Article Text

An international comparison of occupational health guidelines for the management of mental disorders and stress-related psychological symptoms
  1. Margot C W Joosen1,
  2. Evelien P M Brouwers1,
  3. Karlijn M van Beurden1,
  4. Berend Terluin2,
  5. Jani H Ruotsalainen3,
  6. Jong-Min Woo4,
  7. Kyeong-Sook Choi5,
  8. Hisashi Eguchi6,
  9. Jiro Moriguchi7,
  10. Jac J L van der Klink1,8,
  11. Jaap van Weeghel1,9,10
  1. 1Tilburg University, Tilburg School of Social and Behavioral Sciences, Tranzo Scientific Center for Care and Welfare, Tilburg, The Netherlands
  2. 2Department of General Practice and Elderly Care Medicine, VU University Medical Center Amsterdam, EMGO Institute for Health and Care Research, Amsterdam, The Netherlands
  3. 3Cochrane Occupational Safety and Health Review Group, Finnish Institute of Occupational Health, Kuopio, Finland
  4. 4Department of Psychiatry and Stress Research Institute, Seoul Paik Hospital, Inje University School of Medicine, Seoul, Republic of Korea
  5. 5Department of Neuropsychiatry, Eulji University Hospital, Eulji University School of Medicine, Daejeon, Republic of Korea
  6. 6Department of Public Health, Kitasto University School of Medicine, Sagamihara, Kanagawa, Japan
  7. 7Department of Occupational Health Promotion, Kyoto Industrial Health Association, Kyoto, Japan
  8. 8Department of Health Sciences, Division of Community and Occupational Medicine, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
  9. 9Phrenos Centre of Expertise, Utrecht, The Netherlands
  10. 10Parnassia Group, Dijk en Duin Mental Health Center, Castricum, The Netherlands
  1. Correspondence to Dr Margot C W Joosen, Tilburg University, Tilburg School of Social and Behavioral Sciences, Tranzo Scientific Center for Care and Welfare, P.O. Box 90153, Tilburg 5000 LE, The Netherlands; m.c.w.joosen{at}


Background We compared available guidelines on the management of mental disorders and stress-related psychological symptoms in an occupational healthcare setting and determined their development and reporting quality.

Methods To identify eligible guidelines, we systematically searched National Guideline Clearinghouse, Guidelines International Network Library and PubMed. Members of the International Commission on Occupational Health (ICOH), were also consulted. Guidelines recommendations were compared and reporting quality was assessed using the AGREE II instrument.

Results Of 2126 titles retrieved, 14 guidelines were included: 1 Japanese, 2 Finnish, 2 Korean, 2 British and 7 Dutch. Four guidelines were of high-reporting quality. Best described was the Scope and Purpose, and the poorest described were competing interests (Editorial independence) and barriers and facilitators for implementation (Applicability). Key recommendations were often difficult to identify. Most guidelines recommend employing an inventory of symptoms, diagnostic classification, performance problems and workplace factors. All guidelines recommend specific return-to-work interventions, and most agreed on psychological treatment and communication between involved stakeholders.

Discussion Practice guidelines to address work disability due to mental disorders and stress-related symptoms are available in various countries around the world, however, these guidelines are difficult to find. To promote sharing, national guidelines should be accessible via established international databases. The quality of the guideline's developmental process varied considerably. To increase quality and applicability, guideline developers should adopt a common structure for the development and reporting of their guidelines, for example Appraisal of Guidelines for Research and Evaluation (AGREE) criteria. Owing to differences in social systems, developers can learn from each other through reviews of this kind.

  • Mental Health < General expertise
  • Practice guidelines
  • Return-to-work
  • Systematic review

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What this paper adds

  • Mental health problems are among the leading causes of disability worldwide and negatively impact work capacity and productivity.

  • Practice guidelines are important instruments to promote evidence-based practice and increase the quality of care.

  • This paper shows that practice guidelines developed to address work disability due to mental disorders and stress-related symptoms exist in various countries around the world.

  • Occupational health guidelines are rarely available in electronic international databases, which hampers knowledge dissemination and translation.

  • The content of the guideline recommendations is comparable, but not all available guidelines meet current standards for development and reporting quality.


Mental disorders are among the leading causes of disability worldwide.1 These disorders, such as depression, anxiety, adjustment disorders, as well as stress-related symptoms pose an important problem in occupational healthcare because of their negative impact on work capacity and productivity.2 Mental disorders and stress-related symptoms, that is, psychological (work) stress reactions that have caused various health symptoms, can lead to sick leave and long-lasting work disability.3 ,4 In several European countries, Australia and the USA mental disorders are highly prevalent in the working population.5–9 Therefore, mental disorders and stress-related symptoms should not only be considered an individual burden, but also a growing problem for the employers involved and society in general.

In Europe, the total costs of mental disorders (including healthcare costs and work disability costs) are estimated to be €240 billion annually.10 In Europe and the USA, mental health costs mainly arise from productivity losses due to sickness absence or reduction in work functioning.8 ,10 ,11 The latter is considered a largely hidden cost of mental disorders at the workplace.12

Considering the impact of sick leave and reduced work functioning on the individual and society, there is a need for effective management strategies. New evidence is constantly being developed and is usually published in scientific journals. However, for practitioners it is often not feasible to identify, read and interpret the search results for choosing a strategy to problems met in daily practice.13 This can result in large variations in quality of healthcare and can even lead to harmful care.14 Evidence-based practice guidelines are valuable tools to summarise and translate scientific evidence into recommendations that can be used in practice.15 ,16 A guideline is defined as “systematically developed statements to assist practitioner and patient decisions about appropriate healthcare for specific clinical circumstances”.17 The purpose of guidelines is to make explicit recommendations with the intention to influence professional behaviour.18 Therefore, guidelines are important instruments to enhance treatment quality and decrease unnecessary variability in care.19

Owing to the growing impact of mental disorders and stress-related symptoms at work, we can expect more occupational health guidelines to be issued to improve the management of these health problems in the occupational context. These guidelines may be of different content since guidelines are based on the best-available scientific evidence, supplemented with clinical expertise, patient/worker preferences and tailored to local circumstances.20 We know from previous studies that not only content but also the quality of development of guidelines can differ considerably.21 ,22 Recognising the increasing need for quality and transparency, several guideline organisations have set development standards.23 In addition, instruments such as the Appraisal of Guidelines for Research and Evaluation (AGREE) are recognised as valuable tools to evaluate the key aspects of the guideline development process based on what is reported in the guideline.24 As for occupational health guidelines on mental disorders and stress-related symptoms it is not clear what guidelines are used in different countries and if these guidelines meet currently accepted reporting quality criteria. Providing an overview of currently available guidelines can be useful for guideline developers to see how the evidence from literature is used in and adapted to the specific context in different situations.

This study aimed to identify occupational health guidelines focusing on the management of mental disorders and stress-related symptoms from different countries worldwide and to describe them, compare the content and assess their developmental and reporting quality. Specific research questions were: (1) What guidelines can be identified and to what extent are they comparable regarding recommendations for the assessment and treatment of mental disorders and stress-related symptoms, and (2) What is the developmental and reported quality of these occupational health guidelines?


Search strategy

We used two search strategies to identify relevant guidelines: a systematic search in publicly available bibliographic databases and another search by consulting experts, that is, members of the International Commission on Occupational Health (ICOH).

First we searched in two guideline-specific databases: National Guideline Clearinghouse (NGC) and Guidelines International Network Library (G-I-N). In addition, we searched PubMed to trace relevant guidelines in biomedical literature by checking the content and reference lists of relevant reviews on guidelines. To develop a systematic search strategy we first translated our research question according to the PICO method (Patient/population, Intervention/exposure, Control, Outcome).25 This resulted in three relevant groups: (1) Patient/population: Mental disorders and/or stress-related symptoms, (2) Intervention/exposure: Guidelines and (3) Outcome: occupational health outcomes. Including a Control component was not appropriate given our research question. For each search group we included terms and/or synonyms that were used as subject heading and/or text words (see online supplemental file 1). The first group of search terms represented the target population, that is, workers with mental disorders and/or stress-related symptoms. Having a mental disorder according to the Diagnostic and Statistical Manual fourth edition (DSM-IV) classification or suffering from psychological and/or stress-related symptoms were the eligible conditions.8 ,26 The second group included terms concerning guidelines, and the third group embodied occupational health outcomes. We focused on a range of occupational health outcomes such as work participation, work functioning, quality of working life, work resumption and return to work. The selection of search terms was based on the Cochrane OSH group search strategy27 and additional terms relevant for our research question. We combined the three groups of search terms with the operator ‘AND’, and we adjusted the string to function in each of the databases we used.

In the second search we consulted experts. Since many guidelines are not published in international medical journals, we contacted national occupational health organisations to identify guidelines. Sustaining and affiliate ICOH members whose contact details were publicly available on the ICOH website were contacted (see online supplemental file 2). During the period January to June 2012 the organisations from 22 countries spread across the world were contacted by email. Up to three reminders were sent in case of non-response. We asked (1) information on the existence of guidelines focusing on the management of workers with mental health problems in their own country, (2) the language in which the guideline was available, (3) if the contacted person could provide us with the guideline, and (4) or provide information about other organisations or key persons who could supply further information about this topic.

This review was designed and conducted according to the PRISMA statement for reporting systematic reviews.28

Selection of guidelines

To be included in the review guidelines had to meet the following criteria: (1) meet the definition of a guideline by Field and Lohr17 (2) the subject was a mental disorder and/or stress-related symptoms and (3) the guideline addressed the management of the mental health problem primarily targeting occupational health outcomes. Guidelines were excluded if they did not contain specific recommendations for practitioners, focused on primary prevention only, or were not available as full text or comprehensive summary. We applied no language restrictions.

All documents retrieved were evaluated. First, the title and (if available) abstract were reviewed using the aforementioned eligibility criteria. This was performed by two independent reviewers (MJ reviewed 100%; EB, JvW and KvB each reviewed 33.3% of the documents). Disagreements were discussed until consensus was reached or the document was included for full text assessment. In the second step, the full-text documents were assessed by the same four reviewers against the inclusion and exclusion criteria. Disagreements were discussed until consensus was reached.

Data extraction and analysis

Comparison of guidelines

The content of the included guidelines was extracted, summarised and compared regarding the following topics: multidisciplinarity of guideline committee, presentation of the guideline, target population, target users and the evidence level of the recommendations. Recommendations regarding assessment and management were summarised and compared. Only the parts of the guideline that dealt with treatment and management of problems were extracted and not with prevention of problems since this was not the focus of this review. The guidelines were assessed by researchers with relevant language skills (ie, native speakers with excellent command of English).

Assessment of developmental and reporting quality of guidelines

The quality of the guidelines was assessed using the Appraisal of Guidelines for Research and Evaluation (AGREE II) Instrument ( This is a validated generic tool to evaluate the process of guideline development and provides a systematic framework for assessing key components of clinical guideline quality. The instrument consists of 23 items grouped into six domains: (1) scope and purpose (ie, aim and target population), (2) stakeholder involvement (ie, are appropriate stakeholders involved in the development), (3) rigour of development (ie, process of gathering and synthesising the evidence), (4) clarity and presentation (ie, language, structure and format), (5) applicability (ie, likely barriers and facilitators to implementation) and (6) editorial independence (ie, potential competing interests). One item is added to score the overall quality of the guideline. Each item is rated from 1 (strongly disagree or no information provided on this item) to 7 (strongly agree). All information, including guideline documents and available supporting documents, about the development process was gathered prior to the appraisal. Per guideline, two researchers independently assessed the guideline. Three reviewers (JvdK, JvW and BT) were involved in the development of one or two of the included guidelines. To avoid conflict of interest, they were excluded from the appraisal of their own guidelines.

In agreement with the AGREE II manual, domain scores were calculated by summing all scores of the individual items in a domain, and by standardising the total as a percentage of the maximum possible score for that domain: ((Obtained score—Minimum possible score)/(Maximum possible score—Minimum possible score))×100. In line with similar studies, we defined scores above 60% as good, scores of 30–60% as moderate and scores lower than 30% as poor quality.29 ,30


Selection of guidelines

In total, 2126 titles were identified by the international search. After removing 12 duplicates, 2114 documents were reviewed for inclusion. On the basis of title and abstract 2002 documents were excluded. After checking the reference lists of the full-text, seven documents were added.

A total of 119 full-text documents were reviewed. After applying the inclusion criteria, 14 documents were included from five different countries: 1 Japanese,31 2 Finnish,32 ,33 2 Korean,34 ,35 2 British36 ,37 and 7 Dutch.38–44 Table 1 presents the title, country, agencies and year of publication of the included guidelines. The most frequent reasons for excluding full-text references was that the document was not a guideline (n=36), guidelines were developed for diagnostic purposes or focused on primary prevention (n=22), and guideline outcomes were not work related (n=21). Figure 1 is a flow chart of the inclusion process.

Table 1

Included guidelines (country, title, development agency and year)

Figure 1

Flow chart of the inclusion process. Abbreviations: G-I-N, Guideline International Network; NGC, National Guideline Clearinghouse; ICOH, International Committee of Occupational Health.

Characteristics and comparison of recommendations

Guideline characteristics

Table 2 presents background information on the development process of the included guidelines. Below, guideline characteristics are described including references to the specific guideline presented in table 1 (eg, GL 10 refers to the Japanese guideline).

Table 2

Guideline characteristics related to the development of the guideline (Guideline Committee, Target users, Presentation of the guideline and the Evidence level of the recommendations)

The guideline development committee was in all but one case multidisciplinary, including disciplines such as occupational medicine, general practice, psychology, nursing, human resource management, researchers and workers’ representatives. The guideline committee of the Dutch guideline for Psychologists consisted of psychologists only (GL 7). The included guidelines were presented as guideline documents, electronic documents or published in a (scientific) journal. Five guidelines were revised versions of previously developed guidelines (GL 1, 4, 5, 8 and 10). Four Dutch guidelines (GL 1–4) and one of the UK guidelines (GL 9) were developed using comprehensive literature searches to identify relevant literature, and provided information on the weighing of evidence. For three Dutch, two Finnish and two Korean guidelines (GL 5–7, 11–14) the recommendations were based on literature, but no or only limited information was provided on the search strategies and weighing of evidence. In the other UK guideline (GL 8) there were no direct links between recommendations and references. In the Japanese guideline recommendations were based on professional discussion (GL 10).

Target population, objectives, assessment recommendations and management recommendations

Online supplementary table 3 provides details of the target population, guideline objectives and assessment recommendations (diagnostic classification and problem inventory) of the included guidelines. Online supplementary table 4 provides information on management recommendations. These finding are described below (including references to the guidelines).

Target population and guideline objectives

All guidelines focused on workers with mental health problems or psychological symptoms. However, target populations differed regarding the specific diagnosis (depression, anxiety disorders, adjustment disorders, medically unexplained symptoms, mental health symptoms in general, work-related stress symptoms and loss of control due to disaster) and work status (workers on sick leave, workers with participation/performance problems, people who want to work). Depending on the target population and the user group, guideline objectives focused on different (but related) occupational health outcomes. Most guidelines aimed to improve return to work (GL 1, 4, 5–7, 9–11) and/or work retention (GL 1, 3, 6–9, 11, 12 and 14).

Guidelines recommendations regarding assessment

All but one guideline (GL 9) included recommendations on the assessment of workers. Regarding assessment of the individual, most guidelines agreed on assessing mental health symptoms. Some guidelines (GL 1, 3 and 12) recommended assessing symptoms in relation to limitation at work, or the stress process. Only two guidelines did not specifically include symptom assessment (GL 7 and 10). In case of the Dutch guideline for psychologists, this was because this guideline focused on work in addition to care-as-usual, which included an extensive assessment of diagnostics and symptoms.49 The Japanese guideline focused on functioning ability rather than on symptom reduction (GL 10). Classification of diagnosis was recommended by the majority of the guidelines, mostly to assess if the worker was eligible to be treated according to the guideline (GL 1–5, 7, 9 and 11) and/or for assessment reasons (GL 6 and 11). Most guidelines recommended assessing performance problems in the private and/or social life. In addition, all Dutch and both Korean guidelines recommended to examine factors of influence on recovery, such as barriers, perpetuating factors and stressors in private and working life. Four guidelines clearly described how to assess complications, such as suicide risk (GL 1), self-destructiveness (GL 11), and analysis of high-risk groups (GL 13 and 14). Three guidelines included recommendations concerning coping strategies, specifically suggesting assessment of the worker's problem-solving skills (GL 1, 2 and 4).

The importance of assessing workplace factors relevant to mental health and stress-related symptoms and the recovery process was addressed in all the guidelines except for one (GL 9). Mostly this concerned assessment of work context factors such as communication and/or problem-solving skills between worker and supervisor (GL 1, 2, 4, 7, 10, 13 and 14), supportive work environment (GL 1, 7, 10, 11 and 13), competencies and skills at work (GL 3, 10 and 13) and complications/risk factors: for example, work conflict (GL 1) and risks for coworkers (GL 8). Assessment of work content was recommended by four guidelines (GL 5–7 and 10; ie, assessment of workload/stressors and job content). In addition, assessing performance problems at work was recommended in three guidelines (GL 1, 4 and 11) and one guideline recommended assessing work factors that hindered recovery (GL 7). Some guidelines recommended an inventory of the needs for vocational rehabilitation and possible solutions at work (GL 3–5 and 11).

Guideline recommendations regarding management/treatment

We classified management and treatment recommendations into the following categories: Advice/Counselling, Specific mental health treatment, Specific Return-to-Work interventions, Referral to/Collaboration with other healthcare providers and stakeholders and Evaluation. In some guidelines recommendations were made regarding preventive measures, but these were not extracted as this was outside the scope of this review.

With respect to Advice/Counselling, all four of the most recently developed Dutch guidelines recommended a process-based approach of the recovery process (GL 1–4). This involves monitoring the recovery process, and facilitating this process by supportive but careful guidance and only intervening if the recovery process stagnates. The Finnish Depression guideline also included elements of this approach (GL 11). Furthermore, an activating approach (GL 4 and 13), early start of the guidance (GL 1, 4 and 8) and psychoeducation were recommended (GL 4, 5, 7 and 11). In addition, several guidelines agreed on the need to invest in communication with, and support of the worker (GL 5, 6, 11–14). Some guidelines recommended assisting/advising on financial support/grants (GL 3 and 11).

Recommendations on specific mental health treatment concerned mainly psychological interventions, in most cases cognitive (behavioural) interventions, or referring the worker to specialised treatment if the guideline user is not skilled or able to provide psychological care (GL 1, 2, 4–9, 11–13). Other treatment recommendations concerned the use of self-management programmes (GL 3), intervening on precipitating and perpetuating factors relating the worker and their environment (GL 4 and 5), and the use of an Employee Assistance Programme (GL 10). Furthermore, three guidelines (GL 2, 4 and 11) agreed that medication was not (always) indicated, except in cases of severe mental disorders or severe constraints, such as severe depressive disorders or insomnia. The other guidelines did not include any recommendations concerning medication.

Return to work measures were recommended by all guidelines. Half of the guidelines recommended specific work adaptations such as reduction of stressful work conditions, lower work demands, simpler and easier work or prohibition of night shifts (GL 6–8, 10–12 and 14). The remaining guidelines focused on communication and advice for the employer and work setting. Advice consisted of practical problem-solving advice (GL 1), employer being actively involved by tackling precipitating and perpetuating work factors (GL 5 and 8), employer should keep in touch with sick-listed worker (GL 9) and return-to-work meetings with the worker and employer (GL 11). Three guidelines recommended to improve social reintegration at the workplace (GL 11, 13 and 14), by supporting the workplace (GL 12) or by giving instructions to the coworkers and avoiding stigma (Korean guidelines). Furthermore, one Dutch guideline recommended using the Individual Placement and Support model of Supported employment to achieve work participation (GL 3).

With respect to referral/collaboration, in most cases recommendations were related to communication of the treatment plan and/or cooperation between professionals, or involved stakeholders at the workplace (GL 2–5, 7, 8, 13 and 14). Five guidelines advised referral to the psychologist or psychiatrist if recovery stagnates, or exchanging information (GL 2, 4, 8, 10 and 11). Referral to specialised care was also recommended by five guidelines (GL 1, 2, 4, 5 and 11). In addition, four guidelines recommended referral to or discussion with the general practitioner in case of stagnation (GL 1, 2, 4 and 8). Three guidelines did not include specific recommendations concerning referral to or collaboration with other healthcare providers (GL 5, 9 and 12).

Ten guidelines highlighted evaluation recommendations (GL 1, 2, 4, 6, 7, 9–11, 13 and 14), four guidelines did not mention evaluation specifics (GL 3, 5, 8 and 12). Recommendations mainly contained follow-up sessions with the worker, supervisor and/or other care professionals and evaluation of the recovery process (GL 1, 2 and 4), work ability assessment (GL 6, 7 and 11), goals checking (GL 7), and/or exchange of information (GL 10).

Developmental and reporting quality of guidelines

Table 3 presents the AGREE domain scores of the appraised guidelines and the mean scores per domain. The ‘scope and purpose’ domain received the highest scores (73%). Overall, the aim and target population of the guidelines were well documented. Most guidelines (GL 1–5, 7–9, 13 and 14) scored over 60% in this domain.

Table 3

Ratings on AGREE domains and mean scores per domain as a percentage of maximum possible score

On average, the domain ‘Editorial independence’ received the lowest scores (31%). Only one guideline (GL 3) included sufficient information on the independence of the funding body and acknowledgment of possible conflict of interest of the development group. Most guidelines did not explicitly mention this topic. Therefore, six guidelines scored moderate (GL 1, 3, 4, 7, 13 and 14) and seven guidelines (GL 5, 6, 8–12) scored poorly on this domain.

The domain ‘Applicability’, which pertains to the organisational, behavioural and cost implications of applying the guidelines scored only moderate (33%). Seven guidelines had moderate quality scores (GL 1–5, 10 and 13) and five guidelines were considered of poor quality on this domain (6–9, 11, 12 and 14).

As for the ‘Clarity of presentation’, on average the quality was moderate (59.7%). In some guidelines the recommendations were specific and unambiguous (GL 2, 3, 10, 13 and 14). However, in other guidelines recommendations were unclear or ambiguous, mere statements or simply repeated scientific evidence (GL 6, 7, 9, 11 and 12). In addition, key recommendations were often difficult to identify (GL 1, 5–8, 11 and 12).

On average the quality of the ‘Stakeholder involvement’ in the development of the guideline was good (61.5%). The majority of the guidelines had no or limited description of the development of the search methods (ie, ‘Rigour of development’). The two Dutch multidisciplinary guidelines were of good quality in (GL 3 and 4) and provided comprehensive information on the literature search (eg, search terms) selection criteria (eg, weighing of evidence criteria) and links between literature and recommendations.

Regarding the overall assessment of the guidelines, half of the guidelines (7/14; GL 4, 5, 9, 11–14) received a moderate quality score. Three were considered of poor quality (GL 6–8) and four of good quality (GL 1–3 and 10).


Considering the magnitude of the problem that mental disorders and stress-related problems can impose on workers, employers and society it is surprising that, after an extensive search, we found only five countries with one or more occupational health guidelines dealing with these problems. From six other countries, experts confirmed that no occupational health guidelines targeting mental health disorders or stress-related symptoms were available in their country.

The 14 included guidelines were in many ways similar. They had a shared focus in assessment of mental health symptoms and diagnosis, and inventory of performance problems in private and/or social life. All but one guideline addressed the importance of assessing work factors relevant to mental health symptoms and the recovery process. Guideline recommendations mainly focused on advice and counselling methods, and return to work interventions for occupational health professionals. In general, guidelines recommended providing psychological treatment, and several guidelines recommended promoting communication with the worker, and/or cooperation with the employer and other involved stakeholders. The discrepancies between the guidelines were mainly related to the methods used to list work factors and return-to-work interventions and the extent to which these were described.

Our results show that the developmental and reporting quality of occupational health guidelines on mental health problems varies considerably. According to our judgment, the developmental process of three guidelines was of low quality and only four were of good quality when assessed with the AGREE II instrument. The majority of the guidelines missed clearly formulated (key) recommendations. Furthermore, most guidelines inadequately reported editorial independence, barriers and facilitators for implementation and the process to gather and synthesise evidence. Best described was the ‘scope and purpose’ of the guidelines.

Methodological considerations and implications concerning quality and content of guidelines

Of the 14 included guidelines, three were developed in Asia and the remainder was from Europe. From Canada, USA, Australia, New Zealand, South Africa, Germany, Denmark and the Czech Republic we found documents addressing the problem of mental health problems in occupational health. These were often good initiatives in preparation for practice guidelines and might also finds its way to the general public, but did not meet our initial inclusion criteria.

Surprisingly, seven included guidelines were developed in the Netherlands. There are several possible reasons why some many Dutch guidelines were found. First, the organisation of the Dutch occupational healthcare system and its sociopolitical system, in which sick leave guidance by an occupational physician (OP) plays a central role. Over the past decades, mental health problems became the most important category for disability claims in the Netherlands. Consequently, there is a need for effective management strategies for OPs and related professionals. As guideline development is considered an important part of medical professionalism in the Netherlands, medical professional organisations actively participate in guideline development.50 Second, most of the researchers involved in the present study are Dutch experts in the field of occupational medicine and/or mental healthcare and are familiar with Dutch guidelines. However, any researcher from another country, using the same thorough search method, would have found the same results.

Regarding the content, there was some variety between the guidelines. For example, variation in target users (OP, psychologist, manager, general physician), target population (eg, workers with depression, work-related stress problems or medically unexplained symptoms) and the objectives of the guidelines (eg, return to work, work retention, work functioning). These differences might have emerged from differences in healthcare systems, or differences in the membership of the guideline committees.20 Also international variations in sickness and disability systems may play a role. For example, in the Netherlands sickness and disability compensation is provided regardless of the cause of disability, but in Finland only mental disorders are eligible for compensation and symptom diagnoses (such as stress and burnout) are not.9 In addition, in Canada,51 Australia52 and the USA53 no mental health conditions are covered. These differences in systems may impact the content of guidelines; for instance interventions may be successful in one country but totally inappropriate in another given the differences in roles of caregivers and other stakeholders and the legal protections available to workers. As guideline recommendations should not be based on scientific evidence alone, but also take into consideration local circumstances, cross-cultural differences may be reflected in guidelines thereby making it difficult to compare the content of these guidelines.54

Several other reviews have appraised the developmental and reporting quality of occupational health guidelines using the AGREE criteria and obtained similar results to those reported here. Although these studies were not exclusively focused on mental health problems, they also found that the ‘purpose and objective’ was well described in the guidelines, but that the stakeholder involvement, rigour of development, application and editorial independence was poorly reported.21 ,22 ,29 ,30 ,55 We found it difficult to extract the content of recommendations from the guidelines, since the recommendations were often presented in an unclear and/or ambiguous way (see AGREE scores on the domain ‘Clarity and presentation’). Often, recommendations were merely statements or only presented evidence rather than clear recommendations. Moreover, key recommendations were not always easy to identify. In addition, none of the guidelines received a good quality score on the AGREE domain ‘Applicability’, which concerns a description of likely barriers and facilitators to implementation of the guideline. AGREE does not appraise the quality of the content of the guideline, nor does it assess the users’ adherence to it in practice, or its clinical impact, although the AGREE domains ‘stakeholder involvement’ and ‘applicability’ are relevant domains for the usability of the guideline. Moreover, low development and/or reporting quality can have a negative influence on the uptake of guidelines in practice.56 The developing process and reporting of the recommendations is therefore of great importance for a successful implementation.

Strengths and limitations

Mental health disorders are among the leading causes of (work) disability and, according to the WHO depression will become the leading cause of burden of disease worldwide by 2030.1 Given the impact that mental disorders and stress-related symptoms have on the individual, occupational setting and society in general, it is expected that more occupational guidelines in mental health will be developed. Since the medical, social and political context may differ between countries and possibly influence guideline recommendations, reviews such as this may help developers to learn from each other and improve the quality of their guidelines.

The results of this review need to be considered in the light of some methodological limitations. First, the methods we used to identify relevant occupational health guidelines do not guarantee that a representative sample was included. Guidelines were difficult to find since they (generally) are only available in their original language and are rarely indexed in MEDLINE. Moreover, the two guideline-specific databases G-I-N and NGC seldom contain occupational health guidelines. Only two of the included guidelines were found via systematically searching established electronic databases (GL 4 and 9). The remainder was discovered with the help of ICOH members who provided information on the existence of guidelines in their own country. Five representatives of national (ICOH) organisations did not reply to our survey request, preventing inclusion of possible unpublished guidelines from these countries. Despite this limitation, our search method of combining an extensive database search with knowledge from experts all over the world, is an innovative method compared with the search strategies of similar reviews.21 ,57 Although the responses of the ICOH members might not be representative for the entire situation in their respective country, it provided relevant information about the existence and non-existence of national occupational health guidelines, which was not revealed via the globally used databases. In addition, we did not restrict our search to English-language publications, which allowed us to include guidelines written in Finnish, Korean, Japanese and Dutch. To reduce the chance of missing information when translating the guidelines, these non-English guidelines were appraised by native speaking researchers with excellent command of English.

A second limitation might be the inclusion and comparison of four guidelines that were developed more than 6 years ago (GL 5–7 and 9). Assuming that these guidelines were based on the latest scientific evidence available at that time, comparison of the content with recently developed guidelines might provide slightly distorted results. However, since the aim of this review was to collect currently available guidelines, we did not impose any restrictions on publication date.

Recommendations for future guidelines

This review shows that occupational health guidelines on mental health problems are difficult to identify. Only two out of 14 guidelines could be found in electronic databases. To enable guideline developers, implementers and researchers to learn from each other, national guidelines should be accessible via international databases and preferably be available in English. To improve quality, applicability and implementation of guidelines, guideline committees should adopt a common structure for the development and reporting of their guidelines. Preferably, developers should follow currently available minimal quality criteria for the development of guidelines.23 Moreover, we recommend that guideline developers publish their ‘background’ study and their literature study, and clearly describe how they derived recommendations from the available evidence. When high-quality guidelines will be developed, then, other developers can adapt these guidelines, use the same evidence and decide whether the considerations are valid for their context.58

Compared to clinical guidelines, occupational health guidelines are still rarely available in international databases such as G-I-N and NGC. Considering the scope of the problem of sickness absence due to mental health problems and its personal and financial consequences, integration of work-related aspects and occupational health advice in guidelines should be stimulated.59 ,60 So-called multidisciplinary guidelines are good examples of initiatives to close the gap between general healthcare and occupational healthcare.39 ,41 ,59

Finally, for those guidelines that are ‘out of date’ but still relevant for daily practice, we recommend updating them so that the recommendations are consistent with current scientific evidence and expert and worker opinion.


The authors thank the ICOH institutions for providing information on the existence of guidelines in their country. The authors thank Jos Verbeek of the Finnish Institute of Occupational Health for his contribution to this research by appraising the quality and summarising the content of the Finnish guidelines, and Ute Bültmann of UMC Groningen for her advice on the inclusion process.


View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • Contributors MCWJ, EPMB, KMvB, BT, JJLvdK, JvW contributed to the design, analysis and interpretation of the data, draft of the article and its revision for important intellectual content. JHR, J-MW, K-SC, HE, JM analysed and interpretated the data and critically revised the article for important intellectual content. All authors approved the final version of the article to be published.

  • Funding This study was financially supported by the Netherlands Organisation for Health Research and Development (ZonMw; grant number 208030001).

  • Competing interests Four authors were involved in the development of guidelines that were included in this review: JvdK was manager and main author of the Dutch NVAB guideline (GL 1), JvW was chair of the committee and coauthor of the Dutch multidisciplinary guideline for severe mental health problems (GL 3), BT was main author of the Dutch LESA guideline (GL 4), and JR assisted by writing some of the evidence statements of the Finnish stress at work guideline (GL 12). These authors do not receive fees for the use of the guidelines.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles