Proefschrift Wierenga by Nicole Nijhuis

BRAVO@Work

BRAVO@Work The evaluation of the implementation process of a worksite health promotion program

Uitnodiging Voor het bijwonen van de openbare verdediging van mijn proefschrift

BRAVO@Work The evaluation of the implementation proces of a worksite health promotion program Op donderdag 31 maart 2016 om 13.45 uur in de Aula van de Vrije Universiteit aan de Boelelaan 1105 te Amsterdam Na afloop bent u van harte welkom op de receptie

Debbie Wierenga Fluitekruidweg 116 1508AM Zaandam 0655738333 debbie_wierenga@hotmail.com debbie.wierenga@arboned.nl

Debbie Wierenga

Paranimfen â&#x20AC;˘ Ellen Breed

Debbie Wierenga

0643246193 e.e.breed@gmail.com

â&#x20AC;˘ Linda Koopmans

0648032329 lindataylorkoopmans@outlook.com

Body@Work

BRAVO@Work The evaluation of the implementation process of a worksite health promotion program

Debbie Wierenga

The study presented in this thesis was conducted at the Netherlands Organization for Applied Scientific Research TNO, Leiden, the Netherlands. TNO participates in Body@Work, Research Center on Physical Activity, Work and Health, which is a joint initiative of VU University Medical Center (Department of Public and Occupational Health, EMGO+ institute for Health and Care Research), VU University Amsterdam, and TNO. The study was funded by the Netherlands Organization for Health Research Development (ZonMW), project 50-51405-98-019. Financial support for the printing of this thesis has kindly been provided by Body@Work, Research Center on Physical Activity, Work and Health, and the VU University.

English title:

BRAVO@Work: the evaluation of the implementation process of a worksite health promotion program

Nederlandse titel:

BRAVO@Work: de evaluatie van het implementatieproces van een gezondheidsbevorderende interventie op de werkplek

ISBN:

978-94-6233-235-5

Cover design by:

DivingDuck Design (www.divingduckdesign.nl)

Layout and printed by:

Gildeprint, Enschede, the Netherlands.

ÂŠ Copyright 2016, Debbie Wierenga, the Netherlands All rights reserved. No part of this thesis may be reproduced or transmitted in any form of by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without prior written permission from the author, or when appropriate, from the publishers of the papers.

VRIJE UNIVERSITEIT

BRAVO@Work The evaluation of the implementation process of a worksite health promotion program

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 31 maart 2016 om 13.45 uur in de aula van de universiteit, De Boelelaan 1105

door Debbie Wierenga geboren te Zaandam

promotoren:

prof.dr. W. van Mechelen

copromotoren:

dr. L.H. Engbers

dr. P. van Empelen

Contents Chapter 1

General introduction

Chapter 2

The design of a real-time formative evaluation of the

implementation process of lifestyle interventions at two worksites

using a 7-step strategy (BRAVO@Work)

Chapter 3

What is actually measured in process evaluations for worksite

health promotion programs: a systematic review

Chapter 4

The implementation of multiple lifestyle interventions in two

organizations: a process evaluation

Chapter 5

What is the relationship between implementation factors and

employee participation in and satisfaction with a worksite health

promotion program?

111

Chapter 6

A 7-step strategy for the implementation of worksite lifestyle

interventions: helpful or not?

131

Chapter 7

General discussion

151

Summary

175

Samenvatting

179

Dankwoord

183

About the author

187

General Introduction

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

General Introduction

The risks/prevalence of an unhealthy lifestyle Insufficient physical activity, smoking, high levels of alcohol consumption, an unhealthy diet and low levels of relaxation characterize an unhealthy lifestyle. Over the past decades we have learned that an unhealthy lifestyle is associated with chronic diseases, like cardiovascular diseases, type-2 diabetes and certain types of cancer [1, 2]. In addition physical inactivity (decreased energy expenditure) and unhealthy nutrition (increased energy intake) lead to a positive energy balance, which leads to overweight and obesity in the long-run [1, 2]. This is unfortunate, since we can prevent those lifestyle-related conditions by adopting and maintaining a healthy lifestyle [1, 2]. Nonetheless, the prevalence of an unhealthy lifestyle among the Dutch population is high. This is also true for overweight and obesity. In 2014, 49% of the Dutch working population (aged 20 to 65 years) was overweight (Body Mass Index [BMI] â&#x2030;Ľ 25 kg m-2), of which 13% was obese (BMI>30 kg m-2)[3]. In 2013, 58% of the Dutch adult population aged 18 years of older met the guideline for moderate intensity physical activity. This guideline states that adults should engage in physical activity of at least moderate intensity (4.5 to 6 MET; MET stands for metabolic equivalent and expresses the intensity of physical activity in terms of oxygen consumption per kg body weight per minute where 1 MET equals 3,5 ml oxygen per kg body weight per minute)for at least 30 minutes a day on at least five days a week. Additionally, 21% of the Dutch adult population met the guideline for high intensity (â&#x2030;Ľ6.5 MET) physical activity, i.e. adults should engage in physical activity of high intensity for at least 20 minutes a day at least three days a week [4]. In 2012, the prevalence of smoking among Dutch adults was 23% and 7% had high levels of alcohol consumption (i.e., for men more than 2 glasses of alcohol a day and for women more than 1 glass of alcohol a day) [5, 6]. An unhealthy diet is also very common: in 2012 30% of the Dutch adult population met the norms for vegetable consumption (i.e., at least 200 grams of vegetables per day, 7 days a week), and 44% met the fruit consumption norm (i.e., at least 2 pieces of fruit per day, 7 days a week) [4]. Benefits of worksite health promotion programs The WHO states that the workplace directly influences the physical, mental, economic and social well-being of employees and in turn the health of their families, communities and society [7]. Based on this statement, they identified the workplace as one of the priority settings for health promotion in the 21st century [7, 8]. As a consequence, worksite health promotion (WHP) programs have been developed and implemented in research setting (i.e. (randomized) controlled trials) more frequently during the past two decades

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

[9]. In these studies the coordination of the development and implementation was in the hands of Universities or other research institutes. As employees spend up to 60% of their waking hours at the workplace, a large proportion of the adult population can be reached at once, in groups as well as individually [7, 10-12]. In addition, workplaces allow for the implementation of multi-component interventions that not only influence individual health behaviors, but also could address organizational and environmental factors to maximize impact [11, 13]. Furthermore, they offer an existing infrastructure (i.e., communication channels, social networks and organizational structures) and social support system, making the workplace a convenient setting to stimulate a healthy lifestyle [7, 10]. However, besides these opportunities, the worksite can also be a complicated setting for health promotion activities. First, working adults are in general healthier compared to the general population and compared to non-working adults [14]. Second, ethical issues can play a role, since the worksite is primarily intended for working, not for promoting health and a healthy lifestyle of employees [10, 15]. Related to this, an employee is dependent on an employer. So, privacy and autonomy issues could play a role and the general feeling of employees could be that interference of employers with employee’s health might go beyond their ‘jurisdiction’ [10, 16-18]. However, employees are not legally obliged to engage in worksite health promotion initiatives [9]. Additionally their health status may not at all be disclosed to their employer. It is only allowed to provide results of periodic health checks (including information of lifestyle) to an employer on a group level and not on an individual level, in order to avoid discrimination on the basis of health risks [19-21]. The Dutch Ministry of Public Health, Welfare and Sports (i.e. “Ministerie van Volksgezondheid, Welzijn en Sport “) also stipulates that employers have a duty to take care of the health of their employees [22]. In addition, the European occupational health and safety legislation states that employers are obligated to offer their employees a periodic health check to prevent or limit risks related to employees health caused by their work (environment) [23]. However, this legislation does not mean that employers and employees are legally obliged to engage themselves in a WHP program [9]. Implementing a WHP program is not only beneficial for the individual employee. Organizations may also benefit from these programs, as overweight and unhealthy employees have lower productivity levels, decreased work ability and higher sickness absenteeism rates [24-28]. As research has shown that WHP programs can favorably affect important outcomes from an organizational perspective, such as sickness absence, presenteeism and productivity, organizations could benefit in terms of lowering costs [2, 11, 12, 29, 30]. Although the effectiveness of WHP programs is variable across studies on health

General Introduction

and lifestyle related outcomes on the employee level, indeed positive effects have been reported among participants on behaviors such as physical activity, healthy nutrition and smoking cessation. In addition, studies have reported a significant decrease in the number of employees with obesity [1, 31-36]. Several intervention strategies aimed at improving the health and lifestyle of employees have been investigated over the years. However, multicomponent interventions in which several strategies (i.e., environmental, educational and individual) are combined are most effective in improving the lifestyle of employees, provided that they were implemented successfully [37]. However, successful implementation of WHP in daily practice is lacking often [38-42]. Implementation of WHP programs Literature applies many different terms and definitions for ‘implementation research’. Implementation could best be defined as “an effort specifically designed to get best practice findings and related products into routine and sustained use through appropriate uptake/ adoption of interventions” [43]. However, despite the positive effects of WHP programs on both the employee and the organizational level, WHP programs are rarely implemented and used in daily routine or practice. So broad-scale, nationwide implementation of effective WHP programs that promote a healthy lifestyle across a wide range of settings fails [38-42]. An important question is why? Literature has shown that the main focus of researchers in the field of WHP is mostly on determining the effectiveness of these programs in traditional randomized controlled trials (RCT’s). A recent trend is seen in which a more pragmatic design is used. This means that the design could still be a RCT, but that interventions are evaluated under circumstances that resemble routine practice conditions as much as possible [44]. In this type of study the researcher is often merely an observer. Such a pragmatic design makes it possible to evaluate the program under “real world” circumstances. This should facilitate the generalizability of the results (i.e. external validity), but at the same time enhance the internal validity (i.e. the ability to draw true conclusions about causes and effects) of the study, as participants are still randomized [44]. It has also become widely known that higher levels of implementation favorably affect study outcomes [41]. In line with this development, more and more studies evaluate the implementation process of interventions by means of process evaluations. Process evaluations can give valuable insights into the interpretation of the (lack of) effects of an intervention, as it allows researchers to identify successful and unsuccessful program components. Furthermore, a thorough process evaluation allows to map barriers and facilitators affecting

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

the implementation of a program on different organizational levels. These are valuable outcomes, which can be used to improve future program implementation in daily practice. However, up till now, no conceptual framework for process evaluations takes all of these aspects into account. As a consequence many aspects remain unknown. How good do these “pragmatic RCT’s” resemble “real-world” circumstances? How applicable are the results of these process evaluations for other settings and for daily routine and practice? It appears that in order to improve the translation from research into practice we need to open the implementation “black box”. In order to do so, there is a need for more observational study designs, such as a formative evaluation that should go alongside pragmatic trials [45, 46]. A formative evaluation is an assessment that focuses on “the internal dynamics and actual operations of a program, in order to understand its strengths and weaknesses and changes that occur in it over time” [46, 47]. So it gives researchers insight into program implementation over time and employs a mix of qualitative and quantitative research techniques and stimulates real-time monitoring. Real-time monitoring can be obtained by an ‘embedded scientist’, meaning that the researchers is part of the implementation process, but does not actively intervene in this process and acts more as a ‘fly on the wall’. This approach should be helpful in opening the ‘black box’ of the implementation process, as it allows researchers to capture the dynamic process of implementation by means of data triangulation. In doing so, detailed information on the implementation of a specific WHP can be obtained, whereby the focus is not solely on the implementation of the specific interventions by using process evaluations, but also on identifying barriers and facilitators that should be taken into account when implementing the program outside a research setting. 7-step implementation strategy Despite the lack of focus on studying the implementation process of WHP programs and the lack of implementation in practice, researchers do agree that WHP programs should be systematically implemented and should fit the specific context of the worksite in order to enhance success. Additionally, tailoring interventions to the specifics of the target group and the worksite is a successful approach, as each worksite has its own culture and natural social network [48]. In order to successfully implement a WHP program, programs need to go through the four stages of the diffusion of innovations theory: 1) dissemination, 2) adoption, 3) implementation and 4) continuation [49]. The transition from one stage to the next can be influenced by barriers and facilitators on different organizational levels and

General Introduction

can be categorized into five main groups: 1) characteristics of the socio-political context, 2) characteristics of the organization, 3) characteristics of the implementer, 4) characteristics of the program and 5) characteristics of the participant (figure 1) [50, 51]. As research has shown that it is important to take these implementation factors into account when implementing a WHP program, an effective implementation strategy to implement WHP programs within organizations could be a solution [41, 50, 52]. Therefore, Wynne and Clarkin (1992) developed a 7-step strategy of which it is hypothesized that it facilitates a structured and successful implementation of a WHP program that fits the specifics of the worksite. The 7-step strategy is based on the results of a survey among 1500 European companies across seven countries questioning the company’s health policies, worksite health promotion activities and case studies of good practices [53, 54]. The strategy adopts a ‘user-driven’ approach, meaning that employees and managers from different organizational levels are involved and responsible for the development and implementation process. The researchers merely act as observer and do not actively interfere with the process, unless this is specifically asked for by the participating organizations. It is hypothesized that this ‘user-driven’ approach ensures a fit with the wishes and needs of both the employer as well as the employees, thereby enhancing program implementation and maintenance over time. The 7-step strategy mainly aims to systematically develop, implement and maintain health promotion programs at the workplace and consists of the following steps: 1) creating solid support, 2) formation of a project structure, 3) performing a needs assessment, 4) development of interventions, 5) implementation of interventions, 6) evaluation of the implemented interventions, and 7) embedding the interventions in the general occupational health policy of the organization. The 7-step strategy is a general strategy, but it allows a tailored, worksite-specific approach, making it suitable (in theory) for nationwide implementation in different types of organizations.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

Innovation determinants

Implementation strategy

Innovation process

Characteristics of the sociopolitical context Adoption Characteristics of the organization Implementation Characteristics of the innovation

Continuation

Characteristics of the adopting person/user

Figure 1. Framework presenting the innovation process and related innovation determinants (Fleuren et al., 2004)

The 7-step strategy has already been tested in 2007 in daily practice. Stating that a vital, productive and motivated workforce will contribute to maintaining a competitive position in the market, a large national bank implemented an intervention called BRAVO [55]. BRAVO is the Dutch abbreviation for physical activity, smoking, alcohol use, nutrition and relaxation (i.e. â&#x20AC;&#x2DC;Bewegen, Roken, Alcohol, Voeding en Ontspanningâ&#x20AC;&#x2122;). The 7-step strategy was successfully executed in this national bank and seemed effective in positively changing several lifestyle related variables such as increased physical activity levels of moderate intensity and higher levels of relaxation. Although, the strategy seemed promising, secular trends cannot be ruled out, because there was neither a control group, nor a thorough and real-time evaluation of the implementation process. Objectives of this thesis Based on the topics discussed in this introduction, this thesis has two main objectives: 1) To identify implementation determinants (i.e. barriers an facilitators) that either hamper or facilitate the implementation of WHP programs focusing on healthy lifestyle changes, and;

General Introduction

2) To assess whether the use of the 7-step strategy contributes to the successful development, implementation and maintenance of a WHP program aimed at stimulating a healthy lifestyle change among employees. Outline of this thesis In Chapter 2 a conceptual framework is presented for conducting a formative evaluation, which is also used in this study, including the design of the study and a detailed description of the 7-step strategy. Chapter 3 describes a systematic review that was conducted to further our understanding of the quality of process evaluations alongside effect evaluations for WHP program, to identify implementation determinants in WHP program, and to explore the relationship between program effectiveness and program implementation. Chapter 4 describes the process evaluation that was conducted as part of the formative evaluation by applying a mixed methods approach in order to gain insight into the implementation of the lifestyle interventions that were developed in this study, using the 7-step strategy. Additionally, the effectiveness is described of the implemented lifestyle interventions on lifestyle behaviors (physical activity, fruit intake, vegetable intake, smoking, alcohol use and relaxation), in a quasi-experimental controlled trial conducted alongside the formative evaluation. In chapter 5 the results are presented on the association between measured implementation factors identified in our review and employee participation in and satisfaction with the BRAVO@Work program. The usefulness of the 7-step strategy applied in this study is evaluated in chapter 6. Finally, chapter 7 presents a general discussion of the main findings of chapters 2 to 6, methodological considerations, as well as recommendations for daily practice and research. This thesis is concluded with both an English and a Dutch summary.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

References 1.

Pronk NP, Martinson B, Kessler RC, Beck AL, Simon GE, Wang P: The association between work performance and physical activity, cardiorespiratory fitness, and obesity. J Occup Environ Med 2004, 46(1):19-25.

van Duijvenbode DC, Hoozemans MJ, van Poppel MN, Proper KI: The relationship between overweight and obesity, and sick leave: a systematic review. Int J Obes (Lond) 2009, 33(8):807-816.

CBS Statline: Lengte en gewicht van personen, ondergewicht en overgewicht (1981-2014) © Centraal Bureau voor de Statistiek, Den Haag/Heerlen.[accessed on July 27 2015]. URL: http://statline.cbs.nl/ Statweb/publication/?DM=SLNL&PA=81565NED&D1=2,4&D2=0&D3=0,5-10&D4=0&D5=33&HDR=T &STB=G1,G3,G2,G4&VW=T: http://statline.cbs.nl/Statweb/publication/?DM=SLNL&PA=81565NED& D1=2,4&D2=0&D3=0,5-10&D4=0&D5=33&HDR=T&STB=G1,G3,G2,G4&VW=T ed.

van der Klauw M, Verheijden MW, Slinger JD: Monitor Convenant Gezond Gewicht. 2013, TNO 2013 R12037.

van Laar MW, Verdurmen J, Zantinge EM, van Bakel AM: Hoeveel mensen gebruiken alcohol? In: Volksgezondheid Toekomst Verkenning, Nationaal Kompas Volksgezondheid. [accessed on July 27 2015]. URL: http://www.nationaalkompas.nl/gezondheidsdeterminanten/leefstijl/alcoholgebruik/ hoeveel-mensen-gebruiken-alcohol/: Bilthoven: RIVM; 2014.

Monshouwer K, Verdurmen J, Harbers MM: Hoeveel mensen roken? In: Volksgezondheid Toekomst Verkenning, Nationaal Kompas Volksgezondheid. [accessed on July 27 2015]. URL: http://www. nationaalkompas.nl/gezondheidsdeterminanten/leefstijl/roken/hoeveel-mensen-roken/: Bilthoven: RIVM; 2014.

World Health Organization: Workplace health promotion. The workplace: a priority setting for health promotion. Accessed 2011 jan 16.

WHO: Mental health and well-being at the workplace - protection and inclusion in challenging times: Copenhagen: 2010.

Hulshof C, Frings-Dresen MH: International OH systems. Part 2: occupational health delivery in the Netherlands. Occupational health at work 2010, 6:19-23.

10. Robroek SJW, Vathorst SVD, Hilhorst MT, Burdorf A: Moral issues in workplace health promotion. Int Arch Occup Environ Health 2012, 85(3):327-331. 11. Goetzel RZ, Ozminkowski RJ: The health and cost benefits of work site health-promotion programs. In Edited by Anonymous 2008:303-323. [Anonymous, vol 29.] 12. Kuoppala J, Lamminpaa A, Husman P: Work health promotion, job well-being, and sickness absences-a systematic review and meta-analysis. J Occup Environ Med 2008, 50(11):1216-1227. 13. Bull SS, Gillette C, Glasgow RE, Estabrooks PA: Work site health promotion research: to what extent can we generalize the results and what is needed to translate research into practice? Am J Public Health 2003, 30:537-549. 14. Proper KI, Hildebrandt VH: Overweight and obesity among Dutch workers: differences between occupational groups and sectors. Int Arch Occup Environ Health 2010, 83(1):61-68. 15. van den Brink, C.L., Blokstra, A. Hoeveel mensen hebben overgewicht? Nationaal Kompas Volksgezondheid 2014 [accessed 2014 August 17]]; Volksgezondheid Toekomst verkenning Available from: URL: http://www.nationaalkompas.nl/gezondheidsdeterminanten/persoonsgebonden/ overgewicht/hoeveel-mensen-hebben-overgewicht/ [http://www.nationaalkompas.nl/gezondheidsdeterminanten/persoonsgebonden/overgewicht/hoeveel-mensen-hebben-overgewicht/] 16.

Have MT, van der Heide A, Mackenbach JP, de Beaufort ID: An ethical framework for the prevention of overweight and obesity: a tool for thinking through a programme’s ethical aspects. Eur J Public Health 2013, 23(2):299-305.

17. ten Have M, de Beaufort ID, Teixeira PJ, Mackenbach JP, van der Heide A: Ethics and prevention of overweight and obesity: an inventory. Obes Rev 2011, 12(9):669-679.

General Introduction

18. ten Have M, de Beaufort ID, Mackenbach JP, van der Heide A: An overview of ethical frameworks in public health: can they be supportive in the evaluation of programs to prevent overweight? BMC Public Health 2010, 10:638-2458-10-638. 19. Allegrante JP, Sloan RP: Ethical dilemmas in workplace health promotion. Prev Med 1986, 15(3):313320. 20. Rothstein MA, Harrell HL: Health risk reduction programs in employer-sponsored health plans: Part II-law and ethics. J Occup Environ Med 2009, 51(8):951-957. 21. de Laat IJ: Privacy en de zieke werknemer. Tijdschrift Arbeidsrechtpraktijk 2009, June(Special 1):4-9. 22. Ministerie van Volksgezondheid, Welzijn en Sport: Gezond zijn, gezond blijven. Een visie op gezondheid en preventie. 2007, Den Haag. 23.

Ministerie van Justitie: Wet van 18 maart 1999, houdende bepalingen ter verbetering van de arbeidsomstandigheden. Arbeidsomstandighedenwet 1999, Artikel 18.

24. Kirsten W: Making the link between health and productivity at the workplace-a global perspective. Ind Health 2010, 48(3):251-255. 25. Kirsten W: Making the link between health and productivity at the workplace -A global perspective. Ind Health 2010, 48(3):251-255. 26. Jans MP, van den Heuvel SG, Hildebrandt VH, Bongers PM: Overweight and Obesity as Predictors of Absenteeism in the Working Population of the Netherlands. J Occup Environ Med 2007, 49(9):975980. 27. Robroek SJ, van den Berg TI, Plat JF, Burdorf A: The role of obesity and lifestyle behaviours in a productive workforce. Occup Environ Med 2011, 68(2):134-139. 28. Alavinia SM, Molenaar D, Burdorf A: Productivity loss in the workforce: Associations with health, work demands, and individual characteristics. Am J Ind Med 2009, 52(1):49-56. 29. Proper KI, Staal BJ, Hildebrandt VH, van der Beek AJ, van Mechelen W: Effectiveness of physical activity programs at worksites with respect to work-related outcomes. Scand J Work Environ Health 2002, 28(2):75-84. 30.

Cancelliere C, Cassidy JD, Ammendolia C, Cote P: Are workplace health promotion programs effective at improving presenteeism in workers? A systematic review and best evidence synthesis of the literature. BMC Public Health 2011, 11:395-2458-11-395.

31. Ni Mhurchu C, Aston LM, Jebb SA: Effects of worksite health promotion interventions on employee diets: A systematic review. BMC Public Health 2010, 10(10):62. 32. Proper KI, Koning M, van der Beek AJ, Hildebrandt VH, Bosscher RJ, van Mechelen W: The effectiveness of worksite physical activity programs on physical activity, physical fitness, and health. Clin J Sport Med 2003, 13(2):106-117. 33. Jepson RG, Harris FM, Platt S, Tannahill C: The effectiveness of interventions to change six health behaviors: a review of reviews. BMC Public Health 2010, 8(10):538. 34. Verweij LM, Coffeng J, van Mechelen W, Proper KI: Meta-analyses of workplace physical activity and dietary behaviour interventions on weight outcomes. Obesity Reviews 2011, 12(6):406-429. 35. Anderson LM, Quinn TA, Glanz K, Ramirez G, Kahwati LC, Johnson DB, Buchanan LR, Archer WR, Chattopadhyay S, Kalra GP, Katz DL: The Effectiveness of Worksite Nutrition and Physical Activity Interventions for Controlling Employee Overweight and Obesity. A Systematic Review. Am J Prev Med 2009, 37(4):340-357. 36. Groeneveld IF, Proper KI, Van Der Beek AJ, Hildebrandt VH, Mechelen WV: Lifestyle-focused interventions at the workplace to reduce the risk of cardiovascular disease - A systematic review. Scandinavian Journal of Work, Environment and Health 2010, 36(3):202-215. 37. Schroer S, Haupt J, Pieper C: Evidence-based lifestyle interventions in the workplace--an overview. Occup Med (Lond) 2014, 64(1):8-12. 38. Grol R, Wensing M: What drives change? Barriers to and incentives for achieving evidence-based practice. Med J Aust 2004, 180(6 Suppl):S57-S60.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

39. Grol R, Grimshaw J: From best evidence to best practice: effective implementation of change in patientsâ&#x20AC;&#x2122; care. Lancet 2003, 362(9391):1225-1230. 40. Grol R: Implementing guidelines in general practice care. Qual Health Care 1992, 1(3):184-191. 41. Durlak JA, DuPre EP: Implementation matters: a review of research on the influence of implementation on program outcomes and the factors affecting implementation. Am J Community Psychol 2008, 41(34):327-350. 42. Grimshaw J, Eccles M, Tetroe J: Implementing Clinical Guideline: Current Evidence and Future Implication. J Contin Educ Health Prof 2004, 24:S31-S37. 43. Curran GM, Bauer M, Mittman B, Pyne JM, Stetler C: Effectiveness-implementation hybrid designs: combining elements of clinical effectiveness and implementation research to enhance public health impact. Med Care 2012, 50(3):217-226. 44. Patsopoulos NA: A pragmatic view on pragmatic trials. Dialogues Clin Neurosci 2011, 13(2):217-224. 45. Wilson MG, Goetzel RZ, Ozminkowski RJ, DeJoy DM, Della L, Roemer EC, Schneider J, Tully KJ, White JM, Baase CM: Using formative research to develop environmental and ecological interventions to address overweight and obesity. Obesity (Silver Spring) 2007, 15 Suppl 1:37S-47S. 46. Stetler CB, Legro MW, Wallace CM, Bowman C, Guihan M, Hagedorn H, Kimmel B, Sharp N, Smith JL: The role of formative evaluation in implementation research and the QUERI experience. J Gen Intern Med 2006, Feb(21 Suppl 2):S1-8. 47. Patton MQ: Evaluation of program implementation. Eval Stu Rev Annu 1979, 4:318-345. 48. Sparling PB: Worksite health promotion: principles, resources, and challenges. Prev Chronic Dis 2010, 7(1):A25. 49. Rogers E.M.: Diffusion of innovations: New York: Free Press; 2003. 50. Fleuren M, Wiefferink K, Paulussen T: Determinants of innovation within health care organizations: literature review and Delphi study. Int J Qual Health Care 2004, 16(2):107-123. 51. Paulussen TH, Wiefferink K, Mesters I: Invoering van effectief gebleken interventies. In Gezondheidsvoorlichting en gedragsverandering. Edited by Brug, J., Van Asseman, P. & Lechner, L. Assen: van Gorcum: 2007. 52. Wierenga D, Engbers LH, van Empelen P, Duijts S, Hildebrandt VH, van Mechelen W: What is actually measured in process evaluations for worksite health promotion programs: a systematic review. BMC Public Health 2013, 13(1):1190. 53. Wynne R, Clarkin N: Under construction: building for health in the EC workplace: Luxembourg: Officer for Offical Publication of the European Communities; 1992. 54. Wynne R, European foundation for the improvement of living and working conditions: Workplace health promotion in Europe. Research summary. Luxembourg: Office for official publications of the European Communities; 1997. 55. Koenders P: BRAVO: a healthy lifestyle. Fortis puts health management on the agenda (BRAVO: een gezonde leefstijl. Fortis zet gezondheidsmanagement op de kaart). Arbo 2008, 6:16-19.

The design of a real-time formative evaluation of the implementation process of lifestyle interventions at two worksites using a 7-step strategy (BRAVO@Work) Wierenga D, Engbers LH, van Empelen P, Hildebrandt VH, van Mechelen W. BMC Public Health 2012, 12(1):619-629

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Abstract Background: Worksite health promotion programs (WHPPs) offer an attractive opportunity to improve the lifestyle of employees. Nevertheless, broad scale and successful implementation of WHPPs in daily practice often fails. In the present study, called BRAVO@ Work, a 7-step implementation strategy was used to develop, implement and embed a WHPP in two different worksites with a focus on multiple lifestyle interventions. This article describes the design and framework for the formative evaluation of this 7-step strategy under real-time conditions by an embedded scientist with the purpose to gain insight into whether this 7-step strategy is a useful and effective implementation strategy. Furthermore, we aim to gain insight into factors that either facilitate or hamper the implementation process, the quality of the implemented lifestyle interventions and the degree of adoption, implementation and continuation of these interventions. Methods: This study is a formative evaluation within two different worksites with an embedded scientist on site to continuously monitor the implementation process. Each worksite (i.e. a University of Applied Sciences and an Academic Hospital) will assign a participating faculty or a department, to implement a WHPP focusing on lifestyle interventions using the 7-step strategy. The primary focus will be to describe the natural course of development, implementation and maintenance of a WHPP by studying [a] the use and adherence to the 7-step strategy, [b] barriers and facilitators that influence the natural course of adoption, implementation and maintenance, and [c] the implementation process of the lifestyle interventions. All data will be collected using qualitative (i.e. realtime monitoring and semi structured interviews) and quantitative methods (i.e. process evaluation questionnaires) applying data triangulation. Except for the real-time monitoring, the data collection will take place at baseline and after 6, 12 and 18 months. Discussion: This is one of the few studies to extensively and continuously monitor the natural course of the implementation process of a WHPP by a formative evaluation using a mix of quantitative and qualitative methods on different organizational levels (i.e. management, project group, employees) with an embedded scientist on site.

Design of the study

Background An unhealthy lifestyle (e.g. insufficient daily physical activity, unhealthy diet, smoking, high alcohol consumption and low levels of relaxation) is related to several chronic diseases with high prevalence rates in the Netherlands like cardiovascular diseases, diabetes mellitus type II, respiratory diseases (e.g. asthma and COPD), depression and certain types of cancer [1]. A well-known consequence of an unhealthy lifestyle is overweight [2, 3]. Currently, 30.5% of the Dutch working population is overweight and an additional 6% is obese [4]. Furthermore, unhealthy and/or overweight employees show elevated sickness absence rates which significantly increase costs for the company [5]. To ensure long-lasting productivity of employees and to prevent work disability an important component is adopting and maintaining a healthy lifestyle [6, 7]. The WHO states that the workplace directly influences the physical, mental, economic and social well-being of employees and in turn the health of their families, communities and society. Therefore, the workplace offers an ideal setting and infrastructure to support the promotion of health of a large audience [8]. Furthermore, growing evidence is found for the effectiveness of worksite health promotion programs (WHPP) that promote a healthy lifestyle in general [9-12]. Nevertheless, broad scale implementation of these effective WHPP in daily practice and across a wide range of settings often fails [13-17]. In order to improve the implementation of WHPPs into daily practice, it is important to shift the focus from effect evaluations to the evaluation of the implementation process. Hence key determinants of success and failure could be obtained and addressed in future implementation. For this purpose, traditional evaluation designs (i.e. randomized controlled trials) that focus on effect evaluations are not sufficient. These evaluation designs do not provide critical information on the implementation process. So other study designs are required, which focus more on observational strategies. The complementary use of a systematic and real-time formative evaluation within an controlled trial can create a dual style approach whereby critical information on the implementation process over time can be obtained [18, 19]. A formative evaluation is an assessment that focuses on â&#x20AC;&#x153;the internal dynamics and actual operations of a program in order to understand its strengths and weaknesses and changes that occur in it over timeâ&#x20AC;? [18, 20]. It gives researchers insight into program implementation over time and employs a mix of qualitative and quantitative techniques. Formative evaluations emphasize the need for real-time monitoring of the implementation process, but is very time consuming [18-20]. The amount of time that is needed to conduct

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

a real-time formative evaluation could partially be the cause for the lack of such studies. However, investing time in effectiveness studies that are not used in daily practice and only include a posterior process evaluation that does not give insight into the important aspects of the implementation process, is also a waste of money. Despite the lack of focus on studying the implementation process, researchers do acknowledge the fact that for improving the effectiveness and implementation of WHPP in practice, these programs should be systematically implemented in order to achieve successful implementation and continuation. For instance, Durlak and Dupre showed that the level of implementation (i.e. low or high implementation) affects the outcomes obtained by health promotion programs, whereby high implementation increased program success and could lead to greater effects on outcomes for participants [16]. In addition, implementation success is for an important part dependent on an adequate fit of the program with the specific organizational context (i.e. implementation context) in which the program is implemented [16]. The implementation context differs from one worksite to another because of inherent differences between worksites, which makes it difficult to implement effective WHPPs across different worksites [21]. In order to take the implementation context into account it is important to involve the target population (i.e. employees) and implementers within the worksites in the development phase of the WHPP and to keep them involved throughout the whole implementation process. This allows the worksite to incorporate and adjust the WHPP and implementation strategy to their specific needs, interests and the existing setting, thereby increasing the chances of implementation success [22]. Furthermore, in order to successfully implement WHPPs, programs need to pass through the four stages (i.e. dissemination, adoption, implementation and continuation) as stated in the diffusion of innovations theory [23]. Four main categories of innovation determinants may influence the transition process from one stage to the next as potential barriers or facilitators for implementation (see figure 1): 1) characteristics of the sociopolitical context (e.g. fit with existing rules, regulations, and legislation), 2) characteristics of the organization (e.g. hierarchical structure, available expertise), 3) characteristics of the innovation (e.g. compatibility, relevance), and 4) characteristics of the adopting person/user (e.g. self-efficacy, degree of ownership) [24, 25]. The theory described above provides the key elements that should be addressed when implementing a WHPP successfully [23, 25]. However, this theory, along with other implementation theories, does not provide specific strategies or guidelines for implementation.

Design of the study

As such, a new and systematic 7-step implementation strategy was developed that incorporates most of the fore mentioned aspects for successful implementation. This strategy also aims to maintain the implemented programs over time [26, 27]. The 7-step strategy is based on a â&#x20AC;&#x2DC;user-drivenâ&#x20AC;&#x2122; approach towards developing and implementing interventions that specifically address the capacities and needs of the target population at multiple organizational levels (i.e. management, project group, employees). User-driven within this context means that health objectives, interventions and implementation strategies are (co-)developed by members of the target population at different levels of the worksite. The 7-step strategy incorporates planning, implementation, evaluation and maintenance. The strategy ensures that the interventions will be tailored to the specifics of the worksite, thereby ensuring a fit with the implementation context. This increases possibilities for maintenance over time. The 7-step strategy has already been used in practice but whether this strategy is an effective and generic approach for developing and implementing WHPPs has never been studied systematically [28].

Innovation determinants

Implementation strategy

Innovation process

Characteristics of the sociopolitical context Adoption Characteristics of the organization Implementation Characteristics of the innovation

Continuation

Characteristics of the adopting person/user

Figure 1. Framework presenting the innovation process and related innovation determinants (Fleuren et al., 2004)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Therefore the present study, called BRAVO@Work, describes the formative evaluation of this 7-step strategy under real-time conditions by an embedded scientist, with the aim to evaluate and monitor whether this 7-step strategy is a useful and effective strategy to successfully develop and implement a WHPP at two worksites, with a focus on healthy lifestyle changes. Furthermore, we aim to gain insight into factors that either facilitate or hamper the implementation process, the quality of the implemented lifestyle interventions and the degree of adoption, implementation and continuation. This article describes the design and framework for the formative evaluation of the natural course of the development, implementation and maintenance of BRAVO@Work.

Methods Study design, population and setting This study is a formative evaluation, alongside a controlled trial, within two different worksites with an embedded scientist on site to continuously monitor the implementation process. Each worksite (i.e. a University of Applied Sciences and an Academic Hospital) will assign a participating faculty (546 employees) or a department (635 employees) respectively, which will implement a WHPP using the 7-step strategy. Furthermore, both participating worksites will assign a control faculty/department that will not be allowed to participate in the implementation process and use of the 7-step strategy. Employees that are 18 years or older are eligible to participate in the study. Prior to data collection all employees will be informed about the study purposes, after which informed consent will be obtained. All data will be collected using qualitative (i.e. real-time monitoring and semi-structured interviews) and quantitative methods (i.e. process evaluation questionnaires), applying data triangulation. Except for the real-time monitoring, the data collection will take place at baseline and after 6, 12 and 18 months. Employees from different organizational levels of the worksite will be approached to actively participate in the project. The study protocol has been approved by the Medical Ethics Committee of the University Medical Centre of Utrecht (Utrecht, the Netherlands).

Design of the study

The 7-step strategy The 7-step strategy is based on a study in 1992 by Wynne and Clarkin and is supported by the European Foundation for the Improvement of living and Working Conditions [27]. The study of Wynne and Clarkin consisted of two phases. First they conducted a survey among almost 1500 European companies across seven countries questioning their health policies and other activities for worksite health promotion. Second, case studies of good practices were conducted to determine how these companies had organized activities for health promotion at the workplace and how they had integrated these activities in their general occupational health policy. The results of this study showed that the following five aspects were important when implementing a successful health policy at the workplace: A) Needs assessment: for the establishment of a health policy it is important that the wishes and needs of employees are analyzed. In this way the intervention activities can be developed according to their needs; B) Participation: key figures from different levels in the companyâ&#x20AC;&#x2122;s organization need to be involved in the development and implementation of the health promotion program to create a solid support for the health policy. This can be done by means of working groups; C) Flexibility: health promotion programs are similar at some basic points. However, they are not standard programs, since a WHPP needs to fit the specifics of the workplace; D) Integration: the health promotion program needs to include activities that are both aimed at the individual employee and at the work environment; and E) Multidisciplinary: several experts in the fields of human resources, communication, health management, psychology and working environment need to be involved in the development and implementation to increase program effectiveness. These five aspects were translated by Wynne and Clarkin into a new and generic 7-step strategy for the systematic development and implementation of health promotion programs (interventions) at the workplace [26, 27]. The implementation strategy consists of the following 7 steps: 1) creating solid support, 2) formation of basic structures, 3) performing a needs assessment, 4) development of the interventions and health policy, 5) implementation of the interventions, 6) evaluation of the implemented interventions, and 7) embedding the interventions in the general occupational health policy of the organization. A schematic description of the strategy is given in figure 2. The main aspect of this strategy is the active participation of relevant stakeholders (i.e. managers, employees, communications officer, human resources staff, facility management, health and safety executive, company physician) when passing through the 7 steps. Therefore, in step 2, it is recommended that representatives of employees and relevant stakeholders from the participating organizations should be asked to participate in the use and application

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

of the 7-step strategy by taking part in a steering committee (SG) a project group (PG) or a working group (WG). The purpose of these various groups is explained in figure 2, step 2. It is hypothesized that application of the 7-step strategy should ensure that developed and implemented interventions will be tailored to the needs of different stakeholders within the worksite. Furthermore it aims to facilitate the integration of successful interventions in companyâ&#x20AC;&#x2122;s general occupational health policy. In the BRAVO@Work study the 7-step strategy will be used to develop and implement interventions related to multiple lifestyle behaviors (i.e. BRAVO-interventions) at the worksite and to integrate these lifestyle interventions in the companyâ&#x20AC;&#x2122;s general health policy. The way of use and the content of the 7-step strategy will be transferred to representatives within both participating worksites by the researchers. Since we aim to study the natural course of implementation, not the researchers but the representatives (SG, PG, WG) from both participating worksites themselves are set in the lead and made responsible for performing all actions needed according to the 7 steps. However, to ensure the quality of the interventions, the following guidelines are issued: A) only best evidence interventions that fit the worksite should be selected, B) the interventions should be related to at least two BRAVO lifestyle themes, and C) the interventions should involve an environmental component. Framework for evaluation of the implementation process In order to systematically investigate and evaluate the implementation process the four main aspects of the theory described in the introduction (i.e. innovation determinants, adoption, implementation and continuation; figure 1) are operationalized by using a combination of the framework of Steckler and Linnan for process evaluations and the REAIM framework [29-32]. After combining these framework, 8 descriptive components of the implementation process need to be operationalized and subsequently evaluated in this formative evaluation: 1) context, 2) recruitment, 3) reach, 4) dose delivered, 5) dose received, 6) fidelity, 7) satisfaction and 8) maintenance [29-32]. These components will be evaluated at three different levels within the participating worksites: 1) management level, 2) project group level and 3) employee level [24]. Table 1 presents the definition of all 8 components (including data collection method and evaluation level) of the formative evaluation, which together will measure the degree of adoption, implementation and continuation of the 7-step strategy and the BRAVO-interventions within both participating worksites.

Design of the study

Step 1 Creating solid support

Step 2 Formation of basic structure

Step 3 Needs assessment

Step 4 Development phase

Step 5 Implementation phase

Step 6 Evaluation

Step 7 Maintenance

Aim step 1: create solid support of the worksites management team for implementation  Support from higher management will be ensured by signing a letter of intent saying that higher management supports the project and that employees are allowed to participate in the project during working hours  Ensuring management support by providing information on the project and evidence of the expected return of investment of the new interventions and health policy Aim step 2: formation of a project structure with employees from different organizational layers  Installing a steering committee with a chairman, preferably someone from higher management with decision making authority. Other members are also in decision making positions. This board will have monthly meetings to discuss progress and finalize decisions. Furthermore this board is responsible for the embedding of interventions in the general health policy of the company  Appointing a project leader by the chairman. The project leader has authority at the worksite and enthusiasm for the project and will be an intermediate between steering committee and project and working groups  Formation of project group (PG) and, if necessary, working groups (WG) by the project leader. The PG and WG will be responsible for the development and implementation of interventions and both groups will consist of representatives from all layers of the worksite.  The project leader will be advised to include the following relevant stakeholders: managers, employees from different teams, communication officer, human resource staff member, facility management, health and safety executive, company physician. Aim step 3: performing a needs assessment in order to develop interventions according to the needs and characteristics of the worksite  The PG or WG will be advised to perform a needs assessment among all employees by means of a standard needs assessment list. The project group can also decide to use a web-based questionnaire, a physical examination, focus groups or a combination of those methods mentioned. The needs assessment should measure the wishes of worksites management and employees. Aim step 4: develop interventions that match employees needs and characteristics of the worksite  Based on the results of the needs assessment and characteristics of the worksite the interventions will be developed.  The project leader will make a project and communication plan using input from the results of the needs assessment, PG and WG. The project plan should contain the desired changes and goals of the project, an intervention template (detailed description of interventions), communication plan and time-line, budget plan, and a list of involved persons with their tasks and responsibilities. This will facilitate a timed and structured implementation of the interventions.  The project leader will present this project plan to the steering committee, who will decide on the go-no-go criteria of this plan. Aim step 5: implementation of the interventions  The developed interventions will be implemented within the organization by the responsible project members according to the project plan. Aim step 6: evaluation of the implemented interventions and the whole project  During the implementation process an integrative evaluation should be part of the process. This allows for data-driven improvement and adjustments to the project plan as needed.  After the implementation of the interventions, the PG or WG should evaluate the program as a whole in order to determine which interventions will be part of the general occupational health policy. Aim step 7:embedding the 7-step strategy and interventions in the general health policy of the organization  The structural embedding of the use of the 7-step strategy and interventions has to be realized to maintain possible (positive) effects over time. Efforts need to be made by the steering committee to integrate the 7-step strategy and interventions in the worksites general health policy.

Figure 2. Outline of the 7-step strategy for implementation and continuation of a worksite health policy

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Table 1. Definition of the formative evaluation components at three organizational levels, including methods for data collection 4 main aspects of the implementation process

Evaluation component

Definition of evaluation component at relevant organizational levels

Data collection method

Adoption

Recruitment

Management level Sources and procedures used to approach and attract worksites and management to become effective participants of the project

Monitoring

Project group level Sources and procedures used to approach and attract PG and WG members to become effective program members

Monitoring, questionnaire, interviews

Employee level Sources and procedures used to approach and attract employees for participation in BRAVOinterventions to become effective participants

Monitoring, questionnaire, interviews

Project group level Proportion of employees who were approached as PG or WG member

Monitoring, interviews

Implementation

Reach

Employee level Monitoring Proportion of employees who were approached for participation in BRAVO-interventions Dose delivered Management level Providing the 7-step strategy to worksites and PG and WG members Project group level Proportion of intended BRAVO-interventions delivered or provided to employees by PG and WG

Monitoring

Monitoring, questionnaire, interviews

Dose Received Project group level Monitoring Proportion of companies and PG and WG members who received the 7-step implementation strategy

Fidelity

Satisfaction

Continuation

Maintenance

Employee level Proportion of employees who participated in each BRAVO-intervention

Monitoring, questionnaire, interviews

Project group level - Compliance to the 7 steps of the implementation strategy by Steering committee and PG - Compliance of PG to project implementation plan and quality of implementation of BRAVOinterventions Management level Opinion/satisfaction about the project

Monitoring, questionnaire, interviews

Project group level Opinion/satisfaction about the 7-step strategy

Questionnaire, interviews

Employee level Opinion/satisfaction about BRAVO-interventions

Questionnaire, interviews

Interviews

Management level Monitoring, Extend to which the developed BRAVOquestionnaire, interventions and 7-step strategy become routine interviews and part of everyday culture and norms of the organization including the degree to which BRAVOinterventions are continued.

Design of the study

Implementation determinants

Context

All levels Determinants of implementation which can either hinder or facilitate the implementation of the 7-step strategy and BRAVO-interventions. These determinants are subdivided within 4 main categories: 1. Characteristics of the socio-political context 2. Characteristics of the organization 3. Characteristics of the innovation 4. Characteristics of the adopting person/user

Monitoring, questionnaire, interviews

*PG = project group; *WG = working group

Adoption refers to the proportion of worksites and participants who will adopt the 7-step strategy and the BRAVO-interventions [25]. In order to successfully monitor adoption, we specifically examine recruitment. Implementation is the extent to which the intervention has been implemented and received by the intended audience. This will be assessed by examining reach, dose delivered, dose received, fidelity and satisfaction [28, 29]. Continuation is the extent to which the program is sustained over time and has become part of everyday culture of the worksite. It will be operationalized within the component maintenance [25]. The four main categories of innovation determinants are operationalized within the component context and will be called implementation determinants. These implementation determinants could either facilitate or hamper implementation. Table 2 gives an overview of the implementation determinants that will be measured per main category.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Table 2. Implementation determinants measured in this study, sub dived per main category. Main categories of implementation determinants

Measured implementation determinants

Characteristics of the 1. Willingness of participants to cooperate with the innovation socio-political context 2. Degree to which the participant is aware of the health benefits of the innovation 3. The extent to which the innovation fits into existing rules, regulations and legislation Characteristics of the organization

4. Decision making process and procedures in the organization: top-down or bottom-up 5. Hierarchical structure: extent to which decision making process is formalized through hierarchical procedures 6. Formal reinforcement by management to integrate the innovation into organizational policies 7. Organizational size (number of employees): large, medium, small 8. Functional structure (task oriented) versus product structure (output oriented) 9. Staff turnover: high, average, low 10. Degree of staff capacity in the organization or department that implements the innovation 11. Available expertise, in relation to the innovation in the organization or department 12. Number of potential users to be reached 13. Financial resources made available for implementing the innovation 14. Reimbursement for implementers/ organizations to facilitate extra efforts in applying the innovation 15. Other resources made available for implementing the innovation (e.g. equipment, manuals) 16. Administrative support available to the implementers of the innovation 17. Time available to implement the innovation 18. Availability of staff responsible for coordinating implementation in the organization 19. The implementers are involved in the development of the innovation 20. Opinion leaders who influence opinions of others in the organization or department 21. Cooperation with external partners with respect to the implementation of the innovation

Design of the study

Characteristics of the 22. Support from colleagues in implementing the innovation adopting person/user 23. Support from other implementers within the project in implementing the innovation 24. Support from their supervisors in the department with respect to the implementation of the innovation 25. Support from higher management in the organization with respect to the implementation of the innovation 26. Extent to which colleagues implement the innovation (modeling) 27. Extent to which the implementer has the skills needed to implement the innovation 28. Extent to which the implementer has the knowledge needed to implement the innovation 29. Self-efficacy: confidence of the implementer to perform the behavior needed to implement the innovation 30. Extent to which ownership by the implementer is perceived 31. Extent to which the innovation first the perceived task orientation of the implementer 32. Extent to which the implementer expects that the participant will cooperate with the innovation 33. Extent to which the implementer expects that the participant will be satisfied with the innovation 34. Extent to which the goals of the different implementers with respect to the innovation are contradictory 35. Extent to which the implementer has ethical problems with the innovation 36. Attitude of the implementer with respect to the innovation 37. Outcome expectations of the implementer and participants with respect to the innovation 38. Perceived social norm with respect to the innovation by colleagues and supervisors 39. User directed performance feedback: formative or summative feedback 40. Personal benefits for the implementers 41. Extent to which the implementers work as a team Characteristics of the innovation

42. Extent to which the procedures/guidelines of the innovation are clear 43. Extent to which the procedures/guidelines are read by the implementers 44. Extent to which the innovation is complete 45. Extent to which the innovation is too complex to work with 46. Information provided: sufficient, insufficient. 47. Compatibility: degree to which the innovation is perceived as consistent with existing work procedures 48. Triability: extent to which the innovation can be subjected to trial 49. Relative advantage: extent to which the innovation is perceived as advantageous 50. Observability: degree to which the results of the innovations are observable to the implementer 51. Extent to which the innovation is appealing to use 52. Relevance of the innovation for the participant: extent to which the innovation has added value 53. Frequency of use of the innovation: high, low 54. Image of the innovation in the organization: positive, negative

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Data collection procedure All data will be derived from different sources and collected by means of different methods, both qualitative and quantitative (i.e. data triangulation). Data triangulation enables researchers to look for patterns in all collected data in order to develop an overall interpretation including multiple views on the implementation process [33, 34]. The qualitative and quantitative data will complement each other and will thereby give insight into the natural course of implementation. The mix of qualitative and quantitative data will create a rich dataset for interpretation of the implementation process. The real-time formative evaluation was systematically planned prior to the start of the implementation within both participating worksites. This will enable us to collect all data on the 8 process components before, during and after the implementation of the lifestyle interventions in order to better understand the implementation process over time [18]. The evaluation is therefore an integral part of the ongoing implementation process. The primary researcher will continuously monitor the use of the 7-step strategy and the implementation process of the BRAVO-interventions to gain insight into determinants of implementation and all 8 process components. The primary researcher will collect minutes from all project meetings and will document all communication (emails, letters, and phone calls) throughout the implementation process and use of the 7-step strategy within both participating worksites. These minutes and observations will be documented and structured by using monthly predetermined spread sheets. These spreadsheets will be constructed and structured according to our framework for evaluation and the three organizational levels (i.e. management, project group, and employees) of this study and will therefore contain information on all 8 process components related to the 7-step strategy, but also to the lifestyle interventions and the process of adoption, implementation and continuation. Worksite management and team leaders will be asked in semi-structured telephone interviews at baseline (T0), after 6 months (T1) and after approximately 12 months (T2) for A) their experienced barriers and facilitators for the implementation of BRAVO-interventions (Context), B) whether they were aware of the project and if BRAVO-interventions were implemented (Fidelity and Dose delivered), and C) their expectations and satisfaction regarding the complete project (Satisfaction). Data from project group (PG) and working group (WG) members will be collected by means of semi-structured interviews at T0, T1, T2 and after 18 months (T3). Furthermore, a process questionnaire will be distributed at T1 and T2. PG and WG members will be asked in the interviews and process questionnaire for A) their experienced barriers and facilitators

Design of the study

for implementation related to the 7-step strategy and the implementation of the BRAVOinterventions (Context), B) for their expectations, experience and opinion about the use of the 7-step strategy and the implementation of the BRAVO-interventions (Satisfaction), C) for their adherence to the 7-step strategy and project plan (Fidelity), D) whether he or she implemented the intervention they were responsible for (dose delivered), E) which sources and procedures were used to approach and attract PG and WG members for participation in the project, and employees for participation in BRAVO-interventions (Recruitment), F) the intention to use the 7-step strategy and to continue the BRAVO-interventions in the future (Maintenance). Employees from the participating faculty/department will be asked in semi-structured interviews at T0, T1 and T2 and subsequently by a process questionnaire at T1 and T2 for A) their experienced barriers and facilitators for the implementation of BRAVO-interventions (Context), B) whether they were aware of the project and the BRAVO-interventions (i.e. Reach) C) about their expectation and opinion of the project, BRAVO-interventions and ways of recruitment (Satisfaction and Recruitment), and D) their participation to BRAVOinterventions, including reasons for participation and non-participation (Dose received). Analysis Analysis of qualitative data All qualitative data will be analyzed with the software program for qualitative analyses â&#x20AC;&#x2DC;Atlas.tiâ&#x20AC;&#x2122;. All data that is systematically collected or observed during the study is considered to be data [35, 36]. This means that not only the semi-structured, in-depth interviews, but also all data collected during the monitoring process by means of checklists and notes collected during attending project meetings will be regarded and analyzed as qualitative data. All collected data will be marked with a series of codes extracted from the text and from literature about the specific topic (i.e. open coding: describes the content of the material). The codes will be grouped into concepts in order to make them more workable (i.e. selective coding: refer to central concepts underlying the descriptive codes)[35, 36]. These concepts will be categorized to form the basis for the creation of a theory behind the data by using tables and matrices to identify and compare concepts (i.e. theoretical coding: identify patterns and relationships between concepts). All qualitative data will be presented with representative quotes, which cannot be traced back to individual persons.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Statistical analysis of quantitative data For our main objective we aim to observe trends in self-reported and observed implementation of the use of the 7-step strategy and of the implemented lifestyle interventions. That is, at level of project initiation we will examine the results of the project group. At the employee level, we examine self-reported exposure, acceptance and use of lifestyle program components.Â Mixed-effects logistic regression will be used to examine trends. A two-tailed significance level of p < 0.05 is considered to be statistically significant. This analysis will allow for the use of probabilistic or dichotomous data, and will take into account that repeated observations are nested within individuals. Within this context we will also examine the determinants thatÂ may explain exposure, acceptance and uptake level. Furthermore we will analyze demographic variables of non-responders compared to responders the questionnaire. Non-responders are employees who received the questionnaire but did not return it. Analyses will be performed with SPSS 18.0 (SPSS Inc. Chicago, Illinois, USA).

Discussion The purpose of this article was to describe the design and framework for the formative evaluation of the natural course of the development and implementation of BRAVO@Work. Additionally, the seven steps of the applied implementation strategy were presented. This 7-step strategy was designed to successfully develop and implement a WHPP at two work sites, with a focus on healthy lifestyle changes. The rising call to improve the translation of research into daily practice has created a need for a shift in focus towards the evaluation of the implementation process of interventions, rather than the current focus on effect evaluations. Consequently, one of the main strengths of the BRAVO@Work study is that, to our knowledge, this is one of the few studies that systematically monitors and evaluates the natural course of the implementation process prior to, during and after implementation by means of a real-time formative evaluation within a controlled trial. The formative evaluation of the implementation process will be conducted on multiple organizational levels and is systematically planned prior to the start of the implementation within both participating worksites. A well-planned and structured evaluation of the implementation process can provide critical information that could help explain study outcomes on the effectiveness of an intervention but most important,

Design of the study

it could provide a base for enhancing program maintenance [37-40]. The data gathered are essential for generalizing the strategy to other settings and thereby improving the translation of research into daily practice by determining the factors that either facilitate or hamper implementation. The real-time monitoring allows us to gain insight into possible changes over time in the determinants of implementation, such as the attitude towards the 7-strategy/interventions and the intentions to continue. It will allow us to better understand possible failure points in the 7-step strategy. These results will be used to enrich the 7-step strategy, resulting in an implementation strategy suitable for practice. In addition, the 7-step strategy will be evaluated in two different worksites (i.e. education and healthcare), with different organizational structures, cultures, work forces and tasks. This enables us to gain insight into the generalizability of the 7-step strategy across different worksites. As mentioned before the best method to open the ‘black box’ of the implementation process of interventions is to place the researcher on site as an embedded observer (‘fly on the wall’). Our approach fits well with recent calls, such as that of Wandersman et al., (2008) which highlight the need for user-based rather than source-based approaches. Source-based programs follow a linear sequence, meaning that the innovation is directly transferred and implemented from the perspective of the developers (I.e. source) to the users without adaptation to the specific setting in which the innovation will be used [22]. Alternatively, Wandersman et al., (2008) have focused on a user-based model, whereby interventions and implementation strategies are developed by the source but implemented based on the awareness of needs and opportunities for change from the user. This calls for alternative research designs, such as the design of the BRAVO@Work study. Our approach ensures that the dynamic process of implementation is captured by means of data triangulation, in which multiple methodologies are used to examine WHPP assessment, development, implementation and continuation. Another strength of this study is that the formative evaluation will allow us to gain insight into the fit of 7-step strategy and the implemented lifestyle interventions with the worksite. This is an important aspect of the evaluation since the 7-step strategy and literature emphasizes the need for a fit with the worksite in order to achieve successful implementation. A final and most important strength of this study is the use of a mixed methods approach (i.e. observation, monitoring, questionnaires and semi-structured interviews), accompanied by collecting data at different organizational levels in the formative evaluation of the implementation process. This data triangulation is a way of ensuring the integrity of the data since multiple views on the implementation process are mapped [34]. 35

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

However, limitations of this study can also be mentioned. First selection bias due to a selective response could occur. Healthy employees are more likely to fill out the questionnaires and to participate in interviews and interventions. This is the case with instruments that address health behaviors [41]. However, one might hypothesize that this will be less of an issue in formative evaluations, because stating your (positive or negative) opinion on the interventions does not deal with (changing) your poor or good behavior. Hence, it is less personal. To address this potential problem we will analyze demographic variables of nonresponders and invite them for interviews. A second limitation might be that members of the project group (i.e. employees) might be a poor representation of all employees working at the worksite. Ideally, employees participating in one of the three groups are ambassadors of the project at their faculty/department. However, due to the nature of this study we do not control the selection of employees for participation in these groups. Instead, we could only advice the project leader to include employees from all relevant faculties/departments as stated in the 7-step strategy. Since we will conduct a formative evaluation alongside a controlled trial both participating worksites will assign a control faculty/department that will not be allowed to participate in the implementation process. This will allow us to collect data on the effectiveness of the implemented lifestyle interventions by means of a web-based questionnaire distributed in the intervention faculty/department (i.e. department working with the 7-step strategy) and in the control faculty/department. In doing so we will be able to link the outcomes on the effectiveness to the implementation process and will give us insight into the quality of the implemented lifestyle interventions and possibly the separate effects of each implemented lifestyle intervention [30, 42-45]. When BRAVO@Work proves to be successful, the 7-step strategy will be adjusted if necessary and then disseminated nationwide by the Dutch Institute for Sport and Physical Activity, providing companies with an effective strategy to develop and implement a lifestyle policy as part of their health management.

Design of the study

References 1.

Gommer AM, Poos MJJC: Welke ziekten hebben de hoogste prevalentie?In: Volksgezondheid toekomstverkenning. Nationaal Kompas Volksgezondheid. ed. Bilthoven: RIVM; 27 april 2010.

van Duijvenbode DC, Hoozemans MJ, van Poppel MN, Proper KI: The relationship between overweight and obesity, and sick leave: a systematic review. Int J Obes (Lond) 2009, 33(8):807-816.

Proper KI, Hildebrandt VH: Overweight and obesity among Dutch workers: differences between occupational groups and sectors. Int Arch Occup Environ Health 2010, 83(1):61-68.

Jans MP, van den Heuvel SG, Hildebrandt VH, Bongers PM: Overweight and Obesity as Predictors of Absenteeism in the Working Population of the Netherlands. JOEM 2007, 49(9):975-980.

Kirsten W: Making the link between health and productivity at the workplace--a global perspective. Ind Health 2010, 48(3):251-255.

Mills PR, Kessler RC, Cooper J, Sullivan S: Impact of a health promotion program on employee health risks and work productivity. Am J Health Promot 2007, 22(1):45-53.

World Health Organization: Workplace health promotion. The workplace: a priority setting for health promotion. Accessed 2011 jan 16.

Engbers LH, van Poppel MN, Chin A Paw MJ, van Mechelen W: Worksite health promotion programs with environmental changes: a systematic review. Am J Prev Med 2005, 29(1):61-70.

10. Kremers S, Reubsaet A, Martens M, Gerards S, Jonkers R, Candel M, de Weerdt I, de Vries N: Systematic prevention of overweight and obesity in adults: a qualitative and quantitative literature analysis. Obes Rev 2010, 11(5):371-379. 11. Ni Mhurchu C, Aston LM, Jebb SA: Effects of worksite health promotion interventions on employee diets: a systematic review. BMC Public Health 2010, 10(10):62. 12. Jepson RG, Harris FM, Platt S, Tannahill C: The effectiveness of interventions to change six health behaviors: a review of reviews. BMC Public Health 2010, 8(10):538. 13. Grol R, Wensing M: What drives change? Barriers to and incentives for achieving evidence-based practice. Med J Aust 2004, 180(6 Suppl):S57-S60. 14. Grol R, Grimshaw J: From best evidence to best practice: effective implementation of change in patientsâ&#x20AC;&#x2122; care. Lancet 2003, 362(9391):1225-1230. 15. Grol R: Implementing guidelines in general practice care. Qual Health Care 1992, 1(3):184-191. 16. Durlak JA, DuPre EP: Implementation matters: A review of research on the influence of implementation on program outcomes and the factors affecting implementation. Am J Community Psychol 2008, 41:327-350. 17. Grimshaw J, Eccles M, Tetroe J: Implementing Clinical Guideline: Current Evidence and Future Implication. J Contin Educ Health Prof 2004, 24:S31-S37. 18. Stetler CB, Legro MW, Wallace CM, Bowman C, Guihan M, Hagedorn H, Kimmel B, Sharp N, Smith JL: The role of formative evaluation in implementation research and the QUERI experience. J Gen Intern Med 2006, Feb(21 Suppl 2):S1-8. 19. Mittman BS: Creating the evidence base for quality improvement collaboratives. Ann Int Med 2004, 140(11):897-901. 20. Patton MQ: Evaluation of program implementation. Eval Stu Rev Annu 1979, 4:318-345. 21. Leykum LK, Pugh JA, Lanham HJ, Harmon J, McDaniel RRJ: Implementation research design: integrating participatory action research into randomized controlled trials. Implement Sci 2009, 4(69).

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

22. Wandersman A, Duffy J, Flaspohler P, Noonan R, Lubell KS,L., Blachman M, Dunville R, Saul J: Bridging the gap between prevention research and practice: the interactive systems framework for dissemination and implementation. Am J Community Psychol 2008, 41:171-181. 23. Rogers E.M.: Diffusion of innovations: New York: Free Press; 2003. 24. Paulussen TH, Wiefferink K, Mesters I: Invoering van effectief gebleken interventies. In Gezondheidsvoorlichting en gedragsverandering. Edited by Brug,J.,Van Asseman,P.& Lechner,L. Assen: van Gorcum: 2007. 25. Fleuren M, Wiefferink K, Paulussen T: Determinants of innovation within health care organizations: literature review and Delphi study. Int J Qual Health Care 2004, 16(2):107-123. 26. Wynne R, European foundation for the improvement of living and working conditions: Workplace health promotion in Europe. Research summary. Luxembourg: Office for official publications of the European Communities; 1997. 27. Wynne R, Clarkin N: Under construction: building for health in the EC workplace: Luxembourg: Officer for Offical Publication of the European Communities; 1992. 28. Koenders P: BRAVO: a healthy lifestyle. Fortis puts health management on the agenda (BRAVO: een gezonde leefstijl. Fortis zet gezondheidsmanagement op de kaart). Arbo 2008, 6:16-19. 29. Glasgow RE, Vogt TM, Boles SM: Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health 1999, (89):9-1322. 30. Dzewaltowski.D.A., Estabrooks PA, Glasgow RE: The future of physical activity behavior change research: what is needed to improve translation of research into health promotion practice? Exerc Sport Sci Rev 2004, 32(2):57-63. 31. Dzewaltowski.D.A., Glasgow RE, Klesges LM, Estabrooks PA, Brock E: RE-AIM: evidence-based standards and a web resource to improve translation of research into practice. Ann Behav Med 2004, 28(2):75-80. 32. Steckler A, Linnan L: Proces Evaluation for Public Health Interventions and Research: San Francisco: Jossey-Bass; 2002. 33. Mays N, Pope C: Rigour and qualitative research. BMJ 1995, 311(6997):109-112. 34. Mays N, Pope C: Assessing quality in qualitative research. BMJ 2000, 320:50-52. 35. Boeije H: Analyze in qualitative research. (Analyseren in kwalitatief onderzoek. denken en doen): Amsterdam, The Netherlands: Boom onderwijs; 2009. 36. Wester F, Peters V: Qualitative analyse. Principles and procedures (Kwalitatieve analyse. Uitgangspunten en procedures): Bussem, The Netherlands: Coutinho bv; 2004. 37. Steckler A, Ethelbah B, Martin CJ, Stewart D, Pardilla M, Gittelsohn J, Stone E, Fenn D, Smyth M, Vu M: Pathways process evaluation results: a school-based prevention trial to promote healthful diet and physical activity in American Indian third, fourth, and fifth grade students. Prev med 2003, 37:S80-S90. 38. Baranowski T, Stables G: Process evaluations of the 5-a-day projects. Health Educ Behav 2000, 27(2):157-166. 39. Pratt CC, McGuidan WM, Katzev AR: Measuring program outcomes: Using retrospective pretest methodology. Amer J Evaluation 2000, 21(3):341-349. 40. Johnson CC, Lai YL, Rice J, Rose D, Webber LS: ACTION live: using process evaluation to describe implementation of a worksite wellness program. J Occup Environ Med 2010, 52(Suppl 1):S14-S21. 41. Glasgow RE, McCaul KD, Fisher KJ: Participation in worksite health promotion: a critique of the literature and recommendations for future practice. Health Educ Q 1993, 20(3):391-408. 42. Bouffard JA, Taxman FS, Silvermand R: Improving process evaluations of correctional programs by using a comprehensive evaluation methodology. Eval Programm Plann 2003, 26:149-161. 43. Bull SS, Gillette C, Glasgow RE, Estabrooks PA: Work site health promotion research: to what extent can we generalize the results and what is needed to translate research into practice? Am J Public Health 2003, 30:537-549.

Design of the study

44. Craig P, Dieppe P, Macintyre S, Nazareth I, Petticrew M: Developing and evaluating complex interventions: the ned Medical Research Council guidance. BMJ 2008, 337(1655):979-983. 45. Campbell NC, Murray E, Darbyshire J, Emery J, Farmer A, Griffiths F, Guthrie B, Lester H, Wilson P, Kinmonth AL: Designing and evaluating complex interventions to improve health care. BMJ 2007, 334(7591):455-459.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

What is actually measured in process evaluations for worksite health promotion programs: a systematic review

Wierenga D, Engbers LH, van Empelen P, Duijts S, Hildebrandt VH, van Mechelen W. BMC Public Health 2013, 13(1):1190

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Abstract Background: Numerous worksite health promotion program (WHPPs) have been implemented the past years to improve employeesâ&#x20AC;&#x2122; health and lifestyle (i.e., physical activity, nutrition, smoking, alcohol use and relaxation). Research primarily focused on the effectiveness of these WHPPs. Whereas process evaluations provide essential information necessary to improve large scale implementation across other settings. Therefore, this review aims to: (1) further our understanding of the quality of process evaluations alongside effect evaluations for WHPPs, (2) identify barriers/facilitators affecting implementation, and (3) explore the relationship between effectiveness and the implementation process. Methods: Pubmed, EMBASE, PsycINFO, and Cochrane (controlled trials) were searched from 2000 to July 2012 for peer-reviewed (randomized) controlled trials published in English reporting on both the effectiveness and the implementation process of a WHPP focusing on physical activity, smoking cessation, alcohol use, healthy diet and/or relaxation at work, targeting employees aged 18-65 years. Results: Of the 307 effect evaluations identified, twenty-two (7.2%) published an additional process evaluation and were included in this review. The results showed that eight of those studies based their process evaluation on a theoretical framework. The methodological quality of nine process evaluations was good. The most frequently reported process components were dose delivered and dose received. Over 50 different implementation barriers/facilitators were identified. The most frequently reported facilitator was strong management support. Lack of resources was the most frequently reported barrier. Seven studies examined the link between implementation and effectiveness. In general a positive association was found between fidelity, dose and the primary outcome of the program. Conclusions: Process evaluations are not systematically performed alongside effectiveness studies for WHPPs. The quality of the process evaluations is mostly poor to average, resulting in a lack of systematically measured barriers/facilitators. The narrow focus on implementation makes it difficult to explore the relationship between effectiveness and implementation. Furthermore, the operationalization of process components varied between studies, indicating a need for consensus about defining and operationalizing process components.

Systematic review of implementation studies

Introduction Employees with unhealthy lifestyle behaviors and overweight or obese employees are less productive at work, show a decreased work ability and take more sick days compared to employees with a healthy lifestyle [1-4]. An unhealthy lifestyle can be characterized by one or more of the following behaviors; low physical activity levels, an unhealthy diet, smoking, frequent alcohol use and poor levels of relaxation (i.e. mental health and vitality). Because employed adults spend about half of their workday waking hours at their workplace, the worksite may be an effective setting to increase employees health and productivity [5-8]. Therefore, in the past decade, numerous worksite health promotion programs (WHPPs) have been conducted to improve employeesâ&#x20AC;&#x2122; health and lifestyle. However, not all of these programs were successful [5-7]. In order to change employeesâ&#x20AC;&#x2122; health and lifestyle, programs need to work as intended (and therefore avoid theory failure). Additionally, they must be effectively transferred from research to practice and be maintained over time [9]. Programs often fail to reach potential participants adequately due to a lack of adoption, communication, or program sustainability which could lead to low participation levels (in other words, program failure) [10]. Furthermore, research starts to recognize the importance of evaluating different implementation outcomes (such as recruitment, dose delivered, dose received, fidelity, satisfaction, and maintenance) and the contextual factors that hinder or facilitate the implementation of a WHPP [11-14]. These contextual factors are related to characteristics of the context, organization, implementer, program and participants [11, 12, 15]. Since WHPPs are often comprehensive interventions with multiple components, it is difficult to determine the overall level of implementation and investigate which specific intervention components have been successful [14, 16]. A review by Durlak and Dupre (2008) showed that higher levels of implementation improve program outcomes, suggesting that there should be an adequate focus on the implementation process [13]. However, most research focuses primarily on measuring the effects of a WHPP, with an occasional process evaluation performed after implementation. However, systematic process evaluations can produce valuable insights into the interpretation of the (lack of) effects of an intervention by identifying successful and unsuccessful program components, thereby allowing researchers to optimize their program [16-19]. Furthermore, process evaluations can help to identify barriers and/or facilitators influencing the implementation process, while taking into account the different actor levels at which

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

these factors play a role [16, 18]. These are valuable outcomes that can be used to improve program implementation in the future and across other settings. Hence, effect evaluations should be accompanied by systematic and real-time process evaluations [20, 21]. Murta et al. (2007) looked at the quality of process evaluations accompanying controlled trials studying individual based stress management interventions. They showed that process evaluations are often incomplete and not systematically conducted, meaning that they lacked a theoretical framework and were not planned prior to implementation [22]. However, the quality of the process evaluations that were included in this review was insufficient to make sound conclusions in terms of best predictors for successful interventions. Additionally, Murtaâ&#x20AC;&#x2122;s review only focused on evaluations of occupational stress management interventions published between 1977 to 2003. So their conclusion cannot directly be transferred to recent WHPP focusing on other lifestyle behaviors. In their review, Murta also concluded that the framework of Steckler and Linnan proved to be a useful tool to conduct process evaluations. But in this framework little attention has been given to the great variety of contextual factors that can either hinder or facilitate the implementation process. We therefore recently proposed a framework for a systematic and comprehensive process evaluation based on several theoretical frameworks to gain insight into the implementation process (figure 1) [11, 12, 14, 15, 23]. The four main aspects of this framework relate to determinants of implementation that may influence the implementation process, and the implementation process itself (i.e. adoption, implementation and continuation). These four main aspects are operationalized using a combination of the framework of Steckler and Linnan for process evaluations and the RE-AIM framework [14, 23]. So, in order to gain insight in the implementation process, eight different process components at three different actor levels (macro-level: organization and management; meso-level: implementer; microlevel: participant) need to be evaluated using a mixed methods approach [11]. Of these eight components, six components focus on implementation (reach, recruitment, dose delivered, dose received, fidelity, and satisfaction), one component (maintenance) on continuation and the eighth component (context) refers to the determinants of implementation. Since many implementation determinants can be identified, context is further defined by categorizing the barriers and/or facilitators into five main categories: 1) characteristics of the sociopolitical context, 2) characteristics of the organization, 3) characteristics of the implementer, 4) characteristics of the intervention program, 5) characteristics of the participant [11, 12, 15].

Systematic review of implementation studies

Some reviews have focused on process evaluations and relevant implementation barriers and/or facilitators [10, 12, 13, 22, 24, 25]. However, these reviews focused on domains other than WHPPs [12, 13, 24, 25], limited the identification of barriers and/or facilitators to those affecting WHPP participation levels [10], or focused only on process evaluations accompanying stress management interventions [22]. To our knowledge, there has been no systematic review that has examined process evaluations for worksite health promotion programs targeting lifestyle change among employees. The aim of this review was therefore to: (1) further our understanding of the quality of process evaluations alongside effect evaluations for worksite health promotion programs (WHPPs), (2) identify barriers/ facilitators affecting implementation, and (3) explore the relationship between effectiveness and the implementation process.

Innovation determinants

Implementation strategy

Innovation process

Characteristics of the sociopolitical context Adoption Characteristics of the organization Implementation Characteristics of the innovation

Continuation

Characteristics of the adopting person/user

Figure 1. Theoretical framework

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Methods Literature search and study selection For this systematic review, peer-reviewed studies were eligible for inclusion when they reported an effect evaluation as well as a process evaluation for a worksite health promotion intervention focusing on stimulating a healthy lifestyle (physical activity, nutrition, smoking, alcohol use and relaxation) published in English between 2000 and July 2012. The literature search was conducted in two steps. Step 1: a literature search was in the online databases Pubmed, EMBASE, PsycINFO, and the Cochrane Central Register of Controlled trials for peer-reviewed WHPP effect evaluations published in English from 2000 to July 2012. Searches included the following combination of keywords: (Randomized controlled trial OR controlled trial) AND (worker* OR employee* OR worksite OR work environment) AND (worksite health promotion OR lifestyle intervention) AND (absenteeism OR sickness absence OR body mass index OR lifestyle OR health behavior OR cholesterol level). To ensure no studies were overlooked this search string was repeated in each database with the addition of the following keywords for each lifestyle behavior separately: ‘physical activity’, ‘smoking’, ‘alcohol use’, ‘healthy diet’ and ‘relaxation’. Since we were only interested in studies focusing on change in actual behavior rather than change in attitude, social norm or self-efficacy keywords were added on outcome measures related to actual behavior change. All the keywords were specified for each database using Mesh or Thesaurus terms. The complete search strategy for each database and lifestyle behavior are presented in appendices 1. The inclusion criteria for the first stage of the selection process were: (1) a randomized controlled (RCT) or controlled trial (CT) published in English between 2000 and July 2012, (2) an evaluation of the effects of worksite health promotion interventions focusing on physical activity, smoking cessation, alcohol use, healthy diet and/ or relaxation at work, and (3) targeting employees aged 18-65 years. With respect to the second inclusion criteria it should be noted that studies investigating interventions that primarily aim to promote a healthy lifestyle as well as interventions that primarily aim to prevent musculoskeletal disorders of which the intervention contains one of the lifestyle components mentioned above were included. The first author (DW) performed an initial selection based on the titles and abstracts of all papers reporting on effect evaluation. The abstracts of the effect evaluations were presented to the fourth author (SD), who was blinded for authors, affiliations, journal and year of publication. In a consensus meeting between both authors (DW and SD), a final selection of effects evaluations was made. When

Systematic review of implementation studies

an abstract contained insufficient information or when the authors disagreed, the full paper was retrieved and read. When disagreement persisted, the second author (LE) was asked to decide about eligibility. Step 2: DW checked whether the selected effect evaluations were accompanied by a peer-reviewed process evaluation. The inclusion criteria for the second stage of the selection procedure were: (1) published in English between 2000 and July 2012, and (2) reporting on implementation/process outcomes. DW first contacted the corresponding authors by e-mail to ask whether the effect evaluation had been accompanied by a paper on implementation/process outcomes. If the corresponding author did not respond or could not provide the requested information, DW first checked the full text paper of the effect evaluation for references to a related process evaluation (in the reference list and/or text). If nothing was found, DW searched all four databases using the title of the study (if known), trial registration number and the names of all authors for an additional formative, implementation or process evaluation. If based on title and abstract it was unclear whether the paper included information on the implementation process of the intervention, the full text paper was retrieved and read. The full papers were retrieved for all eligible effect evaluation papers paired with a published process evaluation paper. These complete studies were independently assessed for eligibility by the first three authors (DW, LE and PE) before being included in the review. During this final round, LE and PE were blinded for authors, affiliations, journal and year of publication. All references used in eligible effect evaluations and process evaluations were checked by the first author (DW) for other relevant publications that might have been missed in the electronic search (â&#x20AC;&#x2DC;snow ballâ&#x20AC;&#x2122; procedure). Methodological quality assessment In order to answer the first research question, a methodological quality assessment was performed. The methodological quality of all papers included in the review was independently assessed by the first three authors (DW, LE and PE) using a checklist (Table 1). Disagreements between reviewers were discussed and resolved during consensus meetings. The criteria for the assessment of the included effect evaluations were based on the methodological guidelines for systematic reviews developed by the Cochrane Back Review Group [26]. These guidelines were developed for RCTâ&#x20AC;&#x2122;s studying low back pain. Some Cochrane criteria were therefore omitted or adapted to fit the studies included in this

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

review. Other reviews focusing on worksite interventions have used an adapted version of this guideline [5, 27]. The final criteria list consisted of three main categories: internal validity (n=8), descriptive criteria (n=5) and analysis (n=3). Items were scored as positive (+), unsatisfactory (+/-) or negative (-). If an item was not applicable, this was stated. The quality of studies was considered to be ‘above average’ if the overall validity score (V) was above 50%, and a score above 75% was considered to represent relatively good quality [26, 28]. The overall validity score (V) was based on the eight internal validity criteria (V1-V8). When items were not applicable (N/A), they were not included in the percentages. Given the absence of a standardized assessment form, we defined our own criteria for the methodological quality assessment of process evaluations on the basis of our proposed conceptual framework, shortly explained in the introduction and published elsewhere [11]. These criteria are described in detail in table 1. The final criteria list included nine items. Items were scored as positive (+), unsatisfactory (+/-) or negative (-). If an item was not applicable, this was stated. The quality of studies was considered to be ‘above average’ if the overall validity score (T) was above 50%, and a score above 75% was considered to represent relatively good quality. This validity score (T) was based on the nine internal validity criteria (T1-T9). Non-applicable items (N/A) were not included in the percentages. Table 1. Criteria list for the methodological quality assessment of the studies and definitions of the criteria Effect evaluations Internal validity / study design V1 Randomization procedure Positive if a random (unpredictable) assignment procedure sequence of subjects to the study groups was used and if there was a clear description of the procedure and adequate performance of the randomization V2 Similarity of companies V3 Similarity of study groups V4 Dropout

Positive if they controlled for variability in included companies Positive if the study groups were similar at the beginning of the study Positive if the percentage of dropouts during the study period did not exceed 20% for short-term follow-up (≤ 3 months) or 30% for long-term follow-up (> 3 months) and adequately described

V5 Timing of outcome measurement

Positive if timing of outcome assessment was identical for intervention and control groups and for all important outcomes assessments.

V6 Blinding

Positive if the person performing the assessments was blinded to the group assignment Positive if co-interventions were avoided or comparable. Positive when data on outcome was selected with standardized methods of acceptable quality

V7 Co-interventions V8 Outcome

Systematic review of implementation studies

Descriptive criteria D1 Eligibility criteria (in- and exclusion criteria) D2 Baseline characteristics

D3 Company characteristics D4 Intervention

D5 Follow-up Analysis A1 Sample size A2 Confounders A3 Intention to treat

Positive if in- and exclusion criteria of participants were specified Positive if an adequate description of the study groups was given for demographic variables: gender, age, type of work, hours a week working, education level, baseline main outcome measures Positive if an adequate description of the included companies was given (type of industry, organizational characteristics) Positive if an adequate description was given of the interventions(s): number of intervention aspects, type of interventions, frequency of sessions, intensity of intervention(s) Positive if a follow-up of 6 months or longer was described. Positive if an adequate sample size calculation was described Positive if the analysis controlled for potential confounders Positive if the intervention and control subjects were analysed according to the group belonging to their initial assignment, irrespective of non-compliance and co-interventions.

Process evaluations T1 Model used for evaluation Positive if a theoretical framework for the evaluation was used and adequately described. T2 Level of evaluation Positive if implementation was evaluated on 2 or more levels (i.e. macro, meso, micro) T3 Definition of outcome Positive if the definition of the outcome measures (process measure (process variables and barriers and/or facilitators) were accurately components) described T4 Reported process variables a. Positive if four or more process evaluation variables are evaluated (in process evaluation) b. Positive if barriers or facilitators on 1 or more levels are presented T5 Data collection T6 Timing of data collection T7 Quantitative outcome measures

Positive if 2 or more techniques for data collection were used (triangulation). Positive if measurements of barriers and/or facilitators were performed pre-, during and after implementation. Positive if data on quantitative outcome was selected with methods of acceptable quality and data on multiple process components was measured.

T8 Qualitative data

a. Positive if study design for qualitative data (theoretical framework, participant selection, setting, data collection) were adequately described b. Positive if qualitative data was analyzed by two researchers.

T9 Outcome related to implementation of intervention

Positive if outcomes (barriers and/or facilitators) are related to the quality of implementation

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Data extraction The first author (DW) extracted the data using a predefined format. For the effect evaluations, information was extracted about study design, company type, study population, intervention content, and intervention goal. In addition, although not the main aim of this review, the proportion of affected primary outcomes was extracted to provide some information about the effectiveness of the program. For the process evaluations information was extracted about data collection, methods, timing of measurements, level of evaluation, type of evaluation, linking effect to implementation outcomes, model for process evaluation, reported process components and reported barriers and/or facilitators. During data extraction, the process components reported in the studies were classified under the eight process components (recruitment, reach, dose delivered, dose received, fidelity, satisfaction and maintenance, and context) on the basis of the definitions proposed by Wierenga et al. [11, 14]. If a study used another framework/model with additional components, these were classified separately. The barriers and/or facilitators found in the studies were assigned to five categories based on the review by Fleuren et al. (2004): socio-political context, organization, implementer, program and participants [12, 15]. When something was unclear, advice was asked from the second (LE) and third author (PE). The complete data extraction form can be obtained from the corresponding author.

Results Study selection The initial computerized search identified 9112 articles (see flow chart in figure 2). Initial screening of the titles and abstracts produced 704 potentially relevant articles looking at all lifestyle behaviors. 8408 articles were excluded because they were not (R)CT related to a WHPP with the aim of changing employee lifestyles. We excluded 397 of the 704 potentially relevant articles on the grounds that articles were found in multiple databases (Pubmed, EMBASE, PsycINFO, Cochrane Central Register of Controlled Trials), or in the different searches based on lifestyle behavior. Interventions not related to any of the five health behaviors or not targeting employees were also excluded. A total of 307 full-text articles were retrieved and checked for related process evaluation articles, resulting in the exclusion of another 277 articles. The main reasons for exclusion were: no additional published process evaluation for the same study, not targeting employees, or no lifestyle component. Thirty studies reporting on both the effects and the 50

Systematic review of implementation studies

implementation process of the same WHPP were checked for eligibility. After reading the papers, studies were excluded because, contrary to expectations, they did not perform because,eight contrary to expectations, they did not perform a process evaluation. No articles were aadded process evaluation.theNo articlesinwere addedstudies. after Finally, snowballing thestudies references after snowballing references the selected twenty-two met the in the selected studies. Finally, twenty-two met 2; the selection2).criteria and were included in selection criteria and were included in this studies review (table Appendices this review (table 2; Appendices 2). 9112 of records identified through database searching

0 of additional records identified through other sources

8549 records excluded

9112 records screened

307 of records after duplicates removed and full-text articles assessed for eligibility

22 of studies included in the qualitative synthesis

285 of full-text articles excluded: - no additional process evaluation - intervention not related to any of the lifestyle behaviours - target population not employees

Figure ofstudy studyselection selectionprocess process Figure 2. 2. Flowchart Flowchart of Intervention and study characteristics

Intervention and study characteristics The second in table 2 presents the general study characteristics in order to give an overview The secondcolumn column in table 2 presents the general study characteristics in order to give an of the typeof of studies included. Half ofincluded. the studies Half were of conducted in the were USA [29-39], and theinother overview the type of studies the studies conducted the USA half in Europe mostly in the Netherlands [40, 41, 45, 47-49]. Nineteen (86%) studies [29-39], and [40-50], the other half in Europe [40-50], mostly in the Netherlands [40,used 41, a45, 47randomized controlled designused [29, 30, 33-36, 38-50] and three (14%) trial30, design 49]. Nineteen (86%)trial studies a randomized controlled triala controlled design [29, 33-36, 38[31,and 32, 37]. Fourteen studies reported on interventions targeting physical(63%) activitystudies (PA) [29-34, 50] three (14%)(63%) a controlled trial design [31, 32, 37]. Fourteen reported 40-44, 48-50]. Two targeting of these studies focused mainly(PA) on preventing musculoskeletal disorders [40, 44], on interventions physical activity [29-34, 40-44, 48-50]. Two of these studies but the intervention a PA component and so these studies were in the review. focused mainly onincluded preventing musculoskeletal disorders [40,included 44], but the intervention Twelve studies reportedand on the process evaluation of interventions targeting included a PA (55%) component so effect theseand studies were included in the review. Twelve studies healthyreported nutrition [29, 41, 45, 47-49]. Furthermore, five (23%) reported interventions (55%) on32-37, the effect and process evaluation of studies interventions targeting healthy for smoking and32-37, tobacco41, use45, [36-39, 41], and two (9%) studies on relaxation [46, 48]. None nutrition [29, 47-49]. Furthermore, five focused (23%) studies reported interventions of the twenty-two targeted alcohol use. for smoking and studies tobacco use [36-39, 41], and two (9%) studies focused on relaxation [46,

48]. None of the twenty-two studies targeted alcohol use. Methodological quality assessment 49

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Design effect evaluation: RCT Type of company: Construction Study population: Blue collar at risk for CVD Intervention content: Individually based lifestyle intervention by means of motivational interviewing Intervention goal: Change physical activity, diet and smoking behavior Proportion of affected primary outcomes: 8of 15a

Groeneveld et al. (2010, 2011) [32-34]

Netherlands

Model used for evaluation: Not described Measured process components: - Dose delivered: Number of sessions performed by implementer and number of items discussed during session - Dose received: Number of sessions attended by participants - Fidelity: Adherence to intervention protocol by implementers and the quality of the intervention delivery - Satisfaction: Participants opinion on the competence and skills of the implementer. Implementers opinion on their own competence and skills on delivery of intervention and overall difficulty of delivering intervention - Recruitment, Reach, Maintenance, Context: -

Design effect evaluation: RCT Type of company: University, railway transportation, airline, steel Study population: Blue and white collar Intervention content: Worksite participatory ergonomics program Intervention goal: Prevent low back pain and neck pain Proportion of affected primary outcomes: 2 of 14

Driessen et al. (2010,2011) [29-31]

Data collection method PE: Mixed methods Timing PE: Post-intervention PE evaluation levels: Meso, Micro Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Described for body weight

Model used for evaluation: Adjusted version of Linnan and Steckler Measured process components: - Recruitment: intervention departments, working groups formed, working group members for training, response rate baseline questionnaire - Reach: worksite visits by trainer, attendance rates implementers, attendance rates of implementers in training - Dose delivered: perceived implementation of the intervention according to implementers - Dose received: perceived implementation of the intervention of the employees - Fidelity: Extent to which the intervention was delivered as intended - Satisfaction: of implementers about implementation process and intervention, and of employees about the intervention - Context: Perceived barriers and/or facilitators to implementation of intervention - Maintenance: -

Data collection method PE: Mixed methods Timing PE: During and post-intervention PE evaluation levels: Meso, Micro Type of PE: Evaluation of interventions and implementation process Effect of implementation on outcome measure: Not described

Design effect evaluation: Randomized controlled trial (RCT) or Controlled trial (CT) Type of company Study population Intervention content Intervention goal Proportion of affected primary outcomes

Reference

Netherlands

Model used for evaluation Measured process components: Definition used in the study, - =not described

Process evaluation information Data collection method of process evaluation (PE) Timing of process evaluation (PE) Process evaluation (PE) evaluation levels: Macro, meso, micro Type of process evaluation (PE) Effect of implementation on outcome measure: Described or Not described

General study information

Table 2. Characteristics of the studies included in this review

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 Chapter 3

Model used for evaluation: Not described Measured process components: - Dose delivered: Amount of interventions made available during the intervention period - Dose received: Participation rates of employees in intervention - Recruitment, Reach, Fidelity, Satisfaction, Maintenance, Context: -

Model used for evaluation: Durlak & Dupre Measured process components: - Dose delivered: Amount of intervention delivered - Dose received: Amount of intervention received by participants - Fidelity: Degree to which the intervention was implemented as planned - Context: Implementation barriers - Recruitment, Reach, Satisfaction, Maintenance: -

Model used for evaluation: Not described Measured process components: - Dose received: Use of intervention - Satisfaction: Satisfaction with the intervention components - Recruitment, Reach, Dose delivered, Fidelity, Maintenance, Context: -

Model used for evaluation: Not described Measured process components: - Recruitment: Strategies to approach and attract employees - Satisfaction: Experiences during intervention - Context: Benefits/positives and problems/barriers associated with participation in intervention - Reach, Dose delivered, Dose received, Fidelity, Maintenance: -

Data collection method PE: Qualitative Timing PE: Post-intervention PE evaluation levels: Micro Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Not described

Data collection method PE: Qualitative Timing PE: Post-intervention PE evaluation levels: Micro Type of PE: Evaluation of interventions and implementation process Effect of implementation on outcome measure: Not described

Dishman& Wilson Design effect evaluation: RCT et al. (2009, 2010) Type of company: Retail [37, 38] Study population: White collar Intervention content: Social-ecologic intervention by personal and United States team goal-setting Intervention goal: Increase leisure-time physical activity Proportion of affected primary outcomes: 3 of 3

Yap et al. (2009, 2010) [39-41]

Design effect evaluation: CT Type of company: Manufacturing plant Study population: Blue collar Intervention content: Tailored e-mail intervention Intervention goal: Increase intentional physical activity Proportion of affected primary outcomes: 2 of 2

Gilson et al. (2007, Design effect evaluation: RCT 2008) [42, 43] Type of company: University Study population: White collar Intervention content: Unites Kingdom Route and task-based walking Intervention goal: Improve work day step counts and health status Proportion of affected primary outcomes: 1of 4

United States

Design effect evaluation: RCT Type of company: Bus garages Study population: Blue collar Intervention content: Environmental worksite obesity prevention intervention Intervention goal: Improve healthful food choices an physical activity levels Proportion of affected primary outcomes: 4 of 20

Data collection method PE: Quantitative Timing PE: Post-intervention PE evaluation levels: Micro Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Not described

French et al. (2010) [35, 36]

Systematic review of implementation studies

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

54 Model used for evaluation: Not described Measured process components: - Context: Prognostic factors for adherence to intervention - Recruitment, Reach, Dose delivered, Dose received, Fidelity, Satisfaction, Maintenance: -

Data collection method PE: Quantitative Timing PE: Pre-intervention PE evaluation levels: Micro Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Not described

Andersen et al. (2011) [51-53]

Denmark

Design effect evaluation: RCT Type of company: Office Study population: White collar Intervention content: Progressive resistance training (exercise program) Intervention goal: Relieve neck/shoulder pain Proportion of affected primary outcomes: 4 of 6

Model used for evaluation: RE-AIM Measured process components: - Recruitment: Amount of recruited settings for participation and number that agreed to take part, representatives of participating settings (Adoption) - Reach: Participation rates, representativeness of participants, characteristics of participants and non-participants - Dose delivered: Implemented intervention components - Dose received: Use of intervention components - Fidelity: Extent to which the program was delivered as intended at program and individual level (Implementation) - Maintenance: Extent to which the behavior or policy of interest maintained over the long term at individual and institutional level - Context: Management support, job/task factors (physical and psychological demands of specific jobs), environmental factors (physical work environment and the social-organizational environment - Effectiveness: Behavioral outcomes and other outcomes including impact on quality of life - Satisfaction: -

Data collection method PE: Mixed methods Timing PE: Pre-, during, and post-intervention PE evaluation levels: Macro, Meso, Micro Type of PE: Evaluation of interventions and implementation process Effect of implementation on outcome measure: Described for Body Mass Index

Lemon, Estabrook Design effect evaluation: RCT et al. (2010, 2011) Type of company: Hospital [49, 50] Study population: White collar Intervention: Ecological and environmental multilevel United States intervention Intervention goal: Prevent weight gain by targeting healthy eating and physical activity Proportion of affected primary outcomes: 0 of 1

United States

Model used for evaluation: Integrative model Measured process components: - Context: Management support, job/task factors (physical and psychological demands of specific jobs), environmental factors (physical work environment and the social-organizational environment - Recruitment, Reach, Dose delivered, Dose received, Fidelity, Satisfaction, Maintenance: -

Design effect evaluation: CT Type of company: Manufacturing Study population: Blue and white collar Intervention content: Environmental weight management intervention Intervention goal: Improve physical activity and healthy eating Proportion of affected primary outcomes: 6 of 8

Data collection method PE: Mixed methods Timing PE: During and post-intervention PE evaluation levels: Macro, Meso, Micro Type of PE: Evaluation of interventions and implementation process Effect of implementation on outcome measure: Not described

Goetzel, DeJoy, Wilson et al. (2007, 2009, 2010, 2011) [44-48]

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 Chapter 3

Model used for evaluation: Not described Measured process components: - Reach: Seeing or reading the intervention materials (use of intervention) - Dose delivered: Documentation of intervention activities by implementers - Dose received: Average intervention exposure per employee - Context: Worksite characteristics - Recruitment, Fidelity, Satisfaction, Maintenance: -

Model used for evaluation: Not described Measured process components: - Dose delivered: Number of interventions delivered - Dose received: Receiving intervention and materials - Fidelity: Extent to which the intervention was implemented - Satisfaction: Participant satisfaction with intervention - Recruitment, Reach, Maintenance, Context: -

Data collection method PE: Qualitative Timing PE: Post-intervention PE evaluation levels: Micro Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Not described

Beresford et al. Design effect evaluation: RCT (2000, 2001, 2010) Type of company: Hospitals, educational, [58-60] governmental, professional agencies, construction, manufacturing, financial, retail, wholesale, service United States Study population: Blue and white collar Intervention content: Environmental and individual strategies Intervention goal: Increase fruit and vegetable intake Proportion of affected primary outcomes: 2 of 2b

Sorensen, Hunt et Design effect evaluation: RCT al. (2007, 2010) Type of company: Construction [61, 62] Study population: Blue collar Intervention content: Tailored telephonedelivered and mailed intervention United States Intervention goal: Promote smoking cessation and increase fruit and vegetable consumption Proportion of affected primary outcomes: 3 of 3

United States

Model used for evaluation: Not described Measured process components: - Reach: Program awareness - Dose delivered: Delivered interventions - Dose received: Program participation - Context: Management contacts regarding organizational changes, factors that influences implementation - Recruitment, Fidelity, Satisfaction, Maintenance: -

Data collection method PE: Mixed methods Timing PE: Post-intervention PE evaluation levels: Meso, Micro Type of PE: Evaluation of interventions and implementation process Effect of implementation on outcome measure: Not described

Sorensen, Hunt (2005, 2007) [56, 57]

Design effect evaluation: RCT Type of company: Manufacturing Study population: Blue collar Intervention content: Worksite intervention targeting fruit and vegetable consumption, red meat consumption, multivitamin use and physical activity Intervention goal: Cancer prevention Proportion of affected primary outcomes: 2 of 4

Model used for evaluation: Not described Measured process components: - Dose delivered: Time and visits made to worksite - Dose received: Participation rates - Satisfaction: General opinions regarding process, experiences of the intervention, - Context: Benefits or difficulties experiences of the project, facilitating factors for implementation, support and time needed for development, organization of work tasks during intervention - Recruitment, Reach, Fidelity, Maintenance:-

Data collection method PE: Mixed methods Timing PE: Pre-, during, and post-intervention PE evaluation levels: Meso, Micro Type of PE: Evaluation of interventions and implementation process Effect of implementation on outcome measure: Not described

Haukka, Pehkonen Design effect evaluation: RCT et al. (2009, 2010) Type of company: Municipal kitchens [54, 55] Study population: Blue collar Intervention content: Participatory ergonomics intervention Finland Intervention goal: Decrease physical and mental workload Proportion of affected primary outcomes: 0 of 8

Systematic review of implementation studies

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

56 Model used for evaluation: Not described Measured process components: - Dose delivered: Number of delivered interventions - Dose received: Perception of received intervention components and materials (Engagement in intervention components) - Satisfaction: Helpfulness of intervention - Recruitment, Reach, Fidelity, Maintenance, Context: -

Model used for evaluation: Not described Measured process components: - Dose delivered: Amount of intervention delivered - Dose received: amount of intervention received, participation rates, motivation for participation in intervention, program awareness - Recruitment, Reach, Fidelity, Satisfaction, Maintenance, Context: Model used for evaluation: Not described Measured process components: - Reach: Program awareness - Dose delivered: Amount of intervention delivered - Dose received: amount of intervention received, participation rates, motivation for participation in intervention - Recruitment, Fidelity, Satisfaction, Maintenance, Context: -

Data collection method PE: Quantitative Timing PE: Post-intervention PE evaluation levels: Meso, Micro Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Described for smoking cessation

Data collection method PE: Quantitative Timing PE: During and post-intervention PE evaluation levels: Meso, Micro Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Not described Data collection method PE: Mixed methods Timing PE: Post-intervention PE evaluation levels: Micro Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Described for smoking cessation

Stoddard, Hunt et Design effect evaluation: RCT al. (2003, 2005) Type of company: Retail [67, 68] Study population: White collar (teens) Intervention content: Behavioral tobacco control intervention United States Intervention goal: Decrease smoking prevalence Proportion of affected main outcomes: 0 of 2

Volpp, Kim et al. (2009, 2011) [69, 70]

United States

Design effect evaluation: RCT Type of company: Multinational Study population: Not described Intervention content: Offering financial incentives for smoking cessation Intervention goal: Improve smoking cessation rates Proportion of affected primary outcomes: 1 of 1

Design effect evaluation: CT Type of company: Trucking terminals Study population: Blue collar Intervention: Telephone and print-delivered intervention Intervention goal: Promote tobacco use cessation and improve weight management by healthy nutrition Proportion of affected primary outcomes: 1 of 2

Netherlands

Sorensen, Quintiliani et al. (2010) [65, 66]

Model used for evaluation: Not described Measured process components: - Satisfaction: managers opinions on intervention - Context: difficulties with implementation, perceived benefits of intervention - Recruitment, Reach, Dose delivered, Dose received, Fidelity, Maintenance: -

Design effect evaluation: RCT Type of company: Governmental and large companies Study population: White collar Intervention: Environmental interventions including labeling of healthy foods, food supply program and educational program about healthy food Intervention goal: Increase availability and knowledge of healthy foods Proportion of affected primary outcomes: 1 of 3

Steenhuis et al. (2004) [63, 64]

Data collection method PE: Qualitative Timing PE: Post-intervention PE evaluation levels: Meso Type of PE: Evaluation of interventions and implementation process Effect of implementation on outcome measure: Not described

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 Chapter 3

Design effect evaluation: RCT Type of company: Worksite cafeterias of hospitals, universities, police departments, companies Study population: White collar Intervention content: Environmental intervention introducing a small hot meal in addition to the existing size and a proportional pricing strategy in cafeterias Intervention goal: Stimulate workers to replace their larger meal with a smaller meal Proportion of affected primary outcomes: 0 of 2

Vermeer et al. (2011) [73, 74]

Strijk et al. (2011, Design effect evaluation: RCT 2012) [75-77] Type of company: Academic hospital Study population: White collar Intervention content: Vitality intervention Netherlands consisting of weekly yoga and workout session, weekly unsupervised aerobic exercise, free fruit during sessions and three visits of personal vitality coach Intervention goal: Improve lifestyle behaviors Proportion of affected primary outcomes: 3 of 10

Netherlands

Sweden

Design effect evaluation: RCT Type of company: Information technology, media Study population: White collar Intervention content: Web-based health promotion and stress management training Intervention goal: Decrease unwanted stress and promote health and recovery Proportion of affected primary outcomes: 5 of 5

Hasson et al. (2005, 2010) [71, 72]

Model used for evaluation: Not described Measured process components: - Dose received: Frequency of replying to the screening tool - Context: Factors that determine use of program - Recruitment, Reach, Dose delivered, Fidelity, Satisfaction, Maintenance: -

Model used for evaluation: Baranowski & Stables and Rogers Measured process components: - Recruitment: Attracting agencies, implementers, or potential participants for corresponding parts of the program - Dose received: Extent to which participants view or read the materials that reach them - Fidelity: Extent to which the program is implemented as designed - Maintenance: Keeping participants involved in the programmatic and data collection, and extent to which participants continue to do any of the activities - Context: Aspects of the environment of an intervention - Contamination: extent to which participants received interventions from outside the program and the extent to which the control group receives the treatment - Resources: Materials or characteristics of agencies, implementers, or participants necessary to attain project goals - Reach, Dose delivered, Satisfaction:Model used for evaluation: Steckler and Linnan Measured process components: - Reach: proportion of workers participating in intervention - Dose delivered: number of intervention components delivered by implementers - Dose received: Extent to which the workers were engaged in the intervention (attendance rates) - Fidelity: Extent to which the intervention was implemented as planned - Satisfaction: Workers attitude towards the intervention - Context: Organizational and environmental factors concerning the intervention - Recruitment, Maintenance:-

Data collection method PE: Qualitative Timing PE: During PE evaluation levels: Micro Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Not described

Data collection method PE: Mixed methods Timing PE: Pre-, during and post-intervention PE evaluation levels: Meso, Micro Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Not described

Systematic review of implementation studies

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Design effect evaluation: RCT Type of company: Hospitals, cleaning companies, large business with in-house cleaning services Study population: Blue collar Intervention content: Physical coordination training Intervention goal: Improve physical coordination training or cognitive behavioral resources Proportion of affected primary outcomes: 3 of 6

Jorgensen et al.(2011, 2012) [81-83]

Model used for evaluation: Steckler and Linnan Measured process components: - Recruitment: Sources and procedures used to approach and attract potential participants, the number of randomized OPs, and the number of employees that filled out the baseline questionnaire - Reach: Number of employees who attended the counseling sessions, reason for missed counseling sessions and the percentage of drop-outs including reason - Dose delivered: The number of intervention materials or components actually delivered by OPs, and the duration and form of the counseling sessions - Dose received: The extent to which participants use materials, resources, or techniques recommended by program - Fidelity: The extent to which the intervention was delivered as planned: if OPs adhered to the guideline and adequately performed behavior change counseling - Satisfaction: Participants attitudes toward the content, use and limitations of the guideline (OP), or toward the intervention, materials and OP (employee) - Context: Organizational characteristics that affect intervention implementation, including physical, social, political, and economic features - Maintenance: Model used for evaluation: Adapted version Steckler & Linnan Measured process components: - Recruitment: Procedures used to approach and attract participants - Dose delivered: intervention delivered - Dose received: adherence to intervention (attendance rates) - Fidelity: Quality of intervention delivery - Context: Unanticipated events at the work place - Reach, Satisfaction, Maintenance: -

Data collection method PE: Quantitative Timing PE: During intervention PE evaluation levels: Meso Type of PE: Evaluation of interventions Effect of implementation on outcome measure: Not described

RCT: Randomized controlled trial; CT: Controlled trial; PE: process evaluation; Macro level: company/management; Meso level: Implementer; Micro: employee a 3 outcomes were significant at short and long term, 3 outcomes only at short term, 2 outcomes only at long term b both outcomes were significantly improved at short term, however they were not statistically significant at long term but still improved in favor of intervention group

Denmark

Netherlands

Design effect evaluation: RCT Type of company: University, bank, nursing home, spice factory, packaging company, municipality, consumer goods company Study population: Blue and white collar Intervention content: Occupational health guideline Intervention goal: preventing weight gain by increasing PA, decreasing sedentary behavior, increasing fruit consumption or reducing energy intake derived from snacks Proportion of affected primary outcomes: 2 of 6

Verweij et al. (2011, 2012) [78-80]

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 Chapter 3

Systematic review of implementation studies

Methodological quality assessment Table 3 shows the methodological quality scores of the included studies. Initial disagreement between the reviewers about 50 of the 448 effect evaluation criteria (11%) and 46 of the 207 process evaluation criteria (22%) were mainly attributable to the interpretation of the items. The differences concentrated primarily on drop-out (V4), company characteristics (D3), intention to treat (A3), data collection for process evaluation (T5), and the timing of data collection for process evaluation (T6). Full agreement was reached after two discussion sessions, thereby completing the scoring process. The methodological quality scores for the effect evaluations of the studies ranged from 37.5% to 100%. Fifteen (68%) effect evaluations were considered to be â&#x20AC;&#x2DC;above averageâ&#x20AC;&#x2122; (>50%) [29-33, 37, 39-41, 43, 44, 46, 48, 48-50]. Only six of these effect evaluations (27%) were relatively good (quality score >75%) [31, 40, 41, 43, 48, 49]. The quality scores for the process evaluations ranged from 12.5% to 100%. Eight process evaluations (36%) were relatively good (quality score >75%) [51-58]. Only three (14%) studies scored relatively good (quality score >75%) on both the process as well as the effect evaluation with respect to their methodological quality [51, 57, 58]. In four (18.2%) studies the methodological quality of both the effect as well as the process evaluation was poor (quality score 50% or less) [36, 38, 42, 45]. No relation was observed between the quality of effect evaluations and process evaluations.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

60 + +/+ + + + +

+ + + + + + + + + -

+ +/+ + + +/+ +/+/+ + + + +

+ + + +/+/+/+ + + +

+ +/+/+/+ + +/+ +/-

+ + + + +/+ + + + + + + -

+ + + + + +/N/A + + + + + + + +

+ N/A + + N/A N/A N/A N/A + N/A + N/A + N/A

+ + + + +/+/+ + +/+ -

100 31,25 83,3 66,67 75 25 12,5 62,5 50 61,1 68,75 88,89 81,25 100 56,25

Methodological quality assessment criterion – process evaluation T1 T2 T3 T4 T5 T6 T7 T8 T9 Validity score (T) in % + + + + +/- + + + + 94,4 + + + + + N/A + 75 N/A +/- +/- N/A 14,3 + + + + + + + + + 100 N/A + 12,5 + +/- N/A +/- 25 + + + +/- + + + + 83,3

Controlled trial * two effect articles scored together ** two process evaluations scored together *** three effect articles scored together and 2 implementation articles scored together N/A, not applicable. +, positive. +/-, not sufficient; -, negative. All trials are randomized trials except for the trials indicated with a superscript ‘a’. The maximum score for methodological quality of effect evaluations is 8 (based on validity section V1-V8). The maximum score for methodological quality of process evaluation is 9 (based on section T1-T9).

Methodological quality assessment criterion – effect evaluations V1 V2 V3 V4 V5 V6 V7 V8 Validity score (V) in % Driessen et al. (2010, 2011) ** + + +/- +/- + + + + 87,5 Groeneveld et al. (2010, 2011) * + N/A + +/- + +/- N/A + 83,3 French et al. (2010) * +/- + +/- N/A + N/A +/- + 75 Dishman& Wilson et al. (2009, 2010) + +/- +/- + +/- + 56,25 Yap et al. (2009, 2010) * N/A N/A + + + N/A N/A + 100 Gilson et al. (2007, 2008) + N/A + +/- + 50 N/A +/- +/- + N/A +/- + 75 Goetzel, DeJoy, Wilson et al. (2007, 2009-2011) a *** Lemon, Estabrook et al. (2010-2011) +/- + +/- + + + 62,5 Andersen et al. (2011) + N/A + + + + + + 100 Haukka, Pehkonen et al. (2009,2010) + + + + N/A + 71,4 Sorensen, Hunt et al. (2005, 2007) + + +/- +/- + 50 Beresford et al. (2000, 2001, 2010) * + + + + 50 Sorensen, Hunt et al. (2007, 2010) N/A + + + +/50 Steenhuis et al. (2004) +/- +/- + + 37,5 Sorenson, Quintiliani et al. (2010) _ + + + +/- + 56,25 Stoddard, Hunt et al. (2003, 2005) + + N/A + +/50 Volpp, Kim et al. (2009, 2011) + + + + + 62,5 Hasson et al. (2005, 2010) + + + + + 62,5 Vermeer et al. (2011) +/- +/- +/- + 31,25 Strijk et al. (2011, 2012) + + + +/- + + + + 93,75 Verweij et al. (2011, 2012) + + + + + + + 87,5 Jorgensen et al. (2011, 2012) + + + +/- + + 68,75

First author (Year)

Table 3. Overall scores of the methodological quality of the included studies

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 Chapter 3

Systematic review of implementation studies

Process evaluation design The third and fourth columns in table 2 list the characteristics of the included process evaluations. Most studies (50%) used a mixed methods approach (qualitative and quantitative) to look at the implementation process [51-56, 58-62]. With regard to the number of measurements, ten (45%) studies conducted a post-process evaluation only [29, 55, 59, 61-67]. Three (14%) studies collected process evaluation information at three points in time: before, during and after the intervention [54, 56, 60]. Only two (9%) studies evaluated the implementation process at all three actor levels (the macro-, meso- and micro-levels, see above) [53, 60]. The majority (50%) of the studies evaluated the implementation process at both the meso- and micro-levels [51, 52, 54-59, 61, 67, 68]. A minority of the studies (n=8; 36%) reported using a theoretical framework to guide their process evaluation [47, 51-53, 57, 58, 60, 69]. Of these eight studies, four used the framework of Steckler and Linnan (or an adapted version of that framework) [51, 57, 58, 69], and the remaining four studies applied either the integrative model [53], the REAIM model [60], a model based on Durlak and Dupre[52], or a model based on Baranowski and Stables, in combination with Rogersâ&#x20AC;&#x2122; framework for the diffusion of innovations [56]. It should be noted that the integrative model is originally a model for the development of workplace environmental interventions and not specifically for process evaluations [53]. Reporting of process evaluation components The fourth column of table 2 lists the predetermined process components that were measured. Seven of these components (reach, recruitment, dose delivered, dose received, fidelity, satisfaction and maintenance) measure the degree of implementation. The eighth component, context, maps the barriers and/or facilitators that affect implementation. The average number of process evaluation components found in the studies was 3.9, ranging from 1 to 8. Ten (45%) studies evaluated fewer than four process components [29, 53, 6264, 66-68, 70, 71]. Six (27.5%) studies reported five or more process components [51, 56-58, 60, 69]. The studies focused mainly on dose received (82%) [29, 51, 52, 54-63, 65, 67-69, 71], dose delivered (68%) [29, 51, 52, 54, 55, 57-62, 65, 67, 68], and context (68%) [52, 5457, 60, 61, 64, 66, 69-74]. Fewer than half of the studies (41%) looked at fidelity [51, 52, 56-60, 65]. Nine (41%) studies measured satisfaction [51, 54, 57-59, 63-66]. Relatively few studies reported on reach (n=7; 32%) [51, 55, 57, 58, 60, 61], or recruitment (n=5; 23%) [56, 58, 60, 64, 69]. In addition, two (9%) studies looked at maintenance [56, 60].

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Implementation barriers and/or facilitators Next, we identified barriers and/or facilitators which can affect implementation. However, the barriers and/or facilitators were often not systematically measured by means of questionnaires or interviews and often only observed and documented on the basis of the researchers’ experience. Nevertheless, fifty-four different barriers and/or facilitators were obtained from nineteen (86%) studies (table 4) [52, 54-57, 59-64, 66, 68-74]. On average, the studies described 6.5 [range 1-22] barriers and/or facilitators. None of the studies reported a context analysis prior to implementation. The majority of barriers and facilitators in the five categories as a whole were reported only in a single study. Fewest barriers and/or facilitators were reported in the category ‘characteristics of the socio-political context’. Only two (9%) studies reported a barrier or facilitator in this category (these studies are listed in table 4). The largest number of barriers and/or facilitators (n=17) was found in the category ‘characteristics of the organization’. Management support was found most frequently, having been mentioned in eight studies (36%). Strong management support was described as a facilitator, whereas unbalanced or lack of management support was found to be a barrier. The definition of management support varied widely. Another frequently reported barrier was the lack of financial, staffing or material resources (n=5 studies; 26%). In the category ‘characteristics of the intervention’ fifteen barriers and/ or facilitators were identified, with ‘compatibility of intervention with the organization’ being the most commonly reported facilitator in eight studies (42%). The facilitator ‘relative advantage of the intervention’ was reported in five (26%) studies. In the category ‘characteristics of the implementer’ twelve barriers and/or facilitators were identified; the factor ‘available time of the implementer’ was most commonly reported (n=5; 26%). In the category ‘characteristics of the user’, eight barriers and/or facilitators were identified with ‘time constraints’ was frequently reported as a barrier to participation. Furthermore, we found that high work demands and a high workload were also a barrier to participation.

Systematic review of implementation studies

Table 4. Reported barriers (B) and/or facilitators (F) in the studies included in this review Main categories

Description of the determinants for implementation

Characteristics of 1. Compatibility of program with societal developments (attention for health in the socio-political society) [74] context 2. Competitive business environment [53] Characteristics of 3. Organizational reorganization: reorganization due to take over by another the organization company [68] 4. Lack of resources: financial, personnel, material (e.g., equipment, facilities) resources or lack of space or facilities [54, 57, 66, 72, 74] 5. Organizational culture: (a) Senior leaders emphasized the need to implement the intervention keeping the organizational culture in mind [53] (b) Intervention did not fit the organizational culture [55, 72] (c) The organizational culture emphasized goal setting and tracks progress towards achieving goals [53] (d) Worksite culture supported social interaction among workers and between workers and managers [55] 6. Organizational size: (a) In a large organization (1000+ employees) there were numerous competing priorities and it was challenging to maintain visibility [60] (b) In a small organization (<500 employees) it is challenging to assemble a critical mass of potential participants for participation in the intervention [60] (c) Small organizations tend to receive more intervention components per employee than larger organizations [61] 7. Amount of company locations: Different company locations at which the intervention needs to be delivered [74] 8. Organizationâ&#x20AC;&#x2122;s awareness of perceived benefits of investment [74], and awareness of relevance and economics of health and employee wellness [53, 60] 9. Company image: the program gives the organization a positive image since it shows that the organization cares about their employees [66] 10. Perceived responsibility of employer towards workers health and wellbeing [74] 11. High staff turnover rate among employees made it difficult to provide adequate exposure to the intervention [68, 69] 12. Good collaboration between persons/ structures/ services/ collaborative partners within or outside departments and organizations [54, 66, 72] 13. Conflicting relationship between management and researchers[68] 14. General good organizational support for health promotion [53] 15. Poor psychosocial work environment consisting of the following the subcomponents: influence at work, work pace quantitative work demands, interpersonal relations [70] 16. History of social interaction: Worksite has a history of bringing employees together for social activities and a history of positive social interaction between worker and management [55] 17. Management support: (a) Strong (upper) management support for intervention and general health promotion efforts at the organization [55, 60, 68, 72, 73] (b) Unbalanced management support for intervention [55, 68] (c) Managers encouraging workers to attend intervention [55] (d) Experienced management support are different for junior employees and senior employees [64] (e) formal approval of upper management before start of intervention [57] (f) Lack of perceived management support by implementers on site [74] (h) Management commitment and willingness to provide employees with release time from their usual duties to attend intervention [55]

B/F F B B B F B F F

B B F B F F F B F B F B

F B F B F B F

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Table 4. Continued Characteristics of the organization

18. Management participation and engagement: (a) Active management participation and involvement alongside and with workers [55, 73] (b) Active management engagement in planning [55] 19. Relationship between management and employees: Respectful relationship between management and worker [55]

Characteristics of the implementer

20. Job position of implementer:[74] (a) Self-employed (advantage of managing his or her own time) (b) Internal position (facilitating in scheduling appointments) (c) external position 21. (Perceived) Support for implementers:[74] (a)Poor support from co- implementers (b) Support for implementers to change their routines (applicable when implementer is an occupational physician) [74]

Characteristics of the intervention program

F F F

F F B B F

22. Collaboration between implementers: lack of possibility to exchange experiences between implementers [74] 23. Available time of implementer: (a) Sufficient time available to implement intervention [56, 66, 72, 74] (b) The intervention involved extra work on top of the heavy workload of the regular duties of the implementer [66] (c) planning difficulties of implementers with planning al contacts in the intervention period [59] 24. Expectations of implementer: implementers expectations were met [74] 25. Absence of a project leader/ leading person/ ambassador[72] 26. Implementersâ&#x20AC;&#x2122; compliance with intervention protocol [52] 27. Staff turnover among implementers: drop out of implementers (without replacing them) [69, 72] 28. Absence of decision maker among implementers: among the implementers there lacked a person who was entitled to make decision at department level [72] 29. High perceived Level of control for intervention delivery by provider/implementer [60] 30. Low level of engagement of implementers in planning, promoting and providing feedback on intervention activities [55] 31. Personnel characteristics of implementer: sufficient skills, knowledge and competence to implement guideline or intervention correctly [55, 59, 74] 32. Degree of rewards: either financial reimbursement or other incentives [53, 68] 33. Compatibility and alignment of intervention with: (a) organizations mission statement/business goals/ institutional policy change [53, 60, 68, 74] (b) policy, culture, norms and current practices of organization [56, 58, 66, 72] (c) Ease of integration of intervention in working live [64]

34. The intervention fit implementers current work [74] 35. Intervention is part of the worksites integral health policy and seen as a pilot for future health promotion policy instead of independent project [57] 36. Relative advantage: intervention is advantageous compared to the current situation and no negative consequences were observed and the company, managers, implementers and participants benefit from participation [54, 56, 66, 72, 74]

F F

37. Time: Project took more time than expected due to high workload of administration and planning[74] 38. Complexity: Intervention was not too difficult or complex to implement and execute [56, 59, 72, 74] 39. Observability of positive results of the intervention [74]

F B B F B F B B F B F F F F F

F F

Systematic review of implementation studies

Characteristics of the intervention program

Characteristics of the participant

40. Risk and uncertainty level/Triability: the degree to which an innovation can be adopted/implemented with minimal risk [56] 41. Conflicting interest between worksite and intervention [66] 42. Timing of intervention activities: intervention activities coincide with scheduled breaks [68] 43. Technical problems (e.g., equipment breaks down) [54, 69] 44. Degree of incorporation of program communication and interventions into already established communication channels or existing worksite events/meetings [53, 55] 45. Presence of advisory board: well-functioning advisory board [55] 46. Ease of access to the program by bringing the program to participants and making participation free or inexpensive [53] 47. Needs of participants: [a] Positive personal preferences for program [63] [b] No need for intervention (e.g., already being healthy) [74] [c] Positive program expectation [71], [d] Prior failed attempts to maintain a healthy lifestyle [62] 48. Current workload and work structure/schedules: volume of daily tasks, overtime work, shift work, part-time work, irregular work schedules, shifts of different lengths, time-pressures [53, 60, 64, 68] 49. Work demands: Workers were unable to participate since they could not leave their work due to work demands, obligations and limited free time and flexibility to leave immediate work area [55, 57, 60, 64] 50. Time constraints of participants: lack of time, time constraints and willingness to make time to participate at work [53, 54, 57, 62, 63, 74] 51. Amount of peer leaders: Few peer leaders due to geographically separated worksites made it difficult to establish group cohesion[68] 52. Lack of social support: [a] No interaction with the entire workforce to build worksite-wide social norms and social support) [68] [b] Peer support: difficult to engage in behavior not considered normal by peers [64] 53. Lack of motivation of workers to participate in intervention[54] 54. Participants self-efficacy: Low to medium self-efficacy is a barrier for participation [70]

F B F B F F F F B F B B

B B B B B B

Degree of implementation and program effectiveness Only seven (31.8%) studies evaluated the association between implementation and program outcomes [52, 58-62, 67]. These studies generally found that the level of implementation in terms of high fidelity and dose was positively associated with a positive change in their primary outcome measures (body weight, waist circumference, Body Mass Index, fruit and vegetable intake, physical activity levels, smoking cessation). This was analyzed by means of linear or logistic regression analysis [52, 58-62, 67], latent growth modeling [52], dose or as-treated analysis [52, 58-62, 67], analysis of variance [52, 58-62, 67], mixed model logistic regression analysis or linear mixed model regression analysis [52, 58-62, 67], chi square tests [52, 58-62, 67], and multilevel linear regression analysis [52, 58-62, 67]. Of the seven studies, three studies found that higher participation levels (dose received) significantly and

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

positively impacted their primary outcome measures (body weight, waist circumference, physical activity levels and smoking cessation) [52, 58, 62]. In the study on personal and team goal-setting with the aim of increasing leisuretime physical activity, Dishman and Wilson found that high implementation groups had a greater increase in vigorous physical activity over the three time points than did the low implementation sites. For dose received no significant change over time was observed for any of the outcome measures (walking, moderate physical activity, vigorous physical activity) [52]. The results of the study of Volpp and Kim on improving smoking cessation rates by offering financial incentives for smoking cessation, the results showed that the attendance levels to the cessation programs were higher among quitters than among non-quitters in the intervention group. Additionally, quitters in the intervention group attended more than two times as many sessions as non-quitters in the intervention group [52, 58-62, 67]. Finally, in the study of Verweij on preventing weight gain by implementing an occupational health guideline, it was found that employees with higher attendance and satisfaction levels significantly reduced their waist circumference and body weight compared to employees with lower attendance and satisfaction levels[52, 58-62, 67].

Discussion The aims of this review were to (1) further our understanding of the quality of process evaluations alongside effect evaluations for WHPPs, (2) identify barriers/facilitators affecting implementation, and (3) explore the relationship between effectiveness and the implementation process. Prior to discussing the main findings, it should be noted that only a small number of studies that evaluated the effectiveness of a WHPP included a process evaluation relating to the implementation of that WHPP. Of the 307 effect evaluations identified in this review, only twenty-two (7.2%) published an additional process evaluation. With respect to the first aim we can conclude that the quality of process evaluations alongside effect evaluations of WHPPs was generally poor to average and a systematic approach was lacking. This makes it difficult to draw firm conclusion about reliably identifying implementation barriers and/or facilitators (second aim) or about the relation between effectiveness and implementation (third aim). Murta et al. (2007) found the same difficulties in a systematic review. They concluded that process evaluations alongside workplace stress-management interventions

Systematic review of implementation studies

were mainly poor to average, also making it difficult to identify reliable determinants of effective intervention implementation [22]. The process evaluations covered by our review lacked a theoretical basis, and the most frequently measured process components were dose delivered and dose received. This reflects the primary interest of researchers in actual intervention delivery and in participation levels (in other words, quantitative outcomes), rather than in how an intervention is delivered, the quality of delivery (fidelity), maintenance, or the reasons respondents participated or not. Possible explanations may be that researchers are more trained in quantitative methods of research and less in qualitative methods and analysis. Furthermore, good qualitative research takes a lot of time and energy and is therefore more expensive than quantitative research [75]. The included process evaluations tended to operationalize the measured process components in different ways, even when using the same framework for evaluation. For instance, four studies used Steckler and Linnan’s framework as a guideline for their process evaluation [51, 57, 58, 69]. However, in some of these studies, actual attendance levels were placed under reach [51, 57, 58], whereas another study included attendance levels as part of the dose received [69]. The definitions for ‘reach’ and ‘dose received’ as defined by Steckler and Linnan are: ‘proportion of intended target audience that participates in an intervention’ and ‘the extent to which participants actively engage with the intervention’ respectively [14]. The different approaches to operationalization in studies of these components could be explained by these somewhat ambiguous descriptions in Steckler and Linnan’s framework. What is the difference between ‘participating’ and ‘actively engaging with’ an intervention? We can assume that ‘participating in an intervention’ involves a focus on participation and non-participation regardless of the frequency, duration and intensity of participation and that these aspects may actually be taken into account when measuring active engagement with the intervention. Since several studies had different approaches to the operationalization of process components, it was a challenge to translate the definitions used into general terms. We therefore adopted a working definition for each process component based on a previously published framework in order to interpret our findings [11]. For example, we defined reach as ‘the proportion of the target audience that is aware of the intervention’ and dose received as ‘the proportion of participants that actually participates in the intervention’, including frequency, duration and intensity [11]. We therefore distinguished between awareness and actual participation, the latter including level of participation. However, due to the operationalization ambiguity in Steckler and Linnan’s framework, others might define

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

these components, and therefore interpret the findings, differently. This limitation was also mentioned in a comparable review by Durlak and Dupre (2008), revealing that researchers still do not take the time and effort necessary to develop consensus about the best way to perform process evaluations in different settings [13]. Our review of the studies provided evidence of various barriers and/or facilitators that could influence the implementation process. However, partly as a result of the relatively low quality of the process evaluations, no systematic examination of barriers and/ or facilitators affecting implementation was possible. Barriers and/or facilitators were often not systemically measured and only observed and documented on the basis of the researchers’ experience. This raises the question of how reliable the results of this review are with respect to the identified barriers and/or facilitators (table 4). Since the barriers and/or facilitators identified in this review clearly overlap with other reviews focusing on these factors, we can conclude that our results should be reliable [10, 12, 13, 22, 24, 25]. For example, Sangster-Gormley also identified ‘active management participation’ as a facilitator for implementation [25]. Another notable finding was that researchers evaluated barriers and/or facilitators mostly after implementation and only a few studied them during implementation. However, interventions are more likely to be successful if potential barriers and/or facilitators are assessed beforehand so they can be anticipated, facilitating implementation. Although most studies failed to observe barriers and/or facilitators systematically, some barriers and/ or facilitators were more frequently reported in several studies. This suggests that these are important factors which researchers, practitioners and implementers need to take into account during a WHPP. In the category ‘characteristics of the socio-political context’, only one barrier (‘competitive business environment’) and one facilitator (‘compatibility of program with societal developments’) were reported in all twenty-two studies. This could suggest that this category is virtually disregarded, despite its importance for a thorough understanding of these socio-political factors prior to implementation, since the intended user of the intervention is part of an organization, which in turn is part of a wider environment [12]. However, it is also possible that intervention developers and researchers already anticipate on socio-political issues and try to take these into account throughout the development and implementation process. Most barriers and/or facilitators were reported within the category ‘characteristics of the organization’. This is not surprising since most implementation research focuses at this level as this is essential for continuation. Moreover, the studies in this review are selected

Systematic review of implementation studies

because they focus on organizations. Frequently mentioned implementation barriers were: ‘lack of resources’, ‘no fit of the intervention with organizational culture’ and ‘unbalanced or lack of manager support’. For example, when the intervention program requires a fitness center but there is no center nearby, this intervention is not the best fit with the company and other options, such as sports activities (possibly outdoors) in the immediate vicinity of the company, should be explored. Frequently mentioned organizational facilitators for WHPP implementation were: ‘organizations awareness of perceived benefits and relevance’; ‘good collaboration with all persons involved’; ‘strong and formal management support’; and ‘active management participation and engagement’. The researcher or implementer could work on these factors with the aim of optimizing implementation, for example by organizing manager meetings before the start of the intervention so that everyone is informed. However, another frequently mentioned facilitator for implementation was: ‘an organizational culture that emphasizes goal setting and supports social interaction’. Nevertheless, it is unclear what researchers and implementers can do when the organization culture does not fulfill this criterion before the start of the study, especially since organizational culture is a complex phenomenon, which is not easily changed [76]. We suggest that the best way to deal with this ‘problem’ is to adapt the intervention where possible to ensure an optimal fit with the current organizational culture. In addition we would advise researchers and implementers to take the time to assess the organizational culture and include organizational determinants in their process evaluation [77]. Turning to the category ‘characteristics of the implementer’, the results of this review indicate that the most important facilitators for implementers are: ‘sufficient time’, ‘skills’, ‘knowledge’ and ‘competence’. When implementers experienced ‘planning difficulties’ or a ‘heavy workload’, this represented a barrier to implementation. However, these barriers can be overcome by making sure that the implementer receives enough support from stakeholders within the organization and enough administrative support. In the category ‘characteristics of the intervention program’, fifteen barriers and/or facilitators were reported. The three most frequently reported facilitators in this category were: ‘relative advantage’, ‘compatibility’ and ‘complexity’. These three factors are part of Rogers’ Diffusion of Innovations theory, and they are therefore known facilitating factors which could explain why researchers focused on these factors [78]. However, we also identified barriers relating to the intervention characteristics, including: ‘conflicting interests between worksite and interventions’ and ‘technical problems’ which are not part of a known theory. Fewest barriers and/or facilitators were reported in the category ‘characteristics of the

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

participant’. The most commonly reported barrier to participation was ‘time constraints of participants’, suggesting that employers should give employees time off to participate in the interventions. This facilitator is also a reflection of the facilitator ‘management support and engagement’ which was mentioned earlier. Two frequently reported perceived barriers for participants were ‘work demands’ and ‘current workload’. However, as for many of the other found barriers, it is difficult to address these two factors. A possible suggestion to overcome these barriers could be that at onset of the study, employers should stress out that employees are allowed to participate under working hours and that if they experience difficulties in participating due to work demands and workload they should discuss this with their manager. The reported barriers and/or facilitators in this review are comparable with the results of other reviews mentioned in this paper [10, 12, 24, 25]. This suggests that the findings of this review are generalizable to other settings. It would be beneficial to explore all the barriers and/or facilitators listed here before implementation so that researchers and implementers can anticipate possible barriers, incorporate possible facilitators and perhaps adjust their program accordingly. A new Dutch instrument was proposed recently that makes it possible to map barriers and/or facilitators beforehand, and therefore possibly help to facilitate the actual implementation of WHPPs. It also serves as a monitoring instrument enabling program adjustments before and during implementation (TNO R10625; Fleuren et al. 2012). Unfortunately, it was not possible to rank the 54 reported barriers and/or facilitators in order of importance because some determinants were only identified once. This could suggest that the factor was specific to the type of intervention. In addition, the relation between the 54 reported barriers and/or facilitators is unclear. This limitation was also experienced by Fleuren et al. (2004), showing that more research is needed into barriers and/or facilitators affecting the implementation process [12]. Despite the low quality of most included process evaluations, the findings of this review do suggest that higher levels of implementation are associated with better program outcomes. However, the few studies that investigated this association showed a narrow focus on implementation, mainly addressing the link between dose received and/or fidelity in relation to program outcomes. Whereas it is so important to use all aspects of process evaluations for the interpretation of the (lack of) effects of a program [18, 19]. A limitation, which is inherent to writing a systematic review, is publication bias or studies being overlooked. In other words, our overview of the literature may not be complete. We tried to avoid this pitfall by selecting four different databases, both medical

Systematic review of implementation studies

and psychological, using broad search terms and checking each lifestyle (i.e. physical activity, nutrition, smoking, alcohol use and relaxation) separately and by checking the references in the studies we included. Clearly, our decision to restrict our search to process evaluations in combination with effect evaluations in a (randomized) controlled trial design for WHPP has caused a number of process evaluations to fall out of scope for this review. Although we do acknowledge this, our aim was to gain insight into the relationship between implementation quality and effectiveness. Importantly, this is also a tendency that has been suggested by implementation journals, such as Implementation Science. A best-evidence synthesis falls outside the scope of this review. Moreover, it could not be performed since we included only the studies that could be paired with a process evaluation. This means that not all known and relevant controlled trials in the field of WHPPs focusing on healthy lifestyle are included in this review. Conclusion and implications for future research Given the fact that relatively few process evaluations were found by comparison with the high number of RCT’s, and taking into account that not all process evaluations are published, it is safe to conclude that process evaluations are still not as high on the research agenda as effect evaluations. The importance of conducting an effect evaluation as well as a process evaluation is increasingly being advocated [22, 79]. It does appear that lately more process evaluations are being conducted, but this review shows that they are in general of low to average quality due to the lack of a systematic approach. Durlak and Dupre (2008) noted that “implementation matters” and “science cannot study what it cannot measure accurately and cannot measure what is does not define” [13]. Our findings support this observation, and moreover, suggest that we should be asking: what are we measuring, when we measure implementation at all? Without a standardized approach to the operationalization of process evaluation components, it is difficult to faithfully replicate studies, identify implementation barriers and/or facilitators, or assess the methodological quality of process evaluations. This indicates that a general framework for process evaluations is required that researchers, implementers, reviewers and practitioners can use to evaluate and assess the quality of the implementation of a WHPP [80, 81]. In order to create a general framework, it is essential that all relevant stakeholders subscribe to a consensus about the terminology and operationalization of relevant process components by developing a taxonomy for process evaluations. This taxonomy could constitute a fundamental first step towards the standardization of process evaluations in multiple fields,

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

and lead to program reproducibility [80]. Furthermore, it will allow researchers to compare and assess the quality of process evaluations, systematically identify implementation barriers and/or facilitators and possibly link implementation to program outcomes. In short, this review demonstrates the need for a systematic approach to process evaluation as a way of improving WHPP implementation. In order to set the first steps into this direction, we suggest that future process evaluations need to apply the framework Wierenga et al. 2012 we proposed in the introduction of this review and should also use this framework to assess the methodological quality of the studies. This framework allows mapping the complete implementation process and takes into account determinants of implementation and provides the necessary explicit operationalization of each component [11].

Systematic review of implementation studies

Appendices 1 Search strategies used for PubMed Lifestyle behavior

Strategy

Physical activity

Nutrition

Smoking

Alcohol use

Relaxation

((Trial OR randomized controlled trial OR controlled trial) AND (Worker OR workers OR employee OR employees OR workforce OR worksite OR “work site” OR “work environment”) AND (health promotion OR worksite health promotion OR lifestyle intervention OR “stress management” OR “job stress” OR “work stress” OR “psychological stress” OR “life stress” OR “emotional stress” OR distress OR “mental suffering” OR “relaxation therapy” OR relaxation OR meditation OR mind-body OR “mind-body therapies” OR “leisure activities” OR mindfulness OR “mental healing” OR “exercise movement techniques” OR “breathing exercises” OR yoga OR “tai ji” OR pilates OR engagement OR vitality OR detachment) AND (absenteeism OR sickness absence OR sick leave OR body mass index OR body weight OR lifestyle OR life style OR quality of life OR health OR health behavior OR blood pressure OR cholesterol level OR stress OR relaxation)) AND ((English OR Dutch[Language])) AND (“2000/01/01”[Date - Publication]: “2012/08/01”[Date - Publication]))

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Search strategies used for EMBASE Lifestyle behavior

Strategy

Physical activity

‘controlled study’/exp OR ‘randomized controlled trial’/exp OR ‘intervention study’/ exp AND (‘worker’/exp OR ‘employee’/exp OR ‘personnel’/exp OR ‘workplace’/exp OR ‘work environment’/exp) AND (‘health promotion’/exp OR ‘lifestyle modification’/ exp OR ‘physical activity’/exp OR ‘exercise’/exp OR ‘fitness’/exp) AND (‘absenteeism’/ exp OR ‘sickness absence’/exp OR ‘medical leave’/exp OR ‘body mass’/exp OR ‘body weight’/exp OR ‘lifestyle’/exp OR ‘life style’/exp OR ‘quality of life’/exp OR ‘health’/ exp OR ‘health behavior’/exp OR ‘blood pressure’/exp OR ‘cholesterol blood level’/ exp) AND ([Dutch]/lim OR [English]/lim) AND [2000-2012]/py

Nutrition

‘controlled study’/exp OR ‘randomized controlled trial’/exp OR ‘intervention study’/ exp AND (‘worker’/exp OR ‘employee’/exp OR ‘personnel’/exp OR ‘workplace’/exp OR ‘work environment’/exp) AND (‘health promotion’/exp OR ‘lifestyle modification’/ exp OR ‘nutrition’/exp OR ‘food’/exp OR ‘diet’/exp OR ‘catering service’/exp OR ‘nutrition policy’/exp) AND (‘absenteeism’/exp OR ‘sickness absence’/exp OR ‘medical leave’/exp OR ‘body mass’/exp OR ‘body weight’/exp OR ‘lifestyle’/exp OR ‘life style’/exp OR ‘quality of life’/exp OR ‘health’/exp OR ‘health behavior’/exp OR ‘blood pressure’/exp OR ‘cholesterol blood level’/exp) AND ([Dutch]/lim OR [English]/ lim) AND [2000-2012]/py

Smoking

‘controlled study’/exp OR ‘randomized controlled trial’/exp OR ‘intervention study’/ exp AND (‘worker’/exp OR ‘employee’/exp OR ‘personnel’/exp OR ‘workplace’/exp OR ‘work environment’/exp) AND (‘health promotion’/exp OR ‘lifestyle modification’/ exp OR ‘smoking’/exp OR ‘tobacco’/exp OR ‘smoking cessation’/exp OR ‘tobacco dependence’/exp) AND (‘absenteeism’/exp OR ‘sickness absence’/exp OR ‘medical leave’/exp OR ‘body mass’/exp OR ‘body weight’/exp OR ‘lifestyle’/exp OR ‘life style’/ exp OR ‘quality of life’/exp OR ‘health’/exp OR ‘health behavior’/exp OR ‘blood pressure’/exp OR ‘cholesterol blood level’/exp OR ‘stress’/exp) AND ([Dutch]/lim OR [English]/lim) AND [2000-2012]/py

Alcohol use

‘controlled study’/exp OR ‘randomized controlled trial’/exp OR ‘intervention study’/ exp AND (‘worker’/exp OR ‘employee’/exp OR ‘personnel’/exp OR ‘workplace’/exp OR ‘work environment’/exp) AND (‘health promotion’/exp OR ‘lifestyle modification’/ exp OR ‘drinking behavior’/exp OR ‘alcohol’/exp OR ‘alcohol consumption’/exp) AND (‘absenteeism’/exp OR ‘sickness absence’/exp OR ‘medical leave’/exp OR ‘body mass’/exp OR ‘body weight’/exp OR ‘lifestyle’/exp OR ‘life style’/exp OR ‘quality of life’/exp OR ‘health’/exp OR ‘health behavior’/exp OR ‘blood pressure’/exp OR ‘cholesterol blood level’/exp OR ‘stress’/exp) AND ([Dutch]/lim OR [English]/lim) AND [2000-2012]/py

Relaxation

trial OR randomized AND controlled AND trial OR controlled AND trial AND (‘worker’/ exp OR workers OR ‘employee’/exp OR employees OR workforce OR worksite OR ‘work site’ OR ‘work environment’/exp) AND (‘health’/exp AND promotion OR worksite AND ‘health’/exp AND promotion OR ‘lifestyle’/exp AND intervention OR ‘stress management’/exp OR ‘job stress’/exp OR ‘work stress’/exp OR ‘psychological stress’/exp OR ‘life stress’/exp OR ‘emotional stress’/exp OR ‘distress’/exp OR ‘mental suffering’ OR ‘relaxation therapy’/exp OR ‘relaxation’/exp OR ‘meditation’/exp OR ‘mind body’ OR ‘mind-body therapies’/exp OR ‘leisure activities’/exp OR mindfulness OR ‘mental healing’/exp OR ‘exercise movement techniques’/exp OR ‘breathing exercises’/exp OR ‘yoga’/exp OR ‘tai ji’/exp OR ‘pilates’/exp OR engagement OR vitality OR detachment) AND (‘absenteeism’/exp OR ‘sickness’/exp AND ‘absence’/ exp OR sick AND leave OR body AND ‘mass’/exp AND index OR body AND ‘weight’/ exp OR ‘lifestyle’/exp OR ‘life’/exp AND style OR quality AND of AND ‘life’/exp OR ‘health’/exp AND ‘behavior’/exp OR ‘blood’/exp AND ‘pressure’/exp OR ‘cholesterol’/ exp AND level OR ‘stress’/exp OR ‘relaxation’/exp) AND ([Dutch]/lim OR [English]/lim) AND [2000-2012]/py

Systematic review of implementation studies

Search strategies used for PsycINFO Lifestyle behavior

Strategy

Physical activity

(KW=(trial or randomized controlled trial or controlled trial) and (Worker or workers or employee or employees or workforce or worksite or ‘work site’ or ‘work environment’ or workplace or ‘work place’ or personnel or working conditions or business organizations) and (health promotion or worksite health promotion or lifestyle intervention or physical activity or exercise or physical fitness or ‘active living’) and (absenteeism OR sickness absence OR sick leave OR body mass index OR body weight OR lifestyle OR life style OR quality of life OR health OR health behavior OR blood pressure OR cholesterol level)) AND (Limiters - Published Date: 2000010120120731; Language: Dutch, English; Document Type: Journal Article)

Nutrition

KW=(trial or randomized controlled trial or controlled trial) and (Worker or workers or employee or employees or workforce or worksite or ‘work site’ or ‘work environment’ or workplace or ‘work place’ or personnel or working conditions or business organizations) and (health promotion or worksite health promotion or lifestyle intervention or nutrition or food or diets or ‘healthy eating’ or ‘food services’ or ‘nutrition policy’) and (absenteeism OR sickness absence OR sick leave OR body mass index OR body weight OR lifestyle OR life style OR quality of life OR health OR health behavior OR blood pressure OR cholesterol level) AND (Limiters - Published Date: 20000101-20120731; Language: Dutch, English; Document Type: Journal Article) KW = (trial or randomized controlled trial or controlled trial) and (Worker or workers or employee or employees or workforce or worksite or ‘work site’ or ‘work environment’ or workplace or ‘work place’ or personnel or working conditions or business organizations) and (health promotion or worksite health promotion or lifestyle intervention or smoking OR Tobacco or Tobacco smoking or smoking cessation or tobacco use) and (absenteeism OR sickness absence OR sick leave OR body mass index OR body weight OR lifestyle OR life style OR quality of life OR health OR health behavior OR blood pressure OR cholesterol level OR stress) AND (Limiters - Published Date: 20000101-20120731; Language: Dutch, English; Document Type: Journal Article) KW = (trial or randomized controlled trial or controlled trial) and (Worker or workers or employee or employees or workforce or worksite or ‘work site’ or ‘work environment’ or workplace or ‘work place’ or personnel or working conditions or business organizations) and (health promotion or worksite health promotion or lifestyle intervention or ‘drinking behavior’ OR alcohol OR ‘alcohol consumption’ OR ‘alcohol drinking’) and (absenteeism OR sickness absence OR sick leave OR body mass index OR body weight OR lifestyle OR life style OR quality of life OR health OR health behavior OR blood pressure OR cholesterol level OR stress) AND (Limiters Published Date: 20000101-20120731; Language: Dutch, English; Document Type: Journal Article) KW = (trial or randomized controlled trial or controlled trial) and (Worker or workers or employee or employees or workforce or worksite or ‘work site’ or ‘work environment’ or workplace or ‘work place’ or personnel or working conditions or business organizations) and (health promotion or worksite health promotion or lifestyle intervention or “stress management” OR “job stress” OR “work stress” OR “psychological stress” OR “life stress” OR “emotional stress” OR distress OR “mental suffering” OR “relaxation therapy” OR relaxation OR meditation OR mind-body OR “mind-body therapies” OR “leisure activities” OR mindfulness OR “mental healing” OR “exercise movement techniques” OR “breathing exercises” OR yoga OR “tai ji” OR pilates OR engagement OR vitality OR detachment) and (absenteeism OR sickness absence OR sick leave OR body mass index OR body weight OR lifestyle OR life style OR quality of life OR health OR health behavior OR blood pressure OR cholesterol level OR stress OR relaxation)AND (Limiters - Published Date: 20000101-20120731; Language: Dutch, English; Document Type: Journal Article)

Smoking

Alcohol use

Relaxation

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Search strategies used for Cochrane Central Register of Controlled trials Lifestyle behavior

Strategy

Physical activity

(Trial or randomized controlled trial or controlled trial) and (Worker or workers or employee or employees or workforce or worksite or “work site” or “work environment”) and (health promotion or worksite health promotion or lifestyle intervention or physical activity or exercise or physical fitness) and (absenteeism or sickness absence or sick leave or body mass index or body weight or lifestyle or life style or quality of life or health or health behavior or blood pressure or cholesterol level):ti,ab,kw from 2000 to 2012 (Word variations have been searched)

Nutrition

(Trial OR randomized controlled trial OR controlled trial) AND (Worker OR workers OR employee OR employees OR workforce OR worksite OR “work site” OR “work environment”) AND (health promotion OR worksite health promotion OR lifestyle intervention OR Nutrition OR healthy eating OR food OR diet OR food services OR nutrition policy) AND (absenteeism OR sickness absence OR sick leave OR body mass index OR body weight OR lifestyle OR life style OR quality of life OR health OR health behavior OR blood pressure OR cholesterol level):ti,ab,kw from 2000 to 2012 (Word variations have been searched) (Trial OR randomized controlled trial OR controlled trial) AND (Worker OR workers OR employee OR employees OR workforce OR worksite OR “work site” OR “work environment”) AND (health promotion OR worksite health promotion OR lifestyle intervention OR smoking OR Tobacco OR smoking cessation OR tobacco use) AND (absenteeism OR sickness absence OR sick leave OR body mass index OR body weight OR lifestyle OR life style OR quality of life OR health OR health behavior OR blood pressure OR cholesterol level OR stress):ti,ab,kw from 2000 to 2012 (Word variations have been searched) (Trial OR randomized controlled trial OR controlled trial) AND (Worker OR workers OR employee OR employees OR workforce OR worksite OR “work site” OR “work environment”) AND (health promotion OR worksite health promotion OR lifestyle intervention OR Alcohol drinking OR Alcohol intake OR alcohol consumption) AND (absenteeism OR sickness absence OR sick leave OR body mass index OR body weight OR lifestyle OR life style OR quality of life OR health OR health behavior OR blood pressure OR cholesterol level OR stress):ti,ab,kw from 2000 to 2012 (Word variations have been searched) (Trial OR randomized controlled trial OR controlled trial) AND (Worker OR workers OR employee OR employees OR workforce OR worksite OR “work site” OR “work environment”) AND (health promotion OR worksite health promotion OR lifestyle intervention OR “stress management” OR “job stress” OR “work stress” OR “psychological stress” OR “life stress” OR “emotional stress” OR distress OR “mental suffering” OR “relaxation therapy” OR relaxation OR meditation OR mind-body OR “mind-body therapies” OR “leisure activities” OR mindfulness OR “mental healing” OR “exercise movement techniques” OR “breathing exercises” OR yoga OR “tai ji” OR pilates OR engagement OR vitality OR detachment) AND (absenteeism OR sickness absence OR sick leave OR body mass index OR body weight OR lifestyle OR life style OR quality of life OR health OR health behavior OR blood pressure OR cholesterol level OR stress OR relaxation):ti,ab,kw from 2000 to 2012 (Word variations have been searched)

Smoking

Alcohol use

Relaxation

Systematic review of implementation studies

References 1.

Kirsten W: Making the link between health and productivity at the workplace--a global perspective. Ind Health 2010, 48(3):251-255.

Jans MP, van den Heuvel SG, Hildebrandt VH, Bongers PM: Overweight and Obesity as Predictors of Absenteeism in the Working Population of the Netherlands. J Occup Environ Med 2007, 49(9):975980.

Robroek SJ, van den Berg TI, Plat JF, Burdorf A: The role of obesity and lifestyle behaviours in a productive workforce. Occup Environ Med 2011, 68(2):134-139.

Alavinia SM, Molenaar D, Burdorf A: Productivity loss in the workforce: Associations with health, work demands, and individual characteristics. Am J Ind Med 2009, 52(1):49-56.

Engbers LH, van Poppel MN, ChinA.Paw MJ, van Mechelen W: Worksite health promotion programs with environmental changes: a systematic review. Am J Prev Med 2005, 29(1):61-70.

Ni Mhurchu C, Aston LM, Jebb SA: Effects of worksite health promotion interventions on employee diets: a systematic review. BMC Public Health 2010, 10(10):62.

Jepson RG, Harris FM, Platt S, Tannahill C: The effectiveness of interventions to change six health behaviors: a review of reviews. BMC Public Health 2010, 8(10):538.

Proper KI, Koning M, van der Beek AJ, Hildebrandt VH, Bosscher RJ, van Mechelen W: The effectiveness of worksite physical activity programs on physical activity, physical fitness, and health. Clin J Sport Med 2003, 13(2):106-117.

Kristensen TS: Intervention studies in occupational epidemiology. Occup Environ Med 2005, 62:205210.

10. Robroek SJ, van Lenthe FJ, van Empelen P, Burdorf A: Determinants of participation in worksite health promotion programmes: a systematic review. Int J Behav Nutr Phys Act 2009, 6:26. 11. Wierenga D, Engbers LH, van Empelen P, Hildebrandt VH, van Mechelen W: The design of a real-time formative evaluation of the implementation process of lifestyle interventions at two worksites using a 7-step strategy (BRAVO@Work). BMC Public Health 2012, 12(1):619. 12. Fleuren M, Wiefferink K, Paulussen T: Determinants of innovation within health care organizations: literature review and Delphi study. Int J Qual Health Care 2004, 16(2):107-123. 13. Durlak JA, DuPre EP: Implementation matters: A review of research on the influence of implementation on program outcomes and the factors affecting implementation. Am J Community Psychol 2008, 41:327-350. 14. Steckler A, Linnan L: Proces Evaluation for Public Health Interventions and Research: San Francisco: Jossey-Bass; 2002. 15. Paulussen TH, Wiefferink K, Mesters I: Invoering van effectief gebleken interventies. In Gezondheidsvoorlichting en gedragsverandering. Edited by Brug,J.,Van Asseman,P.& Lechner,L. Assen: van Gorcum: 2007:. 16. Craig P, Dieppe P, Macintyre S, Nazareth I, Petticrew M: Developing and evaluating complex interventions: the new Medical Research Council guidance. BMJ 2008, 337(1655):979-983. 17. Saunders RP, Evans MH, Joshi P: Developing a process-evaluation plan for assessing health promotion program implementation: a how-to guide. Health promotion practice 2005, 6(2):134-147. 18. Oakley A, Strange V, Bonell C, Allen E, Stephenson J, RIPPLE Study Team: Process evaluation in randomised controlled trials of complex interventions. BMJ 2006, 332(7538):413-416. 19. Glasgow RE, Klesges LM, Dzewaltowski DA, Bull SS, Estabrooks PA: The future of health behavior change research: what is needed to improve translation of research into health promotion practice? Ann Behav Med 2004, 27(1):3-12.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

20. Stetler CB, Legro MW, Wallace CM, Bowman C, Guihan M, Hagedorn H, Kimmel B, Sharp N, Smith JL: The role of formative evaluation in implementation research and the QUERI experience. J Gen Intern Med 2006, Feb(21 Suppl 2):S1-8. 21. Mittman BS: Creating the evidence base for quality improvement collaboratives. Ann Int Med 2004, 140(11):897-901. 22. Murta SG, Sanderson K, Oldenburg B: Process evaluation in occupational stress management programs: a systematic review. Am J Health Promot 2007, 21(0890-1171; 0890-1171; 4):248-254. 23. Glasgow RE, Vogt TM, Boles SM: Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health 1999, (89):9-1322. 24. Koppelaar E, Knibbe JJ, Miedema HS, Burdorf A: Determinants of implementation of primary preventive interventions on patient handling in healthcare: a systematic review. Occup Environ Med 2009, 66(6):353-360. 25. Sangster-Gormley E, Martin-Misener R, Downe-Wamboldt B, Dicenso A: Factors affecting nurse practitioner role implementation in Canadian practice settings: an integrative review. J Adv Nurs 2011, 67(6):1178-1190. 26. Furlan AD, Pennick V, Bombardier C, van Tulder M, Editorial Board, Cochrane Back Review Group: 2009 updated method guidelines for systematic reviews in the Cochrane Back Review Group. Spine (Phila Pa 1976) 2009, 34(18):1929-1941. 27. Proper KI, Staal BJ, Hildebrandt VH, van der Beek AJ, van Mechelen W: Effectiveness of physical activity programs at worksites with respect to work-related outcomes. Scand J Work Environ Health 2002, 28(2):75-84. 28. van Tulder M, Furlan A, Bombardier C, Bouter L, Editorial Board of the Cochrane Collaboration Back Review Group: Updated method guidelines for systematic reviews in the cochrane collaboration back review group. Spine (Phila Pa 1976) 2003, 28(12):1290-1299. 29. French SA, Harnack LJ, Hannan PJ, Mitchell NR, Gerlach AF, Toomey TL: Worksite environment intervention to prevent obesity among metropolitan transit workers. Prev Med 2010, 50(4):180-185. 30. Dishman RK, DeJoy DM, Wilson MG, Vandenberg RJ: Move to Improve: a randomized workplace trial to increase physical activity. Am J Prev Med 2009, 36(2):133-141. 31. Yap TL, Davis LS, Gates DM, Hemmings AB, Pan W: The effect of tailored E-mails in the workplace. Part I. Stage movement toward increased physical activity levels. AAOHN J 2009, 57(7):267-273. 32. Goetzel RZ, Baker KM, Short ME, Pei X, Ozminkowski RJ, Wang S, Bowen JD, Roemer EC, Craun BA, Tully KJ, Baase CM, DeJoy DM, Wilson MG: First-year results of an obesity prevention program at The Dow Chemical Company. J Occup Environ Med 2009, 51(2):125-138. 33. Lemon SC, Zapka J, Li W, Estabrook B, Rosal M, Magner R, Andersen V, Borg A, Hale J: Step ahead a worksite obesity prevention trial among hospital employees. Am J Prev Med 2010, 38(1):27-38. 34. Sorensen G, Barbeau E, Stoddard AM, Hunt MK, Kaphingst K, Wallace L: Promoting behavior change among working-class, multiethnic workers: results of the healthy directions--small business study. Am J Public Health 2005, 95(8):1389-1395. 35. Beresford SA, Thompson B, Feng Z, Christianson A, McLerran D, Patrick DL: Seattle 5 a Day worksite program to increase fruit and vegetable consumption. Prev Med 2001, 32(3):230-238. 36. Sorensen G, Barbeau EM, Stoddard AM, Hunt MK, Goldman R, Smith A, Brennan AA, Wallace L: Tools for health: the efficacy of a tailored intervention targeted for construction laborers. Cancer Causes Control 2007, 18(1):51-59. 37. Sorensen G, Stoddard A, Quintiliani L, Ebbeling C, Nagler E, Yang M, Pereira L, Wallace L: Tobacco use cessation and weight management among motor freight workers: results of the gear up for health study. Cancer Causes Control 2010, 21(12):2113-2122. 38. Stoddard AM, Fagan P, Sorensen G, Hunt MK, Frazier L, Girod K: Reducing cigarette smoking among working adolescents: results from the SMART study. Cancer Causes Control 2005, 16(10):1159-1164.

Systematic review of implementation studies

39. Volpp KG, Troxel AB, Pauly MV, Glick HA, Puig A, Asch DA, Galvin R, Zhu J, Wan F, DeGuzman J, Corbett E, Weiner J, Audrain-McGovern J: A randomized, controlled trial of financial incentives for smoking cessation. N Engl J Med 2009, 360(7):699-709. 40.

Driessen MT, Proper KI, Anema JR, Knol DL, Bongers PM, van der Beek AJ: Participatory ergonomics to reduce exposure to psychosocial and physical risk factors for low back pain and neck pain: results of a cluster randomised controlled trial. Occup Environ Med 2011, 68(9):674-681.

41.

Groeneveld IF, Proper KI, van der Beek AJ, van Mechelen W: Sustained body weight reduction by an individual-based lifestyle intervention for workers in the construction industry at risk for cardiovascular disease: results of a randomized controlled trial. Prev Med 2010, 51(3-4):240-246.

42. Gilson N, McKenna J, Cooke C, Brown W: Walking towards health in a university community: a feasibility study. Prev Med 2007, 44(2):167-169. 43. Andersen LL, Saervoll CA, Mortensen OS, Poulsen OM, Hannerz H, Zebis MK: Effectiveness of small daily amounts of progressive resistance training for frequent neck/shoulder pain: randomised controlled trial. Pain 2011, 152(2):440-446. 44. Haukka E, Pehkonen I, Leino-Arjas P, Viikari-Juntura E, Takala EP, Malmivaara A, Hopsu L, Mutanen P, Ketola R, Virtanen T, Holtari-Leino M, Nykanen J, Stenholm S, Ojajarvi A, Riihimaki H: Effect of a participatory ergonomics intervention on psychosocial factors at work in a randomised controlled trial. Occup Environ Med 2010, 67(3):170-177. 45. Steenhuis I, Van Assema P, Van Breukelen G, Glanz K, Kok G, De Vries H: The impact of educational and environmental interventions in Dutch worksite cafeterias. Health Promot Int 2004, 19(3):335-343. 46. Hasson D, Anderberg UM, Theorell T, Arnetz BB: Psychophysiological effects of a web-based stress management system: a prospective, randomized controlled intervention study of IT and media workers [ISRCTN54254861. BMC Public Health 2005, 5:78. 47. Vermeer WM, Steenhuis IH, Leeuwis FH, Heymans MW, Seidell JC: Small portion sizes in worksite cafeterias: do they help consumers to reduce their food intake? Int J Obes (Lond) 2011, 35(9):12001207. 48. Strijk JE, Proper KI, van der Beek AJ, van Mechelen W: A worksite vitality intervention to improve older workersâ&#x20AC;&#x2122; lifestyle and vitality-related outcomes: results of a randomised controlled trial. J Epidemiol Community Health 2012, 66(11):1071-1078. 49. Verweij LM, Proper KI, Weel AN, Hulshof CT, van Mechelen W: The application of an occupational health guideline reduces sedentary behaviour and increases fruit intake at work: results from an RCT. Occup Environ Med 2012, 69(7):500-507. 50. Jorgensen MB, Faber A, Hansen JV, Holtermann A, Sogaard K: Effects on musculoskeletal pain, work ability and sickness absence in a 1-year randomised controlled trial among cleaners. BMC Public Health 2011, 11:840. 51. Driessen MT, Proper KI, Anema JR, Bongers PM, van der Beek AJ: Process evaluation of a participatory ergonomics programme to prevent low back pain and neck pain among workers. Implement Sci 2010, 5:65. 52. Wilson MG, Basta TB, Bynum BH, DeJoy DM, Vandenberg RJ, Dishman RK: Do intervention fidelity and dose influence outcomes? Results from the move to improve worksite physical activity program. Health Educ Res 2010, 25(2):294-305. 53. Wilson MG, Goetzel RZ, Ozminkowski RJ, DeJoy DM, Della L, Roemer EC, Schneider J, Tully KJ, White JM, Baase CM: Using formative research to develop environmental and ecological interventions to address overweight and obesity. Obesity (Silver Spring) 2007, 15 Suppl 1:37S-47S. 54. Pehkonen I, Takala EP, Ketola R, Viikari-Juntura E, Leino-Arjas P, Hopsu L, Virtanen T, Haukka E, HoltariLeino M, Nykyri E, Riihimaki H: Evaluation of a participatory ergonomic intervention process in kitchen work. Appl Ergon 2009, 40(1):115-123. 55. Hunt MK, Barbeau EM, Lederman R, Stoddard AM, Chetkovich C, Goldman R, Wallace L, Sorensen G: Process evaluation results from the Healthy Directions-Small Business study. Health Educ Behav 2007, 34(1):90-107.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

56. Vermeer WM, Leeuwis FH, Koprulu S, Zouitni O, Seidell JC, Steenhuis IH: The process evaluation of two interventions aimed at portion size in worksite cafeterias. J Hum Nutr Diet 2012, 25(2):180-188. 57. Strijk JE, Proper KI, van der Beek AJ, van Mechelen W: A process evaluation of a worksite vitality intervention among ageing hospital workers. Int J Behav Nutr Phys Act 2011, 8:58. 58. Verweij LM, Proper KI, Hulshof CT, van Mechelen W: Process evaluation of an occupational health guideline aimed at preventing weight gain among employees. J Occup Environ Med 2011, 53(7):722729. 59. Groeneveld IF, Proper KI, Absalah S, van der Beek AJ, van Mechelen W: An individually based lifestyle intervention for workers at risk for cardiovascular disease: a process evaluation. Am J Health Promot 2011, 25(6):396-401. 60. Estabrook B, Zapka J, Lemon SC: Evaluating the implementation of a hospital work-site obesity prevention intervention: applying the RE-AIM framework. Health Promot Pract 2012, 13(2):190-197. 61. Beresford SA, Shannon J, McLerran D, Thompson B: Seattle 5-a-Day Work-Site Project: process evaluation. Health Educ Behav 2000, 27(2):213-222. 62. Kim A, Kamyab K, Zhu J, Volpp K: Why are financial incentives not effective at influencing some smokers to quit? Results of a process evaluation of a worksite trial assessing the efficacy of financial incentives for smoking cessation. J Occup Environ Med 2011, 53(1):62-67. 63. Yap TL, Busch James DM: Tailored e-mails in the workplace. AAOHN J 2010, 58(10):425-432. 64. Gilson N, McKenna J, Cooke C: Experiences of route and task-based walking in a university community: qualitative perspectives in a randomized control trial. J Phys Act Health 2008, 5 Suppl 1:S176-82. 65. Hunt MK, Harley AE, Stoddard AM, Lederman RI, MacArthur MJ, Sorensen G: Elements of external validity of tools for health: an intervention for construction laborers. Am J Health Promot 2010, 24(5):e11-20. 66. Steenhuis I, van Assema P, Reubsaet A, Kok G: Process evaluation of two environmental nutrition programmes and an educational nutrition programme conducted at supermarkets and worksite cafeterias in the Netherlands. J Hum Nutr Diet 2004, 17(2):107-115. 67. Quintiliani L, Yang M, Sorensen G: A process evaluation of tobacco-related outcomes from a telephone and print-delivered intervention for motor freight workers. Addict Behav 2010, 35(11):1036-1039. 68. Hunt MK, Fagan P, Lederman R, Stoddard A, Frazier L, Girod K, Sorensen G: Feasibility of implementing intervention methods in an adolescent worksite tobacco control study. Tob Control 2003, 12 Suppl 4:IV40-5. 69. Jorgensen MB, Faber A, Jespersen T, Hansen K, Ektor-Andersen J, Hansen JV, Holtermann A, Sogaard K: Implementation of physical coordination training and cognitive behavioural training interventions at cleaning workplaces--secondary analyses of a randomised controlled trial. Ergonomics 2012, 55(7):762-772. 70. Andersen LL: Influence of psychosocial work environment on adherence to workplace exercise. J Occup Environ Med 2011, 53(2):182-184. 71. Hasson H, Brown C, Hasson D: Factors associated with high use of a workplace web-based stress management program in a randomized controlled intervention study. Health Educ Res 2010, 25(4):596-607. 72. Driessen MT, Groenewoud K, Proper KI, Anema JR, Bongers PM, van der Beek AJ: What are possible barriers and facilitators to implementation of a Participatory Ergonomics programme? Implement Sci 2010, 5:64. 73. DeJoy DM, Bowen HM, Baker KM, Bynum BH, Wilson MG, Goetzel RZ, Dishman RK: Management support and worksite health promotion program effectiveness. Ergonomics and Health Aspects 2009, (5624):13-22. 74. Verweij LM, Proper KI, Leffelaar ER, Weel AN, Nauta AP, Hulshof CT, van Mechelen W: Barriers and Facilitators to Implementation of an Occupational Health Guideline Aimed at Preventing Weight Gain Among Employees in the Netherlands. J Occup Environ Med 2012, 54(8):954-960.

Systematic review of implementation studies

75. Pope C, Ziebland S, Mays N: Qualitative research in health care. Analysing qualitative data. BMJ 2000, 320:114-116. 76. Cameron KS, Quinn RE, Robert E: Diagnosing and changing organizational culture: Based on the Competing Values Framework: San Fransico: Jossey-Bass; 2006. 77. Weiner BJ, Lewis MA, Linnan LA: Using organization theory to understand the determinants of effective implementation of worksite health promotion programs. Health Educ Res 2009, 24(2):292305. 78. Rogers E.M.: Diffusion of innovations: New York: Free Press; 2003. 79. Nielsen K, Randall R, Holten A, Gonzรกlezc ER: Conducting organizational-level occupation health interventions: What works? Work & Stress 2010, 24(3):234-259. 80. Abraham C, Michie S: A taxonomy of behavior change techniques used in interventions. Health Psychol 2008, 27(3):379-387. 81. Pedersen LM, Nielsen KJ, Kines P: Realistic evaluation as a new way to design and evaluate occupational safety interventions. Safety Science 2012, 50(1):48-54.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

The implementation of multiple lifestyle interventions in two organizations: a process evaluation

Wierenga D, Engbers LH, van Empelen P, de Moes KJ, Wittink H, Gr端ndemann R, van Mechelen W.

Journal of Occupational and Environmental Medicine 2014, 56(11):1195-1206

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

Abstract Objective: The aim of this study is to systematically evaluate the implementation of multiple lifestyle interventions at two different worksites (i.e. hospital and university) by (1) studying the extent and quality of programme delivery and maintenance, (2) looking at employee recruitment, reach, participation and satisfaction levels, and (3) evaluate the effectiveness of the programme in a quasi-experimental controlled trial conducted alongside the process evaluation. Methods: Quantitative and qualitative data on eight process components were obtained 6 and 12 months after start of implementation. The impact of the intervention on employee lifestyle (physical activity, nutrition, smoking, alcohol use and relaxation) was measured at baseline and at 12-month follow-up using questionnaires. Results: The programme was implemented partly as planned. Compliance to the intervention protocol was higher at the university. Overall, 84.0% (max 25) and 85.7% (max 14) of all planned interventions were delivered at the university and hospital respectively. These were mainly easy to implement environmental and educational interventions. At employee level, high programme reach (96.6%) and overall employee participation (75.1%) was achieved. Satisfaction with the overall programme was moderate (6.8 Âą 1.1). Significant intervention effects were found for days of fruit consumption (Î˛: 0.44 days/week, 95%CI 0.02 to 0.85) in favour of the intervention group. Conclusions: Implementing multiple lifestyle interventions that target different aspects resulted in high awareness and participation levels. However, since both organisations were in charge and the researchers were merely observers, the selected interventions were simple and relatively easy to implement. The inclusion of the topic vitality in annual performance interviews at the hospital and the chair massages, bicycle check and individual coaching programme at the university were included in the organisationâ&#x20AC;&#x2122;s general health policy. Furthermore, the programme was only successful in positively changing days of fruit consumption.

Implementation and effectiveness of lifestyle interventions

Background Unhealthy and overweight employees show elevated sickness absence levels and lower productivity levels, resulting in higher costs for employers [1, 2]. Many worksite health promotion programmes (WHPPs) have been developed in order to improve employee health and to reduce overweight and associated higher sickness absence rates [1, 2]. Reviews have indicated that WHPPs can be effective in changing lifestyle behaviours [3-7]. However, due to the emphasis on outcome evaluations it often remains unclear which problems organisations encounter when implementing a multi-component lifestyle intervention. Understanding what happens during the implementation of a WHPP and how that affects the impact of a programme is an essential step in opening the intervention â&#x20AC;&#x2DC;black boxâ&#x20AC;&#x2122; [8, 9]. Although researchers have started to acknowledge the importance of process evaluations over the past decade, only 7% of the WHPPs evaluating a lifestyle intervention actually have performed a process evaluation. Furthermore, the content, approach and quality of these process evaluations differed greatly between studies [10]. The BRAVO@Work project attempted to remedy this by conducting a controlled trial alongside a comprehensive process evaluation to further our understanding of the natural course of programme implementation. BRAVO@Work is a WHPP in which multiple lifestyle interventions (related to physical activity, smoking, alcohol use, nutrition and relaxation) were implemented at two different worksites using a 7-step implementation strategy that facilitated structured implementation by the organisations themselves [11]. Process evaluations are an important tool for studying the underlying working mechanisms of WHPPs and factors affecting implementation [12, 13]. They provide information on the degree of implementation of a WHPP by giving insight into whether an intervention has been delivered as planned and the extent to which the programme was received by participants [14, 15]. This information can be used by researchers and practitioners for programme improvement and optimization to further future implementation of these programmes in daily practice [8]. WHPPs often report a limited impact on outcome measures such as physical activity, healthy nutrition and obesity, and so process data may teach us more about how programme implementation and reception by participants are linked to outcomes [8, 15, 17]. This study therefore systematically evaluated the implementation of a multicomponent lifestyle intervention at two different worksites by (1) studying the extent and quality of programme delivery and maintenance, (2) looking at employee recruitment, reach, participation and satisfaction levels, and (3) evaluate the effectiveness of the programme in a quasi-experimental controlled trial conducted alongside the process evaluation. 85

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

Methods This process evaluation is part of the systematic formative evaluation of the BRAVO@ Work project. It focused exclusively on evaluating the implementation of multiple lifestyle interventions among employees at two worksites. A researcher on site (referred to as the “embedded” researcher) continuously monitored the natural course of implementation, but emphatically tried not to actively intervene in the development and implementation of the interventions. Details of the study design have been described elsewhere [11]. The study protocol was approved by the Medical Ethics Committee of the University Medical Centre of Utrecht. Study design and population The study was designed as a systematic formative evaluation with a quasi-experimental controlled trial on the side, involving an Academic Hospital and a University of Applied Sciences (hereinafter “the hospital” and “the university”) in Utrecht (Netherlands). One intervention department was recruited at both organisations – a department of gynaecology at the Hospital with 662 employees and the health faculty of the University with 546 employees – in 2009-2010 through the personal network of the researchers. The management of the two intervention departments contacted a comparable control department and faculty (a cardiovascular and lung diseases department with 501 employees at the hospital and the society & law faculty with 484 employees at the university). The upper management of the intervention groups signed a letter of intent stating that they were willing to participate in the study and agreed to the financial and organisational consequences of participating in BRAVO@Work. The control groups were not allowed to implement any interventions during the course of the study. All employees were eligible to participate in the study. The BRAVO@Work study In the BRAVO@Work study a 7-step implementation strategy was used by both intervention departments to develop, implement and maintain interventions targeting multiple lifestyle behaviours. The principal feature of this study is that the intervention departments had sole responsibility for the development, implementation and continuation of the interventions; the researchers acted solely as embedded observers throughout the study. An external advisor informed the project groups of both intervention departments about the 7-step strategy during the first project meetings and supplied them with a list of environmental

Implementation and effectiveness of lifestyle interventions

and individual interventions evidence-based interventions of varying complexity [4-6, 18, 19]. The external advisor was present during most project meetings to answer questions and to provide guidance, but he was specifically briefed to refrain from taking the lead at any time during the project. The 7-step strategy consists of the following steps: 1) creating solid support, 2) formation of a project structure, 3) performing a needs assessment, 4) development of interventions, 5) implementation of interventions, 6) evaluation, and 7) maintenance. Detailed information about the use and interpretation of the specific steps of the strategy can be found elsewhere [11]. In short, the upper management of both participating organisations were required to form their own steering committee and to appoint a project leader. The project leader was then required to establish a project group and was advised to include the following relevant stakeholders: managers and employees from different teams, a communications officer, a human resources officer, a facility manager and a company physician. Managers and employees were eligible for project membership if they were working at the intervention department. With guidance from the external advisor, the project members conducted a needs assessment among all employees in their intervention department, resulting in a list of possible lifestyle interventions fitting the department and the needs of employees. This list was combined with the list of evidence-based interventions referred to above. The project groups needed to choose on a consensus basis the most appropriate and feasible interventions matching employeesâ&#x20AC;&#x2122; needs. The project leader then needed to draw up a project plan with information about the desired changes and project goals, an intervention template, a timeline, a budget plan and a list of involved persons, including their tasks and responsibilities. After approval had been given by the steering committees, the project members were required to implement the interventions during the following 12 months. After the intervention year, the project groups needed to evaluate the project. This evaluation needed to be the basis for the go/no-go decision about whether to continue the programme of specific interventions. The process evaluation To establish a picture of the implementation process of the interventions, eight process components were assessed: recruitment, reach, dose delivered, dose received, fidelity, satisfaction, maintenance and contamination (table 1) [11]. The frameworks of Steckler and Linnan, the RE-AIM framework and the framework of Baranowski et al. (2000) were combined for this purpose [14, 16, 20].

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

- Monitoring records Extent to which the project members delivered the interventions in line with their pre-defined project plan - Semi-structured interviews with project members at 6 and 12 months follow-up

members at 6 and 12 months follow-up

intervention go according to plan?”

- “Did the implementation of the

implementation of the interventions?”

a specific intervention?”

- “What could you tell me about the

- “What were the reasons for not delivering

interventions?”

- “Did the project group deliver all intended

- Monitoring records - Semi-structured interviews with project

Examples of questions

Data collection tool

Proportion of employees who were aware of the project and the interventions follow-up

were you aware of?”

- “Which of the following interventions

- Process questionnaire at 6 and 12 months - “Were you aware of the project?”

Dose received Proportion of employees in the intervention - Process questionnaire at 6 and 12 months - “Could you list the interventions in which group who participated in the project and follow-up you have participated?” interventions - Semi-structured interviews with - “Why did you, or did you not, participate employees at 6 and 12 months follow-up in some of the interventions?”

Reach

Maintenance

Extent to which the developed interventions - Semi-structured interviews with project - “Which interventions will be continued in members at 12 months follow-up the organisation?” were continued in the organisation Employee recruitment for, exposure to, and satisfaction with the interventions - “Which of the following recruitment Recruitment Sources and procedures used to approach - Monitoring records methods were you aware of?” and interest employees for participation - Process questionnaire at 6 and 12 months in the interventions, including employee - “Were you satisfied with the recruitment awareness of and satisfaction with the used follow-up methods used in the project?” recruitment methods - Semi-structured interviews with employees at 6 and 12 months follow-up

Fidelity

Component Definition Implementation of the interventions Dose delivered Proportion of intended interventions actually delivered or provided by project members to employees

Table 1. Process evaluation components and their definition including data collection levels and methods

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 Chapter 4

Program contamination Contamination Extent to which employees in the control departments were aware of the project and received or participated in the interventions departments at 12 months follow-up

- Process questionnaire at control

months follow-up

interventions listed below?”

- “Did you participate in any of the

- “Were you aware of the project?”

image in the organisation”

- “I believe that the program has a positive

desirable”

- “Implementing the program is highly

- “The program matched employee needs”

Aspects of the programme that may have influenced employee participation and satisfaction levels

Context - Process questionnaire at 6 and 12

Satisfaction of employees with the overall - Process questionnaire at 6 and 12 months - “Were you satisfied with the project and project (measured on a 10-point scale; very follow-up interventions?” dissatisfied to very satisfied) and specific interventions (measured on a 5-point Likert - Semi-structured interviews with - “What were positive points or points of scale very dissatisfied to very satisfied) employees at 6 and 12 months follow-up improvement for the implementation of among employees who participated in that the interventions” intervention.

Satisfaction

Implementation and effectiveness of lifestyle interventions

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

The complete theoretical framework has been described in more detail elsewhere [11]. To determine whether the interventions were implemented as planned and which interventions were maintained within the organisations, information was obtained from project members about dose delivered, fidelity and maintenance. The process components ‘recruitment’, ‘reach’, ‘dose received’ and ‘satisfaction’ were used to determine employee exposure to the interventions. More detailed information on the definitions and used methods are presented in table 1. Data collection procedures This process evaluation was conducted within the intervention departments only, except for the component contamination (for which data was gathered at the control departments). Data on the other seven process components were collected on two levels: project group level (i.e. project members from intervention departments) and participant level (i.e., employees from the intervention departments) The data presented in this paper were collected with: [1] Web-based questionnaires distributed among all employees to assess: a) The implementation process at 6-month (hospital n= 622; university n= 504) and at 12-month follow-up (hospital n=613; university n=489). The process questionnaire addressed awareness and satisfaction levels with the implemented interventions and the recruitment methods. The employees were asked to state in the questionnaire whether they were willing to participate in an additional semistructured interview to supplement the results. b) The effectiveness of the implemented interventions at baseline (T0) and 12-month follow-up (T2) in the employees in the intervention group (hospital T0 n=662, T2 n=663; university T0 n=546, T2 n=489) and control group (hospital T0 n=501, T2 n=462; university T0 n=484, T2 n=327). Self-reported physical activity levels, food intake, smoking status, alcohol use and vitality were assessed. [2] Semi-structured interviews with a random convenience sample of employees at 6 months (hospital n=7; university n=10) and at 12-month follow-up (hospital n=7; university n=5). Additionally, most project group members were interviewed at baseline (hospital n=8; university n=8), at 6 months (hospital n=6; university n=11) and at 12-month followup (hospital n=3; university n=7). The interviews were designed to address all nine process components listed above relating to the quality and extent of implementation of the interventions. At 18-months follow-up, the project leaders of both organisations were

Implementation and effectiveness of lifestyle interventions

contacted by telephone to assess programme maintenance. The principal investigator (DW) conducted all interviews during a face-to-face or telephone meeting at a time and location convenient for the participants. Prior to the start of the interview, all participants were informed about the purpose of the study, provided with reassurance about confidentiality, and all participants granted oral permission. All participants were selected by convenience sampling and were invited by email to participate in the interviews. Interviews with employees lasted an average of 24 minutes (range: 8-45 minutes) and those with project members lasted 28 minutes on average (range: 10 – 56 minutes). [3] Onsite monitoring: Throughout the study period, an embedded researcher (DW) continuously monitored the implementation process by documenting relevant email communications, minutes of project meetings and observations in pre-defined spread sheets which were based on the conceptual framework developed prior to the commencement of the study [11] in order to: a) further understand the context in which the project was implemented; b) learn about issues that project members were unaware of or that they were unwilling or unable to discuss candidly in the semi-structured interviews, and; c) assess whether or not the project or aspects of the project were delivered and operated as planned. Effectiveness of BRAVO@Work To assess employees’ physical activity levels, three questions from the validated “Injuries and Physical Activity in the Netherlands” questionnaire were used to measure whether employees had met the physical activity guidelines of the CDC/ACSM, which states that adults should engage in physical activity of at least moderate intensity for at least 30 minutes a day at least five days a week. To assess commuting activity (walking and cycling), one domain from the validated Short Questionnaire to Assess Health Enhancing Physical Activity was used [21, 22]. In addition, questions were added to assess self-reported stair use, elevator use and how often employees spent their lunch break in an active manner. Employees’ food intake was measured as the intake of fruit and vegetables. They were asked on how many days in a normal week they consumed vegetables and fruit and to report the average number of servings (50 grams) of vegetables and the number of pieces of fruit they consumed in a day. In order to assess employee vitality the sum score of four questionnaire items from the Copenhagen Psychosocial Questionnaire was used [23]. In addition, employees were asked to report their body height and body weight, as well as their smoking behaviour and alcohol consumption. 91

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

Data analysis Employees aged 18 years or older and with a contract for at least ten months at the start of the intervention were eligible for inclusion in the analysis. Descriptive statistics (i.e., percentage, mean and standard deviation) were used to analyse recruitment, reach, dose delivered, dose received and satisfaction of employees with the interventions and the overall programme. Differences over time for these process variables were tested for intervention group employees. When variables were continuous, independent t-testing was used, with paired t-testing being conducted for differences over time in the intervention group. Chisquare testing was performed when a variable was dichotomous. To determine effectiveness at 12-month follow-up, linear and logistic regression analyses were conducted with the variable of interest as the outcome, and group allocation (intervention or control) as the independent variable, adjusted for the baseline value of the outcome of interest and relevant covariates (company, gender, educational level and contract hours). The effect analysis included only participants for whom data were present for both time points. Data analyses were also performed to determine significant relationships between the compliance of workers with the interventions and the study outcomes. Compliance with the interventions was defined as (1) low compliance: â&#x2030;¤ 2 interventions and (2) high compliance: > 2 interventions based on the mean number of interventions in which employees had participated. Linear regression analyses were performed to test the differences between these compliance groups. All recorded interviews were transcribed verbatim. Transcripts were then read to establish a general picture of the concepts being studied and of the dynamics of the interviews. Using MAXQDA version 11 (VERBI GmbH, Berlin, Germany), transcripts were marked with open codes (descriptive codes within the immediate domain of the interview questions) and axial codes (analytic codes that represent emerging and overarching themes) [24, 25]. All codes were then grouped into central concepts related to all process components (such as satisfaction, recruitment and fidelity). One meeting was organised with the researchers (DW, LE, and PE) during the data analysis stage, during which they identified codes, concepts and themes, and discussed interpretations of the data to enhance the validity of the interpretation of the findings. In all cases consensus was reached through discussion.

Implementation and effectiveness of lifestyle interventions

Results Questionnaire respondents’ characteristics In the hospital intervention department, 215 of all the eligible employees (34.6%) filled out the process questionnaire at 6 months and 177 employees (28.9%) completed the questionnaire at 12 months. The university response rates were 216 (42.9%) and 174 (35.6%) at 6-month and 12-month follow-up. The demographic characteristics of the completers of the questionnaires in the two organisations were comparable at both time points (p>0.05) (Table 2). The response rates to the 12-months questionnaire for both control worksites were

28.4% (n=131) and 37% (n=121) for the hospital and university respectively. Table 2. Demographic characteristics of questionnaire respondents per company at T1 and T2. Academic Hospital University of Applied Science T1 T2 T1 T2 Gender (male) [% (n)] 9.0% (192) 9.1% (176) 22.4% (214) 21.4% (173) Age (year) [mean ± SD (n)] 41.9 ± 11.4 (206) 43.1 ± 11.1 (172) 45.1 ± 11.8 (208) 45.7 ± 11.8 (167) Work week (hours) 27.8 ± 9.7 (210) 28.0 ± 9.5 (177) 27.7 ± 9.1 (216) 28.1 ± 8.5 (173) [mean ± SD (n)] Work week (days) 3.8 ± 1.4 (211) 3.9 ± 1.3 (175) 3.7 ± 1.1 (216) 3.8 ± 1.0 (174) [mean ± SD (n)] Number of working years N/A 10.5 ± 8.8 (176) N/A 10.2 ± 9.4 (173) [mean ± SD (n)] Abbreviations: SD, standard deviation; N/A, not applicable, n, number of valid cases

Implementation of the interventions Both organisations developed the mandatory project plan. Despite the presence of a project plan at the hospital, delays and incompletion of the execution of some interventions could be observed, which could be related to (1) lack of ownership for the project by project members and (2) the fact that the project plan only contained a list of the general activities, desired changes, overall budget and goals of the project whereas a detailed intervention template, communication plan, budget specification and a list of involved persons with their tasks and responsibilities was lacking. For instance, no interventions were delivered between September and December (2011) and between January and April 2012. Because of the length of time between the interventions, most interviewed employees stated that they often thought that the project had already ended and as a result their willingness to participate was impaired: “I feel that the amount of time between the interventions is too long. So I know there are some more coming. But it’s been so long since there has been 93

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

something that you wonder whether it’s already finished or still on-going.” (UMCU-VB475). Interviews with project members at the university revealed that due to the presence of detailed project plan and a project leader who was assigned 16 hours a week solely for the implementation, programme delivery was structured throughout the academic year (20112012) without any delays in delivery of the intended interventions. Additionally, it should be noted that at the university the project members took one year for writing the project plan and preparing the project. Eventually, at the hospital, 12 out of 14 of the planned interventions (85.7%) were delivered to the employees, and at the university 21 out of 25 (84.0%) planned interventions were delivered (table 3). Main reasons for not delivering an intervention were (1) lack of approval from upper management due to rules and regulations related to the use of the building, (2) time constrains of project members, and (3) excessive costs. Continuous monitoring showed that both project groups mainly developed and implemented relatively simple interventions that did not require active employee participation (such as distribution of free fruit and posters). Most interventions chosen were environmental and educational interventions that could be implemented easily at low costs and effort. For example, one interviewed project member stated: “It (assigning a sports field) looks so simple but then it turns out not to be so simple at all due [rules and regulations regarding] the place and housing and things like that, despite low costs” [HU_P09 T1]. Eventually, at the hospital project members only choose to integrate the topic of vitality in employees’ annual performance interviews since the majority of managers expressed a positive attitude towards this change. This resulted in a pilot project with new funding, which will train managers in the use of the tools and abilities they need to conduct the appropriate interviews with a focus on sustainable productivity related to employees’ lifestyle. If this project proves successful, it will be implemented hospital-wide in early 2014. At the university, a human resources officer was instructed in December 2012 to embed some of the interventions in the organisations general health policy at the intervention faculty (table 3). The coaching programme, yearly bicycle check and stair massages are ongoing. The distribution of the free fruit was stopped one year later (in Fall 2013) due to excessive costs. The structural environmental interventions (like the enlarged bicycle shed, changes to staircases, changes to the routing, sitting balls, standing tables and the stricter smoking policy) were also maintained. Interviews with project members and the field notes revealed that the maintenance decisions were mainly based on the satisfaction rates with an intervention among project members and employees, low intervention costs, perceived success, and ease of implementation. 94

Implementation and effectiveness of lifestyle interventions

Table 3. Overview of the planned and delivered lifestyle interventions per company ACADEMIC HOSPITAL Type Intervention and content Timing GENERAL IBP Health checks: At the start and finish of the project, a health check was (1) May 2011 provided for the employees of the intervention department, and there was a (2) Jun 2012 stand with information leaflets about the project (kick-off). EDU Clinical lesson: Clinical lessons are regularly organised for all doctors within Dec. 2011 the departments. The project and health policy was the topic in one of these lessons. EDU Workshops for managers about vitality interviews: Two workshops were organised for all managers in the department with the aim of debating the importance of discussing vitality and health issues in annual performance interviews. EDU Team meetings: Team managers had the opportunity to invite the project’s external expert to one of their team meetings to supply information about one or more lifestyle themes. NUTRITION EDU Poster on nutrition: Interactive posters with information about the theme of nutrition were placed in the departments. EDU Recipe cards: Free cards with healthy recipes were left in every coffee corner. ENV Free fruit: For three months, a basket of free fruit from a local farmer was placed in the staffroom of the department every two weeks. HABITS (SMOKING & ALCOHOL) IBP Christmas event: Instead of the regular Christmas drinks, a Christmas event was organised with different workshops (including chair massage, mindfulness, cocktail shaking, coffee making, zumba dance) followed by a standing dinner. PHYSICAL ACTIVITY EDU Poster on physical activity and relaxation: Posters with information about the physical activity and relaxation theme were placed in the departments. IBP Lunch walks: Routes in the vicinity of the organisations were mapped out as suggestions for possible lunch walks. Employees were given the opportunity to subscribe to organised lunch walks, including lunch packages. Two lunch walks a week were organised for two weeks. ENV Lines on the floor to encourage stair use: Plans were made for placing lines on the floor of the department to route people via the stairs. This initiative was blocked by higher management. IBP

Pedometer competition: Plans were made for a competition involving the use of the pedometer. Time constraints prevented implementation. MENTAL HEALTH IBP Mindfulness sessions: Two mindfulness sessions were given consisting of exercises in concentration. The aim was to reduce stress, mood changes, fear, depression and concentration problems and to enhance the ability to cope with uncomfortable situations, feelings and thoughts. IBP Peer group counselling (fireplace conversations): Two conversations with a small group of employees (maximum 10) were organised to talk about what generates passion and energy relating to activities at work.

Dec. 2011

Jun. 2012

Jul. – Sep. 2011 Jul. – Sep. 2011 Jul. – Sep. 2011

Dec. 2011

Apr. – Jun. 2012 Apr - Jul. 2012

Not delivered

Jun. - Jul. 2012 Continued Jun. - Jul. 2012

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

UNIVERSITY OF APPLIED SCIENCES Type Intervention and content GENERAL IBP Annual opening ceremony: As part of the annual opening ceremony for 2011, employees were given the opportunity to attend several lifestyle theme related workshops (for example: zumba dance, mindfulness, life hacking, knowledge of foods). IBP Health week: Following the opening ceremony, a health week was organised in which activities, workshops and information on the projects lifestyle themes were provided (such as a lunch concert, yoga and dance workshops, lunch walks, debate on smoking, Nintendo Wii competition, joint breakfast). EDU Meeting for managers about vitality interviews: One meeting was organised for all department managers to debate the importance of discussing vitality and health issues in annual performance interviews.

Timing Sep. 2011

Sep. 2011

Dec. 2011

EDU Team meetings: Team managers were given the opportunity to invite the Apr.- June 2012 external expert to talk at team meetings about the projects’ lifestyle themes and to perform a self-analysis looking at work stress. ENV Quiet room: A room where employees can sit in silence. This did not go Not delivered through in the end because there was no space in the building. NUTRITION ENV Free Fruit: Free fruit was placed once a week in the staffroom over a period Nov.– Dec. 2011 of two consecutive months in 2011 and 2012. and Apr. – Jul. 2012 Continued EDU Superfoods in the canteen: Certain healthy foods were spotlighted by Apr. – May 2012 providing information about the positive qualities of the product on a poster and Sept. 2012and by developing a recipe that incorporated the superfood and selling this Jun. 2013 in the canteen. Jan. 2012 EDU Flyer on exemplary behaviour: A flyer for employees with children was developed including information about the importance of eating a variety of fruit and vegetables with the whole family. ENV Analysis of food in the canteen: The food on offer in the canteen was analysed and displayed on posters which formed the basis for negotiations with the canteen caterer and eventually led to the introduction of a salad bar in the canteen. SMOKING ENV Stricter smoking policy: In front of the building, blue lines were placed on the ground to mark out the non-smoking area. Compliance with the policy was enforced. PHYSICAL ACTIVITY IBP Bicycle check: Two bicycle checks were offered to employees: a mobile cycle repairman was called in to check and repair bikes. ENV Enlarged bicycle shed: The bicycle shed was enlarged to increase the capacity for bicycles and make space for loan bicycles. ENV Changes to staircases: Two staircases in the building were made more attractive. One was decorated (with nature photos) and one was made into a hopscotch game.

Jan. 2012

Nov. 2011 Structural

Nov - Dec 2011 Continued Dec 2011 Structural Feb. – Jul. 2012 Structural

Implementation and effectiveness of lifestyle interventions

ENV New building routing: A new building routing was developed that was placed on the floor and that specifically integrated the staircases that were difficult to find.

Feb. – Jul. 2012 Structural

IBP

Sports materials on loan: A central location was set up with sports materials for loan (such as balls and Frisbees). IBP Route maps for lunch walks:Maps were made with routes in the vicinity of the organisation, including distance and duration. ENV Point of decision prompts to encourage stair use: Posters about the advantages of using stairs were placed in and around elevators and staircases to promote stair use.

Jan. 2012

ENV Sitting balls: Sitting balls were distributed to each department team to encourage an active sitting position. ENV Standing tables: To make it possible to have meetings while standing, standing tables were ordered and a room was assigned to house the tables. IBP Coaching trajectory (pilot project): The employees of one team were given the opportunity to participate in an intensive 15-week coaching project with 4 coaching sessions and 3 workshops on setting goals and stress management, an extensive health check and twice-weekly 1-hour training sessions with a trainer in a local fitness centre.

Sept. 2012 Structural Jul. 2012 Structural Sep. – Nov. 2012 Continued

IBP

Table tennis table: Plans were made for placing a table tennis table outside the building. This did not go through because the location was too windy and there was no alternative.

Not delivered

IBP

Bicycle buddies: To encourage biking for commuting purposes, plans were made for a buddy system. This did not go through because of a lack of interest among employees

Not delivered

IBP

Not delivered Sports in local fitness centre: Sigma is the university fitness hall where employees were given the opportunity to cycle at a discount. This did not go through because it was too expensive.

MENTAL HEALTH IBP Chair massages: During a period of four hours on three afternoons a week, a 15-minute chair massage from physiotherapy students was available to employees. EDU Books on time management: Books on time management and life hacking have been placed in the staffroom.

Apr. 2012 May. – Sep. 2012 Structural

Apr. – Dec. 2012 Continued Apr. – May 2012

Abbreviations: ENV = environmental intervention, EDU = educational intervention, IBP = individuallybased interventions that required committed and active participation.

Employee recruitment for, exposure to, and satisfaction with the interventions Employees from both organisations were recruited for participation in the project by postcards with general project information at the launch of the implementation send to employees’ home addresses (employee awareness T1: 64.3%, n=247). Additionally, at the hospital, a kick-off was organised in the form of health checks with an information stand and, at the university, the project was launched during the annual opening ceremony of the academic year followed by a health week with workshops and an information stand (table

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

4). During the project both organisations send informative e-mails (awareness T2: 76.0%, n=256) and launched an official project website on the organisations’ intranet (awareness T2: 51.2%, n=171). Additionally, other methods were used, such as placing project banners at the entrances of the intervention departments at the hospital and a monthly e-mail update after six months, with specific project information at the university. The interviews with employees from both organisations revealed that most employees were satisfied with the recruitment methods. However, some interviewees said that a personal approach would have been better because they did not read the information on the website and already receive large number of emails every day. Of all respondents, 89.5% (n=383) were aware of the project at 6-month follow-up and 96.6% (n=338) at 12-month follow-up (table 5). Of the latter respondents, 76.6% (n=258) stated that they were also aware of the project goals. Frequently mentioned project goals by the respondents and interviewees at both organisations included: creating awareness of the benefits of a healthy lifestyle, and promoting a healthy lifestyle in order to reduce sickness absence rates and to improve employee performance. Additionally, 75.1% (n=386) of the respondents of both organisations had participated in at least one intervention (excluding environmental interventions and posters). Non-participating respondents worked an average of 4.7 hours less per week than participating respondents at 6-month follow-up (p<0.05). At the university, non-participating respondents worked an average of 24.1 hours in 3.3 days and participating respondents had worked an average of 28.6 hours in 3.8 days (p<0.05). However, this difference was observed only at 6 months. On average, the hospital employees participated in 2.2 ± 1.6 [range 0-8] of nine interventions that were eligible for participation and university employees in 3.0 ± 1.7 [range 0-9] out of eleven interventions eligible for participation. Employees of both organisations who participated in at least one intervention during the study period and responded to both questionnaires (n=164) rated the overall project as moderate at both 6-month followup (6.7 ± 1.6) and 12-month follow-up (6.8 ± 1.1), with no significant change over time (p>0.05) (table 5). Interviewed employees were mainly satisfied with the overall programme because they perceived the programme to be a good initiative which showed that their employer acknowledged the value of employee health and lifestyles, especially since they are working in the health sector. Other reasons mentioned by a majority of the interviewees were ‘the programme generated awareness about the importance of a healthy lifestyle’, and ‘most interventions were perceived to be “pragmatic” and thought to facilitate employees’. Nonetheless, respondents were also very critical about the need for some of the developed

Implementation and effectiveness of lifestyle interventions

interventions and said that, in some cases, a long time interval between interventions had had a negative effect on interest in the project. Additionally, interviewed employees commented that future implementation of health programmes in their organisation should focus more on comprehensive interventions (like the coaching activities at the university) embedded in the companyâ&#x20AC;&#x2122;s health policy. Other suggested improvements were that more efforts should be made to specifically target unhealthy employees and that one employee should be appointed in each department team as an ambassador for the project to raise awareness and participation levels in the organisation. The highest participation and satisfaction rates in both organisations were found for the distributed free fruit. In general, employees were most aware of and participated in easily accessible interventions like environmental interventions and interventions that were part of the existing meetings or events in the organisation that stimulated social interaction with colleagues (table 4). The interviews with employees revealed that employees did not use the materials or actively participated in interventions mainly when the intervention was not easily integrated into their working life (high work demands or inconvenient time and location of intervention) or when the intervention was not perceived as interesting since it did not fit their needs. Interviewed employees were more inclined to be positive about an intervention when they perceived the intervention as relevant, easily accessible, and matching their needs. Contamination Awareness of the project was confirmed by 31.3% of the hospital control respondents and 21.7% of the university control respondents. The questionnaire results showed that almost no employees (a maximum of 2) at the two control worksites were aware of some of the interventions and had also participated in some interventions. However, at the university, the human resources department was so enthusiastic about the free fruit that they decided to implement this intervention at all faculties of the university during the intervention year, despite the objections of the research group. Consequently, 36.8% (n=42) employees of the control group reported that they received the free fruit.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

Table 4. Reach, Dose Received and Employee Satisfaction of the interventions Implemented interventions

Reach Dose Received Satisfaction (1-5 scale) [% aware (n)] [% participated (n)] [mean ± SD (N)]a

ACADEMIC HOSPITAL General Health check (T1) Health check (T2) Managers meeting * Nutrition Poster Nutrition Recipe cards Free fruit Habits Christmas event Clinical lesson * Physical Activity (PA) and mental health Poster on PA and relaxation Lunch walks Mindfulness session Peer group counselling * Team meetings UNIVERSITY OF APPLIED SCIENCEb General Opening ceremony (kick-off) Health week (kick-off) Meeting for managers * Nutrition Free fruit (T1) Free fruit (T2) Superfoods in the canteen Flyer exemplary behaviour * Analysis of the canteen foods Physical Activity Bicycle check Enlarged bicycle shed Adjusted staircases (A: hopscotch, B: nature pictures) Renewed building routing Sport materials on loan Route maps for lunch walks Point of decision prompts to encourage stair use Smoking Stricter smoking policy Mental health Chair massages Books on time management

N/A N/A 10.7% (18)

40.1% (65) 23.5% (38) 4.7% (8)

N/A N/A 3.9 ± 0.6 (8)

79.9% (151) 79.9% (151) 66.5% (125)

N/A 38.1% (72) 55.3% (104)

3.3 ± 0.7 (151) 3.7 ± 0.7 (70) 4.3 ± 0.7(104)

83.5% (157) N/A

30.9% (58) 1.6% (3)

4.0 ± 0.9 (58) 3.3 ± 0.6 (3)

56.2% (95) 75.1% (127) 76.9% (130) 26.6% (45) 21.9% (37)

N/A 5.3% (9) 7.7% (13) 7.1% (12) 9.5% (16)

3.6 ± 0.6 (61) 4.3 ± 0.5 (7) 4.1 ± 0.7 (11) 4.0 ± 0.6 (11) 3.8 ± 0.8 (15)

82.4% (155) 89.0% (170) 12.7% (21)

36.5% (69) 62.3% (119) 8.4% (14)

N/A 3.9 ± 0.8 (116) 4.2 ± 0.6 (11)

89.9% (170) 95.2% (158) 42.2% (70) 11.4% (19) 36.7% (61)

69.8% (132) 84.3% (140) 18.1% (30) 5.4% (9) N/A

4.4 ± 0.6 (132) 4.6 ± 0.6 (131) 4.3 ± 0.9 (27) 3.8 ± 0.7 (8) 4.0 ± 0.6 (43)

57.2% (95) 42.8% (71) 47.6% (79)

21.1% (35) N/A N/A

72.9% (121) 9.0% (15) 39.8% (66) 73.5% (122)

N/A 0.0% (0) 6.6% (11) N/A

3.8 ± 0.6 (17) 4.6 ± 0.6 (60) A: 3.0 ± 1.2 (49) B: 3.8 ± 0.8 (48)** 3.7 ± 1.1 (107) N/A 4.6 ± 0.5 (11) 4.0 ± 0.8 (102)

88.6% (147)

N/A

3.9 ± 1.2 (116)

77.1% (128) 13.3% (22)

26.5% (44) 5.4% (9)

4.7 ± 0.5 (40) 3.8 ± 0.8 (6)

Abbreviations: N/A = not applicable (in the case of an environmental intervention) or not measured, N, number of valid cases for the question; n, number of respondents, M=mean, SD=standard deviation.

100

Implementation and effectiveness of lifestyle interventions

* Only accessible by a specific group of invited employees ** p<0.05: satisfaction rates for A and B significantly differ from each other a Satisfaction rates are only calculated for respondents who stated they participated in the specific individually based or educational intervention or were aware of the environmental intervention b for the coaching trajectory, team meetings, sitting balls and standing tables no data was available since these interventions were implemented after the distribution of the second process questionnaire. Table 5. Reach, Dose Received and Employee Satisfaction of the overall project Academic Hospital T1 T2

University of Applied Science T1 T2

Reach Awareness of project [% (n)] 90.6% (193) 97.7% (172) 88.4% (215) 95.4% (174) Awareness of project goals [% (n)] N/A 76.7% (132) N/A 76.4% (126) Dose receiveda Participation in at least 1 intervention 71.7% (134) 40.7% (66) 80.7% (151) 90.4% (150) [% (n)] Number of interventions received per 1.3±1.0 (187) 0.6±0.9 (162) 1.7±1.1 (187) 1.8±1.2 (166) employee [mean ± SD (N)] Satisfaction Satisfaction with project (grade 1-10) 6.9 ± 2.0 (82) 6.7 ± 1.2 (82) 6.5 ± 1.1 (82) 7.0 ± 0.9 * (82) [mean ± SD (N)] Abbreviations: SD, standard deviation; N/A, not applicable; N, number of valid cases for the question; n, number of respondents * significant difference between T1 and T2 P<0.05 a Environmental interventions are excluded

Effectiveness of BRAVO@Work From 145 respondents data was available from baseline as well as 12 months follow-up and were included in the analyses. Regression analyses (table 6) identified an intervention effect on the number of days of fruit consumption per week (β: 0.44 days/week, 95% CI 0.02 to 0.85) in favour of the intervention group. No other significant effects were found. In both the intervention and control groups, the number of pieces of fruit a day rose from baseline to 12-month follow-up (+1.65 and +1.67 pieces/day respectively), with no significant differences between the groups (β: -0.042 pieces/day, 95% CI -0.291 to 0.208). For the relationship between the outcome measures and participation levels (low participation ≤ 2 and high participation > 2), logistic regression analysis showed a significant relationship between days of vegetable intake and high participation in the intervention group (β: -0.358, 95% CI -0.691 to -0.026), as compared to low participation in favour of the low participation group.

101

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

102 5.6% 46.4% 39.7%

T0m, Percentage 4.2% 53.2% 39.7%

T12m, Percentage 72 73 73

61 72

3.15 (0.39) 2.42 (0.82) 4.90 (4.47) 4.65 (3.88) Control group

n 67 73 73 72 45 72 31 72 70

T12m, Mean (SD) 23.84 (2.94) 5.10 (1.54) 3.34 (2.01) 2.81 (1.96) 3.96 (2.57) 5.39 (1.77) 3.35 (0.75) 6.41 (0.95) 2.87 (0.82)

24.03 (3.21) 5.37 (1.41) 3.62 (2.18) 2.58 (2.35) 3.31 (2.66) 5.32(1.96) 1.68 (0.48) 5.15 (2.50) 2.87 (0.83)

T0m, Mean (SD)

Control group

23.67 (3.53) 5.18 (1.60) 3.49 (1.99) 2.64 (1.88) 4.60 (3.71) 5.89 (1.63) 3.14 (0.87) 6.11 (1.17) 3.08 (0.89)

T12m, Mean (SD)

9.1% 39.7% 35.2%

T0m, Percentage

9.1% 39.4% 44.1%

T12m, Percentage

3.24 (0.41) 2.67 (0.97) 5.37 (5.45) 5.10 (5.00) Intervention group

23.63 (3.51) 4.91 (1.68) 3.54 (2.11) 2.60 (1.85) 4.64 (3.47) 5.50 (1.95) 1.53 (0.50) 5.64 (2.29) 3.15 (0.82)

T0m, Mean (SD)

Intervention group

132 145 8.9%

126 132

127 133 133 133 99 132 70 132 132

0.402 (0.060 to 2.691) 0.482 (0.094 to 2.475) 1.509 (0.787 to 2.894)

OR (95% CI)

0.259 (-0.033 to 0.551) 0.150 (-0.577 to 0.877)

0.006 (-0.007 to 0.019)a 0.304 (-0.119 to 0.727) 0.322 (-0.199 to 0.844) -0.253 (-0.766 to 0.260) 0.253 (-0.591 to 1.096) 0.436 (0.020 to 0.851)* -0.042 (-0.291 to 0.208) 0.278 (-0.016 to 0.571) 0.102 (-0.116 to 0.321)

Î˛ (95% CI)

* p<0.05 For the regression analyses, body mass index was converted into Standard Deviation Scores (SDS) using Dutch growth references [34]. The SDS expresses the measurement relative to a reference population in units of standard deviations above or below the median [35]. Î˛, estimated intervention effect from linear regression analysis adjusted for gender, education level, contract hours, company; OR, odds ratio, estimated intervention effect from logistic regression analysis adjusted for gender, education level, contract hours, company and group allocation (0=control, 1=intervention group); C, control group; I, intervention group; MVPA, moderate-to-vigorous physical activities; VPA, vigorous physical activities.

Smoking (yes) Active commuting (yes) Active lunch break (yes)

Body Mass Index MVPA 30 minutes p/day MVPA 60 minutes p/day VPA 20 minutes p/day Stair use Fruit days/week Fruit pieces/days Vegetables days/week Vegetable servings (=50 g)/ day Vitality Alcohol glasses/week Dichotomous outcome measures

Continue outcome measures

Table 6. Mean and percentages for outcome measures at baseline and 12-months follow-up intervention and control group.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 Chapter 4

Implementation and effectiveness of lifestyle interventions

Discussion The main aim of this study was to systematically evaluate the implementation of a multicomponent lifestyle intervention at two different worksites. To our knowledge, this is the first published process evaluation looking at the natural course of implementation of lifestyle interventions coordinated and implemented by employees and managers without active involvement of the researchers, who acted as embedded observers only. The first aim of this study was to study the extent and quality of programme delivery and maintenance. High rates for dose delivered (84.0% and 85.7%) in the intervention departments were observed and a mix of environmental, educational and individual interventions were implemented. Although in this study the project groups at both organisations were responsible for the implementation, the dose delivered rates we found compare well with other WHPPs in which the researchers were responsible, with dose delivered ranging from 72% to 86% [26, 27]. Hence, rates above 80% can therefore be considered good and are not dependent on who delivers the interventions. However, both project groups mostly opted for relatively simple to implement and â&#x20AC;&#x2DC;funâ&#x20AC;&#x2122; educational and environmental interventions like the free fruit and posters, and did not pay much attention to the evidence base of the interventions that were chosen. The interventions chosen are more likely to create awareness among employees rather than change behaviour. The results indicated that adherence to the original project plan (fidelity) at the university was high, whereas in the hospital more difficulties with programme delivery were experienced. A comparison of the two organisations indicated that programme implementation may be facilitated when, ad forehand, a detailed project plan is developed including the desired changes and goals, an intervention template and a budget per intervention, a description of the explicit responsibilities of each project member and a timeline. This may have helped management staff to take informed decisions when they allocated budgets to the interventions they wanted to have implemented. By contrast, the more ad hoc approach in the hospital has the risk of a decline in participation and satisfaction rates among employees, due to delayed decision-making and the failure to deliver interventions on time despite the rapid launch of the project. This ad hoc approach could also be the reason that none of the interventions in the hospital relating to physical activity, nutrition or habits were maintained after the project had ended. They only opted for the continuation of vitality in employeesâ&#x20AC;&#x2122; yearly performance interviews that, interestingly, was the only intervention that was well thought out and the only intervention

103

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

that the majority of managers and project members were satisfied with. Funding for a pilot project that focuses on training managers in conducting these performance interviews was obtained. If successful, it is expected that, by early 2014, all managers at the hospital will be required to discuss sustainable productivity (and lifestyle-related factors) in annual performance interviews. The university continued three interventions (chair massages, coaching programme and the bicycle check) after the project. This decision was also mainly based on employees’ and managers satisfaction with the interventions. So, the sustainability of ‘simple and fun’ worksite lifestyle interventions, with the exception of environmental interventions, at the university (i.e., the adjusted building routes and standing tables) is can be regarded as low if these do not closely fit employees’ needs or if the costs are excessive. So, the main driver of intervention uptake in daily practice is therefore ease of implementation, degree of satisfaction and a match to employees’ needs regardless of how effective the intervention was in terms of changing the targeted behaviour. With respect to the second aim - establishing a picture of employee recruitment, reach, participation and satisfaction levels - we can conclude that the majority of employees were aware of the programme (96.6%) and had participated in at least one intervention (75.1%) during programme implementation. So the methods used by both organisations for recruiting employees to the programme were apparently effective. However, employees generally reported a preference for a more personal approach, such as an introduction to the project in their monthly team meetings, rather than a dedicated project website and emails which were poorly read. Such a personal approach encourages interaction and would enable employees to discuss and clarify any ambiguities beforehand. This could have facilitated their participation in, and furthered their satisfaction with the programme as a whole. Employees also stressed that more efforts should be made in the future to target relatively unhealthy employees. Nevertheless, the high awareness rates among respondents in this study suggest that both organisations successfully communicated the programme through the organisation. This can be considered a necessary first step in obtaining high participation rates and, ultimately, in changing employee lifestyle behaviour [28]. Furthermore, the participation rates (75.1%) in this study for the programme as a whole can be considered high since, in general, participation levels in WHPPs vary widely (10% to 76%) [28, 29]. A possible explanation for the high overall participation rates in this study is the offering of multiple interventions as shown by a systematic review by Robroek et al. (2009). By addressing multiple lifestyle themes, a programme reaches employees

104

Implementation and effectiveness of lifestyle interventions

with different needs relating to different lifestyle aspects, and this leads to higher overall participation compared to programmes that implement physical activity interventions only [29]. Furthermore, both organisations included a mental health component focusing on work stress and relaxation at work (chair massages, mindfulness sessions, peer group counselling) since the results of previous work satisfaction surveys at both organisations had shown that employees perceived work stress levels as high. Due to the fact that their employer acknowledged this problem by specifically addressing mental health, employees possibly were more appreciative of the entire programme and more inclined to participate. Focusing exclusively on lifestyle aspects like physical activity, nutrition and smoking may lead employees to feel that the employer is intervening too much in the personal lives of the employees. By contrast, adding a mental health component allowed the organisations to show that management cared about the well-being of employees and their working conditions. Differences were found in terms of actual programme participation rates between the two organisations over time and for the specific interventions. Participation rates at the hospital declined over time, possibly due to a very limited project plan. Whereas at the university rates inclined, possibly due to the longer preparation time they took for writing the plan. Despite the decline in participation at the hospital and the rise in participation at the university, employee satisfaction with the programme as a whole was moderate (â&#x2030;Ľ6 and <7.5), even though satisfaction levels for the specific interventions were found to be good. However, in a more comprehensive and individually-based worksite lifestyle intervention by Strijk et al. 2011 that included yoga, workout sessions and a personal vitality coach, satisfaction with the programme was good (â&#x2030;Ľ7.5) [27]. This suggests that, although employees were satisfied with the specific interventions, overall programme satisfaction could have been higher if more individually-based interventions had been implemented. Finally, the effectiveness of the entire programme was evaluated in a quasi-experimental controlled trial conducted alongside the process evaluation. Although multiple lifestyle interventions targeting different behaviours were implemented, only a favourable change over time was observed in the number of days of self-reported fruit consumption but not in the number of pieces of fruit per day. It should be noted that the free fruit was also distributed to the control group at the university due to an over-enthusiastic HRM department. Despite this contamination, an effect was still found in favour of the intervention group. This may be explained by the results on dose received that show that employees from the intervention group (84.3%) had made more use of the fruit than the

105

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

participants in the control group (36.8%). Because of the lack of any effect on other primary outcome measures, only the distribution of the free fruit could be described as effective. The lack of effects on other lifestyle behaviours could possibly be explained by the type of the implemented interventions, as explained earlier. Another explanation could be the lack of power to detect possible effects due low response rate to the questionnaire. Possibly explanations thereof could be the extensive length of the questionnaire, technical problems with filling out the questionnaire and employee suspicion towards their employer about the privacy of the data. The issue of whether it was possible with our chosen study design (quasi-experimental controlled trial) to detect any difference in lifestyle behaviour between the intervention and control groups should be considered also. Although the results did not find any major differences in the demographic characteristics between the two groups, from a methodological perspective it is better to investigate the effectiveness of an intervention in a randomised controlled trial (RCT), with the organisations being randomised. However, a RCT design is not the most ideal design in studies like these where the primary focus is on the natural implementation process of an intervention. A RCT takes place in controlled conditions (research driven), whilst implementation in ‘real life’ practice requires flexibility in the conditions that can be controlled. However the results of the latter probably say more about the generalizability. Strengths and limitations A major strength of this process evaluation is that we evaluated the implementation process on the basis of a comprehensive framework that was developed prior to the start of the implementation. As a result, data was collected continuously from the start of the implementation, which resulted in detailed information about the real-time implementation of the individual interventions. Since data was collected from all relevant stakeholders by combining quantitative and qualitative evaluation methods, more in-depth information was obtained about the reasons for not implementing an intervention participating in an intervention, and about the quality of intervention delivery. The qualitative evaluation therefore helped to interpret the results from the questionnaires [33]. The fact that the “embedded” researcher monitored the implementation process could be a strength, but it may also be a limitation. By being present in the organisation, the researcher may unintentionally affect the implementation process, even when, as here, the researcher does not actively intervene in the process. For example, at the university, the project would probably have terminated prematurely if the researchers and external advisor had not been

106

Implementation and effectiveness of lifestyle interventions

present. As a result of their presence, the project members made the project a priority and did not allow the project to get side-tracked because of high work pressure. Nevertheless, the chosen design produced valuable information about the implementation process, reasons for programme adaptation and the failure to deliver some interventions that might not have been obtained in interviews, during which project members may have given social desirable answers. A limitation in this study is the possibility of selection bias because not all employees completed the questionnaires or participated in the interviews. However, by a random convenience sampling based on employeeâ&#x20AC;&#x2122;s time constraints and work demands for recruiting employees to the interviews, we tried to ensure that we interviewed employees with different participation rates and satisfaction levels. Another strength of this evaluation was the proportion of process components included in this study, which allowed us to acquire a full overview of implementation at all levels. This is a change from other process evaluations, which often focus primarily on dose received [10]. Conclusion and implications for future research This study indicates that the primary precondition for a WHPP in a non-research context is not that an intervention has to be evidence-based, but that it should be easy to implement, with low costs and minimal effort. The intervention departments in this study, and probably most organisations in the Netherlands, are not eager to spend a large amount of their budget and time on implementing more comprehensive â&#x20AC;&#x2DC;evidence-basedâ&#x20AC;&#x2122; individual interventions from the list of known effective interventions. However, this does not seem to have had an effect on employee awareness, participation and satisfaction since the results in these areas were comparable to, if not better than, other more comprehensive lifestyle interventions. As most known evidence-based worksite lifestyle interventions are developed and implemented with coordination from academic researchers, our findings raise questions about the practical applicability of most of these interventions. There is little or no practical information available for organisations about these interventions that they can use as a guide to adopt and implement an evidence-based, complex, lifestyle intervention. This means that they will opt for other, more simple, interventions. Our results suggest that, if interventions are to be implemented in daily practice, they should be visible and easy to implement, require low cost and minimal effort, and match the needs of the organisation and the employee. Furthermore, employees will be more inclined to participate in more simple interventions which do not require active participation and in which they can participate when that is

107

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

convenient for them, as in the case of the free fruit. However, these simple interventions are not continued and in this project only the more elaborate interventions were maintained (i.e., discussing vitality in employees yearly performance interviews, chair massages and the coaching trajectory). The main challenge for upcoming implementation research will be to strike a balance between implementing evidence-based interventions that are known to be effective and to leave the coordination of the implementation to organisationsâ&#x20AC;&#x2122; employees and managers.

108

Implementation and effectiveness of lifestyle interventions

References 1.

van Duijvenbode DC, Hoozemans MJ, van Poppel MN, Proper KI: The relationship between overweight and obesity, and sick leave: a systematic review. Int J Obes (Lond) 2009, 33(8):807-816.

Goetzel RZ, Ozminkowski RJ: The health and cost benefits of work site health-promotion programs. In Edited by Anonymous 2008:303-323. [Anonymous, vol 29.]

Anderson LM, Quinn TA, Glanz K, Ramirez G, Kahwati LC, Johnson DB, Buchanan LR, Archer WR, Chattopadhyay S, Kalra GP, Katz DL: The Effectiveness of Worksite Nutrition and Physical Activity Interventions for Controlling Employee Overweight and Obesity. A Systematic Review. Am J Prev Med 2009, 37(4):340-357.

Jepson RG, Harris FM, Platt S, Tannahill C: The effectiveness of interventions to change six health behaviours: A review of reviews. BMC Public Health 2010, 10.

Ni Mhurchu C, Aston LM, Jebb SA: Effects of worksite health promotion interventions on employee diets: A systematic review. BMC Public Health 2010, 10.

Groeneveld IF, Proper KI, Van Der Beek AJ, Hildebrandt VH, Mechelen WV: Lifestyle-focused interventions at the workplace to reduce the risk of cardiovascular disease - A systematic review. Scandinavian Journal of Work, Environment and Health 2010, 36(3):202-215.

Verweij LM, Coffeng J, van Mechelen W, Proper KI: Meta-analyses of workplace physical activity and dietary behaviour interventions on weight outcomes. Obesity Reviews 2011, 12(6):406-429.

Saunders RP, Evans MH, Joshi P: Developing a process-evaluation plan for assessing health promotion program implementation: a how-to guide. Health promotion practice 2005, 6(2):134-147.

Grimshaw JM, Zwarenstein M, Tetroe JM, Godin G, Graham ID, Lemyre L, Eccles MP, Johnston M, Francis JJ, Hux J, Oâ&#x20AC;&#x2122;Rourke K, Legare F, Presseau J: Looking inside the black box: a theory-based process evaluation alongside a randomised controlled trial of printed educational materials (the Ontario printed educational message, OPEM) to improve referral and prescribing practices in primary care in Ontario, Canada. Implement Sci 2007, 2:38.

10. Wierenga D, Engbers LH, van Empelen P, Duijts S, Hildebrandt VH, van Mechelen W: â&#x20AC;&#x153;What is actually measured in process evaluations for worksite health promotion programs: a systematic review. Accepted for publication in BMC Public Health 2013. 11. Wierenga D, Engbers LH, van Empelen P, Hildebrandt VH, van Mechelen W: The design of a real-time formative evaluation of the implementation process of lifestyle interventions at two worksites using a 7-step strategy (BRAVO@Work). BMC Public Health 2012, 12(1):619. 12. Glasgow RE, Klesges LM, Dzewaltowski DA, Bull SS, Estabrooks PA: The future of health behavior change research: what is needed to improve translation of research into health promotion practice? Ann Behav Med 2004, 27(1):3-12. 13. Bartholomew LK, Parcel GS, Kok G, Gottlieb NH: Planning Health Promotion Programs: An Intervention Mapping Approach. San Francisco, CA: Jossey-Bass; 2006. 14. Steckler A, Linnan L: Proces Evaluation for Public Health Interventions and Research: San Francisco: Jossey-Bass; 2002. 15. Oakley A, Strange V, Bonell C, Allen E, Stephenson J, RIPPLE Study Team: Process evaluation in randomised controlled trials of complex interventions. BMJ 2006, 332(7538):413-416. 16. Baranowski T, Stables G: Process evaluations of the 5-a-day projects. Health Educ Behav 2000, 27(2):157-166. 17. Rongen A, Robroek SJW, van Lenthe FJ, Burdorf A: Workplace Health Promotion: A Meta-Analysis of Effectiveness. Am J Prev Med 2013, 44(4):406-415. 18. Anderson LM, Quinn TA, Glanz K, Ramirez G, Kahwati LC, Johnson DB, Buchanan LR, Archer WR, Chattopadhyay S, Kalra GP, Katz DL: The Effectiveness of Worksite Nutrition and Physical Activity Interventions for Controlling Employee Overweight and Obesity. A Systematic Review. Am J Prev Med 2009, 37(4):340-357.

109

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

19.

Chau JY, der Ploeg HP, van Uffelen JG, Wong J, Riphagen I, Healy GN, Gilson ND, Dunstan DW, Bauman AE, Owen N, Brown WJ: Are workplace interventions to reduce sitting effective? A systematic review. Prev Med 2010, 51(5):352-356.

20. Glasgow RE, Vogt TM, Boles SM: Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health 1999, (89):9-1322. 21. de Hollander EL, Zwart L, de Vries SI, Wendel-Vos W: The SQUASH was a more valid tool than the OBiN for categorizing adults according to the Dutch physical activity and the combined guideline. J Clin Epidemiol 2011, (1878-5921; 0895-4356). 22. Wendel-Vos GC, Schuit AJ, Saris WH, Kromhout D: Reproducibility and relative validity of the short questionnaire to assess health-enhancing physical activity. J Clin Epidemiol 2003, 56(0895-4356; 0895-4356; 12):1163-1169. 23. Kristensen TS, Hannerz H, HĂ¸gh A, Borg V: The Copenhagen Psychosocial Questionnaire--a tool for the assessment and improvement of the psychosocial work environment. Scand J Work Environ Health 31(6):438-449. 24. Pope C, Ziebland S, Mays N: Qualitative research in health care. Analysing qualitative data. BMJ 2000, 320:114-116. 25. Glaser BG: The Constant Comparative Method of Qualitative Analysis. Social Problems 1965, 12:436445. 26. Verweij LM, Proper KI, Hulshof CT, van Mechelen W: Process evaluation of an occupational health guideline aimed at preventing weight gain among employees. J Occup Environ Med 2011, 53(7):722729. 27. Strijk JE, Proper KI, van der Beek AJ, van Mechelen W: A process evaluation of a worksite vitality intervention among ageing hospital workers. Int J Behav Nutr Phys Act 2011, 8:58. 28. Glasgow RE, McCaul KD, Fisher KJ: Participation in worksite health promotion: a critique of the literature and recommendations for future practice. Health Educ Q 1993, 20(3):391-408. 29. Robroek SJ, van Lenthe FJ, van Empelen P, Burdorf A: Determinants of participation in worksite health promotion programmes: a systematic review. Int J Behav Nutr Phys Act 2009, 6:26. 30. Fleuren M, Wiefferink K, Paulussen T: Determinants of innovation within health care organizations: literature review and Delphi study. Int J Qual Health Care 2004, 16(2):107-123. 31. Linnan LA, Sorensen G, Colditz G, Klar DN, Emmons KM: Using theory to understand the multiple determinants of low participation in worksite health promotion programs. Health Educ Behav 2001, 28:591-607. 32. Oude Hengel KM, Blatter BM, Van Der Molen HF, Joling CI, Proper KI, Bongers PM, Van Der Beek AJ: Meeting the challenges of implementing an intervention to promote work ability and healthrelated quality of life at construction worksites: A process evaluation. Journal of Occupational and Environmental Medicine 2011, 53(12):1483-1491. 33. Mays N, Pope C: Qualitative research in health care: Assessing quality in qualitative research. Br Med J 2000, 320(7226):50-52. 34. SchĂśnbeck Y, Talma H, van Dommelen P, Bakker B, Buitendijk SE, Hirasing RA, van Buuren S: Increase in prevelance of overweight in Dutch children and adolescents: a comparison of nationwide growth studies in 1980, 1997 and 2009. PLoS One 2011, 6(11):e27608. 35. Cole TJ, Green PJ: Smoothing reference centile curves: the LMS method and penalized likelihood. Stat Med 1992, 11:1305-1319.

110

What is the relationship between implementation factors and employee participation in and satisfaction with a worksite health promotion program? Wierenga D, Engbers LH, van Empelen P, van Mechelen W. Submitted for publication

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

Abstract The aim of the present study is to gain insight into which implementation factors are associated with employee participation and satisfaction levels of a worksite health promotion program. Thirteen implementation factors derived from literature and participation and satisfaction levels were assessed among employees in the intervention group of two worksites using questionnaires during (T1) and after (T2) implementation. Linear and logistic regression analyses were performed to assess the associations of interest. Results showed that positive attitude towards program implementation at T1 (OR 2.06) was best associated with high participation at T2. The ‘programs match with employee’s needs’ (β: 0.22), ‘positive attitude towards program implementation’ (β: 0.42), and ‘positive attitude towards employer involvement’ (β: 0.20) at T1 was associated best with overall satisfaction at T2. Whereas the ‘programs fit with the organization’ (β: 27), ‘positive attitude towards program implementation’ (β: 32), ‘positive program image’ (β: 0.24), and ‘program notification to new employees’ (β: 0.13) at T2 were best associated with higher overall satisfaction levels. Overall, this study shows that too potentially increase the success of a program, implementers should ensure that the program fits employees needs and that employees have a favourable attitude towards program implementation.

112

Relationship between barriers and facilitators and implementation

Introduction Workplaces offer a unique setting to implement interventions focusing on healthy lifestyle as they offer an efficient structure to reach a large group of people through natural social networks [1, 2]. However, recent reviews have indicated that the effectiveness of worksite health promotion programs (WHPP) on improving employees lifestyle is modest and varies widely across studies [3-5]. One of the reasons thereof could be that WHPPs often report low participation levels, whereas high participation levels are an essential component for program success [6-8]. Low participation levels result in decreased effectiveness as well as decreased cost-effectiveness of the program and hampers the external validity [9]. A review of Bull and colleagues (2003) showed that the range of participation levels in WHPP varied from 8% to 97% across studies, with a median of 61% [6]. Additionally, a recent review by Robroek and colleagues (2009) reported participation levels ranging from 10% to 64%, with a lower reported median of 33% (95%CI: 25 to 42%) [8]. They looked at whether participation in WHPPs, targeting physical activity and nutrition, was influenced by individual, health- and work-related characteristics of the target population or by program aspects. The review showed that participation in WHPPs is in general higher among female employees and for programs that offer incentives, programs consisting of multi-component interventions and programs that target multiple lifestyle behaviors [8]. Additionally, it is assumed that if employees are more satisfied with the overall program they would possible be more inclined to participate [10]. However, evidence was lacking for a considerable number of other factors related to program implementation [8]. Although this might suggest limited importance of these factors, the absence of evidence could possibly be explained by lack of reporting determinants for non-participation of employees in more than 80% of the studies [8]. Another explanation could be related to design issues as most included studies in that review were ad-hoc and cross-sectional and until now no quantitative information about the strength of the associations between implementation factors and participation levels are presented [11, 12]. In light of the review of Robroek and colleagues (2009) we decided to prospectively look at implementation factors that either hinder or facilitate employee participation in the BRAVO@Work [13]. Therefore, the aim of the present study is to gain insight into which implementation factors are associated with program participation and satisfaction.

113

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

Methods This study was carried out as part of the formative evaluation of the BRAVO@Work project with a quasi-experimental design on the side to gain insight into the behavioral effects of the program. The study protocol was approved by the Medical Ethics Committee of the University Medical Centre of Utrecht. Details of the study have been published elsewhere [13]. Study participants and setting The study setting involved an Academic Hospital and a University of Applied Sciences (hereinafter “the hospital” and “the university”) in Utrecht (the Netherlands). In each setting an intervention department was recruited: a department of gynaecology at the Hospital with 662 employees and the health faculty of the University with 546 employees – in 2009-2010 through the personal network of the researchers. The upper management of the intervention groups signed a letter of intent stating that they were willing to participate in the study and that they agreed to the financial and organisational consequences of participating in BRAVO@Work. All employees were eligible to participate in the study, but employees aged 18 years or older, with a contract for at least ten months at the start of the intervention and who responded to both the T1 and T2 questionnaire were eligible for inclusion in the analysis. Detailed information about the intervention and control sites have been published elsewhere [14]. The BRAVO@Work study In the BRAVO@Work study, a 7-step implementation strategy was used by both intervention departments to develop, implement and maintain interventions targeting multiple lifestyle behaviours. The 7-step strategy consists of the following steps: 1) creating solid support, 2) formation of a project structure, 3) performing a needs assessment, 4) development of interventions, 5) implementation of interventions, 6) evaluation, and 7) maintenance. Detailed information about the use and interpretation of the specific steps of the strategy can be found elsewhere [13]. The principal feature of this study was that the intervention departments were primarily responsible for the development, implementation and continuation of the interventions; the researchers acted as embedded observers throughout the study. An external advisor from the Netherlands Institute for Sport & Physical Activity informed the project groups of both intervention departments about the 7-step strategy

114

Relationship between barriers and facilitators and implementation

during the first project meetings and provided a list of environmental and individual evidence-based interventions of varying complexity [4, 15-18]. The external advisor was present during most project meetings to answer questions and to provide guidance, but he was briefed to strongly refrain from taking a leading/coordinative role at any time during the project. In short the 7-step strategy states that the upper management of both participating organisations formed their own steering committee and appointed a project leader. The project leader then established a project group that consisted of managers and employees from different teams within the intervention department (about 6 till 10 people per worksite). The project members conducted a needs assessment among all employees in their intervention department, resulting in a list of possible lifestyle interventions fitting the department and the needs of employees. This list was combined with the list of evidencebased interventions referred to above. Based on this list, the project groups choose, on a consensus basis, the most appropriate and feasible interventions matching employee needs, which they then developed further. Eventually, both companies implemented a mix of environmental, educational and individual interventions. Employees were recruited for participation by means of postcards, a kick-off meeting, informative emails, project website, personal communication and posters. At the hospital 12 interventions and at the university 21 interventions were implemented during the 12 months implementation period. Box 1 gives an overview of the interventions that were implemented in both companies. Detailed information about the implementation of these interventions, intervention content and maintenance of the interventions has been published elsewhere [14].

115

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

Box 1 – Overview of interventions implemented at both companies The hospital - Health checks [IBP] - Educational meetings for doctors (clinical lesson) [EDU] - Workshop for managers about discussing vitality in employees yearly performance Interviews [EDU] - Team meetings with an educational component [EDU] - Poster on nutrition [EDU] - Recipe cards [EDU] - Free fruit [ENV] - Christmas event with a focus on healthy living [IBP] - Poster on physical activity and relaxation [EDU] - Organized lunch walks [IBP] - Mindfulness sessions [IBP] - Peer group counseling [IBP] The university - Annual opening ceremony of the academic year in theme of the project [IBP] - Health week [IBP] - Meeting for managers about discussing vitality in employees yearly performance Interviews [EDU] - Team meetings with an educational component [EDU] - Free fruit [ENV] - - - - - - - - - - - - - - - -

Promotion of ‘Superfoods’ in the organizations’ canteen [EDU] Flyer on exemplary behavior [EDU] Analysis of foods sold in the canteen [ENV] Stricter smoking policy [ENV] Bicycle check [IBP] Enlarged bicycle shed [ENV] Improving the attractiveness of the staircases [ENV] New building routing [ENV] Sports material on loan [IBP] Route maps for lunch walks [IBP] Point of decision prompts to stimulate stair use [ENV] Sitting balls [ENV] Standing tables [ENV] Coaching trajectory [IBP] Chair massages [IBP] Books on time management [EDU]

Timing Month 1 and 14 Month 8 Month 8 Month 14 Month 3 to 5 Month 3 to 5 Month 3 to 5 Month 8 Month 12 to 14 Month 12 to 14 Month 14 to 15 Month 14 to 15 Month 4 Month 4 Month 7 Month 11 to 13 Month 6 to 7 and Month 11 to 13 Month 11 to 12 Month 8 Month 8 Month 6 Month 6 to 7 Month 7 Month 9 Month 9 Month 8 Month 11 Month 12 to 16 Month 16 Month 15 Month 16 Month 11 to 16 Month 11 to 12

Abbreviations: ENV = environmental intervention, EDU = educational intervention, IBP = individuallybased interventions that required committed and active participation.

Data collection Data for this study were collected by means of a web-based questionnaire distributed among all employees at the hospital 9-months (T1=December; n=622) and 15-months (T2=June; n=613) after start of the implementation of the program. At the university the questionnaire was distributed 6-months (T1=January; n=504) and 12-months (T2=July; n=489) after start of the implementation of the program. Survey questions to measure implementation

116

Relationship between barriers and facilitators and implementation

determinants were developed based on a questionnaire called ‘MIDI’ developed by Fleuren and colleagues (2014) [19]. Items were chosen based on their conceptual framework and on Wierenga and colleagues 2012 [12, 13]. Although MIDI is promising, it is not yet a validated instrument [19]. However, were are not aware of any validated instruments for measuring implementation determinants and outcomes. Eventually, based on both conceptual frameworks, 23 survey questions were developed and measured at T1 (representing baseline implementation factors) and T2 (representing follow-up implementation factors). Participants were asked to rate these statements using a 5-point Likert scale (range 1 to 5; totally disagree to totally agree). To group these statements into meaningful categories, we applied factor analysis with Varimax rotation. Combined with our conceptual knowledge we identified thirteen constructs which were found to be sufficiently reliable (Table 1). In both questionnaires employees participation to the specific intervention was assessed for each specific intervention by single ‘yes’ or ‘no’ questions (“could you indicate in which interventions of the following list you have participated? [followed by a list of all implemented interventions]”). Employees satisfaction with the overall project was measured at T1 and T2 using a single question (“were you satisfied with the project?”) on a 10-point scale (range 1 to 10; very dissatisfied to very satisfied) at follow-up. Additionally, semi-structured interviews were conducted at the same time points which are described in more detail in another article [14].

117

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

Table 1. Overview of the 13 measured implementation factors in the questionnaire and their corresponding construct Implementation factors A. Programs’ fit with the organisation

Number of Chronbach’s Items in construct items alpha 2 α= .90 - “BRAVO fits the culture of this organisation” - “BRAVO fits the organisation?”

B. Positive attitude towards program implementation

α = .91

C. Perceived social support of colleagues and supervisor

α= .0.79

- “I find the introduction of BRAVO at the organization highly desirable” - “I find the introduction of BRAVO at the organization good” - “My immediate colleagues support me to participate in BRAVO” - “My supervisor supports me to participate in BRAVO” - “Most of my immediate colleagues are positive about BRAVO” - “My supervisor is positive about BRAVO” - “Most of my immediate colleagues participate in one or more BRAVOinterventions” - “Due to the high pressure of my job I am struggling to participate in the BRAVOinterventions*” - “BRAVO is applicable within my job” - “BRAVO fits well within my daily work tasks” - “BRAVO fits well with my working hours”

D. Perceived social norm among 3 colleagues and supervisor

α = .71

E. Programs’ fit with work demands

α = .73

F. Perceived level of knowledge 2 of lifestyle and health promotion interventions

α = .77

- “I have sufficient knowledge about lifestyle and health promotion interventions” - “My immediate colleagues have sufficient knowledge about lifestyle and health promotion interventions”

G. Programs’ fits employee needs H. Positive program image

N/A

- “BRAVO fits employee needs”

N/A

I. Positive attitude towards employer involvement J. Positive attitude towards program maintenance

N/A

K. Colleagues program 1 awareness L. Program notification to new 1 employees M. Program participation is not 1 perceived as obligatory

N/A

- “I find that BRAVO has a positive image within the organisation” - “I think it is good that my employer thinks about my health” - “When BRAVO is successful, I find that the implemented interventions need to be carried out within the whole organisation in the future” - “Most of my immediate colleagues are aware of BRAVO” - “New employees are actively informed about BRAVO” - “I perceive participation in BRAVO as obligatory”*

N/A N/A

* statement is recoded for all analysis; N/A = not applicable

118

Relationship between barriers and facilitators and implementation

Data analysis Descriptive statistics (i.e., percentage, mean and standard deviation) were used to assess respondentsâ&#x20AC;&#x2122; demographic characteristics at baseline (T1). Overall program participation was computed by adding up the number of each interventions in which an employee participated during the entire study period. Overall program participation was further defined as low versus high participation. Differentiation was based on the median split (median=0.200; n=188). Pearsonâ&#x20AC;&#x2122;s correlations between scores on baseline implementation factors were assessed to determine the degree of coherence between factors. Population effect size was considered small (.10), medium (.30) or large (.50) based on Cohen 1992 [20]. A paired samples t test was used to assess whether scores on implementation factors differed over time. To assess the relationship between the measured implementation factors at baseline (T1) and follow-up (T2) and overall program participation (dichotomous outcome; low participation versus high participation) and satisfaction (continuous outcome) both measured at T2, logistic and linear regression analyses were performed with the variable of interest as outcome and the implementation factors as independent variables. Low participation was defined as participation in â&#x2030;¤ 2 interventions and high participation as participation in >2 interventions. In the first step of the analysis, univariate associations between the implementation factors with program participation and satisfaction were established. Second, multivariate analyses were performed for all variables in the univariate analyses with p<0.05. Additionally, in order to assess the relationship between employees satisfaction at T1 and T2 with overall program participation, logistic regression analyses were performed with overall program participation as the outcome and satisfaction as the independent variable. All analyses included only respondents for whom data were present from both process questionnaires. We controlled for the covariates company and contract hours in all logistic regression analysis and only for company in all linear regression analysis. No differences were found for both worksites separately so only aggregated data are presented in order to maintain power. Statistical significance was defined as p<0.05. All analyses were performed using the Statistical Package of Social Sciences version 20.0 for Windows (SPSS Inc. Chicago, Illinois, USA).

119

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

Results Participants, descriptive statistics and correlations In total, 217 employees of both companies (38.3%) responded to both questionnaires. The majority of the sample was female (84.9%), the mean age was 44.9 (SD 11.5) years with an average workweek of 28.4 (SD 8.6) hours distributed over on average 3.8 (SD 1.1) days a week. Means, SDs and correlations for the measured implementation factors are provided in table 2. In general, the means of the implementation factors were above the midpoint of the scale (above 3.0) and also changed positively over time, which suggested generally favourable evaluations with regard to the program. Correlations between implementation factors were in general medium to strong (Pearsonâ&#x20AC;&#x2122;s correlations ranging from 0.30 to 0.73) and most implementation factors were positively correlated to each other. As expected â&#x20AC;&#x2DC;perceived level of knowledgeâ&#x20AC;&#x2122; was in general negatively correlated to the other implementation factors. Indicating that employees who believe they have a high level of knowledge about a healthy lifestyle and health promotion activities, were less inclined to believe that the implemented program, for example, fits the organization and needs of employees. On average, hospital employees (n=94) participated in 2.2 (SD 1.6; range 0-8) out of ten interventions in which they could have participated. For university employees (n=94) this was 3.0 (SD 1.7; range 0-9) out of eleven interventions. Of all employees of both organisations (n=217), the large majority (75.1%) participated in at least one intervention during the study period. Employees of both organisations who participated in at least one intervention during the study period and responded to both questionnaires (n=164) rated the overall project as moderate at both T1 (6.7; SD 1.6) and T2 (6.8; SD 1.1), with no significant change over time (p>0.05). More detailed information about the participation and satisfaction levels for each intervention have been reported elsewhere [14].

120

(A)

(B)

.11

3.6 (0.9) 3.7 (0.8) 3.4 (0.8) 3.6 (0.8)** .03 3.0 (0.7) 2.9 (0.6) .12 3.6 (1.0) 3.7 (0.9)*

J. Positive attitude towards program maintenance

K. Colleagues program awareness L. Program notification to new employees

M. Program participation is not perceived as obligatory

* p < .05; ** p < .01; *** p < .001; N/A: Not applicable

.25*** .31*** .23**

G. Programsâ&#x20AC;&#x2122; fits employee needs 2.9 (0.8) 3.1(0.8)*** .56*** .66*** H. Positive program image 3.1 (0.9) 3.2 (0.9)** .57*** .69*** I. Positive attitude towards employer 3.8 (0.8) 3.9 (0.8) .35*** .42*** involvement

.22**

.00 .10

.50*** .59***

.43*** N/A

2.9 (0.7) 3.0 (0.7)** .25*** .29*** 3.9 (0.7) N/A N/A N/A

E. Programsâ&#x20AC;&#x2122; fit with work demands F. Perceived level of knowledge of lifestyle and health promotion interventions

.44**

.50**

.25** .22**

(E)

.53** .20** -.23**

-.20** .46** -.09 -

.17*

.13*

.45*** .22** .28*** .18**

.19**

.15* .02

.25**

.11

.43** .40**

(I)

0.12

N/A

.28** -.06 -.03 -.01

-0.01 -

-.08 -.06

.21** .03 .13 0.14* -

.17*

0.21** -0.04

0.01 0.03

0.10 0.13*

0.10

.13* .10

.55** .10 .22** .10 .43** .21** .25** -.05 .46** .27** .21** .14*

.12 -.15*

.24** .49** .29** -.12

N/A N/A

0.12

(M)

.28** -.04 .25** .06

(L)

.23** .33** .26** -.25**

0.41*** 0.41*** 0.37*** 0.18** 0.17** 0.28*** -0.05

(K)

.58** .04 .61** .02

(J)

N/A

.61** .40** 0.65*** .26** 0.34*** 0.27*** -

.34** -.30**

.30**

.38**

-.10

(H) .60** .71**

(G)

-.18* .59** -.30** .64**

(F)

.44*** .31*** N/A .54*** .27*** N/A .22** .20** N/A

.40*** N/A N/A

.67**

.37** .42**

(D)

-0.20** -.11

.28*** .23**

.57***

3.0 (0.7) 3.2 (0.6)** .44*** .38***

D. Perceived social norm among colleagues and supervisor

.22**

2.6 (0.8) 2.8 (0.8)** .12

.30** .29**

(C)

C. Perceived social support of colleagues and supervisor

A. Programsâ&#x20AC;&#x2122; fit with the organisation 3.5 (0.8) 3.7(0.7)*** .73** B. Positive attitude towards program 3.2 (1.0) 3.5 (0.9)*** .62*** implementation

Mean T1 Mean T2 (SD) (SD)

Table 2. Means and correlations for baseline implementation factors (n=164; above diagonal) and follow-up implementation factors (n=184; below diagonal)

Relationship between barriers and facilitators and implementation

121

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

Association between baseline implementation factors and overall employee participation and satisfaction The univariate regression analyses between baseline implementation factors (T1) and selfreported participation rates at follow-up (T2) showed that employees were more likely to participate in more interventions when they perceive that the program fits the organisation, they have a positive attitude towards program implementation and maintenance, they perceive social support and a positive social norm, when the programs fits the needs of employees’ and when the program is perceived to have a positive image (table 3). The results of the multivariate regression analysis in which the seven significantly associated constructs mentioned above were included, showed that the most important correlate with participation was a positive attitude towards program implementation (OR 2.06). None of the other factors contributed towards explaining additional variance in participation rates. The univariate regression analyses showed that eleven baseline implementation factors were positively associated with overall satisfaction (see table 3 for details). Additionally, ‘perceived level of knowledge of lifestyle and health promotion interventions’ was negatively associated with overall satisfaction. Only ‘program participation is not perceived as obligatory’ was not associated with satisfaction (table 3). The multivariate regression analyses showed that the most important correlates with higher satisfaction levels were ‘program matches employee’s needs’ (β: 0.22), ‘positive attitude towards program implementation’ (β: 0.42), and ‘positive attitude towards employer involvement’ (β: 0. 20). None of the other factors contributed to explaining additional variance in satisfaction rates.

122

Relationship between barriers and facilitators and implementation

Table 3. Univariate association between implementation factors and overall program participation (low vs. high) and satisfaction Overall program participation Overall program (T2) [OR (95%CI)]a satisfaction (T2) [β]b Implementation factors T1 (n=173) T2 (n=185) T1 (n=168) T2 (n=188) A. Programs’ fit with the organisation 1.60 1.32 0.52*** 0.63*** (1.05-2.42)* (0.84-2.08) B. Positive attitude towards program 1.94 1.70 0.64*** 0.70*** implementation (1.33-2.84)** (1.15-2.50)** C. Perceived social support of colleagues 1.90 2.08 0.28*** 0.31*** and supervisor (1.26-2.88)** (1.37-3.17)** D. Perceived social norm among 2.11 2.25 0.35*** 0.39*** colleagues and supervisor (1.25-3.58)** (1.36-3.75)** E. Programs’ fit with work demands 1.54 2.74 0.29*** 0.33*** (0.96-2.47) (1.68-4.46)*** F. Perceived level of knowledge of lifestyle 0.77 N/A -0.18* N/A and health promotion interventions (0.49-1.23) G. Programs’ fits employee needs 1.72 1.49 0.56*** 0.60*** (1.15-2.58)** (1.00-2.22)* H. Positive program image 1.50 1.33 0.51*** 0.66*** (1.06-2.14)* (0.94-1.88) I. Positive attitude towards employer 1.21 1.15 0.44*** 0.39*** involvement (0.83-1.77) (0.79-1.66) J. Positive attitude towards program 1.66 1.56 0.42*** 0.40*** maintenance (1.13-2.45)* (1.03-2.37)* K. Colleagues program awareness 1.21 1.54 0.22** 0.11 (0.82-1.78) (0.99-2.38) L. Program notification to new employees 0.98 1.36 0.20** 0.24** (0.62-1.55) (0.82-2.26) M. Program participation is not perceived 1.15 0.96 0.04 0.18* as obligatory (0.84-1.58) (0.66-1.38) * p<0.05, ** p<0.01, *** p<0.001: Determinant concept is significantly associated with outcome measure; a corrected for company and contract hours; b corrected for company; N/A: Not applicable as this factor was only measured at T1.

Association between follow-up implementation factors and overall employee participation and satisfaction The univariate regression analyses between follow-up implementation factors and overall participation generally showed similar results as between the baseline implementation factors and overall participation. Differences were related to the ‘programs fit with the organisation’ and a ‘positive program image’, which were no longer associated with participation. The ‘programs fit with employees’ work demands’, however, now was associated with participation. For the standard multivariate regression analysis no

123

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

significant associations were found. An additional forward selection procedure showed that a ‘positive attitude towards program implementation’ (OR = 2.34) best correlated with overall employee participation. The univariate regression analyses between follow-up implementation factors and overall satisfactions generally showed similar results as between the baseline implementation factors and overall satisfaction. Differences were related to ‘colleagues program awareness’ which was no longer associated with satisfaction. Program participation is not perceived as obligatory’, however, now was associated with satisfaction. The results of the multivariate regression analyses in which all significantly associated concepts mentioned above were included, revealed that the most important correlate with higher satisfaction levels were ‘programs fit with the organization’ (β: 0.27), ‘positive attitude towards program implementation’ (β: 0.32), ‘positive program image’ (β: 0.24), and ‘program notification to new employees’ (β: 0.13). None of the other factors contributed to explaining additional variance in satisfaction rates. Association between employee satisfaction and participation Logistic regression analysis revealed no significant association between employee (n=171) satisfaction at T1 and program participation at T2 (OR: 1.18, 95% CI 0.89 to 1.58). Employees’ overall satisfaction levels at follow-up were positively associated with higher overall employee participation (OR: 1.154, 95% CI 1.07 to 2.22).

Discussion The aim of this study was to gain insight into which implementation factors are associated with program participation and satisfaction. High overall participation was obtained as 75.1% of the employees participated in at least one intervention. The strongest correlate with actual participation in the interventions was having a favorable attitude towards implementation of the project during the project. Other factors that were associated with overall participation were: a perceived fit of the program with the organization and the organizational culture; a perceived fit of the program with employee needs and work demands; a favorable program image; a positive attitude towards program implementation and maintenance; perceived social support among colleagues and supervisors; and a positive social norm.

124

Relationship between barriers and facilitators and implementation

With regard to program satisfaction, employees were moderately satisfied with the project as they rated the project on average with a 6.8 (range 1 to 10). Employees were more likely to be satisfied when they perceived the program fitted employee’s needs and the organizational culture; when they have a positive attitude towards program implementation and employer involvement; when the program has a positive image and when new employees are actively informed about the project. All other measured implementation factors were also correlated with satisfaction, although less strongly as they were not significantly related to satisfaction in the multivariate analyses. Previous implementation frameworks have stressed the importance of ensuring a fit of the program with the users, as well as with the social and organizational context in which an innovation is to be implemented [11, 12, 19, 21]. Our study findings confirm this assumption, showing that when a fit is perceived on all three domains this will most likely contribute to both program participation and satisfaction. Due to the cross-sectional nature of this study we cannot state actual causality. However, the qualitative data obtained during the course of the study supports this notion [14]. Interviews with employees were conducted at baseline and both follow-up moments. These interviews revealed that the programs’ compatibility with societal developments and current media attention regarding the importance of a healthy lifestyle was an important factor underlying their positive attitude towards program implementation and employer involvement. But, it is a major prerequisite that participation is not mandatory. The interviews also revealed that time constraints of employees due to high work demands or due to the fact that an intervention was offered at an inconvenient time or location hampered program participation. Whereas, when the intervention fitted employees specific needs and work agenda, stimulated social interaction with colleagues and was attractive, easily accessible and visible in the department they were more inclined to participate [14]. The implementation factors identified in this article are in line with the results of the reviews by Fleuren and colleagues (2004) and Wierenga and colleagues (2013) [11, 12]. However, both reviews identified over 50 implementation factors using mainly qualitative measurements measured when program implementation has ended [11, 12]. Whereas this study provides a more quantitative ‘evidence based’ overview of ‘buttons to push’ in order to enhance implementation success by enhancing employees participation and satisfaction rates throughout the project. Therefore, we advise that implementers should take the identified implementation factors into account during the development of the program but also continuously monitor these factors during implementation to guarantee high

125

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

participation and satisfaction levels. Explicitly addressing relevant implementation factors in the implementation plan ensures that the factors are continuously taken into account during program implementation. This could be done by using a newly proposed Dutch instrument by Fleuren and colleagues (2014) in the preparation phase of a project [19]. This instrument is intended for researchers who wish to map factors that affect actual use of guidelines related to the implementation of innovations in healthcare settings. As we focus on lifestyle interventions, it is only logical that our study identified additional implementation factors (i.e. favorable program image and a positive attitude towards employer involvement). It is therefore relevant to update the instrument with our found results in the recent study and with relevant implementation factors identified in our review [11]. Strengths and limitations The fact that the quantitative approach for identifying the correlation between implementation factors and employee participation and satisfaction in this study is new can be considered a strength of this study. It provides valuable information about which implementation factors influence program implementation and should be taken into account in future research as well as in implementation programs in daily practice. Another strength of this study is that we evaluated the implementation process on the basis of a conceptual framework that was developed prior to the start of the implementation [13]. This framework was based on several models for process evaluations and included paying specific attention to possible barriers and/or facilitators affecting the development and the implementation process [12, 13, 22-24]. As a result, data were collected continuously from the start of the implementation, which enabled us to measure the implementation factors during and after program implementation. As with most studies using questionnaire data, there is the possibility of selection bias because only a proportion of the total population of employees in each worksite completed the questionnaires. Especially since no data could be gathered about specifics of the non-response group. As a consequence it could be possible that only employees with a favourable opinion about the project and a higher level of participation had filled out the questionnaires. Different assessment time points for the two worksites were used (i.e., web-based questionnaire distributed to hospital employees at 9 months [T1=December] and 15 months [T2= June] after starting the program, yet distributed to university employees at 6 months [T1=January] and 12 months [T2=July] after starting the program). However, the researchers deliberately choose for this approach as at those points in time both worksites were at the same point in the implementation process which

126

Relationship between barriers and facilitators and implementation

was key for the analysis. Also, seasonality does not play a role because data was collected in the same year and only one month apart. A final limitation to this study was that we only measured employee satisfaction with one survey item. However, other recent process evaluations applied same kind of question when measuring employees overall satisfaction rates with an intervention [25-27]. Conclusion The present study provides an important contribution to the scarce availability of systematic evaluations of program implementation, and enhances knowledge on those factors that are associated with increased employee satisfaction and participation levels. Thereby giving insight to researchers and practitioners which â&#x20AC;&#x2DC; buttons they need to pushâ&#x20AC;&#x2122; to enhance successful program implementation and effectiveness. It is important that implementers pay attention to, and anticipate on the mentioned implementation factors before and during the development and implementation of a WHPP. It is especially vital that implementers make sure that employees have a favourable attitude towards program implementation because this factor is best correlated with participation and satisfaction. Additionally, employees with higher satisfaction rates are more likely to participate in the WHPP.

127

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

References 1.

Hunt MK, Stoddard AM, Barbeau E et al. Cancer prevention for working class, multiethnic populations through small businesses: the healthy directions study. Cancer Causes Control 2003;14(8):749-760.

Dishman RK, Oldenburg B, Oâ&#x20AC;&#x2122;Neal H et al. Worksite physical activity interventions. Am J Prev Med 1998;15(4):344-361.

Verweij LM, Coffeng J, van Mechelen W et al.Meta-analyses of workplace physical activity and dietary behaviour interventions on weight outcomes. Obesity Reviews 2011;12(6):406-429.

Groeneveld IF, Proper KI, Van Der Beek AJ, et al.Lifestyle-focused interventions at the workplace to reduce the risk of cardiovascular disease - A systematic review.Scandinavian Journal of Work, Environment and Health 2010;36(3):202-215.

Jepson RG, Harris FM, Platt S et al. The effectiveness of interventions to change six health behaviors: a review of reviews.BMC Public Health 2010;8(10):538.

Bull SS, Gillette C, Glasgow RE et al. Work site health promotion research: to what extent can we generalize the results and what is needed to translate research into practice?Am J Public Health 2003;30:537-549.

Glasgow RE, McCaul KD, Fisher KJ. Participation in worksite health promotion: a critique of the literature and recommendations for future practice. Health Educ Q 1993;20(3):391-408.

Robroek SJ, van Lenthe FJ, van Empelen P et al.Determinants of participation in worksite health promotion programmes: a systematic review. Int J Behav Nutr Phys Act 2009;6:26.

Dzewaltowski DA, Estabrooks PA, Klesges LM et al. Behavior change intervention research in community settings: how generalizable are the results?Health Promot Int 2004;19(2):235-245.

10. Suhre CJM, Jansen EPWA, Harskamp EG. Impact of degree program satisfaction on the persistence of college students.Higher education 2007;54:207-226. 11. Wierenga D, Engbers LH, van Empelen P et al. What is actually measured in process evaluations for worksite health promotion programs: a systematic review. BMC Public Health 2013;13(1):1190. 12. Fleuren M, Wiefferink K, Paulussen T. Determinants of innovation within health care organizations: literature review and Delphi study. Int J Qual Health Care 2004;16(2):107-123. 13. Wierenga D, Engbers LH, van Empelen P et al. The design of a real-time formative evaluation of the implementation process of lifestyle interventions at two worksites using a 7-step strategy (BRAVO@ Work).BMC Public Health 2012;12(1):619. 14. Wierenga D, Engbers LH, van Empelen P et al. The implementation of multiple lifestyle interventions in two organizations: a process evaluation.JOEM 2014;56(11):1195-1206. 15. Anderson LM, Quinn TA, Glanz K et al. The Effectiveness of Worksite Nutrition and Physical Activity Interventions for Controlling Employee Overweight and Obesity. A Systematic Review. Am J Prev Med 2009;37(4):340-357. 16. Ni Mhurchu C, Aston LM, Jebb SA. Effects of worksite health promotion interventions on employee diets: A systematic review.BMC Public Health 2010;10(10):62. 17. Jepson RG, Harris FM, Platt S et al. The effectiveness of interventions to change six health behaviours: A review of reviews. BMC Public Health 2010;8(10):538. 18. Chau JY, der Ploeg HP, van Uffelen JG et al. Are workplace interventions to reduce sitting effective? A systematic review. Prev Med 2010;51(5):352-356. 19. Fleuren M, Paulussen TH, van Dommelen P et al. Towards a Measurement Instrument for Determinants of Innovations (MIDI).International Journal for Quality in Health Care 2014;26(5):501-510. 20. Cohen J. A Power Primer. Psychological Bulletin 1992;112(1):155-159.

128

Relationship between barriers and facilitators and implementation

21. Durlak JA, DuPre EP. Implementation matters: a review of research on the influence of implementation on program outcomes and the factors affecting implementation. Am J Community Psychol 2008;41(34):327-350. 22. Paulussen TH, Wiefferink K, Mesters I. Invoering van effectief gebleken interventies.In Gezondheidsvoorlichting en gedragsverandering. Edited by Brug,J.,Van Asseman,P.& Lechner,L. Assen: van Gorcum; 2007. 23. Rogers EM. Diffusion of innovations: New York: Free Press; 2003. 24. Bartholomew LK, Parcel GS, Kok G et al.Planning Health Promotion Programs: An Intervention Mapping Approach. San Francisco, CA: Jossey-Bass; 2006. 25. Driessen MT, Proper KI, Anema JR et al. Process evaluation of a participatory ergonomics programme to prevent low back pain and neck pain among workers. Implement Sci 2010;5:65. 26. Verweij LM, Proper KI, Hulshof CT et al.Process evaluation of an occupational health guideline aimed at preventing weight gain among employees. J Occup Environ Med 2011;53(7):722-729. 27. Strijk JE, Proper KI, van der Beek AJ et al.A process evaluation of a worksite vitality intervention among ageing hospital workers.Int J Behav Nutr Phys Act 2011;8:58.

129

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

A 7-step strategy for the implementation of worksite lifestyle interventions: helpful or not?

Wierenga D, Engbers LH, van Empelen P, van Mechelen W. Accepted for publication in Journal of Occupational and Environmental Medicine 2016

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

Abstract Objective: To evaluate the use of and adherence toa 7-step strategy for the development, implementation and continuation of a comprehensive, multi-component lifestyle program. Methods: Strategy use and adherence was assessed with twelve performance indicators. Data was collected by combining onsite monitoring with semi-structured interviews at baseline and follow-up (6, 12 and 18 months). Results: Not all performance indicators were met so partial strategy adherence was obtained. The strategy could be improved on the following aspects: support among management, project structure, adaptation to needs of employees, planning and maintenance. Conclusions: The results of this evaluation indicate that strategy adherence facilitated structured development and implementation. Based on the qualitative data this study suggest that when improvements will be made on both the content and performance, the 7-step strategy could be an effective tool to successfully implement a multi-component WHPP.

132

The 7-step strategy: helpful or not?

Background An unhealthy lifestyle is associated with overweight and obesity, which in turn are associated with lower productivity and higher sickness absence [1, 2]. A worksite health promotion program(WHPP) may effectively contribute to favorable lifestyle change and may contribute to company cost cuts, but successful implementation of such a WHPP remains a challenge. However, despite evidence for the effectiveness of WHPP’s in positively changing employees’ lifestyle [3-8], such programs are still not common practice in Dutch organizations. Implementation strategies that address the concerns and barriers of all stakeholders on all organizational levels may increase the uptake and sustainability of WHPP, thereby promoting the actual impact that interventions may have [9-11]. Literature applies many different terms and definitions for ‘implementation research’. In the present study we adopted Curran’s (2012) definition of implementation: “an effort specifically designed to get best practice findings and related products into routine and sustained use through appropriate uptake/adoption of interventions” [12]. Our main objective was to study the use of a 7-step implementation strategy, which was proposed by Wynne and Clarkin (1992).The 7-step strategy is developed based on the results of a survey among 1500 European companies across seven countries questioning the company’s health policies, worksite health promotion activities and case studies of good practices [13, 14]. The 7-step strategy provided means to systematically develop, implement and maintain health promotion programs at the workplace by addressing each of the following elements: 1) creating solid support, 2) formation of a project structure, 3) performing a needs assessment, 4) development of interventions, 5) implementation of interventions, 6) evaluation, and 7) maintenance. The main focal point of this strategy is that it employs a ‘user-driven’ approach towards developing and implementing worksite health interventions. User driven means that employees and managers from different organizational levels are actively involved in the development and implementation process. By as project members going through the steps, ownership is stimulated and a fit with existing capacity, needs and values is ensured [15]. It is hypothesized that application of the 7-step strategy should ensure that developed and implemented interventions will be tailored to the needs of different stakeholders within a worksite. Thereby enhancing implementation success and ensuring the integration of implemented interventions in the organizations general occupational health policy. Currently the strategy is supported by the European Foundation for the Improvement of

133

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

living and Working Conditions [13]. However, to our knowledge it has never been evaluated whether such a strategy actually facilitates implementation. Besides this gap in knowledge, studies evaluating WHPP mainly address the programs’ effectiveness, generally neglecting a systematic effort to gain insight into factors that either hinder or facilitate the implementation process. Therefore, the aim of this process evaluation is to describe the use of and adherence to the 7-step strategy and to assess whether this was associated with the development, implementation and continuation of a comprehensive, multi-component lifestyle program (i.e. physical activity, smoking, alcohol use, nutrition and relaxation).

Methods The process evaluation of the use of and adherence to the 7-step implementation strategy was carried out as part of the BRAVO@Work project, in which by means of a quasiexperimental design also the behavioral effects of the program were examined. The Medical Ethics Committee of the University Medical Centre of Utrecht approved the study protocol. Details of the study have been published elsewhere [16]. Study participants and setting The implementation strategy was evaluated within the context of two different settings: an Academic Hospital and a University of Applied Sciences (hereinafter “the hospital” and “the university”) in Utrecht (the Netherlands). Both worksites were recruited in 2009-2010through the personal network of the researchers. The upper management of the organizations signed a letter of intent stating that they were willing to participate in the study and that they agreed to the financial and organisational consequences of participating in BRAVO@Work. The population under study in this paper are the project members who were responsible for using and adhering to the 7-step strategy and thus for implementing the program in their respective organizations. Data was collected prior to implementation and 6, 12 and 18 months after implementation of the interventions. Interventions were implemented during one year. In addition, a researcher on site (referred to as the “embedded” researcher) continuously monitored the natural course of implementation of the BRAVO interventions, but without being actively involved in the development and implementation of the interventions.

134

The 7-step strategy: helpful or not?

The BRAVO@Work study In the BRAVO@Work study both intervention departments used a 7-step implementation strategy to develop, implement and maintain interventions targeting multiple lifestyle behaviours. As mentioned earlier this strategy is based on a study in 1992 by Wynne and Clarkin [13] in which they determined five important aspects for successfully implementing a health policy at the workplace: A) Needs assessment: for the establishment of a health policy it is important that the wishes and needs of employees are analyzed. In this way, the intervention activities can be developed according to their needs; B) Participation: key figures from different levels in the companyâ&#x20AC;&#x2122;s organization need to be involved in the development and implementation of the health promotion program to create a solid support for the health policy. This can be done by means of working groups; C) Flexibility: health promotion programs are similar at some basic points. However, they are not standard programs, since a WHPP needs to fit the specifics of the workplace; D) Integration: the health promotion program needs to include activities that are both aimed at the individual employee and at the work environment; and E) Multidisciplinary: several experts in the fields of human resources, communication, health management, psychology and working environment need to be involved in the development and implementation to increase program effectiveness. Wynne and Clarkin translated these five aspects into a generic 7-step implementation strategy for the systematic development and implementation of health promotion programs at the workplace [13, 14]. The implementation strategy consists of the following 7 steps: 1) creating solid support, 2) formation of basic structures, 3) performing a needs assessment, 4) development of the interventions and health policy, 5) implementation of the interventions, 6) evaluation of the implemented interventions, and 7) embedding the interventions in the general occupational health policy of the organization (table 1) [16]. The 7-step strategy was used to develop and implement interventions related to multiple lifestyle behaviors (i.e. BRAVO-interventions) at the worksite and to integrate these lifestyle interventions in the companyâ&#x20AC;&#x2122;s general health policy. The main aspect of this strategy is the active participation of relevant stakeholders when passing through the 7 steps. So, the intervention departments had the sole responsibility for the development, implementation and continuation of the interventions and the researchers acted solely as embedded observers throughout the study. To ensure that the companies used the 7-step strategy, an external advisor from the Dutch Institute for Sport and Physical Activity (www.nisb.nl) was appointed. He informed the project groups of both organizations about the content and purpose of the strategy during the first project meetings. Additionally, all members received

135

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

the research proposal outlining the strategy. The external advisor was present during all project meetings to answer questions and to provide guidance, but he was specifically briefed to refrain from taking the lead at any time during the project. In short, the 7-step strategy states that the upper management of both participating organisations was required to form their own steering committee and to appoint a project leader. The project leader was then required to establish a project group and was advised to include the following relevant stakeholders: managers and employees from different teams, a communications officer, a human resources officer, a facility manager and a company physician. Managers and employees were eligible for project membership if they were working at the intervention department. With guidance from the external advisor, the project members needed to conduct a needs assessment among all employees in their intervention department. Based on the outcome of the needs assessment and a list of evidence-based interventions, the project groups needed to select interventions fitting employeesâ&#x20AC;&#x2122; and organizational needs. The project leader then needed to draw up a project plan with the aim to facilitate implementation by means of describing the desired changes and project goals, the interventions, a timeline, a budget plan and persons involved, including their tasks and responsibilities. After approval had been given by the steering committees, the project members were required to implement the interventions during the following 12 months. After the intervention year, the project groups needed to evaluate the project. This evaluation was the basis for the go/no-go decision about whether or not to embed the programme or specific interventions into the organizations general health policy. The process evaluation Following our own conceptual framework for process evaluations, we focussed on assessing fidelity to gain insight in the use and adherence to each step of the 7-step implementation strategy [16-19]. The complete theoretical framework has been described in more detail elsewhere [16]. Fidelity refers to the extent to which the 7-step strategy was used according to protocol. Strategy adherence by project members was assessed from the onsite monitoring and semistructured interviews that included twelve performance indicators (PIs) found relevant by the research team (Table 1). PIs can be used to assess whether the most important recommendations of the 7-step strategy were carried out by the project members. Each PI that was met received a score 1, corresponding with strategy adherence for that PI. Each PI that was partly met received a score between 0 and 1, corresponding with partial strategy adherence in which not all relevant aspects of that PI were met. Each PI that was not met 136

The 7-step strategy: helpful or not?

received a score 0, reflecting no strategy adherence. An overall performance indicator for the whole strategy was calculated based on the average of all individual PIs, for which a higher overall performance score corresponded with higher strategy adherence [20]. The principal investigator (DW) was responsible for the initial scoring of each PI. One meeting was organized between researchers (DW, LE and PE) to discuss interpretation of the data to enhance the validity of the interpretation of the findings. In all cases consensus was reached by discussion. Data collection Data on fidelity were obtained among project members from both worksites by semistructured interviews and by onsite monitoring. The semi-structured interviews with all project members were held at baseline (T0; hospital n=8; university n=8), at T1 (hospital n=6; university n=11) and at T2 (hospital n=3; university n=7). The interviews were designed to gain more in-depth insight into the use of and adherence to the 7-step strategy and to address the PIs. At 18-months follow-up (T3), the project leaders of both organisations were contacted for a telephone interview to assess programme maintenance. All project members were invited through email for participation in the interviews. The principal investigator (DW) conducted all interviews during a telephone interview or face-to-face meeting at a time and location convenient for the participants. Prior to the start of the interview, all participants were informed about the purpose of the study and were reassured about confidentiality. All participants granted oral permission for recording the interview. All interviews were recorded and transcribed verbatim. Interviews with project members lasted 28 minutes on average (range: 10 – 56 minutes). An interview protocol was used including questions and prompts to guide the interview. The first open-ended question in interviews with project members was “How did you experience the implementation of the project so far?” Possible follow-up questions then included “What are in your opinion, positive points within the project?”, “What are points of improvement?”, and “How did you experience your participation in the project group so far?”. Field notes were written during the interviews regarding issues that could be relevant at the analytical stage. Throughout the study period, an embedded researcher (DW) continuously monitored the implementation process by documenting relevant e-mail communications, minutes of project meetings and observations in pre-defined spread sheets (i.e. onsite monitoring). These sheets were developed before onset of the study and were based on the PIs and conceptual framework [16]. 137

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

Data analysis Data from the semi-structured interviews were analyzed using the constant comparative method, in which each item is checked or compared with the rest of the data to inductively establish analytical categories [21, 22]. First, transcripts and field notes were read to get a general understanding of the concepts under study and to get some insight into the dynamics of the interviews. Using MAXQDA version 11 (VERBI GmbH, Berlin, Germany), transcripts were then open-coded by the primary researcher (DW). That is, transcripts were read line by line and relevant passages were marked with a series of codes from text and from literature about the specific topic, with the goal to describe the content of the interviews. Interview codes included both “descriptive” (i.e. within the immediate domain of the interview questions) and “analytic” (i.e. emerging and overarching themes) codes. Throughout the coding process, continuous efforts were made to detect further examples of previously identified codes/themes and, if applicable, to identify new ones [21-25]. Second, all codes were grouped into central concepts underlying the descriptive and analytical codes, thereby making them more workable. These concepts were then categorized into themes in order to identify patterns and relationships between concepts. These themes were partly identified in advance based on literature [9, 10], but were also derived from the data. Various meetings between researchers (DW, LE and PE) were held over the course of data analysis in which identified codes, identified concepts, identified themes, and interpretations of the data were checked and discussed to enhance the robustness of the findings. In all cases consensus was reached through discussion.

Results Strategy adherence was assessed with twelve performance indicators (PIs) that reflect the use of the 7-step strategy. The PIs and their corresponding scores are presented in table 1. Based on the study by Wynne and Clarkin [13, 14] it was hypothesized that each PI would contribute to facilitation of program implementation. The overall performance indicator for the University was 0.76 and for the Hospital 0.54 both representing partial strategy adherence. Overall, lowest scores were found for creating support among middle and lower management (PI 2), formation of project structure (PI 3, PI 4 and PI 5), performing a needs assessment (PI 6), develop a project plan (PI 7), and maintenance (PI 11 and PI 12) (see table 1).

138

The 7-step strategy: helpful or not?

Table 1. Performance indicators of strategy adherence, their description and performance score Performance indicator

Required action to meet the performance indicator

Step 1: creating solid support 1. Ensured upper management Support from upper management was/ ensured by support signing a letter of intent and reserving budget. Support from middle and lower management was 2. Ensured middle and lower ensured at start of the project. management support Step 2: formation of project structure 3. Formation steering committee

A steering committee was installed with a chairman, preferably someone with decisionmaking authority. The steering committee needs to make the go no-go decisions, but is not involved in substantive discussions. A project leader was appointed by the chairman. 4. Appointing a project leader A project group was formed that consisted of employees from different layers of 5. Formation project group and the organization (managers, employees, option working group communications officer, human resource staff member, facility management, health and safety executive, company physician) and optional a working group with only employees. These groups were responsible for the substantive development and implementation of the program. Step 3: needs assessment 6. Perform a needs assessment A needs assessment was performed by the project among employees group in order to map the wishes and needs of employees and managers, by using either a webbased questionnaire, physical examination, focus group sessions or a combination. Step 4: development of interventions 7. Develop a detailed project plan

8. Official decision on which interventions will be implemented

Based on the need assessment and characteristics of the organization, the project leader made a project plan with information about the following aspects: - desired changes - project goals - intervention template - timeline and communication plan - budget plan - list of responsible persons The project plan was presented to the steering committee who decided on the go-no-go criteria of the plan.

Performance Performance score score University Hospital 1

0.5

6 0.5

0.5

0.8

0.3

0 1 1 1 1 1 1

1 1 0 0 0 0 1

139

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

Performance indicator

Step 5: implementation 9. Implementation of the interventions Step 6: evaluation 10. Integrative evaluation

Step 7: maintenance 11. Determine which interventions would be continued 12. Embed the interventions in the organizations general health policy

Required action to meet the performance indicator

All developed interventions have been implemented within the organizations.

Performance Performance score score University Hospital 0.84

0.85

During the project an integrative evaluation was part of the process so adjustments to the project plan could be made if necessary.

The steering committee needed to determine which interventions would be continued and become part of the organizations general health policy. Efforts were made by the steering committee to integrate the implemented interventions in the organizations general health policy.

0.5

0.76

0.54

Total Performance Score

* interpretation of performance indicators: score 0 = no strategy adherence, score between 0 and 1= partial strategy adherence; score 1 = full strategy adherence.

The interviews as well as the data gathered via monitoring showed that the presence of an external advisor facilitated the use of the 7-step strategy, and hence improved the strategy adherence, as evaluated with the performance indicators. As an interviewee stated: â&#x20AC;&#x153;The external advisor from the NISB has added, in my opinion, important information and support. He was the expert. Thatâ&#x20AC;&#x2122;s needed in this organization because otherwise it would not be a priority and not be on the agenda anymore. Meaning that these project group meetings would not happen and subsequently the entire project. It [the presence of the external advisor] was really an extra reassurance, which was needed in light of all other projects in the organizations. Otherwise it would have collapsedâ&#x20AC;?. [University_P10 T1] Data from onsite monitoring and interviews showed that support from middle and lower management (i.e. PI2) could be improved. Both organizations scored a zero on this performance indicator, as middle and lower management were only actively involved in the last four months of the project. Interviewees stated that this hampered program implementation as this resulted in little support for the program from the management at the beginning:

140

The 7-step strategy: helpful or not?

“Well the support from managers is crucial. When you notice that the management is not supporting the program and do not facilitate participation in the program, then it [the program] will not lead to anything. But the fact that management just says ‘we need to do this’ is not enough. That does not ensure that the project becomes part of everyday culture among employees”. [University_P11 T2] The results from the interviews with project members at the end of the program (T2; 12 months after initial start of the implementation) showed that, following the advice of the external consultant, the team meetings held at the end of the project resulted in more support for the program, both among middle and lower management and among employees, as is illustrated by the following quote: “You see that there is a kind of dynamic that occurs when the external advisor and our project leader give a workshop about the program in regular team meetings. As discussed in the project meeting, I believe that when we did these meetings at the beginning of the project, it [the program] would have more support and become part of everyday culture in the organization. Also, actively involving lower management after

these meetings is essential and will lead to higher participation rates and thus program success”. [Hospital_P05 T2] With regard to PI 3 (i.e. installing a steering committee) differences could be observed between both organizations. The University met PI 3 (score=1) because a steering committee was formed with representatives of the upper management with decision-making authority regarding budget allocation facilitated implementation. We observed that each steering committee meeting was well prepared by the project leader with an agenda. The meetings were systematic and focused on making decisions without substantive discussions since these were already held in the project group. The onsite monitoring indicated that this facilitated effective program implementation. However, the University received a score of 0.5 for PI 3 as we observed that the steering committee turned into a project group in which all substantive discussions were held, but no one was responsible for making decisions on go, no-go aspects of the project plan and interventions that could be implemented. The lack of ‘authority’ regarding budget allocations seemed to lead to a considerably longer preparation phase (i.e. 15 months at the University compared to 6 months at the Hospital). Another factor that seemed to contribute to the delay of the initiation of the project at the

141

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

University was a high turnover among project leaders (i.e. PI 4) due to staff turnover and a lack of commitment and time. Hence, their score on PI 4 was 0.5. Observations from both worksites showed that assigning a project leader with 8 to 16 hours a week solely for the implementation of the project seemed to facilitate implementation. The Hospital did meet PI 4. However, onsite monitoring showed that at the beginning of the project, the project leader took sole responsibility, without actively involving project members. However, after a couple of months, she stopped being actively involved and tried to delegate this to the project members. However, the project members felt little responsibility and were not actively taking part in the development and implementation process. It appeared that this was the result of the lack of delegation from the project leader at the beginning of the project and overall hampered program implementation. Both organizations partially met PI 6 (i.e. performing a needs assessment), because observations showed that they only gained insight into the current lifestyle behaviors of their employees, using the baseline web-based questionnaire that was conducted as part of the effect evaluation. No information on the actual needs and wishes of employees was obtained. Instead, only the wishes and needs of project members were assessed during a meeting in which they could indicate their preferences on a list of evidence-based lifestyle interventions of varying complexity [4, 7, 26-28]. The onsite monitoring showed that differences were observed between both organizations regarding the development of a project plan (PI 7). At the University the project plan focused more on the details and included 5 out of the 6 main aspects (i.e., project goals, intervention template, timeline and communication plan, budget per intervention and a list of responsible persons), hence a score of 0.8 was given for PI 7. Interviewees regarded their project plan as facilitating for implementation, which is illustrated by the following quote: â&#x20AC;&#x153;A project plan is necessary; in particular to give direction about where to go and what to do. It [the project plan] can be a) guiding, and b) a tool to see where adjustments in the project were made. Itâ&#x20AC;&#x2122;s like a benchmark. Which does not mean that you necessarily have to follow the project plan from A to Z, but you do need to have a plan in order to be clear on what we want. ... Where you actually want to go are the great ideas and intention, but you need to have a concrete plan in order to actually change behavior. Without a concrete plan you keep having meeting after meeting without concrete actions. Like we had the first 15 months. But from the moment we wrote down concrete actions this gave guidance to the project and we were able to implement actual interventionsâ&#x20AC;?. [Hospital_P08 T2] 142

The 7-step strategy: helpful or not?

At the Hospital the detailed aspects of the project plan (i.e., budget plan, intervention template, timeline and list of responsible persons) were lacking and the plan focused mostly on desired changes and project goals, hence a score of 0.4 was given for PI 7. We observed that it appeared that the lack of translation of the project goals to concrete interventions led to ad hoc intervention development, which is illustrated by the following example: â&#x20AC;&#x153;the Hospital choose to address four lifestyle themeâ&#x20AC;&#x2122;s, one theme every three months. There was a lack of anticipation on the next theme. Planning of the next theme only started after the previous theme had finished, which resulted in gaps in between themes. Hence, no interventions were delivered between September and December (2011) and between January and April 2012. Finally, onsite monitoring and interviews suggested that improvements could be made regarding maintenance (PI 12). Both organizations only had partial strategy adherence (score=0.5). No long-term plans were made regarding follow-up of implemented interventions. Even though monitoring and interviews showed that both organizations decided to continue some of the interventions, no one was made responsible. At the Hospital, the project group choose to start a new pilot project in which managers would be trained in discussing sustainable productivity related to employeesâ&#x20AC;&#x2122; lifestyle in the yearly performance interviews. At the University only relatively simple interventions that did not require active employee participation (such as distribution of free fruit and posters) were continued for half a year.

Discussion The aim of this study was to describe the use of and adherence to the 7-step strategy and to assess whether this strategy had facilitated the development, implementation and continuation of a comprehensive, multi-component lifestyle program. For this purpose, a monitoring instrument with twelve performance indicators was developed which evaluated implementation factors that are targeted by means of the 7-step strategy and that have been found to be effective in implementing effective WHPP. Our findings showed that performance indicators were partially met by the organizations. This suggests lack of completeness of the execution of the 7-step strategy and thus partial strategy adherence, which is also represented in the overall performance score of both organizations. Incompleteness was linked to inadequate support among middle and lower management, project structure,

143

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

adaptation to needs of employees, planning and maintenance. Overall we identified several factors that had facilitated the implementation process: the presence of an external advisor, strong and adequate leadership (i.e., the presence of a project leader), solid support among all organizational layers, and fitting the program to the wishes and needs of employees. Factors that had hampered the implementation process included organizational culture, lack of higher management involvement in the steering committee, and lack of ownership among project members. Based on the results of this study it is reasonable to assume that the 7-step strategy facilitated the formation of a hierarchic project structure with a broad variety of project members in terms of function and department within the organization. Both aspects facilitated the implementation of the program. This is in line with literature [29, 30] and the study of Wynne and Clarkin that showed that key figures from different organizational levels (participation) and employees with different expertiseâ&#x20AC;&#x2122;s like human resources, communication and worksite health promotion (multidisciplinary) need to be part of a projectâ&#x20AC;&#x2122;s structure [13]. The results of this study showed that in order to successfully use the 7-step strategy as intended, the strategy may not have provided sufficient practical guidance or materials for the organizations to use in each step. First, the strategy should be more specific on which strategies to use for inclusion of middle and lower management right from the beginning of the project. In our study, the external advisor visited team meetings at the end of the project. Interviews showed that this personal approach was perceived as beneficial as it led to more support for the program. It is advisable that visiting team meetings as a means to create support among lower and middle management should be included in the 7-step strategy during step 1. By keeping all management levels up to date about upcoming interventions from the beginning, support could be ensured during the whole program. Second, the results showed that due to the lack of a clear format for a needs assessment both organizations were not able to perform a thorough analysis of the wishes and needs of the employees. Literature suggests that ensuring a fit of the program with the wishes and needs of the users and the organization, contributes to program participation and thus to implementation success [9-11, 31-32]. So in order to enhance the effectiveness of the strategy, it is necessary to develop and uptake a specific needs assessment in step 3 that focuses on both the users as well as the organization. For the users a specific needs assessment questionnaire could be included. At the organizational level it is important to leverage existing data to inform this phase. For example, data regarding sickness absence and health risk appraisals can

144

The 7-step strategy: helpful or not?

provide valuable information on the current health status of the organization and main focal areas. The tooling for step 3 should include information on which organizational data could be relevant and on how to make an organizational scan in which all data is integrated. Additionally, organizations would benefit from conducting an assessment in which barriers and facilitators can be identified for the development, implementation and continuation of health promotion activities. For this purpose the Dutch questionnaire MIDI can be used as part of step 3 of the strategy as a supporting tool [31]. The MIDI questionnaire can be used prior to the development and implementation of the program to determine possible barriers and facilitators for implementation of the program and respective interventions. The MIDI questionnaire can also be used during the implementation of the program to adjust the strategy of implementation to gain maximum success. Thirdly, to support the project members and organizations in developing a project plan during step 4, it would be beneficial to add an outline for a project plan that contains all essential paragraphs (i.e. overall goals, detailed intervention template, communication strategy, time-line and budget plan). Additionally, detailed examples of successful worksite health promotion programs could be added to the strategy. By presenting these examples (and potential factors for success) the organizations can have a clear view on whether or not interventions fit the needs and wishes of the employees. As stated by Wynne and Clarkin [13], flexibility and integration are two of the five aspects that are important for successfully implementing a health policy at the workplace. Meaning that WHPPs are similar at some basic points, but that WHPPs need to be adapted to fit the specifics of a worksite (e.g., organizational structure/culture type of workforce and vision), in order to reach its full potential. Finally, the performance indicators show that the low attention to maintenance of the program remains a problem, which has been found also in other studies [9]. Although both worksites in the current study evaluated the development and implementation of the interventions at the end of the program, they did not thoroughly discuss and determine long-term goals. Moreover, monitoring showed that within both worksites nobody felt highly responsible for making sure the program would be maintained over the years. We know that to date it remains difficult to achieve sustained attention for worksite health promotion programs [33]. Organizations need to make an extensive effort to make a considerable impact on the wellness and health of their employees and are thus not always willing to do so, as research has shown that only few programs are cost-effective [34]. Effects will only occur over a long period of time and for that to happen a program needs to be integrated in the organizations HR- policy (i.e., employability) To support organizations in doing so, more guidance seems

145

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

to be needed on how they can successfully achieve maintenance, preferably by examples of successful initiatives of other organizations [33]. Moreover, the 7-step strategy should be linked to existing structures, such as health risk appraisals to further stimulate maintenance of interventions and to encourage organizations to keep developing and implementing new interventions when necessary. Methodological considerations A strength of this study is that we applied qualitative research in a real-world setting, which can provide deeper and more explorative insights into the implementation process in addition to quantitative research. Because we performed our study in a real-world setting (i.e., the researcher was mainly an embedded observer) the generalizability of the results can be regarded as good. The generalizability of qualitative study results is not always good, because the aim of qualitative research is not to find results that are widely applicable, but to go in-depth and show underlying mechanisms [35, 36]. Inherent to qualitative research we need to take recall bias into account. However, we tried to control for this by combining the interviews with onsite monitoring (i.e., data triangulation) and by performing interviews at baseline, halfway through the project, and at the end of the project. Even though we invited all project members to participate in the interviews, not everybody was able to participate. Especially at the end of the project it proved to be difficult to include all project members at the Hospital since their involvement in the project had diminished over time. Furthermore response bias could be a problem with regard to the interviews. Project members may have provided more desirable feedback about the 7-step strategy when faced by the interviewer who was also the embedded researcher in the project. However, prior to starting the interview, we stressed that we were not looking for desirable answers and that it was of great importance to be honest. Conclusion The results of this evaluation indicate that the 7-step strategy was partly used as intended as not all performance indicators were met. So partial strategy adherence was obtained. However, the results of the interviews with the users did indicate that strategy adherence seemed to facilitate the development, implementation and maintenance of lifestyle interventions. Due to the fact that we did not have a control organization that did not use the 7-step strategy to implement lifestyle interventions, it was not possible to link the performance indicators to quantitative outcome measures. Therefore no single component

146

The 7-step strategy: helpful or not?

of the strategy could be identified as more effective. However, based on the qualitative data this study suggest that when the improvements will be made on both the content and performance, the 7-step strategy would be a useful and effective tool to successfully develop, implement and maintain a multi-component WHPP. Suggested improvements include: better practical tools and materials for project members to execute some of the steps and providing examples of successful WHPPs. Moreover, an external advisor with experience is needed to ensure and maintain progress and to make sure that an organization does not drop out during the development phase. As implementing a WHPP by means of an implementation strategy enables successful reach, participation, dose delivered and satisfaction among employees, the 7-step strategy may have good potential for successfully implementing WHPP worldwide in different types of organizations. We believe that, with the suggested improvements, the 7-step strategy will enable organizations to successfully develop, implement and maintain a multi-component WHPP that fits the wishes and needs of that particular organization.

147

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

References 1.

Kirsten W: Making the link between health and productivity at the workplace--a global perspective. Ind Health 2010, 48(3):251-255.

Jans MP, van den Heuvel SG, Hildebrandt VH, Bongers PM: Overweight and Obesity as Predictors of Absenteeism in the Working Population of the Netherlands. J Occup Environ Med 2007, 49(9):975980.

Grimshaw JM, Thomas RE, MacLennan G, Fraser C, Ramsay CR, Vale L, Whitty P, Eccles MP, Matowe L, Shirran L, Wensing M, Dijkstra R, Donaldson C: Effectiveness and efficiency of guideline dissemination and implementation strategies. Health Technol Assess 2004, 8(6):iii-iv, 1-72.

Jepson RG, Harris FM, Platt S, Tannahill C: The effectiveness of interventions to change six health behaviors: a review of reviews. BMC Public Health 2010, 8(10):538.

Ni Mhurchu C, Aston LM, Jebb SA: Effects of worksite health promotion interventions on employee diets: a systematic review. BMC Public Health 2010, 10(10):62.

Verweij LM, Coffeng J, van Mechelen W, Proper KI: Meta-analyses of workplace physical activity and dietary behaviour interventions on weight outcomes. Obesity Reviews 2011, 12(6):406-429.

Wierenga D, Engbers LH, van Empelen P, Duijts S, Hildebrandt VH, van Mechelen W: What is actually measured in process evaluations for worksite health promotion programs: a systematic review. BMC Public Health 2013, 13(1):1190.

10. Fleuren M, Wiefferink K, Paulussen T: Determinants of innovation within health care organizations: literature review and Delphi study. Int J Qual Health Care 2004, 16(2):107-123. 11. Durlak JA, DuPre EP: Implementation matters: a review of research on the influence of implementation on program outcomes and the factors affecting implementation. Am J Community Psychol 2008, 41(34):327-350. 12. Curran GM, Bauer M, Mittman B, Pyne JM, Stetler C: Effectiveness-implementation hybrid designs: combining elements of clinical effectiveness and implementation research to enhance public health impact. Med Care 2012, 50(3):217-226. 13. Wynne R, Clarkin N: Under construction: building for health in the EC workplace: Luxembourg: Officer for Offical Publication of the European Communities; 1992. 14. Wynne R, European foundation for the improvement of living and working conditions: Workplace health promotion in Europe. Research summary. Luxembourg: Office for official publications of the European Communities; 1997. 15. Wandersman A, Duffy J, Flaspohler P, Noonan R, Lubell KS,L., Blachman M, Dunville R, Saul J: Bridging the gap between prevention research and practice: the interactive systems framework for dissemination and implementation. Am J Community Psychol 2008, 41:171-181. 16. Wierenga D, Engbers LH, van Empelen P, Hildebrandt VH, van Mechelen W: The design of a real-time formative evaluation of the implementation process of lifestyle interventions at two worksites using a 7-step strategy (BRAVO@Work). BMC Public Health 2012, 12(1):619. 17. Steckler A, Linnan L: Proces Evaluation for Public Health Interventions and Research: San Francisco: Jossey-Bass; 2002.

148

The 7-step strategy: helpful or not?

18. Baranowski T, Stables G: Process evaluations of the 5-a-day projects. Health Educ Behav 2000, 27(2):157-166. 19. Glasgow RE, Vogt TM, Boles SM: Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health 1999, (89):9-1322. 20. Verweij LM, Proper KI, Hulshof CT, van Mechelen W: Process evaluation of an occupational health guideline aimed at preventing weight gain among employees. J Occup Environ Med 2011, 53(7):722729. 21. Pope C, Ziebland S, Mays N: Qualitative research in health care. Analysing qualitative data. BMJ 2000, 320:114-116. 22. Glaser BG: The Constant Comparative Method of Qualitative Analysis. Social Problems 1965, 12:436445. 23. Mays N, Pope C: Assessing quality in qualitative research. BMJ 2000, 320:50-52. 24. Boeije H: Analyze in qualitative research. (Analyseren in kwalitatief onderzoek. denken en doen): Amsterdam, The Netherlands: Boom onderwijs; 2009. 25. Wester F, Peters V: Qualitative analyse. Principles and procedures (Kwalitatieve analyse. Uitgangspunten en procedures): Bussem, The Netherlands: Coutinho bv; 2004. 26. Ni Mhurchu C, Aston LM, Jebb SA: Effects of worksite health promotion interventions on employee diets: A systematic review. BMC Public Health 2010, 10(10):62. 27. Jepson RG, Harris FM, Platt S, Tannahill C: The effectiveness of interventions to change six health behaviours: A review of reviews. BMC Public Health 2010, 8(10):538. 28. Chau JY, der Ploeg HP, van Uffelen JG, Wong J, Riphagen I, Healy GN, Gilson ND, Dunstan DW, Bauman AE, Owen N, Brown WJ: Are workplace interventions to reduce sitting effective? A systematic review. Prev Med 2010, 51(5):352-356. 29. Kwak L, Kremers SPJ, Van Baak MA, Brug J: Participation rates in worksite-based intervention studies: Health promotion context as a crucial quality criterion. Health Promot Internation 2006, 21(1):66-69. 30. Bartholomew LK, Parcel GS, Kok G, Gottlieb NH: Planning Health Promotion Programs: An Intervention Mapping Approach. San Francisco, CA: Jossey-Bass; 2006. 31. Fleuren M, Paulussen TH, van Dommelen P, van Buuren S: Towards a Measurement Instrument for Determinants of Innovations (MIDI). International Journal for Quality in Health Care 2014, 26(5):501510. 32. Wierenga D, Engbers LH, van Empelen P, de Moes KJ, Wittink H, Grundemann R, van Mechelen W: The implementation of multiple lifestyle interventions in two organizations: a process evaluation. JOEM 2014, 56(11):1195-1206. 33. Heinen L, Darling H: Addressing obesity in the workplace: the role of employers. Milbank Q 2009, 87(1):101-122. 34. van Dongen JM: Economic evaluations of worksite health promotion programs. PhD thesis. VU University Amsterdam, Amsterdam; 2014. 35. Mays N, Pope C: Qualitative research in health care: Assessing quality in qualitative research. Br Med J 2000, 320(7226):50-52. 36. Pope C, Mays N: Reaching the parts other methods cannot reach: An introduction to qualitative methods in health and health services research. Br Med J 1995, 311(6996):42-45.

149

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

General discussion

Chapter 7

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 152

General discussion

An unhealthy lifestyle among employees has a major financial impact on organizations in terms of productivity loss and sickness absence rates [1-4]. Additionally, the prevalence of an unhealthy lifestyle among the Dutch population is high. Therefore, numerous worksite health promotion programs (WHHPâ&#x20AC;&#x2122;s) have been developed and studied intensively on their effectiveness over the past two decades [5-11]. Implementation in daily practice, however, of WHP programs has been limited. High quality process or formative evaluations may improve our understanding of factors that can contribute to the implementation process of such programs. However, only a few studies that have evaluated the effectiveness of a WHP program have also systematically evaluated the implementation process [12]. Hence, the focus of this dissertation was to systematically evaluate the implementation process and to gain insight into implementation factors that contribute to the adoption and uptake of a WHP program in practice. A specific focus was put on the use of a 7-step implementation strategy by two participating worksites of different organizations (hereinafter called worksites), as a means to improve the quality of the implementation and effectiveness of a WHP program. The 7-step implementation strategy was aimed to ensure the: 1) creation of solid support, 2) formation of a project structure, 3) performance of a needs assessment, 4) development of tailored evidence-based interventions, 5) adequate implementation of WHP programs, 6) evaluation of the intervention, and 7) the maintenance of the intervention. These steps were the result of a comprehensive survey including more than 1500 WHP programs. This survey suggested that these 7 steps are crucial for the establishment of a successful WHP program [13]. The two main objectives of this thesis were: 1) To identify implementation determinants (i.e. barriers an facilitators) that either hamper or facilitate the implementation of WHP programs focusing on healthy lifestyle changes, and; 2) To assess whether the use of the 7-step strategy contributed to the successful development, implementation and maintenance of a WHP program aimed at stimulating a healthy lifestyle change among employees. In this general discussion, the main findings of this thesis are presented first. Second, some methodological considerations are addressed. Third, taking the main results into account, an improved 7-step strategy is proposed. Fourth, recommendations and opportunities for future research as well as for daily practice are presented. And finally, this discussion ends with concluding remarks. 153

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

Overview of main findings The main findings of this thesis are presented below by addressing the two main objectives. To identify implementation determinants (i.e. barriers and facilitators) that either hamper or facilitate the implementation of WHP programs focusing on healthy lifestyle changes In order to address this question we systematically reviewed the literature on WHP programs focusing on healthy lifestyle changes (chapter 3), and examined implementation factors within a systematically planned implementation evaluation study (chapters 4 through 6). Based on these combined findings we were able to identify key implementation factors at the organizational, strategy and participant level. At the organizational level, core elements of implementation success were mainly related to management. We observed that (upper) management support and management commitment at the start of the program facilitated program implementation. Active management participation and engagement in the interventions together with employees facilitated participation. Conflicting interests between organizational goals and program goals hampered implementation as it could result in a lack of prioritizing the program. At strategy level several factors seemed to have a noteworthy impact on the implementation process. The following factors could ensure the success of a WHP program: •

Presence of an external supervisor to make sure that the project remains a priority in light of daily working life and ad hoc tasks;

•

Presence of an enthusiastic project leader with sufficient time and skills;

•

Formation of a solid project structure that includes employees from all organizational levels;

•

Installing a steering committee consisting of managers with decision-making authority;

•

Fitting the program to the employee’s needs;

•

Focussing on facilitating employees in adopting and maintaining healthy behaviour and making the program voluntary;

•

Creating solid support among employees by clear communication about the goals and specific interventions;

•

Presence of a detailed project and communication plan that serves as a guide throughout the implementation process;

154

•

Compatibility of the program with the organization and the organizational culture;

•

Ensuring good collaboration between all involved stakeholders;

General discussion

•

Ensuring that the program is beneficial compared to the current situation;

•

Ensuring that no negative consequences are observed and the company, managers, implementers and participants benefit from participation (i.e., relative advantage);

•

Ensuring that implementers have sufficient time, skills, knowledge and competence in order to successfully implement the program.

On the other hand, persistent barriers for the implementation of a WHP program at the strategy level were also identified and included: •

Lack of financial, personnel, and material (e.g., equipment, facilities) resources, or lack of space or facilities;

•

Lack of time and a heavy workload of implementers;

•

High staff turnover among project leaders;

•

Lack of ownership among project members;

•

Use of an ad hoc approach; meaning that interventions were developed on the spot, without thoroughly thinking them through;

•

Lack of financial transparency.

We also identified several implementation determinants at the level of the participants (i.e. employees). First, it is important that organizations focus on positively influencing the employee’s attitude towards the program implementation, meaning that employees positively support the focus of the program and the fact that their employer offers lifestyle interventions. Especially a positive attitude towards employer involvement in employees’ lifestyle choices is important, since employees can perceive this as private and not something their employer should intervene in. Second, in order to increase participation rates, employees need to perceive social support from colleagues and managers during the implementation phase. Finally, the program should be easily integrated in the daily working live and work tasks of employees. Managers need to facilitate participation when time constraints or high work demands are experienced. To assess whether the use of the 7-step strategy contributes to the successful development, implementation and maintenance of a WHP program, with a focus on healthy lifestyle changes Chapter 6 showed that the 7-step strategy was partly used as intended and that moderate strategy adherence was obtained. Both organizations found the 7-step strategy a useful

155

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

tool to implement lifestyle interventions in an organization, as it is similar to structural project management and therefore easy to apply. The strategy focused on those elements that contributed to the implementation process and on key issues within an organization. The strategy also focused on those aspects that contribute to participation and satisfaction of employees with the implemented program. This makes the 7-step strategy in theory a valuable strategy to maximize implementation and impact. However, in order to answer this aim it should be noted that both organizations did not actively use the strategy themselves, as they trusted that the external advisor would monitor the process of following all the 7 steps. And, according to the principles of the 7-step strategy, the external advisor intervened or provided guidance when necessary. However, without his guidance the strategy would probably not have been used. In order to increase use of the strategy in future projects it is advisable to improve the implementation of the 7-step strategy (see improved 7-step strategy: lessons learned). In order to address this second aim of this thesis further, we needed to define success of the 7-step strategy. In this study we struggled with this issue, because no specific outcome variable could be formulated. However, the 7-step strategy was designed to maximize the effectiveness of a WHP program. In order to be effective, evidence-based interventions should be properly implemented and embedded in an organization. So we propose that the 7-step strategy is successful if the following three conditions are met: 1) First, the 7-step strategy should lead to the development and implementation of evidence-based interventions; 2) Second, employees need to adopt and adhere to the implemented interventions, and; 3) Finally, the implemented interventions should be embedded in the organizations health policy. Does the 7-step strategy lead to the development and implementation of evidence-based interventions? Chapter 4 and 6 showed that few evidence-based interventions were implemented in both organizations. This was first of all related to one of the most important shortcomings that users in this study stated: the fact that the strategy did not yield sufficient concrete and practical tools or materials to be able to use the strategy as intended. The strategy did not provide enough guidance in the selection of (existing) evidence-based interventions. The lack of a detailed project and communication plan also contributed to the lack of

156

General discussion

evidence-based interventions. The lack of a project plan resulted in vague goals and unclear or wrongly timed communication to employees (i.e. to early or after the intervention had been implemented). Second, this thesis showed that the organizations under study primarily paid attention to interventions that were aimed at creating awareness about the consequences of unhealthy lifestyle behaviour and work related stress. Both organizations focused on relatively simple and easily implemented interventions, which did not require active employee participations, but which would potentially generate high visibility within the organization. Even though high visibility is a core element in implementation, this seems like a paradox. Organizations seem to be more interested in “quick wins” and whether or not employees would be satisfied with the intervention, rather than the intervention leading to effects on health. Organizations choose mostly environmental and educational interventions that could be implemented easily at low costs and effort with high visibility as a result, regardless of whether these interventions were proven effective. In summary, this thesis showed that the first condition for a successful use of the 7-step strategy was only partially met. Improvements to the strategy in terms of practical tools and type of interventions are necessary to achieve successful development and implementation in the future. Do employees adopt and adhere to the implemented interventions? Chapter 4 showed high overall reach (96.6%) and participation (75.1% participated in at least one intervention). Overall employees were moderately satisfied with the interventions (6.8 ± 1.1), mainly because they perceived the programme to be a good initiative, which showed that their employer acknowledged the value of employee health and lifestyles. On a more critical note, the long-time interval between some interventions reduced employee’s interest in the WHP program and decreased their satisfaction levels. However, most interventions were perceived to be “pragmatic” and were thought to facilitate employees and to fit their needs. These three aspects can be attributed to the 7-step strategy, specifically steps 2 and 3, as the strategy specifically applies a user-centred approach. Aspects of the strategy that contributed to these high participation rates include performing a needs assessment (step 3) and the fact that employees were part of the project group and were actively involved in the development and implementation process. Thus, with regard to the second condition we can conclude that the 7-step strategy was indeed successful, since a high reach and participation rates were obtained and this led to the adoption of the program among most employees in the participating worksites.

157

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

Is the program embedded in the organizations health policy? Chapter 4 and 6 revealed that maintenance proved to be a problem, which is also common in other studies [12, 14]. The first key issue was the lack of discussing long-term goals at the beginning of the project. Both organizations started thinking about maintenance when the WHP program was coming to an end. It is expected that an early discussion about possible maintenance strategies would increase continuation of the WHP program. The second key issue was that both organizations based their maintenance decision on their own and their employee’s satisfaction levels (Chapters 4 and 6), instead of focusing on which interventions were hypothesized to have an effect in the long run. This was related to the fact that the organizations did not pay attention to evidence-based interventions at the beginning of the project. A few environmental interventions (e.g., an enlarged bike shed and new building route) and relatively easily implemented interventions (e.g., chair massages and free fruit) were continued. On a more positive note; at one organization the topic ‘vitality of employees’ and subsequent interventions was integrated in yearly performance interviews. One can argue that integrating the topic of vitality in yearly performance interviews has the potential to have great effects, as it is tailored to the individual worker. Thus, with regard to this third and final condition for the success of the 7-step strategy, we can conclude that the strategy partly led to maintenance. Mainly tools in order to facilitate maintenance of the program in daily practice are currently lacking in the strategy. Methodological contributions In our opinion, this thesis contributed to shedding light on a methodological gap in implementation research. Chapter 3 showed that process evaluations are not systematically performed alongside effectiveness studies and that they are generally rated as being of ‘poor’ to ‘average’ quality. The main reason for this classification was that, in most cases, no conceptual frameworks were used to systematically guide the implementation process. Furthermore, current frameworks do not have a specific focus on implementation factors. Consequently, we highlighted the need for future research with regard to the identification of implementation determinants. This is why we proposed a new conceptual framework for conducting process or formative evaluations in Chapter 2. The proposed conceptual framework was developed, based on literature on process evaluations. We combined three known and frequently used frameworks to capture every aspect of the adoption, implementation and maintenance process, as well as the influence of implementation determinants on the phases of the implementation process. Until now, no such framework

158

General discussion

has been developed by which the implementation process can be studied systematically. The final framework as proposed and applied in this thesis consists of eight components: 1) context, 2) recruitment, 3) reach, 4) dose delivered, 5) dose received, 6) fidelity, 7) satisfaction and 8) maintenance. Each of these components measures a different aspect of the adoption, implementation and continuation of a program. Under the component â&#x20AC;&#x2DC;contextâ&#x20AC;&#x2122;, the implementation determinants and their positive or negative influence on the implementation process were measured. Another positive feature of using the proposed framework is that three different levels are included for each component. These are: the organizational level, the implementer level and the employee level. In this way, we were able to identify successful components and implementation determinants at each level, which is beneficial for future implementation. The proposed conceptual framework greatly facilitated the systematic evaluation of the implementation process in this study. Without this framework we would not have been able to gain insight into the natural process of implementation in daily practice and to assess the use and adherence to the 7-step strategy. Therefore we strongly recommend our framework for use in future research initiatives. Moreover, the framework has already been used by Hoekstra et al. (2014) [15]. Methodological issues Some methodological issues should be considered regarding the strengths and limitations of the BRAVO@Work study. To our knowledge, the BRAVO@Work study is one of the first studies that systematically monitored and evaluated the natural course of the implementation process by means of a real-time formative evaluation within a controlled trial. We measured our process evaluation components prior, during and post implementation, whereas until now implementation studies only have performed post-hoc evaluations. In order to gain a thorough insight into the implementation process and the use of the 7-step strategy, we applied data triangulation (i.e. a combination of quantitative and qualitative methods). Data triangulation enables researchers to look for patterns in the collected data in order to develop an overall interpretation, which includes multiple views on the implementation process [16, 17]. The qualitative and quantitative data complemented each other and created a rich dataset for the interpretation of the natural course of implementation. The results of this study can be regarded as innovative and provide valuable information about the implementation process. This improves the translation from research into practice. Another strength of this study is the stringent focus on capturing and monitoring the implementation process by

159

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

an embedded scientist. Combined with the fact that the worksites themselves were in the lead, we were able to evaluate the use of the 7-step strategy, as well as the dynamic process of implementation of the BRAVO-interventions in real-world circumstances [18]. This is a similar approach as used in action research in which the purpose is to solve a particular problem and to produce guidelines for best practice. As Peter Reason and Hilary Bradbury say: â&#x20AC;&#x153;...it seeks to bring together action and reflection, theory and practice, in participation with others, in the pursuit of practical solutions to issues of pressing concern to people, and more generally the flourishing of individual persons and their communities.â&#x20AC;? [19, 20]. The main difference with our approach is that we did not include a proper feedback loop from the researchers to the worksites. Instead the external advisor was responsible for providing real-time feedback to the project groups, which they could use to alter their implementation plan, based on previous experience. Mainly because we were interested to see how an organization would cope with using the strategy in daily practice without involvement of the scientific world. This is also the reason that we did not apply the traditional randomized controlled trial (RCT) design. The RCT is seen as the golden standard and the preferred design for investigating the effectiveness of an intervention [21, 22]. However, the applicability of the results of a RCT in daily practice are questionable. The design aspects that contribute to the high external validity (e.g. well-defined exclusion criteria) of a RCT may in real life hamper the generalizability of results to other organizations [18, 22]. In contrast, the use of a pragmatic design facilitated the generalizability of results (i.e. enhances the external validity) to other types of organizations than the ones in this study. Since we were interested in the implementation process in daily practice, the choice was made to apply a quasiexperimental design with elements of action research. Some limitations should also be mentioned. Both intervention worksites were active in the health sector. We hypothesize that this implies that participants generally had a higher awareness of the importance of a healthy lifestyle compared to the average Dutch employee. Furthermore, we faced the problem that implementation science is a relatively new field. Hence, psychometrically validated instruments measuring the process of implementation are scarce and are not frequently used. This is largely the result of the challenging nature of the real world setting in which these studies often take place [23-25]. As each setting and intervention is unique, to our knowledge, no valid process evaluation questionnaire was available for use in the BRAVO@Work study. In order to ensure the quality of our outcomes we did incorporate an adapted version of a newly proposed Dutch instrument by Fleuren et al. 2014 for measuring implementation determinants [26, 27]. Finally, the main focus on

160

General discussion

the natural implementation process hampered the ability to obtain sufficient data regarding the effectiveness of implemented programs. Since both worksites were in control, they were involved in the development of the questionnaire. Consequently we needed to make concessions regarding length and content of the questionnaire. Additionally, both worksites choose not to conduct physical examinations. Improved 7-step strategy: lessons learned Taking the results of this thesis into account, the 7-step strategy as presented in Chapter 2 of this thesis needs to be reconsidered. Chapter 6 showed that both worksites experienced difficulties in executing each step as intended. Therefore, in this section of the general discussion, improvements to the overall strategy and to each of the seven steps are presented so it can be used most effectively in future implementation of WHHPâ&#x20AC;&#x2122;s. Improvements to the overall strategy The 7 steps as defined in the BRAVO@Work study remain similar, because project members stated that the steps were logical. But in order to increase the performance of the strategy and to enhance implementation success four main improvements to the overall strategy can be made, namely 1) including practical tools, 2) including a suggested timeline, 3) including a questionnaire to gain insight into implementation determinants, and 4) stimulating an ongoing cyclic process of sustainable employability. Firstly, the strategy needs to contain more (practical) materials to substantiate each step. Our suggested improvements regarding practical materials are included in the improvements of each step below. Secondly, the strategy needs to be optimized by specifying a timeline for successfully executing each of the 7 steps in order to manage expectations. Based on the insights obtained in Chapter 6 it is reasonable to assume that steps 1 through 4 should take on average 6 months. Steps 5 through 7 should take on average 12 months in which implementation, evaluation and maintenance should be integrated. Figure 1 gives an overview of the proposed timeline. Thirdly, the strategy did not explicitly account for all modifiable implementation determinants that could have influenced the development, implementation and maintenance process. Different barriers and facilitators can apply to different organizations. For this purpose the Dutch questionnaire MIDI developed by Fleuren and colleagues (2014) could be used at the initiation of the project [26, 27]. The MIDI questionnaire can be used

161

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

to identify relevant implementation determinants (i.e., barriers and facilitators). In this way, the project group can anticipate the identified barriers and facilitators applicable to their organization, eliminate predictable obstacles and also reinforce already existing facilitators. Finally, this thesis showed that both worksites applied the 7-step strategy as a single project instead of a tool that can be used to stimulate a continuous quality circle to keep improving and to structurally embed an organizations health policy. Every organization is to date. This is also important in light of a letter of the Dutch Ministry of Social Affairs and

dynamic and changes over time. By regularly (e.g. annually) repeating steps 3 through 6, Employment (i.e. "Ministerie van Sociale Zaken en Werkgelegenheid") regarding future work related

the implemented policy can be kept up to date. This is also important in light of a letter of care in January 2015 [28]. In this letter, the ministry states the following: "Prevention and

the Dutch Ministry of Social Affairs and Employment (i.e. “Ministerie van Sociale Zaken en sustainable employability are important objectives for work related care. Good work related care is

Werkgelegenheid”) regarding future work related care in January 2015 [28]. In this letter, aimed at preserving, restoring and improving the health and sustainable employability of workers. the ministry states the following: “Prevention and sustainable employability are important Prevention for andwork improving workers employability are the joint responsibility of employer objectives related care.sustainable Good work related care is aimed at preserving, restoring and employee." this letter, ministry included the advice given by the Dutch Social and improving [28]. theInhealth andthesustainable employability of workers. Prevention and Economic Council (i.e.sustainable "Sociaal EconomischeRaad named care for workers" in improving workers employability(SER)") are the joint"Better responsibility of employer and September 2014 Theyletter, stated the that ministry preventionincluded and working sustainable areSocial the employee.” [28].[29]. In this theon advice givenemployability by the Dutch core of work related(i.e. care. A futureEconomischeRaad system should (more(SER)”) than isnamed currently“Better the case) focus Economic Council “Sociaal care foron workers” in preventing health problems absenteeism. Not only among employees but also among the Dutch September 2014 [29]. Theyand stated that prevention and working on sustainable employability working population. This will result in social costsystem savings according to thethan council [29]. The improved are the core of work related care. A future should (more is currently the case) 7-stepon strategy could help organizations work on preventionNot andonly sustainable of also focus preventing health problemsto and absenteeism. amongemployability employees but their employees. among the Dutch working population. This will result in social cost savings according to the

council [29]. The improved 7-step strategy could help organizations to work on prevention and sustainable employability of their employees. Step 3

Step 4 Step 5 through 7

Step 2

Start

month 1

month2-3

month 4-6

month7 - 18

Step 1 - ongoing

Figure 1. Suggested time line

General discussion

Improved step 1: continuously creating solid support at all organizational levels This thesis showed that, in the current form, the 7-step strategy specifically addresses the importance of creating solid support at the beginning of the development of the WHP program. The strategy states that employees need to be part of the project structure and that upper management needs to ensure their support by signing a letter of intent and reserving budgets. Given the results of this thesis, it should be emphasized that creating solid support should be an ongoing focal point and take place on all organizational levels (upper, middle and lower management and employee). Three additional approaches should be used to ensure ongoing support on all organizational levels. First, in line with the results of the review by Durlak and Dupre (2008), appointing ambassadors for each department is a successful strategy in maintaining support among employees. Ambassadors are employees with a positive attitude, who are actively trying to enthuse their colleagues about the project and who have an intrinsic motivation for the programsâ&#x20AC;&#x2122; subject. They are not necessarily part of the project group, but get detailed briefings on the current status of the program and the interventions [30]. The second approach is ongoing and clear communication. Durlak and Dupre (2008) also showed that employees feel more included and have higher levels of ownership when a transparent and personal communication approach was adopted [30]. This can be accomplished by visiting team meetings at the beginning of the program. In doing so, employees know what to expect and critical notes, questions and ambiguities can be clarified. As a bonus this facilitates a positive employee attitude towards the program, which is important as this enhanced program participation. Throughout the project clear communication on both successes and failures of the project is key [30]. This could be communicated in a regular feature on the organizations intranet or in the staff magazine and complements the personal approach. The third approach is specifying the role of team managers (i.e. lower management). As became clear in this thesis, team leader involvement leads to a direct impact on employeesâ&#x20AC;&#x2122; behavior. Support can be ensured by actively and timely informing lower and middle management on the intervention content and their role in program implementation. This can be accomplished by making the WHP a permanent item on the agenda of management meetings.

163

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

Improved step 2: formation of a project structure with external advisor The 7-step strategy specifies the importance of forming a steering committee, a project group and an internal project leader. However, this thesis showed that an external advisor with experience in applying the 7-step strategy is an essential element within the project structure. The presence of the advisor in this study prevented premature termination of the program. In light of daily activities in an organization, the project members and internal project leader experienced difficulties in prioritizing the program. The combination of an external advisor and internal project leader with a set amount of hours a week to spend on the program proved to be essential to stay on track. Improved step 3: needs assessment and identification of organizational culture A proper needs assessment questionnaire needs to be included in step 3, which gives insight into both A) the current health and lifestyle status of the employees (e.g. a lifestyle mental health and physical health questionnaire), as well as B) the needs and wishes of the employees. This combination is essential. Chapter 5 of this thesis showed that when a program fits the employee’s needs, they have a more positive attitude towards the program, which enhances employee participation. However, from an organizational perspective one wants to intervene on those behaviors that will generate the biggest impact. Hence a combination of A and B is vital. During this third step it is also important to assess the organizational culture as implementation is facilitated when the WHP program fits the culture and core values of an organization [30-33]. Organizational culture is a complex phenomenon. Weiner et al. (2009 and 2011) and Klein and Sorra (1996) showed that there a two aspects that influence the implementation of an intervention: a) the fit of the program with the organizational context and; b) the organizational climate [31, 34, 35]. Weiner et al. (2011) described the fit with the organizational context as how well the intervention is tailored to the wishes and needs of the organization and if the intervention is in line with company values [34]. The second aspect, organizational climate, is ‘the extent to which intended users perceive that innovation is expected, supported and rewarded’ [35]. The latter can influence the implementation through policy, politics and general organizational factors [34]. Both aspects are intertwined within organizational culture as defined by Schein (1985): ‘A pattern of shared basic assumptions that was learned by a group, as it solved its problems of external adaptation and internal integration, that has worked well enough to be considered valid and therefore, to be taught to new members as the correct way to perceive, think and feel in

164

General discussion

relation to those problems.’ [36]. Durlak and Dupre (2008) and Weiner et al. (2011) showed that organizational culture and company values determine the implementation climate [30, 34]. The Competing Values Framework (CVF) developed by Cameron and Quinn is a useful framework for assessing organizational culture. This framework is commonly used to determine the existing organizational culture and includes the validated and reliable Organizational Culture Assessment Instrument (OCAI) [37-39]. It distinguishes four dominant organizational cultures. By identifying the organizational culture before the start of the program, implementers can identify and anticipate on possible obstacles beforehand and act upon them. Improved step 4 and 5: development and implementation of interventions The key to successful implementation (step 5) is a well-developed intervention program (step 4) [12, 30]. However, in the current form, step 4 mainly lacks practical tools to help organizations develop an evidence-based program that fits both the organizations and employee’s needs. Chapter 4 and 6 of this thesis showed that a project plan, which focuses on the bigger picture, as well as details regarding interventions facilitated a structured implementation. For example, one of the worksites did not include a detailed intervention template. As a result they implemented ‘ad hoc’ interventions that did not fit the needs of employees nor were hypothesized to be effective in changing behavior. An empty project plan format must therefore be included in the strategy which contains the following aspects: desired changes and project goals, an intervention template, a timeline, a communication plan, a budget plan, a maintenance plan and a list of persons involved, including their tasks and responsibilities. In addition to this format, a list of evidence based interventions needs to be included in step 4 to make sure organizations do not only choose relatively simple interventions with an ‘ad hoc’ approach, but also to achieve a focus on multi-component interventions that have an effect in the long run. This list needs to include a comprehensible intervention description along with the dosage of the intervention. In this thesis we focussed on interventions to change lifestyle behaviour. However, for an organizations health policy it is also important to focus on other relevant themes like employability, safety, work environment, stress and ergonomics.

165

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

Improved step 6: Integrative evaluation The 7-step strategy states that step 6, the evaluation, consists of an integrative evaluation as well as an overall process and effect evaluation. The results of this thesis showed that the integrative evaluation was perceived as facilitating. It was part of the natural process of the project meetings and led to adjustments to the project plan when necessary, similar as in action research [19, 20]. However, both organizations relied on the process and effect evaluation of the researchers. In light of the results of this thesis it is questionable whether extensive process and effect evaluations are necessary for good corporate HR services. Chapter 4 showed that both worksites based their maintenance decision on satisfaction and awareness levels among employees, without looking at the effects. Therefore we recommend that organizations use their employee satisfaction survey or periodic medical examination to gain insight into behavior change. Additionally, short process evaluation questionnaires for employees regarding dosage and satisfaction for each intervention are beneficial and easily conducted by the organizations themselves. Improved step 7: Embedding WHP in the organizations health policy At the organizational level, maintenance covers the extent to which a program becomes institutionalized or part of the routine practices of an organization.Step 7 currently lacks a specific strategy that organizations can use to ensure embedding of WHP in the organizational health policy. This is essential in order to enhance long-term effectiveness [40]. Unfortunately, to our knowledge, there are no studies that investigated the best approach for maintenance of a WHP program, while the results of our study call for practical tips and tools to ensure maintenance of a WHP program in daily practice. Therefore, based on the results of this thesis and our practical experience we suggest that organizations should focus on three aspects. First, it is important that more than one employee or manager in the organization is made responsible for the organizational health policy and overall vitality of employees. At least one person from the human resource department with enthusiasm and knowledge about the topic should be actively involved. By making it a joint responsibility with (an)other coworker(s) it is less likely that the WHP will end when one of the responsible employees drops out. Second, the health policy should be a permanent item on the agenda of management meetings. It is hypothesized that this will make it more likely that it will become part of mentality and culture throughout the organization. Finally, regularly (e.g. annually) repeating steps 3 through 6 can keep the health policy up to date would ensure that support for the policy on all organization levels remains stable.

166

General discussion

Recommendations and implications for research There are still a lot of challenges and questions in the field of implementation research and WHP that should be answered. More knowledge is needed regarding the following: •

This thesis has clearly demonstrated the need for a systematic approach for process evaluations as a way of improving WHP program implementation. Until now, process evaluations are in general of low to average quality and this needs to be improved, because good quality evaluations will enable researchers to identify effective intervention components. In order to set a first step in this direction, we suggest that future process evaluations need to apply the framework of Wierenga et al. 2012 as proposed in Chapter 2. One of the main reasons for this recommendation is that this framework explicitly takes implementation determinants into account. It also provides a detailed operationalization of each process component at three different organizational levels, which is lacking in other frameworks.

•

In order to successfully conduct process and formative evaluations, questionnaires with good psychometric properties should be developed for collecting such data. An important consideration in this development process should be to minimize the burden that data collection places on study participants. This will help to obtain a higher response rate and thus a more complete view of the process components and the implementation process.

•

Applying a mixed methods approach proved to be very useful. Detailed insight into implementation determinants and into the development and implementation process was obtained. In this study interviews were performed prior, during and after implementation. This comprehensive and time consuming approach was relevant for this thesis as this was one of the first studies that primarily focused on implementation and its determinants. However, for future research, interviews at the end of the project should be sufficient if researchers are present during project meetings. Mainly because the interviews prior and during did not generate many new insights on top of the insights we gathered by attending project meetings. The project meetings give solid information, which could also be used to timely adjust aspects of the implementation strategy. Additionally, the interviews are big burden for participating project members.

•

For studies that focus on implementation, an embedded scientist should be part of the study design. The presence of an embedded scientist allowed for extensive realtime monitoring. This gave insight into the direct influence of barriers hindering

167

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

the implementation process. Hence, a better understanding of the course of the program was obtained. â&#x20AC;˘

Newly designed interventions should be co-developed with organizations. The action research approach is very suitable for this. This approach ensures fitting the program with the organizational needs, as employees are involved. But it also allows researchers to gain insight into the key elements of the intervention program. Since implementation takes place in the real-world setting it is expected that the generalizability of the results is greater.

â&#x20AC;˘

This study showed that simply stating that organizations should focus on maintenance did not lead to the desired effect. Consequently, future research in the field of WHP should focus on identifying effective strategies to ensure embedding of programs in an organizations health policy, in order to enhance longterm effectiveness.

Recommendations and implications for practice During this course of this thesis, prevention and sustainable employability has been gaining more and more interest in organizations and Dutch society. Politics have given it more attention. In 2015 it was officially stated that prevention and working on sustainable employability are the core of work related care and a joint responsibility of employer and worker [28, 29]. Since this thesis focused on implementation of a WHP program in daily practice it is only logical that we identified recommendations for organizations and practitioners. For successful implementation of a WHP program in daily practice, it is important that each initiative is placed under a single umbrella concept focusing on sustainable employability. The improved 7-step strategy can be used to determine the contents of the organizations approach towards sustainable employability. This is vital for successful implementation as the strategy ensures fitting the interventions to the needs and wishes of both the organization and its employees. Additionally, the strategy accounts for modifiable implementation determinants, which can either hinder or facilitate the implementation process. Appointing an external advisor is strongly recommended. He or she can make sure that the program remains a priority in light of daily activities and provides guidance through each step. Additionally, this thesis showed that organizations have the tendency to choose relatively simple and easy to implement interventions. However, in order to ensure longterm effects it is advisable that organizations include known effective interventions. The

168

General discussion

external advisor can play an important role in ensuring this. The external advisor (if he or she has a scientific background) can also be given the role of embedded scientist. This allows for monitoring, proper evaluations and adjustments to the implementation when necessary, which improves implementation success. Finally, in order to ensure long-term effects the vitality management needs to become part of an organizations health and safety policy. Concluding remarks This thesis showed that although the use of the 7-step strategy helps overcome implementation determinants, the strategy only partly led to the successful development, implementation and maintenance of a WHP program. The strategy was not able to induce the implementation and maintenance of evidence-based interventions, which could, in theory, have led to favorable effects on several lifestyle behaviors. The findings as described above make clear that, although the 7-step strategy was perceived as a useful tool to systematically implement lifestyle interventions in an organization, improvements need to be made before nationwide use can be recommended.

169

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

References 1.

Kirsten W: Making the link between health and productivity at the workplace -A global perspective. Ind Health 2010, 48(3):251-255.

Alavinia SM, Molenaar D, Burdorf A: Productivity loss in the workforce: Associations with health, work demands, and individual characteristics. Am J Ind Med 2009, 52(1):49-56.

Jans MP, van den Heuvel SG, Hildebrandt VH, Bongers PM: Overweight and Obesity as Predictors of Absenteeism in the Working Population of the Netherlands. J Occup Environ Med 2007, 49(9):975980.

Goetzel RZ, Ozminkowski RJ: The health and cost benefits of work site health-promotion programs. In Edited by Anonymous 2008:303-323. [Anonymous, vol 29.]

Proper KI, Staal BJ, Hildebrandt VH, van der Beek AJ, van Mechelen W: Effectiveness of physical activity programs at worksites with respect to work-related outcomes. Scand J Work Environ Health 2002, 28(2):75-84.

Kuoppala J, Lamminpaa A, Husman P: Work health promotion, job well-being, and sickness absences-a systematic review and meta-analysis. J Occup Environ Med 2008, 50(11):1216-1227.

Robroek SJ, van den Berg TI, Plat JF, Burdorf A: The role of obesity and lifestyle behaviours in a productive workforce. Occup Environ Med 2011, 68(2):134-139.

Rongen A, Robroek SJW, van Lenthe FJ, Burdorf A: Workplace Health Promotion: A Meta-Analysis of Effectiveness. Am J Prev Med 2013, 44(4):406-415.

10. WHO: World Health Organization: Workplace health promotion. The workplace: a priority setting for health promotion. Accessed 2011 jan 16: Accessed on 1-6-2012 of: http://www.who.int/occupational_ health/topics/workplace/en/index.html. 11. WHO: Mental health and well-being at the workplace - protection and inclusion in challenging times: Copenhagen; 2010. 12. Wierenga D, Engbers LH, van Empelen P, Duijts S, Hildebrandt VH, van Mechelen W: What is actually measured in process evaluations for worksite health promotion programs: a systematic review. BMC Public Health 2013, 13(1):1190. 13. Wynne R, Clarkin N: Under construction: building for health in the EC workplace: Luxembourg: Officer for Offical Publication of the European Communities; 1992. 14. Heinen L, Darling H: Addressing obesity in the workplace: the role of employers. Milbank Q 2009, 87(1):101-122. 15. Hoekstra F, Alingh RA, van der Schans CP, Hettinga FJ, Duijf M, Dekker R, van der Woude L: Design of a process evaluation of the implementation of a physical activity and sports stimulation programme in Dutch rehabilitation setting: ReSpAct. Implement Sci 2014, 9(1):127. 16. Mays N, Pope C: Rigour and qualitative research. BMJ 1995, 311(6997):109-112. 17. Mays N, Pope C: Assessing quality in qualitative research. BMJ 2000, 320:50-52. 18. Patsopoulos NA: A pragmatic view on pragmatic trials. Dialogues Clin Neurosci 2011, 13(2):217-224. 19. Brydon-Miller M, Greenwood D, Maguire P: Why action research? Action Research 2003, 1(1):9-28. 20. Reason P, Bradbury H: Handbook of action research: Participative inquiry and practice. London: Sage Publications; 2001. 21. Moher D, Hopewell S, Schulz KF, Montori V, Gotzsche PC, Devereaux PJ, Elbourne D, Egger M, Altman DG, CONSORT: CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials. Int J Surg 2012, 10(1):28-55.

170

General discussion

22. Kristensen TS: Intervention studies in occupational epidemiology. Occup Environ Med 2005, 62:205210. 23. Martinez RG, Lewis CC, Weiner BJ: Instrumentation issues in implementation science. Implement Sci 2014, 9:118-014-0118-8. 24. Cook DA, Beckman TJ: Current concepts in validity and reliability for psychometric instruments: theory and application. Am J Med 2006, 119(2):166.e7-166.16. 25. Weiner BJ: A theory of organizational readiness for change. Implement Sci 2009, 4:67-5908-4-67. 26. Fleuren M, Paulussen TH, van Dommelen P, van Buuren S: Development measurement instrument for determinants of innovations (MIDI) [Ontwikkeling meetinstrument voor Determinanten van Innovaties (MIDI)]. TNO 2012, TNO 2012 R10625. 27. Fleuren M, Paulussen TH, van Dommelen P, van Buuren S: Towards a Measurement Instrument for Determinants of Innovations (MIDI). International Journal for Quality in Health Care 2014, 26(5):501510. 28. Ministerie van Sociale Zaken en Werkgelegenheid: Toekomst arbeidsgerelateerde zorg (SER-advies). 28 Januari 2015. 29. Sociaal Economische Raad (SER): Beter zorg voor werkenden. Een visie op de toekomst van de arbeidsgerelateerde zorg. 14 September 2014. 30. Durlak JA, DuPre EP: Implementation matters: a review of research on the influence of implementation on program outcomes and the factors affecting implementation. Am J Community Psychol 2008, 41(34):327-350. 31. Weiner BJ, Lewis MA, Linnan LA: Using organization theory to understand the determinants of effective implementation of worksite health promotion programs. Health Educ Res 2009, 24(2):292305. 32.

Ostroff C, Kinicki AJ, Tamkins MM: Organizational Culture and Climate. In Handbook of Psychology. 2003:565-593.

33. Driessen MT, Groenewoud K, Proper KI, Anema JR, Bongers PM, van der Beek AJ: What are possible barriers and facilitators to implementation of a Participatory Ergonomics programme? Implement Sci 2010, 5:64. 34. Weiner BJ, Belden CM, Bergmire DM, Johnston M: The meaning and measurement of implementation climate. Implementation Science 2011, 6(1). 35. Klein KJ, Sorra JS: The challenge of innovation implementation. Acad Manag Rev 1996, 21:1055-1080. 36. Schein EH: Organizational culture and leadership: San Francisco: Jossey-Bass Publishers; 1985. 37. Cameron KS, Quinn RE: Diagnosing and changing organizational culture: Addison-Wesley; 1999. 38. Quinn RE, Rohrbaugh J: Spatial model of effectiveness criteria: towards a competing values approach to organizational analysis. Management Science 1983, 29(3):363-377. 39. Cameron KS, Quinn RE, Robert E: Diagnosing and changing organizational culture: Based on the Competing Values Framework: San Fransico: Jossey-Bass; 2006. 40. Sparling PB: Worksite health promotion: principles, resources, and challenges. Prev Chronic Dis 2010, 7(1):A25.

171

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

summary samenvatting dankwoord about the author

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Summary

Summary An unhealthy lifestyle among employees has a major financial impact on organizations in terms of productivity loss and sickness absence rates. Additionally, the prevalence of an unhealthy lifestyle among the Dutch population is high. Therefore, numerous worksite health promotion programs (WHHPâ&#x20AC;&#x2122;s) have been developed and studied intensively on their effectiveness over the past two decades. However, implementation of WHPPâ&#x20AC;&#x2122;s in daily practice has been limited. Partly because of the lack of high quality process or formative evaluations that may improve our understanding of factors that can contribute to the implementation process of such programs. Hence, the focus of this dissertation was to systematically evaluate the implementation process and to gain insight into implementation factors that contribute to the adoption and uptake of a WHPP in practice. A specific focus was put on the use of an already existing 7-step implementation strategy by two participating worksites of different organizations as a means to improve the quality of the implementation and effectiveness of a WHPP. The 7-step implementation strategy was aimed to ensure the: 1) creation of solid support, 2) formation of a project structure, 3) performance of a needs assessment, 4) development of tailored evidence-based interventions, 5) adequate implementation of WHPP interventions, 6) evaluation of the intervention, and 7) the maintenance of the intervention. The two main objectives of this thesis were: 1) To identify implementation determinants (i.e. barriers an facilitators) that either hamper or facilitate the implementation of WHPPâ&#x20AC;&#x2122;s focusing on healthy lifestyle changes, and; 2) To assess whether the use of the 7-step strategy contributed to the successful development, implementation and maintenance of a WHPP aimed at stimulating a healthy lifestyle change among employees. In order to answer these objectives, we proposed a new conceptual framework for conducting a formative evaluation in Chapter 2. In this chapter we outlined how this framework was used for the evaluation of the use of the 7-step strategy in this study and a detailed description of the background and content of the 7-step strategy was provided. The proposed conceptual framework was developed based on literature on process evaluations. The final framework as applied in this thesis consisted of eight components: 1) context, 2)

175

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Summary

recruitment, 3) reach, 4) dose delivered, 5) dose received, 6) fidelity, 7) satisfaction and 8) maintenance. Each of these component measure a different aspect of the adoption, implementation and continuation of a WHP program. An advantage of the proposed framework is that it takes into account which factors could positively or negatively influence the implementation process. Another plus is that for each component different levels on which the component can act were described, which included the organizational level, implementer level and employee level. Chapter 3 contains a systematic review that summarizes and critically appraises the quality of studies that conducted an effect evaluation as well as a process evaluation for worksite health promotion programs focusing on healthy lifestyle changes and measured implementation determinants. The main results were that, of the 307 identified effect evaluations, 22 (7.2%) published an additional process evaluation. Only eight of these studies based their process evaluation on a theoretical framework. The operationalization of process components varied between studies. Mostly due to the fact that different or no frameworks were used to guide the evaluation. The most frequently reported process components were dose delivered and dose received. Over 50 different implementation determinants were identified. The most frequently reported facilitator for implementation was strong management support. Lack of resources was the most frequently reporter barrier. Seven studies examined the link between implementation and program effectiveness. In general higher levels of implementation favorably affect study outcomes (i.e., effectiveness). In Chapter 4 we evaluated the implementation of multiple lifestyle interventions at the two worksites (i.e. hospital and university) by 1) studying the extent and quality of program delivery and maintenance, and 2) looking at employee recruitment, reach, participation and satisfaction levels. Additionally we evaluated the effectiveness of the interventions in a quasi-experimental controlled trial. A combination of quantitative and qualitative methods were used. Results showed that the WHPPâ&#x20AC;&#x2122;s were implemented partly as planned. 84.9% of the planned interventions were delivered, but these were in general relatively simple and easily implemented interventions which did not require active employee participation (such as distribution of free fruit and posters). Most interventions chosen were environmental changes and educational interventions that could be implemented easily at low costs and effort. Employees showed high reach (96.6%) and overall participation (75.1% participated in at least one intervention). Overall employees were moderately satisfied with the interventions(6.8 Âą 1.1), mainly because they perceived the programme to be a good initiative which showed that their employer acknowledged the value of employee health

176

Summary

and lifestyles. Additionally, most interventions were perceived by the employees to be “pragmatic” and thought to facilitate employees in obtaining a healthy lifestyle. On a more critical note, the long interval between some interventions reduced employee’s interest in the project. The effect evaluation showed a significant intervention effect for days of fruit consumption (β: 0.44 days/week, 95%CI 0.02 to 0.85). In Chapter 5, we explored which implementation determinants were associated with employee participation and satisfaction levels. Thirteen implementation determinants derived from literature were measured during (T1) and after (T2) implementation of the WHPP using a questionnaire. Using linear and logistic regression analyses we found that a positive attitude towards program implementation at T1 (OR 2.06) was best associated with high participation at T2. The ‘programs match with employee’s needs’ (β: 0.22), ‘positive attitude towards program implementation’ (β: 0.42), and ‘positive attitude towards employer involvement’ (β: 0.20) at T1 was associated best with overall satisfaction at T2. Whereas the ‘programs fit with the organization’ (β: 0.27), ‘positive attitude towards program implementation’ (β: 0.32), ‘positive program image’ (β: 0.24), and ‘program notification to new employees’ (β: 0.13) at T2 were best associated with higher overall satisfaction levels. So in order too potentially increase the success of a program, implementers should ensure that the program fits employees needs and that employees have a favourable attitude towards program implementation. Chapter 6 presented the actual use and adherence of both worksites to the 7-step strategy. Strategy adherence was assessed with twelve performance indicators (PIs) that reflect the use of the 7-step strategy. Scores on each PI were calculated based on the data collected by onsite monitoring and thesemi-structured interviews at baseline and followup (6, 12 and 18 months). We calculated an overall performance indicator for the whole strategy based on the average of all individual PIs, for which a higher overall performance score corresponded with higher strategy adherence. The overall performance indicator for the University was 0.76 and for the Hospital 0.54 both representing partial strategy adherence. Overall, lowest scores were found for creating support among middle and lower management, formation of project structure, performing a needs assessment, develop a project plan, and maintenance. The results indicated that strategy adherence facilitated a structured development and implementation. When improvements (as described in chapter 7) will be made to the content and performance of the 7-step strategy, it could be an effective tool to successfully implement a multi-component WHPP.

177

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Summary

Chapter 7, the general discussion, starts with presenting the main findings in the light of the objectives of this thesis, and is followed by the methodological contributions and issues that should be acknowledged when interpreting the findings. An improved 7-step strategy is presented based on the lessons learned in this thesis. Recommendations and implications for future researchers and practitioners are also addressed in this chapter. Overall, this thesis showed that although the use of the 7-step strategy helps overcome implementation determinants, the strategy only partly led to the successful development, implementation and maintenance of a WHPP. The strategy was not able to induce the implementation and maintenance of more comprehensive evidence-based interventions in the current form. The findings as described in this thesis make clear that, although the 7-step strategy was perceived as a useful tool to systematically implement lifestyle interventions in an organization, improvements need to be added before nationwide use can be recommended. For example, the strategy should contain more (practical) materials to substantiate each step. In order to manage expectations, the strategy can be optimized by specifying a timeline for successful execution of each step. As is, the strategy does not account for all possible implementation determinants. A specific questionnaire addressing these determinants for project members can help to anticipate and steer on possible barriers in the implementation. Finally, the strategy should have a bigger focus on repeating the steps to make sure that the policy and interventions are still up to date and that vitality is an ongoing topic which should be addressed regularly. Specific improvements to each of the seven steps are also described in Chapter 7.

178

Samenvatting

Samenvatting Een ongezonde leefstijl van werknemers heeft grote financiële gevolgen voor organisaties in termen van productiviteitsverlies en ziekteverzuim. Daarnaast is de prevalentie van een ongezonde leefstijl onder de Nederlandse bevolking hoog. Daarom zijn er de laatste 20 jaar al tal van gezondheidsbevorderende programma’s op de werkplek (GBW-programma’s) ontwikkeld en onderzocht op hun effectiviteit. Echter valt implementatie van dit soort programma’s in de dagelijkse praktijk tegen. Dit komt mede door het ontbreken van kwalitatief goede procesevaluaties die ons begrip van factoren die het implementatieproces kunnen beïnvloeden, kunnen versterken. Daarom lag de focus van dit proefschrift op het systematisch evalueren van het implementatieproces, om zo inzicht te krijgen in factoren die bijdragen aan de ontwikkeling en implementatie van GBW-programma’s op de werkplek in de dagelijks praktijk. Specifiek lag de focus op het gebruik van een bestaande 7-stappen strategie binnen twee verschillende organisaties als middel ter verbetering van de implementatie om zo de effectiviteit van de interventie te vergroten. De strategie bestaat uit de volgende zeven stappen: 1) creëren van draagvlak, 2) opzetten van een projectstructuur, 3) doen van een behoeftepeiling, 4) ontwikkeling van bewezen effectieve interventies, 5) implementatie van het programma, 6) evaluatie en 7) borging van het programma. De belangrijkste doelen van dit proefschrift waren: 1) het

identificeren

van

determinanten

die

implementatie

van

gezondheidsbevorderende leefstijlinterventies op de werkplek bevorderen dan wel belemmeren en; 2) om te beoordelen of het gebruik van de 7-stappen strategie heeft bijgedragen aan succesvolle ontwikkeling, implementatie en borging van GBW-programma gericht op leefstijl veranderingen onder medewerkers. Om deze doelen te beantwoorden, hebben wij in hoofdstuk 2 een nieuw conceptueel kader voor het uitvoeren van een procesevaluatie voorgesteld. In dit hoofdstuk wordt uiteengezet hoe dit kader werd toegepast voor de evaluatie van het gebruik van de 7-stappen strategie in dit onderzoek. Ook is er een gedetailleerde beschrijving van de achtergrond en de inhoud van de 7-stappen strategie verstrekt. Het voorgestelde conceptuele kader werd ontwikkeld op basis van literatuur over procesevaluaties. Het uiteindelijke kader, zoals toegepast in dit proefschrift, bestaat uit acht componenten: 1) context, 2) recruitment (werking), 3) reach

179

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Samenvatting

(bereik), 4) dose delivered (geïmplementeerde interventies), 5) dose received (participatie aan de interventies), 6) fidelity (strategie trouw), 7) tevredenheid en 8) borging. Bij elk van deze componenten wordt een ander aspect gemeten van adoptie, implementatie en borging van een GBW-programma. Een voordeel van het voorgestelde conceptueel kader was dat het rekening houdt met factoren die het implementatieproces positief of negatief kunnen beïnvloeden (de implementatie determinanten). Een ander pluspunt was dat elke component op verschillende niveaus kan acteren, namelijk: op organisatie, implementeerder en werknemer niveau. Hoofdstuk 3 presenteert de resultaten van een systematische review met daarin een overzicht en kritische beoordeling van studies die zowel een effectevaluatie als procesevaluatie voor een GBW-programma gericht op leefstijl hebben uitgevoerd. Hieruit konden we opmaken dat 22 (7,2%) van de 307 geïdentificeerde effectevaluaties tevens een procesevaluatie hadden gepubliceerd. Slechts acht van deze studies baseerden hun procesevaluatie op een bekend conceptueel kader. Hierdoor varieerde de operationalisering van de gemeten procescomponenten tussen de studies. De procescomponenten waar het vaakst op gerapporteerd werden zijn dose delivered (geïmplementeerde interventies) en dose received (participatie aan de interventies). In totaal werden er meer dan 50 verschillende determinanten voor implementatie geïdentificeerd. De meest frequent gemelde bevorderende factor voor implementatie was management support. Een gebrek aan middelen was de meest frequent gerapporteerde belemmerende factor. Zeven studies onderzochten het verband tussen implementatie en de effectiviteit van het programma. In het algemeen kwam hieruit dat wanneer de implementatie goed en volledig is, dit de effectiviteit van de studie gunstig beïnvloed. Hoofdstuk 4 gaat over de mate waarin de verschillende leefstijlinterventies zijn geïmplementeerd binnen beide werkplekken (ziekenhuis en universiteit) door 1) het bestuderen van de kwaliteit van interventie levering en borging, en 2) te kijken naar de mate van werving, bereik, participatie en tevredenheid van medewerkers. Daarnaast is de effectiviteit van de interventies onderzocht in een quasi-experimenteel gecontroleerd onderzoek. Er is gebruik gemaakt van een combinatie van kwantitatieve en kwalitatieve methoden. De resultaten toonden aan dat de interventies deels uitgevoerd waren zoals gepland. 84,9% van de geplande interventies werden daadwerkelijk geïmplementeerd. Dit waren echter over het algemeen relatief eenvoudig en gemakkelijk te implementeren interventies, welke geen actieve participatie van werknemers behoeften (zoals het verstrekken van gratis fruit en posters om traplopen te stimuleren). De meest gekozen

180

Samenvatting

interventies waren dan ook omgevingsinterventies en educatieve interventies die gemakkelijk en tegen lage kosten en inspanningen geïmplementeerd konden worden. Over het algemeen was het bereik (96,6%) en de totale participatie (75,1% deelgenomen aan ten minste één interventie) goed. Werknemers waren gemiddeld tevreden met de interventies (gemiddelde score 6,8 ± 1,1), vooral omdat ze het programma een goed initiatief vonden omdat het liet zien dat hun werkgever de leefstijl en gezondheid van zijn werknemers belangrijk vond. Bovendien vonden werknemers de meeste interventies praktisch en gericht op het faciliteren van een gezonde leefstijl. Een meer kritische noot was dat het lange tijdsbestek tussen sommige interventies de belangstelling in het programma verminderde. De effectevaluatie liet een significante verandering zien voor aantal dagen fruitconsumptie (β: 0,44 dagen/week, 95% CI 0,02-0,85). Hoofdstuk 5 beschrijft welke implementatie determinanten geassocieerd zijn met participatie- en tevredenheidniveaus van werknemers. Dertien implementatie determinanten afkomstig uit de literatuur zijn tijdens (T1) en na afloop (T2) van het GBWprogramma gemeten met een vragenlijst. Lineaire en logistische regressie analyse lieten zien dat “een positieve houding ten opzichte van de implementatie van het programma” op T1 (OR 2,06) positief geassocieerd was met een hoge deelname aan het programma op T2. “Aansluiting van het programma bij de behoeften van de medewerkers” (β: 0,22), “een positieve houding ten opzichte van de implementatie van het programma”(β: 0,42), en “een positieve houding ten opzichte van de betrokkenheid van de werkgever op het gebied van leefstijl” (β: 0,20) op T1waren positief geassocieerd met tevredenheid op T2. Verder waren “het aansluiten van het programma bij de organisatie” (β: 0,27), “een positieve houding ten opzichte van de implementatie van het programma” (β: 0,32), “een positief programma imago” (β: 0,24), en”aankondiging van het programma richting nieuwe medewerkers” (β: 0.13) op T2 geassocieerd met een hogere algemene tevredenheid op T2. Om het succes van een programma te verhogen, is het dus belangrijk dat implementeerders ervoor zorgen dat het programma aansluit bij de behoeften van werknemers en dat werknemers een positieve attitude hebben voor de implementatie van het programma. In hoofdstuk 6 wordt het feitelijke gebruik en de naleving van de stappen van de 7-stappen strategie binnen beide organisaties geëvalueerd. Het naleven van de strategie werd beoordeeld met twaalf prestatie-indicatoren (PI’s) die het gebruik van de strategie reflecteren. De score voor elke PI werd berekend op basis van data verkregen door monitoring en semigestructureerde interviews bij de 0-meting en de vervolgmetingen na 6, 12 en 18 maanden. Er is een overkoepelende PI voor de hele strategie berekend op

181

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Samenvatting

basis van het gemiddelde van de individuele PIâ&#x20AC;&#x2122;s waarbij een hogere score overeen kwam met een betere naleving van de stappen. Voor de universiteit was deze 0,76 en voor het ziekenhuis 0,54, wat betekend dat beide gemiddeld scoorden ten aanzien van strategie gebruik. De laagste scores werden gevonden voor het creĂŤren van draagvlak onder het midden- en lagere management, de vorming van een projectstructuur, het uitvoeren van een behoeftepeiling, het ontwikkelen van een projectplan, en borging. De resultaten gaven aan dat naleving van de strategie een gestructureerde ontwikkeling en implementatie van een GBW-programma faciliteert. Wanneer de verbeteringen ten aanzien van de inhoud en prestatie van de 7-stappen strategie (zoals beschreven in hoofdstuk 7) zullen worden doorgevoerd, kan de strategie een doeltreffend instrument zijn om succesvol een multicomponent GBW-programma te implementeren binnen organisaties. In hoofdstuk 7, de algemene discussie, zijn de belangrijkste resultaten uit het proefschrift beschreven in het licht van de doelstellingen, gevolgd door de methodologische bijdragen en kanttekeningen die nodig zijn voor de interpretatie van de resultaten. Ook wordt er een verbeterde 7-stappen strategie gepresenteerd op basis van de geleerde lessen uit dit proefschrift. Aanbevelingen en implicaties voor toekomstig onderzoek en praktijkmensen worden tevens besproken in dit hoofdstuk. Dit proefschrift toonde aan dat, hoewel het gebruik van de 7-stappen strategie helpt in het overwinnen van belemmerende factoren voor implementatie, de strategie slechts gedeeltelijk heeft geleid tot succesvolle ontwikkeling, implementatie en borging van een GBW-programma. In de huidige vorm was de strategie niet in staat om de implementatie en borging van meer uitgebreide wetenschappelijk onderbouwde interventies te induceren. De bevindingen zoals beschreven in dit proefschrift maken duidelijk dat, hoewel de 7-stappen strategie werd gezien als een nuttig instrument om systematisch leefstijlinterventies in een organisatie te implementeren, er verbeteringen moeten worden doorgevoerd voordat de strategie landelijk aanbevolen kan worden. De strategie moet bijvoorbeeld meer (praktische) materialen bevatten om elke stap te onderbouwen en goed uit te kunnen voeren. Door een optimale tijdlijn toe te voegen aan de strategie kunnen verwachtingen gemanaged worden. Verder kan het toevoegen van een vragenlijst om implementatie determinanten in kaart te brengen, bijdragen aan het anticiperen en sturen op mogelijke belemmerende factoren bij implementatie. Tot slot moet de strategie een grotere focus hebben op het belang van het herhalen van de stappen om ervoor te zorgen dat het beleid en de interventies up to date blijven en dat vitaliteit een onderwerp is dat regelmatig geadresseerd wordt. Specifieke verbeteringen voor elk van de zeven stappen worden ook beschreven in hoofdstuk 7.

182

Dankwoord

Dankwoord Eindelijk is het zover, na de nodige hoogte en dieptepunten is mijn proefschrift dan echt af! Iedereen die mij kent weet dat met name de laatste loodjes een uitdaging zijn geweest naast mijn nieuwe baan. Ik ben dan ook enorm trots dat er nu een proefschrift ligt waar ik helemaal achter sta, en dat was zeker niet gelukt zonder de steun van vele mensen in de verschillende fasen van het onderzoek. Voordat ik iemand vergeet wil ik allereerst iedereen enorm bedanken! Zonder jullie was het niet gelukt. Natuurlijk begin ik met het bedanken van mijn begeleiders. Luuk, Pepijn en Willem, bedankt voor jullie steun, vertrouwen, input en alle (wetenschappelijke) kennis die ik, grotendeels dankzij jullie, heb opgedaan de afgelopen jaren. Ik was zeker niet de makkelijkste promovendus maar ondanks alles hebben we nu wel met zijn vieren de eindstreep gehaald. Luuk, mijn dankwoord is al een week bijna af en nog steeds kan ik niet de juiste woorden vinden om je te bedanken, maar ik ga het toch proberen. Wat heb ik een geluk gehad dat jij mijn dagelijkse begeleider en copromotor was. Maar je was veel meer dan dat. Je stond altijd voor me klaar, luisterde wanneer dat nodig was en kreeg me altijd weer gemotiveerd om, ondanks kritiek en de nodige “promotietranen”, weer verder te gaan. Je bent een echte mentor voor me geweest en ik heb zowel inhoudelijk als persoonlijk veel van je geleerd. Je praktische manier van denken was zeer welkom en maakt dat ik gegroeid ben ik het vertalen van wetenschap naar praktijk. Je maakte altijd tijd om stukken zo snel mogelijk te lezen en te reageren op een manier dat ik er ook weer mee verder kon. En belangrijker nog, je zorgde dat er ook voldoende ruimte was voor gezelligheid en persoonlijke dingen tijdens de vele overleggen. Van een meisje net klaar met studeren, zoekende naar wat qua werk bij mij zou passen, ben ik nu 5,5 jaar later bijna gepromoveerd en heb ik een baan waar ik heel erg enthousiast over ben. Ik had hier niet gestaan als jij er niet was geweest. Ik ben je enorm dankbaar voor de fijne samenwerking en ik hoop echt dat we deze ooit in de toekomst voort kunnen zetten. Pepijn, het was op zijn zachts gezegd een turbulente samenwerking waarbij we het vaak oneens waren over de inhoud/richting van artikelen. Ik heb heel erg moeten wennen aan de kritische blikken en opmerkingen die je altijd op mijn artikelen wierp. Maar inmiddels ben ik je hier heel erg dankbaar voor. Niet alleen is de inhoud en kwaliteit van mijn proefschrift duidelijk verbeterd dankzij jouw kennis en scherpte, maar ook heb ik persoonlijk veel geleerd van onze samenwerking. Je daagde me uit om kritisch over mijn eigen ideeën en keuzes na te denken en deze altijd te onderbouwen met goede argumenten. Na de soms

183

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Dankwoord

heet oplopende maar altijd leerzame discussies, zijn we er gelukkig altijd goed uitgekomen. Daarnaast waardeer ik het enorm dat je ondanks je overvolle agenda, toch altijd mijn stukken las en van commentaar voorzag. En hebben we ook gezellige momenten gehad waaronder het congres in Gent waar de biertjes toch wel erg goed smaakten. Willem, ongeacht je overvolle agenda heb je altijd tijd gemaakt om, op meestal onmogelijke tijdstippen, zowel inhoudelijk als taalkundig (ik probeer echt om kortere zinnen te maken) feedback te geven om het proefschrift naar een hoger niveau te tillen. Anders dan bij je andere promovendi, zat ik eigenlijk altijd bij TNO waardoor we elkaar wellicht minder vaak hebben gezien, maar ik waardeer je input. Graag bedank ik ook de andere auteurs en in het specifiek Vincent Hildebrandt en Saskia Duijts. Vincent, zonder jouw was dit promotietraject niet mogelijk geweest. Jij bent met het idee voor dit traject gekomen en hebt dit uitgewerkt in een projectvoorstel wat uiteindelijk ook echt gehonoreerd is door ZonMw. Zonder subsidie geen traject en dus ben ik je hier dankbaar voor. Je begon in 2010 samen met Luuk als mijn copromotor maar gezien de richting waar het onderzoek heen ging werd er gezamenlijk besloten dat de kennis van jullie beide teveel op één vlak zat en dus kwam Pepijn erbij. Ik wil je heel erg bedanken voor je bijdrage in het begin en je input voor de eerste twee artikelen. Saskia, jij raakte betrokken bij mijn onderzoek toen je net kwam werken bij de afdeling op de VU en zou samen met mij het review vorm gaan geven. Ook jouw wil ik heel erg bedanken en ik heb je input op het review enorm gewaardeerd. Tijdens mijn promotietraject heb ik ook twee stagiaires mogen begeleiden. Kathelijne de Moes (inmiddels Kathelijne Hazelnoot want in de periode na je stage ging je trouwen) en Maartje Pouw. Kathelijne, bedankt voor je harde werk en input in de proces evaluatie voor werknemers. Uiteindelijk gingen we toch een iets andere richting op met het artikel, maar heeft het wel geleid tot een mooie publicatie. Maartje, ook jij heel erg bedankt voor je inzet en doorzettingsvermogen, ik heb nog steeds de ansichtkaart bewaard die je mij aan het einde van je stage hebt gegeven. Jij hebt tijdens mijn promotietraject een mooi stuk geschreven genaamd “exploring the role of organizational culture on the implementation of a worksite health promotion program”. Helaas, leidde het niet tot een publicatie, maar heb ik de uitkomsten heel erg goed kunnen gebruiken in de general discussion van dit proefschrift. Zonder organisaties en deelnemers had het erg lastig geworden om dit onderzoek uit te voeren. Ik wil dan ook alle deelnemers van het UMC Utrecht divisie Vrouw en Baby en Hogeschool Utrecht Faculteit Gezondheidszorg bedanken voor deelname aan de

184

Dankwoord

interventies en het onderzoek. Speciale dank gaat uit naar de betrokken werknemers in de projectgroepen van beide organisaties. Jullie waren verantwoordelijk voor het doorlopen van de 7-stap strategie. Zonder jullie was het niet mogelijk geweest dit onderzoek uit te voeren. De organisaties waren belangrijk, maar Pieter, jij was net zo belangrijk. Vanuit het NISB (het Nederlands Instituut voor Sport & Bewegen) was jij de externe adviseur die betrokken was bij het project. Jullie ervaring met het BRAVO kompas bracht waardevolle kennis met zich mee. Ook toen je de stap maakte naar “bewegen werkt” bleef je betrokken bij mijn traject. Omdat wij als onderzoekers ons niet mochten bemoeien met de loop van het project zou jij dit voor ons doen. Bedankt voor je tomeloze, enthousiaste en positieve bijdrage. Je kwam bij alle overleggen op je vouwfiets aangesjeesd en hebt ervoor gezorgd dat de projecten ook daadwerkelijk van de grond kwamen binnen de organisaties. Niet voor niks is één van de belangrijkste succesfactoren de aanwezigheid van een externe adviseur! Leden van de leescommissie, prof. A. Burdorf, prof. W. van Rhenen, prof. J.R. Anema, prof. G. Zwetsloot, dr. L. Vaandrager en dr. M. Westerman, hartelijk dank voor de tijd en energie die jullie gestoken hebben in het lezen en beoordelen van mijn proefschrift. Ik kijk uit naar jullie interessante en uitdagende vragen tijdens mijn verdediging. Mijn paranimfen, Ellen en Linda, wat fijn dat jullie twee naast mij op het podium staan op deze voor mij heel speciale dag! Ellen, vanaf het begin van mijn promotietraject heb je altijd de moeite gedaan om echt te begrijpen waar ik mee bezig was ook al zit je als belastingadviseur in een hele andere wereld. Toen mijn eerste artikel gepubliceerd was, vroeg jij gelijk aan mij waar je hele stuk kon lezen en dat betekent nog steeds heel veel. Op het moment dat mijn proefschrift eindelijk echt was goedgekeurd stond je met een lekkere fles wijn voor de deur, die al 2 jaar in de kast stond want we hadden toch echt wel gedacht en ik gehoopt dat ik eerder klaar was. Ik ben blij met je vriendschap en waardeer je steun ontzettend. Dank je wel voor alle gezelligheid, gesprekken, borrels en leuke feestjes! Linda, vanaf het begin van mijn traject zijn we kamergenootjes geweest. Eerst op de grote kamer, later een kamer met zijn tweeën. Je was de beste kamergenoot die ik me kon wensen en ik blij dat we onze vriendschap nog steeds in stand houden. Je was denk ik toch wel de persoon die ik het meest heb gesproken en meegemaakt hebt tijdens mijn promotietraject. Je stond altijd voor me klaar om te luisteren naar van alles en nog wat. Of het nou om privé of werk dingen ging, positief of negatief, het maakte niet uit. Bedankt voor de gezelligheid en het meedenken met o.a. statistische vraagstukken. Ik vond het een eer

185

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Dankwoord

om jouw paranimf te zijn en ik had het niet anders gewild dan dat jij ook op mijn dag een speciale rol hebt. TNO collegae, dank jullie wel voor de gezelligheid in Leiden en voor het feit dat ik me als onofficiële TNO-er zo welkom heb gevoeld. VU-collega’s en specifiek de GH0-gang, ik was niet vaak op de VU, maar als ik er was zorgden jullie ervoor dat ik gelijk bij de GH0-familie hoorde. Het was fijn om op dat soort momenten frustraties te delen en te zien dat ik niet de enige was die ergens tegenaan liep. Esther wil ik graag specifiek bedanken voor de uren die we al kletsend en brainstormend op de kamer hebben doorgebracht. Ook stonden we samen met Femke aan de start van de implementatieclub. ArboNed collegae, inmiddels werk ik hier alweer bijna 2 jaar en heb ik mijn proefschrift afgeschreven terwijl ik hier was. Dat betekend ook dat ik sommige van jullie daar lastig mee heb gevallen. Bedankt voor jullie luisterend oor en de support die ik op allerlei vlakken van alle collega’s, ook buiten het kenniscentrum om, heb gekregen. Lieve vrienden en vriendinnen, ik ga niet iedereen los opschrijven, jullie weten wie ik bedoel. Alle gezellige etentjes, borrels, festivals, feestjes, filmavonden, strandbezoekjes en andere uitstapjes hebben de afgelopen periode onvergetelijk leuk gemaakt. Kleine successen moet je vieren in het leven, bedankt dat jullie er waren om dat te doen als er weer een artikel af, geaccepteerd of gepubliceerd was. Lisa, ik wil je heel erg bedanken voor het tegen lezen van meerdere hoofdstukken in dit proefschrift om zo de Engelse taal te verbeteren. Koen had geen betere vriendin kunnen kiezen ;) Tot slot wil ik natuurlijk mijn ouders bedanken voor jullie onvoorwaardelijke steun en vertrouwen. Het is fijn om te weten dat jullie altijd in mij geloven!

186

About the author

About the author Debbie Wierenga was born on December 24th 1987 in Zaandam, the Netherlands. After completing secondary school (VWO) at St. Michael College in Zaandam in 2006, she started her bachelor education in Biomedical Sciences at the VU University Amsterdam. After she graduated as a Bachelor in Science in 2009, she started a master Public Health Research, with a specialization in Prevention and Public Health. Directly after receiving her Masters degree in July 2010, she started her PhD project (i.e., the BRAVO@Work study) at the EMGO+ Institute within the department of Public and Occupational Health and TNO (Leiden) on the implementation of multi-component lifestyle interventions at two organizations using a 7-step strategy to facilitate structured implementation. The results of this study are presented in this thesis. In 2014, Debbie completed her master’s degree in Epidemiology after following the Postgraduate Epidemiology Program at the VU University Medical Center in Amsterdam and became an epidemiologist. Since 2014 Debbie is working as a (research) consultant at ArboNed where she gives advice to organizations on how to improve overall health and vitality of employees. Based on the results of this thesis she developed a practical strategy (“het Vitaliteitsvliegwiel”) which is now widely used within ArboNed to help organizations implement effective programs that focus on all aspects of a vital organization (i.e., sickness absence, physical and mental health, motivation and occupational hazards). Contact: debbie.wierenga@arboned.nl debbiewierenga@gmail.com

A 187

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

BRAVO@Work

BRAVO@Work The evaluation of the implementation process of a worksite health promotion program

Uitnodiging Voor het bijwonen van de openbare verdediging van mijn proefschrift

Debbie Wierenga Fluitekruidweg 116 1508AM Zaandam 0655738333 debbie_wierenga@hotmail.com debbie.wierenga@arboned.nl

Debbie Wierenga

Paranimfen â&#x20AC;˘ Ellen Breed

Debbie Wierenga

0643246193 e.e.breed@gmail.com

â&#x20AC;˘ Linda Koopmans

0648032329 lindataylorkoopmans@outlook.com

Body@Work