Proefschrift den Boeft

Medically unexplained physical symptoms in primary care Identification, management and societal aspects

Uitnodiging

Medically unexplained physical symptoms in primary care Identification, management and societal aspects

voor het bijwonen van de openbare verdediging van mijn proefschrift getiteld

Medically unexplained physical symptoms in primary care Identification, management and societal aspects

Donderdag 8 september 2016 om 11.45 in het Auditorium Vrije Universiteit De Boelelaan 1105, Amsterdam Na afloop bent u zeer welkom voor een receptie in de receptieruimte Boelelaanzijde (HG-1, C35A) Hopelijk tot dan!

Madelon den Boeft m_denboeft@yahoo.com 06-28402935

Madelon den Boeft

Madelon den Boeft

Paranimf: Vincent Polak v.g.polak@gmail.com 06-52121322 Paranimf: Chantal Gielen c.l.i.gielen@lumc.nl 06-28126400

Medically unexplained physical symptoms in primary care Identification, management and societal aspects

VRIJE UNIVERSITEIT

Medically unexplained physical symptoms in primary care Identification, management and societal aspects ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam, op gezag van de rector magnificus prof.dr. V. Subramaniam, in het openbaar te verdedigen ten overstaan van de promotiecommissie van de Faculteit der Geneeskunde op donderdag 8 september 2016 om 11.45 uur in het auditorium van de universiteit, De Boelelaan 1105

door Madelon den Boeft geboren te Apeldoorn

promotoren:

prof.dr. H.E. van der Horst

prof.dr. M.E. Numans

copromotor:

dr. J.C. van der Wouden

What if I fall?

Oh, my darling, What if you fly - Erin Hanson -

Look at the stars Look how they shine for you And all the things you do Coldplay – Yellow

Voor mijn allerliefste papa, mama en Han

CONTENTS Chapter 1

General introduction

11

Identification of MUPS Chapter 2

Identifying patients with medically unexplained physical symptoms in electronic medical records in primary care: a validation study

27

Chapter 3

Risk assessment models for patients with persistent medically unexplained physical symptoms in primary care using electronic medical records. An observational study

43

Chapter 4

Recognition of patients with medically unexplained physical symptoms by family physicians: results of a focus group study

63

Structuring management of MUPS Chapter 5

Non-pharmacological interventions for somatoform disorders and medically unexplained physical symptoms in adults (concise version)

85

Chapter 6

How should we manage adults with persistent unexplained physical symptoms?

135

Chapter 7

Negotiating explanations: a qualitative analysis of doctor-patient communication in a general practice clinic for patients with medically unexplained physical symptoms

145

Societal aspects of MUPS Chapter 8

The association between medically unexplained physical symptoms and health care use over two years and the influence of depressive and anxiety disorders and personality traits: a longitudinal study

165

Chapter 9

Medically unexplained physical symptoms and work functioning over two years: their association and the influence of depressive and anxiety disorders and job characteristics

185

Chapter 10

General discussion

203

Summary

219

Samenvatting

225

Dankwoord

231

About the author

237

PhD portfolio

239

List of publications

241

1 GENERAL INTRODUCTION

General introduction

Medically unexplained physical symptoms in primary care Experiencing physical symptoms is a part of the daily life of people. Most of these physical symptoms are transient and people usually do not seek medical care (1,2).

When symptoms persist the general practitioner (GP) is often consulted. Up to thirty per cent of all physical symptoms presented to the GP appear to be so called medically unexplained physical symptoms (MUPS)(3,4), physical symptoms for which no sufficient explanation or evidence for an underlying physical disease can be found after adequate medical examination (5,6). Therefore, MUPS are a common phenomenon in primary care. Also in specialist care, MUPS are a frequent reason for encounter with percentages up to 70 per cent depending on the speciality (7,8). In this thesis we especially focus on MUPS in primary care.

MUPS represent a broad spectrum of symptoms in varying degrees of severity (9,10).

Examples of MUPS are pain, fatigue, dizziness, stomach complaints and musculoskeletal complaints. Sometimes several of the presented symptoms seem to cluster.

In that case we can speak of functional somatic syndromes, such as irritable bowel syndrome, chronic fatigue syndrome and fibromyalgia (11–13). Most MUPS are transient and mild, with little impediments for daily life (2). In 2.5% of the patients MUPS become severe and persistent, which may have negative consequences in many domains of daily life. Most patients with persistent MUPS are functionally impaired, have high levels of psychological distress and report a reduced quality of life (14–16).

Also they are at risk for additional unnecessary examinations and treatments that may be potentially harmful (14–16). Furthermore, persistent MUPS put a burden on the doctor-patient relationship. Patients with MUPS are often dissatisfied with the medical care they receive (17). They feel that their doctors do not take them seriously and that their symptoms are unfairly psychologized (17,18). They often feel the need to fight to gain acknowledgement of their symptoms by their GPs. On the other hand, doctors may feel or are frustrated and indicate that effective treatments to manage patients with MUPS adequately are lacking (18–20). Also they struggle to provide symptom explanations for MUPS (21). They only use a limited repertoire in clinical practice (22). When explanations are provided, they often follow psychosocial

13

1

Chapter 1

models of causality (23), which are mostly rejected by patients and can even be felt as threatening (24,25).

In addition, persistent MUPS may have major impact on society due to high healthcare costs (26,27). Several studies show that patients with (persistent) MUPS use disproportionally large amounts of both somatic and mental healthcare services (28–31). This high healthcare use is often attributed to patients pressurizing their GPs for a somatic treatment, while several studies suggest that patients mostly want support and acknowledgement of the reality of their symptoms, but receive interventions initiated by the GP instead (32). Moreover, MUPS can interfere with work functioning, both in terms of disability at work and long-term and short-term absenteeism from work (33–35). Not being able to work or not performing optimally at work is not only a burden for patients and their direct environment, but also leads to an increase in healthcare costs.

Rationale for this thesis Although much research already has been performed on the subject of MUPS, our knowledge on several aspects needs to be increased to address and constrain the consequences of MUPS for patients, for doctors and for the society as I described above. One of the possibilities to improve care especially for patients with MUPS developing towards persistent MUPS is to offer structured care proactively, coordinated by the GP. By attentively exploring MUPS and their impact on the daily life of patients and by discussing interventions in an early stage, persistent MUPS may be prevented. The GP, or multidisciplinary organized primary care centres in which GPs are in the lead, eminently come into sight as the right place from which to coordinate this kind of proactive care as the GP knows his/her patients and usually has a longterm relationship with them within their context. Also the GP is in the position to check on his/her patients regularly and can offer continuity of care.

The first step in structured care is to identify patients with MUPS evolving towards a chronic condition in an early stage. Early identification of patients with MUPS could lead to more attention for better communication during the consultations and providing explanations for patients’ symptoms. Also it urges GPs to consider and discuss

14

General introduction

possible treatment strategies in an earlier stage, taking into account the wishes and needs of the patient. This kind of proactive care, based on risk assessment followed by identification of care gaps among subpopulations of patient who share predefined risks is called panel management. Specific MUPS guidelines provide tools for organizing structured care in clinical practice in this process (5,6).

This thesis is divided into three parts, each focusing on a different part of structured care.

First, I will focus on different methods to identify patients with MUPS and specifically those at risk for persistent MUPS. Second, I will outline the existing treatments and potential gaps in research and practice. Third, I will bring more insight into the relevance of societal consequences of MUPS over time.

Identification of MUPS Early identification of patients with MUPS The identification of patients with MUPS and especially those patients at risk for persistent MUPS is a difficult task both in research and in clinical practice. The biggest challenges for the identification are the lack of knowledge how doctors identify patients with MUPS during consultations, the different MUPS definitions used in literature (11), the heterogeneity among patients with MUPS (9,10) and, although inherent to the definition problem, the lack of a generic MUPS code in GP coding systems like the International Classification of Primary Care (ICPC)(36).

One of the methods for the identification of various patient populations or populations at risk, which is currently increasingly and often used in primary care research, is risk assessment based on advanced analysis of routine electronic medical records (EMRs). Routine EMRs refer to the systematized registration of medical information of patients and populations that are stored electronically during daily practice. They include a broad range of data, namely demographic information, characteristics such as age and gender, medical history, medications, allergies, additional diagnostics and referral correspondence. These EMRs can be used in scientific research when data are anonymised and their use in earlier research has proved to be effective and feasible in various risk populations such as frail elderly and chronically diseased

15

1

Chapter 1

patients (37–40). The advantages of using EMRs are that patient data are directly available, that often no additional data collection is needed and that data of patients usually not being able or willing to agree with analyses in identifiable datasets can also be used. Also, once reliable algorithms become available, individual data can be used for proactive periodical screening purposes as it can provide a quick overview of populations at risk. Regarding MUPS, only a few studies used EMRs for identification. Unfortunately, in these studies either the method was not directly suitable for primary care purposes or it was not useful for screening purposes (41,42). Therefore, better models are needed and should be developed. In this thesis we aimed: • To validate a recently developed EMR screening method from Utrecht, the Netherlands, to identify patients with MUPS (Chapter 2) • To develop risk assessment models using different (advanced) statistical techniques applied to routine primary care EMRs to identify patients at risk for persistent MUPS (Chapter 3) How do doctors identify the heterogeneous group of patients with MUPS? It is known that patients with MUPS constitute a heterogeneous group of patients.

Not only due to the broad range of clinical symptoms, but also due to the variety in sociodemographic characteristics such as age, employment status, educational level and mental health comorbidity (9,10). Therefore they are coded in the routine EMRs

by GPs in different ways. The heterogeneity of MUPS and their registration not only impedes the process of identification but also the fine-tuning of treatment. It may well be possible that the varying and disappointing treatment outcomes are partly due to this heterogeneity, as different groups of patients may benefit from different types of treatment. In previous studies this point was underlined. In patients with fibromyalgia, the authors identified two subgroups: patients with pain avoidance and patients with pain persistence. They concluded that these subgroups benefitted from a different treatment approach (43–45). Two other studies highlighted the relevance of the heterogeneity among patients with chronic fatigue syndrome for their treatment response and the need to explore this heterogeneity more in to depth (46,47).

In the light of these previous studies and the scarcity of effective treatments for

16

General introduction

patients with MUPS (48), more insight should be gained into how doctors identify patients with MUPS during consultations and which subgroups they distinguish. This could help to develop more targeted interventions and is of additional value for research to include the correct population. In this thesis we aimed: • To examine how GPs recognize MUPS in their patients during consultations and which subgroups of patients with MUPS can be recognized (Chapter 4)

Structuring management of MUPS How to structure management of patients identified with MUPS? In the past decades, many strategies for the approach of treatment for MUPS were developed and studied. Up to now, effective treatment strategies in primary care are still lacking and many aspects of the treatments, such as the optimal duration and deliverer of treatment, are still unclear.

In a Cochrane review on enhanced care (where the GP provides cognitive behavioural techniques), the authors concluded that current evidence does not answer the question whether enhanced care delivered by front line primary care professionals has an effect on the outcome of patients with MUPS (49). Another Cochrane review studied the effectiveness of pharmacological interventions, such as antidepressants, anti-epileptic drugs and natural products (50). Although some positive effects were found, caution is needed because of frequently occurring side effects. There are studies investigating the effect of non-pharmacological interventions regarding MUPS (51-55), including two reviews (56,57), but a comprehensive overview of the whole spectrum is missing. In this thesis we aimed: • To perform a Cochrane review in which we will assess the effects of nonpharmacological interventions for somatoform disorders and chronic MUPS to assist health care providers to make optimal treatment decisions and to highlight current gaps in research literature (Chapter 5) • To outline how doctors should manage patients with persistent MUPS in the light of the existing uncertainties (Chapter 6)

17

1

Chapter 1

Providing symptom explanations A specific part of MUPS management is providing acceptable explanations to patients. The explanation phase of the consultation is essential because this builds a bridge between the story of the patient, physical examination and treatment. It is known that most patients seek explanations for their symptoms (58,59), but they often find that their doctors are unable to deliver them (21). Commonly used explanations often follow psychosocial models of causality. An example is reattribution, where symptoms are linked to psychological distress (23). However, these explanations are not fully compatible with current models of symptom persistence (60,61) and are commonly resisted by patients (24,25). Therefore more explanation tools are necessary, as well as insight into how the dialogue between patients and doctors surrounding explanations evolves and how patients subsequently react. In this thesis we aimed: • To carry out detailed analysis of dialogues structure of symptom explanations between patients with MUPS and GPs and reactions of patients (Chapter 7).

Societal consequences of MUPS In an era where healthcare costs are steadily rising, it is important to acknowledge that persistent MUPS have a major contribution on these costs due to high healthcare use. In addition persistent MUPS also have a negative influence on work functioning (26,27). Therefore, persistent MUPS should be prevented or limit their societal consequences as much as possible. In order to do so, more insight is needed into the associations between MUPS and healthcare use and work functioning over time and into potential influencing factors.

There are already several studies exploring these associations, but they are mostly cross-sectional or retrospective (28,29,34,35,62), which limits the interpretability of the results.

We already know that MUPS frequently co-occur with mental health disorders such as depressive and anxiety disorders and that these mental health disorders also have impact on healthcare use and costs (63–66). Therefore it has to be explored

18

General introduction

what the extent is of mental health disorders on the association between MUPS and healthcare use and work functioning. Finally, specifically regarding work functioning, studies have shown that unfavourable job characteristics such as long working hours and a low occupational status may have a negative influence on someone’s work functioning in general. It is yet not know in which extent this also contributes to patients with MUPS. More insight in these relationships including the influencing factors can play a major role in the development of prevention and management strategies for MUPS. In this thesis we aimed: • To examine the association between MUPS and healthcare use over two years and the influence of depressive and anxiety disorders and personality traits on this association (Chapter 8) • To assess the association between MUPS and work functioning over two years and the influence of job characteristics and depressive and anxiety disorders on this association (Chapter 9)

19

1

Chapter 1

Outline of this thesis In Chapter 2 we show the results from a validation study in which we explored the test characteristics of an EMR screening method, developed in Utrecht, the Netherlands, to identify patients with MUPS in routine primary care EMRs. We compared the identified patients with MUPS from the screening method with their scores on the patient health questionnaire-15, which we used as a reference test.

In Chapter 3 we show the results of two risk assessment models that we developed by using data mining and different (advanced) statistical techniques in primary care EMRs with the purpose to identify those patients at risk for persistent MUPS. In Chapter 4 we present the results of a focus group study among GPs who were asked how they recognize patients with MUPS and which subgroups of patients they distinguished.

Chapter 5 is a concise version of a Cochrane review about the effects of non-pharmacological interventions for somatoform disorders and chronic MUPS in adults.

Chapter 6 contains our recommendations based on the four Cochrane reviews about MUPS on how to manage these patients in the light of the existing uncertainties.

Chapter 7 provides insight into how GPs and patients discuss explanations for MUPS. For this study we analysed audiotaped consultations that were held in Scotland.

In Chapter 8 we present the results of the study where we explored the association between MUPS and healthcare use over two years and the influence of depressive and anxiety disorders and specific personality traits on this association. We used data from the Netherlands Study of Depression and Anxiety.

In Chapter 9 we present the results of the study where we explored the association between MUPS and work functioning over two years and the influence of depressive

20

General introduction

and anxiety disorders and job characteristics. Again we used data from the Netherlands Study of Depression and Anxiety.

In Chapter 10 I provide a critical appraisal of our study results and methodological considerations. Also I give recommendations for daily clinical practice and future research.

21

1

Chapter 1

REFERENCES 1. Green L, Fryer G Jr, Yawn B, et al. The ecology of medical care revisited. N Engl J Med. 2001;344:2021–5.

2. van der Linden MW, Westert GP, de Bakker DH, et al. Tweede nationale studie naar ziekten en verrichtingen in de huisartspraktijk: klachten en aandoeningen in de bevolking en in de huisartspraktijk (second national survey of morbidity and interventions in general practice: symptoms and diseases in the population). Utrecht/Bilthoven: NIVEL/ RIVM, 2004.

3. Jackson JL, Passamonti M. The outcomes among patients presenting in primary care with a physical symptom at 5 years. J Gen Intern Med. 2005;20:1032–7.

4. Khan AA, Khan A, Harezlak J, et al. Somatic symptoms in primary care: etiology and outcome. Psychosomatics. 2003;44:471–8.

5. van der Feltz-Cornelis CM, Hoedeman R, Keuter EJW, et al. Presentation of the multidisciplinary guideline medically unexplained physical symptoms and somatoform disorder in the Netherlands: disease management according to risk profiles. J Psychosom Res. 2012;72:168–9.

6. olde Hartman T, Blankenstein N, Molenaar B. NHG-Standaards onvoldoende verklaarde lichamelijke klachten. Huisarts Wet. 2013;5:222–30.

7. Nimnuan C, Hotopf M, Wessely S. Medically unexplained symptoms: An epidemiological study in seven specialities. J Psychosom Res. 2001;51:361–7.

8. Reid S, Wessely S, Crayford T, Hotopf M. Frequent attenders with medically unexplained symptoms: service use and costs in secondary care. Br J Psychiatry. 2002;180:248–53.

9. Weijden T van der, Velsen M van, Dinant GJ, et al. Unexplained complaints in general practice: prevalence, patients’ expectations, and professionals’ test-ordering behaviour.

Med Decis Making. 2003;23:226–31.

10. Aamland A, Malterud K, Werner EL. Patients with persistent medically unexplained physical symptoms: a descriptive study from Norwegian general practice. BMC Fam Pract. 2014;15:107.

11. Fink P, Rosendal M, Olesen F. Classification of somatization and functional somatic symptoms in primary care. Aust N Z J Psychiatry. 2005;39:772–81.

12. Wessely S, Nimnuan C, Sharpe M. Functional somatic syndromes: one or many? The Lancet. 1999;354:936–9.

13. Bazelmans E, Vercoulen J, Swanink CMA, et al. Chronic fatigue syndrome and primary fibromyalgia syndrome as recognized by GPs. Fam Pract. 1999;16:602–4.

14. Barsky AJ BJ. Somatization and medicalization in the era of managed care. JAMA. 1995;274:1931–4.

15. Verhaak PFM, Meijer SA, Visser AP, et al. Persistent presentation of medically unexplained symptoms in general practice. Fam Pract. 2006;23:414–20.

16. Smith GR, Monson RA, Ray DC. Patients with multiple unexplained symptoms. Their characteristics, functional health, and health care utilization. Arch Intern Med. 1986;146:69–72.

17. Page LA, Wessely S. Medically unexplained symptoms: exacerbating factors in the doctor-patient encounter. J R Soc Med. 2003;96:223–7.

22

General introduction

18. Malterud K. Symptoms as a source of medical knowledge: understanding medically unexplained disorders in women. Fam Med. 2000;32:603–11.

19. Reid S, Whooley D, Crayford T, et al. Medically unexplained symptoms--GPs’ attitudes towards their cause and management. Fam Pract. 2001;18:519–23.

20. Woivalin T, Krantz G, Mäntyranta T, et al. Medically unexplained symptoms: perceptions of physicians in primary health care. Fam Pract. 2004;21:199–203.

21. olde Hartman TC, Hassink-Franke LJ, Lucassen PL, et al. Explanation and relations.

How do general practitioners deal with patients with persistent medically unexplained symptoms: a focus group study. BMC Fam Pract. 2009;10:68.

22. olde Hartman TC, van Rijswijk E, van Dulmen S, et al. How patients and family physicians communicate about persistent medically unexplained symptoms. A qualitative study of video-recorded consultations. Patient Educ Couns. 2013;90:354-60.

23. Gask L, Dowrick C, Salmon P, et al. Reattribution reconsidered: narrative review and reflectionson an educational intervention for medically unexplained symptoms in primary care settings. Psychosom Res. 2011;71:325–34.

24. Peters S, Rogers A, Salmon P, et al. What do patients choose to tell their doctors? qualitative analysis of potential barriers to reattributing medically unexplained symptoms. J Gen Intern Med. 2009;24:443–9.

25. Burbaum C, Stresing A-M, Fritzsche K, et al. Medically unexplained symptoms as a threat to patients’ identity?: A conversation analysis of patients’ reactions to psychosomatic attributions. Patient Educ Couns. 2010;79:207–17.

26. Konnopka A, Kaufmann C, König HH, et al. Association of costs with somatic symptom severity in patients with medically unexplained symptoms. J Psychosom Res. 2013;75:370–5.

27. Konnopka A, Schaefert R, Heinrich S, et al. Economics of medically unexplained symptoms: a systematic review of the literature. Psychother Psychosom. 2012;81:265–75.

28. Barsky AJ OE. Somatization increases medical utilization and costs independent of psychiatric and medical comorbidity. Arch Gen Psychiatry. 2005;62:903–10.

29. Fink P, Sørensen L, Engberg M, et al. Somatization in primary care. Prevalence, health care utilization, and general practitioner recognition. Psychosomatics. 1999;40:330–8.

30. Andersen NLT, Eplov LF, Andersen JT, et al. Health care use by patients with somatoform disorders: a register-based follow-up study. Psychosomatics. 2013;54:132–41.

31. Grabe HJ, Baumeister SE, John U, et al. Association of mental distress with health care utilization and costs: a 5-year observation in a general population. Soc Psychiatry Psychiatr Epidemiol. 2009;44:835–44.

32. Ring A, Dowrick C, Humphris G, et al. Do patients with unexplained physical symptoms pressurise general practitioners for somatic treatment? A qualitative study. BMJ. 2004;328:1057.

33. Rask MT, Rosendal M, Fenger-Grøn M, et al. Sick leave and work disability in primary care patients with recent-onset multiple medically unexplained symptoms and persistent somatoform disorders: a 10-year follow-up of the FIP study. Gen Hosp Psychiatry.

2015;37:53–9.

34. Hoedeman R, Krol B, Blankenstein N, et al. Severe MUPS in a sick-listed population: a cross-sectional study on prevalence, recognition, psychiatric co-morbidity and impairment. BMC Public Health. 2009;9:440.

23

1

Chapter 1

35. Hoedeman R, Blankenstein AH, Krol B, et al. The contribution of high levels of somatic symptom severity to sickness absence duration, disability and discharge. J Occup Rehabil. 2010;20:264–73.

36. Lamberts H, Wood M, Hofmans-Okkes IM. International primary care classifications: the effect of fifteen years of evolution. Fam Pract. 1992;9:330–9.

37. Silow-Carroll S, Edwards JN, Rodin D. Using electronic health records to improve quality and efficiency: the experiences of leading hospitals. Issue Brief Commonw Fund.

2012;17:1–40.

38. Loo TS, Davis RB, Lipsitz LA, et al. Electronic medical record reminders and panel management to improve primary care of elderly patients. Arch Intern Med. 2011;171:1552–8.

39. Afzal Z, Engelkes M, Verhamme KMC, et al. Automatic generation of case-detection algorithms to identify children with asthma from large electronic health record databases. Pharmacoepidemiol Drug Saf. 2013;22:826-33 40. Benhamou P-Y. Improving diabetes management with electronic health records and patients’ health records. Diabetes Metab. 2011;37:53–6.

41. Morriss R, Lindson N, Coupland C, et al. Estimating the prevalence of medically unexplained symptoms from primary care records. Public Health. 2012;126:846–54.

42. Smith RC, Gardiner JC, Armatti S, et al. Screening for high utilizing somatizing patients using a prediction rule derived from the management information system of an HMO: a preliminary study. Med Care. 2001;39:968–78.

43. van Koulil S, Kraaimaat FW, van Lankveld W, et al. Screening for pain-persistence and pain-avoidance patterns in fibromyalgia. Int J Behav Med. 2008;15:211–20.

44. van Koulil S, van Lankveld W, Kraaimaat FW, et al. Tailored cognitive-behavioral therapy for fibromyalgia: two case studies. Patient Educ Couns. 2008;71:308–14.

45. Turk DC, Okifuji A, Sinclair JD, et al. Differential responses by psychosocial subgroups of fibromyalgia syndrome patients to an interdisciplinary treatment. Arthritis Care Res Off J Arthritis Health Prof Assoc. 1998;11:397–404.

46. Cella M, Chalder T, White PD. Does the heterogeneity of chronic fatigue syndrome moderate the response to cognitive behaviour therapy? An exploratory study. Psychother Psychosom. 2011;80:353–8.

47. White PD, Goldsmith K, Johnson AL, et al. Recovery from chronic fatigue syndrome after treatments given in the PACE trial. Psychol Med. 2013;43:2227–35.

48. van Dessel N, den Boeft M, van der Wouden JC, et al. Non-pharmacological interventions for somatoform disorders and medically unexplained physical symptoms in adults.

2014;11:CD011142 49. Rosendal M, Blankenstein AH, Morriss R, et al. Enhanced care by generalists for functional somatic symptoms and disorders in primary care. Cochrane Database Syst Rev. 2013;10:CD008142.

50. Kleinstäuber M, Witthöft M, Steffanowski A, et al. Pharmacological interventions for somatoform disorders in adults. Cochrane Database Syst Rev. 2014;11:CD010628.

51. Morriss R, Dowrick C, Salmon P, et al. Cluster randomised controlled trial of training practices in reattribution for medically unexplained symptoms. Br J Psychiatry.

2007;191:536–42.

52. Larisch A, Schweickhardt A, Wirsching M, et al. Psychosocial interventions for somatizing patients by the general practitioner: A randomized controlled trial. J Psychosom Res. 2004;57:507–14.

24

General introduction

53. Schröder A, Rehfeld E, Ørnbøl E, et al. Cognitive–behavioural group treatment for a range of functional somatic syndromes: randomised trial. Br J Psychiatry. 2012;200:499–507.

54. Fjorback LO, Carstensen T, Arendt M, et al. Mindfulness therapy for somatization disorder and functional somatic syndromes: Analysis of economic consequences alongside a randomized trial. J Psychosom Res. 2013;74:41–8.

55. van Ravesteijn HJ, Suijkerbuijk YB, Langbroek JA, et al. Mindfulness-based cognitive therapy for patients with medically unexplained symptoms: process of change. J Psychosom Res. 2014;77:27–33.

56. Burton C. Beyond somatisation: a review of the understanding and treatment of medically unexplained physical symptoms. Br J Gen Pract. 2003;53:231–9.

57. Kleinstäuber M, Witthöft M, Hiller W. Efficacy of short-term psychotherapy for multiple medically unexplained physical symptoms: a meta-analysis. Clin Psychol Rev. 2011;31:146–60.

58. Nettleton S. “I just want permission to be ill”: Towards a sociology of medically unexplained symptoms. SocSciMed. 2006;62:1167–78.

59. Giroldi E, Veldhuijzen W, Mannaerts A, et al. “Doctor, please tell me it’s nothing serious”: An exploration of patients’ worrying and reassuring cognitions using stimulated recall interview. BMC Fam Pract. 2014;15.

60. Deary V, Chalder T, Sharpe M. The cognitive behavioural model of medically unexplained symptoms: A theoretical and empirical review. Clin Psychol Rev. 2007;27:781–97.

61. Rief W, Broadbent E. Explaining medically unexplained symptoms - models and mechanisms. Clin Psychol Rev. 2007;27:821–41.

62. Williams ER, Guthrie E, Mackway-Jones K, et al. Psychiatric status, somatisation, and health care utilization of frequent attenders at the emergency department: a comparison with routine attenders. J Psychosom Res. 2001;50:161–7.

63. Burton C, McGorm K, Weller D, et al. Interpretation of low mood and worry by high users of secondary care with medically unexplained symptoms. BMC Fam Pract. 2011;12:107.

64. van Boven K, Lucassen P, van Ravesteijn H, olde Hartman T, Bor H, van Weel-Baumgarten E, et al. Do unexplained symptoms predict anxiety or depression? Ten-year data from a practice-based research network. Br J Gen Pract. 2011;61:316–25.

65. Kroenke K, Jackson JL, Chamberlin J. Depressive and anxiety disorders in patients presenting with physical complaints: clinical predictors and outcome. Am J Med. 1997;103:339–47.

66. Henningsen P, Zimmermann T, Sattel H. Medically Unexplained Physical Symptoms, Anxiety, and Depression: A Meta-Analytic Review. Psychosom Med. 2003;65:528–33.

25

1

2 Identifying patients with medically unexplained physical symptoms in electronic medical records in primary care: a validation study Madelon den Boeft Johannes C. van der Wouden Trudie R. Rydell-Lexmond Niek J. de Wit Henriette E. van der Horst Mattijs E. Numans

BMC Family Practice 2014;15:109

Chapter 2

ABSTRACT Background When medically unexplained physical symptoms (MUPS) become persistent, it may have major implications for the patient, the general practitioner (GP) and for society.

Early identification of patients with MUPS in electronic medical records (EMRs) might contribute to prevention of persistent MUPS by creating awareness among GPs and providing an opportunity to start stepped care management. Procedures for identification of patients with MUPS in EMRs are not well established yet. In this validation study we explore the test characteristics of an EMR screening method to identify patients with MUPS. Methods The EMR screening method consists of three steps. First, all patients ≥18 years were included when they had five or more contacts in the last 12 months. Second, patients with known chronic conditions were excluded. Finally, patients were included with a MUPS syndrome or when they had three or more complaints suggestive for MUPS. We compared the results of the EMR screening method with scores on the Patient Health Questionnaire-15 (PHQ-15), which we used as reference test. We calculated test characteristics for various cut-off points.

Results From the 1223 patients in our dataset who completed the PHQ-15, 609 (49/8%) scored ≥5 on the PHQ-15. The EMR screening method detected 131/1223 (10.7%) as patients with MUPS. Of those, 102 (77.9%) scored ≥5 on the PHQ-15 and 53 (40.5%) scored ≥10. When compared with the PHQ-15 cut-off point ≥10, sensitivity and specificity were 0.30 and 0.93 and positive and negative predictive values were 0.40 and 0.89, respectively.

Conclusions The EMR screening method to identify patients with MUPS has a high specificity, but many potential MUPS patients will be missed. Before using this method as a

28

Identifying MUPS in electronic medical records

screening instrument for selecting patients who might benefit from structured care, its sensitivity needs to be improved while maintaining its specificity.

BACKGROUND Presentation of medically unexplained physical symptoms (MUPS) is a common phenomenon in primary care. Of all primary care encounters, in up to a third the symptoms presented by the patient remain unexplained [1,2]. In specialist care, these figures may even be higher, depending on the specialty [3]. Although MUPS become persistent in only a minority (2.5%) of patients, the burden of persistent MUPS is high for both patients and doctors and for society [4]. Patients are functionally impaired and may feel that they are not taken seriously by their general practitioner (GP) [5-7]. Furthermore, the doctor-patient relationship is often troubled and many GPs indicate that they find these patients difficult to manage [8,9]. Also persistent MUPS may lead to high and inadequate health care utilization and high associated costs [10-12].

Early identification of patients with a higher risk of developing persistent MUPS in routine electronic medical records (EMRs) could create an opportunity for proactive and structured care, taking into account the severity of MUPS, coordinated by GPs. Awareness among GPs of their population at risk could result in more attention during consultations or in offering effective interventions like cognitive behaviour therapy at an earlier stage if appropriate [13]. The advantage of using EMRs is that the data are directly available and no additional data collection is needed, which saves time-consuming logistical procedures. Furthermore it provides a quick overview of a population at risk.

Early identification in EMRs proved to be feasible and effective for other risk populations, like patients with type 2 diabetes, cardiovascular risks and frail elderly [14-16] as well as for preventive health care [17]. Also Tian et al. developed an applicable EMR algorithm to identify patients with chronic pain [18]. However, identifying patients with MUPS is not an easy task as there is no generally accepted procedure available.

Although some MUPS characteristics, like frequent consultation and referral rate, can be obtained from EMRs, there is no international classification of primary care (ICPC)

29

2

Chapter 2

code available that identifies the combination of symptoms that characterize MUPS of various MUPS subgroups.

Morriss et al. developed an EMR model that estimates the prevalence of MUPS. However, they concluded that the model is not useful for screening purposes due to a low sensitivity [19]. Various other methods for MUPS screening have been developed and studied. Kroenke et al. showed in their validation study that the self-administered Patient Health Questionnaire-15 (PHQ-15) could be used for screening somatization and somatic symptom severity including MUPS [20]. However, the PHQ-15 cannot be easily obtained from EMRs. Verhaak et al. used criteria composed by Robbins et al. to estimate the prevalence of persistent MUPS, but in their study it is about the patients who already suffer from persistent MUPS and not about the patients at risk [4,21].

In 2010, a cross-sectional study focusing on the prevalence of MUPS was conducted in the Utrecht Health Project. Patients with MUPS were identified using EMR data in three subsequent selection steps. In our current study we aim to validate this EMR screening method to identify MUPS patients by comparing it to the commonly used and validated PHQ-15.

METHODS Setting and study population The Utrecht Health Project is a primary care population study with the purpose to enable research into the impact of changes in health care policy, developments in public health and quality management, as well as to support population research into determinants of health and disease [22]. From 2000 on, all inhabitants of the new neighbourhood Leidsche Rijn near Utrecht enlisted with local GPs were invited to participate. Various health measurements and questionnaires were collected after informed consent, including the PHQ-15 since 2005, and they were linked with follow-up data extracted from EMRs. For our current study we were able to use EMRs from 1223 patients 18 years or older who completed the PHQ-15 in 2005–2007. We retrieved EMR data for each patient over the 12 months preceding the completion of the PHQ-15.

30

Identifying MUPS in electronic medical records

Study design and procedure In this validation study we compared the results of the EMR screening method for detecting possible MUPS patients with the PHQ-15 scores of the patients as a reference standard. We defined patients with scores from cut-off point ≥5 on the PHQ-15 as having MUPS, following Kroenke et al. [20].

The EMR screening method The EMR screening method identifies patients with possible MUPS with three subsequent selection steps. First step: Patients ≥18 years with ≥5 general practice consultations during the past 12 months were selected, as it is known that MUPS patients usually have relatively high consultation rates [11]. Second step: patients with chronic obstructive pulmonary disease, hypertension or diabetes mellitus and patients with an established psychiatric diagnosis were excluded, in order to exclude all patients in whom physical symptoms are medically explained and because frequent practice visits might at forehand be assumed for these conditions [23,24].

Third step: In the remaining group, patients were selected who consulted the GP with one of the three MUPS syndromes; irritable bowel syndrome (ICPC D93), fibromyalgia (ICPC L18.01) and chronic fatigue syndrome (ICPC A04.01). This group was called “Syndrome-based Confirmed-MUPS”. Furthermore we selected all patients who had three or more contacts with at least one of a list of 104 ICPC codes suggestive of MUPS, as assessed by the GPs during regular care (symptom diagnoses; for the additional file see full online article). This group was called “High-risk-MUPS”.

“High-risk-MUPS” and “Syndrome-based Confirmed MUPS” together formed the complete MUPS risk population resulting from these selection steps.

Patient health questionnaire-15 Internationally, the PHQ-15 (Figure 1) is a widely used and validated mental health screening instrument to assess the severity of somatic symptoms. It is based on the “Primary Care Evaluation of Mental Disorders” (PRIME-MD), a diagnostic instrument for common mental health disorders and the “PRIME-MD Patient Health Questionnaire” [25,26]. It inquires into 15 symptoms or symptoms clusters that account for more than 90 per cent of physical complaints. Thirteen somatic symptoms of

31

2

instrument to assess the severity of somatic symptoms. It is based on the “Primary Chapter 2

Evaluation of Mental Disorders” (PRIME-MD), a diagnostic instrument for commo

health disorders and the “PRIME-MD Patient Health Questionnaire” [25,26]. It inq

symptoms or symptoms clusters that account for more than 90 per cent of physica

Thirteen somatic symptoms of the PRIME-MD are included in the PHQ-15 and two

are part of the PHQ depression module. For each item, there are three options to s

the PRIME-MD are included in the PHQ-15 and two symptoms are part of the PHQ

severity of complaints; zero (not bothered at all), one (bothered a little) and two (

depression module. For each item, there are three options to score in severity of

lot), resulting in overall score ranging from zero to 30. For all symptoms, a score o

complaints; zero (not bothered at all), one (bothered a little) and two (bothered a

considered severe. twozero studies wasallconcluded theofPHQ-15 is a valid and lot), resulting in overall score rangingInfrom to 30.it For symptoms, that a score

two is considered severe.

In two studies that it wasmay concluded that PHQ-15 is a valid reliable questionnaire be used tothe detect MUPS. Therefore we used this qu and moderately reliable questionnaire that may be used to detect MUPS. Therefore

the reference method in our study [20,27].

we used this questionnaire as the reference method in our study [20,27].

Figure 1. The patient health Figure 1. questionnaire-15 The patient health questionnaire-15

32

Identifying MUPS in electronic medical records

Ethical approval All participants in the Utrecht Health Project gave informed consent for linking their anonymous EMRs to the PHQ-15. The medical ethical committee of University Medical Center Utrecht approved the original protocol of the Utrecht Health Project and its amendments (file#99-240). The current study made use of readily available data and did not require additional informed consent or ethical approval.

Statistical analysis In our dataset consisting of 1223 patients who completed the PHQ-15, we calculated the prevalence of MUPS patients following the EMR screening method. We dichotomized the continuous PHQ-15 outcome by using two cut-off points for mild and medium MUPS; ≥5 and ≥10 respectively, as used by Kroenke et al.[20]. Less than 2% of all entries were missing. Missing data in the PHQ-15 were imputed. Sensitivity analysis showed only minor differences between complete and imputed cases. The multiple imputation model included age, gender, total number of contacts, all 15 PHQ-15 questions and the outcome variable MUPS. Cross tabulations enabled us to calculate sensitivity, specificity, predictive values and likelihood ratios for the two cut-off points including 95% confidence intervals. All analyses were processed with SPSS version 20.0.

RESULTS Prevalence of MUPS We assessed the prevalence of the MUPS risk population in our dataset of 1223 adult patients, consisting of 756 women (61.8%) and 467 men (38.2%) by carrying out the described steps. The mean age was 38.8 years. Twenty-one patients (1.7%) were identified as “Confirmed MUPS”. All 21 were diagnosed with irritable bowel syndrome, for which they had had at least one consultation in the 12 months period.

There were no patients with an ICPC code for chronic fatigue syndrome or fibromyalgia. The EMR screening method identified 126 patients (10.3%) as “High-risk-MUPS”.

Most patients with irritable bowel syndrome also had at least one ICPC code suggestive of MUPS. Together, the total MUPS prevalence of both groups combined in this

33

2

Chapter 2

population according to the EMR method was 131 (10.7%). Of those, 93 (71%) were women, significantly more than men (p = 0.04).

PHQ-15 outcomes In the total population, 609 patients (49.8%) scored ≥5 on the PHQ-15 and 176 (14.4%) ≥10. The PHQ-15 results were skewed (skewness 1.27; Kolmogorov-Smirnov p < 0.001) with a mean of 5.29 and a median of 4.0. In the MUPS group selected by the EMR screening method, 102/131 (77.9%) patients scored ≥5 on the PHQ-15 and 53/131 (40.5%) scored ≥10. Again, the distribution was skewed (skewness 0.51; Kolmogorov-Smirnov P < 0.001) with a mean and median of 8.57 and 8.0, respectively. Of all 21 patients with at least one contact for irritable bowel syndrome, 19 (90.5%) scored ≥5 on the PHQ-15 and 13 (61,9%) scored ≥10.

Test characteristics of the EMR screening method compared with the PHQ-15 For cut-off point ≥5, sensitivity and specificity of the EMR screening method were 0.17 and 0.95, respectively. The likelihood ratios for a positive and negative test were 3.54 and 0.87, respectively. Positive and negative predictive values were 78% and 54%, respectively. For the cut-off point ≥10, sensitivity and specificity were 0.30 and 0.93, respectively. The likelihood ratio for a positive test was 4.29, for a negative test 0.75 and positive and negative predictive values were 40% and 89%, respectively (Tables 1, 2, 3).

Table 1. Two-by-two table of PHQ-15 cut-off point 5 PHQ-15 ≥ 5

PHQ-15 < 5

Total

EMR screening method ‘MUPS’

102

29

131

EMR screening method ‘no MUPS’

507

585

1092

Total

609

614

1223

PHQ-15 < 10

Total

Table 2. Two-by-two table of PHQ-15 cut-off point 10 PHQ-15 ≥ 10 EMR screening method ‘MUPS’

53

78

131

EMR screening method ‘no MUPS’

123

969

1092

Total

176

1047

1223

34

Identifying MUPS in electronic medical records

Table 3. Comparing the EMR screening method with the PHQ-15 cut-off scores PHQ-15 ≥ 5 (95% confidence interval)

PHQ-15 ≥ 10 (95% confidence interval)

Sensitivity

0.17 (0.14 - 0.20)

0.30 (0.24 – 0.38)

Specificity

0.95 (0.93 - 0.97)

0.93 (0.91 – 0.95)

Positive predictive value

0.78 (0.71 – 0.85)

0.40 (0.32 – 0.49)

Negative predictive value

0.54 (0.51 – 0.57)

0.89 (0.87 – 0.91)

Likelihood ratio positive test

3.54 (2.38 – 5.27)

4.29 (2.96 – 5.51)

Likelihood ratio negative test

0.87 (0.84 – 0.91)

0.75 (0.69 – 0.83)

2

DISCUSSION Main findings The aim of our study was to validate the EMR screening method to identify MUPS patients using the PHQ-15 as a reference test in order to map a specific and heterogeneous population at risk that might benefit from structured and stepped care. We found a prevalence of 10.7% with the EMR screening method compared to a high prevalence of 49.8% with the PHQ-15 cut-off ≥5. Most MUPS patients identified by the EMR screening method and patients with IBS scored at least 5 points on the PHQ15. Test characteristics showed a high specificity but a low sensitivity for both PHQ cut-off points, which indicates that about 80% of patients with MUPS were missed.

Interpretation of results The prevalence of MUPS has been frequently studied and varies greatly. In most studies, percentages range around 30 per cent in primary care [1,5,28]. In our study, almost half of all patients scored positive on the PHQ-15 cut-off ≥5, suggesting that many patients in this group of patients probably have incidental complaints and will not benefit from proactive care. The prevalence of 10.7% found by the EMR screening method is lower. The main reason for the difference between our results and existing literature seems to be that the EMR screening method is rather stringent. Various other reasons can also account for the difference. First, the quality of registration in the participating practices may be suboptimal. In this study, only 21 (1.7%) patients

35

Chapter 2

were found with an ICPC code for irritable bowel syndrome, a much lower prevalence than what is known from research, namely 14 to 24 per cent of women and five to 19 per cent of men [29]. However, our findings are consistent with the results from other Dutch studies in routine healthcare data [30]. We did not find patients with a coded diagnosis of chronic fatigue syndrome or fibromyalgia, where one or two could be expected [31]. Patients with chronic fatigue or chronic widespread pain, closely related to fibromyalgia, might have been recorded with other diagnostic terms than L18.01 (fibromyalgia) or A04.01 (chronic fatigue syndrome). These patients will be found in the third step of our selection where MUPS suggestive codes are selected, such as fatigue (A04), general pain (A01) and muscle pain (L18).

Second, we only considered ICPC codes registered during the year preceding the patients’ PHQ-15 score. We did not include patients with a MUPS suggestive or MUPS syndrome ICPC code registered before that time which could have resulted in false negatives. Finally, all patients with known chronic somatic or psychiatric comorbidity were excluded, while studies show that especially those patients more often suffer from unexplained symptoms [23,24].

The high specificity but low sensitivity can partly be explained by the fact that patients do not present all symptoms to GPs and GPs do not code all presented symptoms in their EMRs. Furthermore, the doctor’s diagnostic label is a reflection of the symptom the patient presented and it is his understanding of the situation.

Strengths and limitations Because the only selection criteria of our research population were if they lived in a certain area (Leidsche Rijn) and completed the PHQ-15, we minimized selection bias and response tendencies. We had more women than men in our study because women completed the PHQ-15 more often than men. This is consistent with gender differences in the number of GP encounters in the Netherlands. We also found a significant difference between men and women in the prevalence of MUPS, which is also consistent with other studies [32,33].

Three study limitations should be noted. The first is that no gold standard is available for defining a ‘true’ MUPS population. In the end, only the physician decides whether the patients’ symptoms are medically explained or not, entailing a certain amount of

36

Identifying MUPS in electronic medical records

subjectivity. We chose to use the PHQ-15 because of its availability and as a second best reference standard after the physician’s judgement as this self-administered questionnaire has been validated for clinical practice and research for screening and monitoring MUPS and somatoform disorders by Kroenke and van Ravensteijn [20,27]. Kroenke et al. concluded that high total scores strongly correlate with distress, functional impairment and with increased healthcare use, which supports our choice of the PHQ-15 as a reference standard. Kroenke et al. noted that the PHQ15 could not completely replace the GPs clinical judgment as it cannot distinguish between explained and unexplained symptoms. “Also, obviously using the PHQ-15 as the primary instrument to find MUPS in primary care should not be advised because of the high percentage found when using cut-off point 5 or more.”

Second, when registration by practice employees is not complete and uniform according to existing guidelines and therefore suboptimal, the performance of any EMR search strategy will hamper. Third, MUPS are often associated with frequent attendance, but not always, particularly not in the early stages. By identifying patients with at least five preceding consultations, some patients with MUPS in the earlier stages might be missed.

Implications for research and clinical practice An accurate screening method for retrieving data from EMRs has many advantages for research or care purposes. The identified population can be offered to GPs who should judge if their patients have MUPS or not and should consider proactive and structured stepped care management, depending on the severity of MUPS, for example with panel management to prevent persistence [34, 35]. Looking specifically at this EMR screening method, increasing the sensitivity while maintaining the level of specificity will make it more suitable for proactive panel management. “Potential improvement might be reached with the addition of prescription of analgesics or opiate drugs in patient groups without relevant comorbidity as a predictor. Smits et al. have demonstrated that this kind of prescription is associated with frequent attendance [36,37]. In our relatively small study population relatively few opiate drugs were prescribed without underlying malignancy and many analgesics are freely available and therefore not registered, so we were not able to include prescription of analgesics in our analysis reliably”.

37

2

Chapter 2

CONCLUSION Early identification of MUPS patients in EMRs might support GPs to structure care and to initiate proactive stepped care management. The assessed EMR screening method for the identification of MUPS patients is very specific. However, many patients with MUPS might be missed who scored positive on the PHQ-15, used as a reference test in our dataset. A too stringent search strategy seems the most likely cause. Before using this method, its sensitivity needs to be improved while maintaining its specificity.

ACKNOWLEDGEMENTS The authors would like to thank the general practitioners from the Leidsche Rijn Julius Healthcare Centres in Utrecht and their patients for sharing their anonymous electronic medical records in the Julius General Practitionersâ&#x20AC;&#x2122; Network database and data manager Marloes van Beurden for her contribution to this project.

38

Identifying MUPS in electronic medical records

REFERENCES 1. Jackson JL, Passamonti M. The outcomes among patients presenting in primary care with a physical symptom at 5 Years. J Gen Intern Med. 2005;20:1032–1037.

2. Khan AA, Khan A, Harezlak J, et al: Somatic symptoms in primary care: etiology and outcome. Psychosomatics 2003; 44:471–478.

3. Nimnuan C, Hotopf M, Wessely S. Medically unexplained symptoms: An epidemiological study in seven specialities. J Psychosom Res. 2001; 51:361–367.

4. Verhaak PFM, Meijer SA, Visser AP, et al. Persistent presentation of medically unexplained symptoms in general practice. Fam Pract. 2006;23:414–420.

5. Barsky AJBJ: Somatization and medicalization in the era of managed care. JAMA. 1995;274:1931–1934.

6. Malterud K: Symptoms as a source of medical knowledge: understanding medically unexplained disorders in women. Fam Med. 2000;32:603–611.

7. Salmon P, Peters S, Stanley I: Patients’ perceptions of medical explanations for somatisation disorders: qualitative analysis. BMJ. 1999, 318:372–376.

8. Woivalin T, Krantz G, Mäntyranta T, et al. Medically unexplained symptoms: perceptions of physicians in primary health care. Fam Pract. 2004;21:199–203.

9. Page LA, Wessely S. Medically unexplained symptoms: exacerbating factors in the doctor-patient encounter. J R Soc Med. 2003;96:223–227.

10. Reid S, Wessely S, Crayford T, et al. Frequent attenders with medically unexplained symptoms: service use and costs in secondary care. Br J Psychiatry. 2002;180:248–253.

11. Smith GJ: Patients with multiple unexplained symptoms: Their characteristics, functional health, and health care utilization. Arch Intern Med. 1986;146:69–72.

12. Barsky AJ: Somatization increases medical utilization and costs independent of psychiatric and medical comorbidity. Arch Gen Psychiatry. 2005;62:903–910.

13. Kroenke K, Swindle R: Cognitive-behavioral therapy for somatization and symptom syndromes: a critical review of controlled clinical trials. Psychother Psychosom. 2000; 69:205–215.

14. Feldstein AC, Perrin NA, Unitan R, et al. Effect of a patient panel-support tool on care delivery. Am J Manag Care. 2010;16:256–e266.

15. Drubbel I, Wit NJ D, Bleijenberg N, et al. Prediction of adverse health outcomes in older people using a frailty index based on routine primary care data. J Gerontol A Biol Sci Med Sci. 2013;68:301-8.

16. Loo TS. Electronic medical record reminders and panel management to improve primary care of elderly patients. Arch Intern Med. 2011;171:1552–1558.

17. Ling CYL, Kajioka E, Luu V, et al. A Quality-improvement project use of a computerized medical database and reminder letters to increase preventive care use in kaiser permanente patients. Perm J. 2009; 13:19–24.

18. Tian TY, Zlateva I, Anderson DR. Using electronic health records data to identify patients with chronic pain in a primary care setting. JAMIA. 2013; 20:275-280.

19. Morriss R, Lindson N, Coupland C, et al. Estimating the prevalence of medically unexplained symptoms from primary care records. Public Health. 2012;126:846–854.

39

2

Chapter 2

20. Kroenke K, Spitzer RL, Williams JBW. The PHQ-15: Validity of a new measure for evaluating the severity of somatic symptoms. Psychosom Med. 2002;64:258–266.

21. Robbins JM, Kirmayer LJ, Hemami S. Latent variable models of functional somatic distress. J Nerv Ment Dis. 1997;185:606–615.

22. Grobbee DE, Hoes AW, Verheij TJM, et al. The Utrecht Health Project: Optimization of routine healthcare data for research. Eur J Epidemiol. 2005;20:285–290.

23. Romera I, Fernández-Pérez S, Montejo ÁL, et al. Generalized anxiety disorder, with or without co-morbid major depressive disorder, in primary care: Prevalence of painful somatic symptoms, functioning and health status. J Affect Disord. 2010;127:160–168.

24. Waal MWMD, Arnold IA, Eekhof JAH, et al. Somatoform disorders in general practice Prevalence, functional impairment and comorbidity with anxiety and depressive disorders. Br J Psychiatry. 2004;184:470–476.

25. Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-report version of prime-md: The PHQ primary care study. JAMA.1999; 282:1737–1744.

26. Spitzer RL, Williams JB, Kroenke K. Utility of a new procedure for diagnosing mental disorders in primary care: The prime-md 1000 study. JAMA. 1994;272:1749–1756.

27. Van Ravesteijn H, Wittkampf K, Lucassen P, et al. Detecting somatoform disorders in primary care with the PHQ-15. Ann Fam Med. 2009;7:232–238.

28. Peveler R, Kilkenny L, Kinmonth AL. Medically unexplained physical symptoms in primary care: a comparison of self-report screening questionnaires and clinical opinion. J Psychosom Res. 1997;42:245–252.

29. Webb AN, Kukuruzovic R, Catto-Smith AG, et al. Hypnotherapy for treatment of irritable bowel syndrome. Cochrane Database Systematic Rev. 2007;4:CD0051 30. Van der Linden MW, Westert GP, De Bakker DH, et al. Tweede nationale studie naar ziekten en verrichtingen in de huisartspraktijk. Klachten en aandoeningen in de bevolking en in de huisartspraktijk. Utrecht/Bilthoven: NIVEL/RIVM, 2004.

31. Bazelmans E, Vercoulen JH, Swanink CM, et al. Chronic fatigue syndrome and primary fibromyalgia syndrome as recognized by GPs. Fam Pract 1999, 16(6):602–604.

32. Kingma EM, de Jonge P, Ormel J, et al. Predictors of a functional somatic syndrome diagnosis in patients with persistent functional somatic symptoms. Int J Behav Med. 2013;20:206–212.

33. Kroenke K, Price RK. Symptoms in the community: Prevalence, classification, and psychiatric comorbidity. Arch Intern Med. 1993;153:2474–2480.

34. Neuwirth E, Estee B, Schmittdiel JA, et al. Understanding panel management: a comparative study of an emerging approach to population care. Perm J. 2007;11:12–20.

35. Bodemheimer T, Berry-Millet R. Care management of patients with complex health care needs.

36. Smits FT, Brouwer HJ, ter Riet G, et al. Epidemiology of frequent attenders: a 3-year historic cohort study comparing attendance, morbidity and prescriptions of one-year and persistent frequent attenders. BMC Public Health 2009;9:36.

37. Smits FT, Brouwer HJ, Zwinderman, et al. Morbidity and doctor characteristics only partly explain the substantial healthcare expenditures of frequent attenders: a record linkage study between patient data and reimbursements data. BMC Fam Pract.

2013;14:138.

40

3 Risk assessment models for patients with persistent medically unexplained physical symptoms in primary care using electronic medical records Madelon den Boeft Mark Hoogendoorn Jos WR. Twisk Sjanne Nap Tim van der Neut Johannes C. van der Wouden Henriette E. van der Horst Mattijs E. Numans

Submitted

Chapter 3

ABSTRACT Background Patients with persistent medically unexplained physical symptoms (MUPS) often experience limitations due to their symptoms, have an impaired quality of life and incur substantial health care costs. Early identification of patients with MUPS who have a high risk of persistent MUPS could lead to early intervention by general practitioners (GPs), which could prevent the transition to persistent MUPS. However, there are no effective and efficient screening methods to identify patients with a high risk of persistent MUPS. Therefore, to enhance identification of these risk patients and thus support GPs, we aimed to develop a risk assessment model using a large dataset extracted from primary care electronic medical records (EMRs).

Methods We used anonymised EMRs from 22 Dutch general practices (n=156.176) over a five-year period (2007-2011). We operationalized persistent MUPS by combining international classification of primary care (ICPC) codes for irritable bowel syndrome, fibromyalgia, chronic fatigue syndrome and low back pain without radiation as an endpoint. We balanced the dataset by including all patients with persistent MUPS (n=7840) and a randomly selected sample of patients without persistent MUPS (n=7988). We applied logistic regression analysis and decision tree analysis to assess the persistent MUPS risk on 80% of the dataset. The performances of the models were evaluated with the area under the curve (AUC) on the remaining 20% of the data, the validation set.

Results The variable best discriminating for persistent MUPS in both models was the number of episodes. Risk assessment of persistent MUPS was performed good with the decision tree (AUC = 0.81) and moderate to good with logistic regression (AUC = 0.70).

The validation showed acceptable stability (0.78 and 0.70, respectively).

44

Risk assessment models for persistent MUPS

Conclusion We explored the possibilities of assessment of patientsâ&#x20AC;&#x2122; individual risks for persistent MUPS based on routine EMRs. We developed two moderate to good performing risk assessment models, partly by using relatively new data mining techniques. While our algorithms require further validation and fine-tuning, they provide a starting point from which GPs could evolve towards a more proactive, structured MUPS management together with their patients.

3

BACKGROUND Twenty to thirty per cent of all physical symptoms presented to the general practitioner (GP) remains unsatisfactorily explained after adequate evaluation (1,2). Most of the so called medically unexplained physical symptoms (MUPS) resolve spontaneously. However, a minority of patients in whom MUPS become persistent suffer from major impairments in daily functioning and a reduced quality of life. Also, persistent MUPS lead to high costs due to high health care utilization and decreased work functioning (3,4).

Early assessment of the risk of patients developing persistent MUPS by the GP could create possibilities for offering structured, proactive care and the prevention from the transition of MUPS to persistent MUPS in subgroups of patients with a high risk.

This combination of risk assessment followed by structured, specified proactive care interventions is called panel management (5). Depending on the nature of the symptoms, functional limitations and the preferences of the patient, different treatment options, for example cognitive behavioural therapy or physical therapy, can be discussed and offered at an early stage (6,7).

The identification of patients at risk for persistent MUPS that is needed for panel management is not a trivial task as there are no generally accepted determinants to guide any selection procedure. Other challenges include the heterogeneity among MUPS patients (8), the varying definitions and classifications used in literature (9) and the lack of a generic MUPS code in the international classification of primary care (ICPC) system, a commonly used classification system in primary care (10). Only a few syndromes closely related to the concept of MUPS with a persistent course have a

45

Chapter 3

specific ICPC code: irritable bowel syndrome, chronic fatigue syndrome, fibromyalgia and (persistent or recurrent) low back pain without radiation.

In primary care research, routine electronic medical records (EMRs) can be used for identifying subgroups of patients at risk. There is evidence that suggests that identifying subgroups at risk followed by panel management can improve quality of life, for example in the frail elderly population (11,12). Only a few previous studies developed risk assessment or identification models for MUPS using EMRs, but either the method did not prove to be effective or it was not directly suitable for usage for primary care (13,14).

In the light of the need to identify potentially new risk variables for MUPS in large heterogeneous EMR datasets, more traditional hypothesis driven approaches might not be the most suitable ones. Nowadays, relatively new data mining techniques are available for analysing large heterogeneous databases, such as EMR databases, to unravel relationships between potential risk variables and outcomes and to gain deeper understanding of the data (15â&#x20AC;&#x201C;17). In the present observational study we aimed to explore alternative approaches in developing risk assessment models based on the analysis of primary care EMRs to enhance identification of patients at risk for persistent MUPS using two different statistical techniques.

METHODS Design and study population We used an observational study design. We were able to use a database consisting of fully anonymised extracted routine EMRs from 22 general practice centers (156176 patients) in the area of Utrecht, The Netherlands, covering a five-year period (January 2007-December 2011). All general practice centers represented in this dataset participate in the Julius General Practitionersâ&#x20AC;&#x2122; Network (JGPN) and share their anonymised routine healthcare data through the JGPN database, connected with the Julius Center, University Medical Center Utrecht (UMCU) and approved by the medical ethical committee of the UMCU (file#99-240).

All data are coded using the ICPC for registration of symptoms and diagnoses, the Anatomical Therapeutic Chemical (ATC) classification system for drug prescription

46

Risk assessment models for persistent MUPS

and various measurement codes for biometry, interventions and referrals. All practice centers involved in the JGPN routinely inform their enlisted patients that EMRs are anonymised and shared through the JGPN database and that research is carried out with this coded database. Individual patients who do not want their data to be used for this purpose are enabled to opt-out. Since patient data are completely anonymised, no further informed consent or new approval of the medical ethics committee was needed.

3

Definition of outcome Since MUPS is not a clear homogeneous condition or diagnosis and we aimed at discriminating a subgroup of patients at risk for persistent MUPS, we operationalized the concept of persistent MUPS by choosing any combination of ICPC codes for irritable bowel syndrome (D93), fibromyalgia (L18.01), chronic fatigue syndrome (A04.01) and low back pain without radiation (L03) as the potential outcome. We considered patients registered with any of these codes to closely represent the concept of persistent MUPS, since all four tend to develop into chronic symptomatology without straightforward pathophysiological explanation.

The dataset The database included individual patient records each with the following routine EMR variables: patient characteristics (gender, age), consultation characteristics (date of consultation, ICPC code assigned per consultation, to which episode the consultation belongs), medication (prescription date, type of medication using ATC code), referrals (date of referral to any specialty type) and laboratory results (dates of tests, type of tests, test results, frequency of each test per patient). The first occurrence of any of the four mentioned MUPS ICPC codes was considered to be the first registration of persistent MUPS in our definition. For each patient with persistent MUPS in our definition, all data registered during the year prior to the first defined ICPC code were selected. For the patients in the dataset without one of the persistent MUPS codes, a one-year period was randomly selected. As the number of patients with persistent MUPS in the dataset was small (n=7840), we balanced the dataset by taking a random sample of patients without persistent MUPS (n=7988). From this balanced dataset,

47

Chapter 3

only patients aged between 18 and 65 were included in the analyses (6398 patients with persistent MUPS and 5988 patients without persistent MUPS).

Constructing variables The one-year history per patient is event-based, while the construction of a risk assessment model requires a single record per patient that summarizes the events over the entire year. In order to create such a summarization, we derived variables from the events by counting their frequency within the given period. The included variables are demographic data, summations of ICPC codes belonging to a chapter during the given period (both the number of unique ICPC codes assigned from the chapter and the total number of consultations associated with each chapter), number of episodes, number of prescriptions of medications per ATC group, as well as variables about laboratory results and referrals which are represented as the number of referrals per specialism.

Modelling We performed a pre-selection procedure to reduce the number of variables to be used in the modelling. In this pre-selection first the variables with less than 10 cases were excluded. This was done in the full dataset. Second, because the number of unique ICPC codes belonging to an ICPC chapter and the number of consultations are obviously correlated, we only used the number of consultations or contacts per chapter in the modelling procedure. In exploring possibilities to generate an algorithm, we used two statistical methods to generate a risk assessment model for persistent MUPS. First we used a relatively classical logistic regression analysis approach with a forward selection procedure. Second, we applied a relatively advanced decision tree analysis more commonly used in data mining. The decision tree analysis is based on the method of recursive-partitioning analysis (18,19). In this analysis, the total group is split into subgroups based on the variable, which distinguishes the persistent MUPS from the non-MUPS patients best. These subgroups are again partitioned, until no further significant partitioning is possible. In case of continuous candidate risk variables, groups were split at the cut-off point(s) with the highest discriminative value for persistent MUPS. The terminal nodes in the decision tree

48

Risk assessment models for persistent MUPS

were used to classify the patientâ&#x20AC;&#x2122;s risk for persistent MUPS. For the decision tree analysis, the chi-squared automatic interaction detector (CHAID) algorithm was used (20), with a maximum tree depth of 10 splits. We developed the model in 80% of the dataset and validated the model in the remaining 20%. We assessed the performance of the risk assessment models by computing the area under the curve (AUC) of the receiver operating characteristic curve (ROC), which is also known as the C-statistic.

All analyses were performed in SPSS Modeler Version 16 (IBM).

3

RESULTS The dataset Table 1 shows descriptive information for the variables in the composed dataset, which we used to develop our risk assessment models. In the persistent MUPS group 62% were female and the mean age of the group was 41 years. Forty percent had one episode, 26% had two episodes and 32% had more than two episodes. Medication from ATC chapter S (sensory organs) was prescribed most frequently followed by medication from chapter L (antineoplastic and immune-modulating agents), R (respiratory system) and A (alimentary tract and metabolism). In the non-persistent MUPS groups, 57% were female and the mean age of the group was 40 years. Fiftynine percent had zero episodes, 24% had one episode and 17% had more than one episode. Medication from ATC Chapter A (Alimentary tract and metabolism) was prescribed most frequently, followed by medication from Chapter S and L and R. Both groups were referred in three percent for additional examination or to secondary care specialisms. Most referrals were not included in the prediction modelling because the number of patients referred to secondary care specialisms was less than 10, too low to allow meaningful analyses.

Modelling Table 2 shows the final risk assessment model based on the logistic regression analysis. Based on the Wald statistic it can be seen that the total number of episodes was the most discriminative variable of the logistic model, followed by medication for ATC chapter M (musculoskeletal system). The odds ratio for the number of episodes

49

Chapter 3

Table 1. Descriptive information regarding the variables used to build prediction models for persistent MUPS Variable description

Persistent MUPS (n=6398)

No persistent MUPS (n=5988)

62

57

41 (13)

40 (13)

Patient

Female, %

Age, mean (SD)

Consultations

Total number of episodes, %

0

42

59

1

26

24

>1

32

17

ICPC Chapters, one or more contacts in %

A (General and Unspecified)

14

23

B (Blood and Blood forming organs)

1

2

D (Digestive)

9

10

F (Eye)

3

5

H (Ear)

4

6

K (Circulatory)

5

5

L (Musculoskeletal)

15

21

N (Neurological)

4

4

P (Psychological)

10

8

R (Respiratory)

14

18

S (Skin)

17

21

T (Endocrine, metabolic and nutritional)

4

4

U (Urology)

4

4

W (Pregnancy, child birth, family planning)

6

8

X (Female genital system and breasts)

8

10

Y (Male genital system)

2

3

Z (Social problems)

3

3

Total

3

3

Imaging diagnostics

1

1

Physiotherapy

1

0

Laboratory

2

0

Referrals, one or more in %

50

Risk assessment models for persistent MUPS

Table 1Â continued. Descriptive information regarding the variables used to build prediction models for persistent MUPS Variable description

Persistent MUPS (n=6398)

No persistent MUPS (n=5988)

ATC Medication prescriptions, one or more in %

A (Alimentary tract and metabolism)

24

15

B (Blood and blood forming organs)

4

4

C (Cardiovascular system)

11

11

D (Dermatologicals)

14

15

G (Genito-urinary system and sex-hormones)

11

13

J (Anti-infectives for systemic use)

11

12

H (Systemic hormonal preparations)

3

3

L (Anti-neoplastic and immune-modulating)

1

1

M (Musculoskeletal system)

24

10

N (Nervous system)

21

16

P (Anti-parasitic products)

1

1

R (Respiratory system)

16

16

S (Sensory organs)

6

6

Product pp (number of prescription products with regards to medication)

58

57

Product tp (number of trade products with regards to medication)

15

15

Product mp (number of medical products freely available)

16

9

Product cpp (number of products custom prepared by pharmacist)

2

2

66

64

Total medication

3

Laboratory, one or more in %

Question D (diagnostic questions)

3

2

Question Y (yes/no questions)

2

2

Question MC (multiple choice questions)

1

1

Question M (measurements)

30

29

Question type FT (free text questions)

27

26

Total lab results

44

42

ICPC international classification of primary care; ATC anatomical therapeutic classification

51

Chapter 3

Table 2. Final prediction model for persistent MUPS based on a logistic regression analysis with a forward selection procedure Variable description Female

Odds ratio 1.253

Wald 21.649

P-value <0.001

Age

1.014

51.334

<0.001

Number of episodes

1.486

311.645

<0.001

D (Digestive)

0.974

6.228

0.013

F (Eye)

0.898

8.473

0.004

H (Ear)

0.957

3.202

0.074

L(Musculoskeletal)

0.912

63.265

<0.001

R (Respiratory)

0.981

4.516

0.034

Number of contacts per chapter

S (Skin)

0.952

15.871

<0.001

W (Pregnancy, childbirth, family planning)

0.942

11.119

0.001

X (Female genital system and breasts)

0.912

23.169

<0.001

Y (Male genital system)

0.952

3.582

0.058

Z (Social problems)

0.889

7.833

0.005

Physiotherapy

9.747

24.085

<0.001

Number of referrals

0.754

5.667

0.017

Medication A (Alimentary tract and metabolism)

1.243

51.479

<0.001

J (Anti-infectives for systemic use)

0.859

8.238

0.004

L (Anti-neoplastic and Immune-modulating agents)

0.578

8.438

0.004

M (Musculoskeletal system)

1.935

173.626

<0.001

N (Nervous system)

1.056

9.862

0.002

R (Respiratory system)

1.069

6.756

0.009

Product pp (number of prescription products with regards to medication)

0.934

29.336

<0.001

Product tp (number of trade products with regards to medication)

0.912

21.054

<0.001

Question type FT (free text questions)

1.037

11.905

0.001

Number of laboratory results

0.988

48.002

<0.001

ICPC international classification of primary care; ATC anatomical therapeutic classification

52

Risk assessment models for persistent MUPS

(i.e. 1.486) indicates that when a patient has one more episode, the odds of having persistent MUPS becomes 1.486 higher. The odds ratio for medication for ATC chapter M indicates that one more medication in ATC chapter M is associated with a 1.935 times higher odds of having persistent MUPS. Figure 1 (A till D) shows the most optimal decision tree based on the CHAID algorithm. The first split in the decision tree was also based the total number of episodes. The highest probability of having persistent MUPS (figure 1D), with a probability of persistent MUPS of 96% was obtained via number of episodes more than 3, number of contacts in ICPC Chapter A = zero, number of contacts in ICPC Chapter L = zero and number of prescription products w.r.t. medication = zero. The lowest probability of having persistent MUPS (6%) was obtained by the following steps: number of episodes = one, number of contacts in ICPC Chapter A = zero, number of contacts in ICPC Chapter L more than Figure 1. (A to D) Most optimal risk assessment decision tree for persistent MUPS 1A #episodes=1

A_contacts=0

L_contacts=0

S_contacts=0

R_contacts=0

D_contacts=0

X_contacts=0

H_contacts=0

P_contacts=0

R_contacts>0 (9.9%)

D__contacts>0 (11.1%)

A_contacts>0 (12.9%)

L_contacts>0

S_contacts>0

ATC_N<=2 (9.3%)

Male (5.9%)

Female (17.6%)

ATC_N>2 (22.2%)

X_contacts>0 (13.1%)

H_contacts>0 (6.5%)

P_contacts>0 (11.5%)

W_contacts=0 (77.0%) W_contacts>0 (13.6%)

53

3

Chapter 3

1B #episodes=2

ATC_M=0

R_contacts>0 (32.1%)

R_contacts=0

L_contacts=0

S_contacts=0

S_contacts>0 (51.1%)

L_contacts>0 (50%)

ATC_A=0 (85.6%) ATC_A>0 (70.8%)

A_contacts>0

Male (29.4%)

Female (50.3%)

D_contacts>0 (54.8%)

D_contacts=0

Product pp=0 (83.4%)

L_contacts>0 (43.9%)

L_contacts=0

X_contacts>0 (32.9%)

X_contacts=0

A_contacts=0

ATC_M>0

Product pp>0

#labresults<=5 (74.2%) #labresults>5 (52.2%)

1C

1D #episodes=3

#episodes>3 ATC_A=0

L_contacts=0

ATC_A>0 (72.0%)

A_contacts=0

L_contacts>0 (48.0%)

L_contacts=0 A_contacts=0

A_contacts>0

R_contacts=0 (71.5%)

Age<=35 (43.8%)

R_contacts>0 (54.2%)

Age>35 (64.0%)

Product pp=0 (95.9%)

Product pp>0 (78.7)

L_contacts>0 (67.8%)

A_contacts>0 ATC_A=0 (56.6%) ATCA>0 (76.4%)

MUPS medically unexplained physical symptoms; ICPC international classification of primary care; ATC anatomical therapeutic classification; #episodes = number of episodes; *_contacts = number of contacts in ICPC chapter *; ATC_*; number of medications in ATC chapter *; Product pp = number of prescription products with regards to medication; #labresults = number of laboratory results

54

Risk assessment models for persistent MUPS

zero and gender = male (figure 1A). The AUC computed for the final logistic regression prediction model was 0.70 and for the decision tree, the AUC was 0.81. The validation procedure of both models provided AUC values of 0.70 for the logistic regression model and 0.78 for the decision tree. Figure 2 shows the ROC curves for the two risk assessment models in the validation set.

Figure 2. Receiver operating characteristics for the logistic regression model and decision tree on validation set

Decision tree Logistic regression

----------------------

DISCUSSION Main findings The objective of our study was to develop risk assessment models that potentially could be used to identify patients at risk for persistent MUPS from routine primary care EMRs. We compared the results from a relatively classical logistic regression procedure with those from a decision tree model. We were able to assess the risk of persistent MUPS by using the final logistic regression model with a moderate AUC, while the decision tree analyses performed somewhat better with a moderate to good AUC. The variable most clearly discriminating in both models was the total number of episodes among patients with persistent MUPS.

55

3

Chapter 3

Comparison with literature Only few studies aimed to develop a risk assessment or identification model for patients with MUPS from EMRs. They showed that there is no simple or straightforward method. Smithâ&#x20AC;&#x2122;s model, containing gender, total number of consultations and percentage of consultations with MUPS potential, initially predicted MUPS with an AUC of 0.90. Upon external validation, the value decreased to 0.78 (13). Morriss aimed to estimate the prevalence of MUPS from primary care EMRs (14). Their models, related to MUPS and severe MUPS, had AUCs of 0.70 and 076, respectively.

However, the models also required further validation and they concluded that the models did not yet prove to be effective for clinical screening purposes due to low sensitivity. Rosmalen et al conducted a latent class analysis to identify patients with somatization, a concept closely related to MUPS, and found that a simple symptom count could support the diagnosis somatization (21). Our study is essentially more or less a confirmation of these previous studies because we also found that the number of episodes was the most important discriminative variable for persistent MUPS in both our methods.

Although not entirely comparable with our study, other studies identifying subgroups of MUPS showed varying results. Tian developed a risk assessment model for chronic pain where pain scores, pain medication and specific international statistical classification of diseases and related health problems (ICD-9) scores were combined. They found a high AUC of 0.98 (22). Viniol performed k-means cluster analysis and studied patients with chronic low back pain (23). They found three clusters specifically characterized by age and psychological variables such as distress. Harkness et al used READ codes, standard clinical terminology system codes used in general practice in the United Kingdom, to identify patients at risk for irritable bowel syndrome. They concluded that the results were a large underestimation of the community prevalence of irritable bowel syndrome flawing their analysis (24).

Strengths and limitations The main strength of our study is that we used a large database with routine EMRs, reflecting what is registered in daily clinical practice. To be included in the registration, a patient has to consult the GP and the GP has to code the complaint as a

56

Risk assessment models for persistent MUPS

symptom or a diagnosis. By using data from many practices and from many patients, the risk of selection bias and information bias is reduced and the influence of loss of detail is reduced. This study potentially provides new insights in research methodology because of applying advanced big data analysis on relatively raw data that are gathered from routine primary care.

Our results should be interpreted in the light of some limitations. As with all routinely collected EMRs, we could not verify the correctness and completeness of the coded information and different GPs have different coding patterns. Also, by using four ICPC codes to conceptualize persistent MUPS, we might have missed some potential other characteristics of MUPS patients. But because GPs apply these codes only after medical evaluation thereby ruling out other diseases, we are confident that we really identified patients resembling those with persistent MUPS as the reference, thus enabling us to assess the risk of others to evolve to that condition. Another limitation is that, by balancing the dataset in order to organize our analysis statistically efficient, the prevalence of MUPS changed to more or less 50%, which potentially hampers the generalizability of the results in a way that we look at relative risks and not absolute risks. Finally, in the present study, we used the AUC to quantify the quality of the prediction models. The AUC is an indicator for the internal validity of the model. External validity should be assessed by testing the algorithms in another EMR dataset. In the present study we mimicked external validation by performing a cross-validation of the risk assessment models in an independent part of the original dataset. For the logistic regression model the AUC remained the same, while for the decision tree, the AUC became slightly lower. This is not a big surprise, because the building of the tree is highly data driven. Nevertheless, the validation AUC remained moderate to good.

Implications for clinical practice and research Manuel et al who stated that the use of risk assessment models in clinical medicine has increased over the past 20 years and that they have the potential to support the process of decision making and management of population health (16). Christensen et al confirmed that recognition and management by GPs of common primary care mental health disorders including MUPS can be positively influenced by using

57

3

Chapter 3

screening questionnaires (25). Risk assessment among subgroups of patients potentially developing persistent MUPS could be a useful first step for implementing panel management. By deriving a software algorithm from the risk assessment model to be applied in routine primary care, the tendency to move towards persistent MUPS could be increased. Patients at risk could be approached proactively by their GPs. And when a patient at risk for persistent MUPS presents, GPs are potentially enabled to adjust their consultation strategies by paying attention to empathic communication, exploring the patientsâ&#x20AC;&#x2122; functional limitations and providing rational explanations before evolving targeted interventions in consultation with the patient (26,27).

In future research, first the models have to be externally validated and a sensitivity analyses should be performed where the ratio between persistent MUPS and nonMUPS is varied to increase the generalizability.

CONCLUSIONS We explored the possibilities of assessment of patientsâ&#x20AC;&#x2122; individual risks for persistent MUPS based on routine EMRs and we developed two moderate to good performing risk assessment models, partly by using relatively new data mining techniques. While our algorithms require further validation and fine-tuning, they provide a starting point from which GPs could evolve towards a more proactive and structured MUPS management together with their patients.

ACKNOWLEDGEMENTS The authors would like to thank the general practitioners from the JGPN UMCU and their patients for sharing their anonymous electronic medical records.

58

Risk assessment models for persistent MUPS

REFERENCES 1. Jackson JL, Passamonti M. The outcomes among patients presenting in primary care with a physical symptom at 5 Years. J Gen Intern Med. 2005;20:1032–7.

2. Barsky AJ Borus JF. Somatization and medicalization in the era of managed care. JAMA. 1995;274:1931–4.

3. Verhaak PFM, Meijer SA, Visser AP, et al. Persistent presentation of medically unexplained symptoms in general practice. Fam Pract. 2006;23:414–20.

4. Barsky AJ Orav E. Somatization increases medical utilization and costs independent of psychiatric and medical comorbidity. Arch Gen Psychiatry. 2005;62:903–10.

5. Neuwirth E, B, Schmittdiel JA, Tallman K, et al. Understanding Panel Management: A comparative study of an emerging approach to population care. Perm J. 2007;11:12–20.

6. Burton C. Beyond somatisation: a review of the understanding and treatment of medically unexplained physical symptoms. Br J Gen Pract. 2003;53:231–9.

7. Van Dessel N, den Boeft M, van der Wouden JC, et al. Non-pharmacological interventions for somatoform disorders and medically unexplained physical symptoms (MUPS) in adults. Cochrane Database Syst Rev. 2014;11:CD011142 8. Aamland A, Malterud K, Werner EL. Patients with persistent medically unexplained physical symptoms: a descriptive study from Norwegian general practice. BMC Fam Pract. 2014;15:107.

9. Fink P, Rosendal M, Olesen F. Classification of somatization and functional somatic symptoms in primary care. Aust N Z J Psychiatry. 2005;39:772–81.

10. Lamberts H, Wood M, Hofmans-Okkes IM. International primary care classifications: the effect of fifteen years of evolution. Fam Pract. 1992;9:330–9.

11. Silow-Carroll S, Edwards JN, Rodin D. Using electronic health records to improve quality and efficiency: the experiences of leading hospitals. Issue Brief Commonw Fund.

2012;17:1–40.

12. Loo TS, Davis RB, Lipsitz LA, et al. Electronic medical record reminders and panel management to improve primary care of elderly patients. Arch Intern Med. 2011;171:1552–8.

13. Smith RC, Gardiner JC, Armatti S, et al. Screening for high utilizing somatizing patients using a prediction rule derived from the management information system of an HMO: a preliminary study. Med Care. 2001;39:968–78.

14. Morriss R, Lindson N, Coupland C, et al. Estimating the prevalence of medically unexplained symptoms from primary care records. Public Health. 2012;126:846–54.

15. Cios KJ, William Moore G. Uniqueness of medical data mining. Artif Intell Med. 2002;26:1–24.

16. Manuel DG, Rosella LC, Hennessy D,et al. Predictive risk algorithms in a population setting: an overview. J Epidemiol Community Health. 2012;66:859–65.

17. Parvathi I, Siddharth R. Survey on data mining techniques for the diagnosis of diseases in the medical domain. IJCSIT. 2014;5:838-846.

18. Breiman L, Friedman JH, Olshen RA, et al. Classification and regression trees. Boca Raton. Chapman Hall. 1984.

59

3

Chapter 3

19. Lemon SC, Roy J, Clark MA, et al. Classification and regression tree analysis in public health: methodological review and comparison with logistic regression. Ann Behav Med. 2003;26:172–81.

20. Bhartiya S, Mehrotra D. Applying CHAID algorithm to investigate critical attributes of secured interoperable health data exchange. Int J Electron Heal. 2015;8:25–50.

21. Rosmalen JGM, Tak LM, de Jonge P. Empirical foundations for the diagnosis of somatization: implications for DSM-5. Psychol Med. 2011;41:1133–42.

22. Tian TY, Zlateva I, Anderson DR. Using electronic health records data to identify patients with chronic pain in a primary care setting. J Am Med Inform Assoc JAMIA. 2013;20:e275–280.

23. Viniol A, Jegan N, Hirsch O, et al. Chronic low back pain patient groups in primary care - A cross sectional cluster analysis. BMC Musculoskelet Disord. 2013;14:294 24. Harkness EF, Grant L, O’Brien SJ, et al. Using read codes to identify patients with irritable bowel syndrome in general practice: a database study. BMC Fam Pract. 2013;14:183.

25. Christensen KS, Toft T, Frostholm L, et al. Screening for common mental disorders: who will benefit? Results from a randomised clinical trial. Fam Pract. 2005;22:428–34.

26. Salmon P, Ring A, Dowrick CF, et al. What do general practice patients want when they present medically unexplained symptoms, and why do their doctors feel pressurized? J Psychosom Res. 2005;59:255–60.

27. Burton C, Lucassen P, Aamland A, et al Explaining symptoms after negative tests: towards a rational explanation. J R Soc Med. 2015;108:84–8.

60

4 Recognition of patients with medically unexplained physical symptoms by family physicians: results of a focus group study Madelon den Boeft DaniĂŤlle Huisman Johannes C. van der Wouden Mattijs E. Numans HenriĂŤtte E. van der Horst Peter L. Lucassen Tim C. olde Hartman

BMC Fam Pract. 2016;17:55

Chapter 4

ABSTRACT Background Patients with medically unexplained physical symptoms (MUPS) form a heterogeneous group and frequently attend their family physician (FP). Little is known about how FPs recognize MUPS in their patients. We conducted a focus group study to explore how FPs recognize MUPS and whether they recognize specific subgroups of patients with MUPS. Targeting such subgroups might improve treatment outcomes.

Methods Six focus groups were conducted with in total 29 Dutch FPs. Two researchers independently analysed the data applying the principles of constant comparative analysis in order to detect characteristics to recognize MUPS and to synthesize subgroups.

Results FPs take into account various characteristics when recognizing MUPS in their patients.

More objective characteristics were multiple MUPS, frequent and long consultations and many referrals. Subjective characteristics were negative feelings towards patients and the feeling that the FP cannot make sense of the patientâ&#x20AC;&#x2122;s story. Experience of the FP, affinity with MUPS, consultation skills, knowledge of the patientâ&#x20AC;&#x2122;s context and the doctor-patient relationship seemed to influence how and to what extent these characteristics play a role. Based on the perceptions of the FPs we were able to distinguish five subgroups of patients according to FPs: 1) the anxious MUPS patient, 2) the unhappy MUPS patient, 3) the passive MUPS patient, 4) the distressed MUPS patient, and 5) the puzzled MUPS patient. These subgroups were not mutually exclusive, but were based on how explicit and predominant certain characteristics were perceived by FPs. Conclusions FPs believe that they can properly identify MUPS in their patients during consultations and five distinct subgroups of patients could be distinguished. If these subgroups can be confirmed in further research, personalized treatment strategies can be developed and tested for their effectiveness.

64

Recognition of MUPS by family physicians

BACKGROUND Medically unexplained physical symptoms (MUPS), physical symptoms for which no adequate medical explanation can be found after a proper examination, are common in primary care and may have a major impact on the daily life of patients (1â&#x20AC;&#x201C;3). We know that patients with MUPS constitute a heterogeneous group. This heterogeneity is due to a broad range of clinical symptoms (4), variety in sociodemographic characteristics such as age, employment status and educational level, and lastly to psychiatric comorbidity (5). Almost all kinds of MUPS can be presented to FPs in varying degrees of severity. Functional somatic syndromes such as fibromyalgia (FM), irritable bowel syndrome (IBS) and chronic fatigue syndrome (CFS) are also referred to as MUPS. Currently, few effective interventions for MUPS are available. Up to now, only cognitive behavioural therapy (CBT) has been shown to have a small benefit by reducing symptoms and functional impairments (6). The varying and disappointing treatment outcomes can be due to this heterogeneity, as different subgroups of patients may have different needs and may benefit from personalized and targeted health care. In previous studies among patients with FM the authors identified two subgroups, patients with pain avoidance and patients with pain persistence, and these subgroups benefitted from a different treatment approach (7â&#x20AC;&#x201C;9). Also, several studies highlighted the relevance of the heterogeneity among patients with CFS for their treatment response and the need to explore this heterogeneity more in to depth (10,11). In line with these studies and in the light of the scarcity of effective treatments, identifying distinct subgroups of patients with MUPS might be a way forward to develop more targeted interventions. Even though patients with MUPS are frequently seen by FPs, little is known about the actual process of recognizing MUPS by FPs. With this in mind, we conducted a focus group study that specifically addressed the following two research questions: 1) How do FPs recognize MUPS in their patients and 2) Which distinct subgroups of patients with MUPS do FPs recognize?

65

4

Chapter 4

METHODS Design, setting and participants We chose the focus group method because group dialogues tend to generate rich information as the input of any participant may trigger other participants to share their experiences and thoughts in a natural and dynamic way (12).

We started with analysing three focus groups discussions with FPs from a study previously conducted by olde Hartman et al. (13). In these focus groups many aspects concerning the recognition of MUPS by FPs and delineation of MUPS were addressed.

Therefore we chose to analyse their first three focus groups before initiating new focus groups. Thirteen FPs altogether participated in these first three focus groups and each session lasted approximately one hour and a half. Detailed information is described elsewhere (13). After analysing these focus groups, we organized three additional focus groups with FPs to discuss the recognition of MUPS in more depth and to discuss the existence of possible subgroups of MUPS. For the recruitment of participants, we consulted the staff members of the department of general practice and elderly care medicine of the VU University medical center (VUmc) Amsterdam for names of FPs who might be interested in participating.

We invited the FPs by email, letter and/or phone. Similar to olde Hartman et al, we used a purposive sampling strategy with the aim to increase the external validity of the results. FPs were sampled aiming at variation on the following characteristics: age, gender, working experience, geographic location of practice, academic working career versus non-academic career and affinity with MUPS versus no special affinity with MUPS. Given the qualitative approach, the purposive sampling strategy and the relatively small sample size, we did not take into account the FP characteristics in the analysis.

We invited 52 FPs of whom 32 were willing to participate and 16 actually participated in one of the three focus group discussion. The other 16 FPs were not able to participate due to logistical reasons. Each focus group included four to eight FPs and the sessions lasted approximately one hour and a half. Table 1 summarizes the information of the participants of the six focus group discussions included in our study.

66

Recognition of MUPS by family physicians

Table 1. Participant information of all six focus groups Number of FPs (n=29) Gender

Male

16

Female

13

Age in years (range)

51 (31-67)

Experience as a FP in years (range)

19 (0-34)

Working hours*

Full time

15

Part time

13

Not practicing at the moment

1

4

Type of practice

Solo

1

Pair

8

Group

16

Self-employed**

3

Not practicing at the moment

1

Urbanization

Rural

14

Urban

13

Variable**

1

Not practicing at the moment

1

FP: family physician. *Full time means 80-100% working; Part time means less than 80% working. ** Self-employed FPs work in different family practices

An independent, skilled moderator without any interest in the outcome facilitated the discussions and made sure all themes from the interview guide (Table 2) were discussed. After each focus group discussion, we adapted and refined the interview guide, allowing new ideas and thoughts that emerged in earlier stages of the analysis to be brought forward in subsequent sessions. All discussions were transcribed verbatim according to a transcription guideline and entered in the qualitative software program Atlas.ti version 7.

67

Chapter 4

Table 2. Interview guide Think of a patient with MUPS 1

a. What are characteristics of this patient? b. What are characteristics of the patient’s complaints?

2

Regarding recognizing MUPS in your patients: a. Many doctors say that they know whether they are dealing with a patient with MUPS within a short time. What is your opinion and experience regarding this issue? b. Some of the complaints that you almost instantly consider to be MUPS are indeed MUPS and some are not. When do you adjust your hypothesis? c. Do hunches play a role in the recognition of MUPS? Or feelings that are evoked in you? If so, can you describe these hunches and feelings? d. Does the background of the patient (or the story the patient tells with regard to his complaints) play a role in the recognition of MUPS? How and to what extent? e. Does recognition depend on how much you can empathize with the patient or the complaint? Do you still consider it MUPS when you empathize? f. Does the patient’s insight in social or psychological contributors to his complaints play a role in the recognition of MUPS? How and to what extent? g. Does the quality of the physician-patient relationship play a role in the recognition of MUPS? How and to what extent? h. Do you still consider it to be MUPS when a patient is agreeable and you like him?

3

I would like to hear your opinion on the following statement: “Every doctor has his own type of MUPS patient”. (Does the personality of the doctor influence the recognition of MUPS?)

4

Is there a difference, with regard to the recognition of MUPS, between patients who you have known for long time and patients you hardly know?

5

Are you able to distinguish different subgroups of patients with MUPS? How?

Ethics The medical ethics committee VUmc approved the study (reference number 2015.216). All FPs provided written informed consent to participate, for the usage of the data and to publish this manuscript.

68

Recognition of MUPS by family physicians

Analysis We applied the principles of constant thematic comparative analysis (14). First, two of the authors (MdB, DH) carefully read and familiarized themselves with the transcripts. After that, they independently coded the transcripts and categorized the codes by theme to explore similarities and differences between responses of the FPs regarding the recognition of MUPS and subgroups of patients. While analysing, the researchers had the Dutch College of General Practitioners MUPS guideline in mind (15). This guideline uses a framework that covers specific dimensions of the complaint(s) and general MUPS characteristics and pays attention to the doctorpatient relationship. The complaint dimensions (i.e. the somatic, cognitive, emotional, social and behavioural dimension) are rooted in the biopsychosocial model (16). The biopsychosocial model assumes that perceived health is associated with all the dimensions of human existence and that every human being is in constant interaction with their environment.

All themes were discussed by MdB and DH and refined after each focus group analysis. Disagreements and doubts were frequently discussed with two senior researchers (PL, ToH).

The construction of distinct subgroups of MUPS patients was based on responses and perceptions of the FPs in two consecutive steps: 1) analysis of the direct responses of FPs on the question if they could recognize or distinguish distinct subgroups and more specifically, in what way; and 2) analysis of the responses of FPs that emerged spontaneously during the discussions of other topics and seemed to co-occur or combine into a pattern relating to the recognition of subgroups. The subgroups were organised by describing their most explicit and/or predominant characteristic and the behaviour of the patient during the consultation as reported by the FPs, including their feelings towards these subgroups that were used as signifiers for recognition.

To internally validate our findings we performed a member check among all participating FPs and five FPs who are staff members of VUmc who had not participated in any of the focus groups. In this member check we presented the results of our study and asked the FPs to determine if these were consistent with their perceptions and/ or experience. After the sixth focus group we concluded that saturation was reached because no new themes emerged from the transcripts of this focus group study.

69

4

Chapter 4

RESULTS How do FPs recognize MUPS in their patients All FPs were familiar with the phenomenon of MUPS and thought about patients with MUPS as a heterogeneous group. Most FPs believe that they can easily distinguish between explained and unexplained symptoms when they were familiar with the patientsâ&#x20AC;&#x2122; medical background and context. They stated that they needed follow-up consultations to diagnose MUPS in patients whom they did not know well. Several FPs described that the first signal is often that what patients are telling about their symptoms does not make sense to them.

FPs reported taking diverse characteristics, both objective and subjective, into account when considering MUPS as their working hypothesis. More or less objective characteristics are multiple, non-specific symptoms that remain undiagnosed, frequent and long consultations and a relatively high number of referrals for additional diagnostics or to specialists. Amongst subjective characteristics are the feelings, often negative, that MUPS patients evoke in FPs, like irritation and resistance. Another indication of MUPS seems to be that during the consultation, FPs feel forced to repeatedly switch between the diagnostic and management phase, as the patient does not agree with the actions proposed by the FP. This may result in a struggle. Both the switching between phases and the struggle can make FPs aware that they are dealing with MUPS. Quotes on the recognition of MUPS are summarized in table 3.

The process of recognizing MUPS is not straightforward. Most FPs describe that they recognize MUPS based on subtle feelings. The degree to which FPs are able to empathize with their patients can strengthen them in their idea that symptoms might be unexplained. Most FPs stated that when they did not feel empathy for their patients, they were more often inclined to recognize symptoms as MUPS. Therefore the lack of empathy is often used as a signifier for recognition. The degree of empathy is influenced by mainly two factors. First, FPs reported that when they liked their patient, they could better empathize. Second, when FPs were familiar with the patientâ&#x20AC;&#x2122;s context and when the patient could verbalize their symptoms and relate their symptoms to their context, FPs were also more able to understand and empathize (e.g. low back pain due to heavy physical work or being tired when having a lot

70

Recognition of MUPS by family physicians

Table 3. Quotes of FPs in relation to the recognition of MUPS FG5;FP5: ‘ I believe I know what is going on within 30 seconds, like many of us. When I think within two minutes “I do not have a clue of what is going on here”, then I start to think “This can be MUPS”.’

FG5;FP3: ‘ Well, we all know the consultation where things go as you have planned.

You do what you always do, start with taking history, then physical examination, then you often have a diagnosis and then you discuss the strategy. But with MUPS patients, what I usually notice is that the discussion does not go so well and you switch between phases. And you think, what is going on? That is a first possible recognition clue.’

FG2;FP1: ‘ When someone consults me with chest pain during exercise that disappears after two minutes at rest, that is something completely different from when they present many complaints and we often call them atypical, right? It does not fit with a specific disease. They have a headache, but when you talk about the headache they also have back pain and when you are finished with the back pain, they also feel tingles and with everything together, it just does not make sense.’

FG4;FP5: ‘A long list of episodes.’

FP1 and FP4: ‘Yes.’ FP3: ‘A long history…’

FP5: ’…Without any serious diseases.’

FP2: ‘With many referrals for additional examinations or to specialists.’

FP1: ‘ They remain at the complaint level, like headache or stomach ache or fatigue or dizziness.’

FG1;FP1: ‘I often use it as a diagnostic tool for MUPS, that I get irritated by patients.’

FG4;FP1: ‘ What I notice is that many doctors have the same basic feeling about these patients and how they recognize them: the exhaustion, the desperation of the doctor and the way they easily get into a fight with these patients.’

FG: focus group. FP: family physi cian. The numbers correspond with the focus groups session and the family physician.

of stress makes sense). However, the presence of empathy does not completely rule out the possibility MUPS. FPs may empathise and still consider MUPS. Finally, the personality of the FP, their affinity with MUPS, the years of experience and consultation skills were reported to play an important role in the process of recognition. More work and life experience and better skills in coping with MUPS

71

4

Chapter 4

made recognition easier. Quotes on the contributing factors in recognizing MUPS are summarized in table 4.

Table 4. Quotes of FPS in relation to the contributing factors in the recognition of MUPS FG3;FP2: ‘ For FPs it is important to know the context of a patient. If you know about the busy shoe store, then you can empathise more with MUPS.’ FG4; FP1: ‘So if you know more about the context, you can better empathise.’

FP5: Well, it causes less irritation FP1: W hen you have more experience or when you have a longer relationship, you just have more information and you recognize it sooner.

FG6;FP2: ‘ Some patients with a certain personality structure; it is possible that you are just a bit more sensitive to them. So with some patients you will sooner consider “Could this be MUPS?”’ FG: focus group. FP: family physician. The numbers correspond with the focus groups session and the family physician.

Which distinct subgroups of patients with MUPS do FPs recognize?

Most FPs could not directly describe distinct subgroups when asked during the last three focus group discussions. Apart from this specific question, the subject of distinct subgroups and specific characteristics of patients came up several times. From these characteristics five distinct subgroups emerged from the focus group data.

These subgroups were not mutually exclusive, as patients fulfilling the criteria for one subgroup could also have characteristics belonging to another subgroup (e.g. patients with feelings of anxiety may also have a low mood or can be distressed and these symptoms can be inter-related as well). We discerned the following five subgroups of patients according to FPs: 1) the anxious MUPS patient, 2) the unhappy MUPS patient, 3) the passive MUPS patient, 4) the distressed MUPS patient, and 5) the puzzled MUPS patient. Below we describe each subgroup including (a) the explicit and/or predominant characteristics and the behaviour of the patient during consultations as perceived by the FP and (b) subjective feelings towards the patients in these subgroups that can be signifiers for recognition. The quotes regarding the subgroups are summarized in table 5.

72

Recognition of MUPS by family physicians

Table 5. Quotes of FPs in relation to the different subgroups of patients with MUPS The anxious MUPS patient

FG4;FP4: I often see anxious patients, who have a bad connection with their body and who are in panic because of it and are not easily reassured. They always hope to find reassurance in all kinds of additional examinations.’

FP1: Yes’. FP2: agrees FG5;FP2: Is it not predominantly anxiety or an alarming feeling. It is constantly being overwhelmed by signals, physical signals that they cannot make sense of. I believe that patients are being overwhelmed and do not know what to do with all those symptoms. And therefore they come directly to us, as an authority to tell them what it is.’

FG5;FP3: I am sure that they feel the tingles, the palpitations and the headaches and then they think: “Oh My God, what is this? Let this be nothing serious.” They tend to give a catastrophizing explanation to it.’

The unhappy MUPS patient

4

FG1;FP3: It is the recurrent thing, I mean, the fact that patients come back all the time with headache and then stomach ache and then the next time something else again, while their mood clearly fluctuates, than you do not have the depressive disorder. So it is the time that clarifies it.’

FG2;FP1: I think that in patients with MUPS their mood is not as severely disturbed as in the case of a real depressive disorder. But in some patients with MUPS their mood can be low due to their symptoms. In the consultation room they can be apathetic.’

The passive MUPS patient

FG4;FP4: They hand the problem over to you and you should have the solution. They want to take tablets, but really working on solving their symptoms, they do not want that. They want it all, but not coming from them.’

FG2;FP3: They do not have the coping strategies to get over it. They are powerless.’

FP1: Yes, it just happens to them.’

FP3: It just happens to them and they cannot defend themselves.’

FP2: Yes, that is it. It happens to them and they express it in a certain manner: with a stomach ache at that certain moment.’

FG1;FP4: There is a group that always externalizes problems and symptoms. This can evoke a feeling of irritability in the doctor. It makes me feel powerless, because I do not get a way in.’

FG5;FP5: They look for an explanation, but an external one, not in themselves.’

73

Chapter 4

The distressed FG4;FP1: Patients with moderately severe MUPS are often people who MUPS patient have periods where they are just not so comfortable, because, I do not know, they have troubles at work or in their relationship.

Everyone has phases in their lives when they are feeling suboptimal. These [patients] are the easy ones’.

FG4;FP3: A high stress-level and a high level of expectations of themselves. So with a certain group of patients I often think about a burnout. If you burden yourself long enough, you will eventually get MUPS.’ FG4;FP5: Patients with a different ethnic background have a tough life and when they consult with symptoms from the musculoskeletal system I think “Yes I understand those symptoms, I would have had the same symptoms with that kind of work”.’

FG3;FP2: I think that they sometimes persist in their own model of explanations; they do not want to look in other directions. There are so many psychosocial problems. People do not make choices or they are completely overloaded and then I think “Yes, with three children and this and that, I would be very tired”, but they seem to believe that everything should be possible or something like that.’

FG5;FP1: Sometimes you just have vulnerable patients, who do not have a strong support system and therefore they come to you.’

The puzzling MUPS patient

FG1;FP1: We all know people who present themselves extremely balanced in your consultation room and tell you very clearly that they have symptoms and in whom you find zero abnormality.

Nothing wrong at home, or something like that. Absolutely no abnormality at all.’

FG1;FP3: Of course there are some patients where there is absolutely no explanation at all and I am more inclined to keep searching for one and to refer them for additional examinations.’

FG6; T here are patients where you do not know what is wrong and FP4: where you think “Maybe there is some kind of abnormality that we just have not discovered yet”.’

FP2: I agree that is possible.’

FP4: Yes, that we cannot give an explanation with our current knowledge but in 24 years our diagnosis may be totally different.’

FP3: ’ There are certain things still unexplainable now, but maybe not in another 100 years. Lyme is always a good example of such a thing.’

74

Recognition of MUPS by family physicians

FG6;FP1: O ften they have given you cues in the history taking phase. So you can use physiology and certain explanation models, like adrenaline that is released in a certain situation, which can give you palpitations. But when there is absolutely no clue, then I think “This is really unexplained”.’

FG: focus group. FP: family physician. The numbers correspond with the focus groups session and the family physician.

The anxious MUPS patient According to FPs some patients specifically focus on bodily signals, tend to misinterpret harmless signals as disturbing and alarming and become anxious. Most FPs think that these anxious MUPS patients have an abnormal body experience. Anxiety may also be present in a more generalized way, where patients also worry about other aspects of life.

FPs reported that during consultations, patients tend to express disproportional worry and FPs often observe a discrepancy between the nature of the symptom and the patient’s presentation (e.g. catastrophizing). Also FPs believe that these patients predominantly hope to find reassurance in all kinds of additional examinations and referrals. These factors often negatively affect the empathy FPs feel and therewith clues for MUPS. FPs also mentioned that they know families in which all family members tend to worry about symptoms and have similar coping strategies. When the FP meets a member of such a family, they often consider MUPS early-on.

FPs indicated that they find it difficult to use anxiety as a signifier for MUPS in patients who have had a serious somatic disease (e.g. chest pain in patients with a myocardial infarction in their history). The extent of the patient’s anxiety may currently have no ground, but FPs consider it to be understandable and therefore justifiable which makes FPs also more inclined to keep searching for somatic explanations.

The unhappy MUPS patient FPs stated that there are several patients with MUPS with a low mood but that they find it difficult to distinguish between a primarily depressive disorder, where patients also express physical symptoms, and patients with primarily MUPS that are accompanied by a low mood caused by their symptoms. However, it was clear

75

4

Chapter 4

to them that a psychiatric diagnosis of a depressive disorder rules out MUPS as a primary diagnosis. According to FPs, patients with MUPS with a low mood present themselves as unhappy and sometimes apathetic during consultations. FPs noticed that the mood of these patients is not as seriously disturbed as in depressed patients and fluctuates over the recurrent consultations, where they mostly present physical symptoms as their main problem. FPs indicated that sometimes it takes time and several consultations to distinguish between MUPS and a depressive disorder.

The passive MUPS patient According to the FPs, these patients feel that they have no control over their life and that events in daily life just happen to them. This feeling of having no control might be the result of a traumatic history such as childhood abuse.

FPs stated that these patients present themselves as helpless during consultations, show little capacity for introspection and externalize their problems, which can lead to a difficulty in accepting that social or psychological problems might play a role in their complaints. Several FPs described their behaviour as â&#x20AC;&#x2DC;trash bin emptyingâ&#x20AC;&#x2122;, where patients lean backwards after having spilled their complaints and wait until the FP offers the solution. This often evokes negative feelings, such as irritation or resistance, in FPs. The distressed MUPS patient The FPs believe that they recognize the distressed patient with MUPS quickly. The majority of distressed patients with MUPS verbalizes their symptoms well and are willing and capable to attribute their symptoms to circumstances or psychosocial issues. In some cases, patients lack insight into these psychosocial issues and therefore are more fixed on the physical symptoms, which could lead to conflicts during consultations in which the FP tries to discuss non-physiological causes or influences.

Broadly FPs named three underlying causes for distress: 1) a phase in life where several things come together (e.g. relational/work problems). In this case MUPS are more often acute than chronic, but can be recurrent. FPs consider this combination of life events and symptoms as a variant of normal life; 2) perfectionism and burnout, where patients are often highly educated and successful; 3) poor external life

76

Recognition of MUPS by family physicians

situations (e.g. financial strain/long and heavy working hours and migrants) and no social support system; a general vulnerability.

The puzzling MUPS patient Most FPs believe that there are almost always circumstances, habits, coping strategies or personality traits that sustain or (partially) cause MUPS. According to FPs there is a small group of patients where the FP cannot rely on existing explanatory models to substantiate MUPS as a working hypothesis and therefore has no clue why the patient experiences symptoms. Some FPs suspect a physiological cause that science has not discovered yet.

FPs reported that patients in this subgroup tell a clear story about very specific complaints and FPs indicated that they are more inclined to refer them for additional diagnostics or to a specialist in order to search together for an explanation.

The member check revealed that our findings were consistent between the FPs who participated in our study and FPs who did not participate in our study.

DISCUSSION In our study, we found that FPs not only recognize patients with MUPS by using more or less objective data such as frequent and long consultations, but that subjective feelings such as irritation and resistance contribute to recognition as well. Recognition is also influenced by the patientâ&#x20AC;&#x2122;s ability to tell his/her story, doctor characteristics such as years of life and working experience, the doctor-patient relationship and the knowledge the doctor has of his/her patientsâ&#x20AC;&#x2122; history and context. In these characteristics we recognize the values of family medicine. Five subgroups of patients according to FPs could be distinguished from the focus group discussions data: the anxious, the unhappy, the passive, the distressed and the puzzling patient with MUPS. The last subgroup, the puzzling patient, is somewhat different than the other four and does not include the more general characteristics of patients with MUPS according to FPs.

77

4

Chapter 4

Strengths and limitations Our study has several strong points. As far as we know, our study is the first that addresses recognition of patients with MUPS and distinct subgroups based on perceptions and experience of FPs with this methodology. By starting with the analysis of previously conducted focus group discussions, we laid a strong foundation for the new ones. We validated our results with a member check among participating and non-participating FPs to strengthen our findings. As we assembled and took into account a wide variety of opinions by using a purposive sampling strategy, our results address and illuminate many aspects of the recognition of patients with MUPS. Finally, researchers from several disciplines (FPs, a psychologist and a methodologist) worked together in this study, thereby providing insights from different fields.

Our findings should be interpreted in the light of several limitations. First, the focus groups gave insight in the thoughts of FPs, but the actual clinical practice could be different from what they reported during the focus groups. Besides, as the focus group discussions were audiotaped, we did not take into account non-verbal signals.

Our focus groups consisted of FPs, doctors from other somatic specialties could come up with different subgroups. Finally, we do not know if patients recognize themselves in the emerged subgroups. It is possible that they have other perceptions about their symptoms which could lead to difficulties during consultations or lack of adherence to treatment plans.

Comparison with existing literature Our study results can be compared with those of some other studies. Our findings regarding the recognition of MUPS correspond with a focus group study from Schou Hansen et al (17). Although their study had a broader focus (i.e. employment of the MUPS definition and MUPS management), they also found that the process of recognition is shaped during the consultation. Mik-Meyer et al showed that FPs not only use traditional biomedical diagnostic tools but also rely on their own opinions and feelings and evaluations of a patientâ&#x20AC;&#x2122;s context and circumstances, comparable to what we found (18). Two studies support our findings that a continuous doctorpatient relationship with knowledge of the patient and his/her context facilitates recognition (19,20).

78

Recognition of MUPS by family physicians

We found one study regarding different subgroups of undifferentiated MUPS. Rosmalen et al performed latent class analysis and found two classes, in which the number of symptoms was distinctive (21).

There are several studies that specified subgroups within the specific functional MUPS syndromes. Van Koulil et al found two subgroups of FM patients: patients with pain avoidance and patients with pain persistence, while Turk et al found three groups: the dysfunctional group, the interpersonally distressed group and the adaptive copers group (7,8). Both authors concluded that different groups of patients benefited from different treatments. The characteristics of these subgroups such as passivity and distress are somewhat similar to our findings. Cella et al found with their latent class analysis that one class of patients with CFS, with a predominance of anxiety and a symptom focus, predicted a poor response to CBT (10). Finally, Viniol et al found three clusters of patients with chronic low back pain (CLBP) with a cluster analysis that could influence further treatment (22).

There are some studies that underpin the relationship between MUPS and the characteristics that were predominant in our subgroups. Several studies confirm the comorbidity between anxiety and MUPS and depressive feelings and MUPS (23,24).

Burton et al found that FPs mostly saw worry as a trait coinciding with MUPS rather than as a symptom of an anxiety disorder and low mood as a response to circumstances that could also be a symptom of a depressive disorder (23). Van Gils et al showed that an increase in stress precedes an increase in physical symptoms in some individuals, but is not an universal predictor of MUPS (25). Our findings point in the same direction where FPs considered distress to be a pronounced and sustaining factor. Kempke et al found that perfectionism was related to severity of fatigue and low mood in patients with CFS (26,27).

Finally, the characteristics of our subgroups are not exclusively linked to MUPS. Patients with chronic somatic diseases, such as chronic lung disorders and diabetes, may also suffer from anxiety and depressive symptoms (28â&#x20AC;&#x201C;30). FPs use these mental symptoms as a diagnostic tool for the recognition of MUPS, in contrast to chronic somatic diseases.

Bombardier showed that different psychological types found among CLBP patients are also common among patients with a wider range of chronic medical conditions.

79

4

Chapter 4

And consequently, effective elements of treatment as provided for those conditions could also be used in MUPS management (31).

Implications for clinical practice and future research With our study we gained more in depth insight into the process of recognition of MUPS by FPs. By being alert on signifiers of MUPS during consultations FPs may more quickly recognize MUPS thus allowing them to pay extra attention to the doctorpatient relationship early on. A good and continuous doctor-patient relationship consisting of positive communication, support, empathy is imperative for good MUPS management and leads to better health outcomes. Finally, by quickly recognizing MUPS and adequate treatment, chronicity of MUPS might be prevented.

The subgroups need to be tested and validated. Therefore we suggest both qualitative and quantitative observational studies of consultations between patients and FPs. In these studies, the following research questions should be addressed: what is the prevalence of each subgroup, how do the subgroups overlap, what is the course and outcome per subgroup and which treatments are adequate and acceptable and do treatment effects differ across subgroups? When the subgroups are confirmed and validated, they might be helpful for guiding more personalized treatment.

Furthermore, when our findings are validated, they could contribute to the current discussions about the classification and definition of MUPS in family medicine and could have added value for the MUPS guidelines (15,32). Finally, we believe a future qualitative study should include the perspective of the patient regarding recognition

by FPs. Tschudi-Madsen et al showed that patients frequently consider that they may suffer from MUPS (33). In the proposed future study patients should be asked what their perceptions are regarding these subgroups and whether they believe these subgroups will be helpful for targeting treatment.

CONCLUSION With our study we gained more insight into the complex process of recognition of patients with MUPS by FPs. Recognition is strongly connected with the values of family medicine: the doctor-patient relationship, knowledge of a patientâ&#x20AC;&#x2122;s context

80

Recognition of MUPS by family physicians

and continuous care. We were able to partly unravel the heterogeneity among MUPS patients as five subgroups could be distinguished according to FPs, based on certain explicit or predominant characteristics. It is possible that all patients with MUPS require the same basic treatment, with a focus on good doctor-patient communication and a good relationship, but that the subsequent treatment steps require a different work-up, as different patients may have different needs. Personalizing treatment in this way could improve quality of care. Further research, that should also include the patientâ&#x20AC;&#x2122;s perspective, has to be conducted to confirm and validate the different subgroups.

ACKNOWLEDGEMENTS

4

We would like to thank all FPs who participated in our focus group study, the FPs and FPs in training that participated in our member check and both persons who transcribed all the focus groups for us. Also we would like to thank Saskia Mol for her work as moderator.

81

Chapter 4

REFERENCES 1. Burton C. Beyond somatisation: a review of the understanding and treatment of medically unexplained physical symptoms. Br J Gen Pract. 2003;53:231–9.

2. Kirmayer LJ, Robbins JM. Patients who somatize in primary care: a longitudinal study of cognitive and social characteristics. Psychol Med. 1996;26:937–51.

3. Jackson JL, Passamonti M. The outcomes among patients presenting in primary care with a physical symptom at 5 Years. J Gen Intern Med. 2005;20:1032–7.

4. Van der Weijden T, van Velsen M, Dinant GJ, et al. Unexplained complaints in general practice: prevalence, patients’ expectations, and professionals’ test-ordering behavior.

Med Decis Making. 2003;23:226–31.

5. Aamland A, Malterud K, Werner EL. Patients with persistent medically unexplained physical symptoms: a descriptive study from Norwegian general practice. BMC Fam Pract. 2014;15:107.

6. Van Dessel N, den Boeft M, van der Wouden JC, et al. Non-pharmacological interventions for somatoform disorders and medically unexplained physical symptoms in adults.

Cochrane Database Syst Rev 2014;11:CD011142.

7. Van Koulil S, Kraaimaat FW, van Lankveld W, et al. Screening for pain-persistence and pain-avoidance patterns in fibromyalgia. Int J Behav Med. 2008;15:211–20.

8. Van Koulil S, van Lankveld W, Kraaimaat FW, et al. Tailored cognitive-behavioral therapy for fibromyalgia: two case studies. Patient Educ Couns. 2008;71:308–14.

9. Turk DC, Okifuji A, Sinclair JD, et al. Differential responses by psychosocial subgroups of fibromyalgia syndrome patients to an interdisciplinary treatment. Arthritis Care Res. 1998;11:397–404.

10. Cella M, Chalder T, White PD. Does the heterogeneity of chronic fatigue syndrome moderate the response to cognitive behaviour therapy? An exploratory study. Psychother Psychosom. 2011;80:353–8.

11. White PD, Goldsmith K, Johnson AL, Chalder T, et al. Recovery from chronic fatigue syndrome after treatments given in the PACE trial. Psychol Med. 2013;43:2227–35.

12. Kitzinger J. Qualitative research. Introducing focus groups. BMJ. 1995;311:299–302.

13. Olde Hartman TC, Hassink-Franke LJ, Lucassen PL, et al. Explanation and relations.

How do general practitioners deal with patients with persistent medically unexplained symptoms: a focus group study. BMC Fam Pract. 2009;10:68.

14. Glaser B, Strauss A. The discovery of grounded theory. Chigago: Aldine, 1967.

15. Olde Hartman T, Blankenstein N, Molenaar B. NHG-Standaard somatisch onvoldoende verklaarde lichamelijke klachten. Huisarts Wet. 2013:222–30.

16. Engel G. The need for a new medical model: a challenge for biomedicine. Science 1977;196:129-36.

17. Hansen HS, Rosendal M, Fink P, et al. The general practitioner’s consultation approaches to medically unexplained symptoms: A qualitative study. ISRN Family Med. 2013:541604.

18. Mik-Meyer N, Obling A. The negotiation of the sick role: general practitioners’ classification of patients with medically unexplained symptoms. Sociol Health Illn.

2012;34:1025–38.

82

Recognition of MUPS by family physicians

19. Rask MT, Andersen R, Bro F, et al. Towards a clinically useful diagnosis for mild-tomoderate conditions of medically unexplained symptoms in general practice: a mixed methods study. BMC Fam Pract. 2014;15:118.

20. Ridd M, Shaw A, Salisbury C. “Two sides of the coin”--the value of personal continuity to GPs: a qualitative interview study. Fam Pract. 2006;23:461–8.

21. Rosmalen JGM, Tak LM, de Jonge P. Empirical foundations for the diagnosis of somatization: implications for DSM-5. Psychol Med. 2011;41:1133–42.

22. Viniol A, Jegan N, Hirsch O, et al. Chronic low back pain patient groups in primary care – A cross sectional cluster analysis. BMC Musculoskelet Disord. 2013;14:294.

23. Burton C, McGorm K, Weller D, et al. The interpretation of low mood and worry by high users of secondary care with medically unexplained symptoms. BMC Fam Pract.

2011;12:107.

24. Van Boven K, Lucassen P, van Ravesteijn H, et al. Do unexplained symptoms predict anxiety or depression? Ten-year data from a practice-based research network. Br J Gen Pract. 2011;61:316–325.

25. Van Gils A, Burton C, Bos E, et al. Individual variation in temporal relationships between stress and functional somatic symptoms. J Psychosom Res. 2014;77:34–9.

26. Kempke S, Luyten P, van Wambeke P, et al. Self-critical perfectionism predicts outcome in multidisciplinary treatment for chronic pain. Pain Pract. 14:309–14.

27. Kempke S, Van Houdenhove B, Luyten P, et al. Unraveling the role of perfectionism in chronic fatigue syndrome: is there a distinction between adaptive and maladaptive perfectionism? Psychiatry Res. 2011;186:373–7.

28. Mikkelsen R, Middelboe T, Pisinger C, et al. Anxiety and depression in patients with chronic obstructive pulmonary disease. A review. Nord J Psychiatry. 2004;58:65–70.

29. Anderson R, Freedland K, Clouse R, et al. The prevalence of comorbid depression in adults with diabetes: a meta-analysis. Diabetes Care. 2001;24:1069–78.

30. DeJean D, Giacomini M, Vanstone M, et al. Patient experiences of depression and anxiety with chronic disease: a systematic review and qualitative meta-synthesis. Ont Health Technol Assess Ser. 2013;13:1–33.

31. Bombardier C, Divine GW, Jordan J, et al. Minnesota Multiphasic Personality Inventory (MMPI) cluster groups among chronically ill patients: relationship to illness adjustment and treatment outcome. J Behav Med. 1993;16:467–84.

32. Van der Feltz-Cornelis CM, Hoedeman R, Keuter EJW, et al. Presentation of the multidisciplinary guideline medically unexplained physical symptoms and somatoform disorder in the Netherlands: disease management according to risk profiles. J Psychosom Res. 2012;72:168–9.

33. Tschudi-Madsen H, Kjeldsberg M, Natvig B, et al. Medically unexplained conditions considered by patients in general practice. Fam Pract. 2014;31:156–63.

83

4

5 Non-pharmacological interventions for somatoform disorders and medically unexplained physical symptoms (MUPS) in adults Nikki Claassen- van Dessel Madelon den Boeft Johannes C. van der Wouden Maria Kleinstäuber Stephanie S. Leone Berend Terluin Mattijs E. Numans HenriÍtte E. van der Horst Harm van Marwijk

Cochrane Database Systematic Reviews. 2014;11:CD011142 NB: This chapter is a concise version of the full Cochrane review

Chapter 5

ABSTRACT Background Medically unexplained physical symptoms (MUPS) are physical symptoms for which no adequate medical explanation can be found after proper examination. The presence of MUPS is the key feature of conditions known as ‘somatoform disorders’. Various psychological and physical therapies have been developed to treat somatoform disorders and MUPS. Although there are several reviews on non-pharmacological interventions for somatoform disorders and MUPS, a complete overview of the whole spectrum is missing.

Objectives To assess the effects of non-pharmacological interventions for somatoform disorders (specifically somatisation disorder, undifferentiated somatoform disorder, somatoform disorders unspecified, somatoform autonomic dysfunction, pain disorder, and alternative somatoform diagnoses proposed in the literature) and MUPS in adults, in comparison with treatment as usual, waiting list controls, attention placebo, psychological placebo, enhanced or structured care, and other psychological or physical therapies.

Search methods We searched the Cochrane Depression, Anxiety and Neurosis Review Group’s Specialised Register (CCDANCTR) to November 2013. This register includes relevant randomised controlled trials (RCTs) from The Cochrane Library, EMBASE, MEDLINE, and PsycINFO. We ran an additional search on the Cochrane Central Register of Controlled Trials and a cited reference search on the Web of Science. We also searched grey literature, conference proceedings, international trial registers, and relevant systematic reviews.

Selection criteria We included RCTs and cluster randomised controlled trials which involved adults primarily diagnosed with a somatoform disorder or an alternative diagnostic concept

86

Non-pharmacological interventions for chronic MUPS

of MUPS, who were assigned to a non-pharmacological intervention compared with usual care, waiting list controls, attention or psychological placebo, enhanced care, or another psychological or physical therapy intervention, alone or in combination.

Data collection and analysis Four review authors, working in pairs, conducted data extraction and assessment of risk of bias. We resolved disagreements through discussion or consultation with another review author. We pooled data from studies addressing the same comparison using standardised mean differences (SMD) or risk ratios (RR) and a random-effects model. Primary outcomes were severity of somatic symptoms and acceptability of treatment.

Main results We included 21 studies with 2658 randomised participants. All studies assessed the effectiveness of some form of psychological therapy. We found no studies that included physical therapy.

Fourteen studies evaluated forms of cognitive behavioural therapy (CBT); the remainder evaluated behaviour therapies, third-wave CBT (mindfulness), psychodynamic therapies, and integrative therapy. Fifteen included studies compared the studied psychological therapy with usual care or a waiting list. Five studies compared the intervention to enhanced or structured care. Only one study compared cognitive behavioural therapy with behaviour therapy.

Across the 21 studies, the mean number of sessions ranged from one to 13, over a period of one day to nine months. Duration of follow-up varied between two weeks and 24 months. Participants were recruited from various healthcare settings and the open population. Duration of symptoms, reported by nine studies, was at least several years, suggesting most participants had chronic symptoms at baseline.

Due to the nature of the intervention, lack of blinding of participants, therapists, and outcome assessors resulted in a high risk of bias on these items for most studies. Eleven studies (52% of studies) reported a loss to follow-up of more than 20%.

For other items, most studies were at low risk of bias. Adverse events were seldom reported.

87

5

Chapter 5

For all studies comparing some form of psychological therapy with usual care or a waiting list that could be included in the meta-analysis, the psychological therapy resulted in less severe symptoms at end of treatment (SMD -0.34; 95% confidence interval (CI) -0.53 to -0.16; 10 studies, 1081 analysed participants). This effect was considered small to medium; heterogeneity was moderate and overall quality of the evidence was low. Compared with usual care, psychological therapies resulted in a 7% higher proportion of drop-outs during treatment (RR acceptability 0.93; 95% CI 0.88 to 0.99; 14 studies, 1644 participants; moderate-quality evidence). Removing one outlier study reduced the difference to 5%. Results for the subgroup of studies comparing CBT with usual care were similar to those in the whole group.

Five studies (624 analysed participants) assessed symptom severity comparing some psychological therapy with enhanced care, and found no clear evidence of a difference at end of treatment (pooled SMD -0.19; 95% CI -0.43 to 0.04; considerable heterogeneity; low-quality evidence). Five studies (679 participants) showed that psychological therapies were somewhat less acceptable in terms of drop-outs than enhanced care (RR 0.93; 95% CI 0.87 to 1.00; moderate-quality evidence).

Conclusions When all psychological therapies included this review were combined they were superior to usual care or waiting list in terms of reduction of symptom severity, but effect sizes were small. As a single treatment, only CBT has been adequately studied to allow tentative conclusions for practice to be drawn. Compared with usual care or waiting list conditions, CBT reduced somatic symptoms, with a small effect and substantial differences in effects between CBT studies. The effects were durable within and after one year of follow-up. Compared with enhanced or structured care, psychological therapies generally were not more effective for most of the outcomes.

Compared with enhanced care, CBT was not more effective. The overall quality of evidence contributing to this review was rated low to moderate.

The intervention groups reported no major harms. As most studies did not describe adverse events as an explicit outcome measure, this result has to be interpreted with caution.

88

Non-pharmacological interventions for chronic MUPS

An important issue was that all studies in this review included participants who were willing to receive psychological treatment. In daily practice, there is also a substantial proportion of participants not willing to accept psychological treatments for somatoform disorders or MUPS. It is unclear how large this group is and how this influences the relevance of CBT in clinical practice.

The number of studies investigating various treatment modalities (other than CBT) needs to be increased; this is especially relevant for studies concerning physical therapies. Future studies should include participants from a variety of age groups; they should also make efforts to blind outcome assessors and to conduct follow-up assessments until at least one year after the end of treatment.

BACKGROUND Description of the intervention In previous decades, many pharmacological and non-pharmacological interventions for somatoform disorders and MUPS were developed. The use of antidepressants, in particular, as pharmacological agents for syndromes of MUPS (1,2) or chronic pain (3) was tested. The most relevant groups of antidepressants are the tricyclic antidepressants, selective serotonin reuptake inhibitors, and selective serotonin and noradrenaline (norepinephrine) reuptake inhibitors. In addition to antidepressants, antiepileptic drugs are also commonly used for somatoform disorders (4,5), although they are not advised in guidelines. Pharmacological interventions will be described in a separate forthcoming Cochrane review (6) and this review only focuses on nonpharmacological interventions.

Most non-pharmacological interventions for MUPS focus on addressing cognitions, behaviour, coping styles, and functional consequences of symptoms. These interventions include psychological therapies as well as physical therapies. In the paragraph below, we described examples of several frequently studied forms of psychological and physical therapies.

89

5

Chapter 5

How the intervention might work Psychological therapies - cognitive behavioural therapy

The first and most commonly used and investigated psychological therapy for MUPS is CBT, which is based on the cognitive behavioural model (7). This model proposes that MUPS are caused by a self perpetuating multi-factorial cycle, based on the interaction of different factors in several domains, including somatic (physical) aspects, cognitions (thoughts), behaviour, emotions, and environment (8).

Reattribution is a specific form of CBT (9). This method aims to encourage people to reattribute their MUPS to physiological or psychosocial causes rather than to somatic causes. Reattribution consists of three stages: 1. Making the person feel understood; 2. Changing the agenda of the person, and the doctor, and their mutual agenda during the consultations; and 3. Making the link between physical symptoms and psychosocial problems.

Problem-solving treatment is another form of CBT that has been used for people with MUPS and somatoform disorders. The aim is to reduce complaints associated with unresolved problems in daily life by enhancing a personâ&#x20AC;&#x2122;s problem-solving capacities in a step-by-step manner. This therapy has a positive effect on mental and physical health problems in general (10) Psychological therapies - behavioural therapy

Behavioural therapy, the second group, aims to constructively change a personâ&#x20AC;&#x2122;s behaviour towards their symptoms using operant conditioning - also known as instrumental conditioning - in which a response in a certain context is followed by a reinforcing stimulus or consequence, thereby increasing the likelihood that the same response will follow in future. Biofeedback therapy is an important behavioural intervention relevant to this review. Other forms of behavioural therapy include relaxation therapy (11), and psycho-education (12).

Other psychological therapies

A third group of psychological therapies, more aimed at increasing insight, such as:

90

Non-pharmacological interventions for chronic MUPS

1. Third-wave cognitive behavioural therapy (i.e. the development of a new attitude towards symptoms, based on self-regulation of attention and acceptance) (13); 2. Psychodynamic therapies, a form of depth psychology, which focuses on revealing the unconscious content of a personâ&#x20AC;&#x2122;s psyche in order to alleviate psychological of physical tension (14).

3. Humanistic therapies, focusing on self-development, growth, and responsibilities. Treatment aims to help individuals recognise their strengths, creativity, and choices in the â&#x20AC;&#x2DC;here and nowâ&#x20AC;&#x2122;.

4. Integrative therapies, which integrate components from several theoretical schools, which aims to work with the person to identify procedural sequences, chains of events, thoughts and emotions that explain how a target problem is established and maintained

5

Enhanced care

Another group of therapies offered to people with MUPS is enhanced care. Within these therapies people receive care as usual (mostly by their general practitioners (GP)), enhanced with, for example, participant education or structured counselling moments (15). Within these therapies, there is no specific treatment agenda or structure; the aim is to offer the person some tools to assist in the recovery process, stimulating self-management.

Physical therapies - physical activity training

Several studies have indicated that mental health, including mood, pain thresholds, and sleep, can be improved by low- or moderate-intensity activity (16). Graded activity training is an operant-conditioning behavioural approach in which physical activity is expanded step by step, based on a predetermined time schedule.

Other physical therapies

Other examples of physical therapies for somatoform disorders and MUPS include activation therapy, where physical and behavioural activation is increased in a step-

91

Chapter 5

wise fashion, and running therapy, where running is used therapeutically, mainly to influence the level of stress.

Objectives To assess the effects of non-pharmacological interventions for somatoform disorders (specifically somatisation disorder, undifferentiated somatoform disorder, somatoform disorder unspecified, somatoform autonomic dysfunction, pain disorder, and alternative somatoform diagnoses proposed in the literature) and MUPS in adults in comparison with treatment as usual, waiting list controls, attention placebo, psychological placebo, enhanced or structured care, and other psychological or physical therapies.

METHODS Types of studies We included randomised controlled trials (RCTs) and cluster randomised controlled trials (CRCTs). We also planned to include data from the first phase of crossover trials, but we identified no such trials that met our inclusion criteria. We excluded quasirandomised trials (e.g. allocation to the study group by day of the week).

Types of participants Participant characteristics

Participants had to be at least 18 years old. We applied no maximum age, as the condition can be present at any age. We placed no restriction on gender or culture.

Diagnosis

1. Participants had to meet the criteria for a somatoform disorder or the criteria for one of the alternative somatoform diagnoses proposed in the literature.

The primary diagnosis (a somatoform disorder) had to be made on the basis of a structured clinical interview or diagnostic checklists 2. Participants were characterised with MUPS as their primary problem, on the basis of a validated scale for the assessment of MUPS

92

Non-pharmacological interventions for chronic MUPS

As the subdivision of these two diagnostic concepts (somatoform disorders and MUPS) is based on differences in selection methods used in different research settings rather than on differences between individual people, it is possible that the nature and severity of symptoms may show a certain overlap between the two groups. We disregarded the DSM-5 criteria for somatoform disorders for this version of the review.

Co-morbidities

As we aimed to summarise interventions for multiple symptoms, we excluded studies and reviews that examined participants diagnosed with only one specific functional syndrome or symptom.

Subsets of participants

Some studies could include ‘eligible’ participants as well as ‘ineligible’ participants for this review, for example when an age cut-off was used that was different to the cut-off of this review. When no detailed information was available about these subsets of participants, we requested the data from the trial authors.

Types of interventions Experimental interventions

Eligible studies included one or more of the following experimental interventions.

1. Psychological therapies: CBT, behavioural therapy, third-wave CBT, psychodynamic therapies, humanistic therapies, integrative therapies 2. Physical therapies: physical activity training, other physical therapies We excluded interventions based on complementary medicine from this review. In addition, pharmacological interventions and consultation letter interventions were beyond the scope of this review; they were evaluated in other Cochrane reviews (17, 6). In several of the studies, in both study arms a consultation letter was sent to the primary care physician after baseline assessment, in addition to the planned psychological therapy or comparison condition. Post-hoc, we decided that this was not a reason for exclusion, and we categorised these studies according to the main comparison.

93

5

Chapter 5

Comparator interventions

1. Normal/usual treatment or waiting list procedures.

2. Attention or psychological placebo 3. Enhanced or structured care 4. Other psychological therapies 5. Other physical therapies Types of outcome measures We included studies that met the inclusion criteria described above regardless of whether they reported on the following outcomes.

Primary outcomes 1. Severity/intensity of somatic symptoms 2. Acceptability

Secondary outcomes

1.

2.

3.

4.

5.

6.

Depression and anxiety Dysfunctional cognitions, emotions, or behaviours (participant-rated) Adverse events Treatment response (responder versus non-responder) Functional disability and quality of life Health care use

Hierarchy of outcome measures

If there were multiple instruments measuring the same outcome, we preferred whichever instrument was most commonly used from those listed above.

Timing of outcome assessment

We analysed primary and secondary outcomes at the following time points, if available: immediately post treatment; within 12 months after treatment ending; and more than 12 months after treatment ended.

94

Non-pharmacological interventions for chronic MUPS

Search methods for identification of studies Detailed information from the search methods are described in the full review. In summary, we performed electronic searches in the Cochrane Depression, Anxiety and Neurosis Review Groupâ&#x20AC;&#x2122;s Specialized register (CCDANCTR), conducted complementary searches in Cochrane Central Register of Controlled Trials (CENTRAL) and searched for ongoing clinical trials.

Also we searched the following other resources: grey literature, hand searching for conference proceedings reference lists and correspondence to authors.

Data collection and analysis Selection of studies

In the first step, two review authors (NvD, MdB) independently screened the titles and abstracts of reports identified from the literature search. We discarded studies that obviously did not fulfil the inclusion criteria at this stage of the screening process.

Two review authors (NvD, MdB) retrieved eligible or potentially eligible articles for full-text assessment. We identified and excluded duplicate records and we collated multiple reports that related to the same study so that each study - rather than each report - was the unit of interest in the review. After full-text assessment, the review authors identified studies for inclusion and exclusion. We recorded reasons for exclusion of studies, and resolved disagreements by consensus - if necessary with the involvement of a third review author (JvdW). We listed studies for which additional information was required in order to determine their suitability for inclusion in the review as â&#x20AC;&#x2DC;Studies awaiting assessmentâ&#x20AC;&#x2122;.

Data extraction and management We used a data collection form, piloted on one study in the review, to extract study characteristics and outcome data. Independently, four review authors (NvD, MdB, HvdW, HvM) extracted study characteristics and outcome data from included studies. If necessary, we contacted the authors of trial reports for clarification or for additional information. We organised data using the most recent version of Review

95

5

Chapter 5

Manager 5 software (RevMan 2012). We negotiated disagreements with another review author. We extracted data on the following study characteristics.

1. Trial characteristics 2. Details of methodology 3. Participantsâ&#x20AC;&#x2122; characteristics 4. Intervention characteristics 5. Outcome measures 6. Notes Two review author (NvD, HvdW) entered data into Review Manager 5 for analysis (RevMan 2012). We double-checked that data had been entered correctly by comparing the data presented in the systematic review with the data in the study reports.

A third review author (MdB) spot-checked study characteristics for accuracy against the trial reports.

Main comparisons Based on the available data, we present the following comparisons: 1. Psychological therapy versus usual care (or waiting list procedures) 2. Psychological therapy versus enhanced (or structured) care 3. Psychological therapy versus another psychological therapy Assessment of risk of bias in included studies Independently, two review authors (NvD, MdB) assessed the risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (18). We resolved any disagreements by discussion or by involving another review author (HvM, HvdW). We assessed the risk of bias for the following domains.

1. Random sequence generation 2. Allocation concealment 3. Blinding 4. Incomplete outcome data 5. Selective outcome reporting 6. Other sources of bias

96

Non-pharmacological interventions for chronic MUPS

7. Treatment fidelity 8. Researcher allegiance We judged each potential source of bias as to be of high, low, or unclear risk. Then we summarised the risk of bias judgements across different studies for each of the domains Measures of treatment effect Dichotomous data

For dichotomous outcomes, we used risk ratio (RR) as the summary statistic, together with 95% confidence intervals (CI).

Continuous data

As different measures were used to assess the same outcome, we pooled data using the standardised mean difference (SMD); we calculated 95% CI. Specific attention was paid to the secondary outcome â&#x20AC;&#x2DC;functional disability and quality of lifeâ&#x20AC;&#x2122;, as the direction of scales for these outcomes can differ.

Assessment of heterogeneity We assessed the groups for clinical similarities including elements such as age, gender, and setting.

First, we assessed statistical heterogeneity visually by inspecting forest plots of standardised mean effect sizes and of relative risks. We used the I2 statistic as a second test: I2 describes the percentage of variability in effect estimates that is due to heterogeneity rather than chance.

We used conventions of interpretation defined by Higgins (18). In the case of substantial levels (i.e. where I2 = 50% to 90%) and considerable levels (I2 = 75% to 100%) of heterogeneity, we explored data further by means of subgroup and sensitivity analyses (see below). These were not clear-cut criteria, as the importance of the observed I2 also depends on the magnitude and direction of treatment effects and the strength of evidence for heterogeneity (19; 20); for example: if the I2 value fell slightly below 50% (e.g. 45%) and the direction and magnitude of treatment effects suggested important heterogeneity, we investigated the data further.

97

5

Chapter 5

Assessment of reporting biases We created funnel plots (treatment effect versus standard error of the effect size), if we included at least 10 trials in a meta-analysis, according to the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions (18, 21). When analysing and interpreting a funnel plot, we considered all potential reasons for asymmetry, not just publication bias (e.g. differences in methodological quality, true heterogeneity in intervention effects).

Data synthesis If we found two or more included studies in a comparison category (see Data extraction and management) that used the same outcome construct, we performed a meta-analysis of the results. Two authors (NvD, JvdW) entered data into Review Manager 5 software (RevMan 2012). We expected to find high heterogeneity in nonpharmacological therapy approaches and in symptom severity, duration of symptoms and co-morbidities among the various study populations. Therefore, we analysed dichotomous and continuous treatment effects using a random-effects model. For studies of which data could not be combined, we summarised the results narratively.

RESULTS Description of studies Searches were conducted up to November 2013 (CCDAN registers) and April 2014 (cited reference searches). Figure 1 shows the flow diagram of study selection.

98

RESULTS Description of studies Searches were conducted up to November 2013 (CCDAN registers) and April 2014 (cited Non-pharmacological interventions for chronic MUPS reference searches). Figure 1 shows the flow diagram of study selection.

1. Flow diagram ofstudy study selection Figure 1. Flow diagram of selection

5

Literature database searches In the search of the CDANCTR-Studies and CCDANCTR-Reference Register (from now

Literature searcheswe found 929 abstracts after de-duplication. We on referred to database as CCDAN database),

excluded 842 records, based on the title and abstract, leaving 82 references (65 stud-

In the search the CDANCTR-Studies andthe CCDANCTR-Reference ies) selected for of full-text retrieval. After reading full-text, we judged 27 Register studies (from now on

(49 articles)toeligible for inclusion in thiswe review.

We929 excluded 35 studies articles) referred as CCDAN database), found abstracts after (38 de-duplication.

We excluded 84 99

Chapter 5

and six articles are still awaiting assessment, for example, due to unavailability of a full-text article or difficulties in contacting authors.

The search of CENTRAL database found 995 records. After removing duplicates from the CCDAN search, there were 568 new references. After title and abstract screening, we excluded 560 references, and selected eight articles for full-text reading. After full-text reading, we excluded five articles, and judged three articles eligible for inclusion; however, all three articles described studies already included in the review (e.g. long-term follow-up results) (22, 23, 24). As the Schrรถder article reported a more detailed trial methodology and higher number of participants, we decided to use this article as the main reference of this study (23) instead of Zaby 2008 (25), which was retrieved from the CCDAN search.

We performed a cited reference search on the Web of Science, for citations to primary reports of all studies expected to be included in this review. When hand searching the retrieved articles, we identified three additional relevant references.

After full-text reading, we included one new study (26), and excluded one article due to randomisation method (27). One article (28) described an already included study (29).

Grey literature We performed searches for grey literature but found no new articles. We screened the conference proceedings and found no new articles.

Systematic reviews We found 14 reviews about (specific) non-pharmacological interventions for somatoform disorders or MUPS. After title screening in the reference lists of the reviews, we selected seven additional articles for screening of abstract. After abstract reading, we excluded four articles, and selected three articles for full-text reading. None of these three articles were eligible due to lack of randomisation or inappropriate selection method (30, 31, 32).

100

Non-pharmacological interventions for chronic MUPS

Trial registers We performed ongoing trial searches in the databases of www.clinicaltrials.gov, www.controlled-trials.com, and www.who.int/trialsearch. We found six potentially eligible ongoing trials. As full details of the design and study results were not available, we could not include these studies in the review.

Contacting authors We tried to contact 10 trial authors for missing information regarding the eligibility of studies; four responded and provided the desired information (33, 34, 35, 36). We contacted authors of 20 of the included studies for additional information regarding study design and outcomes, of which 10 provided requested data.

Included studies We included 21 studies, reported in 43 publications, in this review. All included studies concerned psychological interventions.

Design

Twenty of the included 21 studies had a parallel-group, individually randomised design (RCT). One study had a cluster-randomised design (37). We found not trials with a crossover design.

Sample size

The total number of randomised participants was 2658, a mean number of 127 per study (range 32 to 328). Two studies included fewer than 25 participants per arm (26, 38). Most studies reported 25 to 75 participants per arm. Three studies included 75 to 100 participants per arm (39, 40, 41), and two studies included more than 100 participants per arm (29, 37). The largest study was Schaefert 2013 (37), with 328 randomised participants.

Setting

Eight studies recruited participants in primary care only (26, 39, 42, 37, 40, 43, 41; 13). Only two studies recruited in secondary care (e.g. outpatient clinics) (28, 44) and

101

5

Chapter 5

one study recruited inpatients in hospitals (45). Seven studies recruited via medical settings as well as the open population (e.g. through advertisements) (46, 47, 38, 48, 49, 23, 24). Three studies recruited via primary care as well as secondary care (50, 33, 51).

In one study, treatment was performed in group sessions by GPs in primary care who were trained in the specific psychological technique, combined with a psychosomatic specialist (37). In six other studies, treatment took place at a department of psychiatry or psychology (46, 39, 50, 38, 48, 13). Another six studies treated participants in other outpatient clinics (47, 33, 51, 43, 41, 24). Five studies treated participants in specific outpatient symptom clinics or outpatient clinics for psychosomatics (26, 49, 28, 23, 44). One study treated participants as inpatients (45). One study treated participants at home (40). Finally, in one study the treatment setting was unknown (42).

Participants

Most studies recruited women than men, as found in epidemiological studies (52, 53). Only one study reported more men (56%) in the intervention group (44). The proportion of women among all participants in all treatment groups ranged between 66% (26) and 89% (46).

The mean age was 43 years in all included studies, ranging from 35 years (48) to 49 years (49).

Diagnostic criteria and inclusion criteria varied widely between studies. Fourteen studies used standardised diagnostic interviews to establish the diagnosis, the other seven studies used standardised questionnaires. In nine studies, symptoms were referred to as medically unexplained symptoms or unexplained physical symptoms.

Three studies used the diagnoses of somatisation disorder and somatoform disorder to describe the symptoms and two studies used only the term somatisation. One study spoke of abridged somatisation disorder and two other studies spoke of multiple somatoform symptoms.

Exclusion criteria varied between studies, but often included dementia, severe psychopathology such as psychosis, active suicidal thoughts, alcohol dependence, pregnancy, and current psychological therapy.

102

Non-pharmacological interventions for chronic MUPS

Eleven studies reported severity of symptoms at baseline in terms of number of symptoms. This number varied widely, ranging from a lifetime number of seven symptoms (49), to a current number of 32 symptoms (51).

Interventions

As described in the Types of interventions section, we aimed to select studies investigating psychological therapies, as well as studies on physical therapies. We found no studies on physical therapies that were eligible for inclusion. All 21 included studies evaluated a form of psychological therapy. We classified psychological therapies into six subcategories, as pre-defined by the Cochrane Depression, Anxiety and Neurosis Review Group: CBT, behaviour therapy, or other therapies such as third-wave CBT, psychodynamic therapy, humanistic therapy, or integrative therapies. Fourteen studies described certain forms of CBT. Two studies evaluated behaviour therapies (38, 40). Two studies described third-wave CBT (mindfulness) (50, 13), and two studies described psychodynamic therapies (29, 37). In the study of Kolk et al., participants received CBT, client-centred or eclectic therapy, depending on the therapist the participant was assigned to (48); we classified this as integrative therapy. None of the included studies described humanistic therapies.

In eight studies, the participants received group therapy, and in 11 studies they received individual therapy. In one study, participants received both (51), and in one study there were two intervention groups of which one group received group CBT, and one group received personal CBT (42).

The duration of treatment ranged from one day (one single session) (49) to nine months (37), most often between one and three months.

The mean number of sessions varied among studies and ranged from one session (49) to 13 sessions (24). Thirteen studies used five to 10 sessions. Four studies used one to five sessions (26, 49, 40, 45). Four studies used more than 10 sessions (29, 37, 44, 24).

All studies performed follow-up assessment, but one did not report the outcomes of all follow-ups (40). Reported duration of follow-up varied between two weeks (45) and 24 months (40).

103

5

Chapter 5

Comparisons

As described in the Types of interventions section, we aimed to select the following comparator interventions: usual treatment or waiting list, attention or psychological placebo, and other psychological/physical therapies. Fifteen studies compared an intervention to usual treatment or a waiting list. One of these studies had two intervention groups (receiving psychological therapy) and one control group (receiving usual care) (42). None of the included studies described a placebo comparator intervention, but five included studies compared an intervention with enhanced or structured care (i.e. more than just usual care or a waiting list condition) (50, 29, 51, 44, 43). We had not foreseen this comparator at the protocol stage, so we added this later as an additional comparison. Examples of the enhanced care control condition were a basic training for GPs in the detection and management of psychiatric disorders (44).

One study used compared two psychological therapies (23). This study also included a waiting list group, but we excluded data from this group from our analysis as participants were not randomly assigned to this group. We found no studies that compared psychological interventions with physical therapies.

In one study, GPs in both study arms were trained in diagnosis and management of medically unexplained symptoms (37). In addition, the GPs in the intervention group conducted group sessions for people with MUPS, together with a psychosomatic specialist.

In six studies, in both study arms a consultation letter was sent to the primary care physician after baseline assessment, in addition to the planned psychological therapy (46, 39, 50, 38, 42, 51). This was not a reason for exclusion, and we categorised these studies according to the main comparison. In sensitivity analyses, we assessed the effect of the interventions excluding these studies.

Risk of bias in included studies We classified the methodological quality of the 21 studies according to The Cochrane Collaborationâ&#x20AC;&#x2122;s tool for assessing the risk of bias. Figure 2 presents the risk of bias summary figure.

104

Risk of bias in included studies We classified the methodological quality of the 21 studies according to The Cochrane

Non-pharmacological interventions for chronic MUPS

Collaboration's tool for assessing the risk of bias. Figure 2 presents the risk of bias summary figure.

Figure2.2.Risk Risk bias summary: review authors’ judgments’ about each riskitem of bias Figure of of bias summary: review authors’ judgments’ about each risk of bias for item for each included study. each included study.

5

86

105

Chapter 5

Allocation (selection bias) Sequence generation

While all studies specified that participants were randomly allocated to conditions (or GP practices randomised to treatment or control conditions), there were two studied that did not describe how sequence generation was performed (47, 23).

Therefore, we rated them as ‘uncertain’. We rated the other studies as ‘low risk’ as they all used random sequence generation methods, whether by computer or nondigital, for example, using random number tables (48, 49), or a sequence of labelled cards in envelopes or bags (50, 40, 45, 43).

Allocation concealment

For five studies, it was unclear who performed allocation, or whether the person allocating participants to the trial groups was independent. Therefore, we rated these studies ‘unclear’ (46, 39, 47, 38, 23). We rated the remaining 16 studies ‘low risk of bias’ as there was an adequate description of the person performing allocation or the relation to the researchers and therapists(e.g. “randomisation was carried out independently by a nurse who was not participating in the study”) (33).

Blinding (performance bias and detection bias) Blinding of participants and personnel

In 18 studies, blinding of participants and personnel was not possible, due to the nature of the interventions (e.g. psychological group therapy versus waiting list). As this may have influenced the judgement, we rated almost all studies ‘high risk’. We rated two studies ‘unclear’ because one of the two groups (participants or personnel) was blinded and the other was not (43, 37). One study did not describe blinding of personnel (23), and, therefore, we rated it ‘unclear’.

Blinding of outcome assessment

In 19 studies, blinding of outcome assessment was not possible as most outcomes were participant reported. In one study, outcomes were assessed by blinded interviewers, but they did this together with the participants (who were not blinded) (39).

We rated this study ‘unclear’. One study mainly used clinician-rated instruments (42).

106

Non-pharmacological interventions for chronic MUPS

The outcome assessor was blinded, but, as there also were a few participant report instruments (and participants were not blinded), we rated this study ‘unclear’.

Incomplete outcome data (attrition bias) All studies reported follow-up rates; nine (43%) studies reported a loss to followup of 20% or less. We rated these studies ‘low risk’. We rated one study ‘unclear’, because it had a high loss to follow-up, but corrected for this statistically by multiple imputation (50). The remaining 11 studies reported high loss to follow-up (greater than 20%) and, therefore, we rated them ‘high risk’.

Selective reporting (reporting bias) Seventeen studies reported all intended outcomes and, therefore, we rated them ‘low risk’. For one study, a protocol was lacking, therefore it was impossible to evaluate the possibility of selective outcome reporting (50). We rated this study ‘unclear’.

We rated the remaining three studies ‘high risk’. In Kashner 1995, the outcome ‘days in bed’ was described as assessed, but was not reported in the article (47). In Moreno 2013, healthcare use and CGI were mentioned as outcomes in the protocol, but they were not reported (42). Schilte 2001 performed follow-up measurements at six, 12, and 24 months, but only reported outcomes of the last follow-up moment (40).

Treatment fidelity Sixteen studies used a treatment manual or protocol for studied treatments. We rated them ‘low risk’. Three studies did not apply a structured intervention according to a protocol (26, 48, 40), therefore, we rated them ‘high risk’. The two remaining studies did provide information about a form of structure in treatment, but did not mention a protocol or manual for this. We rated them ‘unclear’.

Researcher allegiance In 18 studies, researchers did not report to have a preference for one of the treatment modalities. In the studies of Burton 2012, Lidbeck 1997, and Schaefert 2013, an author was also (one of) the therapist(s), which may have caused some bias.

Therefore, we rated these studies ‘unclear’ (26, 33, 37).

107

5

Chapter 5

Other potential sources of bias We included two multiple intervention studies (42, 23). In the first study, data were presented for each groups to which participants were randomised, so no other potential sources of bias were found (rating: ‘low risk’). In the second study, participants were randomised for CBT or progressive muscle relaxation (PMR) using random sequences. When both groups were full, newly included participants were allocated to the waiting list group. In a later stage, these participants were included in both intervention groups. As participants were their own controls due to this method, we decided to exclude data from the waiting list group from analysis. For this reason, we rated this study ‘unclear’.

One of the studies was a CRCT (37). GPs were randomised, after which individuals were recruited. We considered the randomisation method and statistical analysis appropriate for the study design.

In the studies of Schilte 2001 and Katsamanis 2011, we found considerable baseline imbalances (40, 38). In the study of Schweickhardt 2007, a high percentage (29%) of participants from the control group became involved in psychotherapy (45). This may have influenced the results, although this study provided data for only one outcome (acceptability) and the effects were in the same order of magnitude as in other studies. We rated these three studies ‘unclear’.

Effects of interventions For the description of the results, we stratified the comparisons in the following way (as per the categories of therapies presented in Types of interventions, where data allowed): 1. Psychological therapies versus usual care or waiting list

a.

CBT versus usual care or waiting list

b. Behavioural therapy versus usual care or waiting list

c.

d. Psychodynamic therapy versus usual care or waiting list

e. Integrative therapies versus usual care or waiting list

Third-wave CBT versus usual care or waiting list

2. Psychological therapies versus enhanced or structured care

108

a.

CBT versus enhanced or structured care

Non-pharmacological interventions for chronic MUPS

b. Third-wave CBT versus enhanced or structured care

c.

Psychodynamic therapy versus enhanced or structured care

3. Psychological therapy versus another psychological therapy

a.

CBT versus behavioural therapy

Most studies provided data for some of the outcomes. One study did not provide any data that were suitable for meta-analysis, because the authors only reported change scores (40). Across all comparisons, outcomes, and time points, we created 44 forest plots. Most of these included data from only a limited number of studies: 25 of the forest plots included three or fewer studies, only two included 10 or more studies.

Below, we present the results of the meta-analyses. We also give attention to the subgroups that included a considerable proportion of the studies contributing to the overall comparisons: CBT versus usual care or waiting list and CBT versus enhanced or structured care, because these subgroups were more homogeneous in terms of type of intervention than the overall comparisons. In terms of risk of bias, the studies that provided outcomes for the meta-analyses were representative for the whole group (i.e. covered the broad spectrum of risk of bias assessments across items).

1. Psychological therapy versus usual care or waiting list Fifteen studies, with 1805 randomised participants, compared some form of psychological therapy with usual care or waiting list controls. They addressed the following psychological therapies: 1. CBT versus usual care or waiting list

a. Ten studies, 1037 randomised participants (46, 26, 39, 47, 33, 49, 42, 45, 43, 24)

2. Behavioural therapy versus usual care or waiting list

a.

Two studies, 209 randomised participants (38, 40)

3. Third-wave CBT versus usual care or waiting list

a.

One study, 125 randomised participants (13)

4. Psychodynamic therapy versus usual care or waiting list

a.

One study, 328 randomised participants (37)

5. Integrative therapies versus usual care or waiting list

a.

One study, 106 randomised participants (48)

109

5

Chapter 5

In four of these studies, this was combined with a consultation letter sent to the primary care physician after baseline assessment, in both treatment arms (46, 39, 38, 42). A consultation letter provided recommendations for the primary care physician tailored to the individual personâ&#x20AC;&#x2122;s diagnosis, symptoms, and problems. In one study, the GPs in both treatment groups were trained in diagnosis and management of MUPS (37).

Apart from CBT, for each of the other types of psychological therapy only one study provided outcomes. Hence, for each of these separate treatment types there was insufficient evidence. Below, we described results for the whole group and for the subgroup of studies that compared CBT with usual treatment.

Primary outcomes 1.1 Severity of somatic symptoms

Combining all studies that compared some psychological therapy with usual care or waiting list, psychological therapies were significantly more effective at end of treatment, though the effect was small (SMD -0.34; 95% CI -0.53 to -0.16; 10 studies, 1081 analysed participants) (figure 1). Heterogeneity was moderate (I2 = 49%), and the overall quality of the evidence was low. Compared with usual care, the subgroup of studies that used CBT were also significantly more effective in reducing severity of symptoms at end of treatment (SMD -0.37; 95% CI -0.69 to -0.05; 6 studies, 593 participants, random-effects model). Heterogeneity was substantial (I2 = 70%), and the overall quality of the evidence was low. The point estimates of all but one of the studies favoured the CBT group. The two studies with the smallest effects offered low-intensity CBT (27, 49). A post-hoc analysis without these two studies provided an SMD of -0.58 (a moderate effect size) (95% CI -0.77 to -0.38) and reduced heterogeneity (I2 = 0%.

At follow-up, measurements within one year of follow-up, the effect of psychological therapies remained significant (SMD -0.24; 95% CI -0.37 to -0.11; 7 studies, 950 participants; I2 = 0%). The same was the case for the subgroup of CBT studies (SMD -0.29; 95% CI -0.49 to -0.09; 4 studies, 496 participants). Heterogeneity was low (I2 = 17%).

110

Non-pharmacological interventions for chronic MUPS

Only two studies (all of CBT) with 228 participants provided data for this severity of symptoms beyond one year of follow-up (SMD -0.52; 95% CI -0.80 to -0.24). Heterogeneity was low (I2 = 0%).

Figure 3. Psychological therapies versus usual care or waiting list controls

5

1.2 Acceptability

Compared with usual care, psychological therapies resulted in a higher proportion of drop-outs (RR acceptability 0.93; 95% CI 0.88 to 0.99 favouring usual care; 14 studies, 1644 participants). Heterogeneity was moderate (I2 = 70%). For the studies comparing CBT with usual care, results were of the same magnitude but no longer statistically significant (RR acceptability 0.93; 95% CI 0.85 to 1.01 favouring usual

111

Chapter 5

care; 10 studies, 1037 participants). Heterogeneity was considerable (I2 = 78%). The overall quality of the evidence for this outcome was moderate.

Secondary outcomes 1.3 Severity of anxiety or depressive symptoms (or both)

For participant-rated anxiety symptoms, there was no significant difference at end of treatment (SMD 0.06; 95% CI -0.20 to 0.32, 4 studies, 270 participants). For the studies comparing CBT with usual care, results were similar (SMD 0.07; 95% CI -0.22 to 0.37; 3 studies, 185 participants). Within one year of follow-up only two studies were available (SMD 0.18; 95% CI -0.22 to 0.58; 134 participants).

For clinician-rated anxiety symptoms at end of treatment, there was a statistically significant difference at end of treatment in favour psychological therapies (SMD -0.40; 95% CI -0.63 to -0.17; 3 studies, 320 participants). Within and beyond one year of follow-up, differences remained statistically significant (within one year: SMD -0.66; 95% CI -1.15 to -0.18, 2 studies both CBT, 251 participants; beyond one year: SMD -0.91; 95% CI -1.26 to -0.55; 1 study, 156 participants).

For participant-rated depressive symptoms, there was no significant difference at end of treatment (SMD -0.03; 95% CI -0.22 to 0.16; 6 studies, 661 participants).

Similar results were found for the studies that compared CBT with usual care (SMD 0.09; 95% CI -0.13 to 0.31; 4 studies, 325 participants), and for outcomes after not more than one year of follow-up (SMD 0.04; 95% CI -0.34 to 0.42; four studies, 535 participants).

For clinician-rated depressive symptoms, there was a statistically significant difference at end of treatment in favour of psychological therapies (SMD -0.25; 95% CI -0.48 to -0.02; 3 studies, 316 participants). Within one year of follow-up, the difference was no longer statistically significant (SMD -0.55; 95% CI -1.17 to 0.07; 2 studies, 251 participants). Only one study reported on this outcome beyond one year after treatment (SMD -0.81; 95% CI -1.16 to -0.46; 156 participants).

1.4 Dysfunctional cognitions, emotions, and behaviours

Three studies, two of which compared CBT with usual care, with 440 participants, reported on dysfunctional cognitions, emotions, and behaviours. At end of treatment,

112

Non-pharmacological interventions for chronic MUPS

there was no significant difference between the two groups (SMD -0.11; 95% CI -0.37 to 0.16). The quality of the evidence was moderate. At follow-up within one year, differences remained non-significant (SMD -0.16; 95% CI -0.38 to 0.07).

1.5 Adverse events

Only three studies, all comparing CBT with usual care, reported on adverse events during the treatment period. One study could not be included in the meta-analysis, because no adverse events were found in both groups. The pooled result of the other two studies also showed no significant differences between both conditions (RR 1.31; 95% CI 0.47 to 3.66; 445 participants; I2 = 0%).

1.6 Treatment response (clinician rated)

All four studies addressing clinician-rated treatment response comparing CBT with usual care. At end of treatment, results strongly favoured the treatment group (RR 3.30; 95% CI 2.08 to 5.21; 4 studies, 391 participants; I2 = 19%). We considered the quality of the evidence to be low for this outcome. Three studies provided data for clinician-rated treatment response within one year after end of treatment, still in favour of the treatment group (RR 2.53; 95% CI 1.25 to 5.10; 332 participants; I2 = 59%). At longer follow up (greater than one year after treatment) only two studies reported outcomes, highly favouring the treatment group (RR 10.31; 95% CI 2.95 to 36.02; 240 participants).

1.7 Functional disability and quality of life

Seven studies, of which four addressing CBT reported on functional disability and quality of life, using a variety of instruments. At the end of treatment, a statistically significant effect was found favouring the psychological therapies (SMD 0.17; 95% CI 0.03 to 0.32; 7 studies, 730 participants; I2 = 0%). We judged the evidence to be moderate. At follow-up within one year after treatment, differences were similar but no longer significant (less than one year: SMD 0.16; 95% CI -0.01 to 0.33; 4 studies, 526 participants; I2 = 0%). After one year, only one study provided data for functional disability and quality of life.

113

5

Chapter 5

Four studies compared CBT with usual care. At end of treatment, a non-significant difference was found favouring CBT (SMD 0.15; 95% CI -0.06 to 0.37; 4 studies, 341 participants; I2 = 0%).

1.8 Health care use

Six studies assessed healthcare use, operationalized in different ways, with moderate quality of evidence. During the treatment phase, two studies found a significant difference in the number of participant-initiated doctor visits and medication usage in favour of CBT (SMD -0.68; 95% CI -1.06 to -0.30; 117 participants). In the period less than one year after treatment, perhaps a more relevant timeframe, four studies found no clear evidence of a difference (SMD -0.09; 95% CI -0.31 to 0.12; 532 participants; I2 = 20%). We judged the quality of the evidence to be moderate. For one of the studies, the effect was in the opposite direction, that is, favouring the control group (48). No study provided data for healthcare use beyond one year after treatment. See footnotes of analyses for details about the way healthcare use was assessed.

2. Psychological therapy versus enhanced or structures care Five studies with 680 randomised participants compared a certain psychological therapy with enhanced or structured care. They addressed the following treatments: 1) CBT versus enhanced or structured care

a) Three studies, 349 randomised participants (51, 44, 43)

2) Third-wave CBT versus enhanced or structured care

a) One study, 120 randomised participants (50)

3) Psychodynamic therapy versus enhanced or structured care

a) One study, 211 randomised participants (29).

In two of these studies, treatment was combined with a consultation letter sent to the primary care physician after baseline assessment, in both treatment arms (50, 51).

Below we describe the main results, sorted by outcomes. Apart from CBT, only one or two trials provided data for each of the three other types of psychological therapy; hence, for each of these other treatment types there was insufficient evidence.

114

Non-pharmacological interventions for chronic MUPS

Primary outcomes 2.1 Severity of somatic symptoms

Five studies (with 624 analysed participants) assessed severity of somatic symptoms comparing some psychological therapy versus enhanced care (pooled SMD -0.19; 95% CI -0.43 to 0.04; I2 = 53%). We considered the quality of the evidence to be low.

Within one year of follow-up, this effect was similar but now statistically significant (SMD -0.21; 95% CI -0.40 to -0.02; 5 studies, 593 participants; I2 = 25%). Only two studies each comparing a different psychological therapy to enhanced care, assessed severity of somatic symptoms beyond one year after treatment (SMD -0.32; 95% CI -0.73 to 0.10; 172 participants). The subgroup of studies comparing CBT with enhanced care showed similar results. Heterogeneity was substantial at the end of treatment (I2 = 62%) and moderate within one year after treatment (I2 = 39%).

5

2.2 Acceptability

Five studies, with 679 analysed participants, showed that psychological therapies were less acceptable in terms of drop-outs than enhanced care (RR 0.93; 95% CI 0.87 to 1.00). Heterogeneity was moderate (I2 = 36%), and we judged the quality of the evidence to be moderate. The largest subgroup was CBT. Compared with enhanced care, moderate-quality evidence showed that there was no clear difference between CBT and enhanced or structured care (RR 0.91; 95% CI 0.82 to 1.02; 3 studies, 331 participants). Heterogeneity was moderate to considerable (I2 = 50%).

Secondary outcomes 2.3 Severity of anxiety or depressive symptoms (or both)

Five studies assessed severity of anxiety or depressive symptoms (or both) at end of treatment (SMD -0.14; 95% CI -0.30 to 0.02; 624 analysed participants; I2 = 0%), showing no clear difference. Similar results were found within one year after treatment (SMD -0.13; 95% CI -0.29 to 0.03; 5 studies, 593 participants) and beyond one year after treatment (SMD -0.26; 95% CI -0.55 to 0.03; 2 studies, 184 participants).

The studies investigating CBT showed no significant difference in level of anxiety and depressive symptoms between CBT and enhanced care at end of treatment (SMD

115

Chapter 5

-0.17; 95% CI -0.40 to 0.05; 3 studies. 307 participants) and within one year after treatment (SMD -0.17; 95% CI -0.40 to 0.06; 3 studies, 289 participants). Heterogeneity was low (I2 = 0% at end of treatment and within one year after treatment). Only one CBT study reported on severity of anxiety or depressive symptoms (or both) beyond one year after treatment.

2.4 Dysfunctional cognitions, emotions, and behaviours

Four studies, with 499 analysed participants, provided data for dysfunctional cognitions, emotions, and behaviours at end of treatment, showing no clear evidence of a difference between psychological therapy and enhanced care (SMD -0.09; 95% CI -0.29 to 0.10; I2 = 14%). We judged quality of the evidence to be moderate. At followup within one year after treatment, the difference was statistically significant (P value = 0.05), favouring the psychological therapy over enhanced care (SMD -0.24; 95% CI -0.49 to 0.00; 4 studies, 477 participants; I2 = 42%). Beyond one year of follow-up, only two studies reported on dysfunctional cognitions, emotions, and behaviours and showed no significant difference (SMD -0.58; 95% CI -1.27 to 0.11; 2 studies, 184 participants; I2 = 82%).

The two studies comparing CBT with enhanced care showed no clear evidence of a difference in dysfunctional cognitions, emotions, and behaviours at end of treatment (SMD -0.28; 95% CI -0.57 to 0.01; 2 studies, 182 participants). Heterogeneity was low (I2 = 0%). Within one year after treatment, levels of dysfunctional cognitions, emotions, and behaviours were significantly lower for CBT (SMD -0.45; 95% CI -0.83 to -0.07; 2 studies, 173 participants), though more heterogeneous (I2 = 37%). This effect was even more significant at more than one year after treatment, although this comparison only included one study (SMD -0.94; 95% CI -1.36 to -0.51; 94 participants) (51).

2.5 Adverse events

None of the studies comparing psychological therapy versus enhanced or structures care reported information about adverse events.

116

Non-pharmacological interventions for chronic MUPS

2.6 Treatment response

None of the included studies comparing psychological therapy versus enhanced or structures care reported about treatment response using a standardised method as described in Secondary outcomes. One study reported about treatment response, but for this outcome measure the SF-36 was used (51). In this review, we used the outcomes of this questionnaire in the analyses of functional disability. Another study reported about participants’ perceived change in symptoms (44). At all measurement moments after baseline, participants were asked if their symptoms were “recovered”, “improved”, “the same”, or “worse” since the previous measurement, using a non-standardised questionnaire. At the end of treatment, 32 (82%) participants in the intervention group declared that symptoms were improved or recovered versus 24 (64%) participants in the control group. Six months after treatment, 27 (73%) participants of intervention group reported recovery or improvement versus 23 (59%) participants of the control group.

5

2.7 Functional disability and quality of life

At end of treatment, four studies with 497 analysed participants reporting on functional disability and quality of life, found no significant difference (SMD 0.13; 95% CI -0.05 to 0.30; I2 = 0%). We considered the quality of the evidence to be moderate.

Within one year of follow-up, there was a small effect in favour of psychological therapies (SMD 0.20; 95% CI 0.02 to 0.38; 5 studies, 727 participants; I2 = 0%). Only two studies reported on functional disability and quality of life beyond one year of follow-up and there was no clear evidence of a difference between the interventions (SMD 0.22; 95% CI -0.16 to 0.60; 2 studies, 184 participants).

For the studies comparing CBT with enhanced care, at end of treatment, moderatequality evidence showed no significant difference in terms of level of function/quality of life, with a large CI but homogeneous population (SMD 0.21; 95% CI -0.08 to 0.51; 2 studies, 182 participants; I2 = 0%). There was a small but significant difference in favour of CBT within one year after treatment (SMD 0.30; 95% CI 0.00 to 0.60; 2 studies, 173 participants). At this time point, heterogeneity was low (I2 = 0%). After one year of follow-up, only one study provided data. In this study, CBT resulted in a

117

Chapter 5

significantly higher level of function compared with enhanced care (SMD 0.42; 95% CI 0.01 to 0.83; 94 participants).

2.8 Healthcare use

Only two studies provided usable data for this analysis and quality of the evidence was low (41, 29). There were no significant differences healthcare use between psychological therapies and enhanced care, neither at end of treatment, nor within one year after end of treatment. See footnotes of analyses for details about the way healthcare use was assessed.

3. Psychological therapy versus other psychological therapy Only one included study addressed psychological therapy versus other psychological therapy (23; 173 randomised participants). The study compared CBT with PMR therapy. The study also included a waiting list group, but we excluded this group from analyses as participants in the waiting list group were not randomly assigned.

Primary outcomes and secondary outcomes 3.1 Severity of somatic symptoms

No significant difference was found for severity of somatic symptoms between CBT and PMR at end of treatment (SMD 0.10; 95% CI -0.33 to 0.53; 84 participants).

3.2 Acceptability

There was no significant difference in drop-out rates between CBT and PMR during treatment (SMD 0.98; 95% CI 0.83 to 1.15; 90 participants).

3.3 Severity of anxiety or depressive symptoms (or both) at end of treatment

There was no significant difference in level of depression and anxiety between CBT and PMR at end of treatment (SMD 0.01; 95% CI -0.42 to 0.44; 84 participants).

3.4 Dysfunctional cognitions, emotions, and behaviours

The study did not report about dysfunctional cognitions, emotions, and behaviours.

118

Non-pharmacological interventions for chronic MUPS

3.5 Adverse events

The study comparing CBT with PMR did not report about adverse events.

3.6 Treatment response

The study comparing CBT with PMR did not report about treatment response.

3.7 Functional disability and quality of life

There was no significant difference in level of function between CBT and PMR at end of treatment (SMD 0.28; 95% CI -0.15 to 0.71; 84 participants).

3.8 Healthcare use

The study comparing CBT with PMR did not report about healthcare use.

DISCUSSION

5

Summary of main results Psychological therapy versus usual care

Fifteen studies compared some form of psychological therapy with usual care or a waiting list. Combining 10 of these studies, the psychological therapy was significantly more effective on symptom severity at end of treatment, though the effect was small. Heterogeneity was considerable and the overall quality of the evidence was low. Six of the 10 studies compared CBT with usual care; for this subgroup it was also apparent that CBT was more effective in reducing severity of symptoms at the end of treatment. The treatment effect of psychological therapies as a whole was also noted within one year of follow-up (seven studies). After one year, the evidence was limited to two studies (both CBT), but still in favour of the psychological therapy.

Results for treatment response, one of our secondary outcomes, supported the findings for symptom severity, with moderate-quality evidence.

Regarding the other primary outcome, acceptability, we found a 7% difference in drop-outs, favouring the usual care group. The quality of the evidence was moderate. After we removed an apparent outlier, the result was smaller (5%), but still

119

Chapter 5

statistically significant. There was no significant difference in drop-out rates between CBT and usual care.

For participant-rated symptoms of depression and anxiety, there was no significant difference at the end of treatment or at follow-up. In three studies using clinician-rated instruments, the level of anxiety and depression was slightly lower in the psychological therapy groups at the end of treatment. For anxiety, this difference became larger at follow-up. For clinician-rated depressive symptoms, this effect fluctuated during follow-up (no effect within one year (two studies) and a large effect after one year of follow-up (one study)). Only three studies reported adverse effects and dysfunctional cognitions, emotions, and behaviours. There was no clear evidence of a difference on these outcomes. There was a small difference in functional disability at the end of treatment favouring psychological therapies. This effect was not apparent during follow-up. Two studies (both on CBT) found a small difference in favour of psychological therapies on healthcare use during treatment, four studies found no effect within one year of follow-up. Due to the small number of studies, these results should be considered with caution.

Only two studies compared behavioural therapy with usual care, of which only one provided relevant data (38). In this study, there were no significant differences for any of the outcomes. Only one study compared third-wave CBT (mindfulness therapy) with usual care (13). In this study, mindfulness was more acceptable than usual care, but no evidence of differences was found with respect to other outcomes. One study compared a variety of psychological therapies with usual care (therapy depended on the orientation of the 15 participating therapists) (48). In this study, there was no evidence of differences with respect to any of the outcomes. This comparison had a high external validity as it emulated the way the referral process normally works.

Psychological therapy versus enhanced or structured care

Five studies compared a certain psychological therapy with enhanced or structured care. The quality of the evidence was moderate for most outcomes. At the end of treatment, there was no clear evidence of a difference for symptom severity, but there was a small statistically significant difference within one year after end of

120

Non-pharmacological interventions for chronic MUPS

treatment. The psychological therapy groups had a 7% higher drop-out rate than the control groups.

There was no clear evidence of a difference between the groups in terms of severity of anxiety or depressive symptoms (or both) at the end of treatment and within one year after treatment. There was no clear evidence of a difference between the groups in terms of dysfunctional cognitions, emotions, and behaviours at end of treatment, but at follow-up within one year of treatment there was a small effect in favour of psychological therapy over enhanced care. None of the studies in this comparison reported information about adverse events or treatment response in a standardised way. For functional disability and quality of life, there was no clear evidence of a difference at the end of treatment, but there was a small significant difference within one year of follow-up. There were no significant differences in healthcare use between psychological therapies and enhanced care.

Three of the studies compared CBT with enhanced or structured care. For symptom severity, CBT showed similar results as the whole group. There were no differences in drop-out rates. In addition, there were no significant differences in levels of anxiety and depressive symptoms at the end of treatment and within one year after treatment. Only one study reported data after one year. At the end of treatment, CBT did not result in lower levels of dysfunctional cognitions, emotions, and behaviours, compared with enhanced care. Within one year of treatment, these levels were lower for CBT (two studies). Only one study reported beyond one year of treatment.

The level of functional disability at the end of treatment was comparable for CBT and enhanced care. Within and after one year of treatment there was a small difference in favour of CBT, although only a few studies were included in these analyses. Only one CBT study reported data about healthcare use and found no evidence of difference.

Psychological therapy versus another psychological therapy

Only one study compared two forms of psychological therapy (CBT versus PMR).

There were no differences between the groups for any of the outcomes.

121

5

Chapter 5

Studies A thorough literature search in electronic databases and many other resources such as conference proceedings, international trial registers, grey unpublished literature, and reference lists resulted in 21 studies that could be included in this review. In comparison to other existing reviews about non-pharmacological interventions for MUPS or somatoform disorders (e.g. 54, 15), this number of eligible studies is quite high. Hence, a considerable number of studies was available in order to address our questions. Only a few studies contributed to most of the outcomes. In addition, due to the small number of studies, we were unable to consider the effect of study characteristics (setting, severity, chronicity) on the outcomes.

We believe that the included studies cover a broad spectrum of settings, and both RCTs and CRCTs were included. Participants were recruited in various ways and from various healthcare settings, including primary care, secondary care, tertiary care, and the open population. In the included studies, therapists had different backgrounds (e.g. GPs, psychologists, and other physicians) and different levels of experience. A limitation of the included studies was the relatively low number of included participants per study as most studies only included 25 to 75 participants per study arm.

Participants With only two exceptions (41, 43), studies were performed in developed countries (Western Europe and USA). Most studies randomised more women than men. This is in line with existing reviews, as MUPS and somatoform disorders are more common among women. Included studies cover a broad age range. However, as the mean age of participants was in the 30s or 40s in most of the studies, it may be possible that younger and older people were relatively underpresented. Severity of MUPS at baseline was mostly analysed based on the number of symptoms or duration (or both) of symptoms. The number of symptoms at baseline varied widely, ranging from a lifetime number of symptoms of seven (49) to a current number of symptoms of 32 (51). Baseline duration, only reported in nine studies, ranged on average from four to 25 years. This suggests that most of the included participants may have had chronic symptoms at baseline. Included studies also reported high psychiatric co-morbidity rates, percentages of participants with a current co-morbid axis 1 disorder varied

122

Non-pharmacological interventions for chronic MUPS

between 41% (24) and 92% (39). Taking these findings together, we can say that a limitation may be that participants of included studies were people with relatively severe forms of somatoform disorders and MUPS. The milder forms, with lower levels of co-morbidity may have been underrepresented. In contrast, people with milder symptoms may need less intensive therapy.

Interventions Fourteen of the included studies compared CBT with another intervention. As a result, relatively robust conclusions could be drawn about the effectiveness of CBT. The number of studies describing other psychological therapies (such as behavioural therapies, third-wave CBT, or psychodynamic therapies) was too low to draw conclusions about these forms of therapy. Duration and number of treatment sessions varied widely between the included studies.

It is especially remarkable that we found no studies on physical therapies (such as running therapy). We believe that there is a clear need for this type of research.

Many included studies used forms of enhanced care or other forms of therapy as the control treatment. A limitation of this method may be an underestimation of the treatment effect, due to small inter-group differences. This is illustrated since these studies found fewer and smaller effects than studies comparing a treatment with usual care.

Outcomes In this review, the outcome of functional impairment introduced a problem, as a certain number of studies used SF-36 subscales as the outcome measure. As a result, we reported physical functioning and mental functioning or even subdomains separately. We decided to pool the two main domains into one outcome, but this led to the limitation that differences in effects for physical and mental functioning, as found in some studies (24), disappeared.

Another problem of the current review was that, with one exception, there were not enough studies to assess reporting bias with funnel plots. According to recommendations of the Cochrane Handbook for Systematic Reviews of Interventions, there should be at least 10 studies to perform this (21). In future updates of this review, the

123

5

Chapter 5

addition of new studies may enable us to produce funnel plots for more comparisons and outcomes.

Adverse effects were very infrequently reported and various ways of reporting were used. Therefore, it was impossible to extract these in a standardised way in order to include them in our meta-analytical calculations, except for the first comparison (psychological therapy versus usual care or waiting list). We also have to emphasise again that RCTs and CRCTs are not sufficient to gain information about the more rare or longer-term (or both) adverse events.

Quality of the evidence According to the first quality criterion risk of bias defined by the guidelines by GRADE, Figure 2 and Figure 3 showed that in regard to different types of biases most of our included studies showed a low risk. However, a few specific domains were often rated as being at high risk of bias across the studies. Especially for blinding of the outcome measurement, we identified a high risk of bias in most of the included studies. Most studies could not blind the outcome reporters, mostly the participants, due to the nature of the intervention. A high risk of bias in blinding of participants and personnel was found for the same reason. Nine studies (43% of studies) reported incomplete outcomes, defined as a loss to follow-up of more than 20%. Reasons for loss to follow-up were not systematically described. Another study aspect that affected the quality of the evidence was the generally low number of included participants per study. It has to be taken into account that several of the included studies were performed before the publication and implementation of current quality criteria for conducting and reporting RCTs. The small number of studies did not allow us to assess the effects in subgroups of participants or interventions. Apart from CBT, all other comparisons between specific therapies and usual care or enhanced care, the number of studies was too small (often only one study). We did not consider indirectness (a GRADE item comparing the interventions and outcomes in which we are interested to what was actually studied in the included studies) and publication bias to be important sources of risk of bias, publication bias because of our thorough search process; the overall completeness of reporting, and the fact that several studies that did not find an effect.

124

Non-pharmacological interventions for chronic MUPS

Potential biases in the review process This review has several methodological strengths. The quality of meta-analyses depends on the robustness of the search methods used. In this review, the electronic search was thorough and large in scale with broad parameters. We evaluated published and unpublished studies. The selection criteria were broad, which led to the selection of a relatively high number of studies. We also included non-English studies. As a result, it seems likely that all or almost all evidence in the searched databases that should have been included was included. However, as we did not search Asian databases, this may have led to a potential bias.

The study was performed according to a pre-published protocol. Different review authors performed evaluation of studies for selection, extract data, and assess risk of bias, with the possibility of consulting another review author to resolve disputes.

However, due to the fact that not all choices that had to be made were foreseen, there were also post hoc decisions. Excluding studies that trained GPs to deliver some psychological therapy was one of these decisions. In addition, we performed the allocation of the included studies to the different groups of treatment for analyses post hoc. Another post hoc decision was the addition of enhanced or structured care as a comparator. We made decisions very carefully and included achieving consensus between several review authors with specific knowledge in the field. However, some studies were difficult to categorise, as, for example, treatments included elements of different treatment categories. Therefore, allocation of these studies remained slightly arbitrary. Another post hoc decision was to combine the physical component scale and the mental component scale of the SF-36 into one outcome. Other post hoc decisions were to carry out sensitivity analyses by excluding studies that included consultation letters in both study groups, and by excluding studies with the least intensive interventions.

Although we attempted to obtain missing data from the authors of included studies, it was not possible in every case to obtain these data, and, therefore, the included studies were not represented fully in the meta-analyses. This may also have led to a certain form of bias, although it is difficult to say in what direction this bias would be.

As described in the section Types of outcome measures, we aimed to retrieve data about severity/intensity of MUPS; acceptability; depression and anxiety; dysfunc-

125

5

Chapter 5

tional cognitions, emotions, or behaviours; adverse events; treatment response; functional disability; and quality of life. However, the number of studies reporting on many of our outcomes was relatively low. Results about depression; anxiety; dysfunctional cognitions, emotions, and behaviours; adverse events; and treatment response were frequently lacking. Therefore, we could not draw robust conclusions about these outcome measures.

Although acceptability was a primary outcome of our study, we restricted this to the period from randomisation to the end of treatment. We did not take into account the acceptability of the interventions in the recruitment phase. Participants for whom the intervention or control condition was unattractive probably did not participate.

This affects the external validity of study findings.

In this review, we used point estimates at all follow-up periods to evaluate treatment effect, instead of scores based on change from baseline. We chose this method as these results were retrievable from most of the studies, and combining follow-up outcome data with change from baseline data was considered inappropriate given our choice of SMDs due to the variety of outcome measures that had to be combined.

However, pooling the results of follow-up measurements has the disadvantage that baseline values (and possible baseline differences) are not taken into account. As data were pooled in most analyses, we believe that distortions such as these are generally corrected by the other studies in the analyses. Some studies only reported data about change from baseline (without the actual baseline data) (e.g. 26, 50).

These data could not be used in this review. We contacted authors in order to be obtain the required data, and were successful in many cases though not all.

Agreements and disagreements with other studies or reviews Several systematic reviews have addressed non-pharmacological treatments for participants with some form of somatoform disorder or MUPS (e.g. 54, 55, 15). As many of the included studies in our review were published after 2005, we focused this discussion on the systematic reviews that were published after 2005. In general, we can say that the results of this review are in line with results of existing reviews.

In most reviews, the majority of included studies concerned CBT in some form, and small effect sizes were found. In other reviews, also limited evidence was found for

126

Non-pharmacological interventions for chronic MUPS

other forms of psychological therapies. Studies investigating physical therapies for somatoform disorders or MUPS were also hardly reported in other reviews.

Authorsâ&#x20AC;&#x2122; conclusions Implications for practice

The overall quality of the evidence provided by 21 randomised controlled trials was low to moderate. All psychological therapies combined were superior to usual care or waiting list condition for symptom severity, our first primary outcome, but effect sizes were small. As a single treatment, only cognitive behavioural therapy (CBT) was adequately studied to allow conclusions for practice. Compared with usual care or waiting list conditions, CBT reduced somatic symptoms, with a small effect and substantial differences in effects between CBT studies. The effects were durable within and after one year of follow-up.

Compared with enhanced or structured care, psychological therapies generally were not more effective for most of the outcomes. CBT was also not superior to enhanced care. The question remains how specific CBT is over structured improvements of care. No major adverse events were reported in the intervention groups, although most studies did not describe adverse events as an explicit outcome measure. Apart from CBT, neither psychological nor other non-pharmacological therapies have been adequately studied.

In daily practice, a substantial percentage of people with medically unexplained physical symptoms (MUPS) may not be willing to accept psychologically oriented treatments. Whether such acceptance is associated with the effect of psychological treatments for the total MUPS population was not clear. Due to the small number of studies, we could not draw conclusions about the effect of characteristics such as a profession and experience of the therapist, about treatment intensity and treatment location, on treatment efficacy.

Further optimisation of CBT to target optimal participant profiles and match treatment providers, treatment characteristics, and participants could improve outcomes.

Motivating and preparing people for CBT is important for this participant group. As drop-out rates were not much lower than in control groups, this indicates that when

127

5

Chapter 5

a person has accepted involvement in the treatment, the prospects that the treatment will be completed are good.

Implications for research

Based on the findings in this review, we can make several recommendations for future research. The number of studies investigating various treatment modalities other than CBT needs to increase to build a broader and more varied evidence-base for the treatment of somatoform disorders and MUPS. As physical therapies may offer a more acceptable starting point for treatment for these people than psychological approaches, investigating the effectiveness of physical therapies is to be considered.

We found no such studies.

Most studies in our review focused on chronic manifestations of physical symptoms, often of considerable severity. It is conceivable that interventions were more effective in people with milder symptoms, or of shorter duration, but this needs further testing. A related conceptual issue is that chronic conditions deserve a World Health Organization chronic care or chain care approach as acute treatments will not suffice.

Preventing symptoms from become chronic may be a relevant outcome to be added in future studies.

In future research, more attention should be paid to the impact of interventions on risk factors for recurrence and persistence of symptoms in somatoform disorders and MUPS. These factors include anxiety; depression; and dysfunctional cognitions, emotions, and behaviours. Most included studies in this review did not report on all of these factors. Specific attention to the effect of treatment duration and number of treatment sessions is also needed. In the studies included in this review, duration and number of sessions varied widely, and it is yet unclear which treatment intensities are effective for which participants.

Psychological treatments were not superior to enhanced care. It could be argued that an active comparator such as enhanced care underestimates treatment effects.

However, as this comparative treatment is probably cheaper than more intensive psychological interventions, it would deserve further study.

In our view, teaching people how to tolerate uncertainty and deal with their bodily symptoms can be problematic and will probably always involve high levels of clinical

128

Non-pharmacological interventions for chronic MUPS

skills. One potential intermediate factor is the amount of trust that people have in their therapist or physician, a factor to be taken into account in the design of new studies.

There is a clear need for developing and testing strategies for motivating and preparing people for CBT. CBT is the only evidence-based psychological treatment available at the moment.

A more structural question is, how psychological therapies for participants with somatoform disorders can be better integrated into the healthcare system. Can the healthcare system be restructured in such a way that it facilitates the access of people with somatoform disorders to psychological therapies?

As the cost of treatment can be substantial, but also the cost of the disorder in terms of absenteeism and healthcare use, cost-effectiveness needs to be addressed in future studies.

Future studies should include more participants, preferably use a uniform set of validated outcome measurements, and extend follow-up assessments beyond one year after treatment.

Finally, as newer-generation antidepressants and particularly natural products also reduce somatic symptoms, a preference-led or profile-led approach may be possible.

The aim would be to evaluate to what extent an intervention (consisting of a choice between non-pharmacological and pharmacological therapy combined with chain care), would improve symptoms over usual care.

129

5

Chapter 5

REFERENCES 1. Ford AC, Talley NJ, Schoenfeld PS, et al. Efficacy of antidepressants and psychological therapies in irritable bowel syndrome: systematic review and meta-analysis. Gut 2009;58:367-78.

2. Pae CU, Marks DM, Patkar AA, et al. Pharmacological treatment of chronic fatiguesyndrome: focusing on the role of antidepressants. Expert Opinion on Pharmacotherapy 2009;10:1561-70.

3. Saarto T, Wiffen PJ. Antidepressants for neuropathic pain. Cochrane Database of Systematic Reviews 2007;4: CD005454.

4. Moore RA, Wiffen PJ, Derry S, et al. Gabapentin for chronic neuropathic pain and fibromyalgia in adults. Cochrane Database of Systematic Reviews 2014;4:CD007938.

5. Silberstein SD, Peres MFP, Hopkins MM, et al. Olanzapine in the treatment of refractory migraine and chronic daily headache. Headache 2002;45:515-8.

6. Kleinstäuber M, Witthöft M, Steffanowski A, et al. Pharmacological interventions for somatoform disorders in adults [Protocol]. Cochrane Database of Systematic Reviews 2013;7:CD010628.

7. Deary V, Chalder T, Sharpe M. The cognitive behavioural model of medically unexplained symptoms: a theoretical and empirical review. Clinical Psychology Review 2007;27:78197.

8. Sharpe M, Peveler R, Mayou R. The psychological treatment of patients with functional somatic symptoms: a practical guide. Journal of Psychosomatic Research 1992;36:51529.

9. Goldberg D, Gask L, O’Dowd T. The treatment of somatization: teaching techniques of reattribution. Journal of Psychosomatic Research 1989;33:689-95.

10. Malouff JM, Thorsteinsson EB, Schutte NS. The efficacy of problem solving therapy in reducing mental and physical health problems: a meta-analysis. Clinical Psychology Review 2007;27:46-57.

11. Loew TH, Sohn R, Martus P, et al. Functional relaxation as a somatopsychotherapeutic intervention: a prospective controlled study. Alternative Therapies in Health and Medicine 2000;6:70-5.

12. Guerney B, Stollak G, Guerney L. The practicing psychologist as educator - an alternative to the medical practitioner model. Professional Psychology 1971;2:271-2.

13. van Ravensteijn H, Lucassen P, Bor H, et al. Mindfulness-based cognitive therapy for patients with medically unexplained symptoms: a randomized controlled trial. Psychotherapy and Psychosomatics 2013;82:299-310.

14. Noyes R, Stuart SP, Watson DB. A reconceptualization of the somatoform disorders.

Psychosomatics 2008;49:14-22.

15. Rosendal M, Burton C, Blankenstein AH, et al. Enhanced care by generalists for functional somatic symptoms and disorders in primary care. Cochrane Database of Systematic Reviews 2013;10:CD008142.

16. Weyerer S, Kupfer B. Physical exercise and psychological health. Sports Medicine 1994;17:108-16.

130

Non-pharmacological interventions for chronic MUPS

17. Hoedeman R, Blankenstein AH, van der Feltz-Cornelis CM, et al. Consultation letters for medically unexplained physical symptoms in primary care. Cochrane Database of Systematic Reviews 2010;12:CD006524.

18. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1. The Cochrane Collaboration, 2011. Available from www.cochrane-handbook. org.

19. Deeks JJ, Altman DG, Bradburn MJ. Statistical methods for examining heterogeneity and combining results from several studies in meta-analysis. In: Egger M, Smith GD, Altman DG, editor(s). Systematic Reviews in Health Care: Meta-Analysis in Context. 2nd edition. London: BMJ Publishing Group, 2008.

20. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in metaanalyses. BMJ 2003;327:557-60.

21. Sterne JA, Sutton AJ, Ioannidis JP, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ 2011;343:d4002.

22. Gili M, Magallón R, López-Navarro E, et al. Health related quality of life changes in somatising patients after individual versus group cognitive behavioural therapy: a randomized clinical trial. Journal of Psychosomatic Research 2014;76:89-93.

23. Schröder A, Heider J, Zaby A, et al. Cognitive behavioral therapy versus progressive muscle relaxation training for multiple somatoform symptoms: results of a randomized controlled trial. Cognitive Therapy and Research 2013;37:296-306.

24. Zonneveld LN, van Rood YR, Kooiman CG, et al. Predicting the outcome of a cognitivebehavioral group training for patients with unexplained physical symptoms: a one-year follow-up study. BMC Public Health 2012;12:848.

25. Zaby A, Heider J, Schröder A. Waiting, relaxation, or cognitive-behavioral therapy how effective is outpatient group therapy for somatoform symptoms? Zeitschrift für Klinische Psychologie und Psychotherapie 2008;37:15-23.

26. Burton C, Weller D, Marsden W, et al. A primary care symptoms clinic for patients with medically unexplained symptoms: pilot randomised trial. BMJ Open 2012;2:000513 27. Rembold SM. Somatoform disorders in general practice: construction and evalisation of a psychosocial group program. Gruppenpsychotherapie & Gruppendynamik 2011;47:212.

28. Chernyak N, Sattel H, Scheer M, et al. Economic evaluation of brief psychodynamic interpersonal therapy in patients with multisomatoform disorder. PLoS One 2014;9:83894.

29. Sattel H, Lahmann C, Gündel H, Guthrie E, Kruse J, Noll-Hussong M, et al. Brief psychodynamic interpersonal psychotherapy for patients with multisomatoform disorder: randomised controlled trial. British Journal of Psychiatry 2012;200:60-7.

30. Hellman CJC, Budd M, Borysenko J, et al. A study of the effectiveness of two group behavioral medicine Interventions for patients with psychosomatic complaints. Behavioural Medicine 1990;16:165-73.

31. Lupke U, Ehlert U, Hellhammer D. Behavioural medicine at the general hospital: evaluation of treatment for patients with somatoform Disorders. Verhaltenstherapie 1996;6:22-32.

32. Tschuschke V, Weber R, Horn E, et al. Short ambulant psychodynamic group therapy for patients with somatoform disorders. Zeitschrift für Psychiatrie, Psychologie und Psychotherapie 2007;55:87-95.

131

5

Chapter 5

33. Lidbeck J. Group therapy for somatization disorders in general practice: effectiveness of a short cognitive-behavioural treatment model. Acta Psychiatrica Scandinavica 1997;96:14-24.

34. Gottschalk JM, Bleichhardt G, Hiller W, et al. Cognitive-behavioral therapy enriched with emotion regulation training in patients with multiple somatoform symptoms: preliminary findings. Verhaltenstherapie 2011;21:1-39.

35. Pols RG, Battersby MW. Coordinated care in the management of patients with unexplained physical symptoms: depression is a key issue. Medical Journal of Australia 2008;188:S133-7.

36. Steel Z. Randomised controlled trial of cognitive behavior therapy, structured care and treatment as usual in the management of adult primary care patients presenting with 5 or more chronic medically unexplained symptoms in Ho Chi Minh City, Vietnam. http:// www.anzctr.org.au/ACTRN12611000946910.aspx.

37. Schaefert R, Kaufmann C, Wild B, et al. Specific collaborative group intervention for patients with medically unexplained symptoms in general practice: a cluster randomized controlled trial. Psychotherapy and Psychosomatics 2013;82:106-19.

38. Katsamanis M, Lehrer PM, Escobar JI, et al. Psychophysiologic treatment for patients with medically unexplained symptoms: a randomized controlled trial. Psychosomatics 2011;52:218-29.

39. Escobar JI, Gara MA, Diaz-Martinez AM, et al. Effectiveness of a time-limited cognitive behavior therapy type intervention among primary care patients with medically unexplained symptoms. Annals of Family Medicine 2007;5:328-35.

40. Schilte AF, Portegijs PJ, Blankenstein AH, et al. Randomised controlled trial of disclosure of emotionally important events in somatisation in primary care. BMJ 2001;323:86.

41. Sumathipala A, Siribaddana S, Abeysingha MR, et al. Cognitive-behavioural therapy v. structured care for medically unexplained symptoms: randomised controlled trial.

British Journal of Psychiatry 2008;193:51-9.

42. Moreno S, Gili M, Magallรณn R, et al. Effectiveness of group versus individual cognitivebehavioral therapy in patients with abridged somatization disorder: a randomized controlled trial. Psychosomatic Medicine 2013;75:600-8.

43. Sumathipala A, Hewege S, Hanwella R, et al. Randomized controlled trial of cognitive behaviour therapy for repeated consultations for medically unexplained complaints: a feasibility study in Sri Lanka. Psychological Medicine 2000;30:747-57.

44. Speckens AE, van Hemert AM, Spinhoven P, et al. Cognitive behavioural therapy for medically unexplained physical symptoms: a randomised controlled trial. BMJ 1995;311:1328-32.

45. Schweickhardt A, Larisch A, Wirsching M, et al. Short-term psychotherapeutic interventions for somatizing patients in the general hospital: a randomized controlled study.

Psychotherapy and Psychosomatics 2007;76:339-46.

46. Allen LA, Woolfolk RL, Escobar JI, et al. Cognitive-behavioral therapy for somatization disorder: a randomized controlled trial. Archives of Internal Medicine 2006;166:1512-8.

47. Kashner TM, Rost K, Cohen B, et al. Enhancing the health of somatization disorder patients. Effectiveness of short-term group therapy. Psychosomatics 1995;36:462-70.

48. Kolk AM, Schagen S, Hanewald GJ. Multiple medically unexplained physical symptoms and health care utilization: outcome of psychological intervention and patient-related predictors of change. Journal of Psychosomatic Research 2004;57:379-89.

132

Non-pharmacological interventions for chronic MUPS

49. Martin A, Rauh E, Fichter M, et al. A one-session treatment for patients suffering from medically unexplained symptoms in primary care: a randomized clinical trial. Psychosomatics 2007;48:294-303.

50. Fjorback LO, Arendt M, Ornbøl E, et al. Mindfulness therapy for somatization disorder and functional somatic syndromes: randomized trial with one-year follow-up. Journal of Psychosomatic Research 2013;74:31-40 51. Schröder A, Rehfeld E, Ornbøl E, et al. Cognitive-behavioural group treatment for a range of functional somatic syndromes: randomised trial. British Journal of Psychiatry 2012;200:499-507.

52. Fink P, Sørensen L, Engberg M, et al. Somatization in primary care. Prevalence, health care utilization, and general practitioner recognition. Psychosomatics 1999;40:330-8.

53. Fink P, Hansen MS, Oxhøj ML. The prevalence of somatoform disorders among internal medical inpatients. Journal of Psychosomatic Research 2004;56:413-8.

54. Kleinstäuber M, Witthöft M, Hiller W. Efficacy of short-term psychotherapy for multiple medically unexplained physical symptoms: a meta-analysis. Clinical Psychology Review 2011;31:146-60.

55. Kroenke K, Swindle R. Cognitive-behavioral therapy for somatization and symptom syndromes: a critical review of controlled clinical trials. Psychotherapy and Psychosomatics 2000;69:205-15.

5

133

6 How should we manage adults with persistent unexplained physical symptoms?

Madelon den Boeft Nikki Claassen- van Dessel Johannes C. van der Wouden

Submitted (revised)

How to manage persistent unexplained symptoms?

INTRODUCTION Persistent unexplained physical symptoms, i.e. physical symptoms that exist longer than three months and cannot be (sufficiently) explained by an underlying medical condition after adequate examination, are highly prevalent in all health care settings (1-2). These persistent unexplained physical symptoms may lead to functional impairment, high levels of psychological distress, a troubled doctor-patient relationship and increased health care costs (3-5).

In some cases these symptoms fit criteria of specific functional somatic syndromes such as fibromyalgia, irritable bowel syndrome or chronic fatigue syndrome. But often no specific functional somatic syndrome can be diagnosed.

Many studies have been performed on persistent unexplained physical symptoms, but it remains uncertain how clinicians should manage patients with these symptoms. Therefore, in this paper we will describe which interventions are effective, which are not and how to adequately manage these patients in daily clinical practice.

We will not include the specific functional somatic syndromes. For management of these syndromes, we refer to specific guidelines covering this.

6

WHAT IS THE EVIDENCE OF UNCERTAINTY? In September 2015, we searched the Cochrane Library, including the Cochrane central database of controlled trials, PubMed and clinical trial registers (clinicaltrials. gov, controlled-trials.com, who.int/trialsearch) to identify published and ongoing randomised controlled trials (RCTs) examining all interventions as offered in regular health care for patients with persistent unexplained physical symptoms.

To date, four Cochrane reviews have been published on the subject (6-9). To the best of our knowledge, no relevant new studies have been published since.

A Cochrane review examined the efficacy of different types of antidepressants, antipsychotics, the combination of an antidepressant and antipsychotic and natural products (e.g. St. Johnâ&#x20AC;&#x2122;s wort) (26 RCTs, 2159 participants) (6). This review found a significant positive effect for new generation antidepressants and natural products compared to placebo and for the combination of an antidepressant and an antipsy-

137

Chapter 6

chotic compared to an antidepressant alone for severity of unexplained physical symptoms, the primary outcome. The effect size (standardized mean difference SMD) was in the range of 0.70-0.90, which is considered clinically relevant.

There was no significant effect for tricyclic antidepressants compared to placebo, nor for tricyclic antidepressants compared to new generation antidepressants, nor for the different new generation antidepressants.

The quality of the studies was low due to a high risk of bias in many domains across the studies, strong heterogeneity and small sample sizes. Also, conclusions could only be drawn for the short-term, as follow-up was often only 6 weeks and never longer than 12 weeks. The small beneficial effects should be weighed against often occurring side effects that frequently cause discontinuation. No significant difference in acceptability (drop-out rate) was found between the intervention and comparison groups.

Another Cochrane review examined the efficacy of different forms of non-pharmacological interventions (21 RCTs, 2658 participants) (7). This review found that all non-pharmacological therapies taken together and cognitive behavioural therapy alone (CBT: therapy based on the cognitive behavioural model which proposes that unexplained physical symptoms are caused by self-perpetuating multifactorial cycles based on the interaction in several domains) had a significant beneficial effect compared to usual care (i.e. no active intervention initiated by the researchers/ investigators), or compared to waiting list controls for the primary outcome, severity of unexplained physical symptoms. The effect size was small to moderate (SMD 0.340.37).

Psychological therapies had a higher proportion of dropouts during treatment compared to usual care. No significant difference was found between psychological therapy and â&#x20AC;&#x153;enhanced careâ&#x20AC;? (consisting of reattribution, where symptoms are reframed by making the link between physical symptoms and presumed underlying psychological problems, or of CBT delivered by a general practitioner (GP)), but enhanced care seemed somewhat more acceptable. No major harms or side effects were reported. No studies could be included that addressed physical therapy. The overall quality of the studies was rated low to moderate, due to a high risk of bias

138

How to manage persistent unexplained symptoms?

for several items, such as lack of blinding of the participants, therapists and outcome assessors.

A third Cochrane review examined the efficacy of enhanced care (as explained above) and CBT compared to usual care (6 RCTs, 1787 participants) (8). This review found no significant effect of enhanced care on patient outcomes, such as physical symptoms and quality of life. However, the authors did not calculate a pooled effect due to the small sample of studies that also had high risk of bias and strong heterogeneity.

The fourth Cochrane review examined the effectiveness of consultation letters written by psychiatrists to provide GPs with a diagnosis and treatment advice for patients with unexplained physical symptoms (6 RCTs, 449 participants)(9). This review found that the evidence that a consultation letter is effective in reducing the symptoms is limited, as they found that only six small studies of moderate quality were performed, all of them in the United States.

IS ONGOING RESEARCH LIKELY TO PROVIDE RELEVANT EVIDENCE? In September 2015, we searched the clinical trial registers (clinicaltrials.gov, controlled-trials.com, who.int/trialsearch) and identified ten ongoing RCTs evaluating non-pharmacological interventions (including one on the effect of walking training for patients with somatoform disorders (i.e. psychiatric diagnosis where persistent unexplained physical symptoms are central; Schrรถder 2014; unpublished data only), two evaluating pharmacological treatments, one evaluating enhanced care and none for consultation letters.

139

6

Chapter 6

Box 1: Recommendation for future research Large well conducted RCTs with the following features: Population: Adult patients with persistent unexplained physical symptoms for whom the nature and severity of symptoms are described by a) use of a validated and commonly used measurement instrument and b) clearly defined levels of severity (i.e. duration and number of symptoms). New studies also need to include patients with a low number of symptoms or a recent symptom onset.

Intervention: Different pharmacological and non-pharmacological interventions, including physical therapies (such as walking, running or yoga therapy) and enhanced care (reattribution or CBT delivered by the GP). Treatment characteristics (e.g. duration, intensity, dosage, health care provider) need to be specified, e.g. by using a standardized treatment protocol.

Comparison: Usual care, waiting list controls and head-to-head comparisons of different interventions (including pharmacological and non-pharmacological interventions).

Outcome: Severity of symptoms, functional impairment, mental health (including depressive and anxiety symptoms or disorders), treatment acceptability and side effects.

Outcomes should be measured using validated and commonly used measurement instruments. Particularly for pharmacological interventions, a long follow-up duration (minimum of six months) is needed.

WHAT SHOULD WE DO IN THE LIGHT OF UNCERTAINTY? In consultations where the GP considers unexplained symptoms, we advise the following steps to work out. First, explore all symptom dimensions (somatic, cognitive, emotional, behavioral, social). Next, perform a thorough but focused physical examination (depending on the nature of the symptoms) and if necessary, refer for

140

How to manage persistent unexplained symptoms?

additional diagnostics. If reason arises during the consultation, evaluate if the patient has a mental disorder. When a medical condition can be excluded, determine the severity of the unexplained physical symptoms (e.g. high number of symptoms, impact on daily life) and determine if the duration of the symptom is longer than 3 months.

For patients fitting in the category of persistent unexplained physical symptoms, openly share all findings, explain that there is no underlying medical condition and provide a tangible and constructive explanation for the (persistence of) symptoms, for example by using the vicious circle theory (where pain can lead to less exercise and less exercise can lead to more pain). Discuss the functional impairments and if possible give advice how to cope with these.

We recommend that GPs discuss several non-pharmacological treatment alternatives with the patient. The GP should explain that only CBT has proven to have a beneficial effect, albeit small. Together the patient and the GP can decide which treatment optimally supports the patient. We do not advise pharmacological therapy given the rate of side-effects, non-compliance and low quality of studies. When a specific syndrome exists, such as chronic pain, we advise clinicians to follow specific guidelines for these syndromes.

Box 2: What patients need to know • Unexplained physical symptoms are common and can be invalidating when persistent • Sometimes the symptoms fit criteria of specific syndromes such as fibromyalgia or chronic fatigue syndrome, but often this is not the case • Before establishing the unexplained nature of the symptoms, GPs will explore all symptom aspects and perform necessary examinations • In case of unexplained physical symptoms, GPs will discuss negatively influencing factors and will give advice how to cope with the symptoms • There are many treatment options, but not much research has been performed on their effectiveness. Only cognitive behavioural therapy has a proven beneficial effect • Pharmacological treatment is not recommended

141

6

Chapter 6

Box 3: How patients were involved in the creation of this article Two patients with MUPS were asked if this article covers issues that matter to them and what information might help doctors to better manage them. They suggested to provide explanations of some definitions (such as CBT) and to stimulate doctors to talk openly with their patient about all phases of exploration and treatment of unexplained symptoms. These suggestions were adopted in the paper.

Box 4: What you need to know • If you consider symptoms to be unexplained: explore all symptom dimensions (somatic, cognitive, emotional, behavioural, social) and perform a thorough but focussed physical examination • In case of unexplained symptoms: be open about your findings and provide a tangible and constructive explanation for the (persistence of) symptoms (for example by using the vicious circle theory) • Discuss several treatment options with the patient • Discuss referral for CBT for patients with persistent unexplained physical symptoms • Discourage pharmacological treatment • In case of a specific functional syndrome, follow the specific guideline

142

How to manage persistent unexplained symptoms?

REFERENCES 1. Burton C. Beyond somatisation: a review of the understanding and treatment of medically unexplained physical symptoms. Br J Gen Pract. 2003;53:231–9.

2. Aamland A, Malterud K, Werner EL. Patients with persistent medically unexplained physical symptoms: a descriptive study from Norwegian general practice. BMC Fam Pract. 2014;15:107 3. Jackson JL, Passamonti M. The outcomes among patients presenting in primary care with a physical symptom at 5 Years. J Gen Intern Med. 2005;20:1032–7.

4. Smith GR, Monson RA, Ray DC. Patients with multiple unexplained symptoms: their characteristics, functional health, and health care utilization. Arch Intern Med. 1986; 146:69–72 5. Dirkzwager AJ, Verhaak PF. Patients with persistent medically unexplained symptoms in general practice. BMC Fam Pract. 2007;8:33 6. Kleinstäuber M, Witthöft M, Steffanowski A, et al. Pharmacological interventions for somatoform disorders in adults. Cochrane Database Syst Rev. 2014;11:CD010628.

7. van Dessel N, den Boeft M, van der Wouden JC, et al. Non-pharmacological interventions for somatoform disorders and medically unexplained physical symptoms (MUPS) in adults. Cochrane Database Syst Rev. 2014;11:CD011142.

8. Rosendal M, Blankenstein AH, Morriss R, et al. Enhanced care by generalists for functional somatic symptoms and disorders in primary care. Cochrane Database Syst Rev. 2013;10:CD008142.

9. Hoedeman R, Blankenstein AH, van der Feltz-Cornelis CM, et al. Consultation letters for medically unexplained physical symptoms in primary care. Cochrane Database Syst Rev. 2010;12:CD006524.

143

6

7 Negotiating explanations: a qualitative analysis of doctor-patient communication in a general practice clinic for patients with medically unexplained physical symptoms Madelon den Boeft DaniĂŤlle Huisman LaKrista Morton Peter L. Lucassen Johannes C. van der Wouden Marjan J. Westerman HenriĂŤtte E. van der Horst Christopher D. Burton

Submitted

Chapter 7

ABSTRACT Background Patients with persistent medically unexplained physical symptoms (MUPS) seek explanations for their symptoms, but often find general practitioners (GPs) unable to deliver these. Different methods of explaining MUPS to patients have been proposed for use by GPs. Little is known about how doctor-patient communication evolves around these explanations.

Objective We aimed to examine the dialogue between GPs and MUPS patients related to explanations for symptoms and categorised dialogue types and dialogue outcomes.

Methods We analysed transcripts of 112 audio-recorded consultations (39 patients, 5 GPs) from two studies of the Symptoms Clinic Intervention, a consultation intervention for patients with MUPS in primary care. We used constant comparative analysis to code and classify dialogue types and outcomes in relation to the explanations.

Results We extracted 115 explanation sequences. We identified four dialogue types, which differed in the extent to which the GP and/or patient controlled the dialogue. We categorised eight outcomes of the explanation sequences, ranging from acceptance to rejection by the patient. The commonest outcome was holding (in which the conversation was suspended in an unresolved state) followed by acceptance. Few explanations were directly rejected. Explanations that were negotiated or co-created by patient and GP were most likely to be accepted.

Conclusion We developed a classification of dialogue types and outcomes in relation to explanations offered by GPs for MUPS patients. While it requires further validation, it pro-

146

Negotiating explanations for MUPS

vides a framework of dialogue types and outcomes, which can be used for teaching, evaluation of practice, and research.

BACKGROUND Patients with persistent physical symptoms may experience limitations due to their symptoms, have an impaired quality of life and incur substantial health care costs (1,2). They seek an explanation for their symptoms (3,4), but in the absence of disease they often find doctors unwilling or unable to provide comprehensible explanations (5). When explanations for medically unexplained physical symptoms (MUPS) are provided, they often follow a psychosocial model of causality, for example in early versions of the reattribution approach where symptoms are taken to represent indicators of psychological distress (6). Such explanations are not fully compatible with current models of symptom generation or persistence (7,8) and are commonly rejected by patients (9,10). Rational explanations have recently been proposed, which attempt to make sense of symptoms in a way that is plausible to patients and doctors, blame free, leads to therapeutic action, addresses causation, and ideally is co-created (11).

Relatively few studies have examined how explanations for MUPS are delivered.

Burbaum et al, studying reattribution in extended consultations, concluded that patients often interpret psychosocial attributions of the therapist as threats to their self-identity (9). Peters et al, studying reattribution in routine primary care consultations, identified several barriers, such as patientsâ&#x20AC;&#x2122; unwillingness to discuss (as opposed to acknowledge) the emotional aspects of their problems (10). Furthermore, Aiarzaguena et al examined how GPs presented an explanation in terms of a hormonal imbalance as a means to initiate a discussion on psychosocial aspects, as well as patientsâ&#x20AC;&#x2122; responses to these explanations (12). Patient responses were classified according to the amount of work patients carried out to verbalize a response. One of their main findings was that all responses were either wait-and-see or positive and that resistance was rare. They concluded that symptom explanations paved the way for a psychosocial exploration.

147

7

Chapter 7

The Symptoms Clinic Intervention (SCI) was developed as a consultation-based intervention for patients with persistent MUPS (13), which emphasizes the role of explanations in making sense of symptoms, thus moving forward to the therapeutic phase.

The SCI is designed for patients with moderately severe symptoms and comprises a structured set of four consultations by a trained GP. While substantially longer than conventional consultations, the SCI has a shorter duration and is less psychologically oriented than cognitive behavioural therapy. It seeks to provide patients with: Recognition that their symptoms are legitimate and understood, Explanation of symptoms in terms of biological and psychological processes, Actions to manage or control symptoms and Learning about what they find effective (14). The SCI was developed and preliminary tested in two studies (13,14) in which all consultations were audiorecorded and transcribed for analysis. To examine how symptom explanations were negotiated between GPs and patients with MUPS, we carried out detailed analysis of the dialogue structure of symptom explanations.

METHODS Study Design We conducted a qualitative study using audio-recorded consultations from two earlier studies of the SCI (Multiple Symptoms Studies One and Two). Detailed information is described elsewhere (13,14). Here we provide a brief summary.

Multiple Symptoms Study 1 (MSS1) This was a pilot randomised controlled trial (RCT) of the SCI (13). It was carried out by the main developer of the intervention among patients with moderate MUPS. The pilot was conducted in Edinburgh, Scotland between August 2009 and May 2010. Patients (n=32) were systematically identified and randomised to attend the intervention group (Symptom Clinic) or to usual care. Consultation recordings were available from 16 patients from the intervention group who attended one to four SCI consultations. From this study we retrieved 43 consultations.

148

Negotiating explanations for MUPS

Multiple Symptoms Study 2 (MSS2) This was an observational non-randomised study (14) in which four GPs who had received two days of structured training on the SCI delivered the intervention to five or six patients in their own practice who met the entry criteria used in the first study.

The study took place between August 2014 and June 2015. From MSS2 we retrieved 69 consultations held with 23 patients attending for one to four SCI consultations.

Participating patients All participating patients were adults (age ≥18) with inclusion criteria designed to represent MUPS of moderate intensity. Inclusion criteria were 1) at least two referrals to specialists in the preceding three years, 2) presence of at least one functional syndrome/symptom code in their medical records, 3) current moderate multiple physical symptoms, defined as a score ≥10 on the Patient Health Questionnaire-15 (PHQ15) (15), and 4) GP’s judgement that symptoms were unlikely to be explained by an underlying disease. Exclusion criteria were (a) symptoms of such severity that individuals were not able to leave the house independently and/or (b) active involvement in a programme of rehabilitation or psychotherapy.

Explanations Explanations for symptoms are central to the SCI and seek to make sense of symptoms (11) in ways which emphasise their “bodily nature and cultural meaning” to patients (16). GPs in the SCI constructed, provided and negotiated explanations, which were built around existing models and which tried to avoid psychosocial causality (7,8). A detailed taxonomy of the explanations themselves has been completed separately (Burton, personal communication) and is not considered further here.

Method of analysis All audio-recorded consultations were anonymously transcribed verbatim and entered into the qualitative software package Atlas.ti, Version 7. First, after reading all transcripts, two researchers (MB, DH) independently extracted all sequences in which symptom explanations were given from each consultation. We defined the beginning of an explanation sequence as the point where the GP started to give an explanation.

149

7

Chapter 7

We defined the end of an explanation sequence when the dialogue about that explanation reached a conclusion (agreement, disagreement or something less clear-cut) or terminated for another reason (e.g. change of topic). Contradictions between the researchers were discussed with a third researcher (CB). We restricted analysis of explanation sequences to those that included speech by both the GP and patient and comprised a minimum of four utterances (i.e. statements separated by silence or a change of speaker). When consultations contained several explanation sequences each sequence was extracted and analysed separately.

Second, two researchers (MB, DH) independently read the explanation sequences in detail while developing line-by-line codes, with a focus on the dialogue structure and the content of the explanation sequences. We used a constant comparative approach to identify common thematic features within the codes of the explanation sequences. Following this we identified categories for the explanations at two levels: first a categorisation of the way in which the GP and patient discuss an explanation, which we refer to as dialogue type, and second a categorisation of the outcome of each explanation sequence.

In classifying the explanation sequences into dialogue types, we considered speech as an active dialogue between participants rather than just a means of conveying information from the GP to the patient. Thus, we focused less on the explanation itself and more on the process and outcomes of the interaction around the explanation, and the negotiation of meaning that this entailed. For this, we drew on the idea from Bhaktin of dialogic space (17), meaning the space in which communication around possible explanations occurs. We used this idea of space in two ways. First we considered occupancy or ownership of the dialogic space by patient or GP; this was represented by both the time spent speaking and the apparent control over the content which the patient and doctor exhibited. Second, within this space, we considered the dialogue as either moving towards a common, shared account (“centripetal”) or away from this, towards multiplicity of accounts and disagreement (“centrifugal”) (17). We used these concepts of ownership of the dialogic space and the centripetal or centrifugal direction of the dialogue to differentiate explanation types. We applied them to each whole explanation sequence rather than to indi-

150

Negotiating explanations for MUPS

vidual parts; where sequences included subsections suggestive of different types we selected the most prominent one.

We conducted the analysis iteratively with meetings to discuss coding and evolving classifications between all members of the research team actively involved in the analysis. Once the classification had been developed, two researchers (MB, CB) recoded all explanation sequences separately, with differences in coding resolved by discussion with a third coder (LM).

Ethical approval All participating GPs and patients provided written informed consent. MSS1 was approved by the Lothian Research Ethics Committee (reference 09/S1102/34) and MSS2 by the North of Scotland Research Ethics Committee (reference 14/NS/1014).

Both studies were performed in accordance with the ethical standards of the Declaration of Helsinki.

RESULTS Patient characteristics and elements of analysis The patients participating in MSS1 (n=16) had a mean age of 50 years. Nine were female. The mean PHQ-15 score was 14. The patients participating in MSS2 (n=23) had a mean age of 51 years. Twenty-one were female. The mean PHQ-15 score was 15. We extracted 115 explanation sequences suitable for analysis (61 from MSS1 and 54 from MSS2).

Dialogue type Our final categorisation included four dialogue types: lecture, storytelling, contest, and deliberation. These are described below, with reference to their relationships to dialogic space. Examples are provided in Table 1.

Lecture: The lecture was the simplest dialogue type; indeed it was almost a monologue, as the GP talked in blocks of speech with minimal input from the patient.

The GP occupied the majority of the dialogic space in terms of amount of speech and number of words, but also controlled the content. The GP made little effort to

151

7

Chapter 7

customise the explanation to the patient. There was too little interaction to describe the direction of dialogue as centripetal or centrifugal.

Storytelling: Storytelling was, like the lecture, primarily GP-led. However, compared to the lecture it proceeded in a more interactive way and was more often customised to the patient. The interaction and customisation included (a) inclusion of personally relevant details from the patientâ&#x20AC;&#x2122;s account in the explanation (b) checking back by the GP, thereby ensuring that the patient followed the explanation and (c) more informal language than used in the lecture dialogue type, with popular and sometimes slang terms. While the GP tended to keep control of the direction of the dialogic space, and there was a centripetal drive towards the GPâ&#x20AC;&#x2122;s own interpretation, the patientâ&#x20AC;&#x2122;s responses were more active and the patient appeared to be a willing and interested participant in the dialogic space.

Contest: Contest, on the other hand, represented a struggle for control of the dialogic space. In contrast to the lecture, there was dialogue but it was largely centrifugal. GP and patient both contributed ideas and strove to occupy the dialogic space, putting forward ideas that mattered to them but often were incompatible with the ideas of their interlocutor. For instance the doctor would persevere with the explanation while the patient responded with statements, which emphasised (or protected) their moral status as a legitimate patient (9). Utterances included explicit counter statements and blocking techniques.

Deliberation: Deliberation (18,19) involved greater engagement between GP and patient with both contributing ideas. There was evidence of centripetal forces within the dialogue as both parties worked to find a mutually acceptable explanation.

The most commonly occurring dialogue types were storytelling, which occurred 46 times (40%) and deliberation, which occurred 44 times (38%). Contests occurred 18 times (16%) and lectures were least common with 7 occurrences (6%). The proportion of different dialogue types was similar between the two studies with only contests showing a difference of 5% or more (18% in the first; 13% in the second).

Dialogue outcomes Our final classification of explanation outcomes took a branching structure. We defined outcomes as the state of the dialogue at the end of the explanation sequence.

152

Negotiating explanations for MUPS

Some explanation sequences appeared complete (the dialogue had reached an obvious end, and moved on to a new topic) and had a clear conclusion (such as agreement or rejection). Others also appeared complete, but had no clear conclusion (such as one party electing to “hold” an idea for now while moving the conversation on). Finally some sequences appeared incomplete (and the topic of conversation shifted without completing the current explanation sequence). The classification included eight outcomes, which are described below and shown graphically in Figure 1. Examples of some of the outcomes are also included in Table 1.

Table 1 Examples of dialogue types and final outcomes Sequence 1: dialogue type “lecture” and outcome “passive receipt” GP One of the things that does seem to happen with pain sources is that often after, you seem to get a pattern where repeated pain, kind of, sensitises your pain management system in the, kind of, auto pilot bit of your brain. And sometimes that results then in, kind of, spilling over so that you can end up with one area that is the, kind of, trigger, but a lot of other areas that, kind of, lock down, shut down in relation to that trigger.

So that, and, you know, listening to you describing that pain it sounds as if things start, sinuses whatever, but by the time everything else, kind of, goes, oh god, this is...?

PT The GP did mention that to me.

GP It’s almost like you tighten up, you tighten up here and tighten up there, and you tighten up at the back of the neck. And that, kind of, secondary pain can sometimes just get things running, that, you know, you need, not an awful lot of trigger because those sinuses are sensitive enough without being messed around with. But your brain, kind of, doesn’t like tampering with those, and the reaction is just then progressively other stuff to lock down.

I always say, you know, nature does this, kind of, tortoise thing, my head’s sore, pull it in, scrunch up your neck, tighten everything up there, start this, kind of, secondary thing so that you end up with what should have been a pain that went being a pain that starts something and then is self-perpetuating. And again, one of the simple ways of looking at managing that, of this appointment is to maybe look at ways of, kind of, separating the trigger from the cascade of things that happen downstream, and that might be worth, kind of, concentrating on.

Again it’s about dealing with the symptoms rather than the sinuses trigger, you know, if you say that you kind of accept that you’re going to have trouble with those from time to time, but if you can stop that spilling over into everything else, that’s containment and adaptation and stuff, yes?

PT Yes. MSS1 Explanation 403.1.2

153

7

Chapter 7

Sequence 2: dialogue type “story telling” and outcome “hold” GP So that’s obviously reassuring. So what I’m thinking is that, you know, this represents a sort of chronic pain syndrome that you’re suffering from.

PT Okay.

GP So normally we experience pain when we damage our body, when we injure our body and what happens is that if you imagine you injured your knee, the pains in the knee would send messages initially up to the spinal column and then on up to the brain to say to the brain that the knee has been injured, but what can happen in some people is that the signals from the nerves become amplified, they become louder if you like in the brain, so that things that shouldn’t be uncomfortable become painful.

PT Okay.

GP And there are some theories as to why this happens. We think that some people have a genetic likelihood to experience this chronic pain, sometimes it follows some sort of trauma or physical insult and you mentioned that you thought a lot of this had started after you had your bowel operation which was obviously very traumatic the way it happened and the way it affected you for a number of years, so that may have been...

PT A trigger.

GP ...a trigger for this. The thought is that it’s due to chemical changes in the brain and as a result of those chemical changes, as I mentioned, the messages from the pain nerves are turned up, the volume is much louder than it should be, and there are a number of things that can help to turn it back down again. So one of the things that can help is Amitriptyline and obviously you’re on that at night and that’s one of the medications that can help make these chemical changes in the brain to try and turn down the volume and try and help with your pain. Normal painkillers like Paracetamol and Ibuprofen, things like that, they often are not very helpful in chronic pain. They’re helpful in acute pain, you know, but if you have this pain that’s nagging and there day after day, they sometimes don’t help with that sort of pain.

PT Right. ...

GP What are your thoughts about the suggestion that this may be a chronic pain syndrome?

PT It’s a possibility. I mean, the way I look at it is, I mean, my body at the time of the... oh my brain... the what d’you call it, my intestines, I suppose my body went through a trauma and so anything’s possible. I’m willing to try anything just to start feeling normal again and start getting my memory back.

MSS2 Explanation 615.2.1

154

Negotiating explanations for MUPS

Sequence 3: dialogue type “contest” and outcome “self-affirmation” GP I mean, again if you are a bit tense or you are feeling a little bit anxious or this false alarm system’s kicked in, then you might be holding yourself a bit more tense and that might be contributing to the problems with the shoulder.

PT Yeah, cause my daughter was saying this morning, she said this shoulder blade was sticking way out and I can’t get it to go down, it just won’t!

GP No, and you’ve seen the physio haven’t you?

PT Yeah, I mean, I haven’t been visiting her for quite a while and I did get exercises, it’s just when I get feeling so rotten it’s very difficult to exercise, to make myself exercise when I feel oh [laugh]!

GP Yes, and what we know that these chemicals do in the brain, along with having an exaggerated response to these signals from the body, is that they can also contribute to fatigue and to memory problems and to low mood, but then if you’re in pain every day your mood’s likely to be low anyway.

PT Yeah, and I wouldn’t necessarily say I am really low, when I feel sick and horrible I think ‘oh god’ you know, that does bring me down because you think ‘oh I’m just not feeling well’, I want to feel well and is it just... is there not an end, you know, I just want to think I’ll come out this, you know, or is it just wear and tear and it’s set everything off and I over respond to it, I don’t know?

GP I mean, if it is wear and tear then there’s a very good chance that it will settle down because what we think now is rather than wear and tear, is wear and repair. So if people do damage a joint what happens is that when new bone is laid down, it’s laid down in a slightly unusual way, it can cause some pain and discomfort, but then when the joint repairs itself and that new bone is reabsorbed, things can settle down.

So sometimes people can have a flare up of pain in a joint for two or three months maybe and then it can settle down again.

PT Okay, cause my fingers have been a bit numb, just those two. But I went out in the garden this morning cause I thought ‘well I’m just not going to sit here and feel sorry for myself!’.

MSS2 Explanation 602.2.1

155

7

Chapter 7

Sequence 4: dialogue type “deliberation” and outcome “acceptance” GP I’m not saying it’s imagined, what I’m saying is that I think there is a physical reasons why you’re feeling these aches and pains, that it’s not imagined and that, you know, it’s due to these chemicals in the brain that are making your brain more sensitive to messages it’s receiving. So I’m not saying you’re a hypochondriac or anything like that...

PT I think it’s because of my family and, you know, I’m always the... me and my other sister, we’re the normal ones!

GP Yeah but this can be normal, it’s very common and I think it explains why you are sore In lots of different places and why you’re getting some tummy symptoms and some nausea, You know, because your body is picking up the feedback in a more sensitive way.

PT Yeah. I’ve often had the nausea before any pain because it was the nausea that started before, so is that just still the same thing?

GP I think it probably is the same thing that, you know, if we have acid in the tummy sometimes you’re not aware of it, but if you have this amplification, this increased volume in your brain of the messages, then you might be picking up changes in the stomach more acutely than somebody else. included eight outcomes, which are described below and shown graphically in Figure 1.

PT Yeah, okay.

Yeah that makes sense [laugh].

MSS2 Explanation 602.3.1 Examples of some of the outcomes are also included in Table 1.

Figure ofdialogue dialogueoutcomes outcomes Figure 1.

1. Classification Classification of

Explanation sequence complete, definite conclusion

156

There were two outcome categories representing a definite conclusion: acceptance and

rejection. We categorised an explanatory sequence as ending with acceptance if the patient

either (a) explicitly acknowledged the explanation as helpful or (b) expressed the possibility of adopting the ideas it contained (e.g. hypothetically describing future actions, which would be

Negotiating explanations for MUPS

Explanation sequence complete, definite conclusion There were two outcome categories representing a definite conclusion: acceptance and rejection. We categorised an explanatory sequence as ending with acceptance if the patient either (a) explicitly acknowledged the explanation as helpful or (b) expressed the possibility of adopting the ideas it contained (e.g. hypothetically describing future actions, which would be conditional on the explanation being true).

We categorised an explanatory sequence as ending with rejection if the patient directly and effectively countered the GP’s presented explanation or assertion either by (a) explicit disagreement (e.g. “it is not like that”) or (b) providing an unarguable counterfactual (e.g. “this is not what happened, actually X happened”).

Explanation sequence complete, no definite conclusion We identified three outcome categories for complete sequences with no definite conclusion: self-affirmation, mis-affirmation and holding. In self-affirmation, the patient responded to the explanation sequence with a positively framed statement, but this statement sidestepped the explanation and thus patients maintained their self-integrity (9). In mis-affirmation the patient also concluded the sequence in a positive, affirming way, but with a statement unrelated to the presented explanation. Finally holding represented an outcome in which the dialogue was suspended in an unresolved state. This implied the possibility of accepting the explanation in the future, but that was not yet settled.

Explanation sequence not complete, with engagement We identified two final responses to explanation sequences, which suggested that, while GP and patient were engaged in the dialogue, the sequence was terminated before a conclusion had been reached. In jumping either patient or GP moved abruptly to a new and unrelated topic, effectively restarting the dialogue. In cueing, the patient or GP moved abruptly to a new topic but which was relevant and related to the presented explanation.

157

7

Chapter 7

Explanation sequence without engagement Some explanation sequences involved little patient participation. The patient mostly (or even only) used short affirmations or continuators during the dialogue (e.g. “right”, “hmm”, “okay”).

The most commonly occurring outcomes were holding with 46 instances (40%); accepting with 27 instances (23%); and self-affirmation with 19 (17%) instances. Misaffirmation occurred only once (<1%). The associations between the dialogue and conclusion types are summarized in table 1. Both lecture and storytelling had broadly similar patterns of outcomes, with holding being the most common (3/7 and 22/46 respectively). Deliberation types were more likely than others to end in acceptance (17/44). Frequencies are summarized in table 2.

Table 2 Association of dialogue type and outcome Dialogue type

Lecture

Storytelling

Contest

Deliberation

Outcome

N

%

N

%

N

%

N

%

Total

Accept

1

14

9

20

0

0

17

39

27

Cueing

0

0

2

4

0

0

2

5

4

Hold

3

43

22

48

6

34

15

34

46

Jump

0

0

2

4

2

11

1

2

5

Mis-affirmation

0

0

1

4

0

0

0

0

1

Passive Receipt

3

43

4

9

0

0

0

0

7

Reject

0

0

0

0

4

22

2

4

6

Self-affirmation

0

0

6

13

6

33

7

16

19

Total

7

100

46

100

18

100

44

100

115

DISCUSSION Main findings This study provides a new way of describing and classifying features of the dialogue focussing on the explanations provided by GPs to patients with persistent MUPS in an enhanced care setting. The classification includes a range of different dialogue types and outcomes that describes all episodes of explanation. While most dialogues were

158

Negotiating explanations for MUPS

GP-led, deliberation, with its potential for co-creation of interpretation, occurred within 40% of explanations. Patients explicitly accepted approximately one quarter of explanations and rejected or blocked a similar proportion.

Strengths and limitations This study has several methodological strengths. We systematically analysed a large sample of extracted sequences centring on explanations from structured, long, consultations with patients with persistent MUPS who were systematically identified.

The analysis drew on iterative and reliable coding. Dual independent coding, after agreement of the classification, reduced the risk of bias.

However, we should mention a number of limitations: the nature of the intervention meant that we had multiple explanations but that these came from a small number of GPs discussed with a relatively small number of patients. Another limitation of our analysis is that it takes place at the level of explanation sequences rather than at the level of the whole consultation or even episodes of care comprising several consultations. However, by focusing on clearly defined explanation sequences we were able to examine key elements of the consultation in detail.

Comparison with the literature A recent Cochrane review raised the possibility that more moderate intensive interventions might have added value in the management of MUPS (20) and the SCI was developed as such. It distinguishes itself from other moderately intensive interventions by its central element of the symptom explanations. We are aware of only one study that used symptom explanations as a central intervention for MUPS management (12). However, this study had a smaller sample of 11 patients, had a different methodological approach to analyse patientsâ&#x20AC;&#x2122; responses to symptom explanations and limited the explanations to emotional distress and stress hormones, while the SCI does not require a link with emotional distress.

Burbaum et al focused on responses to reattribution and found three recurring patterns where patients subtly refuted, dropped or undermined the attribution in their reply (9). Even though they used a different sample of patients and a different setting and methodology, we found some similar responses, namely self-affirmation, mis-

159

7

Chapter 7

affirmation and rejection, although these occurred relatively infrequently. Salmon et al found that some consultations involve a certain degree of contest between the patientâ&#x20AC;&#x2122;s authority resting on their knowledge of their own symptoms and the GPâ&#x20AC;&#x2122;s authority based on their professional knowledge (21). We also observed this contest several times in our analysis, where we found that the GP and the patient were occupying a separate dialogic space and did not find common ground.

Implications for further research We have developed a classification of dialogue in response to explanations, which can be used in further studies. Further research and validation is needed to examine this classification of explanation dialogue types and outcomes in larger datasets beyond the SCI and in shorter consultations.

The relatively high rates of accepting and holding responses is encouraging and indicates that the explanations used in the SCI seem to be largely acceptable to patients in an area of practice where patients commonly feel that they are neither understood nor supported (3,4) and where other modes of explanation frequently fail (5, 9,10). Future studies should examine the associations between the elements of explanation dialogue described here and subsequent outcomes of treatment.

CONCLUSION We developed a classification of dialogue types and outcomes in relation to explanations MUPS. While it requires validation in a larger observational study, it provides a framework of dialogue types and patient responses, which can be used for teaching, evaluation of practice, and research.

ACKNOWLEDGEMENTS The authors would like to thank all the participating GPs and patients in this study.

160

Negotiating explanations for MUPS

REFERENCES 1. McGorm K, Burton C, Weller D, et al. Patients repeatedly referred to secondary care with symptoms unexplained by organic disease: prevalence, characteristics and referral pattern. Fam Pract. 2010;27:479–86.

2. Verhaak PFM, Meijer SA, Visser AP, et al. Persistent presentation of medically unexplained symptoms in general practice. Fam Pract. 2006;23:414–20.

3. Nettleton S. “I just want permission to be ill”: Towards a sociology of medically unexplained symptoms. SocSciMed. 2006;62:1167–78.

4. Giroldi E, Veldhuijzen W, Mannaerts A, et al. “Doctor, please tell me it’s nothing serious”: An exploration of patients’ worrying and reassuring cognitions using stimulated recall interview. BMC Fam Pract. 2014;15.

5. olde Hartman TC, Hassink-Franke LJ, Lucassen PL, et al. Explanation and relations.

How do general practitioners deal with patients with persistent medically unexplained symptoms: a focus group study. BMC Fam Pract. 2009;10:68.

6. Gask L, Dowrick C, Salmon P, et al. Reattribution reconsidered: narrative review and reflectionson an educational intervention for medically unexplained symptoms in primary care settings. Psychosom Res. 2011;71:325–34.

7. Deary V, Chalder T, Sharpe M. The cognitive behavioural model of medically unexplained symptoms: A theoretical and empirical review. Clin Psychol Rev. 2007;27:781–97.

8. Rief W, Broadbent E. Explaining medically unexplained symptoms - models and mechanisms. Clin Psychol Rev. 2007;27:821–41.

9. Burbaum C, Stresing A-M, Fritzsche K, et al. Medically unexplained symptoms as a threat to patients’ identity?: A conversation analysis of patients’ reactions to psychosomatic attributions. Patient Educ Couns. 2010;79:207–17.

10. Peters S, Rogers A, Salmon P, et al. What do patients choose to tell their doctors? Qualitative analysis of potential barriers to reattributing medically unexplained symptoms. J Gen Intern Med. 2009;24:443–9.

11. Burton C, Lucassen P, Aamland A, et al. Explaining symptoms after negative tests: towards a rational explanation. J R Soc Med. 2015;108:84–8.

12. Aiarzaguena JM, Gaminde I, Clemente I, et al. Explaining medically unexplained symptoms: somatizing patients’ responses in primary care. Patient Educ Couns.

2013;93:63–72.

13. Burton C, Weller D, Marsden W, et al. A primary care Symptoms Clinic for patients with medically unexplained symptoms: pilot randomised trial. BMJ Open. 2012;2:000513 14. Morton L, Elliott A, Thomas R, et al. Developmental study of treatment fidelity, safety and acceptability of a Symptoms Clinic intervention delivered by General Practitioners to patients with multiple medically unexplained symptoms. J Psychosom Res. 2016;84:37–43.

15. Kroenke K, Spitzer RL, Williams JBW. The PHQ-15: Validity of a new measure for evaluating the severity of somatic symptoms. Psychosom Med. 2002;64:258–66.

16. Kirmayer LJ, Groleau D, Looper KJ, et al. Explaining medically unexplained symptoms.

Can J Psychiatry Rev Can Psychiatr. 2004;49:663–72.

161

7

Chapter 7

17. Shotter J. Moments of common reference in dialogic communication: A basis for unconfused collaboration in unique contexts. Int J Collab Pract. 2009;1:31–9.

18. Emanuel E, Emanuel L. Four models of the physician-patient relationship. JAMA. 267:2221–6.

19. Elwyn G, Lloyd A, May C, et al. Collaborative deliberation: a model for patient care.

Patient Educ Couns. 2014;97:158–64.

20. van Dessel N, den Boeft M, van der Wouden JC, et al. Non-pharmacological interventions for somatoform disorders and medically unexplained physical symptoms in adults.

Cochrane Database Syst Rev 2014;11:CD011142.

21. Salmon P. Conflict, collusion or collaboration in consultations about medically unexplained symptoms: the need for a curriculum of medical explanation. Patient Educ Couns. 2007;67:246–54.

162

8 The association between medically unexplained physical symptoms and health care use over two years and the influence of depressive and anxiety disorders and personality traits: a longitudinal study Madelon den Boeft Jos WR. Twisk Berend Terluin Brenda WJH. Penninx Harm WJ. van Marwijk Mattijs E. Numans Johannes C. van der Wouden Henriette E. van der Horst

BMC Health Services Research. 2016;16:100

Chapter 8

ABSTRACT Background Medically unexplained physical symptoms (MUPS) are highly prevalent and are associated with frequent health care use (HCU). MUPS frequently co-occur with psychiatric disorders. With this study we examined the longitudinal association between MUPS and HCU over two years and the influence of depressive and anxiety disorders and personality traits on this association.

Methods We analysed follow-up data from 2045 to 2981 participants from the Netherlands Study of Depression and Anxiety (NESDA), a multisite cohort study. The study population included participants with a current depressive and/or anxiety disorder, participants with a lifetime risk and/or subthreshold symptoms for depressive and/ or anxiety disorders and healthy controls. HCU, measured with the Trimbos and iMTA questionnaire on Costs associated with Psychiatric illness (TIC-P), was operationalized as the number of used medical services and the number of associated contacts.

MUPS were measured with the Four Dimensional Symptoms Questionnaire, depressive and anxiety disorders with the Composite International Diagnostic Interview and personality traits with the NEO Five-Factory Inventory. Measurements were taken at baseline, one and two years follow-up. We used generalized estimating equations (GEE), using HCU at all three measurements as (multivariate) outcome. GEE also takes into account the dependency of observations within participants.

Results MUPS were positively associated with HCU over two years (medical services: RR 1.020, 95% CI 1.017-1.022; contacts: RR 1.037, 95% CI 1.030-1.044). Neuroticism and depression had the strongest influence on the associations. After adjustment for these factors, the associations between MUPS and HCU weakened, but remained significant (services: RR 1.011, 95% CI 1.008-1.014; contacts: RR 1.023, 95% CI 1.0151.032).

166

MUPS and health care use over two years

Conclusions Our results show that MUPS were positively associated with HCU over two years, even after adjusting for depressive and anxiety disorders and personality traits.

BACKGROUND Medically unexplained physical symptoms (MUPS), physical symptoms that cannot be explained or not sufficiently explained by an underlying medical condition after adequate examination, are highly prevalent in all health care settings [1–5]. MUPS represent a broad spectrum of symptoms in varying degrees of severity, ranging from acute, mild MUPS to severe and chronic MUPS [6,7]. It is known that patients with MUPS have a high health care use (HCU) leading to high costs [8-10]. Therefore, MUPS put a burden not only on patients and physicians, but also on society in a time when health care costs are steadily rising.

This high HCU is regularly attributed to patients pressurizing their general practitioner (GP) for a somatic treatment for their symptoms. However, several studies suggest that most patients do not request somatic interventions but want support and acknowledgement of the reality of their symptoms, but instead receive interventions initiated by the GP [11,12].

Several studies showed that patients with MUPS use disproportionally large amounts of mostly somatic health care services and not particularly mental health care services [13, 14]. Barsky et al for instance found that primary care patients with MUPS had approximately twice the outpatient and inpatient HCU and twice the annual medical costs compared to non-MUPS patients [8]. Studies that have been performed on this topic used different methodological approaches. Many of them used a retrospective design [8,14,15], only included patients from primary care [3,8,13,16] or only included patients with severe MUPS [3,17]. As far as we know, only one recent study used a prospective design with extended follow-up and included a large sample of participants from the general population [18].

It is also known that MUPS frequently co-occur with depressive and/or anxiety disorders [19–21]. This is of great clinical relevance as this ‘cosyndromality’ leads to more disability, impairment and high HCU [8,18,22]. The same applies to some personality

167

8

Chapter 8

traits such as neuroticism. Although in literature most research has been performed on personality disorders that are associated with MUPS [23], Noyes et al suggested that certain personality traits co-occurring with MUPS could lead to increased care seeking behaviour [24]. These findings raise the question what the independent association of MUPS with HCU is and to what extent personality traits and depressive and anxiety disorders add to this association.

For our study, we used data from the Netherlands Study of Depression and Anxiety, a large naturalistic multisite longitudinal cohort. Data on MUPS, HCU, depressive and anxiety disorders and personality traits were all collected over time from a large sample of participants from several health care settings. Therefore, this cohort is ideally suited to investigate the following research questions: What is the association between MUPS and HCU over two years? And, to what extent is the association between MUPS and HCU influenced by depressive and/or anxiety disorders and specific personality traits?

METHODS Design, setting and study sample The Netherlands Study of Depression and Anxiety (NESDA) aims to describe the long-term course and consequences of depressive and anxiety disorders and to examine its predictors. A detailed description of its rationale and design has been published elsewhere [25].

In summary, the study sample consisted of 2981 participants (age 18-65) with current depressive and/or anxiety disorders, participants with a lifetime risk or subthreshold depressive and/or anxiety symptoms and healthy controls. Recruitment took place across primary care practices (n=1610), outpatient secondary mental health care institutions (n=807) and the general population (n=564). Exclusion criteria were not being fluent in the Dutch language and a primary diagnosis of a psychotic, obsessive compulsive, bipolar or severe substance abuse disorder. Baseline data were collected between 2004 and 2007. Assessments, including written questionnaires and interviews, were repeated after one, two, four and six years. Non-response among participants was not significantly related to mental health status, but slightly higher among younger and male respondents. The research protocol was approved centrally by the ethical review board of VU University medical center. Subsequently it was approved by the local ethical review

168

MUPS and health care use over two years

boards of Leiden University Medical Center and University Medical Center Groningen.

The study was performed in accordance with the ethical standards of the Declaration of Helsinki. All participants provided written informed consent.

For the present study, we used data of all participants who completed the questionnaire used for our study, and used the measurements at baseline (T0), one (T1) and two years (T2) of follow-up. Baseline measurements were obtained from 2981 participants. At T1 and T2, 2045 (68.6%) and 2395 (80.3%) participants had a followup assessment, respectively.

Health care use (HCU), the outcome HCU was measured with the Trimbos and iMTA questionnaire on costs associated with psychiatric illness (TIC-P) [26]. The TIC-P is a widely used, feasible and reliable questionnaire on health care consumption and productivity losses for patients with mental health disorders. For this study we focused on the first part of the TIC-P, consisting of dichotomous questions on relevant medical services, followed by a question on the consumption volume (number of contacts) in the past six months; e.g. ‘did you consult with a family physician? No / Yes, namely … times’. We counted the number of medical services used (range 0-14) and additionally categorized these into three subgroups: mental health care services (primary care psychologists, social workers/social psychiatric nurses, secondary mental health care institutions, centers for drugs or alcohol, self-help groups and private psychiatrists/psychotherapists); somatic health care services (family physicians, medical specialists and hospital admissions); and miscellaneous health care services (homecare, complementary alternative professionals, occupational health physicians, physiotherapists). Participants completed the TIC-P at T0, T1 and T2.

Medically unexplained physical symptoms (MUPS), the determinant MUPS were measured with the somatisation scale of the validated Four Dimensional Symptoms Questionnaire (4DSQ) [27]. The self-report 4DSQ has been developed to measure distress, depression, anxiety and somatisation as separate dimensions. The somatisation scale comprises 16 items including physical symptoms that often remain medically unexplained (e.g. dizziness and abdominal pain). The scale highly correlates

169

8

Chapter 8

with instruments used in other countries measuring MUPS; 0.82 in case of the SCL-90 [27,28] and 0.84 in case of the PHQ-15 [29,30]. In the present sample Cronbach’s alpha of the 4DSQ somatisation scale was 0.92, 0.89 and 0.97 at the three measurements, respectively. The items on the somatisation scale are scored on a 5-point Likert scale: “no”, “sometimes”, “regularly”, “often”, and “very often or constantly”. In order to arrive at scale scores, the responses were recoded as 0 for “no”, 1 for “sometimes” and 2 for “regularly”, “often” and “very often or constant” and summated, resulting in a score ranging from 0-32. Additionally, in order to facilitate clinical use and to overcome the fact that there is no linear relation between MUPS and HCU, we repeated the analyses with a dichotomized scale using 11 points as a cut-off score, since a score of 11 or higher is considered to indicate MUPS [27]. Participants completed the 4DSQ at T0, T1 and T2.

Depressive and anxiety disorders The presence of depressive and anxiety disorders was assessed with the validated Composite International Diagnostic Interview (CIDI, WHO 2.1) at T0 and T2. Trained research staff interviewed all participants. Depressive disorders included major depressive disorder and dysthymia. Anxiety disorders included generalized anxiety disorder, panic disorder with or without agoraphobia, social phobia and/or agoraphobia without panic disorder. We only took into account diagnoses established during the previous six months at both assessments.

Personality traits Personality traits were measured with the NEO Five Factor Inventory (NEO-FFI) at T0 and T2. The NEO-FFI measures the five most important personality domains in adults: neuroticism, extraversion, openness, agreeableness and conscientiousness.

Each domain is measured with 12 items, using a five-point Likert response format (sum score: range 12-60). More detailed information about the contents, validity and reliability of the NEO FFI has been published elsewhere [31-33].

Sociodemographic variables and chronic diseases Based on previous studies, we considered the following sociodemographic variables as possible confounders: gender, age, level of education, marital status and the num-

170

MUPS and health care use over two years

ber of chronic diseases [34, 35]. The level of education was derived from the standard classification of education from Statistics Netherlands [36] and categorized into three groups (basic, intermediate, high). Marital status divided participants in those being married/living with a partner and those living alone. Participants were asked if they had one or more diseases from the following chronic diseases categories: respiratory, cardiometabolic, musculoskeletal, digestive, neurological, endocrine and cancer. We only considered and summated the diseases if participants were currently treated with medication and/or under specialist control. All variables were assessed at T0.

Statistical analysis Descriptive statistics are presented as mean with standard deviation for normally distributed continuous data, median and inter-quartile range for skewed continuous variables and as numbers and percentages for dichotomous and categorical variables.

Generalized estimating equations (GEE) with an exchangeable correlation structure were used to assess the relationship between MUPS and HCU longitudinally (fig 1). We used GEE because it takes into account the dependency of repeated observations within the participants and because it is capable of analysing non-complete longitudinal data.

As the total number of consulted medical services showed a Poisson distribution, we used Poisson GEE analysis to assess its association with MUPS. For the total number of contacts, we used negative binomial GEE analysis because the Poisson distribution was skewed to the right (a Poisson distribution with overdispersion). The effect sizes of both the Poisson and the negative binomial GEE analyses are expressed as rate ratios (RRs).

This RR represents the association between MUPS and HCU on average over time and reflects both a within and between subjects interpretation [37]. Besides crude analyses, we adjusted the relationships for the sociodemographic variables, and additionally we examined the influence of depressive and anxiety disorders and personality traits on the association between MUPS and HCU. The influence is expressed as the percentage decrease in the regression coefficient as a result of including each separate variable.

As depressive and anxiety disorders and personality traits were only measured at T0 and T2, these last analyses were based on these two measurements only. We repeated the crude and adjusted analyses for each of the three medical resource subgroups.

Furthermore, as we performed our analyses in a population with an oversampling of

171

8

Chapter 8

analyses in a population with an oversampling of depressive and anxiety disorders, we analysed depressive and anxiety disorders, we analysed whether the effect between MUPS and whether themodified effect between MUPS and/or and HCU was modified byan depressive by HCU was by depressive disorders by adding interactionand/or term todisorders the adding an Finally, interaction term to Finally, we carried thelag same setin oforder analyses with a model. we carried outthe themodel. same set of analyses with aout time model

time model in order to assess if MUPS a certain inone time waslater related toWe HCU one year tolag assess if MUPS at a certain point in timeatwas relatedpoint to HCU year (fig 1).

used our analyses.in our analyses. later (figall1).observations We used allinobservations regression coefficients were considered statistically significant when All All regression coefficients were considered to to be be statistically significant when thethe p-value was p-value was below 0.05. All statistical analyses were performed in SPSS 20.0 for below 0.05. All statistical analyses were performed in SPSS 20.0 for Windows.

Windows.

Figure models Figure1.1.Analysis Analysis models GEE standard analysis T0 MUPS

T1 MUPS

T2 MUPS

T0 HCU

T1 HCU

T2 HCU

GEE time-lag analysis T0 MUPS

T1 MUPS T1 HCU

T2 HCU

T0: baseline; T1: one year follow-up; T2: two years follow-up. MUPS: medically unexplained

T0:physical baseline; T1: oneHCU: yearhealth follow-up; T2: two years follow-up. MUPS: medically unexplained symptoms. care use physical symptoms. HCU: health care use

RESULTS

RESULTS

Table 1 shows the descriptive information of all variables used in this study. At baseline,1 the mean was 42 years and 66%ofwere women. used The mean forAtMUPS Table shows theage descriptive information all variables in thisscore study. baseline, the was 10 at T0, 8.6 at T1 and 8.1 at T2. When using the clinical cut-off point of 11, 42% mean age was 42 years and 66% were women. The mean score for MUPS was 10 at T0, 8.6 at T1 had MUPS at T0, 27% at T1 and 25% at T2. and 8.1 at T2. When using the clinical cut-off point of 11, 42% had MUPS at T0, 27% at T1 and 25% at T2.

172

MUPS and health care use over two years

Table 1. Sample characteristics Baseline

One Year Follow-up

Two year Follow-up

Socio-demographics Females, number (%)

1979 (66.4)

Age in years, mean (SD)

41.9 (13.1)

Level of education, number (%)

Basic

199 (6.7)

Intermediate

1736 (58.2)

High

1046 (35.1)

Number of chronic diseases, mean (SD)

0.6 (0.9)

Married or with partner, number (%)

2066 (69.3)

MUPS (4DSQ somatisation scale) Total score (0-32), mean (SD)

10.0 (7.1)

8.6 (6.7)

8.1 (6.3)

Dichotomized: number with MUPS (â&#x2030;Ľ11) (%)

1237 (42.0)

806 (27.0)

741 (24.9)

Total number of medical services (0-12), mean (SD)

2.4 (1.5)

2.5 (1.6)

2.9 (1.8)

Mental health care services

0.6 (0.8)

0.5 (0.8)

0.6 (0.9)

Somatic health care services

1.2 (0.7)

1.4 (0.8)

1.4 (0.9)

Miscellaneous health care services

0.6 (0.8)

0.7 (0.8)

0.8 (0.9)

Total number of contacts with medical services, median (IQR)

7.0 (14.0)

9.0 (19.0)

12.0 (28.0)

Mental health care contacts

0.0 (5.0)

0.0 (5.0)

0.0 (8.0)

Somatic health care contacts

3.0 (4.0)

4.0 (4.0)

3.0 (5.0)

Miscellaneous health care contacts

0.0 (5.0)

0.0 (6.0)

2.0 (14.0)

Depressive disorders

1158 (38.8)

-

626 (21.0)

Anxiety disorders

1305 (43.8)

-

711 (23.9).

Neuroticism

36.3 (9.4)

-

33.5 (9.0)

Extraversion

36.9 (7.4)

-

37.8 (7.2)

Openness

38.2 (6.0)

-

36.8 (5.3)

Agreeableness

43.8 (5.3)

-

44.5 (5.2)

Conscientiousness

41.7 (6.5)

Health Care Use (TIC-P)

Depressive or anxiety disorders (CIDI), number (%)

8

Personality score (NEO-FFI; 12-60), mean (SD)

42.3 (6.2)

MUPS: medically unexplained physical symptoms. 4DSQ: four dimensional symptom questionnaire.

TIC-P: Trimbos and iMTA questionnaire on costs associated with psychiatric illness. IQR: interquartile range (25th to 75th percentile). CIDI: composite international diagnostic interview. NEO-FFI: NEO fivefactor inventory.

173

Chapter 8

The longitudinal association between MUPS and HCU MUPS were significantly associated with the total number of consulted medical professionals and the total number of associated contacts, respectively, on average over time in both the crude and adjusted GEE analyses (table 2). To illustrate the interpretation of the results: the estimated adjusted RR of 1.02 found for MUPS in relation to the total number of medical resources can be interpreted as follows: for every unit increase in the 4DSQ, a 2% increase in the number of medical services is observed, within and between participants. The small difference between the crude and adjusted analyses was mainly driven by the number of chronic diseases. MUPS defined with the dichotomized 4DSQ showed results in the same direction (services: RR 1.35; 95% CI 1.31-1.40; contacts: RR 1.41; 95% CI 1.28-1.55, not in table). Outcome defined as the three categories of medical services also showed comparable results (table 2). The strongest association was found for both the number of mental health care services used and the number of contacts with these services. Table 3

Table 2. MUPS and HCU over time: GEE standard analyses Crude RR (95% CI)

Adjusted RR (95% CI)

MUPS and total number of medical services

1.022 (1.020-1.024) *

1.020 (1.017-1.022) *

Somatic services

1.013 (1.011-1.015) *

1.009 (1.006-1.011) *

Mental health care services

1.035 (1.030-1.039) *

1.035 (1.030-1.040) *

Miscellaneous health care services

1.027 (1.023-1.031) *

1.026 (1.021-1.030) *

MUPS and total number of contacts

1.045 (1.040-1.052) *

1.037 (1.030-1.044) *

Somatic contacts

1.044 (1.040-1.048) *

1.031 (1.027-1.035) *

Mental health care contacts

1.046 (1.035-1.058) *

1.044 (1.031-1.057) *

Miscellaneous health care contacts

1.042 (1.034-1.050) *

1.029 (1.020-1.038) *

* All RRs including 95% CIs were significant with p-values below 0.001. The adjusted RRs were adjusted for the sociodemographic variables and chronic diseases. All measurements include T0, T1 and T2. Mental health care services: primary care psychologists, social workers/social psychiatric nurses, secondary mental health care institutions, centers for drugs or alcohol, self-help groups and private psychiatrists/psychotherapists. Somatic health care services: family physicians, medical specialists and hospital admissions. Miscellaneous health care services: home care, complementary professionals, occupational health physicians and physiotherapists.

174

MUPS and health care use over two years

Table 3. MUPS and HCU over time, adjusted for depression, anxiety and personality: GEE standard analyses Percentage MUPS and number MUPS and number of decrease in of contacts medical services regression (RR; 95% CI) (RR; 95%CI) coefficient Adjusted RR **

1.021 (1.019-1.024)*

Percentage decrease in regression coefficient

1.041 (1.034-1.048)*

Depressive disorders 1.015 (1.013-1.018)*

29

1.031 (1.024-1.039)*

24

Anxiety disorders

1.018 (1.015-1.020)*

14

1.034 (1.026-1.042)*

17

Neuroticism

1.013 (1.011-1.016)*

38

1.027 (1.018-1.036)*

34

Extraversion

1.018 (1.016-1.021)*

14

1.036 (1.028-1.044)*

12

Openness

1.021 (1.019-1.024)*

0

1.041 (1.034-1.048)*

0

Agreeableness

1.021 (1.019-1.024)*

0

1.039 (1.031-1.047)*

5

Conscientiousness

1.020 (1.017-1.022)*

0

1.037 (1.030-1.045)*

10

Neuroticism & 1.011 (1.008- 1.014)* Depressive disorders

48

1.023 (1.015-1.032)*

44

* All RRs were significant with p-values below 0.001. ** All analyses were based on measurements at T0 and T2, as depression, anxiety and personality were not measured at T1. Therefore, a new rate ratio was calculated, adjusted for sociodemographic variables and chronic diseases.

shows the influence of depressive and anxiety disorders and personality traits on the association of HCU with MUPS. For both HCU outcomes, neuroticism had the strongest influence, followed by depressive disorders. When taking neuroticism and depressive disorders together, the magnitude of the regression coefficient decreased by 48% (services) and 44% (contacts). Despite the contribution of these mental health characteristics, HCU remained significantly associated with MUPS. Adding anxiety disorders and other personality traits did not further affect the association. Also for the dichotomized 4DSQ score, neuroticism and depressive disorders together showed the strongest influence (decrease in regression coefficients of 56% and 59%, respectively) and HCU remained significantly associated with MUPS. When we examined whether the association between MUPS and HCU was modified by depressive and/or anxiety disorders, we found a significant inverse interaction effect (p<0.001),

175

8

Chapter 8

meaning that the association between MUPS and HCU (both services and contacts) was weaker for patients with depressive and anxiety disorders (data not shown).

MUPS related to HCU one year later Table 4 shows the results of the time-lag analyses. When comparing the results of the time-lag analyses with the standard analyses, we found comparable RRs for the number of medical services, but slightly higher RRs for the number of contacts in the time-lag analyses. For the dichotomized 4DSQ score, results were in the same direction (services: RR 1.19; 95% CI 1.15-1.22; contacts: RR 1.64; 95% CI 1.48-1.83).

For the influence of depressive and anxiety disorders and personality traits on the association between MUPS and HCU, we found the same pattern with neuroticism as the strongest influencing variable, followed by depressive disorder and again even stronger when taken together (data not shown). MUPS was still associated with HCU over a longer period of time, as reflected in the time lag analyses (data not shown).

Table 4. MUPS related to HCU one year later: GEE time-lag analyses Crude RR (95% CI)

Adjusted RR (95% CI)

MUPS and number of medical services

1.021 (1.018 - 1.023)*

1.018 (1.015 - 1.020)*

MUPS and number of contacts

1.060 (1.053 - 1.067)*

1.051 (1.043 - 1.058)*

*All RRs were significant with p-values below 0.001. The adjusted RRs were adjusted for the sociodemographic variables and chronic diseases. MUPS were measured at T0 and T1. HCU was measured at T1 and T2.

DISCUSSION In the present study we found a positive association between MUPS and HCU over two years taking into account all measurements, both for the number of medical services as well as the associated contacts. After adjusting for depressive and anxiety disorders and personality traits, the associations weakened, especially due to depressive disorders and neuroticism, but remained statistically significant.

176

MUPS and health care use over two years

Comparison with literature Our findings on the positive association between MUPS and HCU are in accordance with previous studies, irrespective of the differences in methodology [8,10,13,17,18, 22]. However, when we examined the three categories of medical resources, we found the strongest association between MUPS and mental HCU, in contrast to some other studies [8,13,18]. Fink et al concluded in their study among patients with somatoform disorders that these patients used more non-psychiatric health care facilities than patients without somatoform disorders [13]. Also Barsky et al found large amounts of medical, but not mental, health care use among their somatising patients [8]. Their findings support the assumption that patients with MUPS attribute their complaints to their physical symptoms, thereby seeking somatic health care instead of mental health care. The difference with our findings could be explained by the difference in setting (primary care only versus primary care, mental health care and general population).

We found that depressive disorders and neuroticism had the strongest influence on the association between MUPS and HCU over time. As far as we know, no research has been published on the comorbidity of MUPS with psychiatric disorders and personality traits with regard to HCU, especially not in a longitudinal design. De Waal et al found that somatoform disorders and depressive disorders were almost equally associated with HCU, but that the undifferentiated somatoform disorder had an independent effect after adjusting for psychiatric disorders [22]. Noyes et al found that MUPS were associated with specific personality traits as neuroticism and that this led to increased care seeking behaviour, which is in accordance with our findings [24].

However, in contrast with our findings, Carlier et al found no association between somatoform disorders and specific personality traits in their cross-sectional study [38]. As an explanation for this result they argued that their somatoform disorders patients were mostly highly educated and married, indicating a stable personal life.

Generalizability of the results One should realize that we examined the relation between MUPS and HCU in a sample with predominantly depressive and/or anxiety disorders. This may impede the interpretation and generalizability of the results. Therefore, we investigated

177

8

Chapter 8

possible effect modification between MUPS and depressive disorders and between MUPS and anxiety disorders. These analyses showed that within the population without these disorders the relationship between MUPS and HCU was stronger than in the population with these disorders. Based on those analyses, we believe that the observed relations may also hold for the general population.

Strengths and limitations A main strength of our study was that we used the NESDA cohort to answer our research questions. By using this cohort, longitudinal data over two years from a large sample of participants were available, recruited from both primary care, secondary mental health care and the general population. Also, we adjusted all analyses for chronic somatic diseases. Furthermore, we used structured diagnostic interviews and not only self-report questionnaires.

However, our findings should be interpreted in the light of several limitations. First, the NESDA cohort study used the 4DSQ somatisation scale to measure MUPS. As with all existing MUPS questionnaires, it lacks judgement of a clinician to verify that symptoms are really unexplained [39]. However, the 4DSQ highly correlates with the PHQ-15 and SCL-90, questionnaires that are widely used to measure MUPS [27], and may be considered as an adequate proxy measure for MUPS. Second, HCU was measured over the past six months, while MUPS were measured over the past week, leading to incongruence. However, we do not believe that the results were affected much by this. We found a correlation coefficient of 0.72 of MUPS over the two years (data not shown), indicating a quite stable pattern of MUPS over time. Also, our timelag analyses showed results that were similar to those of the standard analyses, with even a slightly higher effect of MUPS on HCU for the number of contacts. Third, as HCU was asked over the past six months, the risk of recall bias exists. However, the direction of this bias is unclear. Other studies have used electronic medical records to assess HCU, but these can also be incomplete [40]. Fourth, we have no information on the actual reasons for health care use and have only assessed the quantity and not the appropriateness of provided health care. Also we did not control for the use of psychopharmacological therapy which could have influenced the results, nor can we determine whether MUPS were the primary problems or symptoms from a

178

MUPS and health care use over two years

psychiatric disorder. Fifth, only the baseline sociodemographic variables and number of chronic diseases were included as covariates in the analyses, although several of these covariates may have changed over the course of two years. However, we do not believe that the results of our adjusted analyses would be different because the influence of these covariates was only marginal. Finally, based on our study we cannot infer causality. Therefore it might at least theoretically be possible that an increase in HCU is leading to more MUPS. Implications for clinical practice and future research As we found that MUPS are independently associated with HCU, attention should be paid to early identification and adequate treatment of MUPS in clinical practice. Also, physicians should be aware of signs of depression and anxiety and personality traits as they have an influence on the association between MUPS and HCU. A possible explanation for prolonged high HCU is that care for patients with MUPS is often fragmented, as diagnostics and treatments are carried out by different health care providers. To reduce high and possibly inadequate health care for these patients, the issue of adequacy and fragmentation of health care patterns should be further examined, also in relation to the patientâ&#x20AC;&#x2122;s quality of life. Guidelines for MUPS across disciplines and different health care settings may be instrumental in this examination [41, 42].

Also, it would be interesting to examine the role of health anxiety as a predictor of HCU in a longitudinal design. As we found neuroticism to be a predictor, it is possible that health anxiety could play an important role as well as these are often related.

CONCLUSIONS Our study showed that MUPS are positively associated with HCU over time, even after adjustment for depressive and anxiety disorders and personality traits. This suggests that good MUPS management is important. Further research is needed to investigate the adequacy of health care use patterns and the association with the patientâ&#x20AC;&#x2122;s quality of life.

179

8

Chapter 8

ACKNOWLEDGEMENTS The authors would like to thank all contributors at the NESDA. The infrastructure of the NESDA is funded through the Geestkracht program of the Netherlands Organisation for Health Research and Development (ZonMw, grant number 10-000-1002) and is funded by participating universities and mental health care organizations (VU University Medical Center, GGZ inGeest, Arkin, Leiden University Medical Center, GGZ Rivierduinen, University Medical Center Groningen, Lentis, GGZ Friesland, GGZ Drenthe, Scientific Institute for Quality of Healthcare (IQ healthcare), Netherlands Institute for Health Services Research (NIVEL) and Netherlands Institute of Mental Health and Addiction (Trimbos).

180

MUPS and health care use over two years

REFERENCES 1. Khan AA, Khan A, Harezlak J, Tu W, et al. Somatic symptoms in primary care: etiology and outcome. Psychosomatics. 2003;44:471–8.

2. Burton C. Beyond somatisation: a review of the understanding and treatment of medically unexplained physical symptoms. Br J Gen Pract. 2003;53:231–9.

3. Jackson JL, Passamonti M. The outcomes among patients presenting in primary care with a physical symptom at 5 Years. J Gen Intern Med. 2005;20:1032–7.

4. Nimnuan C, Hotopf M, Wessely S. Medically unexplained symptoms: An epidemiological study in seven specialities. J Psychosom Res. 2001;51:361–7.

5. Haller H, Cramer H, Lauche R, et al. Somatoform disorders and medically unexplained symptoms in Primary Care: A systematic review and meta-analysis of prevalence. Dtsch Arztebl Int. 2015;112: 279-87.

6. Fink P, Rosendal M, Olesen F. Classification of somatization and functional somatic symptoms in primary care. Aust N Z J Psychiatry. 2005;39:772–81.

7. Smith RC, Dwamena FC. Classification and diagnosis of patients with medically unexplained symptoms. J Gen Intern Med. 2007;22:685–91.

8. Barsky AJ, Orav JE, Bates DW. Somatization increases medical utilization and costs independent of psychiatric and medical comorbidity. Arch Gen Psychiatry. 2005;62:903–10.

9. Konnopka A, Schaefert R, Heinrich S, et al. Economics of medically unexplained symptoms: a systematic review of the literature. Psychother Psychosom. 2012;81:265–75.

10. Escobar JI, Golding JM, Hough RL, et al. Somatization in the community: relationship to disability and use of services. Am J Public Health. 1987;77:837–40.

11. Salmon P, Humphris GM, Ring A, et al. Primary care consultations about medically unexplained symptoms: Patient presentations and doctor responses that influence the probability of somatic intervention. Psychosomatic Medicine. 2007;69:571-577.

12. Ring A, Dowrick C, Humphris G, et al. Do patients with unexplained physical symptoms pressurise general practitioners for somatic treatment? A qualitative study. Br Med J. 2004;328:1057-1060 13. Fink P, Sørensen L, Engberg M, et al. Somatization in primary care. Prevalence, health care utilization, and general practitioner recognition. Psychosomatics. 1999;40:330–8.

14. Hansen MS, Fink P, Frydenberg M, et al. Use of health services, mental illness, and self-rated disability and health in medical inpatients. Psychosom Med. 2002;64:668–75.

15. Williams ER, Guthrie E, Mackway-Jones K, et al. Psychiatric status, somatisation, and health care utilization of frequent attenders at the emergency department: a comparison with routine attenders. J Psychosom Res. 2001;50:161–7.

16. Smits FT, Brouwer HJ, ter Riet G, et al. Epidemiology of frequent attenders: a 3-year historic cohort study comparing attendance, morbidity and prescriptions of one-year and persistent frequent attenders. BMC Public Health. 2009;24;9:36.

17. Andersen NLT, Eplov LF, Andersen JT, et al. Health care use by patients with somatoform disorders: a register-based follow-up study. Psychosomatics. 2013;54:132–41.

18. Grabe HJ, Baumeister SE, John U, et al. Association of mental distress with health care utilization and costs: a 5-year observation in a general population. Soc Psychiatry Psychiatr Epidemiol. 2009;44:835–44.

181

8

Chapter 8

19. Kroenke K, Jackson JL, Chamberlin J. Depressive and anxiety disorders in patients presenting with physical complaints: clinical predictors and outcome. Am J Med. 1997;103:339–47.

20. Henningsen P, Zimmermann T, Sattel H. Medically unexplained physical symptoms, anxiety, and depression: A meta-analytic review. Psychosom Med. 2003;65:528–33.

21. Van Boven K, Lucassen P, van Ravesteijn H, et al. Do unexplained symptoms predict anxiety or depression? Ten-year data from a practice-based research network. Br J Gen Pract. 2011;61:316–25.

22. Waal MWMD, Arnold IA, Eekhof JAH, et al. Somatoform disorders in general practice.

Prevalence, functional impairment and comorbidity with anxiety and depressive disorders. Br J Psychiatry. 2004;184:470–6.

23. Garcia-Campayo J, Alda M, Sobradiel N, et al. Personality disorders in somatization disorder patients: a controlled study in Spain. J Psychosom Res. 2007;62:675–80.

24. Noyes R, Langbehn DR, Happel RL, et al. Personality dysfunction among somatizing patients. Psychosomatics. 2001;42:320–9.

25. Penninx BWJH, Beekman ATF, Smit JH, et al. The Netherlands Study of Depression and Anxiety (NESDA): rationale, objectives and methods. Int J Methods Psychiatr Res. 2008;17:121–40.

26. Bouwmans C, Jong KD, Timman R, et al. Feasibility, reliability and validity of a questionnaire on healthcare consumption and productivity loss in patients with a psychiatric disorder. BMC Health Serv Res. 2013;13:217.

27. Terluin B, van Marwijk HW, Adèr HJ, et al. The four-dimensional symptom questionnaire: a validation study of a multidimensional self-report questionnaire to assess distress, depression, anxiety and somatization. BMC Psychiatry. 2006;6:34.

28. Terluin B, Brouwers EP, van Marwijk HW, et al. Detecting depressive and anxiety disorders in distressed patients in primary care; comparative diagnostic accuracy of the Four-dimensional symptom questionnaire and the hospital anxiety and depression scale. BMC Fam Pract. 2009;10:58.

29. Kroenke K, Spitzer RL, Williams JBW. The PHQ-15: validity of a new measure for evaluating the severity of somatic symptoms. Psychosom Med. 2002;64:258–66.

30. De Vroege L, Emons WHM, Sijtsma K, et al. Validation of the 4DSQ somatization subscale in the occupational health care setting as a screener. J Occup Rehabil. 2015;25:105–15.

31. Costa jr. P, McCrae R. Revised NEO Personality Inventory and NEO Five-Factory Inventory professonial manual. 1992.

32. Murray G, Rawlings D, Allen N, et al. NEO Five-Factor Inventory scores: psychometric properties in a community sample. Meas Eval Couns Dev. 2003;36:140-9.

33. Hoekstra H, Ormel J, de Fruyt F. Handleiding NEO persoonlijkheidsvragen NEO-PI-R en NEO-FFI. (Manual NEO personality questions NEO-PI-R and NEO-FFI). Lisse: Swets & Zeitlinger. 1996.

34. Kroenke K. Symptoms in the community: prevalence, classification, and psychiatric comorbidity. Arch Intern Med. 1993;153:2474–80.

35. Aamland A, Malterud K, Werner EL. Patients with persistent medically unexplained physical symptoms: a descriptive study from Norwegian general practice. BMC Fam Pract. 2014;15:107.

36. Centraal Bureau voor de Statistiek. www.cbs.nl. Access date: October 27th 2015.

182

MUPS and health care use over two years

37. Twisk JWR. Applied longitudinal data analysis for epidemiology (2nd edition). Cambridge: Cambridge University Press. 2013.

38. Carlier IVE, Colijn S, van Rood YR, et al. A comparative analysis of personality pathology profiles among patients with pure depressive-, pure anxiety-, and pure somatoform disorders. J Affect Disord. 2014;168:322–30.

39. Crombez G, Beijrens K, van Damme S, et al. The unbearable lightness of somatisation: a systematic review of the concept of somatisation in empirical studies of pain. Pain.

2009;145:31–5.

40. Bruijnzeels MA, van der Wouden JC, Foets M, et al. Validity and accuracy of interview and diary data on children’s medical utilisation in The Netherlands. J Epidemiol Community Health. 1998;52:65–9.

41. Olde Hartman T, Blankenstein N, Molenaar B. NHG-Standaard somatisch onvoldoende verklaarde lichamelijke klachten. (Dutch college of general practioner’s guideline medically unexplained physical symptoms). Huisarts Wet. 2013;56:222–30.

42. Van der Feltz-Cornelis CM, Hoedeman R, Keuter EJW, et al. Presentation of the multidisciplinary guideline medically unexplained physical symptoms and somatoform disorder in the Netherlands: disease management according to risk profiles. J Psychosom Res. 2012;72:168–9.

8

183

9 Medically unexplained physical symptoms and work functioning over two years: their association and the influence of depressive and anxiety disorders and job characteristics Madelon den Boeft Jos WR. Twisk Berend Terluin Brenda WJH. Penninx Harm WJ. van Marwijk Mattijs E. Numans Johannes C. van der Wouden Henriette E. van der Horst

BMC Family Practice. 2016;17:46

Chapter 9

ABSTRACT Background Medically unexplained physical symptoms (MUPS) are highly prevalent and may affect work functioning. In this study we aimed to assess the longitudinal association between MUPS and work functioning over two years and the influence of job characteristics and depressive and anxiety disorders on this association.

Methods We assessed the longitudinal association between MUPS and work functioning, operationalized in terms of absenteeism and disability at work, in 1887 working participants from the Netherlands Study of Depression and Anxiety (NESDA). The NESDA study population included participants with a current depressive and/or anxiety disorder, participants with a lifetime risk and/or subthreshold symptoms and healthy controls. Absenteeism was assessed with the Health and Labour Questionnaire Short Form and disability with the World Health Organization Disability Assessment Schedule II. MUPS were measured with the Four Dimensional Symptom Questionnaire.

Measurements were taken at baseline and at two years follow-up. We used mixed model analyses to correct for the dependency of observations within participants.

Results MUPS were positively associated with disability (regression coefficient 0.304; 95% CI 0.281-0.327) and with short and long-term absenteeism over two years (OR 1.030, 95% CI 1.016-1.045; OR 1.099, 95% CI 1.085-1.114). After adjusting for depressive disorders, anxiety disorders and job characteristics, associations weakened but remained significant.

Conclusion Our results show that MUPS were positively associated with disability and absenteeism over two years, even after adjusting for depressive and anxiety disorders and job characteristics. This suggests that early identification of MUPS and adequate management is important.

186

MUPS and work functioning over two years

BACKGROUND Medically unexplained physical symptoms (MUPS) are often presented to both primary and secondary health care physicians and may have major impact on daily functioning[1,2]. MUPS are physical symptoms not satisfactorily explained by a somatic underlying condition after an adequate medical examination and represent a broad spectrum of symptoms in varying degrees of severity[3,4] .

It is known that MUPS interfere with work functioning, both in terms of absenteeism and disability[5–7]. Not being able to work or not performing at work optimally is not only a burden for patients and their direct environment, but also for the community due to increasing costs[8,9].

In the Netherlands, studies among workers have shown that high levels of somatic symptoms and distress are determinants for prolonged absenteeism and enduring disabilities[6,10,11]. It was also shown that the prevalence of severe MUPS was higher in the long-term absent employees compared to the non-sick working population[7].

These findings are supported by international studies[12,13] and the relevance of the problem is reinforced by a 10-year follow-up study by Rask et al, who concluded that not only severe MUPS have significant impact on work functioning, but also mild and recent onset MUPS[5]. Despite the relevance of absenteeism and disability from work caused by MUPS, limited research has been performed to assess potential influencing factors on the relationship between MUPS and work functioning and their association over time[14].

It is known that unfavourable job characteristics, such as long working hours and low occupational status, can influence someone’s functioning negatively. Lower graded jobs are associated with more absenteeism and disability, compared to higher graded jobs[15]. The same applies to high demands, whether or not in combination with low support by colleagues, low task control[16–18] and long working hours[19].

The question arises to what extent this also applies to MUPS. Furthermore, as we know that MUPS are often accompanied by depressive and anxiety disorders, it is important to understand the influence of comorbidity of depressive and/or anxiety disorders on work functioning over time in patients considered having MUPS[20].

187

9

Chapter 9

Deeper insight into the association between MUPS and work functioning and influencing factors such as job characteristics and comorbid psychiatric disorders can assist physicians in the early identification of patients at risk and it may provide opportunities to develop adequate management and prevention strategies. Therefore, in this longitudinal study we first assessed the association between MUPS and work functioning, both absenteeism and disability at work, over two years. Second, we investigated the influence of job characteristics and depressive and/or anxiety disorders as potential confounders in this association.

METHODS Design and study population The present longitudinal analysis is part of the Netherlands Study of Depression and Anxiety (NESDA). The NESDA is a multisite naturalistic cohort study, which aims to describe the long-term course and consequences of depressive and anxiety disorders and to examine its predictors. A detailed description of the design and sampling is provided elsewhere[21]. In summary, 2981 participants between 18 and 65 years were included. The NESDA study population consisted of participants with current depressive and/or anxiety disorders, participants with a lifetime risk or subthreshold symptoms and healthy controls. Recruitment of participants took place in primary care practices (n=1610), outpatient mental health care institutions (n=807) and in the general population (n=564). Exclusion criteria were not being fluent in the Dutch language or having a primary diagnosis of a psychotic, obsessive, bipolar or severe substance abuse disorder. Baseline data (T0) were collected between 2004 and 2007.

The research protocol was approved by the ethical committees of participating universities and performed in accordance with the ethical standards of the Declaration of Helsinki. All participants provided written informed consent.

For the present study, we selected working participants (n=1887), i.e. participants who had a paid job for more than eight hours a week at baseline (T0). Of them, 1665 (88.2%) completed the questionnaires regarding work functioning and MUPS and were therefore included in this study (591 male, 1074 female). After two years (T1), 1455 (87.4%) participants had a follow up assessment.

188

MUPS and work functioning over two years

Work functioning We conceptualized work functioning in terms of disability at work and absenteeism from work.

Disability at work was assessed with the validated World Health Organization Disability Assessment Schedule II (WODAS-II) [22]. This is a 36-item instrument, assessing functioning and disability and focusing on six subscales regarding several domains in life. We focused on the domain of work activities which includes four items: 1) difficulty in day-to-day work, 2) difficulty in doing most important work tasks well, 3) difficulty in getting all work done and 4) difficulty in getting work done as quickly as needed. Respondents rated difficulty on a 5-point scale, ranging from 1 (none) to 5 (extreme, cannot do). Therefore the summated work domain scale has a range from 4-20 with higher scores representing a higher level of disability.

Absenteeism from work was assessed with the validated Health and Labour Questionnaire Short Form[23] and was computed by dividing the number of absent days from work during the past six months because of health problems by the number of actual work days per week. As this variable does not meet normality assumptions, we categorized it into three categories: no absenteeism, short-term absenteeism (<2 weeks) and long-term absenteeism (2 weeks or longer), cf. Plaisier et al[24]. The latter two categories made a distinction possible between short-term absenteeism, probably due to self-limiting conditions such as common colds, and long-term absenteeism, which could indicate more chronic conditions.

Medically unexplained physical symptoms MUPS were measured with the somatization scale of the validated Four Dimensional Symptoms Questionnaire (4DSQ), a self-report questionnaire developed to measure distress, depression, anxiety and somatization as separate dimensions in primary care.(25) The somatization scale comprises 16 items, all physical symptoms. The response categories on a 5-point Likert scale are worded as follows: “no”, “sometimes”, “regularly”, “often” and “very often or constantly”. In order to arrive at scale scores, the responses are scored as 0 for “no”, 1 for “sometimes” and 2 for “regularly”, “often” and “very often or constantly” and the item scores were summated into a scale score, as done by Terluin et al. with a range of 0-32[25], indicating that MUPS is considered

189

9

Chapter 9

to be a summation of physical symptoms. Additionally, in order to facilitate clinical use and to overcome the fact that there is no linear relationship between MUPS and work functioning, we repeated the analyses with a dichotomized scale using 11 points as a clinical cut-off score, since a score of 11 or higher is considered to indicate MUPS[25].

Job characteristics Job characteristics were conceptualized as working hours, occupational status and psychosocial working conditions. The occupational status variable was created using the occupational categories provided by Statistics Netherlands and additional self-reported information on employment status. The original eleven categories of Statistics Netherlands were recoded into five categories: 1) high graded non-manual workers, 2) medium or low skilled non-manual workers, 3) self-employed, 4) high skilled manual workers and 5) medium or low skilled manual workers, as was done by Plaisier et al[26]. Psychosocial working conditions consist of job demands, job control and job support and were measured with a questionnaire consisting of dichotomous items, based on the demands/control model[27]. Data on job characteristics were only gathered at baseline (T0). Unfortunately for 30% of the participants no information regarding job characteristics could be obtained.

Depressive and/or anxiety disorders The presence of depressive disorders, including major depressive disorder and dysthymia as well as anxiety disorders, including generalized anxiety disorder, panic disorder with or without agoraphobia, social phobia and/or agoraphobia without panic disorder, were diagnosed with the validated Composite Interview Diagnostic Instrument (CIDI version 2.1). We only took into account diagnoses established during the past six months. Trained research staff interviewed all participants.

Socio-demographic covariates and chronic diseases Based on previous research on the association between MUPS and work functioning, the following covariates were considered as possible confounders: gender, age, level of education, marital status and the number of chronic diseases[5,28]. We divided the level of education into three groups (basic, intermediate, high), derived from the

190

MUPS and work functioning over two years

standard classification of education from Statistics Netherlands[29]. Marital status was divided into two categories: participants married or living with a partner and participants not married or living without a partner. Participants were asked if they suffered from chronic diseases from each of the following categories: respiratory, cardiometabolic, musculoskeletal, digestive, neurological, endocrine and cancer.

Diseases were only included if participants were currently treated with medicines or under specialist care. The number of diseases were summated into a scale score.

Data on the covariates were only gathered at baseline (T0).

Analysis Descriptive statistics are presented as mean with standard deviation for normally distributed continuous data, median and inter-quartile range for skewed continuous variables and as numbers and percentages for dichotomous and categorical variables.

To assess the association between MUPS and work functioning over two years (T0, T1), we used linear mixed model analyses for disability and multinomial logistic mixed model analyses for absenteeism. After crude analyses, we performed adjusted analyses where covariates (sociodemographic variables and the number of chronic diseases) were included in the model. Next, we examined the confounding role of depressive and/or anxiety disorders and job characteristics on the adjusted association between MUPS and work functioning, first separately and then combined.

Effects were expressed as regression coefficients (for disability) and odds ratios (ORs) (for absenteeism) with 95% confidence intervals (CI), representing the longitudinal association between MUPS and work functioning on average over time, reflecting both the within and between subject relationship[30]. For absenteeism, â&#x20AC;&#x2DC;no absenteeismâ&#x20AC;&#x2122; was used as the reference category.

As depressive and anxiety disorders were overrepresented in the NESDA cohort, we added interaction terms to the analysis to assess if depressive and/or anxiety disorders modify the association between MUPS and work functioning. We used all observations in our analyses. As we used a longitudinal design, missing data did not have to be imputed[30].

The statistical analyses were performed in MLwiN (version 2.31) and Stata (version 13).

191

9

Chapter 9

RESULTS Table 1 shows the demographic and clinical characteristics of the study sample. At baseline, the mean age was 40 years and 65% were women. The median score for MUPS was 8 at baseline and 6 at two years follow up (p<0.001). When using the clinical cut-off point of 37% had MUPS at T0 and 22% had MUPS at T1.

The longitudinal association between MUPS and disability at work We found a strong positive association between MUPS and disability at work on average over time (table 2). The estimated crude regression coefficient of 0.295 for MUPS, measured with the continuous 4DSQ somatization scale, in relation to disability at work can be interpreted as follows: for every unit increase/difference in the 4DSQ, there is a 0.295 increase/difference in the severity of disability. Depressive disorders had the greatest influence on the association between MUPS and disability as the regression coefficient became smaller, followed by anxiety disorders. However, the association between MUPS and disability remained statistically significant. Job characteristics did not affect the association. For the dichotomous 4DSQ comparable results were found.

When we examined whether the association between MUPS and disability was modified by depressive and/or anxiety disorders, we found a strong significant negative interaction for depressive disorders (p<0.001) and a borderline significant negative interaction for anxiety disorders (p=0.054), indicating that the association between MUPS and disability was weaker for participants with a depressive and/or anxiety disorder. For the dichotomous 4DSQ comparable results were found.

The longitudinal association between MUPS and absenteeism from work We found a strong positive association between MUPS and short-term absenteeism and an even stronger effect for long-term absenteeism, on average over time, both compared to the reference category â&#x20AC;&#x2DC;no absenteeismâ&#x20AC;&#x2122; (table 3). Depressive disorders had the strongest influence on the association between MUPS (with use of the continuous 4DSQ) and short-term absenteeism (37% decrease in regression coefficient), followed by job characteristics (mainly caused by job support and job control; 33%

192

MUPS and work functioning over two years

Table 1: Demographic and clinical characteristics of research population Baseline (T0) (n=1665)

Two year follow up (T1) (n=1455)

Females, number (%)

1074 (64.5)

941 (64.7)

Age in years, mean (SD)

40.8 (11.7)

40.1 (11.4)

Socio-demographics

Level of education, number (%)

Basic

68 (4.1)

56 (3.8)

Intermediate

886 (53.2)

749 (51.5)

High

711 (42.7)

650 (44.7)

Number of chronic diseases, median (IQR)

0.00 (0.00-1.00)

0.00 (0.00-1.00)

Married or with partner, number (%)

1223 (73.5)

1047 (72.0)

Medically unexplained physical symptoms Total score (0-32), median (IQR)

8.00 (4.00-13.00)

6.00 (3.00-10.00)

Somatisation (cut-off 11 points) (%)

621 (37.3)

322 (22.1)

Work functioning Absenteeism, number (%)

No absenteeism

786 (47.2)

804 (55.3)

Short term absenteeism (<2 weeks)

508 (30.5)

467 (32.1)

Long term absenteeism (â&#x2030;Ľ2 weeks)

371 (22.3)

184 (12.6)

8.5 (4.5)

7.1 (3.7)

Disability (0-20), mean (SD) Job characteristics, median (IQR)

Working hours

32.00 (24.00-38.00) -

Job demands

0.40 (0.20-0.80)

-

Job control

0.79 (0.61-0.93)

-

Job support

0.75 (0.50-1.00)

-

9

Psychiatric comorbidity (number, %)

Depressive disorder

572 (34.4)

285 (19.6)

Anxiety disorder

653 (39.2)

333 (22.9)

Means with standard deviations are given for normally distributed continuous variables, medians and inter-quartile ranges for skewed continuous variables and frequencies with percentages for dichotomous and categorical variables. SD: standard deviation. IQR: interquartile range.

193

Chapter 9

Table 2: Analyses results on the association between MUPS and disability at work over two years Regression coefficient (95% CI) (with continuous 4DSQ)

Regression coefficient (95% CI) (with dichotomous 4DSQ)

Crude

0.30 (0.27-0.32)

3.23 (2.93-3.53)

Adjusted*

0.30 (0.28-0.33)

3.21 (2.91-3.52)

Univariable**

Anxiety disorder

0.26 (0.24-0.29)

2.68 (2.37-3.00)

Depressive disorder

0.23 (0.21-0.26)

2.39 (2.10-2.70)

Job characteristics

Working hours

0.30 (0.28-0.33)

3.21 (2.90-3.52)

Occupational status

0.32 (0.29-0.34)

3.20 (2.83-3.57)

Job demands

0.30 (0.28-0.33)

3.05 (2.71-3.38)

Job control

0.30 (0.28-0.33)

3.00 (2.67-3.34)

Job support

0.30 (0.27-0.32)

2.91 (2.58-3.25)

Multivariable**

0.21 (0.18-0.24)

1.95 (1.59-2.31)

4DSQ: four-dimensional symptom questionnaire *Adjusted for age, gender, level of education, marital status, number of chronic diseases ** Univariable: influence of covariates separately; Multivariable: influence of covariates combined

and 37% decrease in regression coefficient) and anxiety disorders (20% decrease in regression coefficient). Depressive disorders also had the strongest influence on the association between MUPS and long-term absenteeism (40% decrease in regression coefficient), followed by anxiety disorders (20% decrease in regression coefficient).

Job characteristics did not affect the association. Despite adjusting for these confounders, the association between MUPS and absenteeism remained significant. For the dichotomous 4DSQ comparable results were found.

When we examined whether the association between MUPS and absenteeism was modified by depressive and/or anxiety disorders, we only found a negative interaction for anxiety disorders and long-term absenteeism (p=0.022), indicating that the association between MUPS and long-term absenteeism was weaker for participants with an anxiety disorder. We found no significant interaction for depressive disorders.

For the dichotomous 4DSQ comparable results were found.

194

MUPS and work functioning over two years

Table 3: Analyses results on the association between MUPS and absenteeism from work over two years Odds Ratio (95% CI) (with continuous 4DSQ)

Odds Ratio (95% CI) (with dichotomous 4DSQ)

Short-term absenteeism

Short-term absenteeism

Long-term absenteeism

Long-term absenteeism

Crude

1.03 (1.02-1.05)

1.11 (1.09-1.13)

1.41 (1.19-1.67)

3.64 (3.06-4.34)

Adjusted*

1.03 (1.02-1.05)

1.10 (1.09-1.11)

1.37 (1.15-1.62)

3.15 (2.62-3.78)

Univariable**

Anxiety disorder

1.02 (1.01-1.04)

1.08 (1.06-1.09)

1.26 (1.05-1.50)

2.47 (2.04-2.99)

Depressive disorder

1.02 (1.01-1.03)

1.06 (1.04-1.07)

1.21 (1.02-1.45)

2.08 (1.72-2.52)

Job characteristics

Working hours

1.03 (1.02-1.05)

1.10 (1.08-1.11)

1.37 (1.15-1.63)

3.15 (2.62-3.78)

Occupational status

1.03 (1.01-1.04)

1.10 (1.08-1.12)

1.27 (1.03-1.57)

2.81 (2.26-3.50)

Job demands

1.02 (1.01-1.04)

1.09 (1.08-1.11)

1.29 (1.07-1.56)

2.79 (2.28-3.41)

Job control

1.02 (1.01-1.03)

1.09 (1.07-1.11)

1.22 (1.01-1.48)

2.66 (2.17-3.25)

Job support

1.02 (1.01-1.04)

1.09 (1.07-1.11)

1.22 (1.01-1.48)

2.69 (2.20-3.30)

1.01 (1.00-1.03)

1.04 (1.02-1.06)

1.03 (0.82-1.30)

1.61 (1.26-2.05)

Multivariable***

Reference category: â&#x20AC;&#x2DC;no absenteeismâ&#x20AC;&#x2122;. 4DSQ: four-dimensional symptom questionnaire * Adjusted for age, gender, level of education, marital status, number of chronic diseases ** Univariable: influence of covariates separately; Multivariable: influence of covariates combined.

DISCUSSION Summary of results With this study we aimed to assess the association between MUPS and work functioning over two years and the influence of job characteristics and depressive and anxiety disorders on this association. We showed that MUPS was positively associated with disability and absenteeism from work over two years, with a stronger effect for long-term absenteeism than for short-term absenteeism. Depressive and anxiety disorders weakened the association between MUPS and work functioning, but the association remained significant. Job characteristics only weakened the association between MUPS and short-term absenteeism, but again the association remained significant.

195

9

Chapter 9

Comparison with existing literature Little research has been performed on MUPS and work functioning. Even though our study is not directly comparable with previously conducted research because of use of different designs and populations among other things, our study results are in line with other studies.

In a prospective study, Roelen et al. showed that various MUPS were associated with absenteeism and concluded that the more symptoms at baseline, the higher the risk of absenteeism a year later[10]. In two cross-sectional studies, Hoedeman et al. concluded that 15% of employees on sickness leave suffered from severe MUPS and that severe MUPS was associated with four to six times more comorbid depressive and anxiety disorders. The authors used a cut-off score of 15 on the PHQ-15 for the categorisation of severe MUPS. Furthermore, they concluded that employees with severe MUPS had a longer duration of absenteeism[6,7]. In a 5-year follow-up study, Loengaard et al. found that MUPS patients had an increased risk of long-term absenteeism compared to healthy participants[31]. Rask et al performed a 10-year follow-up study and found that recent onset MUPS and somatoform disorders have significant negative long-term impact on patient work functioning[5]. They also concluded that although depressive and anxiety disorders influence the association between MUPS and work functioning, both psychiatric disorders did not fully explain the effect, which corresponds with our findings.

We are not aware of earlier studies that examined the influence of specific job characteristics in patients with MUPS. However, there are some studies that describe the association of job characteristics with work functioning and mental health problems in the general working population. We found that job support and job control weakened the association between MUPS and short-term absenteeism, which is in line with North et al. and Melchior et al [16,18]. Christensen et al. found that absenteeism rates were higher for lower graded jobs compared to higher graded jobs, while Sparks et al. found that the number of working hours might be important for work functioning[19]. We did not find the same results for MUPS.

196

MUPS and work functioning over two years

Strengths and limitations We believe our study has several methodological strengths, in particular the large sample of participants, recruited from primary care and secondary mental health care, and the longitudinal design. Also, all analyses were adjusted for the number of chronic diseases. Finally, for the assessment of depressive and anxiety disorders, structured interviews were used.

Our findings should be interpreted in the light of several limitations. First, as for all questionnaires assessing MUPS, the 4DSQ lacks clinical judgement. However, the 4DSQ highly correlates with the Patient Health Questionnaire-15 and the Symptom Checklist-90, other questionnaires widely used to measure MUPS. In our study we defined MUPS based on the score of the 4DSQ and MUPS is defined as a summation of physical symptoms, which often remain unexplained after appropriate examination. It should be kept in mind that the definition of MUPS is controversial, i.e. whether it represents a specific disorder or whether it is a way of presenting different types of emotional distress. Second, time windows differed between the concepts: disability was measured over the past month and absenteeism over the past six months, while MUPS was measured over the past week. However, we do not believe that the results were much affected by this incongruence. In an additional analysis we found a correlation coefficient of 0.72 between MUPS at T0 and T1, indicating a quite stable pattern of MUPS over two years. Also the additional time lag analysis, relating MUPS at T0 to disability and absenteeism at T1, revealed more or less the same results (data not shown). Third, because job characteristics had 30% missing values, we repeated the crude analyses in a reduced dataset with only complete cases. The results showed that job characteristics still had a weakening influence on the association MUPS and short-term absenteeism, but that this influence was weaker. Fourth, we have no information on the actual reasons for disability and absenteeism. Fifth, participants with depressive and/or anxiety disorders were oversampled in our study population. It is often assumed that the observed strong association between MUPS and work functioning is because of this oversampling.

With the interaction analyses however, we showed that it is actually the other way around: that within participants without these disorders the association between MUPS and work functioning was stronger than in participants with these disorders.

197

9

Chapter 9

Therefore we wanted to emphasize that the observed associations may also hold for the general population. It should be noted that the negative interaction between MUPS and depressive/anxiety disorders is probably due to the strong association between depressive/anxiety disorders and disability.

Finally, the population at follow-up is slightly healthier than the original population.

Although it is possible that those who did not complete follow-up measurements were more severely disturbed, we do not think that this highly influences the results of our study. However, if there is an influence, the magnitude of the observed association between MUPS and work functioning might be an underestimation of the real association due to this phenomenon.

Implication for clinical practice and future research Disability and absenteeism may lead to a decreased quality of life for patients and high direct and indirect costs[8,9,32]. As we found that MUPS have an independent negative association with work functioning, attention should be paid to preventive measures in the work place, early identification of MUPS and adequate treatment.

However, as depressive and anxiety disorders weakened the association between MUPS and work functioning, physicians should be aware of signs and symptoms of depressive or anxiety disorders.

As said, little is known about work environment and only a small number of studies assessed phenomena associated with disability and absenteeism among workers with MUPS[14]. With our analyses, we have shown that job support and job control influence the association between MUPS and mainly short-term absenteeism. More insight is needed about favourable job characteristics to develop interventions for prevention and treatment and return to work programs.

CONCLUSIONS Our results show that MUPS were positively associated with disability and absenteeism over two years, even after adjusting for depressive and anxiety disorders and job characteristics. This suggests that early identification of MUPS and adequate management is important.

198

MUPS and work functioning over two years

ACKNOWLEDGEMENTS The authors would like to thank all contributors at the NESDA. The infrastructure of the NESDA is funded through the Geestkracht program of the Netherlands Organisation for Health Research and Development (ZonMw, grant number 10-000-1002) and is funded by participating universities and mental health care organizations (VU University Medical Center, GGZ inGeest, Arkin, Leiden University Medical Center, GGZ Rivierduinen, University Medical Center Groningen, Lentis, GGZ Friesland, GGZ Drenthe, Scientific Institute for Quality of Healthcare (IQ healthcare), Netherlands Institute for Health Services Research (NIVEL) and Netherlands Institute of Mental Health and Addiction (Trimbos).

9

199

Chapter 9

REFERENCES 1. Burton C. Beyond somatisation: a review of the understanding and treatment of medically unexplained physical symptoms. Br J Gen Pract. 2003;53:231–9.

2. Kroenke K, Spitzer RL, Williams JBW. The PHQ-15: Validity of a new measure for evaluating the severity of somatic symptoms. Psychosom Med. 2002;64:258–66.

3. Fink P, Rosendal M, Olesen F. Classification of somatization and functional somatic symptoms in primary care. Aust N Z J Psychiatry. 2005;39:772–81.

4. Smith RC, Dwamena FC. Classification and diagnosis of patients with medically unexplained symptoms. J Gen Intern Med. 2007;22:685–91.

5. Rask MT, Rosendal M, Fenger-Grøn M, et al. Sick leave and work disability in primary care patients with recent-onset multiple medically unexplained symptoms and persistent somatoform disorders: a 10-year follow-up of the FIP study. Gen Hosp Psychiatry.

2015;37:53–9.

6. Hoedeman R, Blankenstein AH, Krol B, et al. The contribution of high levels of somatic symptom severity to sickness absence duration, disability and discharge. J Occup Rehabil. 2010;20:264–73.

7. Hoedeman R, Krol B, Blankenstein N, et al. Severe MUPS in a sick-listed population: a cross-sectional study on prevalence, recognition, psychiatric co-morbidity and impairment. BMC Public Health. 2009;9:440.

8. Konnopka A, Kaufmann C, König HH, et al. Association of costs with somatic symptom severity in patients with medically unexplained symptoms. J Psychosom Res. 2013;75:370–5.

9. Konnopka A, Schaefert R, Heinrich S, et al. Economics of medically unexplained symptoms: a systematic review of the literature. Psychother Psychosom. 2012;81:265–75.

10. Roelen CAM, van der Pol TR, Koopmans PC, et al. Identifying workers at risk of sickness absence by questionnaire. Occup Med Oxf Engl. 2006;56:442–6.

11. Terluin B, van Rhenen W, Anema JR, et al. Psychological symptoms and subsequent sickness absence. Int Arch Occup Environ Health. 2011;84:825–37.

12. Norrmén G, Svärdsudd K, Andersson D. How primary health care physicians make sick listing decisions: the impact of medical factors and functioning. BMC Fam Pract.

2008;9:3.

13. Hansen A, Edlund C, Bränholm I. Significant resources needed for return to work after sick leave. Work Read Mass. 2005;25:231–40.

14. Aamland A, Malterud K, Werner EL. Phenomena associated with sick leave among primary care patients with Medically Unexplained Physical Symptoms: a systematic review. Scand J Prim Health Care. 2012;30:147–55.

15. Christensen KB, Labriola M, Lund T, et al. Explaining the social gradient in long-term sickness absence: a prospective study of Danish employees. J Epidemiol Community Health. 2008;62:181–3.

16. North FM, Syme SL, Feeney A, et al. Psychosocial work environment and sickness absence among British civil servants: the Whitehall II study. Am J Public Health.

1996;86:332–40.

200

MUPS and work functioning over two years

17. Roelen CAM, Schreuder KJ, Koopmans PC, et al. Perceived job demands relate to selfreported health complaints. Occup Med. 2008;58:58–63.

18. Melchior M, Niedhammer I, Berkman LF, et al. Do psychosocial work factors and social relations exert independent effects on sickness absence? A six-year prospective study of the GAZEL cohort. J Epidemiol Community Health. 2003;57:285–93.

19. Sparks K, Cooper C, Fried Y, et al. The effects of hours of work on health: a meta-analytic review. J Occup Organ Psychol. 1997:391–408.

20. van Boven K, Lucassen P, van Ravesteijn H, et al. Do unexplained symptoms predict anxiety or depression? Ten-year data from a practice-based research network. Br J Gen Pract. 2011;61:316–25.

21. Penninx BWJH, Beekman ATF, Smit JH, et al. The Netherlands Study of Depression and Anxiety: rationale, objectives and methods. Int J Methods Psychiatr Res. 2008;17:121– 40.

22. World Health Organization. Disability Assessment Schedule II. 2000 23. Bouwmans C, Jong KD, Timman R, et al. Feasibility, reliability and validity of a questionnaire on healthcare consumption and productivity loss in patients with a psychiatric disorder. BMC Health Serv Res. 2013;13:217.

24. Plaisier I, Beekman ATF, de Graaf R, et al. Work functioning in persons with depressive and anxiety disorders: The role of specific psychopathological characteristics. J Affect Disord. 2010;125:198–206.

25. Terluin B, van Marwijk HW, Adèr HJ, et al. The four-dimensional symptom questionnaire: a validation study of a multidimensional self-report questionnaire to assess distress, depression, anxiety and somatization. BMC Psychiatry. 2006;6:34.

26. Plaisier I, de Graaf R, de Bruijn J, et al. Depressive and anxiety disorders on-the-job: The importance of job characteristics for good work functioning in persons with depressive and anxiety disorders. Psychiatry Res. 2012;200:382–8.

27. Karasek R, Theorell T. Healthy work, stress, productivity, and the reconstruction of working life. New York: Basic Books, 1990.

28. Aamland A, Malterud K, Werner EL. Patients with persistent medically unexplained physical symptoms: a descriptive study from Norwegian general practice. BMC Fam Pract. 2014;15:107.

29. Centraal Bureau voor de Statistiek. Available from: www.cbs.nl. Access date: October 27th 2015.

30. Twisk JWR. Applied Longitudinal Data Analysis for Epidemiology. Cambridge UK: Cambridge University Press 2013. 2nd edition.

31. Loengaard K, Bjorner JB, Fink PK, et al. Medically unexplained symptoms and the risk of loss of labor market participation - a prospective study in the Danish population. BMC Public Health. 2015;15:844.

32. Aamland A, Werner EL, Malterud K. Sickness absence, marginality, and medically unexplained physical symptoms: a focus-group study of patients’ experiences. Scand J Prim Health Care. 2013;31:95–100.

201

9

10 General discussion

General discussion

In the present chapter, I will discuss the main findings of the papers that are presented in this thesis in the context of the three parts of this thesis: (i) the identification of medically unexplained physical symptoms (MUPS), (ii)structuring management of MUPS and (iii) societal aspects of MUPS. Where appropriate, I will also reflect on the methodological considerations and implications for clinical practice and further research in those three areas.

Identification of MUPS Both in clinical practice and in research, the identification of patients with MUPS is a challenging task. However, the early identification of patients with MUPS and those who are at risk of persisting and debilitating symptoms, may contribute to a timely and appropriate management and thus possibly mitigate the course of MUPS. In the first part of this thesis we tried to tackle the challenge of (early) identification of patients with MUPS from three different angles.

In Chapter 2 and 3, we explored whether it was possible to identify patients with MUPS and those at risk of developing an unfavourable chronic course from routine primary care electronic medical records (EMRs). In current clinical practice and in research this is already applied for other risk populations, such as patients at risk for cardiovascular diseases, patients with diabetes mellitus, frail elderly and patients at risk for the complications of influenza (1â&#x20AC;&#x201C;6). However, using EMRs to identify patients at risk for persistent MUPS is not yet common practice.

Regarding the identification of patients with MUPS, first, we validated an EMR screening method to identify patients with MUPS that was previously developed from data from the Utrecht Health Project (Leidsche Rijn Gezondheidsproject) at the Julius Center related to the University Medical Center Utrecht, the Netherlands (Chapter 2). We compared the results of this EMR screening method (i.e. patients with and without MUPS) with the scores of each patient on the patient health questionnaire-15 (PHQ-15), which we used as a reference test (7). We calculated the test characteristics for various PHQ-15 cut-off points. For the PHQ-15 cut-off point 10 (a moderate somatic symptom severity score, which is a commonly used cut-off), we

205

10

Chapter 10

found a high specificity (0.93), but a low sensitivity (0.30). This means that probably many patients with MUPS were missed when using this selection algorithm.

As this EMR screening method seems not able to adequately select patients with MUPS and was not developed for identifying patients at risk for persistent symptoms, we decided to explore other, more advanced, analytical methods for risk assessment.

The final goal of the development of our risk assessment method was to transform the method into a software algorithm, which then can be implemented in daily clinical practice in order to be part of proactive, structured care for patients with MUPS. In a pre-study, we tried to identify patients with MUPS by performing a latent class analysis in an EMR dataset derived from the academic network of GPs in Amsterdam.

Unfortunately we did not succeed as we could not detect and validate one or more classes of patients with MUPS. Therefore, we decided to change our direction and to use other, more innovative statistical machine learning techniques to support our risk assessment. For the development of the algorithms we used EMRs from 22 Dutch GP practices from Utrecht, the Netherlands (Chapter 3). Both developed models, developed with logistic regression analysis and decision tree analysis, were able to identify patients at risk for persistent MUPS moderate to good, measured with AUCs of 0.70 and 0.81, respectively. Even after cross-validation, which is a particular strength of this study, the AUCs remained stable.

Our study results contribute to the body of knowledge that already exists on identification of patients with MUPS in EMRs. We also found more or less comparable results (8â&#x20AC;&#x201C;10). We can conclude that the (early) identification of patients with MUPS in EMRs is not a trivial task, but we believe that we found a promising perspective for further fine-tuning. Regarding our developed risk assessment models, one of the possible challenges for further fine-tuning could be to include separate International Classification of Primary Care (ICPC) symptom codes (1-29) instead of complete ICPC chapters including those and confirmed diagnose codes (70-99) as well (e.g. complete Chapter A).

An important methodological issue in the development of the identification and risk assessment models is the use of EMRs. The completeness and correctness of the data in an EMR dataset strongly depends on the ICPC coding behaviour of GPs and other practice employees. GPs mostly differ in their coding behaviour in coding

206

General discussion

symptoms (ICPC code 1-29) versus coding diagnoses (ICPC code 70-99). This has to do with the rules of coding on a true level of understanding, which some GPs apply more consistent than others. For example, this means that different GPs can code either constipation (D12), diarrhoea (D11) or irritable bowel syndrome (D93) for the same patient that consults the GP with intestinal complaints. Also, when a patient presents multiple complaints within one consultation, some GPs only code one complaint, often the most severe complaint, while other GPs code all complaints. In the past years the focus on correct and complete coding has increased in clinical practice and research. But it should still receive attention and awareness, for example in forms of structured coding educational activities for GPs in order to further optimise EMRs for use in research as well as for quality management and disease management purposes.

Another methodological issue related to the identification of MUPS is the operationalization of the MUPS outcome variable. In literature there is an on-going discussion about definitions and classifications(11,12). Differentiation is mainly based on the number of complaints and the severity of symptoms, the impact of symptoms on patientâ&#x20AC;&#x2122;s daily life or symptoms from different organ clusters. In the development of our risk assessment models, we chose to operationalize persistent MUPS with the use of the three ICPC codes for MUPS related chronic syndromes (irritable bowel syndrome (D93), fibromyalgia (L18.01) and chronic fatigue syndrome (A04.01)) and chronic or recurrent low back pain without radiation (L03), one of the most common MUPS. Hereby we aimed to capture the largest group of MUPS patients at risk for one of these persistent outcomes. It must me noted that other studies have made other choices when operationalizing persistent MUPS, for example by using the duration of MUPS. Next to the development of the risk assessment models, we performed a qualitative focus group study (Chapter 4). In this study we explored how GPs thought they recognized patients with MUPS during consultations, as little is known about this process. More insight into how GPs recognize patients with MUPS and how GPs code these patients in EMRs could contribute to further fine-tuning and improvement of identification and risk assessment methods. In addition, we asked the GPs participating in the focus groups sessions whether they could recognize different subgroups of

207

10

Chapter 10

patients with MUPS. This might not only contribute to improved identification, but also to a more tailored management because it is possible that different subgroups of patients have different needs in their management.

In the first part of our analysis, which was related to recognition in general, it became clear that most GPs use the core values of family practice in recognizing MUPS. These core values are generalist, personalized and continuous care for patients. Knowledge of the patient and his/her history and context made it easier to recognize them early in the consultation. Not only frequent and long consultations with frequent questions for referrals, but also subjective feelings such as irritability and frustration were clues for recognition. These findings are in line with the findings of previous studies (13â&#x20AC;&#x201C;17). Regarding the differentiation of patients in subgroups, we were able to distinguish five different subgroups of patients with MUPS: the anxious MUPS patient, the unhappy MUPS patient, the distressed MUPS patient, the passive or dependent MUPS patient and the puzzling MUPS patient. Although the subgroups showed overlapping features, they were based on the predominance of specific characteristics of patients. We did not see these five subgroups with their characteristics back in our quantitative risk assessment models.

Two main methodological considerations regarding this qualitative study are important. First, our results only give insight into perceptions of GPs and not into their actual behaviour. It is possible that GPs recognize individual patients with MUPS in clinical practice differently than how they believe they do. Second, we have not yet validated the subgroups. This has to be done in future studies. It is possible that patients do not recognize themselves in the subgroup that they are assigned to by the GP. This could result in a mismatch or struggle between the GP and patient during the consultation. Also because the subgroups show overlapping features this could contribute to the fact that patients do not recognize themselves in the assigned subgroup. When the subgroups are validated in both qualitative and quantitative research, they can support GPs and patients in providing guidance in management and the treatment can be more personalized and tailored to the patient.

208

General discussion

Structuring management of MUPS In general, MUPS management can be divided in different phases. After the first diagnostic phase follows the explanation phase of the consultation. Most patients need an explanation to effectively handle his or her symptoms. Unfortunately, most GPs are trained to specifically provide explanations for symptoms with an underlying cause. They are often not well equipped to explain symptoms without an underlying cause (18). In Scotland the Symptom Clinic Intervention (SCI) was developed, a communication intervention study in primary care where explanations for MUPS were the central element (19). The intervention was taught to Scottish GPs. The study results are presented elsewhere (20). In our qualitative study, we analysed the dialogues between the GP and the patient with MUPS and the responses of the patients in relation to these symptom explanations (Chapter 7). Based on these analyses, we described a range of dialogue types and patientsâ&#x20AC;&#x2122; responses and we presented a classification structure in our paper. This classification can be applied in teaching, evaluation of practice, and research after further validation. We found that deliberative dialogue types(21,22), dialogues with engagement between the GP and patient with both contributing ideas, were associated with acceptance of the explanation by the patient and that explanations were often not directly rejected. For future research, what further needs to be explored is which elements of the explanation are effective for which patients or for which symptoms and which are not.

Much research has been conducted about the effects of different MUPS interventions. Unfortunately intervention effects are often varying and disappointing. Several systematic reviews exist on different forms of MUPS interventions (23,24), but a complete overview for non-pharmacological interventions was missing. For that reason we conducted a Cochrane review, in which we investigated the effects of nonpharmacological interventions for somatoform disorders and chronic MUPS (Chapter 5). We concluded that only cognitive behavioural therapy (CBT) had a small positive effect on the severity of MUPS. When we summarised the main findings of the four currently existing Cochrane reviews about MUPS (i.e. (non-) pharmacological interventions, enhanced care and consultation letters (25â&#x20AC;&#x201C;27)), we again concluded that apart from CBT there is no effective evidence-based treatment available and

209

10

Chapter 10

specifically not in primary care (Chapter 6). Furthermore for CBT, there are many gaps in knowledge regarding the specifics of this treatment. This means that it is not yet clear what the ideal duration of the treatment is, who should provide the therapy and what the content should by all means include. In future research, these issues should be further explored. Also attention should be paid to the quality of the evidence as in all four Cochrane reviews, the quality of the studies was rated very low to moderate.

In order to incorporate the above-mentioned management elements into the concept of proactive, structured care or panel management and implications for clinical practice for patients with MUPS and those at risk for persistent MUPS, GPs should move away from the definition discussion in research literature and should more focus on the patient in the consultation room. First, either by a software reminder or by specific awareness, GPs should identify the patient with MUPS in an early stage in order to prevent persistent symptoms. Just like with every patient, but maybe more specifically for patients with MUPS, in the diagnostic phase GPs should attentively listen with an emphatic attitude to the story of the patient. They have to look for possible clues that arise during this phase and explore the five symptom dimensions according to the Dutch GP Guideline for MUPS and the multidisciplinary guideline for MUPS (28,29). These dimensions are the somatic, cognitive, emotional, behavioural and social dimension. Second, GPs should perform a targeted physical examination and provide a personalized explanation with the classification, when validated, of the different dialogue types and patientsâ&#x20AC;&#x2122; responses in mind. Third, as we know that the heterogeneous group of MUPS patients may be divided into subgroups, personalized and tailored interventions should be discussed with the patient, depending on the impact of the symptoms, possible other characteristics from the subgroups, and wishes of the patient. For example, with a patient with multiple MUPS and predominant anxiety, attention should be paid to effective reassurance(30). With an inactive patient with MUPS, for example with low back pain, the explanation could involve the vicious circle theory (i.e. low back pain leads to inactivity and inactivity leads to stiffness and therefore to more pain, etc.) and this patient should be activated. This was also underlined by van Koulil et al, who differentiated and treated two groups of fibromyalgia patients (31â&#x20AC;&#x201C;33). For a patient with physical complaints related to

210

General discussion

distress or burn-out, attention must be given to decrease the number of stressful activities and/or a reorganisation of the patients personal life (34). An additional advantage of this personalized subgroup approach is that effective strategies from other guidelines, such as the guidelines for depression, anxiety and chronic pain, can be used(35,36). Patients with severe MUPS or somatoform disorders can be referred to secondary health care for instance for CBT (37).

As the GP, compared to other specialists, has the advantage of knowledge about the patient and his or her context and history and also has a continuous and reliable doctor-patient relationship, the GP is eminently the professional who should coordinate care for patients with MUPS. This means that the GP invites the patient for consultations in a proactive, structured manner, whether or not reminded by EMR alerts and that the GP sees to it if the patient improves. Tasks can also be delegated to mental health practice nurses or psychosomatic physiotherapists. Together with the patient, the medical professionals should try to find the best following steps to improve, to prevent worsening and to prevent that the patient gets lost in the medical circuit.

Societal aspects of MUPS In the last part of this chapter, I will focus on the societal aspects of MUPS. These societal aspects become more and more important, because in the last decades costs related to healthcare use are steadily rising. Causes of these costs lie among other things in the consequences of MUPS. Therefore, in Chapter 8 and 9, we investigated the association between MUPS and two important societal aspects: healthcare use and work functioning, thereby highlighting the relevance of the problem of MUPS. We used data collected in the Netherlands Study of Depression and Anxiety (NESDA), a longitudinal cohort study investigating depression and anxiety. In the NESDA dataset, of which we could use three repeated measurements, there was information on MUPS and on the outcomes healthcare use and work functioning. Besides the association between MUPS and healthcare use and work functioning, we also investigated the role of influencing factors, such as depression, anxiety, personality characteristics and, specifically for work functioning, job characteristics. In our analyses, we found

211

10

Chapter 10

that MUPS were significantly and independently related to increased healthcare use and decreased work functioning. These results that are comparable with previously conducted studies. This underlines the importance of good MUPS management.

When interpreting the results, a few important methodological issues must be kept in mind. First, we used the number of consumed medical services and the number of contacts with the medical services as an indicator for healthcare use. However, no information was available about the reason for consultations. Therefore it is possible that the encounters were not related to MUPS but to other health complaints.

Also, we only examined the frequency of healthcare use but we cannot determine whether the healthcare consumption was adequate or inadequate (with inadequate defined as the â&#x20AC;&#x2DC;infiniteâ&#x20AC;&#x2122; search for physical explanations for the symptoms or reassurance). Second, we did not calculate the costs that accompanied high healthcare use or decreased work functioning. This has to be examined in future research, as the actual costs are important for health care policies. Third, in the NESDA study, the four dimensional symptom questionnaire (4DSQ) (38) that we used to measure MUPS, refers to complaints one week before the measurement, while the outcome variables were measured over a longer period of time before the measurement.

Therefore it is difficult to draw strong conclusions about causality. But the stability of the 4DSQ score over time was moderate to high (correlations between 0.7 and 0.8) and the sensitivity analyses that we performed showed comparable results.

Therefore we believe that the results of our analyses give sufficient reliable information about the influence of MUPS on health care use and work functioning. Fourth, the NESDA cohort is not representative for the Dutch general population, because patients with depressive and/or anxiety disorders are highly oversampled. This means that it is questionable whether the results of our analyses hold for the general population. To investigate that, we analysed the associations both in the patients with depressive and/or anxiety disorders and in the patients without those disorders.

Based on the results of those analyses, i.e. the relationships between MUPS and the societal outcomes, were stronger in the population without depressive and/or anxiety disorders, we believe that also in the general population MUPS are related to increased healthcare use and decreased work functioning. Specifically for the association between MUPS and work functioning, it was interesting to find the influence

212

General discussion

of job characteristics (i.e. job control and job support) on the association between MUPS and short-term absenteeism. After adjusting for job control and job support, the association between MUPS and short-term absenteeism became weaker. This suggests that improvement of working conditions may reduce the influence of MUPS on this unfavourable outcome. Future research can address this issue.

As an implication for clinical practice, we believe that it is worthwhile to pay attention to MUPS and their relation with work functioning, both by the GP during consultations (i.e. GPs should ask how work is going in the social dimension) and more directly by employers on the working floor. In addition, future research should be directed at developing preventive and treatment strategies, for example by assessing how optimization of job characteristics can reduce the influence of MUPS on work functioning. Health care insurances companies could play a role in this.

Conclusion In this thesis we have summarized our findings and elements that can be incorporated in proactive, structured care: identification of patients with MUPS (and those at risk for persistent symptoms), explanations and treatment of patients with MUPS. Also we paid attention to two important pillars of society: healthcare use and work functioning and how MUPS relate to them.

To conclude this thesis, I would like to add a personal experience from my own clinical practice about patients with MUPS. Even though I strongly believe that the GP is the best physician to manage MUPS patients and has the best position to structure care, I must sadly admit that in the beginning of my GP career a few years ago, I often found myself frustrated and irritated by this group of patients. I explored all symptom dimensions but following that I struggled with providing a decent explanation and with finding common ground with my patient. More than once my patient left the consultation room and I felt desperate and unsatisfied that both my patient and I were probably (very) unhappy. With time I learned that my feelings of irritation decreased when I got more experienced and I could add different tools to my tool-box. I learned how to apply different tools for different patients and in different situations. But the most important thing that helped me was a thing that one other

213

10

Chapter 10

doctor once had said during my traineeship when he also found me not amused. He told me the following: â&#x20AC;&#x153;Every patient has his own unique story. And there is no patient in the world that consults you with the purpose to annoy you or make you feel bad. Each patient just want your attentive listening, your empathic attitude to see life from his perspective and your devoted time to explain maybe the unexplainable. For you it might be the hundredth time.

But for him, it could be his first time. Do not forget that they have to live with their symptoms, without all the knowledge that you have. So replace irritation with a feeling or a thought of wonder. And then you can be the best physician to guide your patient with his MUPS. And I will reassure you that you will like, or even love it.â&#x20AC;? ...And after quite some years, I still totally agree with him.

214

General discussion

REFERENCES 1. Koopman RJ, Kochendorfer KM, Moore JL, et al. A diabetes dashboard and physician efficiency and accuracy in accessing data needed for high-quality diabetes care. Ann Fam Med. 2011;9:398–405.

2. Benhamou P-Y. Improving diabetes management with electronic health records and patients’ health records. Diabetes Metab. 2011;37:S53–6.

3. Bleijenberg N, Ten Dam VH, Drubbel I, et al. Development of a proactive care program (U-CARE) to preserve physical functioning of frail older people in primary care. J Nurs Scholarsh. 2013;45:230-7 4. Loo TS, Davis RB, Lipsitz LA, et al. Electronic medical record reminders and panel management to improve primary care of elderly patients. Arch Intern Med. 2011 26;171:1552–8.

5. Feldstein AC, Perrin NA, Unitan R, et al. Effect of a patient panel-support tool on care delivery. Am J Manag Care. 2010;16:e256–66.

6. Hak E, Hermens R, van Essen G, et al. Population-based prevention of influenza in Dutch general practice. Br J Gen Pr. 1997;363–6.

7. Kroenke K, Spitzer RL, Williams JBW. The PHQ-15: Validity of a New Measure for Evaluating the Severity of Somatic Symptoms. Psychosom Med. 2002;64:258–66.

8. Smith RC, Gardiner JC, Armatti S, et al. Screening for high utilizing somatizing patients using a prediction rule derived from the management information system of an HMO: a preliminary study. Med Care. 2001;39:968–78.

9. Morriss R, Lindson N, Coupland C, et al. Estimating the prevalence of medically unexplained symptoms from primary care records. Public Health. 2012;126:846–54.

10. Rosmalen JGM, Tak LM, de Jonge P. Empirical foundations for the diagnosis of somatization: implications for DSM-5. Psychol Med. 2011;41:1133–42.

11. Fink P, Rosendal M, Olesen F. Classification of somatization and functional somatic symptoms in primary care. Aust N Z J Psychiatry. 2005;39:772–81.

12. Diagnostic and statistical manual of mental disorders. Am Psychiatr Assoc. 2013; 13. Smith GR, Monson RA, Ray DC. Patients with multiple unexplained symptoms. Their characteristics, functional health, and health care utilization. Arch Intern Med. 1986;146:69–72.

14. Dirkzwager AJ, Verhaak PF. Patients with persistent medically unexplained symptoms in general practice: characteristics and quality of care. BMC Fam Pract. 2007;8:33.

15. McGorm K, Burton C, Weller D, et al. Patients repeatedly referred to secondary care with symptoms unexplained by organic disease: prevalence, characteristics and referral pattern. Fam Pract. 2010;27:479–86.

16. Salmon P, Ring A, Dowrick CF, Humphris GM. What do general practice patients want when they present medically unexplained symptoms, and why do their doctors feel pressurized? J Psychosom Res. 2005;59:255–60.

17. Salmon P. Patients who present physical symptoms in the absence of physical pathology: a challenge to existing models of doctor–patient interaction. Patient Educ Couns.

2000;39:105–13.

215

10

Chapter 10

18. Hartman TC olde, Hassink-Franke LJ, Lucassen PL, Spaendonck KP van, Weel C van. Explanation and relations. How do general practitioners deal with patients with persistent medically unexplained symptoms: a focus group study. BMC Fam Pract. 2009 24;10:68.

19. Burton C, Weller D, Marsden W, Worth A, Sharpe M. A primary care Symptoms Clinic for patients with medically unexplained symptoms: pilot randomised trial. BMJ Open. 2012 ; 9;2:e000513.

20. Morton L, Elliott A, Thomas R, et al. Developmental study of treatment fidelity, safety and acceptability of a Symptoms Clinic intervention delivered by General Practitioners to patients with multiple medically unexplained symptoms. J Psychosom Res. 2016;84:37–43.

21. Emanuel E, Emanuel L. Four Models of the Physician-Patient Relationship. JAMA. 267(16):2221–6.

22. Elwyn G, Lloyd A, May C, et al. Collaborative deliberation: a model for patient care.

Patient Educ Couns. 2014;97:158–64.

23. Burton C. Beyond somatisation: a review of the understanding and treatment of medically unexplained physical symptoms. Br J Gen Pract. 2003;53:231–9.

24. Kleinstäuber M, Witthöft M, Hiller W. Efficacy of short-term psychotherapy for multiple medically unexplained physical symptoms: a meta-analysis. Clin Psychol Rev. 2011;31:146–60.

25. Rosendal M, Blankenstein AH, Morriss R, et al. Enhanced care by generalists for functional somatic symptoms and disorders in primary care. Cochrane Database Syst Rev. 2013;10:CD008142.

26. Kleinstäuber M, Witthöft M, Steffanowski A, van Marwijk H, Hiller W, Lambert MJ. Pharmacological interventions for somatoform disorders in adults. Cochrane Database Syst Rev. 2014;11:CD010628.

27. Hoedeman R, Blankenstein AH, van der Feltz-Cornelis CM, Krol B, Stewart R, Groothoff JW. Consultation letters for medically unexplained physical symptoms in primary care.

Cochrane Database Syst Rev. 2010;(12):CD006524.

28. olde Hartman T, Blankenstein N, Molenaar B. NHG-Standaard Somatisch Onvoldoende verklaarde Lichamelijke Klachten (SOLK). Huisarts En Wet. 2013;5:222–30.

29. van der Feltz-Cornelis CM, Hoedeman R, Keuter EJW, Swinkels JA. Presentation of the multidisciplinary guideline medically unexplained physical symptoms and somatoform disorder in the Netherlands: Disease management according to risk profiles. J Psychosom Res. 2012;72:168–9.

30. Giroldi E, Veldhuijzen W, Mannaerts A, et al. “Doctor, please tell me it’s nothing serious”: An exploration of patients’ worrying and reassuring cognitions using stimulated recall interview. BMC Fam Pract. 2014;15.

31. van Koulil S, Kraaimaat FW, van Lankveld W, et al. Screening for pain-persistence and pain-avoidance patterns in fibromyalgia. Int J Behav Med. 2008;15:211–20.

32. van Koulil S, van Lankveld W, Kraaimaat FW, et al. Tailored cognitive-behavioral therapy for fibromyalgia: two case studies. Patient Educ Couns. 2008;71:308–14.

33. Turk DC, Okifuji A, Sinclair JD, et al. Differential responses by psychosocial subgroups of fibromyalgia syndrome patients to an interdisciplinary treatment. Arthritis Care Res Off J Arthritis Health Prof Assoc. 1998;11:397–404.

34. Bastiaanssen M, Terluin B, Loo M, et al. Landelijke eerstelijns samenwerkingsafspraak (LESA) Overspanning en burn-out. Huisarts Wet 2011;54:10–6.

216

General discussion

35. van Weel-Baumgarten E, van Gelderen M, Grundmeijer H, Licht-Strunk E. NHG Standaard Depressie. Huisarts Wet. 2012;55:252–9.

36. Hassink-Franke, Terluin B, van Heest F, Hekman J. NHG Standaard Angst. Huisarts Wet. 2012;55:68–77.

37. van Dessel N, den Boeft M, van der Wouden JC, Kleinstäuber M, Leone SS, Terluin B, et al. Non-pharmacological interventions for somatoform disorders and medically unexplained physical symptoms (MUPS) in adults. Cochrane Database Syst Rev. 2014 1;11:CD011142 38. Terluin B, van Marwijk HW, Adèr HJ, de Vet HC, Penninx BW, Hermens ML, et al. The Four-Dimensional Symptom Questionnaire (4DSQ): a validation study of a multidimensional self-report questionnaire to assess distress, depression, anxiety and somatization. BMC Psychiatry. 2006 22;6:34.

10

217

Summary

SUMMARY In Chapter 1 I introduced the subject of medically unexplained physical symptoms (MUPS) and how they put a burden on the patient, the GP, the GP-patient relationship and on the society. Also in this chapter I presented the rationale and outline of this thesis. I focused on the identification of MUPS, structuring management of MUPS and societal aspects of MUPS. In Chapter 2 we performed a validation study in which we explored the test characteristics of a screening method in primary care electronic medical records (EMRs) to identify patients with MUPS. This EMR method consisted of three steps: 1) including adult patients (â&#x2030;Ľ18) who had at least five general practitioner (GP) contacts in the past 12 months; 2) patients with known chronic (somatic) diseases were excluded; 3) patients were included who had a MUPS syndrome (irritable bowel syndrome, chronic fatigue syndrome, fibromyalgia) or who had at least three physical symptoms suggestive for MUPS. We compared the identified patients with MUPS from the screening method and those identified without MUPS with their scores on the patient health questionnaire-15. We found a high specificity but a low sensitivity, indicating that many potential MUPS patients will be missed. Therefore, before using this method as a screening method for selecting patients with MUPS, it needs to be improved.

In Chapter 3 we developed two risk assessment models to identify patients at risk for persistent MUPS with two statistical methods in a large primary care EMR database.

We operationalized MUPS as an International Classification of Primary Care (ICPC) code for irritable bowel syndrome, fibromyalgia, chronic fatigue syndrome and low back pain without radiation. The methods we used were a more classic logistic regression analysis and a more innovative decision tree analysis. We found that both models performed moderate to good with areas under the curve of 0.70 and 0.81, respectively. The validation showed acceptable stability (0.78 and 0.70, respectively).

While these models require further external validation and fine-tuning, they can

219

Summary

provide a starting point from which GPs could start evolving MUPS management in a proactive, structured manner together with their patients.

In Chapter 4 we conducted a focus group study with 29 Dutch GPs. We explored how GPs recognize patients with MUPS during their consultations and whether they could recognize subgroups of patients with MUPS. Two researchers independently analysed the data with the constant comparative analysis method. We found that GPs take various characteristics into account when recognizing patients with MUPS. More objective characteristics were multiple physical symptoms, frequent and long consultations and many referrals among other things. Subjective characteristics were negative feelings towards the patient and the feeling that they cannot make sense of the patientâ&#x20AC;&#x2122;s story. Based on the perceptions of the GPs and the predominance of certain characteristics, five subgroups of patients with MUPS were distinguished: the anxious MUPS patient, the unhappy MUPS patient, the passive MUPS patient, the distressed MUPS patient and the puzzling MUPS patient. While these subgroups need further validation, targeting them might improve personalized treatment. For example, an anxious MUPS patient could benefit from extra attention for reassurance and a passive MUPS patient could benefit from an activated treatment.

In Chapter 5 we assessed the effects of non-pharmacological interventions for somatoform disorders and chronic MUPS in adults in comparison with treatment as usual, waiting list controls, attention placebo, psychological placebo, enhanced care and other psychological or physical therapies within a Cochrane review. We searched different databases such as the Cochrane Depression, Anxiety and Neurosis Review Groupâ&#x20AC;&#x2122;s Specialise Register. We also performed several additional searches and we selected both randomized controlled trials (RCTs) and cluster RCTs. Four researchers conducted data extraction and risk of bias assessment. We pooled data from studies addressing the same comparison using standardised mean differences (SMDs) or risk ratios (RRs). The primary outcomes were severity of somatic symptoms and acceptability of the treatment. In total, we included 21 studies (n=2658 patients). All studies assessed the effectiveness of a psychological therapy. Most of them evaluated a form of cognitive behaviour therapy (CBT). We found that when all psychological therapies

220

Summary

were combined they were superior to usual care or waiting list controls in terms of reduction of symptom severity. However, effect sizes were small. Only CBT has been adequately studied which therefore allows a tentative conclusion that it has a small beneficial effect compared to usual care or waiting list controls. CBT was not more effective than enhanced care. The overall quality of evidence was rated low to moderate. Future studies should include various treatment modalities other than CBT, participants from different age groups and with different severity grades and long follow-up assessments. Also researchers should make efforts to blind outcome assessors.

In Chapter 6 we wrote about how to manage adult patients with persistent MUPS in the light of the existing uncertainties. First we summarized the content of the four Cochrane reviews that were published about MUPS until now, including the one we wrote regarding non-pharmacological interventions: one examining the effect of different types of pharmacological treatments, one examining the effect of enhanced care and one examining the effect of consultation letters. After examining the current state of evidence, we gave recommendations for future research and clinical practice. Regarding future research we recommended large, high quality RCTs with adult MUPS patients, whom have different levels of severity of MUPS. Both pharmacological and non-pharmacological interventions in different settings should be assessed and attention should be paid to the treatment characteristics such as the treatment duration, intensity, and dosage. Outcomes should include severity of symptoms and functional impairments among other things but also longer follow-up durations (minimum 6-12 months) are needed. Regarding clinical practice we advised different steps for GPs to undertake to manage patients with MUPS in a structured manner. Exploring all symptom dimensions, explaining MUPS in a constructive and empathic way and discussion of possible treatments personalized to the patientsâ&#x20AC;&#x2122; needs should be central.

In Chapter 7 we examined the dialogue between 39 patients with moderate MUPS and five GPs related to symptom explanations and we explored the patientsâ&#x20AC;&#x2122; responses.

With the constant comparative analysis method, we analysed 112 audio-recorded

221

Summary

consultations from two studies of the Symptoms Clinic Intervention, a moderate intensity consultation intervention for patients with MUPS in primary care. From the consultations we extracted 115 explanation sequences. We identified four dialogue types (i.e. lecture, storytelling, contest, deliberation), which differed in the extent to which the GP and/or patient controlled the dialogue. We identified eight patientsâ&#x20AC;&#x2122; responses, ranging from acceptance to rejection of the explanation by the patient.

From the results we developed a classification of dialogue types and patientsâ&#x20AC;&#x2122; responses. While it requires validation in future studies, it provides a framework of dialogue types and outcomes. This framework can be used for teaching, evaluation of practice and research.

In Chapter 8 we examined the association between MUPS and healthcare use (HCU), operationalized as the number of used medical services and number of health care contacts, over two years. Also we assessed the influence of depressive and anxiety disorders and personality traits on this association. We used data from the Netherlands Study of Depression and Anxiety (NESDA), a multisite cohort study and we included participants with current depressive and/or anxiety disorders, patients with subthreshold symptoms and healthy controls from different settings (n=2981).

Measurements were taken at baseline and at one and two year follow-up. We analysed the data with generalized estimating equations to take into account the dependency of observations within participants with repeated measurements. We found that MUPS were positively associated with HCU over two years. Neuroticism and depressive disorders had the strongest influence on this association, but the association between MUPS and HCU remained significant. This suggests that good MUPS management is important. Also attention should be paid to comorbid depressive disorders and to personality traits. In future research, the adequacy of HCU should be addressed, the initial reason for HCU and a calculation of associated costs should be made that comes with increased HCU. In Chapter 9 we examined the association between MUPS and work functioning, operationalized as disability at work and absenteeism from work, over two years. Again we used data from the NESDA. This this time we only included working participants

222

Summary

(n=1887). We used mixed model analyses to correct for the dependency of observations within participants. We found that MUPS were positively associated with disability and absenteeism over two years. After adjusting for depressive and anxiety disorders, the associations between MUPS and work functioning weakened, but remained significant. Again this suggests that good MUPS management is important.

Attention should be paid to comorbid mental health disorders when patients present themselves with MUPS. More insight is needed in favourable and unfavourable job characteristics to develop prevention and treatment interventions for MUPS. In Chapter 10 I presented an overall summary of our findings, interpreted the findings in the light of current evidence, discussed the most important methodological considerations and I gave recommendations for future research and clinical practice.

223

Samenvatting

SAMENVATTING Het onderwerp van dit proefschrift betreft somatisch onvoldoende verklaarde lichamelijke klachten (SOLK) in de huisartsenpraktijk. We spreken van SOLK als de lichamelijke klachten langer dan enkele weken duren en wanneer er bij adequaat medisch onderzoek geen aandoening is gevonden die de klachten voldoende verklaart. Voorbeelden van SOLK zijn misselijkheid, buikpijn, hoofdpijn en klachten van het bewegingsapparaat.

In het eerste hoofdstuk van dit proefschrift heb ik het onderwerp SOLK uitgebreider beschreven, alsmede welke (negatieve) consequenties SOLK kunnen hebben voor de patiënt, voor de huisarts en voor de relatie tussen huisarts en patiënt. Ook kunnen SOLK gevolgen hebben voor de maatschappij, met name door de hoge kosten die gepaard gaan met frequent doktersbezoek en indirect door verminderd functioneren op het werk of door gedeeltelijk of volledig ziekteverzuim.

In dit proefschrift richt ik me op drie thema’s, namelijk het identificeren van SOLKpatiënten (zowel in digitale zorggegevens als tijdens het consult) (H2, H3, H4), het structureren van de zorg voor SOLK-patiënten (H5, H6, H7) en tenslotte de maatschappelijke aspecten gerelateerd aan SOLK (H8, H9).

In hoofdstuk 2 beschrijf ik een validatiestudie waarin we de testkarakteristieken van een screeningsmethode onderzochten die SOLK-patiënten identificeert in (anonieme) digitale zorggegevens uit de huisartsenpraktijk. De screeningsmethode bestond uit een drietal stappen: 1) alle volwassen patiënten werden geïncludeerd die de huisarts minimaal vijf keer bezochten in het afgelopen jaar; 2) patiënten met bekende chronische ziekten werden geëxcludeerd; 3) alle patiënten met een SOLK syndroom (prikkelbare-darmsyndroom (PDS), chronisch vermoeidheidssyndroom (CVS) en fibromyalgie (FM)) werden geïncludeerd EN patiënten met minimaal drie lichamelijke klachten suggestief voor SOLK. We vergeleken de geïdentificeerde patiënten met en zonder SOLK met hun score op een vragenlijst, de Patient Health Questionnaire-15.

We vonden een hoge specificiteit, maar een lage sensitiviteit. Dit betekent dat veel

225

Samenvatting

potentiële SOLK-patiënten gemist worden. Daarom moet deze screeningsmethode eerst verbeterd worden voordat deze in de praktijk kan worden gebruikt.

In het derde hoofdstuk beschrijven we twee modellen om patiënten te identificeren die risico lopen om persisterende SOLK te ontwikkelen. We gebruikten hiervoor twee statistische methoden (logistische regressie-analyse en een beslisboomanalyse) die we toepasten op een grote database met digitale zorggegevens van huisartsen. We operationaliseerden SOLK met een digitale code voor PDS, CVS, FM en lage rugpijn zonder uitstraling. Beide modellen gaven goede testkarakteristieken (oppervlakte onder de curve van 70% en 81%) die stabiel bleven na validatie. De modellen vereisen nog enige verfijning, maar kunnen een startpunt bieden voor proactieve en gestructureerde zorg voor SOLK-patiënten door hun huisartsen.

In hoofdstuk 4 beschrijven we onze focusgroepstudie onder 29 huisartsen. We onderzochten hoe zij SOLK herkennen bij hun patiënten tijdens consulten en of ze subgroepen van SOLK-patiënten kunnen onderscheiden. De huisartsen betrekken verschillende kenmerken bij de herkenning van SOLK. Objectieve kenmerken zijn meerdere klachten, frequente en lange consulten en veel verwijzingen naar de tweede lijn. Meer subjectieve kenmerken zijn negatieve gevoelens jegens de patiënt en het gevoel dat de huisarts niet wijs wordt uit het verhaal dat de patiënt vertelt.

Gebaseerd op de percepties van de huisartsen en het voorop staan van bepaalde kenmerken kwamen we tot vijf subgroepen: de angstige SOLK-patiënt, de ongelukkige SOLK-patiënt, de gespannen SOLK-patiënt, de passieve SOLK-patiënt en de puzzelende SOLK-patiënt. Deze subgroepen behoeven nog verdere validatie. Na bevestiging van deze modellen kunnen zij een bijdrage leveren aan gepersonaliseerde zorg voor SOLK-patiënten. Ter illustratie zou een angstige SOLK-patiënt baat kunnen hebben bij extra aandacht voor de uitleg en geruststelling terwijl een passieve SOLKpatiënt baat zou kunnen hebben bij een activerend beleid.

In hoofdstuk 5 vatten we de Cochrane review samen naar de effecten van nietfarmacologische interventies voor somatoforme stoornissen en chronische SOLK bij volwassenen in vergelijking met de reguliere behandeling, wachtlijstcontroles,

226

Samenvatting

placebo en andere psychologische of fysieke therapieën. We zochten naar (cluster) gerandomiseerde gecontroleerde studies in verschillende literatuurbestanden. De primaire uitkomst was de ernst van de klachten. In totaal includeerden we 21 studies met 2658 patiënten. Alle studies onderzochten een vorm van psychologische behandeling. De meesten evalueerden een vorm van cognitieve gedragstherapie (CGT). We vonden dat wanneer we alle studies combineerden deze enigszins in het voordeel waren ten opzichte van reguliere zorg of wachtlijstcontroles met betrekking de ernst van de klachten. Qua behandeling werd alleen CGT voldoende onderzocht waardoor we de voorzichtige conclusie kunnen trekken dat CGT ook een voordelig effect heeft op de ernst van de klachten ten opzichte van de reguliere zorg en wachtlijstcontroles.

De algehele kwaliteit van de studies was laag tot matig.

In hoofdstuk 6 beschrijven we een klinische les over de behandeling van patiënten met persisterende SOLK in het licht van alle bestaande onzekerheden. Hiervoor hebben we allereerst de vier Cochrane reviews samengevat die tot nu toe over SOLK zijn verschenen. We beschrijven de huidige stand van zaken van het beschikbare bewijs en doen aanbevelingen voor de klinische praktijk en toekomstig onderzoek.

Wij adviseren dat toekomstige studies patiënten moeten includeren met SOLK in verschillende maten van ernst. Ook dient de opzet van de studies te verbeteren. Zowel farmacologische als niet-farmacologische interventies in verschillende settingen (huisarts, specialist) moeten worden onderzocht waarbij aandacht moet worden besteed aan de kenmerken van de interventies zoals de duur, de specifieke inhoud, de dosering en aan de intensiteit. Ook is een langere follow-up nodig. Voor de huisarts adviseren we verschillende stappen om de zorg voor SOLK-patiënten te structureren.

Het exploreren van alle symptoomdimensies, het uitleggen van SOLK op een constructieve en empathische wijze en het bespreken van mogelijke, persoonsgerichte behandelingen met de patiënt zijn hier centrale elementen in.

In hoofdstuk 7 presenteren we de resultaten van een analyse van gesprekken tussen 39 Schotse patiënten met matig-ernstige SOLK en vijf Schotse huisartsen, specifiek gericht op de uitleg van SOLK gegeven door de huisarts en de reactie hierop van de patiënt. Met de zogenaamde constant vergelijkende analysemethode analyseerden

227

Samenvatting

wij 112 op audiotape opgenomen consulten van twee Schotse studies: the Symptom Clinic Intervention, een communicatie-interventie voor SOLK-patiënten in de eerste lijn. Hieruit extraheerden wij 115 passages waarin de uitleg van SOLK centraal stond.

We identificeerden vier verschillende dialoogtypen, uiteenlopend van een monoloog door de huisarts tot een constructieve conversatie waaraan huisarts en patiënt een gelijke bijdrage leverden. We identificeerden acht patiëntreacties, variërend van acceptatie tot verwerping door de patiënt van de gegeven uitleg. We ontwikkelden een classificatie met dialoogtypes en patiëntreacties. Deze classificatie vereist verder onderzoek, maar kan dan een basis bieden die gebruikt kan worden in onderzoek, opleiding en evaluatie van de praktijk als het gaat om communicatie over SOLK. In hoofdstuk 8 beschrijven we de analyse van het verband tussen SOLK en gezondheidszorggebruik over twee jaar. SOLK werd vastgesteld met een vragenlijst, namelijk de VierDimensionale Klachtenlijst. Gezondheidszorggebruik drukten we uit als het aantal medische disciplines waarmee de patiënt contact had en het totaal aantal contacten hiermee. Daarnaast onderzochten we de invloed van depressieve en/of angststoornissen en persoonlijkheidstrekken op dit verband. We gebruikten data van de Netherlands Study of Depression and Anxiety (NESDA), een cohortonderzoek met deelnemers uit verschillende settingen. Het ging om 2981 deelnemers met een huidige depressieve en/of angststoornis, mensen met een verhoogd risico op zo’n stoornis en een controlegroep van gezonde deelnemers. We vonden dat SOLK onafhankelijk en positief geassocieerd waren met gezondheidszorggebruik. De kenmerken die het meest van invloed waren op het verband tussen SOLK en gezondheidszorggebruik over twee jaar tijd waren neuroticisme en depressieve stoornissen. Deze resultaten suggereren dat er naast goede zorg voor SOLK-patiënten aandacht dient te zijn voor signalen van een depressieve stoornis en neuroticisme. In toekomstig onderzoek moet vooral gekeken worden of er bij een verhoogd zorggebruik sprake is van adequate of inadequate zorg en of de zorg al dan niet versnipperd is. Daarnaast is inzicht nodig in de kosten die gepaard gaan met het bezoeken van verschillende disciplines en verhoogd zorggebruik.

228

Samenvatting

In hoofdstuk 9 beschrijven we het verband tussen SOLK enerzijds en verminderd functioneren op het werk en ziekteverzuim anderzijds over twee jaar tijd. Opnieuw gebruikten we NESDA data. We analyseerden de gegevens van 1887 werkende deelnemers. We vonden dat SOLK onafhankelijk en positief geassocieerd waren met beide werkuitkomsten. Na correctie van depressieve en angststoornissen werd het verband iets zwakker, maar bleef significant. Deze resultaten ondersteunen waarom goede zorg voor SOLK belangrijk is en dat er daarnaast aandacht moet zijn voor signalen van angst en depressie. Meer inzicht is nodig in gunstige en nadelige werkkenmerken om zowel preventieve als therapeutische interventies te ontwikkelen.

Ook moet er meer bewustwording zijn voor SOLK op de werkvloer, niet alleen onder artsen maar ook onder werkgevers en collega-werknemers gezien de invloed die SOLK kunnen hebben.

In het laatste hoofdstuk vat ik alle resultaten samen. Daarbij bespreek ik de belangrijkste methodologische aspecten en doe ik aanbevelingen voor toekomstig onderzoek en de klinische praktijk. Ik sluit het hoofdstuk af met een persoonlijke noot vanuit mijn eigen ervaring als huisarts met SOLK-patiĂŤnten.

229

Dankwoord

DANKWOORD Ik kan wel stellen dat een promotie niet iets is wat helemaal vanzelf gaat. Mijn vader zou waarschijnlijk zoiets zeggen als ‘dat lijkt me wel iets vrij logisch’. En mijn moeder ‘de mooiste dingen in het leven kosten nu eenmaal tijd’. Ze hebben allebei gelijk. Ik denk, wat ik overigens niet volgens de regels der kunsten evidence based heb opgezocht, dat zeker 95 procent van de promovendi ergens in het traject denkt ‘help, waar ben ik aan begonnen!’ (en waarbij dan de andere vijf procent al is afgehaakt).

En wat betreft het schrijven van systematische reviews, suggereerde een Engelsman mij ooit: ‘they should give the researchers a room in the basement of the faculty, behind bars, for their own safety.’ Tja… Goed, zo erg was het bij mij niet. Alleen op sommige momenten kwamen de zen-achtige yoga-oefeningen zeker van pas.

Het was niet alleen afzien. Integendeel. Een promotie is een mooie reis. Een reis waar je de mogelijkheid krijgt jezelf te ontwikkelen. Een brug te slaan tussen de huisartsenpraktijk en de wetenschap. Waar je op je pad inspirerende en gedenkwaardige mensen ontmoet. En last but not least, waarbij je jezelf tegenkomt en veel over jezelf leert. Daarom kan ik terugkijkend zeggen: ik zou het opnieuw doen, met dezelfde mensen. Zonder hen was dit niet mogelijk geweest. Dus daar gaan we!

Promotoren, co-promotor en begeleiders Prof. Dr. Henriëtte van der Horst, beste Henriëtte, dank voor de kans die je me hebt gegeven om bij de afdeling Huisartsgeneeskunde en Ouderengeneeskunde te mogen promoveren. Naast de inhoudgerelateerde zaken vond ik het onder andere erg leuk om samen met jou en Annet Sollie naar Canada te vliegen voor de NAPCRG en in het vliegtuig synchroon film te kijken. Daarnaast hebben we in toenemende mate contact gehad over mijn artikelen. Aan jouw ‘close-reading’ heb ik ontzettend veel gehad. Jij ziet dingen waaraan ik niet dacht en die mijn kijk op het onderwerp hebben verbreed. Jouw prachtige Engelse volzinnen verdienen een aparte vermelding, bij deze. Veel dank voor je betrokkenheid, niet alleen voor mijn promotie, maar ook voor mij als persoon, zeker als het even niet makkelijk was. Dank je wel.

231

Dankwoord

Prof. Dr. Mattijs Numans, beste Mattijs, jij ook bedankt voor deze kans. Het viel niet mee om het project op te starten en gedurende de reis hebben we heel wat tegenslagen samen overwonnen. Je bent een inspirator en jouw enthousiasme heeft mij ondersteund op de juiste momenten. Daarom kijk ik met plezier en voldoening terug op onze samenwerking.

Dr. Hans van der Wouden, beste Hans, mijn wetenschappelijk rots in de branding.

Nog voor ik promovenda werd, hadden wij al samen een CAT geschreven voor Huisarts & Wetenschap. Als derdejaars huisarts in opleiding wilde ik kennismaken met de wetenschap, al had ik er weinig kaas van gegeten. Jij was bereid mij te helpen.

Zie hier de start van mijn onderzoekscarrière. Alle promovendi zijn het eens, jij bent de stabiele factor op de afdeling en jouw deur staat altijd open. Je hebt wel wat gemopper van me moeten aanhoren in de loop der tijd, maar gelukkig mocht ik altijd terug komen en hebben we veel kunnen lachen. We did it! Veel dank!

Prof. Dr. Jos Twisk, beste Jos, ik herinner me nog het eerste wat je tegen me zei over mijn onderzoek. Midden in het voorstelrondje van de multilevelcursus zei jij blij grijnzend en plein public ‘jij bent toch diegene wiens trial niet gelukt is?’ Lekker dan, dacht ik, met rood aangelopen wangen. Gelukkig was ik de enige die me er druk over maakte. En in de pauze zei je dat als je ergens bij kon helpen, je dat graag wilde doen. Zo geschiedde. Wat heb ik je vaak benaderd met statistische vragen. Ik heb bewondering voor jou als docent, hoe je ingewikkelde dingen goed begrijpelijk kunt uitleggen (zelfs aan mij!). Ik bewonder je als mens, met je vrolijke aanwezigheid en je scherpe blik. Dank voor je aanmoedigende woorden, je vertrouwen in mij en al je geduld tijdens mijn promotie en niet te vergeten de master Epidemiologie. Bedankt dat je er was en dat je er bent.

Leescommissie en opponenten Prof. Dr. Henk de Vries, Prof. Dr. Sandra van Dulmen, Prof. Dr. Bert van Hemert, Dr. Peter van de Ven, Dr. Frans Smits, Prof. Dr. Judith Rosmalen, hierbij wil ik jullie allemaal hartelijk bedanken dat jullie mijn manuscript hebben willen beoordelen en voor de vriendelijke woorden daarover.

232

Dankwoord

Co-auteurs Beste co-auteurs, dank voor jullie bijdrage aan alle artikelen. In het bijzonder: Tim en Peter, wat vind ik het toch gezellig dat ik nu bij jullie werk en dat we alle ideeën omtrent SOLK, huisartsgeneeskunde en wetenschap kunnen delen. Chris, thank you for the inspiring time in Aberdeen. And LaKrista, you brought the sun with you during Scottish rainy days.

Afdeling D5 Wat een gave gang. Ik zou iedereen zo’n gang met zulke collega’s aanbevelen. Bedankt allen, en in het bijzonder Annemarie, Floor, Hanneke, Karolien, Anne, Lidy, Marloes, Sandra, Joreintje, Pim, Lishia en Wesley, voor de fijne tijd. Ik mis jullie!

Nog meer in het bijzonder: Lieve Daniëlle van der Laan, jij bent inderdaad the best roomie ever (zoals je zelf ook altijd zo mooi bescheiden over jezelf zegt). Wat hebben we gelachen, de kamer opgefleurd, gekletst, geklaagd, gezongen, en nog meer gelachen. Je bent een topper! En wat je hierna ook gaat doen, je gaat uitblinken!

Lieve Daniëlle Huisman, wat hebben Hans en ik toch een goede keuze gemaakt om jou als onderzoeksassistente aan te nemen. Je ben van onschatbare waarde geweest.

De kwalitatieve projecten hebben ons wel wat zorgen gegeven. En soms moesten we zoeken naar een middenweg in benadering door onze verschillende kijk op zaken.

Maar desondanks, of wellicht juist dankzij, hebben we het mooi afgerond. Ik kijk ook met veel plezier terug op onze treinreizen naar Nijmegen en de werkdag op het Zandvoortse strand vorig jaar zomer. We komen elkaar vast nog weer eens tegen. En indien werk-gerelateerd zeg ik: laten we van Zandvoort een traditie maken. Succes met alles en veel dank.

Lieve Nikki en Kate, mijn SOLK partners-in-crime. Ik denk dat we goed bezig zijn om een steentje bij te dragen aan de kennis over SOLK, in ieder geval binnen het VUmc. Ik kan niet anders zeggen dat ik enorm op jullie gesteld ben, veel met jullie heb gelachen en heb gedeeld, en hoop dat onze vriendschap zich ook buiten het werk

233

Dankwoord

blijft voortzetten. Nikki, voor ons is dat makkelijk. Als buuf ben je namelijk nooit heel ver weg.

Mijn paranimfen Vincent Polak, lieve Vin, als er iets is wat ik iedere vrouw kan aanraden: heb een mannelijke beste vriend. Want die heb ik. Vanaf de brugklas zijn we vrienden en ik denk dat we alle mijlpalen samen hebben meegemaakt. Ik ben blij en dankbaar je in mijn leven te hebben, samen met Deline en kleine Reyn. Trots en gelukkig ben ik dat je naast mij staat tijdens mijn verdediging. Want dan komt het sowieso goed. Hoe vrouwelijk ik mij soms kan gedragen, hoe zeer jouw nuchtere kijk en slappe humor mij vervolgens weer verder helpt. Ik weet dat we vrienden blijven tot we oud en grijs zijn.

Dr. Chantal Gielen, Lieve Tal, een eind van een tijdperk! Onze promoties zijn af! Wat ons onder andere kenmerkt is de wijze waarop ik nu dit dankwoord aan het schrijven ben: achter mijn laptop, in een cafeetje, met jou naast me met een kop thee. Heel erg ‘Sex and the City’. We kennen elkaar sinds onze studie waar we vrolijk doorheen zijn gewandeld. Natuurlijk waren we ook serieus, want zie waar we nu gekomen zijn.

Misschien op andere plekken dan we toen hadden bedacht. Maar net zo mooi en net zo goed. Je bent scherp, je bent pittig, je bent stoer maar bovenal lief. Zoals je weet hou ik van quotes en deze passen bij ons: ‘behind every succesful woman is a best friend giving her crazy ideas’. Of nog beter: ‘a brunette and a blond with an unbreakable bond’.

Mijn lieve vriendinnen en vrienden Mijn allerliefste vriendinnen: Caartje, Sappie, Jacomien, Sanne, Luus, Marieke en Kim. Nog een SATC quote: ‘They say nothing lasts forever, dreams change, trends come and go, but true friendship never goes out of style’. En zo is het. En zo zal het altijd zijn. Natuurlijk ook mijn andere vrienden waaronder Moniek, Matthijs, Thorvald, Annet en Bibian: veel dank dat jullie in mijn leven zijn.

Lieve Peter, de laatste maanden van deze periode heb jij naast mij gestaan. Je blijheid, je gulle lach, je onuitputtelijke optimisme en je vertrouwen hebben me enorm

234

Dankwoord

gesteund. Niets is te gek en alles kan en mag. Bedankt dat je er bent. â&#x20AC;&#x2DC;Add a little confetti to each dayâ&#x20AC;&#x2122;, zo valt het leven te omschrijven met jou erin. Het is een feestje.

Tot slot, mijn allerliefste papa, mama en Han, bedankt dat jullie er altijd zijn. Dat ik altijd, hoe dan ook, in welke situatie dan ook, en bij alles wat ik doe of laat, een thuis bij jullie vind. Jullie hadden gelijk, dat het goed is dat het niet vanzelf gaat en dat alle mooie dingen tijd kosten. Het is gelukt. Dankzij jullie. PS, I love you.

235

About the author

ABOUT THE AUTHOR Madelon den Boeft was born on November 23rd, 1983 in Apeldoorn, the Netherlands, as daughter of Georgy de Mey and Onno den Boeft. After graduating secondary school at the Bonaventura College in Leiden (Gymnasium, 2002) she started studying Medicine at Leiden University.

During her study Medicine she became a board member of the International Federation of Medical Students’ Associations (IFMSA), first as Local Public Health Officer and later as National Public Health Officer. In her functions as board member she developed and organised multiple internships for medical students to developmental countries, organised a national conference for medical students and developed and participated in several local public health projects.

Furthermore she worked as editor for Global Medicine, the bulletin of IFMSA, and took part in a clinical internship in Mangochi, Malawi (2006).

After graduating her Medicine study in 2009, she started working as a resident Pulmonary Medicine at the Spaarne Hospital in Hoofddorp. After a year she decided to become a general practitioner (GP) and started her GP training at VU university medical center (VUmc). During this training she became a board member ‘sponsoring’ for the National Organization for GPs in Training (in Dutch: LOVAH) and editor in chief of the LOVAH bulletin. After graduating her GP training she started working as a GP in Noord-Holland. At the same time she started her PhD project at the Department of General Practice and Elderly Care Medicine, VUmc, named ‘Medically unexplained physical symptoms in primary care; identification, structuring management and societal aspects. During her PhD, she followed the post-initial master Epidemiology at EpidM VUmc, which she completed in March 2016. Also she teached and supervised medical students and interns. In January 2016 she became a postdoctoral researcher at Radboud university medical center in Nijmegen. Currently she works with Tim olde Hartman at the project ‘Empathy and personalized health care’.

Madelon den Boeft lives in Haarlem. In her free time she writes stories for children, rides (western) horses, practices yoga and participates in boxing classes.

237

PhD portfolio

PHD PORTFOLIO

Master of science in Epidemiology, VU university medical center 2014-2016 - V10: Epidemiological research: design and interpretation - V20: Principles of epidemiological data analysis - V30: Regression analysis - WC40: Clinimetrics - V50: Epidemiology in practice: how to design a study - V60: Epidemiology of diseases - K71: Systematic reviews and meta-analysis - K73: Longitudinal data analysis - K74: Multilevel data analysis - K78: Qualitative analysis - WC 80: Clinical prediction models - Internship (Practical period; article M. den Boeft, JWR. Twisk, JC. van der Wouden, B. Penninx, B. Terluin, ME. Numans, HE. van der Horst. Medically unexplained physical symptoms and work functioning over two years: their association and the influence of depressive and anxiety disorders and job characteristics. Supervisor: Prof. Dr. JWR. Twisk. BMC Fam Pract. 2016;17:46.)

National and international conferences - 2015 Symptom Research in Primary Care, Vejle, Denmark. Responses of patients in relation to constructive explanations for medically unexplained physical symptoms. A qualitative study of audiotaped consultations (presentation) - 2015 Symptom Research in Primary Care, Vejle, Denmark. Recognition of patients with medically unexplained physical symptoms by family physicians: results of a focus group study (poster) - 2014 Annual Meeting North American Primary Care Research Group, New York, USA. The association between medically unexplained physical symptoms and health care use and the influence of de-

239

PhD portfolio

pressive and anxiety disorders and personality traits: a longitudinal study (poster) - 2014 Annual Meeting North American Primary Care Research Group, New York, USA. Recognition of patient profiles in medically unexplained physical symptoms: views of family physicians (poster) - 2013 NHG Wetenschapsdag, Leiden. Identifying patients with medically unexplained physical symptoms in primary care electronic medical records, a validation study (presentation) - 2013 WONCA 2013, Prague, Czechia. Identifying patients at risk for chronic medically unexplained physical symptoms in primary care electronic medical records with data mining techniques (presentation) - 2013 Annual Meeting North American Primary Care Research Group 2013, Ottawa, Canada. Identifying patients at risk for chronic medically unexplained physical symptoms in primary care electronic medical records with data mining techniques (poster) Grants

ÂŁ5000, National Institute for Health Research Funding for Cochrane review proposal: Non pharmacological interventions for somatoform disorders and medically unexplained physical symptoms in adults. Published: Cochrane Database Syst Rev. 2014;11:CD011142

Foreign cooperation

Visiting researcher University of Aberdeen (May-June 2014). Article: Negotiating explanations: a qualitative analysis of doctor-patient communication in a general practice clinic for patients with medically unexplained physical symptoms. Supervisor: CD. Burton. Submitted.

240

List of publications

LIST OF PUBLICATIONS

International publications M. den Boeft, JC. van der Wouden, TR. Rydell-Lexmond, NJ. de Wit, HE. van der Horst, ME. Numans. Identifying patients with medically unexplained physical symptoms in electronic medical records in primary care: a validation study. BMC Fam Pract.

2014;15:109 N. van Dessel, M. den Boeft, JC. van der Wouden, M. Kleinsteuber, SS. Leone, B. Terluin, HE. van der Horst, ME. Numans, HMJ. van Marwijk. Non-pharmacological interventions for somatoform disorders and medically unexplained physical symptoms in adults. Cochrane Database Syst Rev. 2014;11:CD011142 M. den Boeft, JWR. Twisk, JC. van der Wouden, HMJ. van Marwijk, B. Penninx, B. Terluin, ME. Numans, HE. van der Horst. The association between medically unexplained physical symptoms and health care use over two years and the influence of depressive and anxiety disorders and personality traits: a longitudinal study. BMC Health Serv Res. 2016;16:100 M. den Boeft, JWR. Twisk, JC. van der Wouden, B. Penninx, B. Terluin, ME. Numans, HE. van der Horst. Medically unexplained physical symptoms and work functioning over two years: their association and the influence of depressive and anxiety disorders and job characteristics. BMC Fam Pract. 2016;17:46.

M. den Boeft, D. Huisman, JC. van der Wouden, ME. Numans, HE. van der Horst, PL. Lucassen, TC. olde Hartman. Recognition of patients with medically unexplained physical symptoms by family physicians: results of a focus group study. Accepted by BMC Fam Pract. May 2016 Submitted for publication M. den Boeft, M. Hoogendoorn, JWR. Twisk, S. Nap, T. van der Neut, JC. van der Wouden, HE. van der Horst, ME. Numans. Risk assessment models for patients with

241

List of publications

persistent medically unexplained physical symptoms in primary care using electronic medical records. An observational study.

M. den Boeft, N. van Dessel, JC. van der Wouden. How should we manage adults with persistent unexplained physical symptoms? (revised) M. den Boeft, D. Huisman, L.M. Morton, P.L. Lucassen, J.C. van der Wouden, M.J. Westerman, H.E. van der Horst, C.D. Burton. Negotiating explanations: a qualitative analysis of doctor-patient communication in a general practice clinic for patients with medically unexplained physical symptoms.

Additional national publications M. den Boeft. Cognitieve gedragstherapie lijkt effectief bij fibromyalgie. Huisarts Wet. 2014-4 M. den Boeft. Smartphone-interventie bij chronische pijn. Huisarts Wet. 2013-9 M. den Boeft, JC van der Wouden. IJzersuppletie bij moeheid en een laag ferritine.

Huisarts Wet. 2012-9

242

Medically unexplained physical symptoms in primary care Identification, management and societal aspects

Uitnodiging

Medically unexplained physical symptoms in primary care Identification, management and societal aspects

voor het bijwonen van de openbare verdediging van mijn proefschrift getiteld

Medically unexplained physical symptoms in primary care Identification, management and societal aspects

Donderdag 8 september 2016 om 11.45 in het Auditorium Vrije Universiteit De Boelelaan 1105, Amsterdam Na afloop bent u zeer welkom voor een receptie in de receptieruimte Boelelaanzijde (HG-1, C35A) Hopelijk tot dan!

Madelon den Boeft m_denboeft@yahoo.com 06-28402935

Madelon den Boeft

Madelon den Boeft

Paranimf: Vincent Polak v.g.polak@gmail.com 06-52121322 Paranimf: Chantal Gielen c.l.i.gielen@lumc.nl 06-28126400