Review of layperson screening tools and model for a holistic mental health screener in lower and middle income countries

Background The needs of people diagnosed with Mental Neurological and Substance-Use (MNS) conditions are complex including interactions physical, social, medical and environmental factors. Treatment requires a multidisciplinary approach including health and social services at different levels of care. However, due to inadequate assessment, services and scarcity of human resource for mental health, treatment of persons diagnosed with MNS conditions in many LMICs is mainly facility-based pharmacotherapy with minimal non-pharmacology treatments and social support services. In low resource settings, gaps in human resource capacity may be met using layperson health workers. A layperson health working is one without formal mental health training and may be equivalent to community health worker (CHW) or less cadre in primary health care system. Objectives This study reviewed layperson mental health screening tools for use in supporting mental health in developing countries, including the content and psychometric properties of the tools. Based on this review this study proposes recommendations for the design and effective use of layperson mental health screening tools based on the Five Pillars of global mental health. Methods A systematic review was used to identify and examine the use of mental health screening tools among laypersons supporting community-based mental health programs. PubMed, Scopus, CINAHL and PsychInfo databases were reviewed using a comprehensive list of keywords and MESH terms that included mental health, screening tools, lay-person, lower and middle income countries. Articles were included if they describe mental health screening tools used by laypersons for screening, delivery or monitoring of MNS conditions in community-based program in LMICs. Diagnostic tools were not included in this study. Trained research interviewers or research assistants were not considered as lay health workers for this study. Results There were eleven studies retained after 633 were screened. Twelve tools were identified covering specific disorders (E.g. alcohol and substance use, subcortical dementia associated with HIV/AIDS, PTSD) or common mental disorders (mainly depression and anxiety). These tools have been tested in LMICs including South Africa, Zimbabwe, Haiti, Malaysia, Pakistan, India, Ethiopia and Brazil. The included studies show that simple screening tools can enhance the value of laypersons and better support their roles in providing community-based mental health support. However, most of the layperson MH screening tools used in LMICs do not provide comprehensive information that can inform integrated comprehensive treatment planning and understanding of the broader mental health needs of the community. Conclusion Developing a layperson screening tools is vital for integrated community-based mental health intervention. This study proposed a holistic framework which considers the relationship between individual’s physical, mental and spiritual aspect of mental health, interpersonal as well as broader contextual determinants (community, policy and different level of the health system) that can be consulted for developing or selecting a layperson mental health screening instrument. More research are needed to evaluate the practical application of this framework.


24
The needs of people diagnosed with Mental Neurological and Substance-Use (MNS) conditions are 25 complex including interactions physical, social, medical and environmental factors. Treatment requires a 26 multidisciplinary approach including health and social services at different levels of care. However, due 27 to inadequate assessment, services and scarcity of human resource for mental health, treatment of persons 28 diagnosed with MNS conditions in many LMICs is mainly facility-based pharmacotherapy with minimal 29 non-pharmacology treatments and social support services. In low resource settings, gaps in human 30 resource capacity may be met using layperson health workers. A layperson health working is one without 31 formal mental health training and may be equivalent to community health worker (CHW) or less cadre in 32 primary health care system.  where treatment of persons diagnosed with MNS conditions is mainly facility-based pharmacotherapy 93 with minimal non-pharmacology treatments and social support services.

94
The lack of integrated assessment tools could explain the gap in their treatment plan because it is training of non-specialist and various innovation to engage lay person in self-care and informal 112 community care in order to enhance access, reduce cost and reduce stigma (7). These strategies create the 113 opportunity to expand the integration of mental health into existing health care system, strengthen human 114 resources, improve delivery of services and care reaching more persons with MNS conditions.

116
Engagement of non-professionals, or lay persons, in the screening and delivery of mental health may be a 117 promising mechanism to improve support for persons with MNS conditions. Non-specialist mental health 118 workers have been classified as health professionals/workers that may have received general mental 119 health training but are not specifically trained as mental health professionals (e.g. doctors, nurses, para-120 professionals and nonprofessional lay providers) (8,9). Non-specialist mental health workers may also assistants were also not considered as lay assessors for this study because we could not verify their 192 professional or prior training. Studies in which lay persons were only engaged in providing community-193 based mental health interventions but not in the assessment or screening of the participants were also 194 excluded from this review. Studies that did not specify who administered the tools were also excluded.

195
Articles published before 2008 and in languages other than English were also excluded.

216
The holistic policy and intervention framework (HPIF) also known as the five pillars of global mental 217 health and addiction, provides guidance on the analysis, evaluation, and sustainability of global mental 218 health capacity building interventions (Khenti et al, 2015). and their references scanned to identify the final 11 articles reviewed for this study.

272
One systematic review was identified focusing on the review of mental health screening tools that are 273 validated for use in LMICs. This review did not focus specifically on tools validated for lay person use.

274
However, three articles cited in the review that involved laypersons in the validation studies were 275 included in this study. The eleven articles that were reviewed for this study were those which describe

283
The majority of studies focused on the description or application of laypersons service provision rather 284 than specific evaluation of screening tools. Although the primary focus of these studies was not on the use 285 of screen tools per se, these studies were included as they provided a description of a number of screening  (14).

294
This review included tools that were designed for use by lay-persons as well as other tools. We also 295 searched the references of this review to determine if there were additional relevant studies to those 296 identified in our review. We identified four additional studies that described the validation of several tools 297 for use by lay persons in a community or PHC setting(26-29). Find the information in Annex II.

298
Across all studies, twelve screening tools were identified. Table 1 below provides a description of the 299 identified mental health screening tools validated for layperson use in low and middle income countries.  (14). Table 3 provides a summary classification of the identified screening tools.

308
Analysis of the identified tools shows that most tools were being used as interviewer-administered tools 309 even though they were originally developed and validated for self-report. Information was not available 310 as to whether this difference in mode of administration affects the measured outcome or score on these 311 instruments.

312
For this study, we included mental health screening tools that have been developed or validated and used 313 at the community level by lay persons that are non-mental health professionals and non-clinicians or non-   problem over the past few weeks using a 4-point Likert-type scale (from 0 to 3). The score was used to 396 generate a total score ranging from 0 to 36. The positive items were corrected from 0 (always) to 3 397 (never) and the negative ones from 3 (always) to 0 (never

449
There are a number of strengths of the lay person screening tools identified in this review. These include 451 minimal training requirements, low literacy requirements for completion, and the ability some tools to 452 detect psychiatric conditions in physically ill patients. They also have strong psychometric properties in 453 the study populations. In terms of limitations, most of these tools assess psychiatric symptoms alone and 454 are restricted to the somatic manifestations usually ignoring the cognitive and emotional domains. Also, 455 most of these tools were originally developed for use in the western world (except for SSQ and 456 AKUADS) and translated into other languages to be used in other countries. Therefore care should be 457 taken in interpreting the psychometric properties of the translated tools whose content might not 458 necessarily be the appropriate cultural or indigenous idioms for mental distress in that population.

459
Furthermore, the validation of some of these tools were done by health or mental health professionals in   determine whether the items all seem to be measuring the same thing. If the mode of administration of the 529 screening tool would be self-administered, the intraclass Correlation Coefficient (ICC) will be used for 530 test-retest or intra-rater reliability in addition to internal consistency (Cronbach's alpha). For interviewer-531 administered mode of administration, the inter-rater reliability (Interclass Correlation Coefficient) will 532 also be checked. Pearson correlation coefficient could also be used but while ICC gives consideration to 533 errors/biases that two raters might introduce into the measure, Pearson correlation has been found to be 534 theoretically incorrect in this aspect(37). Percentage agreement is also commonly used, as is kappa and 535 weighted kappa statistics. While ICC and Kappa yield identical results, ICC might be easier to calculate.

536
There is also the alternate form reliability testing which requires creating another version of the tool 537 although this is rarely used(37) . Reliability of the screening tool could also be tested or piloted in an 538 heterogeneous sample to evaluate the reliability of the tool to detect the defined attributes in people at risk 539 and those not at risk of developing MNS condition. Evaluation of the psychometric properties of the tools 540 should be done at similar settings in which the tools will be used. Since the screening tool will used 541 among general population or those at risk of MNS conditions to make a decision whether an individual 542 should be referred for further evaluation and possible treatment, the result of the reliability testing 543 (Cronbach's alpha, the intraclass or interclass correlation coefficient) higher than 0.7 might be considered 544 good reliability of the tool in the target population compared to diagnostic tools that might require higher 545 level of reliability. The ability of the tool to correctly identify those that are actually at risk of or with 546 MNS condition (sensitivity) and those that are actually not at risk of or without MNS condition

547
(specificity) could also be tested. Effort should be aimed at achieving high sensitivity of the measure to 548 minimize false positive result due to high stigma associated with MNS condition in the target population.

587
The development and the use of mental health screening tools should provide opportunities for training 588 lay community workers that will administer these tools. In addition, the use of the tools with the general 589 population or people affected by MNS conditions should provide information on awareness and improve 590 their knowledge of MNS issues.

591
Self-report tools usually include instructions or come with separate instruction manuals for completing 592 and scoring them and so does not require training. In the reviewed studies interviewers were trained on 593 how to administer the tools and score the responses. Mental health screening tools should have 594 accompanying instruction manual on how to complete and score them. These manual can be used for 595 training the users, especially the computer-based algorithm type.

596
The assessors should be trained to use various information sources such as observation, interview with the 597 person and those accompanying them (friends/family).

599
Sustainability will be the outcome of a tool that has been developed with adequate consideration of the 600 initial four pillars. These pillars should not be considered as isolated pillars but all inclusive. The 601 development of a holistic screening tool for a mental health intervention that is sustainable requires that 602 the tool is accepted and demand for use which will depend on the ease of use (length of the tools, 603 language used and cost). And finally the tool should be easy to adapt to different culture or setting.

604
Acceptability and Utility: One way to ensure acceptability and utility of a layperson mental health 605 screening is to engage the community from the design/development of the tools through its validation, 606 provide feedback on its use, contribution to evaluation of its effectiveness and review. This will 607 promote community ownership. Also the use of the tool should provide interpretation of results or 608 score and the next step to take. The data should be collected such that it can be easily integrated into

621
People should be screened free of charge therefore the tool should be licensed for free and easily 622 accessible.

623
The incorporation of the five pillars into the design of layperson screening tools is expected to improve 624 the reach, utility, and impact of the lay person screening process supporting a more responsive health 625 system. Therefore, in order to reach more people and create a more responsive mental health system, the 626 effort needs to include developing a holistic tool that is of cultural relevance. This as a matter of fact 627 cannot be achieved by a handful of people. Rather there is need to partner with different stakeholder or 628 partners for knowledge exchange, and pool different innovations and experiences that these diverse ideas 629 can bring. Capacity building is also not just focused on the researchers but also on the users of the tools.

630
The use of the tools should create opportunities for creating or raising awareness about mental issues,

631
increase the knowledge of the participants and their caregivers and help them in the understanding of the 632 determinants of mental health. The interviewer should also use different sources to gather as much 633 information as possible. There should be room for interaction between the clients and the assessor. There

634
should be opportunities to ask questions and also receive feedback. And the feedback should be used to 635 improve the design or the delivery of the tools.

637
This study did not do technical review and analysis of pooled psychometric data of the identified tools 638 because this is not the purpose of this review. The purpose of this study is to conduct a qualitative review 639 and examine the characteristics of mental health screening tools proposed for use by layperson as 640 described above. This study also did not review grey literature reports which could also have biased the 641 selection of the studies chosen for these review.

643
The needs of people living with MNS conditions are multifaceted and interlinked in a complex manner.

644
Inability to accurately identify these needs is a major contributor to the treatment gap in their 645 management. Screening tools can provide comprehensive information about these needs to inform holistic 646 care and responsive health system. Community layperson can reach more people in needs with 647 information and access to care. Developing a layperson screening tools is vital for integrated community-648 based mental health intervention. This study has proposed a holistic framework that can be consulted for 649 developing or selecting a layperson mental health screening instrument.

650
More research are needed to evaluate the practical application of this framework. Other research 651 questions unanswered by this study include whether or not there are impact or effect on measured 652 /outcome if self-report tools are used as interviewer-administered tool against being used as self-653 administered tool in order to know which version is more effective in low resource settings.