Development and implementation of guidelines for the management of depression: a systematic review

Abstract Objective To evaluate the development and implementation of clinical practice guidelines for the management of depression globally. Methods We conducted a systematic review of existing guidelines for the management of depression in adults with major depressive or bipolar disorder. For each identified guideline, we assessed compliance with measures of guideline development quality (such as transparency in guideline development processes and funding, multidisciplinary author group composition, systematic review of comparative efficacy research) and implementation (such as quality indicators). We compared guidelines from low- and middle-income countries with those from high-income countries. Findings We identified 82 national and 13 international clinical practice guidelines from 83 countries in 27 languages. Guideline development processes and funding sources were explicitly specified in a smaller proportion of guidelines from low- and middle-income countries (8/29; 28%) relative to high-income countries (35/58; 60%). Fewer guidelines (2/29; 7%) from low- and middle-income countries, relative to high-income countries (22/58; 38%), were authored by a multidisciplinary development group. A systematic review of comparative effectiveness was conducted in 31% (9/29) of low- and middle-income country guidelines versus 71% (41/58) of high-income country guidelines. Only 10% (3/29) of low- and middle-income country and 19% (11/58) of high-income country guidelines described plans to assess quality indicators or recommendation adherence. Conclusion Globally, guideline implementation is inadequately planned, reported and measured. Narrowing disparities in the development and implementation of guidelines in low- and middle-income countries is a priority. Future guidelines should present strategies to implement recommendations and measure feasibility, cost–effectiveness and impact on health outcomes.

Furthermore, the growing awareness of the social determinants of mental disorders provides the impetus to prioritize the development and implementation of evidence-based practices for depression management in low-and middleincome countries.
Clinical practice guidelines translate research into recommendations to standardize care, improve health outcomes and reduce morbidity and mortality. 13, 14 We conducted a systematic review of existing guidelines for the management of depression in adults with major depressive or bipolar disorder. We compared guidelines from low-, middle-and high-income countries to characterize disparities in the development and implementation of guidelines globally.

Methods
We conducted a systematic review concordant with Preferred Reporting Items for Systematic Reviews and Meta-Analyses recommendations. 15 Our protocol was registered in the International Prospective Register of Systematic Reviews (CRD42019124759). 16

Search strategy
We searched the following online databases from 1994 to January 2019, without language restrictions: Ovid®, MEDLINE® PubMed®, Embase®, ProQuest PsycIN-FO®; Web of Science, KCI-Korean Journal, Russian Science Citation Index, and SciELO Citation Index; African Journals Online; PakMediNet; EBSCO CINAHL Plus; and Cochrane Library. We searched titles and abstracts using medical search heading terms and keywords. Text keywords used include, for example: bipolar disorder, depressive disorder, mood disorders, depressi*, practice guidelines, evidence-based medicine, guideline*, (medical OR psychiatric association) AND (treatment OR management OR clinical recommendation*). The full search records and details of the grey literature and manual searches are available in the data repository. 17

Inclusion and exclusion criteria
We included national and international guidelines for the management of depression in adults (aged approximately 18-70 years) with major depressive or bipolar disorder defined by standardized diagnostic criteria. Diagnostic criteria included the International statistical classification of diseases and related health problems, 10th edition (ICD-10) and the Diagnostic and statistical manual of mental disorders (DSM-IV, DSM-IV-TR and DSM-5). We excluded guidelines published exclusively for the treatment of depressive symptoms in the absence of major depressive or bipolar disorder; developed for use in local regions, hospitals, states or provinces; developed before 1994 (based on, for example, ICD-9 or DSM-III); or with inaccessible full-texts (we approached authors for access to full-text publications of relevant guidelines). Guidelines with original and updated recommendations were considered duplicates (the most recent update was reviewed). Additional selection and data extraction processes are available in the data repository. 17

Quality assessment
We evaluated the quality of the guideline development process by assessing compliance to the Institute of Medicine's eight standards for clinical practice guidelines: (i) transparency in guideline development processes and funding; (ii) disclosure, management and divestment of conflicts of interest; (iii) multidisciplinary and balanced composition of development group; (iv) recommendations based on a systematic review; (v) rating of evidence quality and strength of recommendation grading; (vi) articulation of recommendations; (vii) external review process; and (viii) schedule for guideline update. 14 A guideline was externally reviewed if it was made available to the general public or target users and relevant stakeholders for comment before its publication. A guideline development group was multidisciplinary and balanced if it included subject-matter experts, clinicians and patient representatives. A guideline met the standard for strength of recommendation grading if all of the following were included for at least three quarters of its recommendations: evidence, harms, benefits, and level of confidence. A guideline clearly articulated its recommendations if each stated recommendation was specific, unambiguous and actionable.

Systematic reviews
Depression guidelines in low-and middle-income countries Yena Lee et al.
We adopted measures from the GuideLine Implementability Appraisal and other published criteria to evaluate how amenable each guideline was to implementation. 18, 19 We assessed characteristics of the guideline development process that facilitate the adoption and application of guideline recommendations: attention to ease of implementation; consideration of economic, legal, social and ethical issues; appraisal of economic or resource implications; evaluation of patient preferences; assessment of implementation enablers and barriers; credibility of authoring individuals or organizations; and the provision of tools to facilitate guideline adoption. We assessed a guideline as having considered the ease of implementation if recommendations requiring minimal resources were presented before those requiring more intensive resources. Less intensive interventions were, for example, those with minimal need for highly skilled personnel, medications, and expensive facilities or infrastructure. 18 We determined the individuals or organizations who authored the guidelines as having credibility if their expertise was concordant with the target audience (for example, a psychiatric association had published recommendations intended for psychiatrists).
The data extraction form is available in the data repository. 17

Critical appraisal
We compared outcome measures between guidelines from high-, upper-middle-and low-or lower-middle-income countries, as classified by the World Bank for the fiscal year of the publication date. 20 We pooled guidelines from low-and lower-middle-income countries for the analysis as there was only one national guideline from a low-income country. Other low-income countries without national guidelines had guidelines as part of international guidelines. We excluded international guidelines from comparisons between income classifications, 21-28 unless they were developed for countries uniformly belonging to a single income classification. 29- 33 We evaluated to what extent differences in access to quality health care predict disparities in the quality of guideline development processes observed across income classifications. The median Healthcare Access and Quality index was computed for each income classification group using the most re-cently published index estimates. 34 The global Healthcare Access and Quality index was 54.4 in 2016; higher indices indicate greater access to quality health care (range: 0-100). 34 We present numbers and percentages of total number of guidelines across or within income groups, as relevant. We conducted statistical analyses using R software version 3.4.4 (R Foundation for Statistical Computing, Vienna, Austria), with α = 0.05. 35 We compared outcomes between income classifications using glm for logistic regressions. We computed incident rate ratios (IRRs) using msm::deltamethod and robust (White-Huber) standard errors (SE) using sandwich::vcovHC to evaluate the association between Healthcare Access and Quality index and guideline development quality.

Results
Our database searches yielded 9833 records. After screening the titles and abstracts of non-duplicate records, we reviewed the full texts of 312 records for eligibility ( Fig. 1) Of the 13 international guidelines, five were from countries in the same income group. [29][30][31][32][33] Of the eight guidelines from countries in different income groups, five covered low-or lowermiddle income countries that lacked national guidelines. [21][22][23][24]28 However, only two international guidelines included at least one author from a low-or lowermiddle-income country. 21,22 Guidelines were published in 27 languages, predominantly in the English language (40 guidelines; 42%).

Target populations and users
Fifty-two guidelines were specifically developed for major depressive disorder and 33 for bipolar disorder. One guideline was developed for the treatment of mood disorders, four for psychiatric disorders, one for psychiatric and neurological disorders, one for medical and psychiatric disorders, and three for depression in special populations (perinatal depression, major depressive disorder with chronic

Scope and intent
The majority of guidelines recommended the use of the two-item Patient Health Questionnaire 125 to screen for depressive symptoms. Fifty-seven guidelines (60%) provided recommendations related to depression screening, 51 of which supported screening for depression in the target setting (such as primary care), either systematically or selectively (such as in high-risk populations, postpartum women or settings with resources available for managing depression). The majority of these guidelines recommended the use of the two-item Patient Health Questionnaire 125 to screen for depressive symptoms. Six guidelines recommended against screening for depression, citing insufficient evidence supporting its effectiveness. Most guidelines with screening recommendations were developed in high-or upper-middleincome countries (53/57 guidelines). The majority of guidelines included recommendations for screening, measuring or treating cardiovascular and metabolic comorbidities (58 guidelines; 61%) or psychiatric comorbidities (65 guidelines; 68%; Table 2).
Recommendations for the primary prevention of depression were included in 16 guidelines (17%), most of which were developed in high-income (10 guidelines) or upper-middle-income countries (four guidelines). These guidelines described risk factors, strategies for reducing risk (such as lifestyle modification, managing stress, psychoeducation or psychosocial support) and methods for early detection. Few guidelines evaluated the literature on the effectiveness of different interventions for primary prevention or cited limitations of current evidence.
Fifteen guidelines (16%) provided decision support or recommendations for assessing work ability, sick leave or return to work; all were published by high-income or upper-middle-income countries ( Table 2). Eleven of these guidelines originated in Europe, 42,58,59,65,92,114,120 while four of these guidelines originated in Canada, Chile, Colombia and Japan. 66,68,102,103,115,117,119,126 The recommendations were often limited to the discussion of standardized scales for measuring work-related impairment, factors moderating patients' return to work, resources for supporting patient employment or occupational rehabilitation and regional disability legislations. The guidelines from Colombia, Finland, Netherlands and Sweden recommended that patients continue to work, unless otherwise indicated, and advised patients and clinicians to discuss work-related factors that may hinder recovery. 66,92,102,103,114,115 Notably, the Swedish bipolar disorder guideline listed an employment rate of 50% among patients as a national target. 122

Development processes
The quality of the guideline development processes varied across country income classifications. The median number of standards met was five (interquartile range: 3-7) for high-income country guidelines, two (interquartile range: 1-4) for upper-middle-income country guidelines and one (interquartile range: 0-1.5) for low-or lower-middle-income country guidelines. The World Health Organization (WHO) guidelines, developed specifically for low-and lowermiddle-income countries, 21 met all but one Institute of Medicine-defined standard (systematic review of cost-effectiveness).
The guideline development processes and funding sources were explicitly specified in 51 guidelines (54%), only two of which originated in low-or lower-middle-income countries. Potential conflicts of interest were openly declared and managed in a higher proportion of guidelines from high-income (36 guidelines; 62%) versus upper-middleincome (six guidelines; 27%) and low-or lower-middle-income countries (two guidelines; 29%).
Only 25 guidelines (26%) were developed by a multidisciplinary group The majority of guidelines recommended the use of two diagnostic questions to screen for depressive symptoms: "During the past two weeks, have you often been bothered by little interest or pleasure in doing things" and "During the past two weeks, have you often been bothered by feeling down, depressed or hopeless. " 125 Notes: Country income groups are World Bank classifications. 20 International guidelines are from countries in different income groups. Total number of guidelines included: high-income countries: 58; upper-middleincome countries: 22; low-or lower-income countries: 7; international (different income groups): 8. Depression guidelines in low-and middle-income countries Yena Lee et al.
comprising subject experts, clinicians and patients or patient advocates. Development groups often lacked patient or patient advocacy representation. None of the low-or lower-middle-income country guidelines had a multidisciplinary development group.
A systematic review of comparative effectiveness of interventions being recommended had been carried out by 57 guidelines (60%), all of which were developed by international authorship groups or in high-or upper-middleincome countries. Some guidelines from low-or middle-income countries were based on recommendations of other published international guidelines. Only 10 guidelines (11%), all from high-and upper-middle-income countries, had conducted a systematic review of cost-effectiveness of a particular intervention or set of recommendations.
Forty guidelines (42%) included with their recommendations ratings of evidence, harms, benefits, and confidence level. More guidelines from highincome countries (27 guidelines; 47%) met the Institute of Medicine's standard for strength of recommendation grading. Thirty-five guidelines (37%) had been externally reviewed (for example, by being posted for public comment or reviewed by stakeholders external to the development group); none of these guidelines originated in low-or lowermiddle-income countries.
Thirty-eight guidelines (40%) stated plans to renew or update their recommendations, excluding three guidelines that were withdrawn past the scheduled updating date. 36,54,78 Fewer guidelines provided a scheduled date for renewal (26 guidelines; 27%). The scheduled renewal date of these guidelines was often within 3-5 years of the publication date (mean: 4 years; standard deviation; SD: 2). Notwithstanding, 49 guidelines (52%) were revisions, of which 17 guidelines were published within 5 years of the previous iteration. On average, guidelines were revised within 7 years (SD: 3). How frequently and how recently revisions were published were similar between high-and middle-income countries (Fig. 3). None of the guidelines from low-or lowermiddle-income countries stated plans to revise recommendations or included a renewal date.
The median Healthcare Access and Quality index was significantly greater among guidelines from high-income

Facilitators and barriers of implementation
The target patient population and intended users were clearly defined in 93 (98%) and 79 (83%) guidelines, respectively. The authors of 75 guidelines (79%) met criteria for credibility with the intended audience ( Fig. 5). Most of these guidelines originated in highincome countries (52 guidelines). Target users or patient representatives evaluated enablers and barriers to the implementation of 24 guidelines (25%); 11 guidelines involved both target users and patient representatives, 11 guidelines involved target users without patient representatives and two guidelines involved only patient representatives in the evaluation of enablers and barriers. None of the guidelines from low-or lower-middle-income coun-tries evaluated enablers and barriers to implementation.
Twenty-one guidelines (22%) evaluated patient preferences by conducting literature reviews of patient preferences or by including patient representatives in the guideline development group, as external reviewers or as members of focus groups. None of the low-or lowermiddle-income country guidelines evaluated patient preferences.
Twenty-four guidelines (25%) ordered their recommendations by ease of use (for example, using a steppedcare model). For management of mild depression, these guidelines recommended low-intensity psychosocial and psychological interventions (for example, physical activity, psychoeducation, sleep modification or computerized cognitive behavioural therapy) before pharmacological interventions (for example, selective serotonin reuptake inhibitors) or high-intensity psychological interventions (for example, cognitive behavioural therapy or interpersonal psychotherapy). Whether a guideline had ordered recommendations by ease of use varied across income classifications.
Eighteen guidelines (19%), mostly from high-income countries (15 guidelines), evaluated the resource implications of implementing guideline recom-mendations. Five guidelines described personnel, infrastructure and training requirements for each recommendation in detail. 21,41,66,80,114 Costs and other economic considerations informed the development of 29 guidelines (30%), 24 of which were from high-income countries. Several European guidelines conducted modelling analyses to project the cost-effectiveness and budget impact of their recommendations. 41,80,102,114,115,119 The number of guidelines that considered legal or ethical issues did not vary across income classifications. Twenty-five guidelines (26%) discussed various legal aspects of patient care, such as involuntary treatment of psychiatric patients, certification requirements for professionals providing psychotherapy, availability of antidepressants across national regulatory agencies, national work or disability legislations and statutory patient rights. Twenty-one guidelines (22%) discussed ethical considerations relevant to care provision, such as risks versus benefits of taking medications while pregnant or breastfeeding and obtaining informed patient consent before initiating electroconvulsive therapy or off-label drug usage.
Thirty-nine guidelines (41%) discussed social aspects affecting patient care or illness presentation, such as race or ethnicity, and advised clinicians to Notes: Country income groups are World Bank classifications. 20 The numerator is given above each bar. International guidelines are from countries in different income groups. We adopted measures from the GuideLine Implementability Appraisal and other published criteria to evaluate how amenable each guideline was to implementation. 18,19 Systematic reviews Depression guidelines in low-and middle-income countries Yena Lee et al.
consider patient factors, such as social support availability, interpersonal relationship quality, workplace or other factors influencing recovery, childhood trauma and developmental disabilities. Other guidelines, for example, emphasized the importance of adapting guidelines to local contexts and training endusers to be culturally sensitive. Some guidelines commented on the lack of availability of personnel with sufficient training in some areas of the country and the implications of this for clinical care. More guidelines from high-income countries (29 guidelines; 50%) were informed by social considerations when compared to upper-middle-income (seven guidelines; 32%) and low-or lower-middle-income (none) countries.

Monitoring implementation
Thirty-three guidelines (35%), mostly from high-income countries (25 guidelines), operationalized monitoring or auditing criteria for assessing the implementation of guidelines. These guidelines suggested quality indicators or measures of guideline concordance, such as the proportion of patients prescribed lithium or a selective serotonin reuptake inhibitor for at least four weeks.
Fifteen guidelines (16%), none of which were from low-or lower-middle income countries, described plans for assessing implementation of guidelines or adherence to guideline recommendations (Fig. 4). However, none of these guidelines provided plans to assess whether these actions would improve health or functional outcomes or cost-effectiveness.
Guidelines described, for example, available health administrative data sets or national electronic medical records that could be used to assess measures of guideline implementation and quality indicators. The Swedish National Quality Register for Bipolar Disorder included longitudinal data from 244 active health-care providers and approximately 30% of patients with bipolar disorder in Sweden. 127 Quality indicators included the percentages of patients diagnosed with a structured diagnostic instrument, receiving psychoeducation, currently employed or who relapsed with a recurrent mood episode in 12 months, as well as sex and regional differences in lithium prescription. 127 The National Institute for Health and Care Excellence in England measured the adoption of some recommendations across mental health guidelines, such as the proportion of people with subthreshold or mildto-moderate depression receiving lowintensity psychosocial interventions. 41 WHO described the adoption of the Mental Health Gap Action Programme in 18 Member States, with a focus on informing future implementation plans and characterizing implementation enablers and barriers. 128 Sixty-five guidelines (68%) provided tools for guideline application, such as a quick reference summary. More high-income country (47/58) and international (6/8) guidelines provided implementation tools. Twenty-four guidelines (25%) described plans for disseminating guidelines, 29 of which originated in high-income countries.

Discussion
We found that many low-and lowermiddle-income countries, especially in Africa, lacked published clinical practice guidelines for the management of depression. However, international guidelines exist that cover or specifically target these countries. [21][22][23][24]28 While the overarching aim of guidelines is to improve health outcomes and cost-effectiveness, it remains unclear to what extent guidelines for the management of depression are being implemented and improving health outcomes, particularly in low-and lower-middle-income countries. 128 Most guidelines lacked plans to assess quality indicators or recommendation implementation. We were unable to identify any national guidelines that included government-sanctioned incentives, such as remuneration, for adhering to guideline recommendations or penalties for not implementing recommendations at point-of-care. A notable exception, not included in the present review, is a guideline for adults with mood disorders from Florida, United States of America. 129 The guideline is integrated into an e-health infrastructure and mandated to be implemented with practitioner concordance monitoring. Government policies that require health-care providers to adhere to recommendations, via health insurance disbursement for example, may facilitate the implementation of guidelines and monitoring of effectiveness.
The disparities in availability, development processes and quality of guidelines underscore an unmet need for decision support in low-and middleincome countries. 1,130 Due to limitations in access to resources, health-care personnel in low-and middle-income countries are additionally constrained in their ability to provide timely and appropriate patient care. 34,131,132 Barriers to the application of standard interventions in many low-resource settings include limitations in the availability of interventions (for example, regulatory approval of certain medicines or acquisition costs) and patient access to health-care professionals (for example, specialist fees, rural regions and private versus public clinics). Limitations in the availability of facilities and resources to monitor serum drug levels and liver or renal function (for example, with lithium treatment) may further limit access to treatments in low-resource settings. 18, [131][132][133][134][135] Recommendations to implement guidelines must be sufficiently contextualized with relevant ethical, legal, social and economic considerations. [136][137][138][139] Low-and middle-income countries are differentially affected by multimorbidity, which drastically reduces life expectancy and increases personal, social and economic burden. 8,12,140 Not only is the prevalence of noncommunicable diseases escalating globally, but the risks of infectious diseases have not declined in low-and middle-income countries, further increasing the burden and complexity of managing chronic conditions in these countries. 141 However, only 50-67% of low-and middle-income country guidelines provided recommendations for the assessment and management of psychiatric or cardiometabolic comorbidities in depression. Future guidelines should provide guidance for screening and managing multimorbidity in adults with depression.
Most guidelines for the management of depression provided tools for the application of guideline recommendations, such as a summary document or a quick-reference guide. However, less than one fifth of the guidelines we identified provided materials for patients; fewer targeted policy-makers or payers. Guideline implementation requires diversity in the engagement of target audiences and stakeholders, as well as realistic and relevant implementation plans. 142 Future guidelines, therefore, need to be developed collaboratively by a broader collective of stakeholders. 137

Systematic reviews
Depression guidelines in low-and middle-income countries Yena Lee et al.
Guideline development groups should include experts in experimental, observational and contextual evidence and knowledge users (such as clinicians and patient advocates). 14,137,143,144 However, less than one third of guidelines for depression globally included a multidisciplinary development group; in comparison, approximately 64% (36/56) of guidelines for diabetes mellitus and 52% (12/23) of guidelines for hypertension were developed by a multidisciplinary authorship group. 145,146 Many guidelines for depression identified in our study were developed without target-user representatives or patient advocates who would be able to provide guidance on the appropriateness, translatability, feasibility and acceptability of guideline recommendations.
Guidelines endeavour to comprehensively review and corroborate knowledge of intervention efficacy, effectiveness, safety and tolerability. Guidelines must also be informed by an evaluation of the determinants, processes and outcomes of implementing evidence-based recommendations. 137 However, while 60% of guidelines for the management of depression identified herein were based on a systematic review of intervention efficacy and effectiveness literature, only 25% of guidelines evaluated enablers and barriers to implementation. Such gaps in the development processes of existing guidelines may limit the implementation of guidelines for mood disorders. [147][148][149] Future guidelines for the management of depression should involve a combination of international and local collaboration, taking into consideration contextual factors that may facilitate or hinder access to health services or treatments. Contextual factors that may be relevant include, for example, structural or policy aspects of the health-care system, education and training; access to treatment methods for depression; and availability of modern technology.
The main aim of our initiative was not to synthesize a consensual set of implementation measures across low-and middle-income countries. However, lessons learnt from implementation science across other noncommunicable diseases could be a starting point for determining policy and implementation principles for depression management. For example, internet access may be needed to facilitate guideline dissemination, especially in low-and middle-income countries. The integration of technology may also facilitate chronic disease management. The guiding principles include prioritizing the involvement of stakeholder and end-user input in any policy around implementation, identification of those people most at risk, and appraisal of local health-care resources.
The paucity of depression guidelines from low-income countries may reflect limitations in our search strategy (for example, the African Journals Online database primarily includes articles published in English). We were more likely to identify guidelines available online than in print only. To mitigate this possibility, we contacted members of the Global Alliance for Chronic Diseases and members of national psychiatric or other medical associations across geographical and linguistic world regions. Database searches may miss guidelines published as government reports or in formats other than peer-reviewed journal articles or meeting abstracts. To improve the likelihood of detecting such guidelines, we manually searched the websites of multiple national and international medical associations and ministries of health and included experts from 27 countries across all continents in our collaboration. Thus, the possible selection bias in our search is unlikely to confound our findings of differences in guideline quality and development across economic strata.
Our large number of evaluators may have resulted in differences in data extraction. However, we completed blinded evaluations in duplicate using structured evaluation forms; a third reviewer independently evaluated all forms. In addition, guidelines were evaluated by two or three reviewers who had not been involved in their development.
The focus of our analysis on guidelines may inadequately capture separate implementation studies of guidelines. Future research should primarily evaluate implementation studies of guidelines. We limited our inclusion criteria to national and international guidelines, which may not capture more regional or local differences in guideline development or implementation. Our comparison of guidelines by country-level income classification and Healthcare Access and Quality index did not consider differences in the availability and accessibility of health care within individual countries. Much of the available research informing guideline recommendations has been conducted in high-income countries, with an overrepresentation of Caucasian groups, often overestimating patient access to expensive medications and specialized care.
In conclusion, the implementation of guidelines for the management of depression is inadequately planned, reported and measured. As a result, it remains unknown to what extent guidelines are acceptable to patients and other target users; are feasible and cost-effective; and improve health outcomes. Narrowing the disparities in the development and implementation of guidelines, particularly in low-and middle-income countries, is a priority. Refinement of decision support processes in depression is a critical first step towards the aim of reducing morbidity, especially in low-and middle-income countries. Future guidelines should present strategies to implement recommendations and measure feasibility, cost-effectiveness and impact on health outcomes, co-designed by stakeholders and experts with practical (experiential) knowledge from low-and middleincome countries. ■ Depression guidelines in low-and middle-income countries Yena Lee et al. и средним уровнем доходов к таковым в странах с высоким уровнем доходов. В будущих клинических рекомендациях должны быть представлены стратегии по внедрению рекомендаций и оценке их осуществимости, экономической эффективности и воздействию на результаты в отношении здоровья.

Systematic reviews
Depression guidelines in low-and middle-income countries Yena Lee et al. 18. Owolabi   Member States without at least one national or international guideline (other than the WHO guideline) were excluded from the total count of 83 countries and not shaded in the map.