Implementation of the Swedish Guideline for Prevention of Mental ill-health at the Workplace: study protocol of a cluster randomized controlled trial, using multifaceted implementation strategies in schools

Globally, common mental disorders (CMDs), such as depression, anxiety and adjustment disorders, are highly prevalent [1]. The Organization for Economic Cooperation and Development (OECD) has estimated that approximately 20% of the working-age population is living with a CMD at any given time [2]. The consequences of the high prevalence of CMDs for employers are enormous, including high rates of sickness absence, significant losses in productivity at work and individual suffering [2]. A professional group that is especially vulnerable for CMDs are schoolteachers [3‐5]. Teachers have been identified as being the professional group with the highest risk of poor mental health compared to other professions [6]. Evidence shows that, in addition to personal factors, social and organizational risk factors at the workplace, such as high demands at work, low job-control and low support from colleagues and superiors increase the risk for CMDs [7, 8]. The systematic management of social and organizational risk factors at the workplace is therefore commonly recommended as a strategy to prevent CMDs [9, 10]. The social and organizational work environment of teachers is characterized by high workload/work intensity, role overload, increased class size per teacher, unacceptable pupils’ behavior, bad school management and/or lack of support from management, resulting in a high risk for mental ill-health and consequently sick leave [11].

Few workplaces to date, however, systematically modify or eliminate sources of social or organizational risks inherent in the work environment. When it comes to schools, a recent report by the Swedish Work Environment Authority showed that most Swedish schools had severe shortages in their social and organizational risk management, including a lack of assessment of unhealthy work demands [12]. Rather, the focus is often primarily on remedial measures for employees affected by CMDs [13, 14]. Even though evidence suggests favorable effects at both individual and organizational level of implementing organizational level interventions aimed at addressing the problems at source [15]. In order to support workplaces with the management of their social and organizational work environment and prevent CMDs, we launched the first Swedish occupational health guideline to support the prevention and treatment of work-related mental ill-health at the workplace [16]. The guideline is based on the best available evidence [e.g. [17‐19]] and has been compiled in a cross-disciplinary collaboration between employers, occupational health service practitioners and researchers.

The mere existence of guidelines does, however, not guarantee their usage [20]. Efforts to disseminate guidelines via e.g. websites, publication of consensus statements, or mass mailings often result in relatively low adoption, resulting in only small changes in practice. There is therefore a clear need for well-designed studies that aim to facilitate the implementation of guidelines into practice, including occupational health guidelines aimed at the prevention of CMDs. Various models demonstrate that multiple contextual factors, so-called barriers and facilitators, affect how efficiently guidelines and recommendations are implemented. These factors can be related to characteristics of the guidelines themselves (e.g. complexity), characteristics of the users (e.g. knowledge), the internal environment (e.g. the organization’s capacity for change) and the external environment (e.g. existence of national policies) [21]. In 2012, Eurofond and the European Agency for Safety and Health at Work (EU-OSHA) undertook a European assessment of barriers and facilitators for managing social and organizational risks at the workplace [22, 23]. According to their findings and supported by others [24‐26], factors that can hinder the management of these risks were primarily related to characteristics of the users (insufficient education and knowledge) and to the internal environment (lack of technical support and guidance, as well as lack of resources) [22‐26].

In order to facilitate guideline implementation and adherence to recommendations, implementation strategies are required that target identified barriers and facilitators [27]. To date, there is no consensus on which implementation strategies are most effective in facilitating guideline implementation. It is, however, recognized that active strategies are needed to improve guideline adherence [28]. Research has also shown that multifaceted strategies appear to be more effective than single strategies [29], because multiple implementation barriers are targeted at the same time. However, previous studies examining the effectiveness of multifaceted strategies are inconclusive [28, 30, 31]. One reason for this is that the multifaceted strategies evaluated often were aimed at only one barrier (e.g. lack of knowledge) and therefore not successful in changing behavior [28]. Behavior change is more likely to occur if implementation strategies are targeting different types of barriers simultaneously (e.g. lack of knowledge and lack of organizational support). Another limitation of previous studies evaluating implementation strategies’ effectiveness is their lack of a clear theory-base [32]. A recent review on the use of theory (or models and frameworks) to plan or evaluate guideline implementation concluded that only half of the guideline implementation studies were based on theory, models or conceptual frameworks and only few provided details about how they were used [33]. In this study, we employ Michie and colleagues’ COM-B model to inform the design of implementation strategies [34]. COM-B is used to identify which components of the behavior system (motivation, capability or opportunity) need to be changed to achieve a behavior change [34]. The COM-B model posits that behavior is a function of three components: Capability (C), Opportunity (O), and Motivation (M). Capability refers to the ability to engage in the cognitive or physical processes necessary for the behavior, e.g. knowledge and skills. Motivation refers to those brain processes that direct behavior, and include reflective and automatic motivation, e.g. analytical decision-making and emotional responses. Opportunity refers to those factors that lie outside the individual and influence behavior, e.g. social support and prompts [34]. Implementation strategies are chosen based on Powell and colleagues’ refined compilation of implementation strategies. The compilation has been recommended as a guide to facilitate the development of multifaceted, multilevel implementation strategies that are tailored to local contexts [27]. To evaluate guideline implementation, we use Proctor and colleagues’ conceptual framework for dissemination and implementation [35]. Proctor’s framework defines evidence-based intervention strategies (e.g. guidelines) and separate strategies for implementing those intervention strategies into practice. Moreover, the framework defines different levels of change that an intervention is addressing: the larger system or environment, the organization, a group or team, or the individual. Finally, we will use the Consolidated Framework for Implementation Research (CFIR) [21] to identify barriers and facilitators that may influence the implementation process. CFIR is a meta-theoretical framework that consists of five domains (the intervention, inner setting, outer setting, the individuals involved and the implementation process) that can be used to identify barriers and facilitators [21].

A randomized controlled trial with two arms will be conducted comparing an implementation intervention with a single strategy versus one with multifaceted strategies (educational strategy, implementation teams and workshops). Information will be collected on both implementation effectiveness, and intervention (guideline) effectiveness (i.e. reduction of risk factors for mental ill-health). This will provide valuable knowledge on whether adherence to the recommendations in the guideline will result in reductions in employees’ perceived social and organizational risk factors, mental ill-health and sick-leave. The programme theory is depicted in Fig. 1. The trial will be conducted within schools with the aim to prevent mental ill-health and sick leave among teachers. Globally, teachers have a high prevalence of CMDs and related sick-leave [37‐39].

The hypothesis is that schools that receive support in implementing the guideline through a multifaceted implementation strategy are more responsive to working in a structured and systematic manner with the management of social and organisational risks at their workplace and consequently reduce risk factors for mental ill-health, compared to schools that receive support through a single implementation strategy based on education.

The secondary objective is to evaluate the implementation process of the guideline within schools. This will be done by collecting qualitative and quantitative data on implementation outcomes as defined by Proctor and colleagues [35]. Moreover, contextual factors (barriers and facilitating factors) that may influence the implementation process will be identified. This is done by collecting qualitative and quantitative data on barriers and facilitators in accordance with the CFIR [21]. This will give a better understanding of factors that may influence the process and will inform future necessary adaptation of the implementation strategies.

The study is a 12-month cluster-randomized waiting list-controlled trial with before and after measurements involving 20 primary schools in two municipalities in Sweden. Randomization (1:1 allocation to intervention and waiting list control) is conducted at the school level. Primary schools are randomized by a computer-generated randomization-list to either the ARM 1 or ARM 2 after stratification in blocks by municipality and school size (number of students per school). Schools in ARM 1 will receive all three implementation strategies during the first year of the study period while schools in ARM 2 will receive only the educational strategy in year 1 and continue with the other strategies during year 2 (Fig. 2). The implementation strategies target the managerial level of first line managers i.e. school-principals. All employees of the participating schools will be invited to complete outcome-questionnaires at baseline and 6-, 12-, 18- and 24-months follow-up during working hours.

This trial has been registered in ClinicalTrials.​gov (NCT03322839, date assigned: August 2017). Moreover, this study has been approved by the Ethical Committee of Stockholm (2017/984–31/5).

It is not possible to modify the allocated implementation strategies for a given school. Individual participants can discontinue their participation during the study trial and have their results removed. Two persons from the research-team will deliver the intervention strategies to the schools, reminding each other to adhere to the protocol. Detailed schedules for the educational-meeting and for each workshop are developed to further improve adherence to the protocol. The schedule includes a description of the time line, different components (e.g. group work, presentation) and who is responsible for which components. The schedule includes a fidelity column where any deviations are noted. The research-team will be responsible for monitoring whether the implementation interventions are delivered according to the protocol. Procedures for monitoring adherence include making notes in the fidelity column when deviations occur, for example regarding time or content. Moreover, a research-log is kept describing deviations and possible reasons for deviations.

To reduce the risk of contamination between schools of ARM 1 and 2 within the same municipality, schools of ARM 1 are encouraged not to discuss the project with schools of ARM 2. Every 6 months during the trial period, the research team has a telephone meeting with the school principals to collect data regarding schools’ exposure to other interventions and potential contamination for control schools.

An overview of the measurement variables, method of data-collection, data-source and time-point for each measure is described in Table 4.

Figure 3 describes the time schedule of enrolment, implementation strategies and assessments.

Change from baseline in self-reported demands at work during 6-, 12-, 18- and 24- months follow-up is assessed with the Copenhagen Psychosocial Questionnaire (COPSOQ) II [42]. The COPSOQ II is a well-established instrument aimed at assessing the psychosocial work environment. COPSOQ has been widely used to assess teacher’s psychosocial work environment [39]. Three different types of demands at work are assessed: quantitative demands, work pace and emotional demands. Quantitative demands (e.g. Do you get behind with your work?) are assessed with four items, work pace is assessed with three items (e.g. Do you have to work very fast?) and emotional demands (e.g. Is your work emotionally demanding) are assessed with four items. All items have 5-point response anchors ranging from (1) “always” to (5) “never/hardly ever”.

Nine items of the COPSOQ II are used to assess three domains, namely social support from superior, social support from colleagues and recognition by management. Change from baseline social support from superior (e.g. How often do you talk with your superior about how well you carry out your work?) during 6-, 12-, 18- and 24- months follow-up is assessed with three items with 5-point response anchors ranging from (1) “always” to (5) “never/ never/hardly ever”. Change from baseline social support from colleagues (e.g. How often do you talk with your colleagues about how well you carry out your work?) during 6-, 12-, 18- and 24-months follow-up is assessed with three items with 5-point response anchors ranging from (1) “always” to (5) “never/hardly ever”. Change from baseline recognition by management (e.g. Is your work recognised and appreciated by the management?) during 6-, 12-, 18- and 24-months follow-up is assessed with three items with 5-point response anchors ranging from (1) “to a very large extend” to (5) “to a very small extend”.

Change in baseline presenteeism during 6-, 12-, 18- and 24-months follow-up is assessed with a single item developed and validated by Aronsson and colleagues [43]. Presenteeism is defined as coming to work despite health problems with answering categories “yes” or “no”.

Change in baseline work performance during 6-, 12-, 18- and 24-months follow-up is assessed by using a modified version of the Work Productivity and Impairment Scale General Health Questionnaire (WPAI: GH [44, 45]). The questions concern the extent to which the individual experiences that their work performance is affected by health problems and the extent to which the individual experiences that their work performance is affected by work-related problems. The response anchors range from “health problems/work-related problems have not influenced my work” (0) to “health problems/work-related problems have completely prevented me from working” (10).

Change in baseline recovery during 6-, 12-, 18- and 24-months follow-up is assessed with a single item [46] with 5-point response anchors ranging from (1) “never” to (5) “very often”.

Change in baseline work-life balance (e.g. my work drains so much of my energy that it has a negative effect on my private life) during 6-, 12-, 18- and 24- months follow-up is assessed with 4-items of the COPSOQ with 4-point response anchors ranging from (1) “yes, definitely” to (4) “no, not at all”.

Change in baseline work engagement during 6-, 12-, 18- and 24-months follow-up is assessed with the three statements of the Utrecht Work Engagement Scale (UWES) [47] with six response anchors ranging from (0) never to (5) always, every day.

Change in self-reported stress during 6-, 12-, 18- and 24-months follow-up is assessed with a single question [48] with five response anchors ranging from (1) “not at all” (5) “very much”. In addition to assessing change in baseline self-reported stress by questionnaire we will also assess stress by sending a validated question by text message by mobile phone with SMS-track (https://​sms-track.​com/​) every 4th week over 12 months. This question has been used in our previous studies on stress prevention in the workplace and involves sending a Short Message Service (SMS) every week over 12 months with a high response rate and gives a good indication of changes in stress over time [49].

Change in baseline self-perceived health during 6-, 12-, 18- and 24- months follow-up is assessed with a single question from the SF-12 Health Survey (Short-Form Health Survey [50] with 5-point response anchors ranging from (1) excellent to (5) bad.

Change in self-report baseline short-term sick leave (last 7-days) during 6-, 12-, 18- and 24-months follow-up is assessed with two items [45]. One item assesses the number of hours worked per week; the other item assesses how many days in the last 7 days the respondent has been away from work due to sickness. Change in long-term sick leave is assessed by register-data from the Swedish Social Insurance Agency. Information is collected on long-term sick leave (> 2 weeks) due to mental ill-health (F diagnoses according to ICD-10 (International Statistical Classification of Diseases and Related Health Problems). Data is collected for the entire study period (24 months) and 1 year prior to baseline. Only data on sick leave related to depression, anxiety, or stress problems (light form of mental illness) is collected. Employees provide a separate written consent for the collection of their personal sickness data from the Swedish Social Insurance Agency.

Change in baseline self-reported psychosocial safety climate (PSC) during 6-, 12-, 18- and 24-months follow-up is assessed with 12 items of the Psychosocial Safety Climate Scale [52] with 5-point response anchors ranging from (1) strongly disagree to (5) strongly agree. For example, my workplace management acts quickly to correct problems/issues that contribute to employees’ psychosocial health. The items measure the concept of psychosocial safety climate, which refers to an organizational climate for psychological health, wellbeing and safety and reflects the extent to which the senior management, according to the employees, commits to stress prevention at work [52].

The Regional Ethical Board in Stockholm (dr. nr. 2017/984–31/5) has approved the proposed study. The study complies fully with current ethical requirements regarding the handling and storage of personal data and regarding the written informed consent process in accordance with Sweden’s Personal Data Act and Secrecy Act. Any amendments made to the protocol are communicated to the Regional Ethical Board in Stockholm, the International Standard Randomized Controlled Trial Register and if applicable to scientific journals, where the results of the study are published. Potential participants are given adequate information on both the possible risks and the potential benefits of their involvement in the study in order to allow them to make informed decisions about whether or not to participate. First, an oral presentation is given at each school during which the research team inform all employees about the study. Second, participants will need to provide a written informed consent. Before completing the informed consent form, potential participants receive an information letter describing the purpose of the study, research approach, voluntary participation and data collection process. The information letter also explains that participants can withdraw from participation at any time of the study. Some people could consider the information collected a breach of integrity, even though the collected information is not linked to individual’s personal identification number. Moreover, there may be concerns that the schools will have access to individual answers. The information letter therefore clearly states that data is only collected for the purpose of the study, data is only presented on group-level and no personal data will be shared with the school. The personal identification number is coded in accordance with a serial number so that these are de-identified. The code key is saved, which will enable individuals to request an extract of the collected data and demand that information on him/her to be destroyed without any given reason. One possible risk is that the participants in the focus group interviews could potentially perceive the interview questions as sensitive. This risk is dealt with by stating that participation is voluntary and that participants decide for themselves what information they wish to share. If the participant wishes, he or she can read through and comment on the interview printout. On demand of the participant, information on him/her can be destroyed without any given reason.

Several methodological matters need to be taken into consideration. One of the study’s weaknesses is that we cannot exclude the risk of contamination between schools within the same municipality, i.e. that the group receiving only the educational meeting may receive information and/or material from the workshops. This concern is discussed with the schools and educational board of the municipalities that have received clear instructions not to discuss the content of the workshops or share any workshop documents. Another weakness is that we had to randomize schools before the baseline measurement was performed. This was necessary for the schools to plan the activities during their busy academic school year. Finally, we cannot exclude that the constantly changing school-context (e.g. high turnover of personnel and school principals, reorganizations, national and local policy changes etc.) may influence effectiveness. It is likely that schools in both arms are equally affected by the changing school-context. In order to consider these contextual changes, information is collected from the municipalities’ educational board and school principals during the entire study period.

However, the study also has several strengths, such as the structured and theory-based development of the implementation strategies. The structured approach includes the identification of barriers and facilitators, implementation strategies that target identified barriers and facilitators and the use of model (COM-B) in the development of the strategies. Moreover, two planning-workshops with, among others, school-principals were held to tailor the content of the strategies to the school-context. A second strength is the systematic approach used to plan the evaluation. A recent review on the use of theory (or models and frameworks) to plan or evaluate guideline implementation concluded that only half of the guideline implementation studies were based on theory and many provided little details about how theory was used [33]. To evaluate guideline implementation, we use Proctor and colleagues’ framework for dissemination and implementation research [35]. In addition, CFIR [21] is used to identify barriers and facilitators that may have influenced the implementation process. By doing this the study will generate new insight into the effectiveness of theory-informed implementation. Another strength of this study is that we will collect information on implementation effectiveness and intervention effectiveness (i.e. risk factors for mental ill-health), in addition to conducting a thorough process evaluation, including the identification of barriers and facilitators. This will provide valuable knowledge on whether successful implementation of the recommendations in the guideline will result in changes in the employees’ perceived social and organizational risk factors, mental ill-health and sick leave. The process evaluation will in turn help to explain the effectiveness of the chosen strategies and any contextual factors that might have influenced effectiveness.

Mental ill-health is today the leading cause of sick leave among the working population in Sweden, a professional group that is especially vulnerable are teachers. The problems related to mental ill-health are extensive for the individual due to suffering and stigmatization, and expensive for society. There is therefore an urgent need to gain more knowledge on effective strategies that prevent the occurrence of CMDs. If the multifaceted implementation intervention proves successful, schools will be able to systematically manage social and organisational risks at their workplace and reduce the risk for CMD and sick leave among its employees.