angle plate (Aculoc or Aculoc 2; Acumed, Hillsboro, OR, USA) with proximal 3.5-mm locking screws and distal 2.3-mm locking cortical pegs. A dorsal plaster cast for pain relief was applied for 10 days, after which the patients received formal instructions for active mobilization of the wrist. All procedures were performed by experienced surgeons familiar with the operative technique being used. Control Patients allocated to the control group underwent closed reduction in the emergency department, and were then immobilized in a below-elbow arm cast. They were scheduled to attend follow-up visits at 1 and 2 weeks after treatment. If the reduction met the acceptable criteria, immobilization was continued for 6 weeks, after which the cast was removed and


Acta Orthopaedica 2019; 90 (2): 123–128

active mobilization of wrist was started. Operative treatment with a palmar plate was offered when there was greater than 10° dorsal angulation of the articular surface on the lateral radiograph, less than 15° of radial inclination or greater than 2 mm ulnar positive variance on the posteroanterior radiograph within 2 weeks after injury. If patient declined surgery, nonoperative treatment was continued until fracture union. In the case of delayed surgery in the control group, the postoperative follow-up and rehabilitation program was identical to that of the intervention group. Outcome measures The primary outcome measure was the Disabilities of the Arm, Shoulder, and Hand (DASH) score (Hudak et al. 1996, Aro et al. 2008) at 2 years. Secondary outcome measures included range of motion (ROM), grip strength, subjective assessment of wrist function, radiological results, complications, and rate of re-operations. Follow-up visits for both groups were scheduled at 3, 6, 12, and 24 months. Function At each follow-up visit, the patients completed DASH questionnaires. All objective functional measurements were performed by a physiotherapist not involved in patient care. Dorsal and volar flexion and ulnar and radial deviation were measured using a manual goniometer. Pronation and supination were measured using a Myrin compass goniometer (Follo Futura AS, Ås, Norway). Grip strength (kg) was measured using a calibrated Hydraulic Hand Dynamometer Model SH5001 (Saehan Corporation, Changwon, South Korea). The best result of 3 attempts was used for analysis. Patients were also asked to subjectively grade the function of their injured wrist as poor, fair, good, or excellent using the uninjured side as reference. Radiography In both treatment groups standard posteroanterior (PA) and lateral wrist radiographs were taken immediately after primary injury and after closed reduction of the fracture. In the control group, radiographs were taken at 1, 2, 6, and 12 weeks and at 6, 12, and 24 months after injury. In the operative treatment group, radiographs were taken at postoperative day 1 and at 6 weeks, 12 months, and 24 months after the injury. The uninjured side was radiographed at 6 weeks after injury. KS made digital measurements from the radiographs. Fracture types were classified according to AO classification, using the main types A and C to maintain reproducibility (Flinkkilä et al. 1998). Dorsal angulation was measured from lateral radiographs and radial inclination from posteroanterior radiographs. Ulnar variance (in mm) was measured from PA radiographs with reference to the uninjured side. Adverse effects An adverse event was defined as any unfavorable or unintended sign that could affect the results or that necessitated

125

secondary surgery. Malunion was defined as dorsal angulation of > 15° and radial inclination of < 15°. Radiocarpal osteoarthritis was assessed and graded according to Knirk and Jupiter (1986). Re-operation was defined as secondary surgery related to the primary injury or operation. Sample size Sample size was calculated using the DASH score, assuming a clinically relevant 15-point difference between treatments and standard deviation (SD) of 22, based on previous studies (Gummesson et al. 2003, Anzarut et al. 2004, Rozental and Blazar 2006). The calculation indicated a need for 35 patients per group (SD = 22, α = 0.05, power = 0.80), with an estimated 10% dropout rate. We decided to include 80 patients in the study. Randomization Patients were randomly allocated into study groups based on a computer-generated list. Randomization was performed in blocks, with block sizes randomly varying between 4, 6, 8, and 12. Separate lists were created for age groups of < 65 and ≥ 65 years, and for type A and C fractures. Randomization lists were sealed into numbered opaque envelopes. After confirmation of patient eligibility and obtaining the patient’s written informed consent, the treating surgeon opened a numbered envelope revealing the method of treatment. Statistics The patients were analyzed primarily on an intention-to-treat basis and secondarily “as treated.” Missing data in our primary outcome variable DASH at 24 months’ follow-up of randomized patients were imputed using a multiple imputation (MI) method. The missing data pattern was non-monotone and therefore we used MI with fully conditional specification using variables DASH (3, 6, and 12 months) and age, fracture type, and randomization group to model data for DASH at 24 months. 50 different data sets were created and the pooled result is presented. Summary measurements are presented as mean (SD) or as median with 25th and 75th percentiles. Comparisons between study groups were performed using Student’s t-test or the Mann–Whitney U-test for continuous variables, and by the chi-square or Fisher’s exact test for categorical variables. The Kruskal–Wallis test was used to compare DASH results at the 2-year follow-up between the operative, nonoperative to the end, and delayed operation groups. If p < 0.05 according to Kruskal–Wallis test then Mann–Whitney U-test was used for comparisons between the 2 groups. Repeatedly measured data were analyzed using a linear mixed model (LMM) assuming patients as random effects. The covariance pattern was chosen according to Akaike’s information criteria. The p-values reported for LMM are ptime for the change over time, pgroup for average treatment difference, and ptime x group for interaction between time and treatment.


126

Acta Orthopaedica 2019; 90 (2): 123–128

Table 2. Demographics. Values are number of patients unless otherwise indicated Factor

Early surgery group (n = 38)

Control group (n = 42)

Age in years, mean (range) 62 (50–79) 64 (50–82) Age ratio, < 65 / ≥ 65 years 23 / 15 23 / 19 Sex, female / male 37 / 1 39 / 3 Dominant hand involved 15 17 AO classification A 23 25 C 15 17

The results of between-group comparisons are presented as the difference between means and 95% confidence interval (CI). Comparisons of functional outcome scores between age groups were preplanned subgroup analyses, while comparisons between early surgery, delayed surgery, and nonoperative treatment (“as treated”) were decided post hoc. Ethics, registration, funding and potential conflicts of interest All patients gave written informed consent, and the study was approved by the Oulu University Hospital Ethics Committee (number EETTMK: 143/2007) and registered at Clinicaltrials. gov (NCT02990052). The authors received no financial support for the research, authorship, and/or publication of this article and the authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Results The main study groups had comparable baseline demographic characteristics (Table 2). 16 patients in the control group underwent delayed operation at 1 to 3 weeks after the initial trauma due to early loss of reduction (Figure 1). 2 patients in the control group declined delayed surgery and were treated nonoperatively. Function Mean DASH scores at the 2-year follow-up differed statistically significantly between study groups, favoring early operation: 7.2 vs. 14.4, p = 0.005 (difference between means, −7; CI −13 to −1.5) (Figure 2). According to MI analysis the mean difference reduced being –6 points (8 vs. 14; CI –11 to –0.12, p = 0.05). At the 2-year follow-up, statistically significant differences favoring early surgery were also detected in flexion (71° vs. 64°; p = 0.002; difference between means, 7°; CI 3–12) and ulnar deviation (28° vs. 25°; p = 0.009; difference between means, 3°; CI 0.9–6). In terms of ROM, only the recovery rate of extension was faster in the early surgery group. Grip strength at 2-year follow-up was comparable between study groups (Table 3, Supplementary data). Wrist

DASH score 50 Control group Early surgery group 40

30

20

10

0 3

6

12

24

Months of follow-up

Figure 2. DASH scores for main study groups at follow-up points. Scores are presented as median with 25th and 75th percentiles. The p-values were determined according to a linear mixed model (LMM): ptime < 0.001, pgroup = 0.04, and ptime x group = 0.6.

function was self-assessed to be excellent or good by 30/33 patients in the early surgery group compared with 23/35 patients in the control group (p = 0.01). In analysis limited to patients of ≥ 65 years of age, DASH scores at the 2-year follow-up did not differ statistically significantly between the early surgery and control groups: 11 vs. 17 (difference between means, −6; CI −18 to 4; p = 0.2). For the patients below 65 years of age the mean DASH for early surgery and control groups was 6 vs. 11 (difference between means, −5; CI −12 to –1; p = 0.01). In “as treated” analysis, the mean DASH scores at the 2-year follow-up were 7 (SD 10) in the early surgery group, 13 (SD 12) in the nonoperative to the end group, and 17 (SD 16) in the delayed operation group (p = 0.02). A statistical, and probably also clinically significant difference between early and delayed surgery groups was found (difference between means, –9; CI –19 to –0.3; p = 0.02). Radiography All radiographic parameters were statistically significantly better in the early surgery group compared with the control group (Table 4, Supplementary data). Of the 42 patients in the control group, 18 exhibited secondary loss of reduction at the 1- or 2-week follow-up visits, and 16 of these patients underwent delayed operation at between 1 and 3 weeks after primary injury. In the conversion group mean dorsal angulation was −1.1° (SD 5), radial inclination 24° (SD 4), and ulnar variance −0.6 mm (SD 1.2). Loss of acceptable alignment after 2 weeks’ follow-up visits was observed in 14/42 patients in the control group at final follow-up. Complications and secondary operations 1 case of carpal tunnel syndrome, 1 patient with flexor tenosynovitis, and 1 case of post-traumatic radiocarpal arthritis


Acta Orthopaedica 2019; 90 (2): 123–128

was noted in the early operation group. Only the patient with carpal tunnel syndrome required secondary operation. In the control group 4 patients with a carpal tunnel syndrome and 1 case of grade I post-traumatic radiocarpal osteoarthritis were noted. 3 patients in the control group had secondary operation, 2 with carpal tunnel syndrome and 1 with malunion.

Discussion Our results showed that DASH score at 2 years favored early surgery over initial nonoperative treatment of displaced DRFs. Wrist flexion and ulnar deviation at 24 months favored early surgery, but other between-group differences in ROM and grip strength after 6 months were minor. Surgery resulted in earlier recovery of wrist extension than achieved with primary nonoperative treatment. Compared with the control group, the early surgery group showed better patient satisfaction with superior radiographic results and fewer complications and secondary operations. Delayed operation in cases of secondary displacement after initial nonoperative treatment did not provide comparable results to early surgery in terms of DASH score. Arora et al. (2011) performed an RCT comparing palmar plating and nonoperative treatment for unstable DRFs in patients ≥ 65 years old, and found no statistically or clinically significant between-group differences in DASH score or ROM at 12 months. Their study differed from ours, as we included only fractures that were acceptably reduced. Nevertheless, our findings regarding functional results in an elderly cohort support those of Arora et al. (2011). Additionally, Bartl et al. (2014) performed a multi-center RCT comparing open reduction and internal fixation with cast treatment of primary unstable intra-articular DRFs in patients of ≥ 65 years of age with 12 months of follow-up, and reported no significant betweengroup differences in DASH scores or ROM. In the control group, almost half of the fractures lost alignment within 2 weeks after acceptable primary reduction, in line with previous findings (Mackenney et al. 2006, Bartl et al. 2014, Martinez-Mendez et al. 2018). Yamashita et al. (2015) performed a retrospective study of extra-articular DRFs, and found no difference in functional results between early and delayed fixation at 1 year of follow-up. Our study included intra-articular fractures and had a longer follow-up time, possibly explaining the difference. Weil et al. (2014) performed a retrospective study, including mostly type C fractures, and showed statistically worse Quick-DASH scores at one year in the delayed surgery group (operated > 21 days after injury) compared with the historical cohort of primarily operated patients. In our study, delayed operation was performed in more than one-third of patients in the control group because of secondary displacement, with inferior results compared with early surgery. The data suggest that the common treatment protocol of initial closed reduction and secondary surgery of partially healed fracture after secondary displacement

127

is not optimal. Hence the treatment method should probably be decided at an early stage after injury. Also, late instability after 2 weeks in our study population was common, comprising one-third of nonoperatively treated patients. This supports previous findings (Mackenney et al. 2006, Wadsten et al. 2018). It could be useful to apply treatment algorithms based on predictors of fracture instability (Mackenney et al. 2006, Walenkamp et al. 2016), but the value in clinical practice is not yet proven. Although the functional scores favored the early surgery group, the higher patient satisfaction in the early operation group was unexpected and may have been due to the higher rate of malalignment and secondary operations in the control group. Unstable fractures that lose alignment had inferior results in general compared with stable ones in control group, which could explain this difference. Although DASH and patient-rated wrist evaluation (PRWE) are valid outcome measures, they probably cannot capture all impairments, as shown in this and previous studies (Plant et al. 2017). Our study has several strengths. The RCT design of this study comparing 2 commonly used treatment protocols generated high-quality evidence regarding treatment of this common fracture. The investigated fracture types are the most challenging in terms of decision-making. Both groups had similar baseline demographics, and the catchment of patients at the 1- and 2-year follow-up points was sufficient. Moreover, one experienced surgeon treated most of the cases. This study also has several weaknesses. Due to limited resources we could screen only a proportion of all DRF patients treated at the emergency clinic during the study period. There was some patient dropout during the study and in accordance with our primary study protocol we also operated on a few stable fractures in the early surgery group, which may have biased the results. Type C3 fractures were not included, which we considered to be unsuitable for nonoperative treatment. Therefore results cannot be generalized to all fracture types. Considering our primary intention and sample size analysis to compare 2 main study groups, caution should be used when analyzing the results of subgroups of the study. In summary, early surgery after primary closed reduction of DRF could be recommended to physically active patients, if secondary displacement is imminent. High displacement rate after primary nonoperative treatment was observed and delayed surgery in these cases was not beneficial in terms of function when comparing with early surgery. The decision between surgery and nonoperative treatment should be made at a very early stage, and delayed operation avoided in cases of secondary displacement of DRFs in elderly people. Supplementary data Tables 3 and 4 are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/17453674. 2018.1561614


128

KS designed the study, reviewed the literature, was responsible for patient recruitment and randomization, analyzed the data, and prepared the manuscript. JL and TF designed the study and prepared the manuscript. PO contributed to study design and performed the statistical analyses. Acta thanks Hebe Désirée Kvernmo and Mats Wadsten for help with peer review of this study.

Anzarut A, Johnson J A, Rowe B H, Lambert R G W, Blitz S, Majumdar S R. Radiologic and patient-reported functional outcomes in an elderly cohort with conservatively treated distal radius fractures. J Hand Surg Am 2004; 29: 1121-7. Aro H, Hacklin E, Madanat R, Strandberg N. Finnish DASH. (http://www. dash.iwh.on.ca/translate.htm). Orthopaedic Research Unit, University of Turku, Turku, Finland; 2008. Arora R, Lutz M, Deml C, Krappinger D, Haug L, Gabl M. A prospective randomized trial comparing nonoperative treatment with volar locking plate fixation for displaced and unstable distal radial fractures in patients sixty-five years of age and older. J Bone Joint Surg Am 2011; 93: 2146-53. Bartl C, Stengel D, Bruckner T, Gebhard F. The treatment of displaced intraarticular distal radius fractures in elderly patients. Dtsch Arztebl Int 2014; 111: 779-87. Chung K C, Shauver M J, Birkmeyer J D. Trends in the United States in the treatment of distal radial fractures in the elderly. J Bone Joint Surg Am 2009; 91: 1868-73. Costa M L, Achten J, Parsons N R, Rangan A, Griffin D, Tubeuf S, Lamb S E. DRAFFT Study Group. Percutaneous fixation with Kirschner wires versus volar locking plate fixation in adults with dorsally displaced fracture of distal radius: randomised controlled trial. BMC Musculoskelet Disord 2014; 349: g4807. Diaz-Garcia R J, Oda T, Shauver M J, Chung K C. A systematic review of outcomes and complications of treating unstable distal radius fractures in the elderly. J Hand Surg Am 2011; 36: 824-35. Figl M, Weninger P, Jurkowitsch J, Hofbauer M, Schauer J, Leixnering M. Unstable distal radius fractures in the elderly patient: volar fixed-angle plate osteosynthesis prevents secondary loss of reduction. J Trauma 2010; 68: 992-8. Flinkkilä T, Nikkola-Sihto A, Kaarela O, Pääkkö E, Raatikainen T. Poor interobserver reliability of AO classification of fractures of the distal radius: additional computed tomography is of minor value. J Bone Joint Surg Br 1998; 80: 670-2.

Acta Orthopaedica 2019; 90 (2): 123–128

Grewal R, MacDermid J C. The risk of adverse outcomes in extra-articular distal radius fractures is increased with malalignment in patients of all ages but mitigated in older patients. J Hand Surg Am 2007; 32: 962-70. Gummesson C, Atroshi I, Ekdahl C. The disabilities of the arm, shoulder and hand (DASH) outcome questionnaire: longitudinal construct validity and measuring self-rated health change after surgery. BMC Musculoskelet Disord 2003; 4: 11. Hudak P L, Amadio P C, Bombardier C. Development of an upper extremity outcome measure: the DASH (disabilities of the arm, shoulder and hand). Am J Ind Med 1996; 29: 602-8. Knirk J L, Jupiter J B. Intra-articular fractures of the distal end of the radius in young adults. J Bone Joint Surg Am 1986; 68: 647-659. Mackenney P J, McQueen M M, Elton R. Prediction of instability in distal radial fractures. J Bone Joint Surg Am 2006; 88: 1944-51. Martinez-Mendez D, Lizaur-Utrilla A, de-Juan-Herrero J. Intra-articular distal radius fractures in elderly: a randomized prospective study of casting versus volar plating. J Hand Surg Eur 2018; 43(2): 142-7. Mattila V M, Huttunen T T, Sillanpää P, Niemi S, Pihlajamäki H, Kannus P. Significant change in the surgical treatment of distal radius fractures: a nationwide study between 1998 and 2008 in Finland. J Trauma 2011; 71: 939-43. Mellstrand-Navarro C, Pettersson H J, Tornqvist H, Ponzer S. The operative treatment of fractures of the distal radius is increasing: results from a nationwide Swedish study. Bone Joint J 2014; 96-B: 963-9. Orbay J L, Fernandez D L. Volar fixed-angle plate fixation for unstable distal radius fractures in the elderly patient. J Hand Surg Am 2004; 29: 96-102. Plant C E, Parsons N R, Costa M L. Do radiological and functional outcomes correlate for fractures of the distal radius? Bone Joint J 2017; 99-B: 376-82. Rozental T D, Blazar P E. Functional outcome and complications after volar plating for dorsally displaced, unstable fractures of the distal radius. J Hand Surg Am 2006; 31: 359-65. Wadsten M Å, Sjödén G O, Buttazzoni G G, Buttazzoni C, Englund E, SayedNoor A S. The influence of late displacement in distal radius fractures on function, grip strength, range of motion and quality of life. J Hand Surg Eur 2018; 43(2):131-6. Walenkamp M M, Aydin S, Mulders M A, Goslings J C, Schep N W. Predictors of unstable distal radius fractures: a systematic review and metaanalysis. J Hand Surg Eur 2016; 41(5): 501-15. Weil Y A, Mosheiff R, Firman S, Liebergall M, Khoury A. Outcome of delayed primary internal fixation of distal radius fractures: a comparative study. Injury 2014; 45: 960-4. Yamashita K, Zenke Y, Sakai A, Oshige T, Moritani S, Maehara T. Comparison of functional outcome between early and delayed internal fixation using volar locking plate for distal radius fractures. J UOEH 2015; 37: 111-19.


Acta Orthopaedica 2019; 90 (2): 129–134

129

Patient-reported outcomes after a distal radius fracture in adults: a 3–4 years follow-up Roderick H VAN LEERDAM 1, Floortje HUIZING 1, Frank TERMAAT 1, Sanne KLEINVELD 2, Steven J RHEMREV 3, Pieta KRIJNEN 1, and Inger B SCHIPPER 1 1 Department 3 Department

of Surgery, Leiden University Medical Center (LUMC), The Netherlands; 2 Department of Surgery, Haga Hospital, The Netherlands; of Surgery, The Hague Medical Center (HMC), The Netherlands Correspondence: r.leerdam@gmail.com Submitted 2017-11-13. Accepted 2018-11-12.

Background and purpose — There are few reports on the outcome of distal radius fractures after 1 year. Therefore we investigated the long-term patient-reported functional outcome and health-related quality of life after a distal radius fracture in adults. Patients and methods — We reviewed 823 patients, treated either nonoperatively or operatively in 2012. After a mean follow-up of 3.8 years 285 patients (35%) completed the Patient-Rated Wrist Evaluation (PRWE) and EuroQol5D. Results — The mean PRWE score was 11. The mean EQ-5D index value was 0.88 and the mean EQ VAS for selfrated health status was 80. Nonoperatively treated type A and type B fractures had lower PRWE scores compared with operatively treated patients, whereas the EQ-5D was similar between groups. The EQ VAS for patients aged 65 and older was statistically significantly lower than that of younger patients. Interpretation — Patients had a good overall long-term functional outcome after a distal radius fracture. Patients with fractures that were possible to treat nonoperatively had less pain and better wrist function after long-term follow-up than patients who needed surgical fixation.

The initial treatment of distal radius fractures can be operative or non-operative. The functional outcome after these fractures may be impaired (Rozental et al. 2002, Chung et al. 2007, Ranjeet and Estrella 2012). Most studies describe the functional outcome over a 1-year period. The recovery is characterized by an initial decline in function followed by improvement after 6–12 months. Though functional recovery is known to often take more than 1 year, studies that describe functional recovery beyond 1 year are scarce and report predominantly on certain subgroups (Brogren et al. 2011, Landgren et al. 2011, Williksen et al. 2013, Lalone et al. 2017). In recent decades there has been a shift in how to assess the functional outcome after distal radius fractures. Previously, the outcome was predominantly determined by physicianreported parameters such as the radiographic properties of the fracture or range of motion of the injured joint or limb. However, these clinical parameters do not represent patients’ perspectives and seem less relevant for outcome evaluation (Fujii et al. 2002, Anzarut et al. 2004, Wilcke et al. 2007, Finsen et al. 2013, Plant et al. 2017). Nowadays, patient-reported outcomes (PROs) seem to be more relevant and are used as standard for this purpose. In the initial phase of the treatment, patients are focused on pain relief rather than long-term expectations and consequences. Patient-relevant long-term results include functional recovery, return to work, and the ability to perform daily activities. For distal radius fractures, several PROMs have been developed and validated such as the Patient Rated Wrist Evaluation (PRWE) (Abramo et al. 2008, Kleinlugtenbelt et al. 2016, Landgren et al. 2017, MacDermid 2011) and the Disability of the Arm, Shoulder and Hand score (DASH) (Hudak et al. 1996). To assess a patient’s general health and quality of life a generic PROM like the EuroQol-5D (EQ-5D) or the Short Form 36 (SF-36) can be used (Brooks 1996, Warek and

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2019.1568098


130

Acta Orthopaedica 2019; 90 (2): 129–134

Falkinham 1996). So far, the available validated PROMs for distal radius fractures have been mainly used for research purposes. Although not yet implemented in daily clinical practice, it is assumed that PROMs can be a valuable tool for monitoring outcome during routine clinical follow-up (MacDermid et al. 2004). Most studies on determining the functional outcomes of distal radius fractures focus on the outcome of PROMs within the first year. We assessed long-term patient-reported outcome after a nonoperative or operatively treated distal radius fracture. 

To identify risk factors for a worse patient-reported functional outcome we recorded age (< 65 years or ≥ 65 years), sex, fracture side, whether the dominant hand was fractured, fracture type according to the AO classification (type A: extraarticular fracture; type B: partial articular fracture; type C: complete articular fracture) and treatment (nonoperative or operative) (Müller et al. 1990, Marsh et al. 2007). Patients who were eligible according to hospital records received the questionnaires and a written consent form by post in January 2016. To increase the response rate, all nonresponders were additionally contacted by phone up to 3 months after the initial questionnaires were sent.

Patients and methods

Statistics The patient characteristics were described using summary statistics and the response group was compared with the nonresponse group to study whether the respondents were representative of the total study population. PRWE and EQ-5D scores were compared between patient groups (sex, treatment, dominant hand fractured, AO classification) using t Students’ t-test or 1-way ANOVA. Multiple linear regression analysis was conducted to identify which patient and fracture characteristics (age, sex, fracture type, dominance of the fractured hand, treatment) were associated with the PRWE score. All statistical analyses were performed using SPSS Statistics for Windows, version 23 (IM Corp, Armonk, NY, USA). P-values < 0.05 were considered statistically significant.

Hospital patient records were reviewed to identify all patients with a distal radius fracture (DRF) that were treated in 1 of the 3 regional Level-1 trauma center locations in the Western part of the Netherlands between January 2012 and December 2012. Yearly, an estimate of 900 distal radius fractures are treated in these centers. Inclusion criteria were completed treatment for a DRF of any type with a minimum follow-up of 3 years and age ≥ 18 years. Patients were excluded if they could not be followed for at least 3 years (e.g., deceased or living abroad), were mentally incompetent, had more than one (contralateral or ipsilateral) or an open DRF, or had other diseases or injuries that interfered with the normal function of the wrist (e.g., rheumatic diseases or neurodegenerative diseases, multiple injuries). Treatment was either operative or nonoperative. Type of treatment was based on the national treatment guideline for DRFs (Dutch Surgical Society 2010) and supervised by the attending surgeon. At the time of the study (January–April 2016) all included patients had completed their treatment. Functional outcome was measured using the PRWE. The PRWE is a 15-item questionnaire on wrist pain and disability in activities of daily living as perceived by the patient. Scores on 2 subscales (pain and function) are combined in a total score ranging from 0 (no pain nor disability) to 100 (severe pain and disability) (MacDermid et al. 1998). The total PRWE score was calculated using the published algorithm (sum of the 5 pain items plus the sum of the 10 function items divided by 2) (MacDermid 2011). Questions not answered by patients were replaced with the mean score of the subscale as specified in the PRWE user manual. The patient-perceived health-related quality of life was measured using the EQ-5D questionnaire, a generally accepted and validated PROM (Brooks 1996). The EQ-5D descriptive system has 5 questions regarding mobility, self-care, daily activities, pain/discomfort, and anxiety/ depression. The scores for these 5 health dimensions are converted to a utility score that ranges from –0.33 to 1; a lower score reflects a poorer quality of life (Lamers et al. 2005). The EQ Visual Analogue Scale (a vertical 100 mm scale from zero to 100) is a self-rated health status measure.

Ethics, funding, and potential conflicts of interest Prior to the start of the study, the institution’s ethics committee approved the study (P15.311). No financial support for the research was obtained. No conflicts of interest were declared.

Results 823 patients met the inclusion criteria and were sent the questionnaires by mail. 364 (44%) did not respond and were considered lost to follow-up. Another 174 patients (21%) were unwilling to participate. The response rate was thus 35%. 13 of the 285 responders were excluded: 6 patients suffered from other diseases/injuries that interfered with proper function of the wrist, 5 had more than one DRF, and 2 completed less than 50% of the questionnaire. 207 (73%) responded by mail and 78 by phone, after initially not having responded. 243 (89%) patients filled in the questionnaire completely. Of the incomplete questionnaires no patient left more than 4 of the 33 questions unanswered. Responders and non-responders were similar regarding age, sex, fracture side, and AO fracture classification. The proportion of operatively treated fractures was higher in the responders’ group compared with the nonresponders’ group: 32% versus 20% (p < 0.01) (Tables 1 and 2). Mean follow-up was 46 months (42–50).


Acta Orthopaedica 2019; 90 (2): 129–134

131

Table 1. Patient characteristics of response and non-response groups

Table 2. Characteristics of included participants

Factor

Response Non-response p-value

Patient characteristics

Number, n (%) Mean age (SD) (years) Sex, n (%) Male Female AO fracture type, n (%) A B C Treatment group, n (%) Nonoperative Operative Fractured side, n (%) Left Right

285 (35) 62 (16) 71 (25) 214 (75)

538 (65) 61 (21) 157 (29) 381 (71)

113 (40) 90 (32) 82 (29)

250 (47) 161 (30) 127 (24)

0.1

195 (68) 90 (32) 159 (56) 126 (44)

431 (80) 107 (20) 286 (53) 250 (47)

< 0.001

Number, n Mean age (SD) (years) Sex, n (%) Months of follow-up (SD) AO fracture type, n (%) Treatment group, n (%) Fractured side, n (%) Dominant side fractured, n (%)

0.7 0.2

0.6

Table 3. Patient-reported outcomes for PRWE and EQ-5D. Values are mean (SD)

Total Age

PRWE p-value EQ-5D p-value

11 (18) < 65 10 (16) 0.2 ≥ 65 12 (20) Sex Male 8 (13) 0.2 Female 12 (20) AO fracture type A 9 (18) 0.6 B 11 (19) C 12 (18) Treatment group Nonoperative 8 (15) < 0.01 Operative 17 (22) Dominant side fractures Yes 10 (16) 0.3 No 13 (21)

EQ VAS p-value

0.88 (0.20) 0.90 (0.20) 0.85 (0.20) 0.92 (0.16) 0.87 (0.21)

0.06 0.09

80 (15) 82 (15) 78 (16) 83 (14) 79 (16)

0.88 (0.20) 0.89 (0.16) 0.86 (0.23)

0.6

80 (15) 79 (15) 80 (16)

0.9

0.89 (0.19) 0.86 (0.22)

0.4

80 (15) 79 (16)

0.7

0.87 (0.21) 0.86 (0.22)

0.9

78 (15) 80 (16)

0.3

Patient-reported outcome Patient-Rated Wrist Evaluation The mean PRWE score was 11 (SD 18, range 0–96). There was no statistically significant difference in PRWE scores regarding sex (p = 0.2) or AO fracture type (p = 0.6). Wrist pain and function were similar for patients who fractured their dominant hand (p = 0.3). PRWE scores also did not statistically significantly differ between elderly (65 years and over) and younger patients (p = 0.2). The patients in the nonoperative treatment group had lower PRWE scores, indicating less pain and better wrist function, compared with the patients in the operative treatment group (p < 0.01) (Tables 3 and 4). In the multiple linear regression analysis the nonoperatively treated patients with a type A fracture scored better than operatively treated type A fractures; this was borderline significant (p = 0.06). The PRWE score for type B fractures was favorable in the nonoperative group (p < 0.01). No significant difference between PRWE score of non-operatively or operatively treated C fractures was noted (Table 5, see Supplementary data).

0.04 0.09

Male Female A B C Non-operative Operative Left No Missing

272 62 (16) 69 (25) 203 (75) 46 (4) 107 (39) 86 (32) 79 (29) 185 (68) 87 (32) 149 (55) 123 (45) 53 (20)

Table 4. Mean PRWE score for conservative and operatively treated type A, B, and C fractures. Values are mean (SD) Operative Non-operative treatment PRWE n PRWE n PRWE difference p-value AO fracture type A 93 8 (15) B 67 7 (15) C 25 10 (14)

14 20 (29) 19 25 (24) 54 13 (19)

–12 0.002 –18 < 0.001 –2.9 0.2

EuroQoL-5D The mean EQ-5D score after follow-up was 0.88. EQ-5D scores were similar between AO fracture type groups or between the treatment groups. The mean EQ-5D score for patients aged 65 and older was lower than that of younger patients (p = 0.06). The mean EQ VAS was 80 (range 30–100; 95% CI 78–82). The EQ VAS score for patients aged 65 and older was lower than that of younger patients (p = 0.04). There were no significant differences in the EQ VAS between other subgroups (Table 3). 

Discussion The patients in this study, with a mean age of 62 years, had a good overall long-term functional outcome after a DRF. Nonoperatively treated type A and B fractures had better PRWE


132

scores than operatively treated type A and B fractures after a mean follow-up of almost 4 years indicating better wrist function with less pain. In multiple linear regression analysis this difference was only statistically significant for type B fractures. We found similar patient-reported functional outcome between nonoperatively treated type A, B, and C fractures. Higher scores for health-related quality of life as measured by the EQ-5D did not go hand in hand with better wrist function, as measured by PRWE. Most previous studies on patient-reported functional outcome presented results obtained after 3, 6, or 12 months of follow-up. Our study assessed patient-reported outcomes at a mean of 4 years. We found a significant difference in functional outcome between the 2 treatment groups, in favor of nonoperative treatment. In the subgroup analysis this functional benefit was most obvious in type A and B fractures. The minimum clinically relevant difference of the PRWE score for patients with a DRF has been reported to vary from 4 to 20 (Kim and Kang 2013, Walenkamp et al. 2015, Kleinlugtenbelt et al. 2018). We found a mean difference of 12 and 18 between operatively and nonoperatively treated type A and B fractures respectively, which could indicate a clinically relevant difference. A number of papers have compared patient-reported functional outcome between nonoperative and operative treatment in elderly patients. Arora et al. (2011) found significantly better patient-reported wrist function for type A and C fractures in the operative treatment group in the early postoperative time period, yet after 6 and 12 months only grip strength was statistically significantly better. Egol et al. (2010) did not report the functional benefit for operatively treated patients. At 3, 6, and 12 months there was no significant difference in patient-reported wrist function. Research by Bartl et al. (2014) concerning type C fractures only found a marginal and clinically unimportant difference in functionality in favor of operatively treated patients at 3 and 12 monthsâ&#x20AC;&#x2122; follow-up. Similar to our results, quality of life was similar between operative and nonoperative treatment. In contrast with these 3 studies and our findings, Sharma et al. (2014) found statistically significantly better wrist function in operatively treated type B and C fractures at the final follow-up, after 24 months. To our knowledge no other studies have reported on the long-term follow-up patient-reported functional outcome in favor of nonoperative treatment. In our study the EQ-5D showed no statistically significant differences in outcome for age, sex, treatment group, or fracture type. An earlier study implied that the EQ-5D values might normalize to pre-fracture values at 6â&#x20AC;&#x201C;12 months (Hagino et al. 2009). Because of these contradictory results we feel that the EQ-5D may not be the most discriminating and appropriate instrument in evaluation of long-term follow-up in DRFs. One of the strengths of this study was the large number of patients who participated. This provided considerable statisti-

Acta Orthopaedica 2019; 90 (2): 129â&#x20AC;&#x201C;134

cal power to the analysis (Table 2). Our population included all types of and differently treated distal radius fractures. In the subgroup analysis, however, the group sizes were relatively small and hence statistical analysis was not meaningful. Another limitation is a greater proportion of operatively treated patients in the response group, which introduces a potential bias. The over-representation may indicate that the operated patients had worse results compared with the nonoperatively treated patients. The willingness to respond to the questionnaire might be influenced by the residual level of pain and dysfunction; operatively treated patients might be more eager to share their experiences. Another explanation could be that having surgery is a big life event, triggering a higher rate of response. Also, more type C fractures were present in the operated group. This is merely due to the national guideline, which advises that complex fractures be treated more often with internal or external fixation. However, we did not find statistically significant differences in functional outcome between fracture types, indicating that in our study function did not correlate with fracture type. Another shortcoming was the low response rate of 35% due to a high number of non-responding or unwilling-to-participate patients. To increase the response rate, all non-responders were additionally contacted by phone up to 3 months after the initial questionnaires were sent. This introduces a risk of bias. Finally, the type of surgery and experience of the surgeon was not investigated. Both parameters could also affect outcome (Landgren et al. 2017). They were, however, beyond the scope of our study. Operative treatment for type A and B wrist fractures was associated with worse patient-reported wrist function at prolonged follow-up compared with nonoperative treatment. We speculate that this is possibly due to a higher complication rate after operative treatment, such as symptomatic hardware and hardware-related tendon irritation or even injury that can occur up to 4 years after surgery (Soong et al. 2011). Most of the published studies evaluate complications up to one year after operation. We did not, though, have the data in this study to substantiate this speculation. However, further analysis of the records of operatively treated patients showed that, in 9 out of the 87 patients, hardware was removed because of functional limitations and pain. This could even be an underestimation if patients sought care elsewhere. The mean PRWE score of patients in this study who had their hardware removed was 24 compared with 16 in patients who did not. Although this group was too small to perform statistical analysis, it should be noted that after reoperation patient wrist function was still not satisfactory. An alternative explanation for the difference between the 2 treatment groups could be that patients who were treated operatively have different expectations than patients who were treated nonoperatively. The treating surgeon must be aware of this: expectation management plays a crucial role in satisfaction with final functional outcome.


Acta Orthopaedica 2019; 90 (2): 129–134

In summary, after a mean follow-up of 4 years, patients generally perceived the outcome after a distal radius fracture as good. Patients with fractures that it was deemed possible to treat nonoperatively had better PRWE scores compared with those needing operative treatment independent of sex and affected side. Quality of life between the treatment groups was similar. Supplementary data Table 5 is available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/17453674. 2019.1568098

Study conception and design: RL, FT, IS. Acquisition of data: FH, RL. Analysis and interpretation of data: RL, FH, FT, PK, IS. Drafting of manuscript: RL, FH, FT, SK, SR, PK, IS. Critical revision: RL, FT, SK, SR, PK, IS. Acta thanks Marcus Landgren, Henrik Sandelin and Niels Schep for help with peer review of this study.

Abramo A, Kopylov P, Tagil M. Evaluation of a treatment protocol in distal radius fractures: a prospective study in 581 patients using DASH as outcome. Acta Orthop 2008; 79(3): 376-85. doi: 10.1080/1745367071 0015283. Anzarut A, Johnson J A, Rowe B H, Lambert R G, Blitz S, Majumdar S R. Radiologic and patient-reported functional outcomes in an elderly cohort with conservatively treated distal radius fractures. J Hand Surg Am 2004; 29(6): 1121-7. doi: 10.1016/j.jhsa.2004.07.002. Arora R, Lutz M, Deml C, Krappinger D, Haug L, Gabl M. A prospective randomized trial comparing nonoperative treatment with volar locking plate fixation for displaced and unstable distal radial fractures in patients sixtyfive years of age and older. J Bone Joint Surg Am 2011; 93(23): 2146-53. doi: 10.2106/JBJS.J.01597. Bartl C, Stengel D, Bruckner T, Gebhard F, Group O S. The treatment of displaced intra-articular distal radius fractures in elderly patients. Dtsch Arztebl Int 2014; 111(46): 779-87. doi: 10.3238/arztebl.2014.0779. Brogren E, Hofer M, Petranek M, Dahlin L B, Atroshi I. Fractures of the distal radius in women aged 50 to 75 years: natural course of patientreported outcome, wrist motion and grip strength between 1 year and 2–4 years after fracture. J Hand Surg Eur 2011; 36(7): 568-76. doi: 10.1177/1753193411409317. Brooks R. EuroQol: the current state of play. Health Policy 1996; 37(1): 53-72. Chung K C, Kotsis S V, Kim H M. Predictors of functional outcomes after surgical treatment of distal radius fractures. J Hand Surg Am 2007; 32(1): 76-83. doi: 10.1016/j.jhsa.2006.10.010. Dutch Surgical Society. Richtlijn Distale radius fracturen: Diagnostiek en behandeling. Nederlandse Vereniging voor Heelkunde 2010 [Treatment guidelines for distale radius fractures]. https://heelkunde.nl/sites/ heelkunde.nl/files/richtlijnen-definitief/Richtlijn_Distale_radius_fracturen_definitieve_versie_0511.pdf (accessed June 1, 2018). Egol K A, Walsh M, Romo-Cardoso S, Dorsky S, Paksima N. Distal radial fractures in the elderly: operative compared with nonoperative treatment. J Bone Joint Surg Am 2010; 92(9): 1851-7. doi: 10.2106/JBJS.I.00968. Finsen V, Rod O, Rod K, Rajabi B, Alm-Paulsen P S, Russwurm H. The relationship between displacement and clinical outcome after distal radius (Colles’) fracture. J Hand Surg Eur 2013; 38(2): 116-26. doi: 10.1177/1753193412445144.

133

Fujii K, Henmi T, Kanematsu Y, Mishiro T, Sakai T, Terai T. Fractures of the distal end of radius in elderly patients: a comparative study of anatomical and functional results. J Orthop Surg (Hong Kong) 2002; 10(1): 9-15. doi: 10.1177/230949900201000103. Hagino H, Nakamura T, Fujiwara S, Oeki M, Okano T, Teshima R. Sequential change in quality of life for patients with incident clinical fractures: a prospective study. Osteoporos Int 2009; 20(5): 695-702. doi: 10.1007/ s00198-008-0761-5. Hudak P L, Amadio P C, Bombardier C. Development of an upper extremity outcome measure: the DASH (disabilities of the arm, shoulder and hand) [corrected]. The Upper Extremity Collaborative Group (UECG). Am J Ind Med 1996; 29(6): 602-8. doi: 10.1002/(SICI)10970274(199606)29:6<602::AID-AJIM4>3.0.CO; 2-L. Kim J K, Kang JS . Evaluation of the Korean version of the patient-rated wrist evaluation. J Hand Ther 2013; 26(3): 238-43; quiz 44. doi: 10.1016/j. jht.2013.01.003. Kleinlugtenbelt Y V, Nienhuis R W, Bhandari M, Goslings J C, Poolman R W, Scholtes V A. Are validated outcome measures used in distal radial fractures truly valid? A critical assessment using the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) checklist. Bone Joint Res 2016; 5(4): 153-61. doi: 10.1302/2046-3758.54.2000462. Kleinlugtenbelt Y V, Krol R G, Bhandari M, Goslings J C, Poolman R W, Scholtes V A B. Are the patient-rated wrist evaluation (PRWE) and the disabilities of the arm, shoulder and hand (DASH) questionnaire used in distal radial fractures truly valid and reliable? Bone Joint Res 2018; 7(1): 36-45. doi: 10.1302/2046-3758.71.BJR-2017-0081.R1. Lalone E, MacDermid J, Grewal R, King G. Patient reported pain and disability following a distal radius fracture: a prospective study. Open Orthop J 2017; 11:589-99. doi: 10.2174/1874325001711010589. Lamers L M, Stalmeier P F, McDonnell J, Krabbe P F, van Busschbach J J. [Measuring the quality of life in economic evaluations: the Dutch EQ-5D tariff]. Ned Tijdschr Geneeskd 2005; 149(28): 1574-8. Landgren M, Jerrhag D, Tagil M, Kopylov P, Geijer M, Abramo A. External or internal fixation in the treatment of non-reducible distal radial fractures? Acta Orthop 2011; 82(5): 610-13. doi: 10.3109/17453674.2011.618910. Landgren M, Abramo A, Geijer M, Kopylov P, Tagil M. Similar 1-year subjective outcome after a distal radius fracture during the 10-year-period 2003–2012. Acta Orthop 2017; 88(4): 451-6. doi: 10.1080/17453674.2017.1303601. MacDermid J. The patient-rated wrist evaluation (PRWE) user manual. June 2011. https://srs-mcmaster.ca/wp-content/uploads/2015/05/EnglishPRWE-User-Manual.pdf MacDermid J C, Turgeon T, Richards R S, Beadle M, Roth J H. Patient rating of wrist pain and disability: a reliable and valid measurement tool. J Orthop Trauma 1998; 12(8): 577-86. MacDermid V E, Neuber-Hess M S, Rose P K. The temporal sequence of morphological and molecular changes in axotomized feline motoneurons leading to the formation of axons from the ends of dendrites. J Comp Neurol 2004; 468(2): 233-50. doi: 10.1002/cne.10966. Marsh J L, Slongo T F, Agel J, Broderick J S, Creevey W, DeCoster T A, Prokuski L, Sirkin M S, Ziran B, Henley B, Audige L. Fracture and dislocation classification compendium 2007: Orthopaedic Trauma Association classification, database and outcomes committee. J Orthop Trauma 2007; 21(10 Suppl): S1-133. Müller M E, Nazarian S, Koch P, Schatzker J. The comprehensive classification of fractures of long bones. Berlin/New York: Springer-Verlag; 1990. Plant C E, Parsons N R, Costa M L. Do radiological and functional outcomes correlate for fractures of the distal radius? Bone Joint J 2017; 99-B(3): 37682. doi: 10.1302/0301-620X.99B3.35819. Ranjeet N, Estrella E P. Distal radius fractures: does a radiologically acceptable reduction really change the result? J Clin Diagnostic Res 2012; 6(8): 1388-92. doi: 10.7860/JCDR/2012/4567.2366. Rozental T D, Beredjiklian P K, Steinberg D R, Bozentka D J. Open fractures of the distal radius. J Hand Surg Am 2002; 27(1): 77-85.


134

Sharma H, Khare G N, Singh S, Ramaswamy AG, Kumaraswamy V, Singh A K. Outcomes and complications of fractures of distal radius (AO type B and C): volar plating versus nonoperative treatment. J Orthop Sci 2014; 19(4): 537-44. doi: 10.1007/s00776-014-0560-0. Soong M, van Leerdam R, Guitton T G, Got C, Katarincic J, Ring D. Fracture of the distal radius: risk factors for complications after locked volar plate fixation. J Hand Surg Am 2011; 36(1): 3-9. doi: 10.1016/j.jhsa.2010.09.033. Walenkamp M M, de Muinck Keizer R J, Goslings J C, Vos L M, Rosenwasser M P, Schep N W. The minimum clinically important difference of the patientrated wrist evaluation score for patients with distal radius fractures. Clin Orthop Relat Res 2015; 473(10): 3235-41. doi: 10.1007/s11999-015-4376-9.

Acta Orthopaedica 2019; 90 (2): 129â&#x20AC;&#x201C;134

Warek U, Falkinham J O, 3rd. Action of clofazimine on the Mycobacterium avium complex. Res Microbiol 1996; 147(1-2): 43-8. Wilcke M K, Abbaszadegan H, Adolphson P Y. Patient-perceived outcome after displaced distal radius fractures: a comparison between radiological parameters, objective physical variables, and the DASH score. J Hand Ther 2007; 20(4): 290-8; quiz 9. doi: 10.1197/j.jht.2007.06.001. Williksen J H, Frihagen F, Hellund J C, Kvernmo H D, Husby T. Volar locking plates versus external fixation and adjuvant pin fixation in unstable distal radius fractures: a randomized, controlled study. J Hand Surg Am 2013; 38(8): 1469-76. doi: 10.1016/j. jhsa.2013.04.039.


Acta Orthopaedica 2019; 90 (2): 135–142

135

The design of the cemented stem influences the risk of Vancouver type B fractures, but not of type C: an analysis of 82,837 Lubinus SPII and Exeter Polished stems Georgios CHATZIAGOROU 1,2, Hans LINDAHL 1,3, and Johan KÄRRHOLM 1,2 1 The Swedish Hip Arthroplasty Register, Gothenburg; 2 Department of Orthopaedics, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg; 3 Department of Orthopaedics, Lidköping Hospital, Sweden Correspondence: g.chatziagorou@gmail.com Submitted 2018-09-30. Accepted 2019-01-07.

Background and purpose — In total hip replacements, stem design may affect the occurrence of periprosthetic femoral fracture. We studied risk factors for fractures around and distal to the 2 most used cemented femoral stems in Sweden. Patients and methods — This is a register study including all standard primary Lubinus SPII and Exeter Polished stems operated in Sweden between 2001 and 2009. The outcome was any kind of reoperation due to fracture around (Vancouver type B) or distal to the stem (Vancouver type C), with use of age, sex, diagnosis at primary THR, and year of index operation as covariates in a Cox regression analysis. A separate analysis of the primary osteoarthritis patient group was done in order to evaluate eventual influence of the surgical approach (lateral versus posterior) on the risk for Vancouver type B fractures. Results — The Exeter stem had a 10-times (95% CI 7–13) higher risk for type B fractures, compared with the Lubinus, while no statistically significant difference was noticed for type C fractures. The elderly, and patients with hip fracture or idiopathic femoral head necrosis, had a higher risk for both fracture types. Inflammatory arthritis was a risk factor only for type C fractures. Type B fractures were more common in men, and type C in women. A lateral approach was associated with decreased risk for Type B fracture. Interpretation — Stem design influenced the risk for type B, but not for type C fracture. The influence of surgical approach on the risk for periprosthetic femoral fracture should be studied further.

Periprosthetic femoral fracture (PPFF) is more common in uncemented stems (Hailer et al. 2010, Thien et al. 2014, Abdel et al. 2016). In cemented stems, higher risk for fracture has been reported for “force-closed” (e.g., Exeter Polished, CPT), compared with “shape-closed” stems like Lubinus SPII and Charnley (Lindahl et al. 2005, Thien et al. 2014, Broden et al. 2015, Palan et al. 2016). However, most of the previous studies have focused on fractures treated with stem revision (Thien et al. 2014, Palan et al. 2016), and hence mainly fractures around a loose stem. It is probable that the shape and the surface finish of the stem contribute to the risk for Vancouver type B fractures (fractures around or close to a femoral stem) (Broden et al. 2015, Palan et al. 2016). Little research has been done to investigate whether the design of the stem can affect the risk for suffering a fracture distal to the stem (Vancouver type C) (Lowenhielm et al. 1989). The majority of hip arthroplasty registries report only primary procedures and revisions. Therefore, type C fractures, treated in principle with open reduction and internal fixation (ORIF) without revision, are not reported. A recent register study from Sweden (Chatziagorou et al. 2018), revealed that only 17% of these fractures were reported to the Swedish Hip Arthroplasty Register (SHAR). Type C fractures were numerically more common among Lubinus SPII stems, while type B fractures predominated after insertion of an Exeter stem. These figures were, however, not related to the numbers at risk in each group. We are not aware of any study where the majority of type C fractures treated without revision were included. Nothing is known regarding the influence of surgical approach on the risk for postoperative periprosthetic fracture around a total cemented hip prosthesis. Both the Exeter and the Lubinus stems are frequently used in Sweden. Between 2000 and 2016, 104,081 Lubinus

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2019.1574387


136

Acta Orthopaedica 2019; 90 (2): 135–142

Eligible primary THRs n = 82,837 Lubinus SP II stems, n = 55,026 (66%) Exeter polished stems, n = 27,811 (34%) Excluded THRs (n = 3,024): – with primary diagnosis of tumor, 548 – stems shorter or longer than standard length, 844 – previous history of hemiarthroplasty, 30 – THRs with uncemented cups, 1,602 Analyzed THRs n = 79,813 Lubinus SP II stems, n = 52,625 (66%) Exeter polished stems, n = 27,188 (34%)

Table 1. Periprosthetic fractures (n = 31) primarily excluded Intraoperative fracture Malignancy at the time of reoperation Active deep infection Perforation only Fracture occurred during TKR surgery Vancouver type A Sawing (non-iatrogenic)

10 5 5 3 3 4 1

length is one of these parameters, and during the study period (2001–2011), the standard length for both involved stems was 150 mm. Stem lengths other than 150mm were excluded. Further excluFigure 1. Flow chart. Of the 73,630 originally included patients, 70,981 remained for analysis. Standard stem length was 150 mm for both Exeter and Lubinus SPII. sion criteria are presented in a flow chart (Figure . 1). Surgical treatment of fracture types excluded SPII and 53,358 Exeter stems were used in primary total hip for various reasons was labelled as reoperation due to causes replacements (THR) (Karrholm et al. 2017). This corresponds other than PPFF in the analyses (Table 1). to two-thirds of all primary THRs during this period. PreviData for the primary THRs and the reoperations were ous studies have shown that Exeter stems run increased risk of derived from the SHAR. The reporting of primary hip arthrorevision due to periprosthetic fractures, whereas less is known plasties is almost complete (98%) (Karrholm et al. 2017), about the risk for reoperation including also operative treat- whereas the reporting of reoperations is poorer (Söderman et ment with osteosynthesis and without stem exchange. al. 2000, Lindgren et al. 2014). Therefore, data linking was We compared the Lubinus SPII and the Exeter Polished done between the SHAR and the National Patient Register stem as risk factors for Vancouver type B and C fractures. (NPR), in order to detect even PPFFs not registered with the Other risk factors studied were age, sex, diagnosis at the pri- SHAR. Cross-matching for the other types of reoperations mary THR, year of index operation, and surgical approach. was not done. The NPR holds information on all inpatient care We hypothesized that Lubinus stems might run an increased since 1987, and all outpatient care since 2001. Both private risk of type C fractures because of the high resistance of this and public healthcare providers have had to report to the NPR stem to undergoing type B fractures ending up in a revision. since 2001. All medical records of reoperations due to fracture To include all types of surgical procedures of the operated were collected and scrutinized to detect all femoral fractures femur with relation to the hip prosthesis inserted, our primary in patients with a primary THR. The information provided in outcome was any reoperation due to periprosthetic fracture. the case records was also used for fracture classification by GC, according to the Vancouver classification system (Brady et al. 1999). A detailed description of the classification process, as well as its validation, is described in a previous publiPatients and methods cation (Chatziagorou et al. 2018). Bilateral observations were All primary standard Lubinus SPII and Exeter Polished included as previous studies have indicated that this will not stems used in THRs between 2001 and 2009, and reported cause significant problems related to dependency (Ranstam et to the SHAR, were included. We studied reoperations for al. 2011). any reason, and specifically due to PPFF between 2001 and 2011, to include a minimum of 2 years’ follow-up (maximum Statistics follow-up 11 years). Follow-up ended at the date of reopera- Statistical calculations were done using IBM SPSS statistion for any reason, death, emigration, or on December 31, tics 23 (IBM Corp, Armonk, NY, USA). To identify eventual 2011, whichever came first. Reoperation was defined as any demographic differences between the Lubinus and the Exeter further surgical intervention related to the index hip arthro- group, a chi-squared test and Mann–Whitney test were used. plasty irrespective of whether the prosthesis or parts of it have The 10-year survival was calculated with Kaplan–Meier analbeen exchanged, extracted, or left untouched. All type A frac- ysis (log rank test). We plotted survival curves for the covaritures (fractures of the greater and lesser trochanter), conserva- ates included, and log–log plots to test that the Cox proportively treated periprosthetic fractures, and fractures occurring tional hazard model was fulfilled. A Cox regression model during insertion of a primary stem (intraoperative fractures) was used to analyze the relative risk for reoperation due to were excluded. The SHAR records stem characteristics pro- PPFF. Adjustment for age, sex, type of stem, and diagnosis at spectively, for all primary and secondary arthroplasties. Stem the time of primary THR, as well as the year of index opera-


Acta Orthopaedica 2019; 90 (2): 135–142

tion, was performed. The distribution of the population into age groups was done according to the age at the time of the primary operation. The aim was to have as equally sized groups as possible. Diagnosis was separated into primary osteoarthritis (OA), inflammatory arthritis, hip fracture, idiopathic femoral head necrosis, and various (including sequel to childhood hip disease). Censored were cases with cause of reoperation other than PPFF, excluded cases (Table 1), patients who died without any reoperation, or those who had not been reoperated until the end of 2011. The surgical approach (lateral versus posterior), as a risk factor for Vancouver type B fracture, was studied in a subgroup analysis. Complete information on surgical approach was available in 43,639 Lubinus and in 22,271 Exeter cases with primary OA. Missing data were 9 cases for each stem (Table 6). In the other groups of diagnoses, up to 97.5% (hip fracture) had missing information. Therefore, we chose to include only those patients operated due to primary OA and with a lateral or posterior incision. P-values were 2-sided with a significance level < 0.05, and 95% confidence intervals (CI) were calculated.

137

Table 2. Patient demographics and reoperations Item

Lubinus SPII

Exeter

All

p-value

Primary THRs, n (%) 52,625 (66) 27,188 (34) 79,813 (100) Male sex, n (%) a 20,870 (40) 10,322 (38) 31,192 (39) < 0.001 b Median age (interquartile range): at primary THR 72.2 (13) 72.1 (13) 72.1 (13) at reoperation for any reason 73.0 (13) 74.7 (14) 74.0 (13) 0.001 c at reoperation due to PPFF 80.3 (14) 79.8 (13) 79.9 (13) Age group, n (%) a < 64 10,613 (20) 5,824 (22) 16,437 (20) 64–69 11,015 (21) 5,502 (20) 16,517 (21) 70–74 10,826 (21) 5,505 (20) 16,331 (20) 75–79 10,451 (20) 5,172 (19) 15,623 (20) 80–100 9,720 (18) 5,185 (19) 14,905 (19) Diagnoses, n (%) a < 0.001 b Primary OA 43,648 (83) 22,280 (82) 65,928 (83) Hip fracture 6,181 (12) 2,794 (10) 8,975 (11) Idiopathic femoral head necrosis 1,148 (2) 968 (4) 2,116 (3) Inflammatory arthritis 1,162 (2) 654 (2) 1,816 (2) Various d 486 (1) 492 (2) 978 (1) a Reoperations, n (%) All reasons 1,660 (3.2) 966 (3.6) 2,626 (3.3) 0.003 b Due to PPFFs 167 (0.3) 298 (1.1) 465 (0.6) < 0.001 b Revisions, n (%) a All reasons (revision of any part) 1,095 (2.1) 595 (2.2) 1,690 (2.1) All reasons (revision of the stem) 544 (1.0) 343 (1.3) 887 (1.1) 0.004 b Due to PPFFs e 18 (0.03) 131 (0.5) 149 (0.2) < 0.001 b P-value is referred to only in cases with statistically significant difference. a % of all primary Lubinus, Exeter, and both stems, respectively. b Pearson chi-squared test. c Mann–Whitney test. d Other reasons including sequel after childhood hip disease. e The number of revisions of any part due to fracture was the same as the number of stem revisions due to fracture.

Ethics, funding, and potential conflicts of interest The study was approved by the Central Ethical Review Board in Gothenburg (Entry number: 198-12, Date: 2012-04-05). There was no financial support for this research. The authors declare no conflict of interest.

Results Study population Between 2001 and 2009, 82,837 primary Lubinus SPII and Exeter Polished femoral stems were inserted in 73,630 patients. The data linking with the NPR resulted in a total of 626 PPFFs (295 of these were registered only in NPR), giving 4,233 reoperations between 2001 and 2011. After the exclusions (Figure 1 and Table 1), there were 79,813 primary THRs (70,981 patients), with 2,626 first-time reoperations (2,597 patients) left for analysis. 465 of the reoperations (462 patients) were due to periprosthetic femoral fracture. The mean follow-up time was 5.6 years. A slightly higher proportion of men was noted in the Lubinus SPII group (Table 2). The Exeter group had, proportionally, more patients classified as idiopathic femoral head necrosis and “various.” The cups

used with the Lubinus and the Exeter stems are presented in Table 3. Vancouver type and risk factors The proportion of reoperations due to PPFF was higher in the Exeter than in the Lubinus group, as reflected in the survival analyses (Table 2, Figure 2). The commonest fracture type observed after insertion of a Lubinus SPII stem was Vancouver type C (74%), whereas type B fractures were more common after use of Exeter Polished stems (73%, Table 4). The Exeter stem had a 3.5-times higher risk for PPFF (B or C), and a 9.6times higher risk for type B fracture when compared with the Lubinus SPII (Table 5). There was no statistically significant difference between the 2 groups regarding the risk of type C fracture. Overall, women more frequently sustained fractures distally to the stem, whereas men had a higher risk for fracture around the stem, and a slightly higher risk for PPFF in general (type B or C). The risk for fracture increased with age, irrespective of whether age was studied as a continuous or a categorical variable. Patients aged 80 years and older had the highest risk for both type B and C fractures, compared with patients younger than 64 years (Table 5).


138

Acta Orthopaedica 2019; 90 (2): 135–142

Figure 2. Cumulative survival (unadjusted) for periprosthetic femoral fracture. Numbers at risk at the end of 10 years’ follow-up were: 2,903 for the Lubinus SPII group, and 1,518 for the Exeter Polished group. 2(a): All fractures studied (Type B and C fractures). Mean survival at 10 years was 99.4% (SE 0.06) for the Lubinus SPII, and 97.9% (SE 0.17) for the Exeter Polished (log rank test p < 0.001). 2(b): Type B fractures. Mean survival at 10 years was 99.8% (SE 0.04) for the Lubinus SPII, and 98.6% (SE 0.11) for the Exeter Polished (log rank test p < 0.001). 2(c): Type C fractures. Mean survival at 10 years was 99.6% (SE 0.05) for the Lubinus SPII, and 99.3% (SE 0.11) for the Exeter Polished (log rank test p = 0.08).

Table 3. Type of cups (all cemented) used with Lubinus SPII and Exeter Polished stems. Values are frequency (%) Cups used with Lubinus SPII stems, total number Lubinus FAL Charnley Elite ZCA XLPE Exeter Duration OPTICUP Contemporary Hooded Duration Avantage Cemented Reflection Various Exeter Polished stems, total number Exeter Duration Charnley Elite Contemporary Hooded Duration Charnley Marathon XLPE Cenator ZCA XLPE Exeter Various

All THRs

PPFFs

52,625 (100) 44,620 (85) 5,075 (9.6) 943 (1.8) 809 (1.5) 674 (1.3) 158 (0.3) 111 (0.2) 93 (0.2) 55 (0.1) 87 (0.2) 27,188 (100) 9,157 (34) 8,308 (31) 6,454 (24) 2,041 (7.5) 714 (2.6) 194 (0.7) 168 (0.6) 68 (0.3) 84 (0.3)

167 (100) 139 (83) 16 (9.6) 5 (3) 0 2 (1.2) 3 (1.8) 0 1 (0.6) 0 1 (0.6) 298 (100) 110 (37) 97 (33) 48 (16) 22 (7.4) 12 (4.0) 4 (1.3) 1 (0.3) 1 (0.3) 3 (1.1)

22 different types of cups were used in conjunction with Lubinus SPII, and 17 with Exeter Polished stems.

Table 4. Distribution of periprosthetic femoral fractures according to the Vancouver classification system. Values are frequency (%) Vancouver Lubinus B1 B2 B3 C Total

27 (16) 15 (9) 2 (1) 123 (74) 167 (100)

Exeter 55 (19) 157 (53) 4 (1) 82 (28) 298 (100)

Type A fractures were excluded from this study.

All 82 (18) 172 (37) 6 (1) 205 (44) 465 (100)

Table 5. Risk factors, adjusted hazard ratios (HR), and 95% confidence intervals (CI) for reoperation due to periprosthetic femoral fracture Risk factors

Vancouver B&C Vancouver B Vancouver C HR (CI for HR) HR (CI for HR) HR (CI for HR)

Stem Lubinus SPII (ref.) 1 1 1 Exeter Polished 3.5 (2.9–4.2) 9.6 (7.0–13) 1.3 (0.95–1.7) Sex Men (reference) 1 1 1 Women 0.7 (0.6–0.8) 0.4 (0.3–0.5) 2.0 (1.4–2.8) Age groups < 64 (reference) 1 1 1 64–69 1.1 (0.8–1.5) 1.0 (0.6–1.5) 1.3 (0.7–2.2) 70–74 1.5 (1.1–2.1) 1.4 (0.9–2.2) 1.7 (1.0–2.8) 75–79 2.0 (1.5–2.7) 2.1 (1.4–3.1) 1.9 (1.2–3.1) 80–100 3.1 (2.3–4.2) 2.9 (2.0–4.3) 3.4 (2.1–5.4) Diagnoses Primary OA (ref.) 1 1 1 Inflam. arthritis 3.6 (2.3–5.5) 1.9 (0.9–4.2) 5.6 (3.3–9.6) Hip fracture 3.6 (2.9–4.5) 3.3 (2.4–4.4) 4.2 (3.0–5.7) Idiopathic femoral head necrosis 3.5 (2.4–5.0) 3.0 (1.9–5.0) 4.1 (2.4–7.1) Various a 2.0 (1.02–3.9) 1.6 (0.6–3.9) 2.8 (1.01–7.6) Calendar year for primary THR 1.1 (1.0–1.1) 1.1 (1.1–1.2) 1.0 (0.97–1.1) a Other

reasons including sequel after childhood hip disease.

Inflammatory arthritis, when compared with primary OA, did not affect the risk for fracture around a stem, but distal to it. Patients with hip fracture or idiopathic femoral necrosis had approximately 3 times higher risk for type B fractures, and 4 times for type C (Table 5). The later the year for the index operation, the more likely the patient would suffer a type B fracture. No corresponding time-related change in risk was observed as regards type C fractures. The subgroup analysis (lateral versus posterior approach) was done in 43,271 Lubinus SPII stems and 21,562 Exeter Polished stems, inserted


Acta Orthopaedica 2019; 90 (2): 135–142

139

Table 6. Distribution of surgical approach among hips with primary OA. Values are frequency (%) Approach a Item Lateral Posterior Other

All

Lubinus SPII Primary THRs 12,540 (100) 30,731 (100) 368 (100) 43,639 (100) Type B fractures 6 (0.05) 21 (0.07) 0 27 (0.06) Type C fractures 16 (0.13) 47 (0.15) 1 (0.27) 64 (0.15) Exeter Polished Primary THRs 10,471 (100) 11,091 (100) 709 (100) 22,271 (100) Type B fractures 49 (0.47) 89 (0.80) 8 (1.13) 146 (0.66) Type C fractures 14 (0.13) 27 (0.24) 2 (0.28) 43 (0.19) a9

Exeter and 9 Lubinus stems with unknown approach were excluded. Only patients operated with lateral or posterior approach were included in the separate regression analysis.

due to primary OA (Table 6). 1,077 stems operated with other surgical approaches were excluded from this analysis. Stems inserted with the posterior approach had a 1.6-times higher risk for suffering a Vancouver type B fracture compared with those inserted with a lateral approach (Table 7).

Discussion Several previous studies have demonstrated an increased risk for periprosthetic fracture of the Exeter when compared with the Lubinus stem (Lindahl et al. 2005, Thien et al. 2014). To our knowledge, this is the first study that distinguished between Vancouver type B and type C fractures, based on extensive research to include all reoperations. Earlier studies have either looked at the overall risk for periprosthetic fracture (Lindahl et al. 2005, Palan et al. 2016), or the risk for revision due to fracture (Cook et al. 2008, Thien et al. 2014) (mainly Vancouver type B2 and type B3 fractures), for one or both of these stems. Our main finding is that the Lubinus SPII did not have a higher risk for type C fractures, despite the fact that almost 3 out of 4 fractures around this stem were located distal to it (see Table 4). The finding that the Exeter Polished stem had a higher risk for fracture (type B and overall), confirms earlier publications (Lindahl et al. 2005, Thien et al. 2014). The commonest fracture type in this material was, however, type C (see Table 4). This observation results from an almost complete registration of fractures treated with osteosynthesis only, and without any stem revision (Chatziagorou et al. 2018). In Sweden, type B fractures are more common in uncemented stems, and type C fractures in cemented stems (Chatziagorou et al. 2018), in contrast to a previous study from the Mayo Clinic (Abdel et al. 2016). The cemented Lubinus SPII stem (Waldemar Link, Hamburg, Germany) is a shape-closed, CoCrMo, tapered, and anatomically s-shaped stem, with a collar, a matte finish, and a 19° built-in anteversion of the femoral neck. Its shape allows

Table 7. Risk factors, adjusted hazard ratios (HR), and 95% confidence intervals (CI) for reoperation due to Vancouver type B fracture Risk factors

Vancouver B HR (CI for HR)

Stem Lubinus SPII (reference) 1 Exeter Polished 11.4 (7.5–17) Sex Men (reference) 1 Women 0.4 (0.3–0.5) Age groups < 64 (reference) 1 64–69 1.3 (0.7–2.4) 70–74 1.9 (1.1–3.3) 75–79 3.1 (1.8–5.3) 80–100 4.5 (2.7–7.7) Calendar year for primary THR 1.1 (1.04–1.2) Surgical approach: Lateral (reference) 1 Posterior 1.6 (1.2–2.2) Only patients with the diagnosis of primary OA, and a lateral or posterior approach were included in this analysis.

neutral positioning in the femoral canal and resists rotational forces (Sesselmann et al. 2017), while the collar is claimed to restrict the distal migration of the stem (Catani et al. 2005). The anatomical shape of this stem probably facilitates an adequate cement mantle (Broden et al. 2015). The cemented Exeter stem (Stryker Howmedica, Mahwah, NJ, US) is a forceclosed, straight, collarless, double-wedge tapered, highly polished stem. It does not bond to the cement and is designed to subside into the cement mantle as a wedge (Palan et al. 2016). Both stems are well documented with excellent outcomes in the short and long term (Murray et al. 2013, Prins et al. 2014, Sesselmann et al. 2017, Westerman et al. 2018). It is postulated that the subsidence of the Exeter stem into the cement mantle will create an axial loading effect within the cement mantle, resulting in hoop stresses in the adjacent bone, which might increase the risk of sustaining a PPFF. As soon as a periprosthetic fracture occurs close to an Exeter stem, the stem is by definition loose (Broden et al. 2015). The reason why forceclosed cemented stems have a higher risk for periprosthetic fractures has been reported previously (Broden et al. 2015, Palan et al. 2016). The higher percentage of type C fractures within Lubinus SPII stems possibly has to do with the relative lower risk for fractures close to it (type B). We assume that these fractures are, rather, secondary to an osteoporotic femur (elderly, women, inflammatory arthritis, and previous history of hip fracture), than secondary to the stem’s design. Previous comparisons between the posterior and the lateral approach showed superior results for the former regarding the thickness of the cement mantle (Hank et al. 2010), the alignment of the stem (Vaughan et al. 2007, Broden et al. 2015), and revision risk due to aseptic loosening of the stem (Lind-


140

gren et al. 2012). Femurs with a loose stem are more prone to suffer a periprosthetic fracture (Lindahl et al. 2005). Thus, the posterior surgical approach should be beneficial regarding the risk for aseptic loosening and, hence, the risk for type B fractures around a loose stem. We are not aware of any publication where the surgical approach is studied as a risk factor for postoperative periprosthetic femoral fractures, secondary to a primary cemented THR. Berend et al. (2006) found that an anterolateral approach was associated with intraoperative fracture of the proximal femur, in both cemented and uncemented stems. A more recent study showed that patients older than 85 years, with hemiarthroplasty, had 2 times higher risk for postoperative PPFF if operated with a posterior approach, compared with those operated via a direct lateral approach (Rogmark et al. 2014), but this observation might have been confounded by inclusion of both cemented and uncemented stems of various designs. Our finding, that use of a posterior approach is associated with a higher risk for PPFF, is difficult to explain. A radiostereometric study (Glyn-Jones et al. 2006) observed slightly increased retrotorsion of the Exeter stem if inserted through a posterior compared with an anterolateral approach, suggesting a less secure stem fixation in the former group. Gore et al. (1982) showed less prosthetic femoral anteversion and more inward rotation of the operated hip with the posterior approach. A Cochrane review (Jolles and Bogoch 2006) also reported increased internal rotation of the hip joint in extension with use of the posterior approach, suggesting that implant loading might differ depending on the approach used. The influence of potential risk factors for PPFF such as age, sex, and diagnosis at the time of primary THR vary depending on the type of the stem (cemented/uncemented, primary/ secondary), the outcome measure (revision, reoperation, nonoperative treatment), and whether the fracture is intra- or postoperative (Berend et al. 2006, Cook et al. 2008, Meek et al. 2011, Abdel et al. 2016). We studied risk factors in patients with cemented Lubinus or Exeter stems and only those suffering a postoperative PPFF on the same side, and without any history of previous reoperation. Therefore, a generalization of our results for the whole population of patients with THR would be unreliable. High age, as well as the diagnosis of hip fracture or idiopathic femoral head necrosis, implied an increased risk for both type B and C fracture. Men had a higher risk for type B fractures, probably because of younger age with increased daily activity level (Witte et al. 2009) and higher risk for aseptic loosening than women (Hailer et al. 2010). Conversely, women, with more osteoporotic femoral bone and higher mean age at the time of primary THR, more frequently suffered type C fractures. The year of index operation influenced the risk for type B and not for type C fractures. This is probably the result of the increasing mean age at the index operation during the study period, from 71 years in 2001 to 72 years in 2009. Another reason could, theoretically, be a trend toward a decrease in

Acta Orthopaedica 2019; 90 (2): 135â&#x20AC;&#x201C;142

postoperative clinical and radiological controls after primary THR. This could lead to more cases with â&#x20AC;&#x153;unknown stem loosening,â&#x20AC;? and thus an increased risk for Vancouver type B fractures. The addition of the surgical approach in the subgroup analysis did not alter the relation of the other risk factors (age, sex, stem type, and year of primary THR). Inflammatory arthritis did not have a higher risk for fractures around the stem when compared with primary OA. This finding is in line with a previous report (Thien et al. 2014). On the other hand, femurs with inflammatory arthritis run a 6-times higher risk for distal femoral fractures. This is in accordance with a previous publication that reported a higher risk for osteoporotic fractures (Yamamoto et al. 2015) in patients with rheumatoid arthritis. There are limitations to our study. The linkage between the SHAR and the NPR included only reoperations due to periprosthetic fracture and not all other reasons for reoperation (aseptic loosening, infection, dislocation, other). These reoperations are recorded in the SHAR, but could be underreported, especially those performed owing to infection (Lindgren et al. 2014). Therefore, the real number of all reoperations could be slightly higher than found by us. Reoperations that took place before the PPFF were detected when the case records were scrutinized. All other reoperations not reported to the SHAR could most probably be expected to be equally distributed between the 2 groups studied. Another limitation, however, is that we did not include the presence of a total knee replacement (TKR) as a risk factor. Total hip replacements with an ipsilateral TKR have a higher risk for proximal femoral fracture (Katz et al. 2014), and total knee replacement is associated with distal femoral fracture (McGraw and Kumar 2010). We do not, however, think that the relative number of patients with TKR differs between those who have been operated with a Lubinus and those who have received an Exeter stem. Hips with primary osteoarthritis and inflammatory arthritis had almost the same share in the 2 groups. It is also important to underline that the classification process was based on reading of medical records. A better optimized way would be to define the fracture type based on information from both the medical records and the radiographs. In a previous validation of the classification process (Chatziagorou et al. 2018) we did, however, observe good agreement corresponding to previous validations of the Vancouver classification (Brady et al. 2000, Rayan et al. 2008). In addition, our analysis was based only on fractures classified as either B or C, without any further analyses of the sub-categories in the type B group. The methodological strength of this study was the relatively good data quality of a large volume of material, and its high external validity regarding PPFFs in the Swedish population. The hip prostheses studied in our report have a long tradition in Sweden with excellent implant survival (Junnila et al. 2016). The volume of our data was big enough to analyze only stems of the same length (150 mm). A difference in stem length can potentially affect the risk of periprosthetic fracture and its classification into type B or C. We also excluded unce-


Acta Orthopaedica 2019; 90 (2): 135–142

mented cups, since use of such implants was shown to result in a higher rate of femoral lysis when used with Exeter V40 stems (Westerman et al. 2018). Furthermore, we investigated all kind of reoperations due to PPFF, and not only revisions, which is the contemporary standard in other arthroplasty registries. This, in addition to the cross-matching with the NPR, gave us the unique opportunity to study an almost complete data set of fractures treated surgically with other methods such as ORIF and without concomitant revision of the stem (mostly type B1 and C fractures). Overall, the Exeter stem had almost an 3.5-times increased risk to suffer a periprosthetic fracture and about 10 times increased risk to suffer a PPFF leading to revision, which for the patient usually is a more demanding procedure than operation with osteosynthesis. According to our findings and previous studies the difference in risk ratio will increase further with increasing age and in patients with secondary OA. Lindahl et al. reported 23% reoperation rate (235 of 1,002) in patients treated surgically for type B and C fractures (Lindahl et al. 2005). We therefore think that our findings have clinical relevance and especially in the older population with a high incidence of osteoporosis. In summary, the Exeter Polished stem had a higher risk for postoperative periprosthetic femoral fractures of type B compared with the Lubinus SPII. As regards type C fractures there was no difference. The relative increased proportion of type C versus type B fractures in the Lubinus group might indicate that, after the insertion of a Lubinus stem, the distal femur will constitute the weakest part as long as the stem has not loosened. Our study suggested that the posterior approach may not be beneficial regarding the risk of PPFF in cemented THRs, but this observation needs to be studied further.

GC: Planning of the research, collection of the material, analysis of the material, manuscript. HL: Planning of the research, collection of the material, manuscript. JK: Planning of the research, analysis of the material, manuscript. Acta thanks Søren Overgaard for help with peer review of this study.

Abdel M P, Watts C D, Houdek M T, Lewallen D G, Berry D J. Epidemiology of periprosthetic fracture of the femur in 32 644 primary total hip arthroplasties: a 40-year experience. Bone Joint J 2016; 98-b(4): 461-7. Berend M E, Smith A, Meding J B, Ritter M A, Lynch T, Davis K. Long-term outcome and risk factors of proximal femoral fracture in uncemented and cemented total hip arthroplasty in 2551 hips. J Arthroplasty 2006; 21(6 Suppl 2): 53-9. Brady O H, Garbuz D S, Masri B A, Duncan C P. Classification of the hip. Orthop Clin North Am 1999; 30(2): 215-20. Brady O H, Garbuz D S, Masri B A, Duncan C P. The reliability and validity of the Vancouver classification of femoral fractures after hip replacement. J Arthroplasty 2000; 15(1): 59-62. Broden C, Mukka S, Muren O, Eisler T, Boden H, Stark A, Skoldenberg O. High risk of early periprosthetic fractures after primary hip arthroplasty in elderly patients using a cemented, tapered, polished stem. Acta Orthop 2015; 86(2): 169-74.

141

Catani F, Ensini A, Leardini A, Bragonzoni L, Toksvig-Larsen S, Giannini S. Migration of cemented stem and restrictor after total hip arthroplasty: a radiostereometry study of 25 patients with Lubinus SP II stem. J Arthroplasty 2005; 20(2): 244-9. Chatziagorou G, Lindahl H, Garellick G, Karrholm J. Incidence and demographics of 1751 surgically treated periprosthetic femoral fractures around a primary hip prosthesis. Hip Int 2018; Jul 1: doi: 10.1177/ 1120700018779558. [Epub ahead of print]. Cook R E, Jenkins P J, Walmsley P J, Patton J T, Robinson C M. Risk factors for periprosthetic fractures of the hip: a survivorship analysis. Clin Orthop Relat Res 2008; 466(7): 1652-6. Glyn-Jones S, Alfaro-Adrian J, Murray D W, Gill H S. The influence of surgical approach on cemented stem stability: an RSA study. Clin Orthop Relat Res 2006; 448: 87-91. Gore D R, Murray M P, Sepic S B, Gardner G M. Anterolateral compared to posterior approach in total hip arthroplasty: differences in component positioning, hip strength, and hip motion. Clin Orthop Relat Res 1982; 165: 180-7. Hailer N P, Garellick G, Karrholm J. Uncemented and cemented primary total hip arthroplasty in the Swedish Hip Arthroplasty Register. Acta Orthop 2010; 81(1): 34-41. Hank C, Schneider M, Achary C S, Smith L, Breusch S J. Anatomic stem design reduces risk of thin cement mantles in primary hip replacement. Arch Orthop Trauma Surg 2010; 130(1): 17-22. Jolles B M, Bogoch E R. Posterior versus lateral surgical approach for total hip arthroplasty in adults with osteoarthritis. Cochrane Database Syst Rev 2006; (3): Cd003828. Junnila M, Laaksonen I, Eskelinen A, Pulkkinen P, Havelin L I, Furnes O, Fenstad A M, Pedersen A B, Overgaard S, Karrholm J, Garellick G, Malchau H, Makela K T. Implant survival of the most common cemented total hip devices from the Nordic Arthroplasty Register Association database. Acta Orthop 2016; 87(6): 546-53. Katz J N, Wright E A, Polaris J J, Harris M B, Losina E. Prevalence and risk factors for periprosthetic fracture in older recipients of total hip replacement: a cohort study. BMC Musculoskelet Disord 2014; 15: 168. Karrholm J, Lindahl H, Malchau H, Mohaddes M, Nemes S, Rogmark C, Rolfson O. The Swedish Hip Arthroplasty Register 2017. Annual Report 2016. Lindahl H, Malchau H, Herberts P, Garellick G. Periprosthetic femoral fractures: classification and demographics of 1049 periprosthetic femoral fractures from the Swedish National Hip Arthroplasty Register. J Arthroplasty 2005; 20(7): 857-65. Lindgren V, Garellick G, Karrholm J, Wretenberg P. The type of surgical approach influences the risk of revision in total hip arthroplasty: a study from the Swedish Hip Arthroplasty Register of 90,662 total hip replacements with 3 different cemented prostheses. Acta Orthop 2012; 83(6): 559-65. Lindgren J V, Gordon M, Wretenberg P, Karrholm J, Garellick G. Validation of reoperations due to infection in the Swedish Hip Arthroplasty Register. BMC Musculoskelet Disord 2014; 15: 384. Lowenhielm G, Hansson L I, Karrholm J. Fracture of the lower extremity after total hip replacement. Arch Orthop Trauma Surg 1989; 108(3): 141-3. McGraw P, Kumar A. Periprosthetic fractures of the femur after total knee arthroplasty. J Orthop Traumatol 2010; 11(3): 135-41. Meek R M, Norwood T, Smith R, Brenkel I J, Howie CR. The risk of periprosthetic fracture after primary and revision total hip and knee replacement. J Bone Joint Surg Br 2011; 93(1): 96-101. Murray D W, Gulati A, Gill H S. Ten-year RSA-measured migration of the Exeter femoral stem. Bone Joint J 2013; 95-B(5): 605-8. Palan J, Smith M C, Gregg P, Mellon S, Kulkarni A, Tucker K, Blom A W, Murray D W, Pandit H. The influence of cemented femoral stem choice on the incidence of revision for periprosthetic fracture after primary total hip arthroplasty: an analysis of national joint registry data. Bone Joint J 2016; 98-B(10): 1347-54. Prins W, Meijer R, Kollen B J, Verheyen C C, Ettema H B. Excellent results with the cemented Lubinus SP II 130-mm femoral stem at 10 years of follow-up: 932 hips followed for 5–15 years. Acta Orthop 2014; 85(3): 276-9.


142

Ranstam J, Karrholm J, Pulkkinen P, Makela K, Espehaug B, Pedersen A B, Mehnert F, Furnes O. Statistical analysis of arthroplasty data, II: Guidelines. Acta Orthop 2011; 82(3): 258-67. Rayan F, Dodd M, Haddad F S. European validation of the Vancouver classification of periprosthetic proximal femoral fractures. J Bone Joint Surg Br 2008; 90(12): 1576-9. Rogmark C, Fenstad A M, Leonardsson O, Engesaeter L B, Karrholm J, Furnes O, Garellick G, Gjertsen J E. Posterior approach and uncemented stems increases the risk of reoperation after hemiarthroplasties in elderly hip fracture patients. Acta Orthop 2014; 85(1): 18-25. Sesselmann S, Hong Y, Schlemmer F, Wiendieck K, Soder S, Hussnaetter I, Muller L A, Forst R, Wierer T. Migration measurement of the cemented Lubinus SP II hip stem: a 10-year follow-up using radiostereometric analysis. Biomed Tech (Berl) 2017; 62(3): 271-8. SĂśderman P, Malchau H, Herberts P, Johnell O. Are the findings in the Swedish National Total Hip Arthroplasty Register valid? J Arthroplasty 2000; 15(7): 884-9.

Acta Orthopaedica 2019; 90 (2): 135â&#x20AC;&#x201C;142

Thien T M, Chatziagorou G, Garellick G, Furnes O, Havelin L I, Makela K, Overgaard S, Pedersen A, Eskelinen A, Pulkkinen P, Karrholm J. Periprosthetic femoral fracture within two years after total hip replacement: analysis of 437,629 operations in the Nordic Arthroplasty Register Association database. J Bone Joint Surg Am 2014; 96(19): e167. Vaughan P D, Singh P J, Teare R, Kucheria R, Singer G C. Femoral stem tip orientation and surgical approach in total hip arthroplasty. Hip Int 2007; 17(4): 212-17. Westerman R W, Whitehouse S L, Hubble M J W, Timperley A J, Howell J R, Wilson M J. The Exeter V40 cemented femoral component at a minimum 10-year follow-up. Bone Joint J 2018; 100-b(8): 1002-9. Witte D, Klimm M, Parsch D, Clarius M, Breusch S, Aldinger P R. Ten-year survival of the cemented MS-30 femoral stem: increased revision rate in male patients. Acta Orthop Belg 2009; 75(6): 767-75. Yamamoto Y, Turkiewicz A, Wingstrand H, Englund M. Fragility fractures in patients with rheumatoid arthritis and osteoarthritis compared with the general population. J Rheumatol 2015; 42(11): 2055-8.


Acta Orthopaedica 2019; 90 (2): 143–147

143

Varying but reduced use of postoperative mobilization restrictions after primary total hip arthroplasty in Nordic countries: a questionnaire-based study Kirill GROMOV 1,2, Anders TROELSEN 1, Maziar MODADDES 3,4, Ola ROLFSON 3,4, Ove FURNES 5,6, Geir HALLAN 5,6, Antti ESKELINEN 7,8, Perttu NEUVONEN 7,8, and Henrik HUSTED 1 1 Department

of Orthopaedic Surgery, Copenhagen University Hospital Hvidovre, Denmark; 2 Danish Hip Arthroplasty Registry; 3 Department of Orthopaedics, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden; 4 Swedish Hip Arthroplasty Register; 5 The Norwegian Arthroplasty Register, Department of Orthopaedic Surgery, Haukeland University Hospital, Bergen, Norway; 6 Department of Clinical Medicine, University of Bergen, Norway; 7 Coxa Hospital for Joint Replacement, and Faculty of Medicine and Life Sciences, University of Tampere, Tampere, Finland; 8 Finnish Hip Arthroplasty Registry Correspondence: kirgromov@gmail.com Submitted 2018-11-07. Accepted 2019-01-03.

Background and purpose — Mobilization has traditionally been restricted following total hip arthroplasty (THA) in an attempt to reduce the risk of dislocation and muscle detachment. However, recent studies have questioned the effect and rationale underlying such restrictions. We investigated the use of postoperative restrictions and possible differences in mobilization protocols following primary THA in Denmark (DK), Finland (FIN), Norway (NO), and Sweden (SWE). Patients and methods — All hospitals performing primary THA in the participating countries were identified from the latest national THA registry report. A questionnaire containing questions regarding standard surgical procedure, use of restrictions, and postoperative mobilization protocol was distributed to all hospitals through national representatives for each arthroplasty registry. Results — 83% to 94% (n = 167) of the 199 hospitals performing THA in DK, FIN, NO, and SWE returned correctly filled out questionnaires. A posterolateral approach was used by 77% of the hospitals. 92% of the hospitals had a standardized mobilization protocol. 50%, 41%, 19%, and 38% of the hospitals in DK, FIN, NO, and SWE, respectively, did not have any postoperative restrictions. If utilized, restrictions were applied for a median of 6 weeks. Two-thirds of all hospitals have changed their mobilization protocol within the last 5 years—all but 2 to a less restrictive protocol. Interpretation — Use of postoperative restrictions following primary THA differs between the Nordic countries, with 19% to 50% allowing mobilization without any restrictions. There has been a strong tendency towards less restrictive mobilization over the last 5 years.

Dislocation following primary total hip arthroplasty (THA) has been reported to occur in 1–10% of patients (Meek et al. 2006, Patel et al. 2007, Kotwal et al. 2009, Jørgensen et al. 2014). Both surgery- and patient-related factors have been shown to affect the risk for dislocation, including surgical approach, implant position, implant type, implant fixation, femoral head size, age, sex, comorbidities, and cognitive function (Jolles et al. 2002, Byström et al. 2003, Brooks 2013, Seagrave et al. 2017a, 2017b, Miller et al. 2018, Tsikandylakis et al. 2018). Movement restrictions and other hip precautions following THA have commonly been practiced to prevent dislocation and muscle detachment (Husted et al. 2014)—especially if a lateral transgluteal approach has been used. However, recent studies have questioned this rationale, as liberal postoperative mobilization protocols have been demonstrated not to increase the risk for dislocation (Peak et al. 2005, Restrepo et al. 2011, Gromov et al. 2015, Allen et al. 2018). This was confirmed by a recent systematic review, which concluded that a more liberal lifestyle restriction and precaution protocol did not increase the dislocation rates after THA (van der Weegen et al. 2015). Despite increasing evidence that postoperative restrictions may be unnecessary, a recent study from the UK showed that 97% of physiotherapists and occupational therapists routinely prescribed hip precautions (Smith and Sackley 2016). Later, a national survey from the Netherlands found that restrictions were recommended to between 69% and 100% of patients following primary THA depending on the surgical approach used (Peters et al. 2017). Little is known about the use of postoperative restrictions in the Nordic countries. Such knowledge of the utilization of postoperative restrictions would make it easier to compare and interpret studies from the different Nordic countries. This

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2019.1572291


144

Acta Orthopaedica 2019; 90 (2): 143–147

Table 1. Coverage (n/N) of hospitals and procedures Coverage hospitals Country n a N b (%) Denmark Finland Norway Sweden Total

24 29 27 33 42 58 74 79 167 199 (84)

Coverage procedures n c N d (%) 10,470 10,708 (98) 10,125 10,657 (95) 7,431 8,881 (83) 17,414 18,140 (96) 45,440 48,386 (94)

a

number of hospitals performing primary THA that returned correctly filled out questionnaires. b number of hospitals performing primary THA countrywide. c number of primary THA performed annually in the hospitals that returned correctly filled out questionnaires. d number of primary THA performed annually countrywide.

would also facilitate the investigation of complications and functional outcome following THA. Finally, with increasing evidence on limited benefits with postoperative restrictions, updated national guidelines are needed to reduce inequity in postoperative care following primary THA. This questionnaire-based study investigated the use of postoperative restrictions and describes differences in mobilization protocols following primary THA in Denmark (DK), Finland (FIN), Norway (NO), and Sweden (SWE).

Patients and methods We identified all hospitals in DK, FIN, NO, and SWE performing primary THA from the respective national arthroplasty register’s most recent annual report. A survey with questions regarding standard surgical procedure, use of restrictions, and postoperative mobilization was designed according to guidelines presented by Sprague et al. (2009). Besides asking if the hospitals used any mobilization restrictions at all following primary THA, we also asked about specific restrictions employed and specific aids given to the patients as a part of the standard mobilization protocol (Kornuijt et al. 2016, Lee et al. 2017). Authors KG, AT, and HH drafted the questionnaire, whereafter all other authors were asked to review it. Subsequently, we revised the questionnaire according to comments to increase clarity and face validity. The questionnaire was designed in English and is presented in Supplementary data. The questionnaire was distributed to all identified hospitals through national representatives for each participating arthroplasty registry by email or regular mail. The questionnaire was sent to the head of the arthroplasty department, who was asked to fill out the questionnaire on behalf of the department. If the head of the arthroplasty department was not identified, the questionnaire was sent to the head of the orthopedic department. Approximately 1 month after sending out the questionnaire, a letter or email was sent out to all nonrespondents with a reminder to complete and return the ques-

Table 2. Demographics of hospitals that returned the correctly filled out questionnaire

DK FIN NO SWE Total n = 24 n = 27 n = 42 n = 74 n = 167

Approach used Direct anterior – 1 5 1 7 Anterolateral 1 9 8 36 54 Direct lateral – 1 8 19 28 Posterolateral 23 27 31 48 129 Articulation Neutral 9 17 23 59 108 Highwall 15 10 19 15 59 Dual mobility 1 - - - 1 Standard protocol Yes 24 22 39 69 154 No – 5 3 5 13 Full weight-bearing) Yes 23 27 40 74 164 No 1 – 2 – 3 Restrictions Yes 12 10 27 28 77 No 12 11 8 28 59 Depending – 6 7 18 31 Aids Yes 10 18 20 52 100 No 14 9 22 22 67 Restriction changed within 5 years Yes 17 19 25 51 112 No 7 8 17 23 55 Supervised physiotherapy Yes 11 14 26 51 102 No 1 5 2 6 14 Individual 12 8 14 17 51 DK = Denmark; NO = Norway; FIN = Finland; SWE = Sweden

tionnaire. Descriptive statistics were applied using IBM SPSS Statistics v25 (IBM Corp, Armonk, NY, USA). Ethics, funding, and potential conflict of interest. No approval from the National Ethics Committee was necessary as this was a non-interventional observational study. No funding was received for this work. The authors declare no conflict of interest.

Results 29, 33, 58, and 79 hospitals performing primary THA were identified in DK, FIN, NO, and SWE, respectively. 24/29, 27/33, 42/58, and 74/79 of the hospitals in DK, FIN, NO, and SWE, respectively, returned complete questionnaires giving a coverage of 84% (167/199). The hospitals responding to the questionnaire included 94% (45,440/48,386) of all primary THAs performed in Nordic countries in 2017 (Table 1). A posterolateral approach was the most common surgical approach in all countries. DK was the only country using elevated-rim acetabular components more frequently than neutral acetabular components (Table 2). 92% of the hospitals had a


Acta Orthopaedica 2019; 90 (2): 143–147

Table 3. Restrictions utilized depending on the approach Restrictions

Direct Antero- Direct Postero anterior lateral lateral lateral n = 7 n = 54 n = 28 n = 129

145

Distribution of restrictions in hospitals that used restrictions (%) Do not bend the hip over 90° Do not cross your legs Do not internally rotate the hip

Yes No

2 5

28 14 86 26 14 43

Do not twist your upper body Do not roll or lie on the non-operated side Do not use a bathtub

standardized mobilization protocol and 98% allowed immediate full weight-bearing (Table 2). 12/24, 11/27, 8/42, and 28/74 of the hospitals in DK, FIN, NO, and SWE, respectively, did not have any postoperative restrictions. For 31 hospitals that used different restrictions depending on the approach, 28 used restrictions for a posterolateral approach, 11 used restrictions for an anterolateral approach and 7 used restrictions for a direct lateral approach. 98% of the hospitals allowed immediate full weight-bearing. For hospitals applying restrictions (n = 108), these were used for 2 weeks in 1%, for 4 weeks in 6%, for 6 weeks in 47%, and for a minimum of 12 weeks in 45% of the hospitals. Restrictions were used for a median of 6 weeks. As regards the approach used, 71% of the hospitals using a direct anterior approach did not use restrictions compared with 33% of hospitals using a posterolateral approach (Table 3). Avoiding bending the hip over 90 degrees and not crossing the legs were the most commonly employed restrictions (Figure 1). Aids (other than walking aids) were routinely used by 60% of the hospitals, with aids for putting on socks and an elevated toilet seat being the most common (Figure 2). 67% (112/167) of all hospitals had changed their mobilization protocol within the last 5 years—all but 2 to a less restrictive protocol.

Discussion In this questionnaire-based study, we found that one-third of all participating hospitals did not use any postoperative restrictions following primary THA, while one-fifth imposed restrictions depending on the surgical approach. Denmark was the most liberal country with half of the hospitals not using any restrictions while Norway was the most restrictive country with one-fifth of the hospitals not employing any restrictions. Very few previous studies have investigated the use of restrictions on a national level. Recently, Peters et al. (2017) performed a national survey investigating use of postoperative restrictions following primary THA in the Netherlands and found that restrictions were applied for between 69% and 100% of the patients depending on the surgical approach used. Based on our results, even the most conservative country in our study (NO) had a more liberal approach than the Netherlands. Our results also showed a more liberal approach compared with a survey from the UK, which found hip precautions to

0

10

20

30

40

50

60

70

80

90 100

Figure 1. Restrictions used in hospitals that used restrictions.

Distribution of aids in hospitals that used aids (%) Sock aid Elevated toilet seat Elevated chair Abduction pillow 0

10

20

30

40

50

60

70

80

90 100

Figure 2. Aids used in hospitals that utilized aids on discharge.

be routinely prescribed by 97% of health professionals (physiotherapists and occupational therapists) participating in the study (Smith and Sackley 2016). The median use of restrictions for 6 weeks reported in our study is in agreement with Peters et al. (2017), and Smith and Sackley (2016). We found that restrictions were most frequently used with a posterolateral approach and least with a direct anterior approach. This conforms with the survey results by Peters et al. (2017) reporting a 100% restriction use with the posterolateral approach compared with 69% with the direct anterior approach. This difference is most likely explained by higher dislocation rates for a posterolateral approach compared with the direct anterior approach reported by some authors (Hailer et al. 2012, Zijlstra et al. 2017, Miller et al. 2018). An important finding in our study is a strong trend towards a less restrictive mobilization protocol in recent years: twothirds of hospitals had changed their mobilization protocol to a less restrictive one in the last 5 years. This is supported by the emerging evidence that removal of postoperative restrictions does not seem to lead to an increased risk for dislocation following primary THA (Peak et al. 2005, Restrepo et al. 2011, Gromov et al. 2015, van der Weegen et al. 2015, Kornuijt et al. 2016, Allen et al. 2018). Furthermore, a recent study showed that while most patients can remember all of the restrictions recommended at 8 weeks after surgery only one-fifth adhere to all restrictions, suggesting that even if restrictions are prescribed most patients do not adhere to them (Lee et al. 2017). To our knowledge only 2 studies have found a correlation between removal of restrictions and increased risk of dislocation. In a registry-based study, Jørgensen et al. (2014) found


146

that departments that did not use restrictions had a higher dislocation rate compared with departments that applied restrictions. However, the study was not designed to analyze restrictions, and the difference in dislocation rate could potentially be explained by a number of factors other than the use of restrictions. Mikkelsen et al. (2014) found a statistically nonsignificant increase in the dislocation rate in patients who were mobilized without restrictions, while no difference was seen in patient-reported outcomes after 6 weeks. However, this study investigated only 365 patients and was not powered to investigate dislocations following THA. Conversely, several authors have suggested that a liberal mobilization protocol following primary THA may lead to earlier return to work and higher patient satisfaction (Peak et al. 2005, Barnsley et al. 2015, van der Weegen et al. 2015). Our study has several limitations. First, we asked about standard protocols in the participating hospitals; we do not know to what extent the individual surgeons adhered to these protocols. Second, while the response rate was excellent, 16% of the hospitals (accounting for 6% of annually performed THAs) did not respond, allowing for some degree of bias. There is, however, no obvious reason to think that the departments that did not reply systematically were less or more restrictive than the those that did so. Also, although we included a question regarding articulation size in our questionnaire, we did not investigate other factors that could influence the risk of dislocation such as the type of implant and fixation. These factors could affect the postoperative protocols (Kim et al. 2009, Seagrave et al. 2017a). In recent years there have been a trend towards increased use of larger femoral heads, as this has been suggested to reduce the risk for dislocation. A recent study from the Nordic Arthroplasty Registry Association found that the risk for revision due to dislocation was lower when comparing 32-mm heads to 28-mm heads, while no benefit with use of 36-mm heads over 32-mm heads was found (Tsikandylakis et al. 2018). We did not investigate the femoral head size used by the individual departments, and this could potentially affect the surgeons in regard to use of postoperative restrictions. Further, we did not investigate whether or not the use of restrictions differed depending on the diagnosis. While we have investigated the restrictions recommended by the hospitals, we do not know if the physiotherapists working with the patients after discharge adhered to those protocols. Finally, while we investigated the use of restrictions, we did not investigate the dislocation rates, patient-reported outcomes, or patient satisfaction in departments using restrictions or in those with a more liberal mobilization protocol. Thus, no conclusion can be drawn from this study on the potential association between the use of postoperative restrictions and dislocation rates following primary THA. In summary, we found that use of postoperative restrictions following primary THA differed between the Nordic countries with 19–50% allowing mobilization without any restrictions.

Acta Orthopaedica 2019; 90 (2): 143–147

There has been a strong tendency towards less restrictive mobilization over the last 5 years. Whether this trend has had any affect on the dislocation rates and whether the dislocation rates differ between the less and the more restrictive hospitals in Nordic countries is unknown. Supplementary data The questionnaire is available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/ 17453674.2019.1572291 KG and HH planned the study. KG, AT, and HH designed the questionnaire and collected the data from DK. MM, OR, OF, GH, AE, and PT distributed and collected the questionnaires in the respective countries. KG analyzed the data and drafted the manuscript. All authors reviewed the manuscript. Acta thanks Anil Peters and Stephan Vehmeijerfor help with peer review of this study.

Allen F C, Skinner D L, Harrison J, Stafford G H. The effect of precautions on early dislocations post total hip arthroplasty: a retrospective cohort study. Hip Int 2018; 28(5): 485-90. Barnsley L, Barnsley L, Page R. Are hip precautions necessary post total hip arthroplasty? A systematic review. Geriatr Orthop Surg Rehabil 2015; 6(3): 230-5. Brooks P J. Dislocation following total hip replacement. Bone Joint J 2013; 95-B(11_Suppl_A): 67-9. Byström S, Espehaug B, Furnes O, Havelin L, Norwegian Arthroplasty Register. Femoral head size is a risk factor for total hip luxation: a study of 42,987 primary hip arthroplasties from the Norwegian Arthroplasty Register. Acta Orthop Scand 2003; 74(5): 514-24. Gromov K, Troelsen A, Otte K S, Orsnes T, Ladelund S, Husted H. Removal of restrictions following primary THA with posterolateral approach does not increase the risk of early dislocation. Acta Orthop 2015; 86(4): 463-8. Hailer N P, Weiss R J, Stark A, Kärrholm J. The risk of revision due to dislocation after total hip arthroplasty depends on surgical approach, femoral head size, sex, and primary diagnosis. Acta Orthop 2012; 83(5): 442-8. Husted H, Gromov K, Malchau H, Freiberg A, Gebuhr P, Troelsen A. Traditions and myths in hip and knee arthroplasty: a narrative review. Acta Orthop 2014; 85(6): 548-55. Jolles B M, Zangger P, Leyvraz P-F. Factors predisposing to dislocation after primary total hip arthroplasty: a multivariate analysis. J Arthroplasty 2002; 17(3): 282-8. Jørgensen C C, Kjaersgaard-Andersen P, Solgaard S, Kehlet H. Hip dislocations after 2,734 elective unilateral fast-track total hip arthroplasties: incidence, circumstances and predisposing factors. Arch Orthop Trauma Surg 2014; 134(11): 1615-22. Kim Y-H, Choi Y, Kim J-S. Influence of patient-, design-, and surgery-related factors on rate of dislocation after primary cementless total hip arthroplasty. J Arthroplasty 2009; 24(8): 1258-63. Kornuijt A, Das D, Sijbesma T, van der Weegen W. The rate of dislocation is not increased when minimal precautions are used after total hip arthroplasty using the posterolateral approach. Bone Joint J 2016; 98-B(5): 58994. Kotwal R S, Ganapathi M, John A, Maheson M, Jones S A. Outcome of treatment for dislocation after primary total hip replacement. J Bone Joint Surg Br 2009; 91-B(3): 321-6. Lee G R H, Berstock J R, Whitehouse M R, Blom A W. Recall and patient perceptions of hip precautions 6 weeks after total hip arthroplasty. Acta Orthop 2017; 88(5): 496-9.


Acta Orthopaedica 2019; 90 (2): 143–147

Meek R M D, Allan D B, McPhillips G, Kerr L, Howie C R. Epidemiology of dislocation after total hip arthroplasty. Clin Orthop Relat Res 2006; 447: 9-18. Mikkelsen L R, Petersen M K, Søballe K, Mikkelsen S, Mechlenburg I. Does reduced movement restrictions and use of assistive devices affect rehabilitation outcome after total hip replacement? A non-randomized, controlled study. Eur J Phys Rehabil Med 2014; 50(4): 383-93. Miller L E, Gondusky J S, Kamath A F, Boettner F, Wright J, Bhattacharyya S. Influence of surgical approach on complication risk in primary total hip arthroplasty. Acta Orthop 2018; 89(3): 289-94. Patel P D, Potts A, Froimson M I. The dislocating hip arthroplasty: prevention and treatment. J Arthroplasty 2007; 22(4 Suppl 1): 86-90. Peak E L, Parvizi J, Ciminiello M, Purtill J J, Sharkey P F, Hozack W J, Rothman R H. The role of patient restrictions in reducing the prevalence of early dislocation following total hip arthroplasty: a randomized, prospective study. J Bone Joint Surg Am 2005; 87(2): 247-53. Peters A, Veldhuijzen A J H, Tijink M, Poolman R W, Huis In ’t Veld R M H A. Patient restrictions following total hip arthroplasty: a national survey. Acta Orthop Belg 2017; 83(1): 45-52. Restrepo C, Mortazavi S M J, Brothers J, Parvizi J, Rothman R H. Hip dislocation: are hip precautions necessary in anterior approaches? Clin Orthop Relat Res 2011; 469(2): 417-22.

147

Seagrave K G, Troelsen A, Madsen B G, Husted H, Kallemose T, Gromov K. Can surgeons reduce the risk for dislocation after primary total hip arthroplasty performed using the posterolateral approach? J Arthroplasty 2017a; 32(10): 3141-6. Seagrave K G, Troelsen A, Malchau H, Husted H, Gromov K. Acetabular cup position and risk of dislocation in primary total hip arthroplasty: a systematic review of the literature. Acta Orthop 2017b; 88(1): 10-17. Smith T O, Sackley C M. UK survey of occupational therapists’ and physiotherapists’ experiences and attitudes towards hip replacement precautions and equipment. BMC Musculoskelet Disord 2016; 17(1): 228. Sprague S, Quigley L, Bhandari M. Survey design in orthopaedic surgery: getting surgeons to respond. J Bone Joint Surg Am 2009; 91(Suppl 3): 27-34. Tsikandylakis G, Kärrholm J, Hailer N P, Eskelinen A, Mäkelä K T, Hallan G, Furnes O N, Pedersen A B, Overgaard S, Mohaddes M. No increase in survival for 36-mm versus 32-mm femoral heads in metal-on-polyethylene THA. Clin Orthop Relat Res 2018; 476(12): 2367-78. van der Weegen W, Kornuijt A, Das D. Do lifestyle restrictions and precautions prevent dislocation after total hip arthroplasty? A systematic review and meta-analysis of the literature. Clin Rehabil 2015; 30(4): 329-39. Zijlstra W P, De Hartog B, Van Steenbergen L N, Scheurs B W, Nelissen R G H H. Effect of femoral head size and surgical approach on risk of revision for dislocation after total hip arthroplasty. Acta Orthop 2017; 88(4): 395-401.


148

Acta Orthopaedica 2019; 90 (2): 148–152

An international comparison of THA patients, implants, techniques, and survivorship in Sweden, Australia, and the United States Elizabeth W PAXTON 1,3, Guy CAFRI 1, Szilard NEMES 2,3, Michelle LORIMER 5, Johan KÄRRHOLM 2,3,4, Henrik MALCHAU 2,3,4, Stephen E GRAVES 5, Robert S NAMBA 6, and Ola ROLFSON 2,3,4 1 Department

of Clinical Analysis, Surgical Outcomes and Analysis, Southern California Permanente Medical Group, San Diego, CA, USA; 2 Swedish Hip Arthroplasty Register, Gothenburg, Sweden; 3 Department of Orthopaedics, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden; 4 Sahlgrenska University Hospital, Gothenburg, Sweden; 5 Australian Orthopaedic Association National Joint Replacement Registry, Adelaide, Australia; 6 Southern California Permanente Medical Group, Irvine, CA, USA Correspondence: Liz.w.paxton@kp.org Submitted 2018-08-13. Accepted 2018-12-18.

Background and purpose — International comparisons of total hip arthroplasty (THA) practices and outcomes provide an opportunity to enhance the quality of care worldwide. We compared THA patients, implants, techniques, and survivorship in Sweden, Australia, and the United States. Patients and methods — Primary THAs due to osteoarthritis were identified using Swedish (n = 159,695), Australian (n = 279,693), and US registries (n = 69,641) (2003– 2015). We compared patients, practices, and implant usage across the countries using descriptive statistics. We evaluated time to all-cause revision using Kaplan–Meier survival curves. We assessed differences in countries’ THA survival using chi-square tests of survival probabilities. Results — Sweden had fewer comorbidities than the United States and Australia. Cement fixation was used predominantly in Sweden and cementless in the United States and Australia. The direct anterior approach was used more frequently in the United States and Australia. Smaller head sizes (≤ 32 mm vs. ≥ 36 mm) were used more often in Sweden than the United States and Australia. Metal-onhighly cross-linked polyethylene was used more frequently in the United States and Australia than in Sweden. Sweden’s 5- (97.8%) and 10-year THA survival (95.8%) was higher than the United States’ (5-year: 97.0%; 10-year: 95.2%) and Australia (5-year: 96.3%; 10-year: 93.5%). Interpretation — Patient characteristics, surgical techniques, and implants differed across the 3 countries, emphasizing the need to adjust for demographics, surgical techniques, and implants and the need for global standardized definitions to compare THA survivorship internationally.

Arthroplasty registries provide a mechanism for evaluating patient, surgical, and implant characteristics associated with revision surgery (Paxton et al. 2012, 2015, Khatod et al. 2014) and to identify clinical best practices for enhancing quality of care (Herberts and Malchau 1999, 2000, Graves 2010, Paxton et al, 2010, 2012). In addition, identification of variations between countries provides an opportunity to evaluate similarities and differences in practices and outcomes. Investigation of total hip arthroplasty (THA) variation between countries has been examined in Scandinavia (Havelin et al. 2011, Makela et al. 2014) but has been limited in other countries. However, with an increased focus on the need for worldwide evidence on THA implant performance, international collaborations have increased (Sedrakyan et al. 2011, 2014). Despite an increased focus on such collaborations, variations in US, Australian, and Swedish THA patients, practices, and outcomes have not been fully examined. Therefore, we investigated similarities and differences in patient characteristics, surgical techniques, implant selection, and implant survival rates in THA patients across the 3 countries to identify future areas of research based on the country comparisons.

Patients and methods Primary THAs due to osteoarthritis were identified using national and regional registries in Sweden (n = 159,695) (SHAR 2014), Australia (n = 279,693) (AOANJRR 2015), and the United States (Paxton et al. 2012) (n = 69,641) from 2003 to 2015. The capture rate of these registries exceeds 95% and loss to follow-up is less than 8% over the study period. Validation and quality control methods of these registries have been previously published (Soderman et al. 2000, Paxton et al. 2010, 2012, AOANJRR 2016). Bilateral procedures were included in the study. Hip resurfacing procedures were excluded.

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2019.1574395


Acta Orthopaedica 2019; 90 (2): 148–152

149

Percent of primary THA – Kaiser Permanente

Percent of primary THA – Sweden

Percent of primary THA – Australia

100

100

100

90 80 70

metal-on-conventional metal-on-HXLPE metal-on-metal ceramic-on-conventional ceramic-on-HXLPE ceramic-on-ceramic

metal-on-conventional metal-on-HXLPE metal-on-metal ceramic-on-conventional ceramic-on-HXLPE ceramic-on-ceramic

90 80 70

90 80 70

60

60

60

50

50

50

40

40

40

30

30

30

20

20

20

10

10

10

0 2003

2005

2007

2009

2011

2013

2015

Year

0 2003

2005

2007

2009

2011

2013

2015

Year

0 2003

metal-on-conventional metal-on-HXLPE metal-on-metal ceramic-on-conventional ceramic-on-HXLPE ceramic-on-ceramic

2005

2007

2009

2011

2013

2015

Year

Figure 1. THA implant bearing surfaces by country.

Patient characteristics (i.e., age, sex, BMI, ASA score), surgical techniques (i.e., surgical approach, type of cement fixation), implant types (i.e., bearing surface, femoral head size), and 5and 10-year implant survival were reported from the registries. Sweden’s BMI and ASA were available from 2008 to 2015. Australia’s BMI was available only for 2015 and their ASA scores from 2012 to 2015. For the US cohort, BMI and ASA were available for the entire time period. Tables with aggregate-level data were shared across countries. Descriptive statistics were used to compare and contrast patients, practices, and implant usage. Kaplan–Meier survival curves were used to evaluate time to allcause revision across the countries. Chi-square tests of survival probabilities were used to assess differences in THA survival probabilities between the countries at 5- and 10-year follow-up. 95% confidence intervals (CI) are also presented. Ethics, funding, and potential conflicts of interest Approval from the Institutional Review Board was obtained prior to the start of this study (#5488 approved on August 27, 2009) and from the Regional Ethical Review Board in Gothenburg (entry number 271-14 approved on April 7, 2014 with amendment T-609-17 approved on July 10, 2017). There is no funding. There are no conflicts of interest.

Results Incidence rates of primary THA The volume of primary THAs for OA increased each year in all 3 countries during the study period. The 2015 incidence rates of THA with an OA diagnosis were higher in Australia and Sweden then in the US cohort (Table 1, see Supplementary data). Patient characteristics THA sex was predominantly female in all 3 countries. However, Australia had a higher proportion of males

than the US and Swedish cohorts. The US cohort was younger than both Australian and Swedish cohorts. Sweden had the lowest proportion of obese patients and lowest ASA scores of the 3 cohorts (Table 2, see Supplementary data). Surgical techniques Cement fixation was used predominantly in Sweden while cementless fixation was used more frequently in the United States and Australia. The percentage of hybrid fixation was higher in Australia than for the US cohort and in Sweden. The posterior approach was the main surgical approach for all countries. However, the direct anterior approach used in Australia and the United States was not adopted in Sweden during the study period (Tables 3 and 4, see Supplementary data). Implant characteristics While metal-on-highly cross-linked polyethylene (HXLPE) was used more frequently in the United States and Australia, metal-on-conventional bearing surface was used more often in Sweden, especially during the early part of the period studied. Ceramic-on-ceramic was used more frequently in Australia and rarely in Sweden. Ceramic-on-HXLPE was used more frequently in the US cohort (Table 4, see Supplementary data). In all countries, metal-on-metal bearing surfaces decreased, in Sweden from a few hundred to zero. In all 3 countries, the use of ceramic-on-HXLPE increased during the study period (Figure 1). Femoral head size Femoral head size use differed across countries with Sweden using smaller head sizes (i.e., ≤ 28 mm and 32 mm) whereas the US and Australian cohorts had a greater proportion of 36 mm and larger head sizes (Table 4). The use of 32 mm femoral head size became more prominent in Sweden during the study period (Figure 2).


150

Acta Orthopaedica 2019; 90 (2): 148–152

Percent of primary THA – Kaiser Permanente

Percent of primary THA – Sweden

Percent of primary THA – Australia

100

100

100

90

head size 28 mm head size 32 mm head size ≥ 36 mm

head size 28 mm head size 32 mm head size ≥ 36 mm

90

90

80

80

80

70

70

70

60

60

60

50

50

50

40

40

40

30

30

30

20

20

20

10

10

10

0 2003

2005

2007

2009

2011

2013

2015

Year

0 2003

2005

2007

2009

2011

2013

2015

Year

0 2003

head size 28 mm head size 32 mm head size ≥ 36 mm

2005

2007

2009

2011

2013

2015

Year

Figure 2. THA implant head size by country.

Hospital annual volume Hospital volume was similar across US and Swedish cohorts. However, Australia had a higher percentage of low volume facilities compared with the United States and Sweden (Table 5, see Supplementary data). Outcomes THA survival at 5 years was higher in Sweden (97.8%, CI 97.8–97.9) than the US (97.0%, CI 96.7–97.2), and Australian (96.3%, CI 96.2–96.4) cohorts. The US cohort had a higher 5-year THA survivorship than the Australian. THA survival at 10 years was higher in Sweden (95.8%, CI 95.6–95.9) than for the US cohort (95.2%, CI 94.7–95.6) and in Australia (93.5%, CI 93.4–93.7). The US cohort had a higher 10-year THA survivorship than Australia (Table 6, Figure 3, see Supplementary data). The most frequent reasons for revisions differed across the countries. Aseptic loosening was higher in Sweden along with infection (Figure 4, see Supplementary data). Instability was a more common reason for revision in the US cohort than in Sweden and Australia.

Discussion This study provides the first comprehensive assessment of US, Swedish, and Australian THA practice patterns and outcomes, and identifies variation between countries in patient characteristics, fixation, implant characteristics, and THA implant survival. Patient characteristics First, the study highlights variation in the incidence of primary THA for OA with Sweden and Australia having a higher annual incidence rate than the US cohort. Our findings are consistent with other reports of variation in international total hip incidence rates (Merx et al. 2003). Differences in incidence rates

may reflect the younger population in the US health system, an actual variation in diagnoses leading to THA, differences in diagnostic accuracy and indications for surgical treatment, varying access to care in the different healthcare systems, or possibly variation in population demand. Second, patient characteristics appear to differ across countries with Sweden reporting lower BMI and ASA scores than Australia and the United States. This finding is consistent with other studies that have indicated a higher BMI in the US population (ProCon.org 2011). Differences in ASA scores could reflect a healthier population in Sweden but could also represent variations in coding practices across the countries. In addition to differences in BMI and comorbidities, cohorts also differed in age distribution with the US group having a younger THA population. This may reflect differences in thresholds for operating on younger patients or varying access to care across the different healthcare systems. The differences in patient characteristics emphasize the need to adjust for this variation when examining THA outcomes across countries. Surgical techniques Although the posterior approach was used most frequently in all 3 countries, Sweden did not utilize the direct anterior approach. Several systematic reviews and registry studies suggest that the anterior approach is associated with lower dislocation and revision rates (Barrett et al. 2013, Higgins et al. 2015, Sheth et al. 2015, Miller et al. 2018a, b). However, these studies focused on short-term and functional outcomes without any certain conclusions concerning longer term revision rates. Despite differences in surgical approaches, Sweden had lower THA revision rates than the US cohort and Australia. Fixation and implant characteristics Another identified difference was type of THA fixation used in the different countries. While Sweden used cement fixation more frequently, US and Australian practices were predominantly cementless. Studies examining fixation seem to suggest


Acta Orthopaedica 2019; 90 (2): 148â&#x20AC;&#x201C;152

an advantage for cement fixation, especially in older patients (SHAR 2009, Hailer et al. 2010, AOANJRR 2017, Phedy et al. 2017). Future studies comparing international variations should account for fixation types and implants to fully evaluate differences in THA implant survival. Type of THA bearing surface also differed across countries with Sweden adopting more metal-on-conventional polyethylene. Several large registry studies have identified a higher revision rate in metal-on-conventional bearing surface than metalon-HXLPE (AOANJRR 2013, Paxton et al. 2015). However, this difference may be prosthesis-dependent (Johanson et al. 2017). Swedenâ&#x20AC;&#x2122;s use of metal-on-conventional has decreased and the use of metal-on-HXLPE increased during the study period. Sweden also adopted less large head size metal-onmetal, which has been reported as having a higher risk of THA revision (AOANJRR 2008). The use of large head size metalon-metal decreased in both the United States and Australia. In all 3 countries, the use of ceramic-on-HXLPE bearing surface increased. While the Australian registry reports lower revision rates in ceramic-on-HXLPE compared with metal-onHXLPE, a recent US study indicated similar revision rates among these bearing surfaces but higher dislocation rates in ceramic versus metal femoral heads of < 32 mm, suggesting both head size and bearing surface material influence risk of revision (AOANJRR 2017, Cafri et al. 2017). Hospital volume Annual hospital volume also differed across countries. While hospital volumes were similar in Sweden and the US cohort, the Australian registry had a higher number of cases performed at low-volume hospitals. Lower hospital volume has been identified in relationship to higher complication rates and readmissions (Dy et al. 2014, Laucis et al. 2016, Sibley et al. 2017). Evaluating further the effects of hospital volume may identify potential areas of focus for quality improvement within healthcare systems. THA survival In evaluating THA survival, all 3 countries had 5- and 10-year THA implant survival estimates above 95%. 5- and 10-year implant survival was highest in Sweden and lowest in Australia. Differences in survival could possibly be related to different thresholds for revision THA surgery in those countries. Most likely, however, the difference THA survival is related to the degree of variation in implant selection between the countries. While Sweden and the US cohort used a limited number of implants during this timeframe, Australia had much greater variation in THA implant models. In Australia alone, over 2,000 cup and stem combinations were used. 78 different THA acetabular cups and stem model combinations have been used with 10-year follow-up and cumulative percentage revision rates ranging from 2% to 46%. Only 35% of these combinations had less than a 5% 10-year cumulative percentage revision (AOANJRR 2017). In comparison, Swedenâ&#x20AC;&#x2122;s 10-year

151

THA survival ranged from 94.4% to 98.1% based on a more restricted use of cup/stem combinations. In Sweden, 6 stems and 15 cups accounted for over 90% of the implant usage (SHAR 2016). This suggests that implant selection plays a key role in THA survival. The comparison of specific implant performance in similar patients with similar techniques must be evaluated to understand the underlying source of this international variation in THA survival. In addition to differences in revision rates, the reasons for revision also differed across countries. Sweden had a higher percentage of aseptic loosening than the US and Australian cohorts, which could be related to the higher percentage of metal-on-conventional polyethylene use in Sweden during the study period. The US cohort had a higher percentage of pain as the revision diagnosis. Revision due to pain and aseptic loosening combined in the US group was comparable to aseptic loosening diagnosis in both Sweden and Australia, suggesting aseptic loosening maybe the underlying diagnosis of pain in the US cohort. Revision diagnosis of infection was higher in Sweden than in the other countries. The US cohort had a higher rate of revision due to instability despite the use of larger femoral head sizes and the direct anterior approach. This most likely reflects the predominant use of uncemented cups in the US cohort, which has been reported to have more instability than cemented cups (Conroy et al. 2008). Differences in revision diagnoses may be related to the different underlying mechanisms of failure related to different implant usage and indications but could also be related to variation in surgeon documentation and definitions across registries, again emphasizing the need for standardized, global revision definitions to conduct international comparisons of THA outcomes. This study has both strengths and limitations. The strengths of this study include the large registry data sets with highquality data and minimal loss to follow-up. In addition, registries provide real-world data with high generalizability/external validity. Limitations include the descriptive nature of the study, which has not been adjusted for confounders, the observational study design limiting causality, and the use of only one US integrated healthcare system. In addition, although countries differed in patient, implant, and surgical factors, the difference in THA survival may be interpreted as not being clinically relevant due to the small differences. However, this study emphasizes there are differences in THA survival overall and further research needs to be conducted to evaluate THA outcomes by specific types of prostheses while controlling for patient, surgical, and hospital factors. In summary, patient characteristics, surgical techniques, and implant selection differs across the 3 countries, emphasizing the need to address regional and national differences in demographics, surgical techniques, implants, and the need for global standardized definitions to compare results across existing registries and to develop international THA benchmarking standards.


152

Supplementary data Tables 1–6 and Figures 3–4 are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/17453674.2019.1574395 EP: conception of study, interpretation of data and manuscript preparation. SN, JK, HM, SG, RS, OR: interpretation of data and manuscript preparation. ML, GC: statistical analyses, interpretation of data, and manuscript preparation. Acta thanks Antti Eskelinen and Sarunas Tarasevicius for help with peer review of this study.

AOANJRR. Australian Orthopaedic Association National Joint Replacement Register Annual Report. Adelaide, SA, Australia; 2008, 2013, 2015, 2016, and 2017. https://aoanjrr.sahmri.com/annual-reports-2017. Barrett W P, Turner S E, Leopold J P. Prospective randomized study of direct anterior vs postero-lateral approach for total hip arthroplasty. J Arthroplasty 2013; 28(9): 1634-8. Cafri G, Paxton E W, Love R, Bini S A, Kurtz S M. Is there a difference in revision risk between metal and ceramic heads on highly crosslinked polyethylene liners? Clin Orthop Relat Res 2017; 475(5): 1349-55. Conroy J L, Whitehouse S L, Graves S E, Pratt N L, Ryan P, Crawford R W. Risk factors for revision for early dislocation in total hip arthroplasty. J Arthroplasty 2008; 23(6): 867-72. Dy C J, Bozic K J, Pan T J, Wright T M, Padgett D E, Lyman S. Risk factors for early revision after total hip arthroplasty. Arthritis Care Res (Hoboken) 2014; 66(6): 907-15. Graves S E. the value of arthroplasty registry data. Acta Orthop 2010; 81(1): 8-9. Hailer N P, Garellick G, Karrholm J. Uncemented and cemented primary total hip arthroplasty in the Swedish Hip Arthroplasty Register. Acta Orthop 2010; 81(1): 34-41. Havelin L I, Robertsson O, Fenstad A M, Overgaard S, Garellick G, Furnes O. A Scandinavian experience of register collaboration: the Nordic Arthroplasty Register Association (NARA). J Bone Joint Surg Am 2011; 93(Suppl 3): 13-19. Herberts P, Malchau H. Many years of registration have improved the quality of hip arthroplasty. Läkartidningen 1999; 96(20): 2469-73, 75-6. Herberts P, Malchau H. Long-term registration has improved the quality of hip replacement: a review of the Swedish THR Register comparing 160,000 cases. Acta Orthop Scand 2000; 71(2): 111-21. Higgins B T, Barlow D R, Heagerty N E, Lin T J. Anterior vs. posterior approach for total hip arthroplasty, a systematic review and meta-analysis. J Arthroplasty 2015; 30(3): 419-34. Johanson P E, Furnes O, Ivar Havelin L, Fenstad A M, Pedersen A B, Overgaard S, Garellick G, Makela K, Karrholm J. Outcome in design-specific comparisons between highly crosslinked and conventional polyethylene in total hip arthroplasty. Acta Orthop 2017; 88(4): 363-9. Khatod M, Cafri G, Namba R S, Inacio M C, Paxton E W. Risk factors for total hip arthroplasty aseptic revision. J Arthroplasty 2014; 29(7): 1412-7. Laucis N C, Chowdhury M, Dasgupta A, Bhattacharyya T. Trend toward high-volume hospitals and the influence on complications in knee and hip arthroplasty. J Bone Joint Surg Am 2016; 98(9): 707-12.

Acta Orthopaedica 2019; 90 (2): 148–152

Makela K T, Matilainen M, Pulkkinen P, Fenstad A M, Havelin L I, Engesaeter L, Furnes O, Overgaard S, Pedersen A B, Karrholm J, Malchau H, Garellick G, Ranstam J, Eskelinen A. Countrywise results of total hip replacement: an analysis of 438,733 hips based on the Nordic Arthroplasty Register Association database. Acta Orthop 2014; 85(2): 107-16. Merx H, Dreinhofer K, Schrader P, Sturmer T, Puhl W, Gunther K P, Brenner H. International variation in hip replacement rates. Ann Rheum Dis 2003; 62(3): 222-6. Miller L E, Gondusky J S, Bhattacharyya S, Kamath A F, Boettner F, Wright J. Does surgical approach affect outcomes in total hip arthroplasty through 90 days of follow-up? A systematic review with meta-analysis. J Arthroplasty 2018a; 33(4): 1296-302. Miller L E, Gondusky J S, Kamath A F, Boettner F, Wright J, Bhattacharyya S. Influence of surgical approach on complication risk in primary total hip arthroplasty. Acta Orthop 2018b; 89(3): 289-94. Paxton E W, Inacio M C, Khatod M, Yue E J, Namba R S. Kaiser Permanente National Total Joint Replacement Registry: aligning operations with information technology. Clin Orthop Relat Res 2010; 468(10): 2646-63. Paxton E W, Inacio M C, Kiley M L. The Kaiser Permanente Implant Registries: effect on patient safety, quality improvement, cost effectiveness, and research opportunities. Perm J 2012; 16(2): 36-44. Paxton E W, Inacio M C, Khatod M, Yue E J, Funahashi T, Barber T. Risk calculators predict failures of knee and hip arthroplasties: findings from a large health maintenance organization. Clin Orthop Relat Res 2015; 473(12): 3965-73. Phedy P, Ismail H D, Hoo C, Djaja Y P. Total hip replacement: a meta-analysis to evaluate survival of cemented, cementless and hybrid implants. World J Orthop 2017; 8(2): 192-207. ProCon.org. US and global obesity levels: the fat chart. Santa Monica, CA; 2011. Sedrakyan A, Paxton E W, Phillips C, Namba R, Funahashi T, Barber T, Sculco T, Padgett D, Wright T, Marinac-Dabic D. The International Consortium of Orthopaedic Registries: overview and summary. J Bone Joint Surg Am 2011; 93(Suppl 3): 1-12. Sedrakyan A, Paxton E, Graves S, Love R, Marinac-Dabic D. National and international postmarket research and surveillance implementation: achievements of the International Consortium of Orthopaedic Registries initiative 2014. J Bone Joint Surg Am 2014; 96(Suppl): 1-6. SHAR. Swedish Hip Arthroplasty Register Annual Report 2009. https://shpr. registercentrum.se/shar-in-english/annual-reports-from-the-swedish-hiparthroplasty-register/p/rkeyyeElz. DOI: 10.18158/Hysl6bA0Z SHAR. Swedish Hip Arthroplasty Register Annual Report 2014. https://shpr. registercentrum.se/shar-in-english/annual-reports-from-the-swedish-hiparthroplasty-register/p/rkeyyeElz. DOI: 10.18158/B1OyzZ00Z SHAR. Swedish Hip Arthroplasty Register Annual Report 2016. https://shpr. registercentrum.se/shar-in-english/annual-reports-from-the-swedish-hiparthroplasty-register/p/rkeyyeElz. DOI: 10.18158/SJy6jKyrM Sheth D, Cafri G, Inacio M C, Paxton E W, Namba R S. Anterior and anterolateral approaches for THA are associated with lower dislocation risk without higher revision risk. Clin Orthop Relat Res 2015; 473(11): 3401-8. Sibley R A, Charubhumi V, Hutzler L H, Paoli A R, Bosco J A. Joint replacement volume positively correlates with improved hospital performance on Centers for Medicare and Medicaid Services quality metrics. J Arthroplasty 2017; 32(5): 1409-13. Soderman P, Malchau H, Herberts P, Johnell O. Are the findings in the Swedish National Total Hip Arthroplasty Register valid? A comparison between the Swedish National Total Hip Arthroplasty Register, the National Discharge Register, and the National Death Register. J Arthroplasty 2000; 15(7): 884-9.


Acta Orthopaedica 2019; 90 (2): 153–158

153

High annual surgeon volume reduces the risk of adverse events following primary total hip arthroplasty: a registry-based study of 12,100 cases in Western Sweden Per JOLBÄCK 1,2,3,5, Ola ROLFSON 1,3, Peter CNUDDE 1,3,4, Daniel ODIN 3, Henrik MALCHAU 1,3, Hans LINDAHL 1,2,3, and Maziar MOHADDES 1,3 1 Department

of Orthopaedics, Institute of Clinical Sciences, The Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden; 2 Department of Orthopaedics, Skaraborgs Hospital, Lidköping, Sweden; 3 Swedish Hip Arthroplasty Register, Gothenburg, Sweden; 4 Department of Orthopedics, Hywel Dda University Healthboard, Prince Philip Hospital, Bryngwynmawr, UK; 5 Research and Development Centre, Skaraborgs Hospital, Skövde, Sweden Correspondence: per.jolback@vgregion.se Submitted 2018-04-30. Accepted 2018-11-17.

Background and purpose — Most earlier publications investigating whether annual surgeon volume is associated with lower levels of adverse events (AE), reoperations, and mortality are based on patient cohorts from North America. There is also a lack of adjustment for important confounders in these studies. Therefore, we investigated whether higher annual surgeon volume is associated with a lower risk of adverse events and mortality within 90 days following primary total hip arthroplasty (THA). Patients and methods — We collected information on primary total hip arthroplasties (THA) performed between 2007 and 2016 from 10 hospitals in Western Sweden. These data were linked with the Swedish Hip Arthroplasty Register and a regional patient register. We used logistic regression (simple and multiple) adjusted for age, sex, comorbidities, BMI, fiation technique, diagnosis, surgical approach, time in practice as orthopedic specialist and annual volume. Annual surgeon volume was calculated as the number of primary THAs the operating surgeon had performed 365 days prior to the index THA. Results — 12,100 primary THAs, performed due to both primary and secondary osteoarthritis by 268 different surgeons, were identified. The median annual surgeon volume was 23 primary THAs (range 0–82) 365 days prior to the THA of interest and the mean risk of AE within 90 days was 7%. If the annual volume increased by 10 primary THAs in the simple logistic regression the risk of AE decreased by 10% and in the adjusted multiple regression the corresponding number was 8%. The mortality rate in the study was low (0.2%) and we could not find any association between 90-day mortality and annual surgeon volume. Interpretation — High annual surgical activity is associated with a reduced risk of adverse events within 90 days. Based on these findings healthcare providers should consider planning for increased surgeon volume.

In order to improve the outcomes after total hip arthroplasties (THA) and thereby reduce the burden of complications (Lawson et al. 2013), it is crucial to identify factors influencing adverse events (AE) associated with surgery. Earlier studies have shown that patient comorbidities, ASA classification, age, sex, BMI, and smoking increase the risk of complications and reoperations (Bozic et al. 2012, Lalmohamed et al. 2013, Arsoy et al. 2014, Duchman et al. 2015, Singh et al. 2015, Kallio et al. 2015, Bohl et al. 2016, Lubbeke et al. 2016). Procedure -related factors such as surgical approach, type of implant, fixation technique, and surgery time (Yasunaga et al. 2009, Lindgren et al. 2012) as well as hospital- and/or surgeon volume (Kreder et al. 1997, Solomon et al. 2002, Kaneko et al. 2014, Glassou et al. 2016, Kurtz et al. 2016, Laucis et al. 2016) are also suggested to influence outcomes after THA. The association between annual volume for both hospitals and individual surgeon and AE and reoperations have been discussed during the last decade, not only for primary THAs, but also in knee arthroplasty surgery (Kreder et al. 2003), vascular procedures (Pearce et al. 1999), general surgical procedures and gynecological interventions (Muilwijk et al. 2007). Few studies have investigated the relation between surgeon’s annual volume and outcomes (both medical and surgical complication, reoperations, mortality, and patient-reported outcomes) following primary THAs. Most of these studies report an association between a higher annual volume and fewer AE (Kreder et al. 1997, Lavernia and Guzman. 1995, Katz et al. 2001, 2003, Losina et al. 2004, Paterson et al. 2010, Camberlin et al. 2011, Ravi et al. 2014, Koltsov et al. 2018). All of these published reports are based on patient cohorts in North America with the exception of Camberlin et al. (2011) who studied a Belgian cohort of patients. There are, however, differences between countries with regards to training programs and level of surgeon activity. Second, there is a lack of publications adjusting for important confounders such as type of

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2018.1554418


154

Acta Orthopaedica 2019; 90 (2): 153–158

exceeds 98% during the last 10 years (Kärrholm et al. 2016). Patient data, age, sex, height, weight, ICD-10 diagnoses, fixation technique, surgical approach, and type of implant are registered in the Excluded (n = 2,986): SHAR. – reason for surgery not OA in SHAR, 120 Vega, a regional patient register, was initiated – other incision than posterior or direct lateral in SHAR, 140 – data on operating surgeon not available in local medial in 2000. It is an aggregated database, containing records, 29 records concerning all healthcare contacts (both – no information on volume 365 days prior to index THA, 1,756 public and privately funded) for all residents in – missing data on BMI in SHAR, 941 the region. In 2006 this regional patient register contained records of about 12 million healthcare THAs included in the analysis n = 12,100 contacts for the population of approximately 1.3 Figure 1. Flow chart. million people. Vega provides information to fixation, surgical approach, and time as orthopedic specialist. the National Patient Register (NPR). The PIN is used as the We evaluated possible associations between the surgeon’s unique identifier for all entries in Vega. The regional patient annual volume and the risk of AE and mortality within 90 days register contains details on: depiction of the caregiver at the following primary THA. We used data from a national qual- point of contact such as, for example, level of hospital or elecity register as well as hospital administrative data in Western tive care, diagnoses, and interventions such as, for example, Sweden, the second largest region in Sweden.  type of surgery, and length of stay in the hospital. Annual surgeon volume was defined as the number of primary THAs the operating surgeon performed in the 365 days prior to the index THA of interest (Ravi et al. 2014). Annual Patients and methods hospital volume was calculated as annual surgeon volume but Patient selection based on number of primary THAs in the 365 days prior to the Inclusion criteria for the study were: a primary THA either index THAs. with a cemented, uncemented, or hybrid fixation technique in A direct acyclic graph was used to visualize and determine patients with index diagnosis osteoarthritis (OA) of the hip covariates of interest based on previous publications. The foldefined by the International Classification of Diseases (ICD)- lowing covariates were identified as confounders and included 10 codes M16.0–M16.7 or M16.9. All patients underwent sur- in the multiple logistic regression analysis: age, sex, BMI, gery using a posterior or a direct lateral approach. We selected comorbidities, years in practice as orthopedic specialist at the all surgeries performed in all hospitals managed by the county time of the index THA, fixation technique, diagnostic indicacouncil of Western Sweden between 2007 and 2016 reported tion for implantation, surgical approach, and annual hospital to the Swedish Hip Arthroplasty Register (SHAR) and the volume. Smoking was also identified as a confounder but was regional patient register, Vega (Figure 1). not included in the multiple logistic regression analysis due to lack of information on patient smoking habits over the whole Sources of data investigated period (i.e., SHAR has not collected information Hospital medical records, SHAR, and the regional patient reg- during the entire investigated period). ister were used as data sources. The linking between hospital The years in practice for each orthopedic specialist at the time medical records and SHAR was done using the 10-digit per- of the index THA was defined as the difference between date sonal identity number (PIN) (Ludvigsson et al. 2009), name for surgery and date of certification as orthopedic specialist. The Elixhauser comorbidity index (ECI) is a comprehensive of the hospital, and date of surgery. The linked dataset, containing information from hospital medical records and SHAR, set of 30 comorbidities associated with substantial increase in was subsequently forwarded to the regional patient register length of stay, hospital charges, and mortality (Elixhauser et to add all adverse events and the data were pseudonymized al. 1998, van Walraven et al. 2009). The ECI has been considreplacing the PIN with a unique identifier. For each operat- ered as a superior predictor for long-term outcomes (beyond ing surgeon involved, data on the year for license to practice 30 days) to the Charlson comorbidity index (Sharabiani et al. and/or specialist certificate in orthopedics were obtained from 2012). The period used for calculating ECI in this study was publicly available data at the Swedish National Board of 365 days prior to the index THA. Comorbidities present in the Health and Welfare’s register of licensed healthcare profes- 365 days prior to the index THA were used for calculating ECI. sionals (HOSP). The variable sources are detailed in Table 1, An AE was defined as a readmission for a predefined set see Supplementary data. of World Health Organization International Classification of The SHAR’s aim is to register all primary THAs and reop- Diseases (ICD) and the Nordic Medico-Statistical Committee erations performed in Sweden. The coverage has been 100% (NOMESCO) Classification of Surgical Procedures codes for over the last 25 years and the completeness of primary THAs interventions (Appendix, see Supplementary data). Death for Primary THAs performed 2007–2016 extracted from hospital medical records n = 15,086


Acta Orthopaedica 2019; 90 (2): 153–158

155

Frequency

Frequency

800

400

600

300

400

200

200

100

0

0

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 39

0

10

20

30

40

50

60

70

80

Number of primary THAs 365 days prior to index THA

Years of experience

Figure 2. Distribution of the experience of the surgeon at the time of the index THA. Experience is computed as years between orthopedic specialist certification and surgery. Note: There are 2 THAs for year 39 and 1 for year 40, not visible in the graph.

Figure 3. Distribution of annual volume 365 days prior to the index THA.

any reason was also included in the definition of AE. The code list for AE has been elaborated by the Swedish Knee Arthroplasty Register (SKAR) in collaboration with the National Board of Health and Welfare to be used after knee replacements. Based on the same principles SHAR elaborated a code list adapted for elective hip replacements. The codes were classified into the following groups; A = surgical procedure codes that include reoperations of THA implants and other procedures that may represent a complication, DA = diagnostic codes that imply surgical complications, DB/DB 2 = diagnostic codes that cover hip-related diseases that may have been used for complications after THA surgery, DC = diagnostic codes covering cardiovascular events that may be related to the surgery, DM/DM 2 = diagnostic codes concerning other medical events not related to the THA surgery but that may be related to the surgery if they occur shortly afterwards. A, DA, BD, and BD 2 in the Appendix are surgical complications (i.e., hip-related complications) and DC, DM, and DM 2 medical complications (i.e., serious cardiovascular or medical complications).

determined limits (0, 10, 20, 30, 40, and 50) for annual surgeon volume in Table 4. A sensitivity analysis was performed according to guidelines in statistical analysis of arthroplasty data to evaluate the consequence of violating the assumption of independent observation (i.e., analysis when the second hip was excluded in patients with bilateral THAs) (Ranstam et al. 2011). Patients operated with simultaneous bilateral THAs were captured as 1 surgery in the study. As Ranstam et al. (2011) concluded based on a literature survey, there is little practical consequence of analyzing bilateral prostheses—at least with knee and hip data. We expect that the dependency structure of 2 hips from the same patients is stronger and of more consequence that the dependency structure of different patients having the same surgeon. THA surgery is a highly standardized procedure and as such we do not expect surgeon-related base risks and modelling approaches did not indicate such results. Our primary outcome was AE within 90 days following the index THA surgery and our secondary outcome was mortality within 90 days following the index THA surgery.

Statistics SPSS version 25 (IBM Corp, Armonk, NY, USA) and R version 3.2.3 (R Foundation for Statistical Computing, Vienna, Austria) was used for the statistical analysis. We used both simple and multiple logistic regression. Data from the regressions are presented with regression coefficient (β-coefficient), 95% confidence interval (CI), and p-value. P-value for statistical significance was set at < 0.05. A predictive model was created to analyze risk of AE and mortality. Predicted risk was calculated using a fitted simple logistic regression model. Prediction intervals (PI) were calculated to see the prediction strength with a 95% prediction interval. The predicted risk of AE within 90 days is presented unadjusted with arbitrarily

Ethics, funding, and potential conflicts of interests The study was approved by the Central Ethical Review Board in Stockholm (DNR Ö 9-2016). A research grant for the project was received from Skaraborgs Hospital research foundation. There is no conflict of interest.  

Results 268 different surgeons performed the 12,100 operations of which 8% (989) were performed by orthopedic trainees. The median years in practice as an orthopedic specialist at the time of the index THA was 12 (0–40) (Figure 2). The median


156

Acta Orthopaedica 2019; 90 (2): 153–158

Table 2. Patient characteristica and surgical data Age, years, mean (SD) All Male Female Sex, n (%) Male Female BMI, mean (SD) Diagnostic indication for implantation, n (%) Primary OA Secondary OA Fixation technique, n (%) Cemented Uncemented Hybrid Reverse hybrid Surgical approach, n (%) Posterior incision in lateral position (Moore) Lateral position (Gammer)

Table 3. Elixhauser comorbidity index 365 days prior to the index THA 69 (11) 68 (11) 70 (10) 5,101 (42) 6,999 (58) 28 (5) 11,414 (94) 686 (6) 8,820 (68) 2,330 (19) 620 (5) 930 (8) 4,201 (35) 7,899 (65)

annual surgeon volume was 23 primary THAs (0–82) 365 days prior to the THA of interest (Figure 3). Mean age for all patients was 69 years (SD 11) and the proportion of females was 58%. Primary OA was the most common diagnosis (94%). Some 68% of the patients received a cemented THA followed by uncemented (Table 2); 45% of the patients had no comorbidities according to ECI 365 days preceding the index surgery (Table 3). Outcomes Readmissions for any cause within 90 days occurred in 1,019 patients (8%) and with the AE definition used (see Appendix) the rate decreased to 818 (7%). In all, 69% of all AE could be classed as surgical complications and 31% as medical complications. For AE within 90 days the simple logistic regression showed a statistically significant reduced risk with increasing annual surgeon volume (regression coefficient = 0.990, CI 0.986–0.995). The corresponding numbers in the multiple regression were: regression coefficient = 0.992, CI 0.987–0.998. According to the predictive model the risk of an AE decreased by more than 35% if the surgeon had performed 50 or more THAs compared with 0 THAs during the 365 days preceding the index surgery (Table 4). A total of 28 patients died within 90 days. The annual surgeon volume did not influence the risk of mortality in the simple regression (regression coefficient = 0.999, CI 0.974– 1.022) or the multiple regression (regression coefficient = 1.000, CI 0.978–1.031). The prediction interval for mortality could not be calculated due to the low mortality rate. The result of the sensitivity analysis is similar to the result including both hips. 1,093 surgeries were excluded and the sensitivity analysis contained 11,007 THAs. 70 patients were operated with simultaneous bilateral THAs. Data for the sensitivity analysis are not shown.

Elixhauser comorbidity index 0 1 2 3 4 5 6 7 8 9 10 11

Total

n (%)

5,474 (45) 3,254 (27) 1,789 (15) 890 (7) 409 (3) 171 (1) 70 27 11 3 0 2

Table 4. Predicted risk of AE within 90 days for annual surgeon volume of primary THAs Annual Mean 95% prediction surgeon risk interval volume (%) (%) 0 10 20 30 40 50

8 7–10 8 6–9 7 5–9 6 5–8 6 4–7 5 4–7

12,100 (100)

Discussion We found that higher caseloads of annual THAs were associated with decreased level of AE within 90 days after surgery. This finding is supported by previous publications (Lavernia and Guzman 1995, Kreder et al. 1997, Katz et al. 2001, Losina et al. 2004, Paterson et al. 2010, Camberlin et al. 2011, Ravi et al. 2014). Based on previous publications it is difficult to understand what the optimal annual surgeon volume is in order to achieve low levels of AE and reoperation. Furthermore, annual surgeon volume can vary over time and by calculating the annual surgeon volume as the number of primary THAs performed 365 days prior to the index surgery we were able to capture this variation. This method has been used by Ravi et al. (2014) in their study and might be a more correct estimation than using all THAs during a calendar year where all surgeries are attributed with the same volume regardless of whether the actual surgery being analyzed is the first or the last one during the measured year. 90-day mortality is rare following primary THA surgery in Sweden. The 0.2% mortality rate in our study is lower than the average mortality rates following primary THAs in 2 published systemic reviews (0.7% and 0.5%) (Singh et al. 2011, Berstock et al. 2014). Berstock et al. (2014) included 7 studies on mortality within 90 days in their systemic review and in these the mortality rates varies between 0.1% and 0.7%. Ravi et al. (2014) (not included in any of the systemic reviews) did not find any obvious relation between mortality within 90 days and surgeon volume despite higher mortality rates. An explanation of the lower mortality rates in our study compared with the systemic reviews might be that the Swedish THA patients are healthier than patients included in systemic reviews (i.e., a selection bias of patients undergoing THA surgery between countries and hospitals). Hence, mortality rates between


Acta Orthopaedica 2019; 90 (2): 153–158

different studies might not be generalized depending on differences in the organization of healthcare and individual surgical practices. Our study has some limitations. We have not adjusted the multiple logistic regression for smoking, despite the knowledge of its negative influence on AE (Singh et al. 2015, Duchman et al. 2015). In our dataset, during the years 2013–2016, around 5% of patients were reported as smokers (information from the SHARs PROM program). Furthermore, data are missing on 17% included procedures, and finally the frequency of smoking is decreasing during the years 2013–2016. In spite of the fact that we have some information on smoking behavior in our study, we decided not to include smoking in the regression analysis because of the high amount of missing values. A second limitation is that only primary THAs performed within the region of Western Sweden were included. Some of the surgeons involved in the study might have had a temporary or partial employment, having performed primary THAs outside the investigated region. In Sweden there is no central dataset on surgeons, regarding their employment and activity. We presumed that the limited number of surgeons operating on cases outside the region of Western Sweden not would influence our conclusions. Finally, we share the same limitation as in all observational studies using administrative data. Both change of practice during the study period and local trends but also differences in registration might occur between the included hospitals. The regional patient register we used is not validated on its own but it provides data to the NPR. The Swedish National Inpatient Register (IPR) is part of the NPR. The IPR has been validated and contains 99% of all hospital discharges (Ludvigsson et al. 2011). In this study we used a definition of adverse events requiring hospital admission. Hence, we believe our data are robust and our conclusions are valid. One strength is that we could control for the surgeon’s experience (i.e., years as orthopedic specialist) at the time of the index surgery. The Swedish National Board of Health and Welfare register of licensed healthcare professionals has the exact date of certification for all doctors applying for licenses to practice and orthopedic specialist certification. We decided to include years in practice in the regression model. We have previously shown that surgeons with longer experience operate on patients with different diagnoses, patient characteristics, and using other implants compared with less experienced surgeons (Jolbäck et al. 2018). Years as a recognized specialist in orthopedics might also be considered as a proxy for surgical skills accumulated by the experience of previous procedures during the surgeon’s career. But, also, the knowledge gained and experience of preparing patients both physically and mentally prior to the surgery can be of importance. More experienced surgeons are likely to make more appropriate decisions regarding the indication for surgery, the operative details (technical aspects), and other perioperative factors that could result in an improved outcome. By including the years in practice at the time of the index surgery in the analysis we were

157

able to adjust for the above confounders. Another strength of the study is that we have been able to adjust for both surgical approach and type of fixation. To our knowledge, this is the first publication analyzing the risk of adverse events and mortality based on annual surgeon volume, adjusting for important confounders such as type of fixation, surgical approach, and time as orthopedic specialist. Finally, we used an administrative database registering all healthcare including readmission to hospitals in the whole of Sweden for the inhabitants of Western Sweden. This means that the risk of not collecting all readmissions within 90 days following the index THA is near to non-existent. Analyzing 12,100 surgeries reported to the SHAR, we conclude that high annual surgical activity is associated with a reduced risk of AE within 90 days following primary THAs. Based on these findings, healthcare providers should consider planning for an increased surgeon volume. Supplementary data Table 1 and the Appendix are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/ 17453674.2018.1554418

PJ had the original idea for the study, processed the data, and prepared the first version of the manuscript. All authors took part in the planning of the study, analysis, and interpretation of the data, and in writing of the manuscript. All authors read and approved the final manuscript. Acta thanks Sarah Whitehouse for help with peer review of this study.

Arsoy D, Woodcock J A, Lewallen D G, Trousdale R T. Outcomes and complications following total hip arthroplasty in the super-obese patient, BMI > 50. J Arthroplasty 2014; 29(10): 1899-905. Berstock J R, Beswick A D, Lenguerrand E, Whitehouse M R, Blom A W. Mortality after total hip replacement surgery: a systematic review. Bone Joint Res 2014; 3(6): 175-82. Bohl D D, Sershon R A, Fillingham Y A, Della Valle C J. Incidence, risk factors, and sources of sepsis following total joint arthroplasty. J Arthroplasty 2016; 31(12): 2875-9 e2. Bozic K J, Lau E, Kurtz S, Ong K, Rubash H, Vail T P, Berry D J. Patientrelated risk factors for periprosthetic joint infection and postoperative mortality following total hip arthroplasty in Medicare patients. J Bone Joint Surg Am 2012; 94(9): 794-800. Camberlin C, Vrijens F, De Gauquier K, Devriese S, Van De Sande S. Provider volume and short term complications after elective total hip replacement: an analysis of Belgian administrative data. Acta Orthop Belg 2011; 77(3): 311-19. Duchman K R, Gao Y, Pugely A J, Martin C T, Noiseux N O, Callaghan J J. The effect of smoking on short-term complications following total hip and knee arthroplasty. J Bone Joint Surg Am 2015; 97(13): 1049-58. Elixhauser A, Steiner C, Harris D R, Coffey R M. Comorbidity measures for use with administrative data. Med Care 1998; 36(1): 8-27. Glassou E N, Hansen T, Mäkelä K, Havelin L I, Furnes O, Badawy M, Kärrholm J, Garellick G, Eskelinen A, Pedersen A B. Association between hospital procedure volume and risk of revision after total hip arthroplasty: a population-based study within the Nordic Arthroplasty Register Association database. Osteoarthritis Cartilage 2016; 24(3): 419-26.


158

Jolbäck P, Rolfson O, Mohaddes M, Nemes S, Karrholm J, Garellick G, Lindahl H. Does surgeon experience affect patient-reported outcomes 1 year after primary total hip arthroplasty? Acta Orthop 2018; 89(3): 265-71. Kallio P J, Nolan J, Olsen A C, Breakwell S, Topp R, Pagel P S. Anesthesia preoperative clinic referral for elevated hba1c reduces complication rate in diabetic patients undergoing total joint arthroplasty. Anesth Pain Med 2015; 5(3): e24376. Kaneko T, Hirakawa K, Fushimi K. Relationship between peri-operative outcomes and hospital surgical volume of total hip arthroplasty in Japan. Health Policy 2014; 117(1): 48-53. Kärrholm J, Lindahl H, Malchau H, Mohaddes M Rogmark C, Rolfsson O. The Swedish Hip Arthroplasty Register Annual Report; 2016, 2017. Katz J N, Losina E, Barrett J, Phillips C B, Mahomed N N, Lew R A, Guadagnoli E, Harris W H, Poss R, Baron J A. Association between hospital and surgeon procedure volume and outcomes of total hip replacement in the United States Medicare population. J Bone Joint Surg Am 2001; 83-A(11): 1622-9. Katz J N, Phillips C B, Baron J A, Fossel A H, Mahomed N N, Barrett J, Lingard E A, Harris W H, Poss, R, Lew R A, Guadagnoli E, Wright E A, Losina E. Association of hospital and surgeon volume of total hip replacement with functional status and satisfaction three years following surgery. Arthritis Rheum 2003; 48(2): 560-8.[AQ3] Koltsov J C B, Marx R G, Bachner E, McLawhorn A S, Lyman S. Risk-based hospital and surgeon-volume categories for total hip arthroplasty. J Bone Joint Surg Am 2018; 100: 1203-8. Kreder H J, Deyo R, Koepsell T, Swiontkowski M F, Kreuter W. Relationship between the volume of total hip replacements performed by providers and the rates of postoperative complications in the state of Washington. J Bone Joint Surg Am 1997; 79(4): 485-94. Kreder H J, Grosso P, Williams J I, Jaglal S, Axcell T, Wal E K, Stephen D J. Provider volume and other predictors of outcome after total knee arthroplasty: a population study in Ontario. Can J Surg 2003; 46(1): 15-22. Kurtz S M, Lau E C, Ong K L, Adler E M, Kolisek F R, Manley M T. Hospital, patient, and clinical factors influence 30- and 90-day readmission after primary total hip arthroplasty. J Arthroplasty 2016; 31(10): 2130-8. Lalmohamed A, Vestergaard P, Javaid M K, de Boer A, Leufkens H G, van Staa T P, de Vries F. Risk of gastrointestinal bleeding in patients undergoing total hip or knee replacement compared with matched controls: a nationwide cohort study. Am J Gastroenterol 2013; 108(8): 1277-85. Laucis N C, Chowdhury M, Dasgupta A, Bhattacharyya T. Trend toward high-volume hospitals and the influence on complications in knee and hip arthroplasty. J Bone Joint Surg Am 2016; 98(9): 707-12. Lawson E H, Hall B L, Louie R, Ettner S L, Zingmond D S, Han L, Rapp M, Ko C Y. Association between occurrence of a postoperative complication and readmission: implications for quality improvement and cost savings. Ann Surg 2013; 258(1): 10-18. Lavernia C J, Guzman J F. Relationship of surgical volume to short-term mortality, morbidity, and hospital charges in arthroplasty. J Arthroplasty 1995; 10(2): 133-40. Lindgren V, Garellick G, Kärrholm J, Wretenberg P. The type of surgical approach influences the risk of revision in total hip arthroplasty: a study from the Swedish Hip Arthroplasty Register of 90,662 total hip replacements with 3 different cemented prostheses. Acta Orthop 2012; 83(6): 55965.

Acta Orthopaedica 2019; 90 (2): 153–158

Losina E, Barrett J, Mahomed N N, Baron J A, Katz J N. Early failures of total hip replacement: effect of surgeon volume. Arthritis Rheum 2004; 50(4): 1338-43. Lubbeke A, Zingg M, Vu D, Miozzari H H, Christofilopoulos P, Uckay I, Harbarth S, Hoffmeyer P. Body mass and weight thresholds for increased prosthetic joint infection rates after primary total joint arthroplasty. Acta Orthop 2016; 87(2): 132-8. Ludvigsson J F, Otterblad-Olausson P, Pettersson B U, Ekbom A. The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research. Eur J Epidemiol 2009; 24(11): 659-67. Ludvigsson J F, Andersson E, Ekbom E, Feychting M, Kim J-L, Reuterwall C, Heurgren M, Otterblad-Olausson P. External review and validation of the Swedish national inpatient register. BMC Public Health 2011; 11: 450. Muilwijk J, van den Hof S, Wille J C. Associations between surgical site infection risk and hospital operation volume and surgeon operation volume among hospitals in the Dutch nosocomial infection surveillance network. Infect Control Hosp Epidemiol 2007; 28(5): 557-63. Paterson J M, Williams J I, Kreder H J, Mahomed N N, Gunraj N, Wang X, Laupacis A. Provider volumes and early outcomes of primary total joint replacement in Ontario. Can J Surg 2010; 53(3): 175-83. Pearce W H, Parker M A, Feinglass J, Ujiki M, Manheim L M. The importance of surgeon volume and training in outcomes for vascular surgical procedure. J Vasc Surg 1999; 29(5): 768-76. Ranstam J, Kärrholm J, Pulkkinen P, Makela K, Espehaug B, Pedersen A B, Mehnert F, Furnes O, Nara Study Group. Statistical analysis of arthroplasty data, II: Guidelines. Acta Orthopaedica 2011; 82(3): 258-67. Ravi B, Jenkinson R, Austin P C, Croxford R, Wasserstein D, Escott B, Paterson J M, Kreder H, Hawker G A. Relation between surgeon volume and risk of complications after total hip arthroplasty: propensity score matched cohort study. BMJ 2014; 348: g3284. Sharabiani M T, Aylin P, Bottle A. Systematic review of comorbidity indices for administrative data. Med Care 2012; 50(12): 1109-18. Singh J A, Kundukulam J, Riddle D L, Strand V, Tugwell P. Early postoperative mortality following joint arthroplasty: a systematic review. J Rheumatol 2011; 38(7): 1507-13. Singh J A, Schleck C, Harmsen W S, Jacob A K, Warner D O, Lewallen D G. Current tobacco use is associated with higher rates of implant revision and deep infection after total hip or knee arthroplasty: a prospective cohort study. BMC Medicine 2015; 13: 283. Solomon D H, Losina E, Baron J A, Fossel A H, Guadagnoli E, Lingard E A, Miner A, Phillips C B, Katz J N. Contribution of hospital characteristics to the volume–outcome relationship: dislocation and infection following total hip replacement surgery. Arthritis Rheum 2002; 46(9): 2436-44. van Walraven C, Austin P, Jennings A, Quan H, Forster A J. A modification of the Elixhauser comorbidity measures into a point system for hospital death using administrative data. Med Care 2009; 47(6): 626-33. Yasunaga H, Tsuchiya K, Matsuyama Y, Ohe K. High-volume surgeons in regard to reductions in operating time, blood loss, and postoperative complications for total hip arthroplasty. J Orthop Sci 2009; 14(1): 3-9.


Acta Orthopaedica 2019; 90 (2): 159–164

159

Estonian hip fracture data from 2009 to 2017: high rates of non­ operative management and high 1-year mortality Pärt PROMMIK 1,2, Helgi KOLK 1,2, Pirja SARAP 1,2, Egon PUUORG 2, Eva HARAK 1,2, Andres KUKNER 2, Mati PÄÄSUKE 1, and Aare MÄRTSON 1,2 1 University of Tartu, 2 Tartu University Hospital, Estonia Correspondence: part.prommik@ut.ee Submitted 2018-09-05. Accepted 2018-12-08.

Background and purpose — There are no national guidelines for treatment of hip fractures in Estonia and no studies on management. We assessed treatment methods and mortality rates for hip fracture patients in Estonia. Patients and methods — We studied a populationbased retrospective cohort using validated data from the Estonian Health Insurance Fund’s database. The cohort included patients aged 50 and over with an index hip fracture diagnosis between January 1, 2009 and September 30, 2017. The study generated descriptive statistics of hip fracture management methods and calculated in-hospital, 1-, 3 , 6-, and 12-month unadjusted all-cause mortality rates. Results — 91% (number of hips: 11,628/12,731) of the original data were included after data validation. Median patient age was 81 years, 83 years for women and 74 years for men. 28% were men. Treatment methods were: total hip arthroplasty 7%; hemiarthroplasty 25%; screws 6%; sliding hip screw 25%; intramedullary nail 27%; and nonoperative management 10%. Unadjusted all-cause mortality rates for in-hospital, 1, 3, 6, and 12 months were: 3%, 9%, 18%, 24%, and 31% respectively. The 12-month mortality rate for nonoperative management was 58%. Interpretation — High rates of nonoperative management and overall high 1-year mortality rates after an index hip fracture indicate the need to review exclusion criteria for surgery and subacute care in Estonia.

There has been no change in mortality rates for hip fracture (HF) over the last 3 decades, regardless of advancements in surgical solutions and regardless of the fact that the majority of patients now are operated on (Mundi et al. 2014, Johansen et al. 2017). With the total number of HFs predicted to increase, the associated socioeconomic burden will become an even more challenging problem in the future (Gullberg et al. 1997, Cheung et al. 2018). The most appropriate healthcare strategies to address this challenging problem will be those based on the valid conclusions of high-quality research, which requires accurate data. Quality of administrative data can be improved with validation and this may lead to more accurate conclusions, which are needed for effective treatment guidelines. Currently there are no national guidelines for HF in Estonia and its management is unstudied. The management-specific outcomes may contribute towards the development of healthcare strategies and clinical practice. Therefore we assessed the relative prevalence of HF management methods in Estonia and calculated the mortality rates.

Patients and methods We conducted a population-based retrospective cohort study based on validated medical bill data from the Estonian Health Insurance Fund (EHIF), which insures 94% of the Estonian population. The data of HF patients in Estonia without EHIF insurance was also included. The cohort included those patients aged 50 and over with an index HF diagnosis between January 1, 2009 and September 30, 2017. HF diagnosis was based on the International Classification of Diseases (ICD-10) codes: S72.0—fracture of femoral neck; S72.1—pertrochanteric fracture; and S72.2—subtrochanteric fracture.

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2018.1562816


160

The following data were extracted for analysis: anonymized patient identification number (ID); age at hospitalization; sex; admission date; discharge date; fracture type; death date; date(s) of operation(s) within a year of hospitalization; and the Nordic Medico-Statistical Committee’s Classification of Surgical Procedures code (NCSP). Patients’ comorbidities (ICD-10 codes) were queried 4 years prior to index HF to calculate the Charlson comorbidity index (CCI) using updated weights and coding algorithms according to Quan et al. (2005, 2011). Only codes that appeared at least twice and at least 7 days apart were included to increase the method’s validity. Data on any HF diagnoses prior to the study period were also extracted. For patients without operative management data, personal identification codes and healthcare service codes used for orthopedic operations’ funding were extracted. Patients’ operative and mortality statuses were finalized as at December 31, 2017 and October 10, 2018, respectively. We used a multi-step strategy to validate the data. First, patients with a known prior HF were excluded. Second, a logic check was used and patients with both an HF diagnosis and an appropriate NCSP code within 3 months of diagnosis were included. The appropriate NCSP codes for operative management (OM) methods were: total hip arthroplasty, THA (NFB20, NFB30, NFB40, NFB99); hemiarthroplasty, HA (NFB00-9; NFB10-9); screws (NFJ70-3); sliding hip screw, SHS (NFJ603, NFJ80-3); intramedullary nail, IMN (NFJ50-3). Finally, the digital imaging and reports of those patients without operation information were reviewed, as well as the controlled operation codes used for funding. If this review did not show that a patient met the inclusion criteria, then the patient’s medical records were examined. For validation purposes, 2 national databases were used: Foundation of Estonian PACS (an image archiving and communication system database) and Electronic Health Record (e-Health Record). Digital imaging was reviewed by a radiologist (PS) and an orthopedic surgeon (EP). Fractures were initially classified by EHIF database and confirmed by an agreement as follows: the radiologist and report; the orthopedic surgeon and report; radiologist and orthopedic surgeon (no report). Medical records were reviewed by an orthogeriatrician (HK). For comparisons of sex proportions in age subgroups with the general population, the Statistics Estonia online database with summary-level data was used (http://pub.stat.ee/). However, only years 2016 and 2017 were included, since only these contained sufficiently detailed information for age subgroups. Statistics For continuous variables with normal distribution, mean and standard deviations (SD) are shown. For continuous variables with non-normal distribution, median (range) is shown. Categorical variables are shown as proportions. Multiple analyses were performed on the following variables: age, divided into 10-year subgroups; fracture type, grouped as intracapsular (S72.0) and extracapsular (S72.1 and S72.2); management method, grouped as nonoperative management (NOM) and

Acta Orthopaedica 2019; 90 (2): 159–164

OM; OM, grouped as THA, HA, screws, SHS, IMN; and temporal change, divided into two 4-year periods (2009–2012 and 2013–2016). In-hospital mortality was calculated only for stationary acute care patients. Patients hospitalized in 2017 were excluded from any analyses that required full-year data. Age and CCI were non-normally distributed by the Kolmogorov-Smirnov test (p < 0.001 for both) and were therefore analyzed using a Mann–Whitney U-test (Wilcoxon rank-sum test). Sex proportions within age subgroups were compared with those of the general population using a binominal test. The Mantel–Haenszel test for trend (Linear-by-Linear Association test in IBM SPSS Statistics software; IBM Corp, Armonk, NY, USA) was used to find linear associations. The Pearson chi-square test was used for proportional comparisons. Kaplan–Meier unadjusted cumulative all-cause mortality analyses were conducted at the end of hospitalization and at 1, 3, 6, and 12 months afterwards. The log-rank test was used to compare cumulative mortality between groups. Cox proportional hazard regression analysis was used to estimate age, sex, and CCI adjusted differences in mortality risk between groups. Hazard ratios are presented with 95% confidence intervals (CI). Statistical significance was defined as p < 0.05 and all tests were 2-sided. All statistical analyses were conducted using statistical software IBM SPSS Statistics for Windows, version 25 (IBM Corp, Armonk, NY, USA). Ethics, registration, funding, and potential conflicts of interest The study was approved by the Research Ethics Committee of the University of Tartu on June 17, 2013 (reference 227/T12). Additional approval from the Estonian Data Protection Inspectorate for the use of personalized data was received on December 1, 2017 (reference 2.2.-1/17/47). This work was supported by the following projects: Interreg Baltic Sea Region Programme 2014-2020 (grant number: #R001); Estonian Science Agency project IUT20-46 (TARBS14046I); HypOrth Project funded by the European Union’s 7th Framework Programme grant agreement no. 602398; Institutional Research Funding IUT20-58 of the Estonian Ministry of Education and Research. No conflicts of interest were declared.

Results Data validation After data validation 91% (11,628/12,731) of the original population data were included (Figure 1). Almost all patients (99%; 11,500/11,628), had health insurance. In the period between 2009 and 2016, a mean of 1,328 (SD 66) patients per year received an HF diagnosis. Patients and management Median patient age was 81 years (50–104). The proportion of men was 28% (3,287/11,628) and they were 9 years


Acta Orthopaedica 2019; 90 (2): 159–164

Figure 1. Data validation process. HF= hip fracture; NCSP = the Nordic MedicoStatistical Committee’s Classification of Surgical Procedures code.

Appropriate NCSP code available indicating HF diagnosis and operative management n = 10,116

161

HF on medical bill January 1, 2009 to September 30, 2017 n = 12,731

No NCSP code available. Digital images and healthcare service codes were reviewed n = 1,748

Excluded: known HF before study period n = 656

No NCSP code, digital images and healthcare service codes were available. Medical records were reviewed n = 211

Excluded: diagnosis other than HF n = 298

Excluded (n = 149): – diagnosis other than HF, 124 – no records found, 25

Total number of index HF patients included n = 11,628

Baseline characteristics of patients presented as total, operative (OM), and nonopera­ tive management (NOM) and statistical difference between the last two. Values are n (%) unless otherwise specified

Total OM NOM n = 11,628 n = 10,431 n = 1,197

p-value

Men 3,287 (28) 2,912 (28) 375 (31) 0.01 Median age (range) 81 (50–104) 81 (50–102) 82 (50–104) < 0.001 Age subgroups: < 0.001 50–59 804 (6.9) 720 (6.9) 84 (7.0) 60–69 1,422 (12) 1,298 (12) 124 (10) 70–79 2,984 (26) 2,702 (26) 282 (24) 80–89 4,913 (42) 4,442 (43) 471 (39) ≥ 90 1,505 (13) 1,269 (12) 236 (20) Fracture type: < 0.001 Femoral neck 5,988 (52) 5,145 (49) 843 (70) Pertrochanteric 4,967 (43) 4,663 (45) 304 (25) Subtrochanteric 673 (5.8) 623 (6.0) 50 (4.2) CCI, mean (SD) 1.7 (1.7) 1.6 (1.6) 2.0 (1.8) < 0.001 Comorbidities: Myocardial infarction 809 (7.0) 725 (7.0) 84 (7.0) 0.9 Congestive heart failure 5,097 (44) 4,520 (43) 577 (48) 0.001 Peripheral vascular disease 1,219 (10) 1,060 (10) 159 (13) 0.001 Cerebrovascular disease 2,504 (22) 2,225 (21) 279 (23) 0.1 Dementia 1,121 (9.6) 913 (8.8) 208 (17) < 0.001 Chronic pulmonary disease 1,259 (11) 1,117 (11) 142 (12) 0.2 Rheumatic disease 388 (3.3) 349 (3.3) 39 (3.3) 0.9 Peptic ulcer disease 550 (4.7) 488 (4.7) 62 (5.2) 0.4 Mild liver disease 175 (1.5) 154 (1.5) 21 (1.8) 0.4 Diabetes without chronic complication 1,264 (11) 1,132 (11) 132 (11) 0.8 with chronic complication 688 (5.9) 619 (5.9) 69 (5.8) 0.8 Hemi- or paraplegia 536 (4.6) 475 (4.6) 61 (5.1) 0.4 Renal disease moderate/severe 473 (4.1) 414 (4.0) 59 (4.9) 0.1 Any malignancy 1,193 (10) 1,042 (10) 151 (13) 0.005 Moderate/severe liver disease 36 (0.31) 27 (0.26) 9 (0.75) 0.004 Metastatic solid tumor 42 (0.36) 38 (0.36) 4 (0.33) 0.9 AIDS/HIV 1 (0.01) 1 (0.01) 0 (0) 0.7

tion of men in each age subgroup differed from that of the general population as follows: 17% higher for 50–59 years; 11% higher for 60–69 years; 3.8% lower for 70–79 years; 8.1% lower for 80–89 years; and 4.9% lower for 90 years and over (p < 0.001 for all) (Figure 2). 90% (10,431/11,628) of patients received operative management. Operative methods were: THA 7.4% (861/11,628); HA 25% (2,949/11,628); screws 5.8% (679/11,628); SHS 25% (2,856/11,628); IMN 27% (3,086/11,628). The operation date was available for 99% (10,372/10,431) of operated patients: 72% (7,461/10,431) were operated on within the first 2 days of hospitalization. Temporal changes in management methods of note (> 1%) between 2009–2012 and 2013–2016 for intracapsular fractures were: 3.9% increase for THA; 1.1% increase for HA; 1.7% decrease for screws; 3.9% decrease for SHS (p < 0.001). The same estimates for extracapsular fractures were: 29% decrease for SHS; 28% increase for IMN (p < 0.001) (Figure 3). In comparison with OM, NOM patients had higher median age; a higher proportion of patients aged 50–59 or ≥ 90 years; a higher proportion of femoral neck fracture; more comorbidities (Table).

CCI = Charlson Comorbidity Index

younger (Table). Men were in the majority in the 2 youngest age subgroups. There was a statistically significant linear trend for the proportion of men to decrease as age increased; the reverse was found for women (p < 0.001). The propor-

Mortality In-hospital, 1-, 3-, 6-, and 12-month unadjusted all-cause mortality rates were: 3.2%; 8.6%; 18%; 24%; and 31%, respectively. Unadjusted all-cause cumulative 1-year mortality rates


162

Acta Orthopaedica 2019; 90 (2): 159–164

Male sex (%)

Intracapsular fracture type

100

2009 2010 2011 2012 2013 2014 2015 2016

Study sample General population

80

60

Extracapsular fracture type 2009 2010 2011 2012 2013 2014 2015 2016

40

20

0

50–59

60–69

70–79

80–89

≥90

Age groups Figure 2. Relative proportion of men by age as compared with the general population. *p < 0.001 difference between populations’ age subgroups.

0

20

40

60

80

100

Distribution of management methods (%) Figure 3. Distribution of management methods by study year and fracture type. THA = total hip arthroplasty, HA = hemiarthroplasty, SHS = sliding hip screw, IMN = intramedullary nail, NOM = nonoperative management.

for women and men were similar (31%). However, age and comorbidity adjusted analysis showed the 1-year mortality risk for men to be 1.6 times higher (HR 1.6 [CI 1.5–1.7]). Overall mortality rate for the combined operative management methods was 28%, as compared with the NOM rate of 58% (p < 0.001). When adjusted for age, sex, and CCI, the 1-year mortality risk for NOM was 2.6 times higher than for operative management (HR 2.6 [CI 2.4–.9]) (Figure 4).

Discussion Distribution of operative management methods and time trends were similar to those of other studies (Gjertsen et al. 2017, Johansen et al. 2017). However, the NOM rate was unexpectedly high, being 1.6–10 times higher than NOM rates reported in other general population studies (Neuman et al. 2010 [USA], Cram et al. 2017 [Canada], Johansen et al. 2017 [Sweden, England, Wales, Northern Ireland, Scotland, Ireland, New Zealand, Australia]). The rate is comparable to that reported for nursing home residents (Berry et al. 2009, Neuman et al. 2014). Furthermore, the 12-month mortality rates for NOM patients were 10-17% higher than those of comparable studies (Cram et al. 2017, Ree et al. 2017). Multiple factors are generally associated with NOM: older age; male sex; more comorbidities; residence in a rural area; femoral neck fractures; residence at baseline long-term care; lower income; and black race (Neuman et al. 2010, Cram et al. 2017). Our findings were consistent with these factors: NOM patients were older; had more comorbidities; and included a higher proportion of men and femoral neck fractures. The study data did not provide information on race, residence, or income. The relatively high NOM rate in Estonia may also be attributable to country-specific factors: traditions and expectations of patients and family; absence of national guidelines;

Figure 4. Cox survival curves adjusted for age, sex, and Charlson comorbidity index score. For abbreviations, see Figure 3 caption.

and differences in short- and long-term healthcare support for OM and NOM patients. Mortality rates for in-hospital and at 1 month were consistent with those of earlier studies (Medin et al. 2015, Johansen et al. 2017). However, mortality rates at 3, 6, and 12 months were higher. For example, the Estonian 3-month mortality rate is comparable to the highest reported rates of a systematic review of 63 studies (Abrahamsen et al. 2009). Multiple studies, including a systematic review, have reported lower mortality rates at 1 year than the Estonian rate at 6 months (Kurtinaitis et al. 2012, Diamantopoulos et al. 2013, Brozek et al. 2014, Klop et al. 2014, Mundi et al. 2014, Poenaru et al. 2014). However, relatively similar 12-month mortality rates have been reported in Denmark, Hungary, and Scotland (Medin et al. 2015, Jantzen et al. 2018). Delayed surgery may also contribute to the high mortality rates; however, our result is similar to the findings of a recent study (Johansen et al. 2017). Relatively high mortality rates from the 3rd month onwards may be attributable in part to high rates of NOM that showed a higher 1-year mortality risk compared with every operative management type. Also, the crude 1-year mortality rate for operatively managed patients was lower than that of the overall study population. Differences in mortality from the third month onwards may be related to shortcomings in subacute care, such as accessibility of rehabilitation. This is supported by a study that compared the Estonian HF group with a nonfracture reference group using age and comorbidity adjusted relative risk ratios. The relative risk ratios are higher, especially for HF women at 3 and 12 months in Estonia, compared with the findings of a systematic review (Haentjens et al. 2010, Jürisson et al. 2017). The same estimates for men are near the upper confidence limits reported in the review article. On the other hand, previous studies have shown various preoperative indicators to be associated with increased HF mortality


Acta Orthopaedica 2019; 90 (2): 159–164

risk: advanced age; male sex; pre-fracture functional status; residence in an institutional care home; presence of an intracapsular fracture; cognitive impairment; depression; and more comorbidities (Hu et al. 2012, Liu et al. 2017). We found similar unadjusted mortality rates for men and women. However, the median age of men was lower, and ageadjusted analyses showed a higher mortality risk for men. These results are consistent with those of other studies (Abrahamsen et al. 2009, Kannegaard et al. 2010). The median age of men at the time of HF was relatively low. Other studies have reported a similar median age for female HF, but with a smaller age difference between the sexes (Kannegaard et al. 2010, Kurtinaitis et al. 2012, Diamantopoulos et al. 2013, Klop et al. 2014). The Estonian HF population has 5% more patients below the age of 80 years compared with an Austrian study by Brozek et al. (2014). There also was a difference in the proportion of sexes in age subgroups. In Estonia, men were highly prevalent in each of the 2 youngest age subgroups, whereas in Austria 10% fewer men were in the 50–59 and 60–69 age subgroups and more men were older (Brozek et al. 2014). The prevalence of men was also higher in the 2 youngest age subgroups, as compared with the general population. Jürisson et al. (2015) proposed that the incidence of Estonian male HF may be explained by relatively high rates of alcohol consumption, with a consequently greater risk of alcohol-related falls and injuries. The problem of flawed data in databases has been reported in previous studies on HF populations (Cundall-Curry et al. 2016). They suggested that data should not be used from administrative databases without validation, because conclusions based on inaccurate data may be erroneous and may misinform clinical practice and policy development. Our study followed a novel strategy of multi-step data validation to improve the quality of data extracted from a large administrative database. This strategy enabled a relatively high proportion of unsuitable cases (8.7%) to be excluded from analysis. Our study has multiple strengths thT increase the generalizability of the results: validated high-quality, whole-population data; unbiased and standardized data collection; and up-to-date data, with a long (9-year) study period. However, some limitations must be acknowledged. First, the data validation process may not have yielded the same level of accuracy as would the review of individual patient data case-by-case; however, individualized review is time-consuming and would involve processing an unnecessary amount of personalized data. In contrast, our study’s logical data validation process enabled the reduction of personal data use to just 14%. Second, the EHIF database does not contain information on patients who pay for their own care. The number of these patients in Estonia would be negligible, however, since emergency medical care is guaranteed and all Estonian citizens have health insurance on retirement. Third, the EHIF database dates back only to 2004. Study data may therefore have included patients with secondary HF. Secondary HF is associated with increased risk

163

of death and may therefore have affected the results (Sobolev et al. 2015). Finally, the study data did not provide information on patients’ residence and lifestyle factors, which could have informed some of the issues raised in the discussion. In summary, this is the first study to report managementspecific outcomes for HF in Estonia. The study identified several issues that merit further attention in clinical practice and research. Clinical practice should be reviewed with an aim to lower NOM and the 1-year mortality rate. Further research is already underway on the NOM decision-making process and on the long-term use of rehabilitation after HF. Contributions of the authors were as follows: PP, HK, MP, and AM designed the study; PP, PS, EP, and HK validated data; PP and EH performed data analysis and wrote the first draft; HK, AK, MP, and AM jointly revised the manuscript to its final form. Acta thanks Jan-Erik Gjertsen and Cecilia Rogmark for help with peer review of this study.

Abrahamsen B, Staa T van, Ariely R, Olson M, Cooper C. Excess mortality following hip fracture: a systematic epidemiological review. Osteoporos Int 2009; 20(10): 1633-50. Berry S D, Samelson E J, Bordes M, Broe K, Kiel D P. Survival of aged nursing home residents with hip fracture. J Gerontol A Biol Sci Med Sci 2009; 64A(7): 771-7. Brozek W, Reichardt B, Kimberger O, Zwerina J, Dimai H P, Kritsch D, Klaushofer K, Zwettler E. Mortality after hip fracture in Austria 2008– 2011. Calcif Tissue Int 2014; 95(3): 257-66. Cheung C-L, Ang S B, Chadha M, Chow E S-L, Chung Y-S, Hew F L, Jaisamrarn U, Ng H, Takeuchi Y, Wu C-H, Xia W, Yu J, Fujiwara S. An updated hip fracture projection in Asia: the Asian Federation of Osteoporosis Societies study. Osteoporos Sarcopenia 2018; 4(1): 16-21. Cram P, Yan L, Bohm E, Kuzyk P, Lix L M, Morin S N, Majumdar S R, Leslie W D. Trends in operative and nonoperative hip fracture management 19902014: a longitudinal analysis of Manitoba administrative data. J Am Geriatr Soc 2017; 65(1): 27-34. Cundall-Curry D J, Lawrence J E, Fountain D M, Gooding C R. Data errors in the National Hip Fracture Database. Bone Joint J 2016; 98-B(10): 1406-9. Diamantopoulos A P, Hoff M, Skoie I M, Hochberg M, Haugeberg G. Shortand long-term mortality in males and females with fragility hip fracture in Norway: a population-based study. Clin Interv Aging 2013; 8: 817-23. Gjertsen J-E, Dybvik E, Furnes O, Fevang J M, Havelin L I, Matre K, Engesæter L B. Improved outcome after hip fracture surgery in Norway. Acta Orthop 2017; 88(5): 505-11. Gullberg B, Johnell O, Kanis J A. World-wide projections for hip fracture. Osteoporos Int J Establ Result Coop Eur Found Osteoporos Natl Osteoporos Found USA 1997; 7(5): 407-13. Haentjens P, Magaziner J, Colon-Emeric C, Vanderschueren D, Milisen K, Velkeniers B, Boonen S. Meta-analysis: excess mortality after hip fracture among older women and men. Ann Intern Med 2010; 152(6): 380-90. Hu F, Jiang C, Shen J, Tang P, Wang Y. Preoperative predictors for mortality following hip fracture surgery: a systematic review and meta-analysis. Injury 2012; 43(6): 676-85. Jantzen C, Madsen C M, Lauritzen J B, Jørgensen H L. Temporal trends in hip fracture incidence, mortality, and morbidity in Denmark from 1999 to 2012. Acta Orthop 2018; 89(2): 170-6. Johansen A, Golding D, Brent L, Close J, Gjertsen J-E, Holt G, Hommel A, Pedersen A B, Röck N D, Thorngren K-G. Using national hip fracture registries and audit databases to develop an international perspective. Injury 2017; 2174-9.


164

Jürisson M, Vorobjov S, Kallikorm R, Lember M, Uusküla A. The incidence of hip fractures in Estonia, 2005–2012. Osteoporos Int 2015; 26(1): 77-84. Jürisson M, Raag M, Kallikorm R, Lember M, Uusküla A. The impact of hip fracture on mortality in Estonia: a retrospective population-based cohort study. BMC Musculoskelet Disord 2017; 18(1): 243-52. Kannegaard P N, Mark S van der, Eiken P, Abrahamsen B. Excess mortality in men compared with women following a hip fracture: national analysis of comedications, comorbidity and survival. Age Ageing 2010; 39(2): 203-9. Klop C, Welsing P, Cooper C, Harvey N, Elders P, Bijlsma J, Leufkens H, de Vries F. Mortality in British hip fracture patients, 2000–2010: a populationbased retrospective cohort study. Bone 2014; 66: 171-7. Kurtinaitis J, Dadonienė J, Kvederas G, Porvaneckas N, Butėnas T. Mortality after femoral neck fractures: a two-year follow-up. Med Kaunas Lith 2012; 48(3): 145-9. Liu Y, Wang Z, Xiao W. Risk factors for mortality in elderly patients with hip fractures: a meta-analysis of 18 studies. Aging Clin Exp Res 2017; 30(4): 323-30. Medin E, Goude F, Melberg H O, Tediosi F, Belicza E, Peltola M, on behalf of the EuroHOPE study group. European regional differences in all-cause mortality and length of stay for patients with hip fracture. Health Econ 2015; 24: 53-64. Mundi S, Pindiprolu B, Simunovic N, Bhandari M. Similar mortality rates in hip fracture patients over the past 31 years. Acta Orthop 2014; 85(1): 54-9.

Acta Orthopaedica 2019; 90 (2): 159–164

Neuman M D, Fleisher L A, Even-Shoshan O, Mi L, Silber J H. Non-operative care for hip fracture in the elderly: the influence of race, income, and comorbidities. Med Care 2010; 48(4): 314-20. Neuman M D, Silber J H, Magaziner J S, Passarella M A, Mehta S, Werner R M. Survival and functional outcomes after hip fracture among nursing home residents. JAMA Intern Med 2014; 174(8): 1273-80. Poenaru D V, Prejbeanu R, Iulian P, Haragus H, Popovici E, Golet I, Vermesan D. Epidemiology of osteoporotic hip fractures in Western Romania. Int Orthop 2014; 38(11): 2329-34. Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi J-C, Saunders L D, Beck C A, Feasby T E, Ghali W A. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care 2005; 43(11): 1130-9. Quan H, Li B, Couris C M, Fushimi K, Graham P, Hider P, Januel J-M, Sundararajan V. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol 2011; 173(6): 676-82. Ree C L P van de, Jongh M A C D, Peeters C M M, Munter L de, Roukema J A, Gosens T. Hip fractures in elderly people: surgery or no surgery? A systematic review and meta-analysis. Geriatr Orthop Surg Rehabil 2017; 8(3): 173-80. Sobolev B, Sheehan K J, Kuramoto L, Guy P. Excess mortality associated with second hip fracture. Osteoporosis Int 2015; 26(7): 1903-10.


Acta Orthopaedica 2019; 90 (2): 165–171

165

Superior fixation and less periprosthetic stress-shielding of tibial components with a finned stem versus an I-beam block stem: a randomized RSA and DXA study with minimum 5 years’ follow-up Maiken STILLING 1,3,5, Inger MECHLENBURG 1,2, Claus Fink JEPSEN 3, Lone RØMER 4, Ole RAHBEK 3,5, Kjeld SØBALLE 3,5, and Frank MADSEN 3 1 Orthopaedic Research Unit, Aarhus University Hospital; 2 Centre of Research in Rehabilitation (CORIR), Department of Clinical Medicine, Aarhus University Hospital and Aarhus University; 3 Department of Orthopaedic Surgery, Aarhus University Hospital; 4 Department of Radiology, Aarhus University Hospital; 5 Department of Clinical Medicine, Aarhus University, Denmark Correspondence: maiken.stilling@clin.au.dk Submitted 2018-06-06. Accepted 2018-11-07.

Background and purpose — The stem on the tibial component of total knee arthroplasty provides mechanical resistance to lift-off, shear forces, and torque. We compared tibial components with finned stems (FS) and I-beam block stems (IS) to assess differences in implant migration. Patients and methods — In a patient-blinded RCT, 54 patients/knees (15 men) with knee osteoarthritis at a mean age of 77 years (70–90) were randomly allocated to receive tibial components with either a FS (n = 27) or an IS (n = 27). Through 5 to 7 years’ follow-up, implant migration was measured with RSA, periprosthetic bone mineral density (BMD) was measured with DXA, and surgeons reported American Knee Society Score (AKSS). Results — At minimum 5 years’ follow-up, maximum total point motion (MTPM) was higher (p = 0.04) for IS (1.48 mm, 95% CI 0.81–2.16) than for FS (0.85 mm, CI 0.38–1.32) tibial components. Likewise, total rotation (TR) was higher (p = 0.03) for IS (1.51˚, CI 0.78–2.24) than for FS (0.81˚, CI 0.36–1.27). Tibial components with IS externally rotated 0.50° (CI –0.06 to 1.06) while FS internally rotated 0.09° (CI –0.20 to 0.38) (p = 0.03). Periprosthetic bone stress-shielding was higher (p < 0.01) up to 2 years’ follow-up for IS compared with FS in the regions medial to the stem (–13% vs. –2%) and posterior to the stem (–13% vs. –2%). Below the stem bone loss was also higher (p = 0.01) for IS compared with FS (–6% vs. +1%) up to 1-year follow-up. Knee score improved similarly in both groups up to 5 years’ follow-up. Interpretation — Periprosthetic bone stress-shielding medial and posterior to the stem until 2 years, and tibial component migration at 5 years, was less for a finned compared with an I-shaped block stem design.

Fixation of the tibial component in total knee arthroplasty (TKA) can be augmented with different stem designs. The classic stem shape is either a central block or cylinder with or without fins (Hernandez-Vaquero et al. 2008), which provides mechanical resistance to lift-off, shear forces, and torque during knee kinematics (Grupp et al. 2017). The smaller the tibial stem, the less bone is lost at implantation, and potentially preserved for later revisions (Molt and Toksvig-Larsen 2015). A finned stem design provides a greater mechanical resistance to torque than a block stem (Hernandez-Vaquero et al. 2008). A further advantage of a finned stem compared with a cylindrical or a block stem is that less bone volume is removed and the periprosthetic bone is exposed to less stress (HernandezVaquero et al. 2008, Molt and Toksvig-Larsen 2015). Therefore, improved fixation and survival could be expected with a finned stem design. With cemented TKA fixation primary stability of the components is secured initially and longer term fixation relies on the quality of the fixation interfaces, including cement penetration and bone quality (Vertullo and Davey 2001, Andersen et al. 2017). Early migration of the tibial component measured with radiostereometric analysis (RSA) is predictive for an increased risk for subsequent loosening at a later stage (Ryd et al. 1995, Pijls et al. 2012). We compared tibial components with an I-beam stem (IS) and a finned stem (FS) at 2 and minimum 5 years’ follow-up in a randomized trial evaluating RSA-measured implant migration and polyethylene wear, changes in periprosthetic bone mineral density (BMD), and differences in clinical outcome measured by the American Knee Society Score (AKSS).  

Patients and methods Between January 2005 and December 2007, 54 patients with primary osteoarthritis of the knee were assessed for eligibility

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2019.1566510


166

ENROLLMENT

Acta Orthopaedica 2019; 90 (2): 165–171

Assessed for eligibility n = 54 Randomized n = 54

ALLOCATION

Allocated to I-beam stem (n = 27) Received allocated intervention (n = 27)

Allocated to Finned stem (n = 27) Received allocated intervention (n = 27)

FOLLOW-UP

Lost to follow-up at 5 years (n = 12): – due to comorbidity, 6 – dead, 4 – declined to participate, 2

Lost to follow-up at 5 years (n = 11): – due to comorbidity, 2 – dead, 7 – declined to participate, 2

A

B

Figure 2. The Maxim Total Knee (Zimmer Biomet, Warsaw, Indiana) with cobalt-chromium Tibial Tray Interlok components with (A) an I-beam stem and (B) a finned stem.

ANALYSIS

Analyzed with RSA and DXA: – at 2 years (n = 22) – at 5–7 years (n = 15)

Analyzed with RSA and DXA: – at 2 years (n = 27) – at 5–7 years (n = 16)

Figure 1. CONSORT flow diagram showing the inclusion/exclusion process and follow-up until minimum 5 years

in this single-center patient-blinded randomized controlled clinical trial. Randomization in blocks of 6 patients (3 IS and 3 FS) was done by drawing labels from a box, and the labels were then concealed in 54 consecutively numbered closed envelopes. All eligible patients gave their informed consent to participation to the surgeon and received allocation intervention at Aarhus University Hospital, Denmark (Figure 1, Table 1). The inclusion criteria were primary osteoarthritis of the knee, age above 70 years, informed consent, and only one knee operated. The exclusion criteria were severe neuromuscular or vascular disease of the lower extremities, insufficient bone quality for a TKA, known osteoporosis, previous proximal tibial osteotomy, or other major knee surgery. Calculation of sample size The primary end point was 2 years’ follow-up, and with minimal relevant difference of 0.5 mm maximum total point motion (MTPM) (power 90%, alpha 0.05, standard deviation 0.6 mm) (Ryd et al. 1995) this study was powered for 22 patients per group. To compensate for eventual dropouts, 27 patients per group were included (Figure 1). Implants Maxim Total Knee (Zimmer Biomet, Warsaw, IN, USA) cruciate-retaining components were used. The cobalt-chromium modular Tibial Tray Interlok had either an I-beam block stem or a finned stem (Figure 2). Both stem types were 4 cm long and fixed to the tibial baseplate (non-modular). Both components were fixed in the bone by vacuum-mixed Palacos R bone cement (Heraeus Medical GmbH, Wehrheim, Germany) applied under the baseplate while the stem was fixed press-fit (without cement) in the proximal tibia. The femoral component Maxim cobalt-chromium (Biomet Inc, Warsaw, IN, USA) and the patella resurfacing polyethylene was fixed by Palacos

Table 1. Demographics, surgical and clinical data at baseline Factor Men / women Operated side (right / left) Age at surgery (mean, range) BMI at surgery (mean, range) Number of surgeons Implant size (range) Polyethylene thickness (mm) Surgery time (min) AKSS (max 100) (mean, range) Knee Score Function Score)

I-beam stem (n = 27)

Finned stem (n = 27)

8 / 19 12 / 15 77 (70–90) 28 (20–37) 4 71–83 10 (8–12) 63 (45–85)

7 / 20 12 / 15 77 (70–85) 29 (21–37) 4 69–83 10 (8–12) 48 (48–90)

34 (13–70) 45 (0–70)

36 (10–62) 54 (15–90)

R bone cement. The polyethylene insert was a modular component of gamma sterilized Arcom (Zimmer Biomet, Warsaw, IN, USA) ultra-high molecular weight polyethylene fixed with similar locking splits in IS and FS components (Figure 2). Surgery All patients were operated in a theatre with laminar airflow by 4 experienced knee surgeons. A tourniquet was applied and an anterior midline incision was used. The posterior cruciate ligament was retained in all cases. In both groups the proximal tibia was cut using the same extra-medullary guide, aiming for a perpendicular cut in the frontal plane and a posterior slope of 3°. The cut surfaces of the patella and femur were cleaned by high-pressure lavage before cementation. 5–6 tantalum beads (1 mm) (Wennbergs Finmek AB, Gunnilse, Sweden) were inserted in the proximal tibia intraoperatively. All patients received a draining tube in the joint for approximately 24 hours. All patients were treated prophylactically with a preoperative single dose i.v. 2 g dicloxacillin and all received prophylactic thrombotic medication with 1 daily dose subcutaneous 2.5 mg fondaparinux for 5 to 7 days. The patients were mobilized on the first postoperative day and allowed weight-bearing as tolerated, but with the assistance of 2 crutches for the first 6 weeks. The in-hospital stay varied between 4 and 6 days.


Acta Orthopaedica 2019; 90 (2): 165–171

Radiosteometric analysis (RSA) Stereoradiographs were obtained within the first 2–3 days after surgery (reference examination). Subsequent examinations were obtained at 6 weeks, 3 months, at 1 and 2 years, and cross-sectionally at 5 to 7 years with the patient supine and the operated knee aligned parallel to the calibration box (y-axis) in a foam positioner. The position and orientation of the global coordinate system in the reference examination defined the direction of implant migration in the follow-up examinations. At 5 years standing stereoradiographs (30° knee flexion) were also obtained. We used a fully digitized standard RSA setup (FCR Profect CS; Fujifilm, Vedbaek, Denmark) with 2 synchronized ceiling-fixed roentgen tubes (Arco-Ceil/Medira; Santax Medico, Aarhus, Denmark) angled 40° on each other and an unfocused uniplanar carbon calibration box (Box 24, Medis Specials, Leiden, the Netherlands). Analysis was performed with Model Based RSA vs. 3.21 (RSAcore, Leiden, The Netherlands) using CAD implant models (Kaptein et al. 2007). The signed migrations described the general migration of the implants, and the maximum total point motion (MTPM) described the vector of the point of the implant model that migrated the most. Implants were classified as stable or continuously migrating based on the MTPM, as described by Ryd et al. (1995). Further, we calculated the total rotation (absolute implant rotation) using the Pythagorean theorem (TR = √X2+Y2+Z2). Polyethylene wear was measured as loss of joint space width by calculating the migration difference on the y-axis between the femoral and tibial model components at 5 years (standing) stereoradiographs with postoperative (supine) stereoradiographs as baseline. The condition number of the bone marker model was mean 43 (SD 22) and the rigid body error was mean 0.17 (SD 0.1). Guidelines for maximum CN (< 150) and ME (0.35) were used as upper limits (Valstar et al. 2005), and no RSA analyses were excluded by these criteria. The repeatability of the RSA measurements was calculated based on double stereoradiographic examinations of 29 patients (15 IS and 14 FS) at the last follow-up (Valstar et al. 2005). The postoperative stereoradiograph was used as the reference in migration analysis of the double examinations (Table 2, see Supplementary data). Dual-energy X-ray absorptiometry (DXA) DXA scans of the periprosthetic bone were performed within the first postoperative week (baseline), and at 1 and 2 years, and cross-sectionally at 5 to 7 years using a Lunar Prodigy DXA Scanner (GE Healthcare, Waukesha, WI, USA). The patients were positioned supine with the leg in a foam frame to keep the knee semi-flexed by approximately 25° and the lower leg in neutral rotation (Stilling et al. 2010). Rice was applied around the knee as tissue-equivalent material and scans were performed with the “spine” mode. Analysis was performed in 3 regions of interest (ROI) (Figure 3) with a precision range from 1.8% to 3.7% for the anteroposterior scans, and from 3.4% to 6.2% for the lateral scans (Stilling et al. 2010).

167

A

B

Figure 3. DXA scans showing the implant detection of the finned stemmed implant (blue), bone borders (yellow line), and 3 regions of interest (ROI) around the stem. A. Anterior/posterior view. B. Lateral view.

Clinical follow-up The patients were seen for clinical examination preoperatively, and at 1 and 2 years, and cross-sectionally at 5 to 7 years postoperatively. Clinical data collection was conducted unblinded by the 4 surgeons. The AKSS was used to quantify the functional result, patient satisfaction, and pain (Insall et al. 1989). Statistics All continuous variables were tested for normality (Shapiro Wilk test). The groups were then compared by a 2-sample t-test or a 2-sample t-test with unequal variances as appropriate. Non-normal data were tested by a 2-sample Wilcoxon rank-sum (Mann–Whitney) test. Spearman’s rho was used to test correlations between implant migration, bone-density changes, patient factors, and clinical outcome. The statistical analyses were performed using the STATA 14.0 (StataCorp, College Station, TX, USA) software package. The significance level was set at 0.05. Ethics, funding, and potential conflicts of interest The Central Denmark Region Committee on Biomedical Research Ethics approved the protocol (Journal Number: M-20030239), which was registered with ClinicalTrials.gov (NCT00175136), and performed in compliance with the Helsinki Declaration. Informed consent was obtained from all participants. Zimmer Biomet, the Danish Rheumatism Association, and the A.P. Møller Foundation unconditionally sponsored the study. The authors declare no conflicts of interest.

Results RSA 1, 2, and minimum 5 years’ RSA data are shown in Figure 4 and Table 3. There was similar implant migration between groups until 2 years’ follow-up. On group level, there was no measurable migration between 1 and 2 years for IS (p = 0.3) or for FS (p = 0.7). Between 1 and 2 years 21 tibial components (12/22 IS and 9/27 FS) were classified with continuous migration (MTPM > 0.2mm) and 28 were stable (10/22 IS and


168

Acta Orthopaedica 2019; 90 (2): 165–171

Mean (SD) MTPM (mm) I-beam stem Finned stem

2.5

Mean (SD) (+) internal / (–) external rotation) (°)

Mean (SD) total rotation (°) I-beam stem Finned stem

2.5

2.0

2.0

1.5

1.5

0.5

0.0

–0.5 1.0

1.0

0.5

0.5

0.0

0.0

–1.0

0

10

20

30

40

50

60

Months after index operation

I-beam stem Finned stem

–1.5 0

10

20

30

40

50

60

Months after index operation

0

10

20

30

40

50

60

Months after index operation

Figure 4. Line plot summarizing the MTPM, total rotation (TR), and rotation about the y-axis (internal–external rotation) of the tibial components in the 2 stem groups at 6 weeks, 3 months, 1 year, 2 years, and minimum 5 years. Between 2 and minimum 5 years’ follow-up there was a statistically significant difference in all 3 migration parameters showing higher migration in the tibial components with I-beam stem (red) compared with finned stem (blue). The dots mark the means and the error bars are standard deviations.

Table 3. Signed migrations of the tibial components at 1, 2, and minimum 5 years’ follow-up. Values are mean (95% CI)

I-beam stem

Finned stem

p-value c

x–translation (+medial/–lateral): 1 year 0.05 (–0.12 to 0.22) –0.02 (–0.12 to 0.07) 2 years 0.04 (–0.18 to 0.27) –0.03 (–0.12 to 0.07) 5 years –0.20 (–0.50 to 0.09) 0.02 (–0.06 to 0.10) y–translation (+lift-off/–subsidence): 1 year 0.09 (0.01 to 0.17) 0.09 (0.01 to 0.17) 2 years 0.12 (0.06 to 0.18) 0.10 (0.01 to 0.19) 5 years 0.15 (0.05 to 0.25) 0.08 (–0.05 to 0.21) z–translation (+anterior/–posterior): 1 year 0.07 (–0.23 to 0.37) –0.22 (–0.45 to 0.01) 2 years –0.07 (–0.43 to 0.29) –0.16 (–0.36 to 0.04) 5 years –0.39 (–0.89 to 0.10) –0.07 (–0.39 to 0.25) MTPM a 1 year 1.02 (0.65 to 1.40) 0.94 (0.65 to 1.24) 2 years 1.17 (0.72 to 1.61) 0.91 (0.65 to 1.17) 5 years 1.48 (0.81 to 2.16) 0.85 (0.38 to 1.32) x–rotation (+anterior tilt/–posterior tilt): 1 year 0.15 (–0.34 to 0.64) –0.33 (–0.64 to –0.02) 2 years –0.05 (–0.52 to 0.41) –0.22 (–0.51 to 0.07) 5 years –0.52 (–1.33 to 0.29) –0.12 (–0.66 to 0.42) y–rotation (+internal rotation/–external rotation): 1 year –0.05 (–0.37 to 0.26) 0.03 (–0.19 to 0.19) 2 years –0.31 (–0.59 to –0.04) –0.02 (–0.27 to 0.22) 5 years –0.50 (–1.06 to 0.06) 0.09 (–0.20 to 0.38) z–rotation (+varus/–valgus): 1 year 0.02 (–0.21 to 0.26) 0.04 (–0.10 to 0.17) 2 years 0.00 (–0.32 to 0.32) 0.03 (–0.10 to 0.16) 5 years 0.39 (0.01 to 0.76) –0.01 (–0.14 to 0.13) b TR 1 year 1.12 (0.75 to 1.48) 0.90 (0.66 to 1.14) 2 years 1.12 (0.73 to 1.50) 0.88 (0.67 to 1.08) 5 years 1.51 (0.78 to 2.24) 0.81 (0.36 to 1.27) a MTMP: maximum total point motion. b The total rotation (TR) was calculated

theorem.

c Difference

0.5 0.5 0.5 1.0 0.6 0.3 0.5 0.4 0.5 0.8 0.8 0.04 0.2 0.4 0.9 0.7 0.2 0.03 1.0 0.7 0.1 0.5 0.5 0.04

using the 3D Pythagorean

between groups (two-sample Wilcoxon rank-sum (Mann–Whitney) test).

18/27 FS) (p = 0.1) (Ryd et al. 1995). At 2 years, 5/21 of IS and 5/27 of FS had rotation of more than 0.7° about the y-axis (p = 0.7) (Gudnason et al. 2017), and at 5 years this was 4/14 in both groups. Between 2 and minimum 5 years there was no measurable migration at group level (p > 0.09); however, 19 tibial components (12/15 IS and 7/16 FS) had MTPM > 0.2 mm and 12 were stable (3/15 IS and 9/16 FS) (p = 0.04). We found no clinically relevant or statistically significant correlations between implant migration (MTPM) and sex (p = 0.9) and BMI (p = 0.6). There was a correlation between MTPM at minimum 5 years’ follow-up and periprosthetic bone loss percentage medially (rho –0.38; p = 0.03) under the tibial component. There was 68% more (p = 0.04) polyethylene wear of tibial components with IS (1.05 mm, CI 0.64–1.46) compared with FS (0.48 mm, CI 0.06–0.89). The wear-rate (mean and (standard deviation)) of polyethylene inserts in tibial components with IS and FS was 0.21 (0.15) mm/year and 0.10 (0.15) mm/ year, respectively. DXA IS had more periprosthetic stress-shielding (bone loss) (p < 0.01) up to 2 years’ follow-up in the regions medial to the stem (–13% vs. –2%) and posterior to the stem (–13% vs. –2%) compared with FS (Table 4, see Supplementary data). Below the stem, bone loss was also more pronounced (p = 0.01) for IS compared with FS (–6% vs. +1%) up to 1-year follow-up. Between 2 and minimum 5 years the bone loss medially, laterally, and below the stem decreased markedly for FS (p < 0.002) whereas there was no statistically significant difference for IS (p > 0.2). Clinical results There were no intraoperative complications. 1 patient (IS) had a postoperative superficial infection that resolved after anti-


Acta Orthopaedica 2019; 90 (2): 165–171

biotic treatment, and 1 patient (FS) had an early deep infection that was treated with reoperation including lavage and exchange of the polyethylene liner plus 6 weeks of antibiotic treatment, and healed without further complications. There were no revisions within 5 years’ follow-up. Clinical results were similar between the 2 groups of patients at minimum 5 years of follow-up (Table 5, see Supplementary data). Knee score and function score improved until 1 year but not thereafter. 27 patients had an extension deficit at baseline; all except 2 patients regained full extension (5 and 10 degrees extension deficit).

Discussion This is the first study to compare tibial components with different stem designs in a randomized study. The key findings were that tibial components with finned stem in comparison with I-beam block stem had less periprosthetic bone stressshielding up to 2 years, and maintained better fixation and had less polyethylene wear at 5 years. Fixation and survival Although cemented components achieve stable fixation during surgery, they normally present with a pattern of some initial migration until 3 months and thereafter stabilization until small individual component migration become indicative of later aseptic loosening (Ryd et al. 1995). Classically, cemented metal-backed tibial components migrate via tilting subsidence with lift-off and loosening (Gudnason et al. 2017). The summed migration measures (MTPM and TR, Figure 4) in our study indicated more migration in the IS group already at 6 weeks, but there was no progressive migration at group level until after 2 years’ follow-up. At 1-year followup, the mean MTPM in both groups was slightly higher than the reported mean of 0.7 mm for the AGC TKA in the metaanalysis study of Pijls et al. (2012) which placed the implants in our study in the “at risk” group (0.45–1.6 mm MTPM at 1 year), with a higher than 5% revision risk at 10 years. Ryd et al. (1995) reported MTPM around 1 mm at 5 years and 1.2 mm at 10 years in non-revised cemented tibial components. Between 2 and minimum 5 years, the IS group had mean MTPM of 1.48 mm in 12 of 15 patients displaying continuous tibial component migration, whereas the FS group had mean MTPM of 0.85 mm in 7 of 16 patients displaying continuous tibial component migration. We did not have any revisions until the cross-sectional 5- to 7-year follow-up in this study apart from a liner exchange during 1 infection reoperation. The Finnish Knee Arthroplasty Registry have shown a 94% 10-year survival rate in osteoarthritic knees operated with the AGC TKA (Biomet), a cemented tibial platform with an I-shaped block stem, which is the predecessor of the Maxim TKA (Biomet) of this study (Himanen et al. 2005). In a large case series, Faris et al. (2015) reported aseptic 10-year survival

169

of 98.7% for the AGC TKA system (n = 5,972) and 98.5% survival rate for the Vanguard TKA system (n = 1,209). The Vanguard TKA system has a finned stem and is the successor of the Maxim TKA (Biomet) with a finned stem used in this study. The arthroplasty registries do not report the stem design of the tibial components, and it is unclear how many tibial components have a fixed or modular IS or FS stem besides the AGC TKA system, the Maxim TKA system, and the Vanguard system. Further, it is unclear whether survival of IS and FS in these TKA brands are different. At minimum 5 years we found higher total rotation for IS compared with FS, which showed as a difference in signed rotation about the y-axis, as 0.5° external rotation for IS and 0.09° internal rotation for FS. This indicates tibial components with IS being less resistant to torque over time as compared with tibial components with FS. With a sensitivity of 50% and specificity of 90%, y-axis rotation at a threshold of 0.7° is also predictive of aseptic loosening of the tibial component (Gudnason et al. 2017). Periprosthetic stress-shielding Periprosthetic stress-shielding and bone loss is well known in the tibial metaphysis after TKA. Minoda et al. (2010) showed an approximate 40% relative BMD change medial and lateral to a tibial component cemented stem 2 years after surgery. Varus knees have been shown to have higher BMD in the more loaded medial tibial metaphysis prior to TKA intervention; however, knee alignment correction with TKA resulted in similar medial and lateral BMD values suggesting bone remodeling responding to load (Soininvaara et al. 2004). Heterogeneous BMD loss in relation to a central stem has been described up to 7 years’ follow-up (Hernandez-Vaquero et al. 2008). At 2 years’ follow-up we found periprosthetic bone loss of mean 10–14% of postoperative BMD values medial, lateral, anterior, and posterior to the IS. In comparison, periprosthetic bone changes around the FS was between mean –7% (anterior) and +1% to –2% medial, lateral and posterior to the stem. The reason was probably higher bone stress around the I-beam stem by torsional forces. After 2 years the bone stress-shielding eased off in the IS group but increased in the FS group. We saw a correlation between MTPM at 5 years’ follow-up and the periprosthetic bone loss percentage under the medial as well as the lateral part of both types of tibial components at 2 years indicating a clinical significance of bone loss. Formerly only a relationship between the preoperative BMD in the knee region and postoperative cementless tibial component migration has been shown (Andersen et al. 2017). Polyethylene wear We found a mean polyethylene insert wear-rate of 0.21 mm/ year and 0.10 mm/year in tibial components with IS and FS, respectively. This wear may arise from articulate wear as well as backside wear on the metal-backed tibial components


170

(Conditt et al. 2004, Collier et al. 2007). The locking mechanism (an anterior split) of the polyethylene insert as well as the average PE thickness was the same and could not explain the wear difference found between stem groups. Our method could not differentiate between polyethylene wear in the medial and lateral compartments but the magnitude was similar to other reports, though these generally showed a bit more wear in the medial compartment and a bit less in the lateral compartment (Conditt et al. 2004, Collier et al. 2007). Clinical outcomes The AKS scores in healthy persons without TKA in the same age range as the patients in our study are a knee score of median 100 (70–100) and a function score of median 71 (17– 100) (Bremner-Smith et al. 2004). At 5 years, our patients had a function score of mean 58 (49–67) with a marginal and statistically insignificant improvement of 5 points since baseline. This was far below the clinically relevant change of 35 points (Jacobs and Christensen 2009) but, in comparison with other TKA studies with 6–11 years’ follow-up, our patients scored only slightly lower (Ladermann et al. 2008) or comparably (Arthur et al. 2013, Breugem et al. 2014). The knee score of our patients increased by more than 50 points until 5 years to a mean 89 (84–94) points, which was only slightly lower than for healthy individuals. Knee flexion of 115° was similar between groups and similar to other reports (Faris et al. 2015). Limitations and strengths Due to aging and morbidity, there was a marked dropout of 23 patients (of a total 54) at 5 years’ follow-up in this elderly patient group (mean age 77 years at time of inclusion). However, the dropouts were similarly distributed between groups (12 IS, 11 FS), and although it was higher than the anticipated sample size at 2 years, the study had sufficient power to detect group differences at 5 years. The strength of the study was that only the stem design differed between groups; the tibial components were otherwise alike and fixed with cement in the same manner and by the same surgeons. In summary, at minimum 5 years’ follow-up there was less periprosthetic bone stress-shielding, superior fixation, less polyethylene wear, and similar clinical outcome on AKKS for tibial components with finned stem compared with tibial components with I-beam stem. Supplementary data Tables 2, 4, and 5 are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/ 17453674.2019.1566510

The authors wish to thank consultant orthopedic surgeons Anders Odgaard and Andreas Kappel for performing some of the surgeries. They thank project coordinator Rikke Mørup for performing the radiostereometric analyses and Peter Bo Jørgensen for analysing the DXA scans.

Acta Orthopaedica 2019; 90 (2): 165–171

FM, MS, KS, and OR designed the study. MS, FM, and LR gathered the data. MS analyzed the data. IM and MS wrote the initial draft. MS, IM, KS, OR, FM, and LR revised the final draft. Acta thanks Stephan Maximilian Röhrl for help with peer review of this study.   Andersen M R, Winther N S, Lind T, Schroder H M, Flivik G, Petersen M M. Low preoperative BMD is related to high migration of tibia components in uncemented TKA-92 patients in a combined DEXA and RSA study with 2-year follow-up. J Arthroplasty 2017; 32(7): 2141-6. Arthur C H, Wood A M, Keenan A C, Clayton R A, Walmsley P, Brenkel I. Ten-year results of the Press Fit Condylar Sigma total knee replacement. Bone Joint J 2013; 95(2): 177-80. Bremner-Smith A T, Ewings P, Weale A E. Knee scores in a ‘normal’ elderly population. Knee 2004; 11(4): 279-82. Breugem S J, van Ooij B, Haverkamp D, Sierevelt I N, van Dijk C N. No difference in anterior knee pain between a fixed and a mobile posterior stabilized total knee arthroplasty after 7.9 years. Knee Surg Sports Traumatol Arthrosc 2014; 22(3): 509-16. Collier M B, Engh C A Jr, McAuley J P, Engh G A. Factors associated with the loss of thickness of polyethylene tibial bearings after knee arthroplasty. J Bone Joint Surg Am 2007; 89(6): 1306-14. Conditt M A, Ismaily S K, Alexander J W, Noble P C. Backside wear of modular ultra-high molecular weight polyethylene tibial inserts. J Bone Joint Surg Am 2004; 86(5): 1031-7. Faris P M, Ritter M A, Davis K E, Priscu H M. Ten-year outcome comparison of the anatomical graduated component and vanguard total knee arthroplasty systems. J Arthroplasty 2015; 30(10): 1733-5. Grupp T M, Saleh K J, Holderied M, Pfaff A M, Schilling C, Schroeder C, Mihalko W M. Primary stability of tibial plateaus under dynamic compression-shear loading in human tibiae: influence of keel length, cementation area and tibial stem. J Biomech 2017; 59:9-22. Gudnason A, Adalberth G, Nilsson K G, Hailer N P. Tibial component rotation around the transverse axis measured by radiostereometry predicts aseptic loosening better than maximal total point motion. Acta Orthop 2017; 88(3): 282-7. Hernandez-Vaquero D, Garcia-Sandoval M A, Fernandez-Carreira J M, Gava R. Influence of the tibial stem design on bone density after cemented total knee arthroplasty: a prospective seven-year follow-up study. Int Orthop 2008; 32(1): 47-51. Himanen A K, Belt E, Nevalainen J, Hamalainen M, Lehto M U. Survival of the AGC total knee arthroplasty is similar for arthrosis and rheumatoid arthritis. Finnish Arthroplasty Register report on 8,467 operations carried out between 1985 and 1999. Acta Orthop 2005; 76(1): 85-8. Insall J N, Dorr L D, Scott R D, Scott W N. Rationale of the Knee Society clinical rating system. Clin Orthop Relat Res 1989; (248): 13-14. Jacobs C A, Christensen C P. Correlations between knee society function scores and functional force measures. Clin Orthop Relat Res 2009; 467(9): 2414-9. Kaptein B L, Valstar E R, Stoel B C, Reiber H C, Nelissen R G. Clinical validation of model-based RSA for a total knee prosthesis. Clin Orthop Relat Res 2007; 464: 205-9. Ladermann A, Lubbeke A, Stern R, Riand N, Fritschy D. Fixed-bearing versus mobile-bearing total knee arthroplasty: a prospective randomised, clinical and radiological study with mid-term results at 7 years. Knee 2008; 15(3): 206-10. Minoda Y, Kobayashi A, Iwaki H, Ikebuchi M, Inori F, Takaoka K. Comparison of bone mineral density between porous tantalum and cemented tibial total knee arthroplasty components. J Bone Joint Surg Am 2010; 92(3): 700-6. Molt M, Toksvig-Larsen S. 2-year follow-up report on micromotion of a short tibia stem: a prospective, randomized RSA study of 59 patients. Acta Orthop 2015; 86(5): 594-8. Pijls B G, Valstar E R, Nouta K A, Plevier J W, Fiocco M, Middeldorp S, Nelissen R G. Early migration of tibial components is associated with late


Acta Orthopaedica 2019; 90 (2): 165â&#x20AC;&#x201C;171

revision: a systematic review and meta-analysis of 21,000 knee arthroplasties. Acta Orthop 2012; 83(6): 614-24. Ryd L, Albrektsson B E, Carlsson L, Dansgard F, Herberts P, Lindstrand A, Regner L, Toksvig-Larsen S. Roentgen stereophotogrammetric analysis as a predictor of mechanical loosening of knee prostheses. J Bone Joint Surg Br 1995; 77(3): 377-83. Soininvaara T A, Miettinen H J, Jurvelin J S, Suomalainen O T, Alhava E M, Kroger H P. Periprosthetic tibial bone mineral density changes after total knee arthroplasty: one-year follow-up study of 69 patients. Acta Orthop Scand 2004; 75(5): 600-5.

171

Stilling M, Soballe K, Larsen K, Andersen N T, Rahbek O. Knee flexion influences periprosthetic BMD measurement in the tibia: suggestions for a reproducible clinical scan protocol. Acta Orthop 2010; 81(4): 463-70. Valstar E R, Gill R, Ryd L, Flivik G, Borlin N, Karrholm J. Guidelines for standardization of radiostereometry (RSA) of implants. Acta Orthop 2005; 76(4): 563-72. Vertullo C J, Davey J R. The effect of a tibial baseplate undersurface peripheral lip on cement penetration in total knee arthroplasty. J Arthroplasty 2001; 16(4): 487-92.


172

Acta Orthopaedica 2019; 90 (2): 172–178

Equivalent 2-year stabilization of uncemented tibial component migration despite higher early migration compared with cemented fixation: an RSA study on 360 total knee arthroplasties Elise K LAENDE 1,2, Janie L ASTEPHEN WILSON 1,3, Joanna MILLS FLEMMING 4, Edward R VALSTAR 5, C Glen RICHARDSON 2, and Michael J DUNBAR 1,2 1 School

of Biomedical Engineering, Dalhousie University, Halifax, Nova Scotia, Canada; 2 Division of Orthopaedics, Department of Surgery, Dalhousie University and QEII Health Sciences Centre, Nova Scotia Health Authority, Halifax, Nova Scotia, Canada; 3 Department of Surgery, Dalhousie University, Halifax, Nova Scotia, Canada; 4 Department of Mathematics and Statistics, Dalhousie University, Halifax, Nova Scotia, Canada; 5 Department of Orthopedics, Leiden University Medical Center, Leiden, The Netherlands Correspondence: elaende@dal.ca Submitted 2018-06-27. Accepted 2018-10-26.

Background and purpose — Thresholds of implant migration for predicting long-term successful fixation of tibial components in total knee arthroplasty have not separated cemented and uncemented fixation. We compared implant migration of cemented and uncemented components at 1 year and as the change in migration from 1 to 2 years. Patients and methods — Implant migration of 360 tibial components measured using radiostereometric analysis was compared at 1 year and as the change in migration from 1 to 2 years in 222 cemented components (3 implant designs) and 138 uncemented components (5 implant designs). Results — 1-year maximum total point motion was lower for the cemented tibial components compared with the uncemented components (median = 0.31 mm [0.03–2.98] versus 0.63 mm [0.11–5.19] respectively, p < 0.001, mixed model). The change in migration from 1 to 2 years, however, was equivalent for cemented and uncemented components (mean [SD] 0.06 mm [0.19] and 0.07 mm [0.27] mm respectively, p = 0.6, mixed model). Interpretation — These findings suggest that current thresholds of acceptable migration at 1 year may be better optimized by considering cemented and uncemented tibial components separately as higher early migration of uncemented components was not associated with decreased stability from 1 to 2 years.

Cemented fixation in total knee arthroplasty (TKA) is currently the most common method of fixation, but there is increasing interest in uncemented TKA in an effort to provide longer lasting constructs to the young, active patient through osseointegration of the tibial component (Drexler et al. 2012, Brown et al. 2013, Cherian et al. 2014, Mont et al. 2014). A concern with uncemented TKA is that failure to achieve initial fixation may lead to revisions due to aseptic loosening. Early patterns of implant migration measured with radiostereometric analysis (RSA) have been shown to predict long-term implant outcomes. In particular, 2 studies have demonstrated the predictive value of migration 1-year post-operation (Pijls et al. 2012b), and the change in migration between 1 and 2 years postoperatively (Ryd et al. 1995) in determining long-term survivorship. Notably, both of these studies pooled cemented and uncemented tibial components in their analyses. In contrast, a Cochrane Review (Nakama et al. 2012) concluded that although cemented tibial components had lower initial migration, uncemented fixation provided a lower risk of future aseptic loosening, as measured indirectly as a change in migration between 1 and 2 years, despite higher early migration. While cemented fixation depends on an immediate mechanical interlock provided by cured bone cement, uncemented fixation requires bone in-growth into the implant surface, which occurs in the early postoperative period (Freeman and Tennant 1992, Dalury 2016). Because of these fundamental differences in the mechanisms of early fixation for cemented and uncemented components, it is unclear if it is appropriate to evaluate cemented and uncemented tibial components under the same thresholds of early migration for prediction of successful fixation. In this study we compared the magnitudes of implant migration of cemented and uncemented tibial components at 1 year postoperatively and between 1 and 2 years postoperatively. We hypothesized that the uncemented components would have

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2018.1562633


Acta Orthopaedica 2019; 90 (2): 172â&#x20AC;&#x201C;178

higher migration levels at 1 year postoperatively, but similar migration magnitudes between 1 and 2 years postoperatively, indicating good long-term performance. The secondary objective of this analysis was to examine the effect of implant design on tibial component migration.

Methods This study included RSA data on subjects who received a primary TKA between 2002 and 2015 at 2 institutions (Halifax Infirmary, Halifax, Nova Scotia and St. John of God Hospital Subiaco, Perth, Australia). The source of data for this study is the Halifax RSA Database, which was created with the aim of collecting RSA outcome data on a wide range of arthroplasty implants. Subjects were included in the database if they were part of implant-specific RSA protocols (completed or ongoing) or were enrolled in an implant-generic RSA protocol for any subject undergoing primary or revision knee or hip arthroplasty (Figure 1 and Tables 1 and 2, see Supplementary data). All subjects had tantalum RSA markers inserted into the proximal tibia and into the non-articulating periphery of the polyethylene component at the time of surgery. Subjects received postoperative care that included antibiotics, anticoagulation medication, and physiotherapy in hospital and after discharge. All subjects were mobilized to immediate full weight-bearing postoperatively. Subjects were followed for 2 years and had RSA exams immediately postoperatively (reference exam) and at a minimum of 1 and 2 years postoperatively. Details of the RSA equipment are included in the Supplementary data (Table 3). Inclusion criteria for this analysis were a primary diagnosis of osteoarthritis, no previous knee replacement, and RSA migration data at both 1 and 2 years postoperatively. Exclusion criteria included severe joint deformity requiring revision components in primary cases, revision of the tibial component, and technical problems with the RSA analysis (insufficient markers visible, condition number > 150, or mean error of rigid body fitting > 0.35 mm) (Valstar et al. 2005). The primary outcome measure was RSA-defined implant migration calculated as maximum total point motion (MTPM), the vector length of the point on the implant that moved the most (Ryd et al. 1995). All analyses used fictive markers at standardized locations for MTPM calculations (Nilsson et al. 1991). Rigid body motions were calculated using markerbased methods (Selvik 1989) to eliminate any differences due to model fitting that may occur with model-based RSA. Migrations at 1 and 2 years were calculated relative to the immediate postoperative reference exam. Statistics Mixed models were fitted to determine whether fixation (cemented or uncemented) had a significant effect on (i) migration at 1 year (relative to the immediate postoperative

173

reference exam) and (ii) the change in migration between 1 and 2 years postoperatively. The models included patients as random effects (to account for bilateral cases) along with fixed effects for sex, age, and BMI using the lme4 package (Bates et al. 2015) in R (R Core Team 2015). For 1-year migrations, log(MTPM) was taken as for the outcome variable (Astephen et al. 2010, Pijls et al. 2012a). For the change in migration from 1 to 2 years, the proportion of subjects for which it exceeded the 0.2 mm threshold (indicative of continuous migration) was also calculated (Ryd et al. 1995). Similar mixed models were fitted to the cemented and uncemented groups separately in order to investigate the influence of implant design. Significance was set at p < 0.05. Confidence intervals for statistical analyses are included in the Supplementary data (Table 5). Ethics, funding, and potential conflict of interest Ethics approval was obtained (local REB approval number 1020265) and subjects provided written consent. Funding was provided by the Atlantic Canada Opportunities Agency. Authors MJD, CGR, and JLAW have consultancy agreements with Stryker, a commercial party indirectly related to this article. Previous unrestricted research grants have been received from Stryker, Zimmer, and Wright Medical Technologies Inc. by the institution with which MJD, CGR, and EKL are associated.

Results Subjects 518 primary TKA with RSA markers inserted were available from the Halifax RSA Database. After cases were removed under the exclusion criteria or due to missed follow-up visits (Figure 1 and Table 1, see Supplementary data) 360 primary TKA in 333 individuals were analyzed; 222 knees had cemented tibial baseplates and 138 were uncemented. For bone and implant markers, conditions numbers were 36 (19) and 37 (20) (mean [SD]) respectively while mean error of rigid body fitting was 0.13 mm (0.07) mm and 0.10 mm (0.07) respectively. Double exam precision was calculated for all available cases (Table 4, Supplementary data). Surgeries were performed by 7 surgeons and 8 implant designs were used (5 uncemented, Table 6). Simplex P bone cement (Stryker, Mahwah, NJ, USA) was used for all cemented components. Comparing demographics between the cemented and uncemented groups (Table 6), the uncemented group had a lower mean BMI (p < 0.001, t-test) and a higher proportion of male subjects (p < 0.001, Fisherâ&#x20AC;&#x2122;s exact test). 1-year migration Tibial component migration measured as MTPM at 1 year was lower for the cemented group compared with the uncemented group (p < 0.001 unadjusted and adjusted for age, sex, BMI; Figure 2, Table 7).


174

Acta Orthopaedica 2019; 90 (2): 172–178

Table 6. Subject demographics by fixation (cemented and uncemented) and by implant design Fixation Insert a

Implant

All implants All cemented implants Advance b MP, PS NexGen c PS Triathlon d PS, CR, CS All uncemented implants b Advance Biofoam porous coated, no screws MP Advance Biofoam b porous coated + screws MP TM Monoblock c trabecular metal PS, CR TM Modular c trabecular metal PS Triathlon PA-Coated d porous coated + periapatite PS, CR, CS

n 360 222 59 30 133 138 22 20 48 16 32

Mean age (SD) Mean BMI (SD) % female 64 (7.8) 64 (8.3) 64 (7.8) 66 (8.7) 64 (8.4) 65 (7) 69 (5.2) 69 (5.2) 64 (7.7) 62 (8.7) 65 (8.4)

32 (6.1) 33 (6.5) 32 (5.6) 32 (5.7) 34 (6.9) 31 (5.2) 30 (3.8) 31 (4.6) 32 (5.4) 35 (4.8) 29 (5.0)

61 68 69 60 70 49 55 35 60 63 31

a

MP = medial pivot (posterior cruciate ligament resected), PS = posterior stabilized (posterior cruciate ligament resected), CR = cruciate retaining, and CS = cruciate stabilized. Inc., Arlington, TN

b Wright Medical Technology, c Zimmer, Warsaw, IN d Stryker, Mahwah, NJ

Table 7. Tibial component 1-year MTPM migration and change in MTPM migration from 1 to 2 years by fixation and implant groups Implant

n

All cemented implants Advance NexGen Triathlon All uncemented implants Biofoam Biofoam + screws TM Monoblock TM Modular Triathlon PA

222 59 30 133 138 22 20 48 16 32

1-year migration (MTPM, mm) Mean (SD) Median (range) 0.41 (0.36) 0.40 (0.24) 0.49 (0.34) 0.40 (0.41) 0.98 (0.94) 1.08 (1.05) 0.82 (0.75) 0.89 (0.77) 1.52 (1.24) 0.85 (1.00)

0.31 (0.03–2.98) 0.30 (0.16–1.13) 0.38 (0.14–1.66) 0.28 (0.03–2.98) 0.63 (0.11–5.19) 0.68 (0.18–4.09) 0.66 (0.29–3.80) 0.55 (0.14–3.06) 1.20 (0.31–5.19) 0.50 (0.11–4.17)

1-year MTPM (mm)

1-year MTPM (mm)

5

5

4

4

3

3

2

2

1

1

0

0

Cemented

Uncemented

Figure 2. 1-year MTPM migration by fixation (cemented, n = 222; uncemented, n = 138). Boxes enclose 25th–75th percentiles with internal horizontal line at the median, whiskers extend a further 1.5 times the inter-quartile range and points beyond this range are plotted individually.

Change in MTPM from 1 to 2 years (mm) Mean (SD) Median (range) 0.06 (0.19) 0.04 (0.14) 0.13 (0.38) 0.05 (0.14) 0.07 (0.27) 0.05 (0.32) 0.04 (0.31) 0.05 (0.21) 0.32 (0.35) –0.01 (0.19)

0.04 (–0.38 to 1.76) 0.06 (–0.31 to 0.36) –0.01 (–0.16 to 1.76) 0.03 (–0.38 to 0.53) 0.04 (–0.76 to 1.30) 0.01 (–0.62 to 1.25) 0.04 (–0.55 to 0.94) 0.04 (–0.76 to 0.78) 0.19 (–0.10 to 1.30) 0.00 (–0.72 to 0.43)

Cemented Uncemented

Advance NexGen Triathlon Biofoam Biofoam TM TM Triathlon + screws Modular Monoblock PA

Figure 3. 1-year MTPM migration for cemented and uncemented tibial components by implant design. Boxes enclose 25th–75th percentiles with internal horizontal line at the median, whiskers extend a further 1.5 times the inter-quartile range and points beyond this range are plotted individually.


Acta Orthopaedica 2019; 90 (2): 172–178

2-year – 1-year MTPM (mm)

175

2-year – 1-year MTPM (mm)

1.5

1.5

1.0

1.0

0.5

0.5

0

0

–0.5

–0.5

Cemented

Uncemented

Figure 4. Change in MTPM migration from 1 to 2 years by fixation (cemented, n = 222; uncemented, n = 138). Boxes enclose 25th–75th percentiles with internal horizontal line at the median, whiskers extend a further 1.5 times the inter-quartile range and points beyond this range are plotted individually.

Cemented Uncemented

Advance NexGen Triathlon Biofoam Biofoam TM TM Triathlon + screws Modular Monoblock PA

Figure 5. Change in MTPM migration from 1 to 2 years for cemented and uncemented tibial components by implant design. Boxes enclose 25th–75th percentiles with internal horizontal line at the median, whiskers extend a further 1.5 times the inter-quartile range and points beyond this range are plotted individually.

Within the cemented group, the NexGen implant group had greater 1-year migration (p = 0.03 unadjusted and adjusted for age, sex, BMI; Figure 3). For the uncemented implants, the TM Modular implant demonstrated higher 1-year migration (unadjusted: p-value < 0.001; adjusted for age, sex, BMI: p-value = 0.006; Figure 3). Change between 1- and 2-year migration The change in MTPM migration between 1 and 2 years was not statistically significantly different between the cemented and uncemented groups (unadjusted: p-value = 0.7; adjusted for age, sex, BMI: p-value = 0.6; Figure 4, Table 7). The proportion of implants with continuous migration between 1 and 2 years of more than 0.2 mm was similar between groups, with 29/221 (13%) in the cemented group and 21/138 (15%) in the uncemented group. When comparing the change in migration of individual implant designs in the group between 1 and 2 years, the NexGen group had a higher change in migration (p = 0.03 unadjusted and adjusted for age, sex, BMI; Table 7, Figure 5). The NexGen group contained 1 outlier, defined as having a change in MTPM of more than 2 SD from the mean. Removing this subject did not alter the overall conclusion of the analysis (cemented and uncemented implants had similar change in migration between 1 and 2 years), but within the cemented group the NexGen group no longer had statistically significantly different migration for both the 1-year migration value and the change in migration from 1 to 2 years. Of the 5 implant designs in the uncemented group, the TM Modular group had a greater change in migration from 1 to 2 years compared with the other implant designs (p < 0.001 unadjusted and adjusted for age, sex, BMI; Table 7, Figure 5). Additionally, of the continuous migrators, 8 were TM Modular implants, representing half of this implant group. As the

results for the TM Modular group suggested poorer long-term outcomes, there was concern that this group was influencing the overall results for the uncemented group. To investigate this, the TM Modular group (n = 16) was excluded and the data reanalyzed. Using a modified uncemented group did not alter the conclusions: MTPM migration at 1 year remained significantly higher for the modified uncemented group (median = 0.57 mm, range 0.11–4.17 mm, n = 122) compared with the unchanged cemented group (p < 0.001 unadjusted and adjusted for age, sex, BMI) and the change in MTPM migration between 1 and 2 years for the modified uncemented group (mean [SD] 0.03 mm [0.24]) was similar to that for the cemented group (p = 0.4 unadjusted and adjusted).

Discussion The application of equivalent thresholds of acceptable RSA migration at 1 year for cemented and uncemented TKA appears to be suboptimal, as higher initial migration seen in the first postoperative year for uncemented components was not associated with greater migration between 1 and 2 years, which is an established criterion for predicting longer-term fixation (Ryd et al., 1995). Pooling RSA data of both cemented and uncemented tibial components has been employed in 2 important previous studies using early RSA data to predict long-term implant outcomes. In the first study, Ryd et al. (1995) found that MTPM migration between 1 and 2 years postoperatively of greater than 0.2 mm was predictive of later loosening with 85% predictive power. 158 cases were included in that analysis, composed of 103 cemented components and 55 uncemented components. In the second study, Pijls et al. (2012b) in a meta-analysis concluded that mean MTPM migration at 1 year of greater than


176

0.5 mm put an implant design “at risk” and greater than 1.6 mm was “unacceptable” based on the predicted revision rate at 5 years. In that meta-analysis, 847 subjects were included from 50 RSA studies that were matched to survival studies of the same implant designs (20,599 subjects in 56 studies). Of the 28 implant designs included, 18 had cemented fixation and 10 were uncemented. The differences found in our study at 1 year are statistically significant and clinically relevant because the differences in means place the cemented group, as well as each cemented implant design, in the “stable” category and the uncemented group, and all individual uncemented implant designs, in the “at risk” category according to Pijls et al. (2012b). However, our findings of equivalent change in migration between 1 and 2 years indicate no greater risk for uncemented implants despite greater uncertainty at 1 year based on the current threshold. The value of increased certainty by 1 year is that in a model of phased innovation (Malchau 2000, Nelissen et al. 2011, Pijls 2014), an early time point for safety thresholds substantially reduces the follow-up time required, providing more timely assessment of implant designs and limiting exposure. Higher 1-year migration for uncemented implants may be due to a “settling” period prior to bone ingrowth (Onsten et al. 1998, Molt and Toksvig-Larsen 2014, Henricson and Nilsson 2016). Once osseointegration is achieved, the potential for long-term fixation is good for uncemented tibial components while cemented components are susceptible to cement-related complications such as cement delamination (Dalury 2016). Previous RSA studies comparing cemented and uncemented implants have reported higher early migration for the uncemented components while achieving good long-term performance with contemporary uncemented fixation, including hydroxyapatite coatings and trabecular metal monoblock components (Hilding et al. 1995, Nilsson et al. 1999, Regner et al. 2000, Toksvig-Larsen et al. 2000, Carlsson et al. 2005, Nilsson et al. 2006, Pijls et al. 2012a, Wilson et al. 2012, van Hamersveld et al. 2017, Pijls et al. 2018). Review papers of cemented versus uncemented fixation have been inconclusive, citing a lack of long-term follow-up studies, but do conclude that there are promising results, especially with hydroxyapatite coatings and trabecular metal in short-term and RSA studies (Nakama et al. 2012, Brown et al. 2013, Mont et al. 2014). The 1-year MTPM migrations in the current study are similar to those found in a meta-analysis of tibial component migration for cemented (mean MTPM of 0.44 mm) and uncemented components (mean MTPM of 1.09 mm) (Pijls et al. 2018). The authors additionally proposed using the previously published 1-year MTPM migration thresholds at 6 months instead due to minimal migration between 6 months and 1 year (Pijls et al. 2018), but do not suggest differing thresholds for cemented and uncemented fixation as we and other authors have previously suggested (Henricson and Nilsson 2016). We found statistically significant differences in 1- to 2-year migration between different types of uncemented fixation,

Acta Orthopaedica 2019; 90 (2): 172–178

suggesting that not all uncemented fixation is equivalent. The TM Modular implant group had the least favorable migration results with significantly greater migration at both 1 year and between 1 and 2 years, indicating that this implant design is at greater risk of poor long-term survivorship. While both the TM Modular and TM Monoblock tibial components rely on bone in-growth into a porous trabecular metal structure, the benefits of the lower modulus monoblock component may be compromised with the addition of a stiff baseplate in the modular component to allow polyethylene inserts to be locked in place. Previous studies on the TM Modular component have reported 4 failures due to aseptic loosening in 167 cases (2%, all within the first postoperative year) (Zandee van Rilland et al. 2015); 7 revised or radiographically loose components in a series of 51 subjects (Behery et al. 2017); 1 revision for subsidence out of 50 cases (Fricka et al. 2015); and statistically significantly higher overall migration compared with the TM Monoblock component, but no difference between groups in change in migration from 1 to 2 years in 53 subjects (Andersen et al. 2016). It has not been possible to date to identify the TM Modular component in isolation in any national knee registry reports, so the survivorship of this implant in general use remains to been seen. Notably, a similar uncemented implant design by the same manufacturer employing trabecular metal and a modular tibial tray was recalled in 2015 due to an increase in complaints of loosening and radiolucent lines (FDA 2015). Additionally of note, the uncemented group with screw fixation performed equivalently to the same implant without screw fixation, although the intention of screw fixation is to provide immediate stability. Lack of immediate stability with screw fixation has been seen in previous RSA studies (Nilsson et al. 2006, Stilling et al. 2011). The differences in magnitudes between the uncemented subgroups may offer a preview of refined thresholds for 1-year screening of uncemented implants: the median 1-year migration of the TM Modular group was 1.2 mm compared with 0.5–0.7 mm for the other four uncemented groups. Matching of RSA and survivorship studies will be required to perform the robust analysis of Pijls et al. (2012b) to determine if a separate early (6-month or 1-year) threshold for uncemented components is valid. Alternatively, using a later reference exam (such as 6 weeks or 3 months) may permit determination of an early threshold that is valid for both cemented and uncemented fixation, but may be more difficult to find evidence for from the available literature. A limitation of our study is that subjects were not randomly assigned to the cemented and uncemented groups. However, this study represents analysis of 1 of the largest datasets of TKA postoperative RSA data to date. The results therefore provide important insight mechanisms into early migration depending on implant fixation. The demographic data show statistical differences between groups, although the clinical relevance of a BMI difference of 2 (with both groups >


Acta Orthopaedica 2019; 90 (2): 172–178

30) is likely negligible. The proportion of females in the cemented group (68%) versus the uncemented group (49%) was unexpected and may reflect an unconscious bias by operating surgeons in not using uncemented implants in women due to bone quality concerns. These demographic variables were accounted for in the statistical models, so the differences between fixation methods cannot be attributed to mismatched demographic factors between the cemented and uncemented groups. Although demographic factors are eclipsed by the differences due to implant fixation method in the overall group, it is likely that demographic factors do influence implant migration and may account for some of the variability in early migration, especially within the uncemented group. We excluded revised implants in this study to allow a comparison of the 2 methods for thresholds of allowable motion. Of the 14 total revisions, only 3 were performed for reasons related to mechanical loosening (1 peri-prosthetic fracture and 2 for aseptic loosening) and all 3 revisions were performed within the first 2 postoperative years so these cases would have been excluded from the analysis by default as the change in migration from 1 to 2 years could not be evaluated. Excluding the remaining cases ensured that no additional cases of mechanical loosening were included as our data capture only the most responsible reason for revision in what may be a multifactorial process. In summary, our study finds that the pattern of migration between 1 and 2 years was similar between cemented and uncemented groups and therefore supports both the use of uncemented fixation and the previous findings that this metric is appropriate to evaluate all tibial component fixations (Ryd et al. 1995). However, the magnitudes of migration at 1 year are higher for the uncemented group suggesting that thresholds at 1 year may not apply equally to cemented and uncemented implants for predicting revision rates as suggested by Pijls et al. (2012b). A further refinement of the 1-year threshold may be appropriate for uncemented implants to enable more conclusive evaluations of uncemented implant designs. Supplementary data Figure 1 and Tables 1–5 are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/ 17453674.2018.1562633

EKL, MJD, JLAW, and ERV conceived of the study. MJD and CGR performed the surgeries. EKL compiled and analyzed the data and wrote the manuscript. JMF provided statistical consultation and reviewed the analyses. All authors provided critical review and approved the final manuscript. Edward Valstar passed away before the final version of manuscript was completed but provided significant contributions during the course of the study. As other authors have previously described, Dr Valstar was a dedicated researcher and teacher and his loss to the research community is great (Nelissen et al. 2017).

177

The authors gratefully acknowledge the participation of the orthopedic surgeons who performed surgeries in Halifax, NS, Canada: David Amirault, Gerry Reardon, Michael Gross, Michael Biddulph; and Perth, Australia: Dermot Collopy; the research staff: Sue Moore, Allan Hennigar, Brittany Scott, James Edwards, Michaela Wallace, Jo-Anne Douglas, and Elise McNeill (Perth, Australia); and the research study participants. Acta thanks Koen van Hamersveld and Anders Henricson for help with peer review of this study.

Andersen M R, Winther N, Lind T, Schr Der H, Flivik G, Petersen M M. Monoblock versus modular polyethylene insert in uncemented total knee arthroplasty. Acta Orthop 2016; 87(6): 607-14. Astephen J L, Wilson D A, Dunbar M J, Deluzio K J. Preoperative gait patterns and BMI are associated with tibial component migration. Acta Orthop 2010 Aug; 81(4): 478-86. Bates D, Maechler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Statistical Software 2015; 67(1): 1-48. Behery O A, Kearns S M, Rabinowitz J M, Levine B R. Cementless vs cemented tibial fixation in primary total knee arthroplasty. J Arthroplasty 2017; 32(5): 1510-5. Brown T E, Harper B L, Bjorgul K. Comparison of cemented and uncemented fixation in total knee arthroplasty. Orthopedics 2013; 36(5): 380-7. Carlsson A, Bjorkman A, Besjakov J, Onsten I. Cemented tibial component fixation performs better than cementless fixation: a randomized radiostereometric study comparing porous-coated, hydroxyapatite-coated and cemented tibial components over 5 years. Acta Orthop 2005; 76(3): 362-9. Cherian J, Banerjee S, Kapadia B, Jauregui J, Harwin S, Mont M. Cementless total knee arthroplasty: a review. J Knee Surg 2014; 27(03): 193-8. Dalury D F. Cementless total knee arthroplasty: current concepts review. Bone Joint J 2016; 98-B(7): 867-73. Drexler M, Dwyer T, Marmor M, Abolghasemian M, Sternheim A, Cameron H U. Cementless fixation in total knee arthroplasty: down the boulevard of broken dream—opposes. J Bone Joint Surg Br 2012; 94(11 Suppl A): 85-9. FDA. Class 2 Device Recall Persona Trabecular Metal Tibial Plate/Persona TM Tibia; 2015. Freeman M A, Tennant R. The scientific basis of cement versus cementless fixation. Clin Orthop Relat Res 1992; (276): 19. Fricka K B, Sritulanondha S, McAsey C J. To cement or not? Two-year results of a prospective, randomized study comparing cemented vs. cementless total knee arthroplasty (TKA). J Arthroplasty 2015; 30(9 Suppl): 55-8. Henricson A, Nilsson K G. Trabecular metal tibial knee component still stable at 10 years. Acta Orthop 2016 Jul 27; 87(5): 504-10. Hilding M B, Yuan X, Ryd L. The stability of three different cementless tibial components: a randomized radiostereometric study in 45 knee arthroplasty patients. Acta Orthop Scand 1995; 66(1): 21. Malchau H. Introducing new technology: a stepwise algorithm. Spine 2000; 25(3): 285. Molt M, Toksvig-Larsen S. Peri-Apatite™ enhances prosthetic fixation in TKA: a prospective randomised RSA study. J Arthritis 2014; 03(03): 134. Mont M, Pivec R, Issa K, Kapadia B, Maheshwari A, Harwin S. Long-term implant survivorship of cementless total knee arthroplasty: a systematic review of the literature and meta-analysis. J Knee Surg 2014; 27(05): 36976. Nakama G Y, Peccin M S, Almeida G J, Lira Neto Ode A, Queiroz A A, Navarro R D. Cemented, cementless or hybrid fixation options in total knee arthroplasty for osteoarthritis and other non-traumatic diseases. Cochrane DB Syst Rev 2012; 10: CD006193. Nelissen R G, Pijls B G, Karrholm J, Malchau H, Nieuwenhuijse M J, Valstar E R. RSA and registries: the quest for phased introduction of new implants. J Bone Joint Surg Am 2011; 93 (Suppl 3): 62-5.


178

Nelissen R, Kaptein B, Veeger D. Edward Valstar (1970–2017). Acta Orthop 2017; 88(6): 701-2. Nilsson K G, Karrholm J, Ekelund L, Magnusson P. Evaluation of micromotion in cemented vs uncemented knee arthroplasty in osteoarthrosis and rheumatoid arthritis: randomized study using roentgen stereophotogrammetric analysis. J Arthroplasty 1991; 6(3): 265. Nilsson K G, Karrholm J, Carlsson L, Dalen T. Hydroxyapatite coating versus cemented fixation of the tibial component in total knee arthroplasty: prospective randomized comparison of hydroxyapatite-coated and cemented tibial components with 5-year follow-up using radiostereometry. J Arthroplasty 1999; 14(1): 9. Nilsson K G, Henricson A, Norgren B, Dalen T. Uncemented HA-coated implant is the optimum fixation for TKA in the young patient. Clin Orthop Relat Res 2006; 448: 129. Onsten I, Nordqvist A, Carlsson A S, Besjakov J, Shott S. Hydroxyapatite augmentation of the porous coating improves fixation of tibial components: a randomised RSA study in 116 patients. J Bone Joint Surg Br 1998; 80(3): 417. Pijls B G. Evidence based introduction of orthopaedic implants: RSA, implant quality and patient safety. Doctoral Thesis, Leiden University 2014. p 1-194. Pijls B G, Valstar E R, Kaptein B L, Fiocco M, Nelissen R G. The beneficial effect of hydroxyapatite lasts: a randomized radiostereometric trial comparing hydroxyapatite-coated, uncoated, and cemented tibial components for up to 16 years. Acta Orthop 2012a; 83(2): 135-41. Pijls B G, Valstar E R, Nouta K A, Plevier J W, Fiocco M, Middeldorp S, et al. Early migration of tibial components is associated with late revision: a systematic review and meta-analysis of 21,000 knee arthroplasties. Acta Orthop 2012b; 83(6): 614-24. Pijls B G, Plevier J W M, Nelissen R. RSA migration of total knee replacements: a systematic review and meta-analysis. Acta Orthop 2018; 89(3): 320-8.

Acta Orthopaedica 2019; 90 (2): 172–178

R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2015. Regner L, Carlsson L, Karrholm J, Herberts P. Tibial component fixation in porous- and hydroxyapatite-coated total knee arthroplasty: a radiostereometric evaluation of migration and inducible displacement after 5 years. J Arthroplasty 2000; 15(6): 681. Ryd L, Albrektsson B E, Carlsson L, Dansgard F, Herberts P, Lindstrand A, et al. Roentgen stereophotogrammetric analysis as a predictor of mechanical loosening of knee prostheses. J Bone Joint Surg Br 1995; 77(3): 377. Selvik G. Roentgen stereophotogrammetry: a method for the study of the kinematics of the skeletal system. Acta Orthop Scand Suppl 1989; 232: 1. Stilling M, Madsen F, Odgaard A, Romer L, Andersen N T, Rahbek O, et al. Superior fixation of pegged trabecular metal over screw-fixed pegged porous titanium fiber mesh: a randomized clinical RSA study on cementless tibial components. Acta Orthop 2011; 82(2): 177-86. Toksvig-Larsen S, Jorn L P, Ryd L, Lindstrand A. Hydroxyapatite-enhanced tibial prosthetic fixation. Clin Orthop Relat Res 2000(370): 192. Valstar E R, Gill R, Ryd L, Flivik G, Börlin N, Kärrholm J. Guidelines for standardization of radiostereometry (RSA) of implants. Acta Orthop 2005; 76(4): 563-72. van Hamersveld K T, Marang-van de Mheen P J, Tsonaka R, Valstar E R, Toksvig-Larsen S. Fixation and clinical outcome of uncemented peri-apatite-coated versus cemented total knee arthroplasty: five-year follow-up of a randomised controlled trial using radiostereometric analysis (RSA). Bone Joint J 2017; 99-B(11): 1467-76. Wilson D A, Richardson G, Hennigar A W, Dunbar M J. Continued stabilization of trabecular metal tibial monoblock total knee arthroplasty components at 5 years measured with radiostereometric analysis. Acta Orthop 2012; 83(1): 36-40. Zandee van Rilland E D, Varcadipane J C, Geling O, Murai Kuba M, Nakasone C K. A minimum 2-year follow-up using modular trabecular metal tibial components in total knee arthroplasty. Recons Rev 2015; 5(3): 23-8.


Acta Orthopaedica 2019; 90 (2): 179–186

179

Predicting individual knee range of motion, knee pain, and walking limitation outcomes following total knee arthroplasty Yong-Hao PUA 1, Cheryl Lian-Li POON 1, Felicia Jie-Ting SEAH 2, Julian THUMBOO 3, Ross Allan CLARK 4, Mann-Hong TAN 5, Hwei-Chi CHONG 1, John Wei-Ming TAN 1, Eleanor Shu-Xian CHEW 1, and Seng-Jin YEO 5 1 Department 3 Department

of Physiotherapy, Singapore General Hospital, Singapore; 2 Department of Physiotherapy, Sengkang General Hospital, Singapore; of Rheumatology and Immunology, Singapore General Hospital, Singapore; 4 Research Health Institute, University of the Sunshine Coast, Sunshine Coast, Australia; 5 Department of Orthopaedic Surgery, Singapore General Hospital, Singapore Correspondence: pua.yong.hao@sgh.com.sg Submitted 2018-08-04. Accepted 2018-11-18.

Background and purpose — Up to 20% of patients are dissatisfied after total knee arthroplasty (TKA), mainly because of pain and restricted physical function. We developed a prediction model for 6-month knee range of motion, knee pain, and walking limitations in patients undergoing TKA surgery. Patients and methods — We performed a prospective cohort study of 4,026 patients who underwent elective, primary TKA between July 2013 and July 2017. Candidate predictors included demographic, clinical, psychosocial, and preoperative outcome measures. The outcomes of interest were (i) knee extension and flexion range of motion, (ii) knee pain rated on a 5-point ordinal scale, and (iii) self-reported maximum walk time at 6 months post TKA. For each outcome, we fitted a multivariable proportional odds regression model with bootstrap internal validation. Results — At 6 months post TKA, around 5% to 20% of patients had a flexion contracture ≥ 10°, range of motion < 90°, moderate to severe knee pain, or a maximum walk time ≤ 15 minutes. The model c-indices (the probabilities to correctly discriminate between 2 patients with different levels of follow-up TKA outcomes) when evaluating these patients were 0.71, 0.79, 0.65, and 0.76, respectively. Each postoperative outcome was strongly influenced by the same outcome measure obtained preoperatively (all p-values < 0.001). Additional statistically significant predictors were age, sex, race, education level, diabetes mellitus, preoperative use of gait aids, contralateral knee pain, and psychological distress (all p-values < 0.001). Interpretation — We have developed models to predict, for individual patients, their likely post-TKA levels of knee extension and flexion range of motion, knee pain, and walking limitations. After external validation, they can potentially be used preoperatively to identify at-risk patients and to help patients set more realistic expectations about surgical outcomes.

Some 11% to 20% of patients are dissatisfied following total knee arthroplasty (TKA) mainly because of knee motion limitations (Matsuda et al. 2013, Huang et al. 2017), knee pain (Gunaratne et al. 2017), and functional limitations (Gunaratne et al. 2017). Early and accurate identification of patients at risk for poor post-TKA outcomes would better direct resources toward preventive care for them. Furthermore, to facilitate shared decision-making, providing patients preoperatively (Barlow et al. 2016) with individualized information on expected post-TKA outcomes may help them set more realistic expectations about surgical outcomes (Husain and Lee 2015, Volkmann and FitzGerald 2015), which in turn improves patient satisfaction. Few studies have systematically combined multiple predictors to generate individualized outcome predictions. These studies have focused singularly on combined pain and physical function outcomes (Dowsey et al. 2016, Sanchez-Santos et al. 2018), health-related quality of life (Gutacker and Street 2017), or patient satisfaction (Van Onsem et al. 2016). Although consensus statements (Singh et al. 2016, Lange et al. 2017) have advocated knee range of motion, knee pain, and physical function as distinct and important post-TKA outcomes, no studies have developed a prediction model to provide individualized predictions on these outcomes separately. Thus, we aimed to create a prediction model for post-TKA knee range- of motion, knee pain, and walking limitations.

Patients and methods Between July 2013 and July 2017, we identified 5,491 patients aged ≥ 50 years who underwent a unilateral primary TKA for knee osteoarthritis (OA) in Singapore General Hospital—the largest tertiary teaching hospital in Singapore, which performed half of all knee arthroplasties in the nation of 5.6 million people (Singapore Ministry of Health [last accessed Janu-

© 2018 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2018.1560647


180

ary 19, 2018]). We excluded patients who underwent revision knee surgery within 6 months post TKA (n = 16). We also excluded patients who had a history of rheumatoid arthritis (n = 58) and patients with stroke or Parkinson’s disease (n = 108). For patients with consecutive admissions for TKA (n = 863), only data from their first admission were used. Of the remaining 5,309 patients, we selected a cohort of 4,026 patients with non-missing 6-month follow-up outcomes (Figure 1, Supplementary data). Included patients were similar to those who were excluded because of missing data (Table 1, Supplementary data). All data were collected by physiotherapists and data technicians trained in the testing procedures and entered into an electronic registry database as per routine practice policies of our institution. Study design and reporting was based on the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) Statement (Collins et al. 2015). Candidate predictors We selected candidate predictors based on clinical expertise, literature review (Dowsey et al. 2016, Van Onsem et al. 2016, Gutacker and Street 2017), and data availability in our databases. To improve the practicality of the prediction models, we considered variables that were less equipment-dependent and were routinely and easily measured in the clinical setting. We identified 14 predictors, which included demographic, clinical, psychosocial, and preoperative outcome measures (Table 1, Supplementary data). Of interest, the presence of contralateral knee pain was measured by the “Patient Category” item (response choice b) from the Knee Society Clinical Rating System (Insall et al. 1989). For the type of walking aids used preoperatively, we coded the responses into 4 categories: (1) none, (2) walking stick or umbrella, (3) quadstick, and (4) walking frame or 2 canes or crutches. To assess selfreported depression, a single question (Q28) from the SF-36 (“How much of the time during the past four weeks have you felt downhearted and depressed?”) was used (Pomeroy et al. 2001). Outcome measures The outcomes of interest were the 6-month postoperative knee range of motion, knee pain, and walking limitations. Notably, we have chosen an intermediate (6-month) postoperative timepoint because (i) model prediction accuracy may decrease with a longer time horizon, (ii) knowledge of intermediateterm (6-months) risk for poor TKA outcomes will aid patient education and assist in rehabilitation planning, and (iii) TKA outcomes such as knee range of motion were reportedly nearing their peak at the 6-month timepoint (Stratford et al. 2010). A long-arm goniometer was used to measure active-assisted knee extension and flexion range of motion with the patients in supine position. To measure knee pain, patients were asked to describe their usual knee pain during the past 4 weeks. This variable, taken from Q1 of the Oxford Knee Questionnaire

Acta Orthopaedica 2019; 90 (2): 179–186

(Dawson et al. 1998), has 5 categories: (i) none, (ii) very mild, (iii) mild, (iv) moderate, and (v) severe. To measure walking limitations, patients were asked to estimate the time they were able to walk (without a rest) before they had severe difficulty with the operated knee. This variable had 4 categories: (i) > 30 min, (ii) 16–30 min, (iii) 5–15 min, and (iv) around the house only.   Statistics Data are expressed as means (SD) and medians with quartiles for continuous variables and as counts with percentage for categorical variables. We used proportional-odds ordinal regression models, which examined the multivariable associations of the predictors listed in Table 1 (Supplementary data) with 3 outcomes—namely, knee flexion and extension range of motion, knee pain, and walking limitations. We used proportional-odds ordinal regression because (i) it can handle both ordinal and clumped continuous outcomes (Chang and Pocock 2000, Liu et al. 2017) and (ii) it preserves the information in ordinal outcomes and has greater precision compared with binary logistic regression. To avoid assuming linearity, we modelled all continuous predictors as restricted cubic splines (Durrleman and Simon 1989, Harrell Jr 2015). All other predictors were included in the models as binary or categorical variables. Apart from the “education-level” variable, which was missing in 7.7%, all other predictors were missing at very low levels (0.02% to 0.5%). Thus, we used the transcan function in the R Hmisc package (R Foundation for Statistical Computing) to singly impute missing values. To account for model overfitting, we shrank the odds ratios (ORs) in the models using penalized regression (Moons et al. 2004, Harrell Jr 2015). To account for the clustering of patients within surgeons, which may bias the confidence intervals toward being too narrow, we calculated Huber–White robust estimates of standard errors and confidence intervals (White 1980). We assessed model performance in 2 ways. First, we measured model discrimination by the concordance index (c-index). Similar to an AUC statistic, the generalized c-statistic may be interpreted as the probability to correctly discriminate between 2 patients with different levels of follow-up TKA outcomes, where a value of 1 represents perfect discrimination and 0.5 represents no discrimination (“coin flip”) (Harrell Jr 2015). We computed the generalized c-indices of our ordinal models and the c-indices for the predictions of poor post-TKA outcomes (defined as (i) a knee flexion contracture ≥ 10° (Ritter et al. 2008), (ii) a knee flexion range < 90°, (iii) a knee pain rating of “moderate” or “severe,” and (iv) a maximum walk time ≤ 15 min). Because a prediction model is expected to perform better (“optimistically”) in the development sample than in new (but similar) samples, bootstrap internal validation (Austin and Steyerberg 2017) with cluster sampling (Bouwmeester et al. 2013) was performed to shrink the c-indices for “optimism” (Harrell Jr 2015). Second, we assessed model calibration using loess-smoothed calibration plots.


Acta Orthopaedica 2019; 90 (2): 179â&#x20AC;&#x201C;186

181

Figure 2. Part 1. Forest plot of odds ratios (ORs) and 95% confidence intervals (CIs) from proportional-odds ordinal regression models predicting (i) knee extension and flexion range of motion, (ii) knee pain intensity, and (iii) walking limitations at 6 months post-TKA. Values to the immediate right of continuous predictors are quartiles, and the ORs estimate the odds of better TKA outcomes (i.e., greater knee range of motion, lower knee pain, and greater walking ability) at the 75th vs. the 25th percentile values. This scaling is done to facilitate the interpretation and comparison of effect sizes of continuous predictors that are often measured on different units. As an illustrative example, other variables being equal, patients with a BMI of 30 (75th percentile) had, on average, 0.89 times (95% CI, 0.80â&#x20AC;&#x201C;0.99) the odds of having greater walking ability at 6 months post-TKA, relative to patients with a BMI of 24 (25th percentile). TKA: total knee arthroplasty. Notably, as the goal of this analysis was prediction, the ORs are a measure of predictive effects and they should not be interpreted as causal effects (Steyerberg 2009).

In sensitivity analyses, we examined potential variations in temporal effects (year of knee surgery) and predictor effects over time (Austin et al. 2017), and we found no statistically significant overall time effect (data not shown). We also calculated sensitivity, specificity, positive and negative predictive values (with Wilson 95% confidence intervals [CI]) for various risk thresholds of our prediction models. We assessed and confirmed the appropriateness of our prediction model using residual plots and empirical cumulative logit curves, and all analyses (including the computation of 95% CI) and graphing were done with the rms (Harrell Jr 2017), ROCR (Sing et al. 2005), and ggplot2 (Wickham 2009) R packages (http:// www.r-project.org). The web-based application was developed with the R shiny (Chang et al. 2017) package.

Ethics, funding, and potential conflicts of interest The institutional review board approved the study with a waiver of informed consent (SingHealth CIRB 2014/2027, Singapore). This work was supported by the Singhealth Allied Health Research Publication Grant and the Singapore General Hospital SMART II Centre Grant. No conflicts of interest were declared.

Results The mean age of all 4,026 patients was 68 years (SD 7.5) and women accounted for three-quarters of the sample (Table 1, Supplementary data). Figure 2 shows the partial effects of each predictor for all outcomes.


182

Acta Orthopaedica 2019; 90 (2): 179–186

Figure 2. Part 2. See Figure legend on previous page.

Knee range of motion Preoperatively, mean (SD) knee extension and flexion range were 7° (7°) and 118° (18°), respectively. At 6 months postsurgery, around a fifth of patients (20%; CI, 18–21) had a knee flexion contracture ≥ 10° while 5.6% of the patients (4.9–6.4) had a knee flexion range < 90°. The generalized, optimismcorrected c-indices for the knee extension and flexion models were 0.65 and 0.70, respectively. The c-indices for the prediction of a postoperative knee flexion contracture ≥ 10° and a knee flexion range < 90° were 0.71 and 0.79. For knee extension range of motion (Figure 2), preoperative knee extension range was the strongest predictor. Additional statistically significant predictors (p < 0.001) of greater (better) postoperative knee extension range were younger age, absence of diabetes mellitus, and lower preoperative walking limitations (better walking ability). For knee flexion range of motion, preoperative knee flexion range was the strongest predictor. Additional statistically significant predictors (p < 0.001) of greater (better) postoperative knee flexion range were male

sex, absence of diabetes mellitus, and greater preoperative knee extension range. Knee pain Preoperatively, over four-fifths of patients (83%, CI 82–84) reported at least moderate knee pain; at 6 months post-surgery, the figure was around 1 in every 10 patients (8.7%, CI 7.9–9.7). The generalized, optimism-corrected c-index for the knee pain model was 0.58. The c-index for the prediction of a postoperative knee pain rating of at least “moderate” was 0.65. Beside lower preoperative knee pain levels, additional statistically significant predictors (p < 0.001) of lower (better) postoperative knee pain levels were lower preoperative depression levels, lower preoperative knee flexion range, and Chinese race. Walking limitations Preoperatively, half of the patients (55%, CI 53–57) reported an inability to walk for more than 15 minutes; at 6 months


Acta Orthopaedica 2019; 90 (2): 179–186

183

Table 2. Operating characteristics of prediction models to predict poor post-TKA outcomes at various risk thresholds Cutpoint (%)

TN (n)

FP (n)

FN (n)

TP (n) Sensitivity (95% CI) Specificity (95% CI)

PPV (95% CI)

Knee extension model: 0.05 219 3,002 15 773 0.98 (0.97–0.99) 0.07 (0.06–0.08) 0.20 (0.19–0.22) 0.10 844 2,377 60 728 0.92 (0.90–0.94) 0.26 (0.25–0.28) 0.23 (0.22–0.25) 0.15 1,490 1,731 141 647 0.82 (0.79–0.85) 0.46 (0.45–0.48) 0.27 (0.25–0.29) 0.17 a 1,790 1,431 183 605 0.77 (0.74–0.80) 0.56 (0.54–0.57) 0.30 (0.28–0.32) 0.20 2,070 1,151 261 527 0.67 (0.64–0.70) 0.64 (0.63–0.66) 0.31 (0.29–0.34) Knee flexion model: 0.05 2,586 1,144 71 207 0.74 (0.69–0.79) 0.69 (0.68–0.71) 0.15 (0.13–0.17) 0.06 a 2,833 897 84 194 0.70 (0.64–0.75) 0.76 (0.75–0.77) 0.18 (0.16–0.20) 0.10 3,213 517 130 148 0.53 (0.47–0.59) 0.86 (0.85–0.87) 0.22 (0.19–0.26) 0.15 3,472 258 166 112 0.40 (0.35–0.46) 0.93 (0.92–0.94) 0.30 (0.26–0.35) 0.20 3,607 123 196 82 0.29 (0.24–0.35) 0.97 (0.96–0.97) 0.40 (0.34–0.47) Knee pain model: 0.05 379 3,294 12 340 0.97 (0.94–0.98) 0.10 (0.09–0.11) 0.09 (0.08–0.10) 0.08 a 1,995 1,678 107 245 0.70 (0.65–0.74) 0.54 (0.53–0.56) 0.13 (0.11–0.14) 0.10 2,679 994 185 167 0.47 (0.42–0.53) 0.73 (0.71–0.74) 0.14 (0.12–0.17) 0.15 3,455 218 303 49 0.14 (0.11–0.18) 0.94 (0.93–0.95) 0.18 (0.14–0.23) 0.20 3,621 52 337 15 0.04 (0.03–0.07) 0.99 (0.98–0.99) 0.22 (0.14–0.34) Walking limitations model: 0.05 668 2,819 19 514 0.96 (0.94–0.98) 0.19 (0.18–0.20) 0.15 (0.14–0.17) 0.10 1,801 1,686 88 445 0.83 (0.80–0.86) 0.52 (0.50–0.53) 0.21 (0.19–0.23) 0.15 2,553 934 188 345 0.65 (0.61–0.69) 0.73 (0.72–0.75) 0.27 (0.25–0.29) a 0.17 2,705 782 205 328 0.62 (0.57–0.66) 0.78 (0.76–0.79) 0.30 (0.27–0.32) 0.20 2,931 556 263 270 0.51 (0.46–0.55) 0.84 (0.83–0.85) 0.33 (0.30–0.36)

NPV (95% CI) 0.94 (0.90–0.96) 0.93 (0.92–0.95) 0.91 (0.90–0.93) 0.91 (0.89–0.92) 0.89 (0.87–0.90) 0.97 (0.97–0.98) 0.97 (0.96–0.98) 0.96 (0.95–0.97) 0.95 (0.95–0.96) 0.95 (0.94–0.96) 0.97 (0.95–0.98) 0.95 (0.94–0.96) 0.94 (0.93–0.94) 0.92 (0.91–0.93) 0.91 (0.91–0.92) 0.97 (0.96–0.98) 0.95 (0.94–0.96) 0.93 (0.92–0.94) 0.93 (0.92–0.94) 0.92 (0.91–0.93)

TN = true negative, FP = false positive, FN = false negative, TP = true positive, PPV = positive predictive value, NPV = negative predictive value. a Optimal cut-off at Youden index. Poor post-TKA outcomes were defined as (i) a knee flexion contracture ≥ 10°, (ii) a knee flexion range < 90°, (iii) a knee pain rating of “moderate” or “severe,” and (iv) a maximum walk time ≤ 15 minutes.

post-surgery, this was just over 1 in every 10 patients (13%, CI 12–14). The generalized, optimism-corrected c-index for the walking limitations model was 0.70. The c-index for the prediction of a maximum walk time ≤ 15 minutes was 0.76. Besides lower levels of preoperative walking limitations, additional statistically significant predictors (p < 0.001) of lower (better) levels of postoperative walking limitations were younger age, the use of a smaller or no gait aid preoperatively, lower preoperative depression levels, the absence of contralateral knee pain, and Chinese race. The prediction models Figure 3 (Supplementary data) shows the calibration plots of all prediction models. Table 2 shows the test characteristics of the prediction models at various risk thresholds and suggests that the prediction models tend to be adept in identifying patients at low risk of poor TKA outcomes: its negative predictive values were ≥ 94% for identifying true low-risk patients at a 5% risk threshold. To facilitate the use of the prediction models in clinical practice, we created a web application (https://sgh-physio. shinyapps.io/predicTKR/) that shows the expected distributions of the TKA outcomes for individual patients based on their preoperative demographic and clinical characteristics. To provide a “bottom-line” prediction of outcome (Barlow

et al. 2016) and to avoid setting universal (fixed) outcome thresholds which do not account for patients’ baseline (preoperative) levels (Hildon et al. 2012), the app also computes the predicted probabilities of achieving patient-defined acceptable levels of outcomes.

Discussion We developed 4 models to predict, for individual patients, their likely levels of knee extension and flexion range of motion, knee pain, and walking limitations at 6 months post TKA. Our models showed adequate calibration (Figure 2) and modest to moderately good predictive discrimination when evaluating patients with poor postoperative outcomes, with c-indices ranging between 0.65 (knee pain model) and 0.79 (knee flexion model). Across all models, we found that the postoperative outcomes were strongly influenced by the same outcome measure obtained preoperatively—an unsurprising finding that is consistent with the literature (Gandhi et al. 2006, Stratford et al. 2010, Lewis et al. 2015, Harmelink et al. 2017, SanchezSantos et al. 2018). Nonetheless, preoperative outcomes are not the only predictor of postoperative outcomes (Figure 2), which supports the need to consider several factors when predicting postoperative outcomes.


184

Interestingly, we observed that greater preoperative knee flexion range of motion was associated with lower odds of better (lower) postoperative knee pain rating (IQR-OR, 0.82, CI, 0.74–0.91; Figure 1). To our knowledge, this association has not previously been examined and may seem paradoxical at first. However, it is plausibly explained by the emerging evidence of the inverse associations (i) between greater preoperative knee radiographic severity and lower (better) postoperative pain severity (Dowsey et al. 2012, Valdes et al. 2012) and (i) between greater knee flexion range of motion and lower radiographic severity of knee OA (Holla et al. 2011). Accordingly, it is possible that knee pain in patients with mild knee radiographic OA (and good flexion range of motion) is not directly driven by knee joint damage, but rather by chronic pain mechanisms such as higher pain sensitivity and/or central sensitization (Valdes et al. 2012). In our study, knee radiographic severity, indexed by the Kellgren–Lawrence grade in previous studies (Dowsey et al. 2016), was not included because it is not readily available for incorporation into realtime prediction given its semi-quantitative nature (Wright 2014). Thus, it would be of interest for future prediction modelling studies to compare the predictive information provided by Kellgren–Lawrence grade with that provided by knee flexion range of motion. In our study, patients’ preoperative depression level was an important predictor of postoperative 6-month pain and walking limitations. Reviewing the literature, recent systematic reviews have reported that greater preoperative psychological distress is associated with worse pain and physical function post TKA (Lewis et al. 2015, Bletterman et al. 2017). In terms of intervention studies, Riddle et al. (2011) reported that patients who were preoperatively instructed in pain coping skills reported better Month-2 self-reported pain and physical function outcomes. Similarly, Cai et al. (2017) demonstrated, in a recent randomized clinical trial, that providing targeted cognitive behavioral therapy in the early postoperative care resulted in better Month-6 self-reported pain and physical function outcomes. Thus, taken together, our results indicate that at-risk patients may benefit from targeted behavioral interventions during the preoperative and early postoperative periods to modify this risk factor and improve outcomes. In our multi-racial sample of Chinese, Malay, and AsianIndian patients, we observed racial variation in post-TKA outcomes: Chinese race was associated with significantly lower (better) postoperative levels of knee pain and walking limitations, and these findings persisted after adjustment for multiple demographic, clinical, psychosocial, and preoperative outcome measures (Figure 1). At a time when race has been generally inadequately studied in orthopedics (Somerson et al. 2014), our findings provide useful timely information and, potentially, differences in perceived pain thresholds (Tan et al. 2008) and sociocultural expressions of pain (Campbell and Edwards 2012) may help explain them. Although we

Acta Orthopaedica 2019; 90 (2): 179–186

found important variations in outcomes according to patient race, patients belonging to a racial group that is considered as having poor post-TKA outcomes may still achieve a good outcome prognosis if other variables are favorable. We emphasize the need to consider multiple factors when providing individualized outcome predictions. Also, we caution that predictions from our models should serve as a starting point for shared decision-making, and not as a definitive or final recommendation. Limitations Our study has limitations. First, our data come from only 1 institution, but it delivers care to a large segment of the nation’s population. Having a large and representative population-based sample improves the stability of our model predictions and their applicability to institutions with patients who have similar characteristic to our patients. Second, although our prediction models have satisfactory discrimination, their performance is not optimal. Furthermore, our models could be criticized for not including potentially important predictors such as the severity of radiographic knee OA and a comprehensive list of comorbidities and psychosocial factors. Nevertheless, as we continue to grow our database and refine the variables collected, we will be able to update our prediction models and improve their prediction accuracy. Third, although we have implemented our prediction models in a web application to improve their accessibility in the clinical setting, we acknowledge that prediction models are unlikely to be widely used unless they can be incorporated into electronic medical records systems. As our prediction models comprise routinely and easily measured variables in the clinical setting, it is possible to integrate and externally validate them in electronic medical records systems. Future studies should explore this possibility. Fourth, we studied an Asian sample so the extent to which our results may apply to non-Asians is unknown. Finally, although we believe that knowledge of intermediateterm (6 months) risk for poor TKA outcomes will aid patient education and assist in rehabilitation planning, our models do not predict longer-term outcomes; TKA outcomes such as self-reported knee pain and physical function may require as long as 2 years to reach a plateau (Giesinger et al. 2014, Lim et al. 2015). In summary, at 6 months post TKA, around 5% to 20% of patients had knee range limitations, moderate to severe knee pain, or walking limitations (maximum walk time ≤ 15 minutes). Using data that are routinely collected and available from our database, we have developed prediction models that can potentially complement clinical and shared decisionmaking by providing personalized risk estimates of these important outcomes. The next step in translating this work is to perform external testing before evaluating the impact of individualized risk predictions on improving patient outcomes and satisfaction.


Acta Orthopaedica 2019; 90 (2): 179–186

Supplementary data Table 1 and Figures 1 and 3 are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/17453674.2018.1560647 The authors acknowledge the support from Jennifer Liaw, the head of the Department of Physiotherapy, Singapore General Hospital. They also thank William Yeo from the Orthopaedic Diagnostic Centre, Singapore General Hospital, for his assistance. Finally, they thank Nai-Hong Chan and the therapy assistants (Penny Teh, Hamidah Binti Hanib, Bamgbang Heryanto, and Shela Devi D/o Perumal) for their kind assistance. YHP wrote the first draft of the manuscript and performed the data analysis. All authors contributed to conception and design of the study, critical analysis of the data, interpretation of the findings, and critical revision of the manuscript.  Acta thanks Daan Nieboer and Annette W-Dahl for help with peer review of this study.

Austin P C, Steyerberg E W. Events per variable (EPV) and the relative performance of different strategies for estimating the out-of-sample validity of logistic regression models. Stat Methods Med Res 2017; 26(2): 796-808. Austin P C, van Klaveren D, Vergouwe Y, Nieboer D, Lee D S, Steyerberg E W. Validation of prediction models: examining temporal and geographic stability of baseline risk and estimated covariate effects. Diagn Progn Res 2017; 1: 12. Barlow T, Scott P, Griffin D, Realpe A. How outcome prediction could affect patient decision making in knee replacements: a qualitative study. BMC Musculoskelet Disord 2016; 17: 304. Bletterman A N, de Geest-Vrolijk M E, Vriezekolk J E, Nijhuis-van der Sanden M W, van Meeteren N L, Hoogeboom T J. Preoperative psychosocial factors predicting patient’s functional recovery after total knee or total hip arthroplasty: a systematic review. Clin Rehabil 2017; 32(4): 512-25. Bouwmeester W, Moons K G, Kappen T H, van Klei W A, Twisk J W, Eijkemans M J, Vergouwe Y. Internal validation of risk models in clustered data: a comparison of bootstrap schemes. Am J Epidemiol 2013; 177(11): 1209-17. Cai L, Gao H, Xu H, Wang Y, Lyu P, Liu Y. Does a program based on cognitive behavioral therapy affect kinesiophobia in patients following total knee arthroplasty? A randomized, controlled trial with a 6-month follow-up. J Arthroplasty 2017; 33(3): 704-10. Campbell C M, Edwards R R. Ethnic differences in pain and pain management. Pain Manag 2012; 2(3): 219-30. Chang B H, Pocock S. Analyzing data with clumping at zero: an example demonstration. J Clin Epidemiol 2000; 53(10): 1036-43. Chang W, Cheng J, Allaire J J, Xie Y, McPherson J. shiny: web application framework for R. R package version 1.0.5. https://CRAN.R-project.org/ package = shiny 2017. Collins G S, Reitsma J B, Altman D G, Moons K G. Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD): the TRIPOD statement. Ann Intern Med 2015; 162(1): 55-63. Dawson J, Fitzpatrick R, Murray D, Carr A. Questionnaire on the perceptions of patients about total knee replacement. J Bone Joint Surg Br 1998; 80(1): 63-9. Dowsey M M, Nikpour M, Dieppe P, Choong P F. Associations between preoperative radiographic changes and outcomes after total knee joint replacement for osteoarthritis. Osteoarthritis Cartilage 2012; 20(10): 1095-102. Dowsey M M, Spelman T, Choong P F. Development of a prognostic nomogram for predicting the probability of nonresponse to total knee arthroplasty 1 year after surgery. J Arthroplasty 2016; 31(8): 1654-60.

185

Durrleman S, Simon R. Flexible regression models with cubic splines. Stat Med 1989; 8(5): 551-61. Gandhi R, de Beer J, Leone J, Petruccelli D, Winemaker M, Adili A. Predictive risk factors for stiff knees in total knee arthroplasty. J Arthroplasty 2006; 21(1): 46-52. Giesinger K, Hamilton D F, Jost B, Holzner B, Giesinger J M. Comparative responsiveness of outcome measures for total knee arthroplasty. Osteoarthritis Cartilage 2014; 22(2): 184-9. Gunaratne R, Pratt D N, Banda J, Fick D P, Khan R J K, Robertson B W. Patient dissatisfaction following total knee arthroplasty: a systematic review of the literature. J Arthroplasty 2017; 32(12): 3854-60. Gutacker N, Street A. Use of large-scale HRQoL datasets to generate individualised predictions and inform patients about the likely benefit of surgery. Qual Life Res 2017; 26(9): 2497-505. Harmelink K E M, Zeegers A, Hullegie W, Hoogeboom T J, Nijhuis-van der Sanden M W G, Staal J B. Are there prognostic factors for one-year outcome after total knee arthroplasty? a systematic review. J Arthroplasty 2017; 32(12): 3840-53 e1. Harrell Jr F E. Regression modeling strategies: with applications to linear models, logistic and ordinal regression, and survival analysis. New York: Springer; 2015. Harrell Jr F E. rms: Regression Modeling Strategies. R package version 5.1-1. http://CRAN.R-project.org/package = rms 2017. Hildon Z, Neuburger J, Allwood D, van der Meulen J, Black N. Clinicians’ and patients’ views of metrics of change derived from patient reported outcome measures (PROMs) for comparing providers’ performance of surgery. BMC Health Serv Res 2012; 12: 171. Holla J F, Steultjens M P, van der Leeden M, Roorda L D, Bierma-Zeinstra S M, den Broeder A A, Dekker J. Determinants of range of joint motion in patients with early symptomatic osteoarthritis of the hip and/or knee: an exploratory study in the CHECK cohort. Osteoarthritis Cartilage 2011; 19(4): 411-19. Huang Y, Lee M, Chong H C, Ning Y, Lo N N, Yeo S J. Reasons and factors behind post-total knee arthroplasty dissatisfaction in an Asian population. Ann Acad Med Singapore 2017; 46(8): 303-9. Husain A, Lee G C. Establishing realistic patient expectations following total knee arthroplasty. J Am Acad Orthop Surg 2015; 23(12): 707-13. Insall J N, Dorr L D, Scott R D, Scott W N. Rationale of the Knee Society clinical rating system. Clin Orthop Relat Res 1989; (248): 13-14. Lange T, Schmitt J, Kopkow C, Rataj E, Gunther K P, Lutzner J. What do patients expect from total knee arthroplasty? A Delphi consensus study on patient treatment goals. J Arthroplasty 2017; 32(7): 2093-9 e1. Lewis G N, Rice D A, McNair P J, Kluger M. Predictors of persistent pain after total knee arthroplasty: a systematic review and meta-analysis. Br J Anaesth 2015; 114(4): 551-61. Lim J B, Chi C H, Lo L E, Lo W T, Chia S L, Yeo S J, Chin P L, Tay K J, Lo N N. Gender difference in outcome after total knee replacement. J Orthop Surg (Hong Kong) 2015; 23(2): 194-7. Liu Q, Shepherd B E, Li C, Harrell F E, Jr. Modeling continuous response variables using ordinal regression. Stat Med 2017; 36(27): 4316-35. Matsuda S, Kawahara S, Okazaki K, Tashiro Y, Iwamoto Y. Postoperative alignment and ROM affect patient satisfaction after TKA. Clin Orthop Relat Res 2013; 471(1): 127-33. Moons K G M, Donders A R, Steyerberg E W, Harrell F E. Penalized maximum likelihood estimation to directly adjust diagnostic and prognostic prediction models for overoptimism: a clinical example. J Clin Epidemiol 2004; 57(12): 1262-70. Pomeroy I M, Clark C R, Philp I. The effectiveness of very short scales for depression screening in elderly medical patients. Int J Geriatr Psychiatry 2001; 16(3): 321-6. Riddle D L, Keefe F J, Nay W T, McKee D, Attarian D E, Jensen M P. Pain coping skills training for patients with elevated pain catastrophizing who are scheduled for knee arthroplasty: a quasi-experimental study. Arch Phys Med Rehabil 2011; 92(6): 859-65.


186

Ritter M A, Lutgring J D, Davis K E, Berend M E. The effect of postoperative range of motion on functional activities after posterior cruciate-retaining total knee arthroplasty. J Bone Joint Surg Am 2008; 90(4): 777-84. Sanchez-Santos M T, Garriga C, Judge A, Batra R N, Price A J, Liddle A D, Javaid M K, Cooper C, Murray D W, Arden N K. Development and validation of a clinical prediction model for patient-reported pain and function after primary total knee replacement surgery. Sci Rep 2018; 8(1): 3381. Sing T, Sander O, Beerenwinkel N, Lengauer T. ROCR: visualizing classifier performance in R. Bioinformatics 2005; 21(20): 7881. Singapore Ministry of Health. Knee replacement surgery. https://www.moh. gov.sg/content/moh_web/home/costs_and_financing/hospital-charges/ Total-Hospital-Bills-By-condition-procedure/Knee_Knee_replacement_ surgery.html (last accessed January 19, 2018). Singh J A, Dowsey M M, Dohm M, Goodman S M, Leong A L, Scholte Voshaar M, Choong P F. Achieving consensus on total joint replacement trial outcome reporting using the OMERACT filter: endorsement of the final core domain set for total hip and total knee replacement trials for endstage arthritis. J Rheumatol 2016; 44(11): 1723-6. Somerson J S, Bhandari M, Vaughan C T, Smith C S, Zelle B A. Lack of diversity in orthopaedic trials conducted in the United States. J Bone Joint Surg Am 2014; 96(7): e56. Steyerberg E. Clinical prediction models: a practical approach to development, validation and updating. New York: Springer; 2009.

Acta Orthopaedica 2019; 90 (2): 179â&#x20AC;&#x201C;186

Stratford P W, Kennedy D M, Robarts S F. Modelling knee range of motion post arthroplasty: clinical applications. Physiother Can 2010; 62(4): 37887. Tan E C, Lim Y, Teo Y Y, Goh R, Law H Y, Sia A T. Ethnic differences in pain perception and patient-controlled analgesia usage for postoperative pain. J Pain 2008; 9(9): 849-55. Valdes A M, Doherty S A, Zhang W, Muir K R, Maciewicz R A, Doherty M. Inverse relationship between preoperative radiographic severity and postoperative pain in patients with osteoarthritis who have undergone total joint arthroplasty. Semin Arthritis Rheum 2012; 41(4): 568-75. Van Onsem S, Van Der Straeten C, Arnout N, Deprez P, Van Damme G, Victor J. A new prediction model for patient satisfaction after total knee arthroplasty. J Arthroplasty 2016; 31(12): 2660-7 e1. Volkmann E R, FitzGerald J D. Reducing gender disparities in post-total knee arthroplasty expectations through a decision aid. BMC Musculoskelet Disord 2015; 16: 16. White H. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica 1980; 48: 817-30. Wickham H. ggplot2: elegant graphics for data analysis. New York: Springer; 2009. Wright R W. Osteoarthritis classification scales: interobserver reliability and arthroscopic correlation. J Bone Joint Surg Am 2014; 96(14): 1145-51.


Acta Orthopaedica 2019; 90 (2): 187–189

187

Custom-made asymmetric polyethylene liner to correct tibial component malposition in total knee arthroplasty — a case report Andreas KAPPEL 1,2, Claes Sjørslev BLOM 1, and Anders EL-GALALY 1,2 1 Orthopaedic Research Unit, Aalborg University Hospital, Hobrovej, 9000 Aalborg; 2 Department of Clinical Medicine, Aalborg University, Søndre Skovvej 15, 9000 Aalborg, Denmark Correspondence: andreas.kappel@rn.dk Submitted 2018-11-28. Accepted 2018-12-11.

A 56-year-old woman presented with knee pain, bow-leggedness, and instability following revision total knee arthroplasty 11 years previously. A complex surgical history related to her right knee was revealed. At the age of 19, she suffered a midshaft tibial fracture treated non-operatively resulting in a sagittal bowing deformity. The anterior cruciate ligament was reconstructed at the age of 37 and a proximal bony correction using Ilizarov external fixation was done to correct recurvatum at the age of 38. In addition, 10 arthroscopic procedures were performed on the knee from the age of 20 to 34 years. A primary cemented TKA was performed at the age of 44 (NexGen CR, femur size C, tibia size 3, polyethylene 12 mm and patella size 29; Zimmer Biomet, Warsaw, IN, USA). Due to instability a partial revision was done 5 months later where the femoral component was brought distally and the polyethylene liner changed to LPS (NexGen LCCK femur size C, stem 12×100 mm, medial and lateral augments size 5mm, polyethylene size 14 LPS). However, pain, malalignment, and instability persisted.

Figure 1. Before liner exchange.

Physical examination revealed a varus leg with varus thrust and lateral laxity of 5–10° in both extension and flexion, and limited knee hyperextension with flexion to 120°. There was no pathological medial or sagittal laxity, normal patellar tracking and no signs of malrotation. Radiographs revealed wellfixed components (Figure 1). Supplementary CT scan showed correct rotational placement of components. An EOS scan revealed coronal malposition of the tibial component with with mechanical tibiofemoral angulation of 9° varus (mechanical lateral distal femoral angle (mLDFA) = 91°, mechanical medial proximal tibial angle (mMPTA) = 82°) and a sagittal deformity of the tibia with posterior translation of the plateau and increased posterior slope (Figure 2). Since only the tibial component was malpositioned and the soft-tissue envelope was intact, we therefore decided to correct the malalignment with a custom-made asymmetric polyethylene liner that was designed in cooperation with the manufacturer (Figure 3). The design incorporated a medial build-up of 6 millimeters to correct the 9° of varus and a slight

Figure 2. EOS scan before liner exchange.

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2018.1561384


188

Acta Orthopaedica 2019; 90 (2): 187–189

Figure 3. Custom liner construct.

Figure 4. After liner exchange.

Figure 5. EOS scan after liner exchange.

posterior build-up of 3° to diminish the excessive slope of the tibial plateau and component. The revision surgery was uneventful. Following moderate medial and posteromedial release the liner was inserted, and both alignment and stability was found to be satisfactory (Figure 4). At 1-year follow-up the pain had decreased significantly and the patient had no complaints of malalignment or instability. Range of motion was from full extension to 120° flexion. We found no medial or sagittal laxity but still a lateral laxity of 5–10° in both extension and flexion. Outcome scores showed improvements from preoperative to 1-year follow-up: Oxford Knee Score (OKS) (from 12/48 to 31/48), EQ-5D-3L (from 0.3 to 0.7). The EOS scan at 1-year follow-up demonstrated neutral mechanical alignment (Figure 5).

In this case, malpositioning of the tibial component caused varus malalignment. Tibial component revision would be the standard treatment to address this. However, this otherwise straightforward procedure was considered rather complicated, as a standard stemmed component, due to the posterior translation of the tibial plateau, would not fit the actual anatomy (Figure 6). We considered the use of a very short cemented stem, but due to the bony deformity the risk of repeated malpositioning and risk of difficulties in balancing the knee gave reason for concern. A hinged implant was less tempting due to the young age of the patient. Correction of the sagittal deformity with one or more osteotomies was also considered, but the complexity and high risk of complications caused concern. The decision to use a custom-made liner was preceded by thorough physical and radiological examination. Coronal mechanical alignment, sagittal alignment and component rotation was examined with EOS and CT scan. Coronal and sagittal malposition of the tibial component was evident while the femoral component was well placed and the soft-tissue envelope intact. Changes in liner symmetries may affect alignment, softtissue tension, and ROM. The insertion of an asymmetric liner to restore mechanical alignment affects soft-tissue balance throughout the whole ROM and in our opinion requires that the femoral component is positioned absolutely correctly.

Discussion Bony deformity might complicate both primary and revision TKA thus meticulous pre-surgical planning of both bony resection and choice of implant are advisable to secure the mechanical axis and stability of the joint (Xiao-Gang et al. 2012, Loures et al. 2018).

Figure 6. Templating tibia.


Acta Orthopaedica 2019; 90 (2): 187â&#x20AC;&#x201C;189

While the mechanical alignment is restored with the custom implant, asymmetric stresses are introduced to both the tibial bony fixation and the tibial component polyethylene locking mechanism. Whether these asymmetric stresses do affect longevity by increasing the risk of aseptic loosening or tibial backside wear might be a reason for concern (Rao et al. 2002, Gromov et al. 2014). Our patient had a satisfactory 1-year follow up with no radiographic loosening and improved patient-reported outcomes. The use of a custom-made liner has been previously described by Sah et al. (2008) who used this technique to correct excessive slope of the tibial component in a complex primary TKA; however, in that case only sagittal correction was intended. To our knowledge, no previous reports have described the use of a custom liner to correct coronal or combined coronal and sagittal alignment. In our opinion, the use of a custom-made polyethylene liner offers a simple alternative to more complex revision knee surgery with good short-term follow up in this case. However, the

189

technique is an option only in isolated tibial malposition and is dependent on soft-tissue stability and/or level of constraint. Acta thanks Kaj Knutson for help with peer review of this study.

Gromov K, Korchi M, Thomsen M G, Husted H, Troelsen A. What is the optimal alignment of the tibial and femoral components in knee arthroplasty? Acta Orthop 2014; 85(5): 480-7. Loures F B, Correia W, Reis J H, Pires E Albuquerque R S, de Paula Mozela A, de Souza E B, Maia P V, Barretto J M. Outcomes after knee arthroplasty in extra-articular deformity. Int Orthop 2018; Sep 14. doi: 10.1007/s00264018-4147-9. [Epub ahead of print] Rao A R, Engh G A, Collier M B, Lounici S. Tibial interface wear in retrieved total knee components and correlations with modular insert motion. J Bone Jointt Surg Am 2002; 84(10): 1849-55. Sah A P, Scott R D, Iorio R. Angled polyethylene insert exchange for sagittal tibial malalignment in total knee arthroplasty. J Arthroplasty 2008; 23(1): 141-4. Xiao-Gang Z, Shahzad K, Li C. One-stage total knee arthroplasty for patients with osteoarthritis of the knee and extra-articular deformity. Int Orthop 2012; 36(12): 2457-63.


190

Acta Orthopaedica 2019; 90 (2): 190–190

Correspondence

Early recovery trajectories after fast-track primary total hip arthroplasty: the role of patient characteristics Sir,—It is with great interest that I read the recently published paper by Porsius and co-workers (Porsius et al. 2018). However, I am worried that the authors may not have obtained a valid estimate of hip function and pain using the Oxford Hip Score (OHS) (Dawson et al. 1996), which the authors define as their main outcome measure. Using the OHS, the authors have assessed hip function and pain 1 week before surgery and every week for the first 6 weeks after surgery in patients recovering from a total hip replacement. However, the OHS uses a recall period of 4 weeks (Dawson et al. 1996). I cannot find information in the paper regarding a potential modification of the originally validated version of the OHS. Could the authors please help explain? Thomas Bandholm Clinical Research Center, Amager-Hvidovre Hospital, University of Copenhagen, Denmark Department of Orthopedic Surgery, Amager-Hvidovre Hospital, University of Copenhagen Department of Physical and Occupational Therapy, AmagerHvidovre Hospital, University of Copenhagen, Denmark Email: Thomas.Quaade.Bandholm@regionh.dk

Sir,—We thank T Bandholm for his interest in our study especially concerning the weekly use of the Oxford Hip Score (OHS) as main outcome measure. In our study we used data that was obtained using a diary as reported in another study (Klapwijk et al. 2017). The OHS was assessed pre-operatively, and post-operatively on a weekly basis. To align with the goals of a diary study, we omitted the usual 4-week time frame used in the postoperative OHS questionnaire. In line with the other questions that were assessed daily in the diary, patients filled out the OHS for their current situation. We are not aware of studies reporting on the effects of the specific recall period on the OHS. In our opinion, a change of the recall period, which we believe was necessary for our research goal, does not lead to an invalid estimate of hip function and pain. However, it would be prudent not to compare our absolute post-operative scores directly to other studies using a 4-week time frame OHS. Our results should only be interpreted in line with the goal of our study, which was to characterize subgroups of patients according to their hip function trajectory in the first 6 weeks after primary THA. On behalf of all authors Jarry Porsius Department of Plastic and Reconstructive Surgery & Department of Rehabilitation Medicine, Erasmus University Medical Center Rotterdam, the Netherlands Email: j.porsius@erasmusmc.nl Dawson J, Fitzpatrick R, Carr A, Murray D. Questionnaire on the perceptions of patients about total hip replacement J Bone Joint Surg Br 1996; 78(2): 185-90. Klapwijk LC M, Mathijssen N M C, van Egmond J C, Verbeek B M, Vehmeijer S B W. The first 6 weeks of recovery after primary total hip arthroplasty with fast track. Acta Orthop. 2018;89(1):140. Porsius J T, Mathijssen N M C, Klapwijk-Van Heijningen L C M, Van Egmond J C, Melles M, Vehmeijer S B W. Acta Orthop 2018; 89(6): 597-602.

© 2019 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2019.1576339


Format: 215 × 280 mm

1 Einheit: 7 mm

2/19 ACTA ORTHOPAEDICA

Preventing and reducing periprosthetic joint infections

For anaerobes

Bone cement with 2 antibiotics Broad spectrum of activity For MRSA/MRSE

Infection prevention Implant fixation High stability

High initial antibiotic release

Revision

www.heraeus-medical.com 1810_15791_AZ_COPAL_Produktfamilie_Acta_Orthop_Europe_215x280_EN.indd 1

Vol. 90, No. 2, 2019 (pp. 97–190)

PRODUCTS & SOLUTIONS YOU CAN TRUST.

Volume 90, Number 2, April 2019

ISSN 1745-3674

13.03.19 12:31

Profile for Acta

Acta Orthopaedica Vol. 90 Issue 2, April 2019  

Acta Orthopaedica Vol. 90 Issue 2, April 2019  

Profile for montzka