The Mental Health Outcomes of Income Generating Activity in the Friendship Bench Program

From the 2022 issue of Advocates' Forum

By Armando Garcia, Nicholus Tint Zaw, Gracia Lee, Adrian Hernandez


Abstract

In urban Zimbabwe, there are high correlations between mental health distress, HIV trauma, and economic hardship, leaving a significant number of people there living with a common mental disorder. At the same time, psychiatric treatment in hospitals and community settings is inadequate and inconsistent. The Friendship Bench is an intervention that seeks to fill this service gap through problem-solving therapy sessions with lay health workers and, for some advanced participants, an income generating activity. This protocol proposes the following questions: (1) What effect does participation in the income-generation activity of the Friendship Bench in Zimbabwe have on recipients’ (18-44 years of age) mental health outcomes? (2) With attention to age, gender, ethnicity, religion, and health status, what socio-demographic group experiences the most and least improvement in mental health? It uses a difference-in-difference approach to estimate the effects of the intervention’s economic component.


The Friendship Bench is a therapeutic intervention that was developed by Dr. Dixon Chibanda in Harare, Zimbabwe. While the intervention has evolved since its inception in 2006, many fundamental principles remain. The bench in the Friendship Bench program is literal—a wooden construct placed on the grounds of a primary care clinic where lay health workers (LHWs), primarily composed of grandmothers from the community, sit and talk with people in need (Abas et al., 2016). The intervention focuses on problem-solving, based on cognitive behavioral therapy (CBT). The LHWs are therefore trained in problem-solving therapy (PST), psychoeducation, basic counseling skills, and referral provision.

Beginning with the 2006 pilot study, Chibanda’s program has sought to fill an extreme treatment gap: 68% of Zimbabwe’s urban population living with a common mental disorder (CMD), high correlations between mental health distress, HIV trauma, and economic hardship (Verhey et al., 2020) with inadequate and inconsistent funding for human resources, psychiatric treatment, and mental health promotion in hospitals and community settings (e.g., only ten psychiatrists for a population of 13 million) (Mangezi & Chibanda, 2010; World Health Organization, 2020).

In practice, primary care clinics and community agencies first refer participants experiencing mental distress to the LHW on-duty. The LHW administers a psychosocial assessment, a mental health screener on the locally validated, Shona Symptom Questionnaire (SSQ-14), and conducts semi-structured PST. Finally, the LHW invites participants to return for five further sessions (Abas et al., 2016). If at any point or at the end of six sessions, recipients report acute mental distress or risk, they are assigned to the LHW’s supervisor for mental health aid, referral to psychiatric care, and follow-up care.

After attending four PST sessions, patients are eligible to participate in what is called the Circle Kubatana Tose—a support group led by women who had themselves taken part in the Friendship Bench program and went on to receive basic group management training. During weekly meetings, the group shares personal experiences while participants busily crochet a bag using recycled plastic materials. The crocheting keeps them active so they do not, as participants say, “think too much.” At the same time, in crafting bags that can be sold, they are developing a skill to generate income (Chibanda et al., 2016b; Circle Kubatana Tose, n.d.a).

Research has established the effectiveness of the Friendship Bench in improving mental health outcomes (Abas et al., 2016; Chibanda et al., 2017; Munetsi et al., 2018). In 2016, Chibanda et al. (2016b) published an RCT of the Friendship Bench. There were 573 patients in the study (286 in the intervention, 287 in the control group), all of whom were 18 years and older and screened positive for a common mental disorder (CMD) on the SSQ-14. Of the 573 participants, 495 (86.4%) were women with a median age of 33, 238 (41.7%) were living with HIV, and 521 (90.9%) completed follow-up at six months. The authors indicate that amongst individuals with CMDs, LHW-administered PST with education and support, compared to enhanced treatment as usual, resulted in improved mental health. Abas et al. (2016) conducted a process evaluation and found that the intervention was well-received by participants, provided rewarding work for the LHWs, and was financially sustainable over time. Lastly, in 2020, the intervention adapted to the COVID-19 pandemic by moving sessions to phone calls and WhatsApp. They also provided educational material to promote mental health and wellbeing for those quarantining. As of January 2022, the Friendship Bench has trained 1,391 LHWs, serviced 143,353 participants, implemented 88 active, Circle Kubatana Tose groups, and cumulatively resulted in a 61% decrease in depression and suicidal ideation, as well as a 60% increase in quality of life amongst recipients (Circle Kubatana Tose, n.d.a).

At this time, however, there has yet to be an evaluation of the impact of the making and selling of bags by members of the Circle Kubatana Tose groups. A comprehensive research and program evaluation that considers the psychosocial model of illness is needed. Given the degree to which studies have demonstrated that low socioeconomic status (SES) is associated with mental health disorders (Lund, Stansfeld & Da Silva, 2015), and that economic packages in other countries and have shown to be associated with improved economic and mental health outcomes (Lussi & Morato, 2012; McDonald & Bertram, 2018), an examination of the economic programs that strategically address people’s social determinants of mental health is warranted.

This protocol has been developed by authors trained in different disciplines (social work, public policy, and sociology) who found they shared an interest in the research and evaluation of mental health interventions in international settings. While none have worked with Chibanda or on Zimbabwean-based interventions, they bring unique insights from their cumulative work in countries like Singapore, Myanmar, and Mexico, which they use to think about the way that culture informs localized understandings, experiences, and access of health services and alternative healing practices. Particularly invested in global mental health, health policy, and economic sociology, they are drawn to the Friendship Bench’s model of task-shifting, scaled-up interventions in global mental health, and those with economic components designed to address socioeconomic determinants of (mental) health. They see the Friendship Bench as an important object of inquiry to analyze the role that economic activity plays in therapeutic settings.

This protocol is hypothetical, since the authors lack a Memorandum of Agreement with Chibanda and the Friendship Bench; however, it proposes a study to better understand the relationship of income generation activities and mental health outcomes in the context of global mental health. It proposes, to that end, the following questions: (1) What effect does participation in the income generation activity of the Friendship Bench in Zimbabwe have on the mental health outcomes of participants who are 18–44 years old? (2) With attention to age, gender, ethnicity, religion, and health status, what socio-demographic group experiences the most and least improvement in mental health? This proposal argues that if implemented, the evaluation’s results will demonstrate the extent to which targeted forms of behavioral and economic activation influence mental health outcomes. Moreover, it will provide insights as to how administrators in global mental health can utilize this approach to determine whether they should scale-up economic programs for particular socio demographic groups in their localities.


Literature Review

Eaton et al. (2014) studied the science of implementing health care interventions at large scales, as was the case with the Friendship Bench. They focused on how interventions can be integrated into routine care settings and determined that the benefits of those interventions are dependent on context when scaled up to meet the needs of more and more people. In this way, they showed the possibility of creating comprehensive medical services wherein medical, behavioral, and essential support services could operate as a single care system. When Abas et al. (2016) studied the impact of the Friendship Bench program, they noted its steady impact and delivery of care—highlighting that the patients interviewed reported benefitting from the approach and that the LHWs conveyed a feeling of making a positive contribution.

Through a mixed methods case study, Abas et al. (2016) were able to expand on the importance of context noted by Eaton et al. (2016). They found that the consistent presence and social position of the LHWs (how they were seen by the community, their maturity, their ethos), the intervention’s employment of a locally validated screening tool, and the accepted relevance of problem-solving therapy, were seemingly vital in the program’s success.

The comprehensive medical care suggested by Eaton et al. (2016) requires adherence to the psychosocial model of illness because there is an important correlation between a person’s social environment and their neurocognitive, physical, emotional, and psychological development. Utilizing Bronfrenbrenner’s ecological systems theory (EST) allows for understanding how social and environmental factors—at the levels of the micro-, mezzo-, macro- and exo-system—impact the individual (Bronfrenbrenner, 1994). For example, it is understood that financial deprivation can generate stress, stimulating the hypothalamic-pituitary-adrenal axis to secrete excessive levels of cortisol. Thus, low socio-economic status (SES) is associated with illness and mental health disorders, including schizophrenia, depression, personality disorders, and substance use (Lund, Stansfeld & Da Silva, 2015).

Amongst patients in the Friendship Bench, the experience of mental illness is multi-faceted and interrelated. Verhey et al. (2020) note these social-environmental influences. There is distress associated with socioeconomic factors such as poverty, food, and housing stability. Moreover, they reported that approximately 80% are living with HIV, and of those, 30% present with a CMD. The authors attribute this comorbidity to HIV stigma as infection is seen as society’s punishment for immoral behavior and lack of access to antiretroviral medication. This protocol’s authors see, therefore, that the Friendship Bench’s problem-solving approach is an effective tool to promote coping and solution-focused practices (Abas et al., 2016; Chibanda et al., 2017; Munetsi et al., 2018), and the gap in the literature is how the income generation component of the program affects the mental health outcomes of participants.

The literature on global mental health programming argues for the inclusion of economic components that help offset vulnerability to mental health distress (Kola et al., 2021; Maes, 2015; Abas et al., 2016). A systematic analysis in low- and middle-income countries showed that financial interventions alleviate stress responses in individuals with low socioeconomic status (SES), improving childhood and adult development, behavioral health, and self-esteem (Lund et al., 2011). For example, it has been shown that participation in the Mexican Oportunidades resulted in decreased depression in the women participants (Ozer et al., 2011). In the case of the Ecuadorian Bono de Desarrollo Humano, children who received social grants showed significant improvements in overall development (Fernald & Hidrobo, 2011). There is also substantial evidence on the impact of job training on reducing mental health distress (Lund, Stansfeld & Da Silva, 2015).

In the case of the crocheting that occurs in the Circle Kubatana Tose sessions, however, the movement of hands and focus on a task is more than skills training—it provides its own benefits through behavioral activation. Through interviews with the LHWs in the Friendship Bench program, Verhey et al. (2020) found the following: (1) the closest term for “trauma” is the Shona word njodzi, which describes the perception of temporally adverse events related to death or near-death exposure, serious injury, and sexual violence; (2) njodzi progresses from the incurrence of a traumatic event to kuendereramberi kwe njodzi (ongoing difficulties), such as generational poverty, chronic illness, and lack of medical care, and (3) this process is mediated by kufungisisa kwe njodzi (excessive thinking due to trauma).

The crocheting of the bags during the peer-led group sessions allows patients to experientially detach from memories, thoughts, and behavior related to stressors in their social environment. Participants say that while engaged in the task, they “don’t think too much” (Circle Kubatana Tose, n.d.a). Chibanda et al. (2016b) relate this to CBT’s cognitive restructuring. CBT has been widely adapted in global mental health with great success as patients are encouraged to replace negative self-talk with positive thoughts that bring them closer to enacting behaviors that are better aligned with their values (Zhang, Zhiang, Lin, & Huang, 2020; Andersen, et al., 2018; Boisits, et al., 2021).


Proposed Methodology

Study Design

Suggested is a quasi-experimental evaluation, with a nonequivalent control group and a pre-test-post-test design. The evaluation aims to answer the following questions: (1) What effect does participation in the income-generation activity of the Friendship Bench in Zimbabwe have on the mental health outcomes of participants 18-44 years of age? (2) With attention to age, gender, ethnicity, religion, and health status, what socio-demographic group experiences the most and least improvement in mental health? The nature of the program does not allow for the use of an RCT because randomization was enacted at its inception. While a quasi-experimental design cannot ascertain causation, it can provide useful information to examine the feasibility of producing future randomized studies to evaluate the impact of income generating activities on mental health, whether in Zimbabwe or other parts of the Global South.

The evaluation study specifies the consideration of two classes of study participants: those who participated in the income generation component and those who did not. Propensity score matching will be applied to ensure the groups share similar characteristics. This method will mitigate bias in random sampling and promote internal validity. The program effect will be examined using the Difference in Difference (DiD) approach. A detailed description of the statistical model appears in Appendix A: Model for DiD Approach.

Sample and Sampling Procedures

The list of the Friendship Bench’s program beneficiaries will serve as the primary source for the sample frame. The following inclusion criteria will be applied to screen potential participants for propensity score matching: (1) participants (18-44 years of age) who met clinical criteria for a CMD before entrance into the income generation activity, based on measures derived from the SSQ-14 (i.e., those who scored nine or higher); Chibanda et al. (2016a) has noted that this cutoff mark allows for a sensitivity and specificity level of 84% for depression and/or general anxiety, with the scale affording a high internal reliability (Cronbach α=0.74). (2) Participants will be recruited if they have been with the Friendship Bench for no more than two years and participated in the income generation component for a minimum of one year and a maximum of two. (3) Propensity score-matching will be conducted to ensure the comparison group shares the aforementioned clinical and programmatic characteristics. With regard to the exclusion criteria, the evaluative study will not include individuals who have been diagnosed or meet criteria for a serious mental disorder, including but not limited to: schizophrenia, bipolar, borderline personality disorder, and those who have suicidal ideation. The study will also not consider those who are minors or who have cognitive and/or developmental disabilities, as these groups deserve specific considerations.

Based on that sample frame, the list of beneficiaries will be organized into two strata: people who are and are not participating in the income generation activity. Propensity score matching will then be applied to identify similar sets of people from each group, pairing them based on their scores. From the available matches, 90 pairs will be randomly selected. The final sample size will therefore consist of 180 participants (90 for each study class, which represent 11% of the beneficiaries of the income generation component). A detailed description of this process is described in Appendix B: Methodology Flow Chart.

There might be some specific demographic characteristics (e.g., residents of particular geographical areas) that cannot be matched, though that is a marginal risk. The larger challenge will be ensuring that group compositions afford external validity with regard to the broader RCT and service population—that is, that the quasi-experimental design mirrors the RCT and total service population’s social composition. The evaluative study will conduct descriptive statistical analysis to continuously assess demographic characteristics, then apply stratified sampling as necessary to ensure external validity to the RCT and service population.

Recruitment

The evaluation team will work with the Friendship Bench’s staff to access the list of study participants and gather existing data on income generation activities. Other pertinent information will be requested to facilitate adherence to the inclusion and exclusion criteria. After that, the team will implement the propensity score method to consolidate matches from the two groups, and among that available list, 90 pairs will be randomly selected.

Data Collection Procedures

Data collection will occur in two ways. For the pre-test measurement, the team will use the Shona Symptom Questionnaire (SSQ-14) score that was acquired before they participated in the income generation component, pulling it from administrative data. For the post-test measurement, the evaluation team will adhere to the current protocol for acquisition of new measures. The LHWs will administer the SSQ-14 with the study participants. Thus, no further training processes or validation tests of the tool will be required.

Before pursuing any data collection activities (including obtaining secondary data), the informed consent-taking process will be done through program staff to explain the evaluation study purpose, process, the detailed usage of their therapeutic data, and potential threats and harms. Only participants who provide consent will be included in the final sample list. Their information will be exported from the Friendship Bench’s database to prepare the dataset for analysis. Carefully crafted data security and protection protocols will be followed throughout the study.

Hypotheses

There are two main hypotheses this evaluative study aims to test: (H1) participation in the income generation activity enhances mental health, and (H2) some social characteristics moderate effects on mental health outcomes if they participate in the income generation activity. With regard to the second hypothesis, the following social characteristics will be examined: age, gender, ethnicity, religion, and health status. They will demonstrate the relationship, if any, between the independent variables and mental health outcomes (dependent variable) (see the detailed description included Appendix D: Data Analysis Plan). Since this inquiry is exploratory, no data exists to propose more specific hypotheses. Nonetheless, if they are true, the data will have the potential to suggest a correlation between the income generating activity and mental health outcomes, particularly with regard to the population’s most vulnerable individuals. Additionally, they will identify areas of inquiry concerning groups for whom the intervention is not as effective.

Measures

The Shona Symptom Questionnaire (SSQ-14) was designed by Vikram et al. (1997) with key input from local stakeholders, health systems leaders, and users of mental health services in Zimbabwe, as the first indigenous mental health scale for use in sub-Saharan Africa (see Appendix C). It will serve as the survey instrument given its use of local idioms of distress to measure mental health. Some of its questions include: During the course of the week… “Did you sometimes think deeply or think about many things?”; “did you find yourself sometimes failing to concentrate?” Chibanda et al. (2016a) notes the scale affords a high internal reliability (Cronbach α=0.74). The authors, furthermore, report its high level of validity when used in primary health populations in Zimbabwe with a considerable prevalence of HIV. They recommend it for research and clinical care capacities to identify individuals who may benefit from treatment.

Potential Challenges and Solutions

While the COVID-19 pandemic has significantly altered the Friendship Bench’s activities, information on how it has been adapted to the varying conditions of the crisis has not been made fully public (The Friendship Bench, n.d.b). There might be missing data, a disruption of services, and alternate implementation mechanisms that need to be accounted for in the data analysis. Hatry (2014) notes that an evaluation study can take a temporal approach by focusing on participants enrolled during a certain time period. In this case, these would be individuals who were enrolled in the income generation component from 2018 to the first quarter of 2020, provided they otherwise meet the inclusion criteria discussed in previous sections.

Another challenge will be ensuring that group compositions can provide external validity with regard to the broader RCT and service population. The evaluative study will conduct descriptive statistical analysis to continuously assess the groups’ demographic characteristics, then apply stratified sampling as needed to ensure external validity to the RCT and service population.


Conclusion

Zimbabwe has very limited mental health services and the Friendship Bench has proven to contribute to mitigating treatment gaps. The program includes PST sessions carried out by LHWs, who are trained and supervised by mental health professionals, as well as an optional income-generating component to alleviate stressors associated with CMDs. There have been no studies on the Friendship Bench’s income-generating activity on beneficiaries. Hence, this study aims to address this gap by using a quasi-experimental design with a nonequivalent control group and a pre-test-post-test design. This evaluation also seeks to identify the socio-demographic groups that have benefitted the most and the least, with respect to mental health outcomes. As an exploratory study, this type of study can facilitate future research and project planning to ensure that key groups are positively impacted by the program.


Authors Note

No evaluation without programs and no programs without evaluation! The global mental health movement has been characterized by a disruption of the West’s legacies of colonialism, imperialism, and hyper-capitalism on continents beyond Europe. For over 500 years, indigenous groups in Africa, the Americas, and Asia have suffered due to the material conditions brought about by slavery, conquest, and systemic racism. This paper is therefore built on two implicit, yet fundamental, principles. First, the incorporation of strategies to address social determinants in global health is a required response to white supremacy. Second, knowledge is power and praxis; which is to say, knowledge is filtered through the systems that sustain white supremacy, yet there is power in representation, cultivation, cultural humility, and participatory action towards the disruption of social inequities. It is in this sense that we invite you to think about what this means in your practice.


References

Abas, M., Bowers, T., Manda, E., Cooper, S., Machando, D., Verhey, R., Lamech, N., Araya, R., & Chibanda, D. (2016). “Opening up the mind”: Problem-solving therapy delivered by female lay health workers to improve access to evidence-based care for depression and other common mental disorders through the Friendship Bench Project in Zimbabwe. International Journal of Mental Health Systems, 10(39), 1–8.

Andersen L. S., Magidson J. F., O’Cleirigh C., Remmert. J. E., Kagee, A., Leaver, M., Stein, D. J., Safren, S. A., Joska, J. (2018). A pilot study of a nurse-delivered cognitive behavioral therapy intervention (Ziphamandla) for adherence and depression in HIV in South Africa. Journal of Health Psychology, 23(6), 776–787.

Boisits, S., Abrahams, S., Schneider, M., Honikman, S., Kaminer, D., & Lund, C. (2021). Developing a task-sharing psychological intervention to treat mild to moderate symptoms of perinatal depression and anxiety in South Africa: a mixed-method formative study. International Journal of Mental Health Systems, 15(23), 1–14.

Bronfenbrenner, U. (1994). Ecological models of human development. In International Encyclopedia of Education, Vol. 3, 2nd Ed. Oxford: Elsevier. Reprinted in: Gauvain, M. & Cole, M. (Eds.), Readings on the development of children, 2nd Ed. (1993, pp. 37–43). Freeman.

Chibanda, D., Cowan, F., Verhey, R., Machando, D., Abas, M., & Lund, C. (2017). Lay health workers’ experience of delivering a problem-solving therapy intervention for common mental disorders among people living with HIV: A Qualitative Study from Zimbabwe. Community Mental Health Journal, 53, 143–153.

Chibanda, D., Verhey, R., Gibson, L. J., Munetsi, E., Machando, D., Rusakaniko, S., Munjoma, R., Araya, R., Weiss, H. A. and Abas, M. (2016a). Validation of screening tools for depression and anxiety disorders in a primary care population with high HIV prevalence in Zimbabwe. Journal of Affective Disorders, 198, 50–55.

Chibanda, D., Weiss, H.A., Verhey, R., Simms, V., Munjoma, R., Rusakaniko, S., Chingono, A., Munetsi, E., Bere, T., Manda, E., & Abas, M. (2016b). Effect of a primary care–based psychological intervention on symptoms of common mental disorders in Zimbabwe. JAMA, 316(24), 2618–2626.

Columbia Public Health. (2022). Public health methods: Difference-in-difference estimation. https://www.publichealth.columbia.edu/research/population-health-methods/difference-difference-estimation

Eaton, J., De Silva, M., Rojas, G., & Vikram, P. (2014). Scaling-up services for mental health. In V. Patel, H. Minas, A. Cohen, & M. J. Prince, Global Mental Health: Principles and Practice (pp. 297–334). Oxford University Press.

Fernald, L. C. & Hidrobo M. (2011). Effect of Ecuador’s cash transfer program (Bono de Desarrollo Humano) on child development in infants and toddlers: a randomized effectiveness trial. Social Science & Medicine, 72(9), 1437–46.

Friendship Bench. (n.d.a). CKT: Circle kubatana tose. Retrieved May 2021 from https://www.friendshipbenchzimbabwe.org/ckt

Friendship Bench. (n.d.b). Covid-19. Retrieved June 2021, from https://www.friendshipbenchzimbabwe.org/covid19response

Friendship Bench. (n.d.c). Friendship bench. Retrieved March 2022, from https://www.friendshipbenchzimbabwe.org

Hatry, H. P. (2010) Using agency records. In H. Wholey, H. Hatry, & K. Newcomer (Eds.), Handbook of practical program evaluation (pp. 243–261). Jossey-Bass.

Kola, L., Kohrt, B. A., Hanlon, C., Naslund, J. A., Sikander, S., Balaji, M., Benjet, C., Cheung, E. Y. L., Eaton, J., Gonsalves, P. & Hailemariam, M. (2021, February 24). COVID-19 mental health impact and responses in low-income and middle-income countries: reimagining global mental health. The Lancet Psychiatry. https://doi.org/10.1016/S2215-0366(21)00025-0

Lund, C., Stansfeld, S., & De Silva, M. (2014). Global mental health: principles and practice (pp. 116–136). In V. Patel, H. Minas, A. Cohen, & M. J. Prince (Eds.), Social Determinants of Mental Health. Oxford University Press.

Lund, C., De Silva, M., Plagerson, S., Cooper, S., Chisholm, D., Das, J., Knapp, M. & Patel, V. (2011). Poverty and mental disorders: breaking the cycle in low-income and middle-income countries, Lancet, 378(9801), 1502–14. https://doi.org/10.1016/S0140-6736(11)60754-X

Lussi, I. A. D. O., & Morato, G. G. (2012). The meaning of work for users of mental health services included in income generation projects linked or not to the movement of solidarity economy. Cardenos de Terapia Occupacional, 20(3), 369–380.

Maes, K. (2015). Task-shifting in global health: mental health implications for community health workers and volunteers (pp. 291–308). In B. J. Kohrt, & E. Mendenhall (Eds.), Global Mental Health: Anthropological Perspectives. Routledge.

Mangezi, W., & Chibanda, D. (2010). Mental health in Zimbabwe. International Psychiatry, 7(4), 93–94.

McDonald, S., & Bertram, M. (2018). Job creation through income generation: an evaluation of re-cover, a decorating project developed with forensic mental health service users. The Journal of Mental Health Training, Education and Practice, 13(3), 148–156.

Munetsi, E., Simms, V., Dzapasi, L., Chapoterera, G., Goba, N., Gumunyu, T., Weiss, H. A., Verhey, R., Abas, M., Araya, R., & Chibanda, D. (2018). Trained lay health workers reduce common mental disorder symptoms of adults with suicidal ideation in Zimbabwe: A cohort study. BMC Public Health, 18(227), 1–7.

Ozer E. J., Fernald L. C., Weber A., Flynn E. P., VanderWeele T. J. (2011). Does alleviating poverty affect mothers’ depressive symptoms? A quasi-experimental investigation of Mexico’s Oportunidades program. International Journal of Epidemiology, 40(6), 1565–1576.

Verhey, R. (2016, April). Talking your way back into life: Zimbabwe’s Friendship Bench. The Friendship Bench.

Verhey, R., Chibanda, D., Vera, A., Manda, E., Brakarsh, J., & Seedat, S. (2020). Perceptions of HIV-related trauma in people living with HIV in Zimbabwe’s Friendship Bench Program: A qualitative analysis of counselors’ and clients’ experiences. Transcultural Psychiatry, 57(1), 161–172.

Vikram, P., Simunyu, E., Gwanzura, F., Lewis, G., & Mann, A. (1997). The Shona Symptom Questionnaire: The development of an indigenous measure of common mental disorder in Harare. Acta Psychiatrica Scandinavaca, 95, 469–475.

World Health Organization (WHO). (2020). Zimbabwe: WHO special initiative for mental health situational assessment. https://www.who.int/docs/default-source/mental-health/special-initiative/who-special-initiative-country-report---zimbabwe---2020.pdf

Zhang, H., Zhiang, D., Lin, H., & Huang, L. (2020). Problem Management Plus in the treatment of mental disorders in patients with multiple myeloma. Supportive Care in Cancer, 28, 4721–4727.


Appendix A

Statistical Model for Difference-in-Differences Approach

The Difference-in-Difference (DiD) statistical model analyzes quasi-experimental designs. It uses longitudinal data between treatment and control groups to estimate a causal effect. DiD is ideal for studies where randomization is not possible because it relies on a less strict assumption of exchangeability. The approach removes potential biases in post-intervention comparison between the treatment and control group that could be due to permanent differences between them and/or alternate causes of the outcomes. For more on DiD, please refer to Columbia Public Health (2022).

The program’s impact on mental health outcomes will compare the differences between treatment and comparison groups, before and after the program intervention.

Table 1: Difference-in-Differences Approach

 

Treatment

Comparison

Difference-in-Differences

Before

A

B

 

Intervention

After

C

D

 

Difference

(YC -YA)

(YD -YB)

(YC -YA) - (YD -YB)

 

The following DiD regression model will be applied to investigate the significance level and magnitude of program effects on mental health outcomes. The detailed model description is explained below.

Y = β0 + β1Treated * After + β2 Treated + β3 After + u

Where:

β0=YB = expected mental health outcomes of comparison group before intervention

β2=(YA- YB) = expected differences of mental health outcomes between treatment and comparison groups before the intervention

β3=(YD - YB) = expected differences of mental health outcomes of comparison group before and after intervention

β1=[(YC -YA) - (YD -YB)] =Difference-in-Difference


Appendix B

Methodology Flow Chart


Appendix C

Shona Symptom Questionnaire–14 (Vikram et al., 1997)


Appendix D

Data Analysis Plan

The data processing and analysis tasks will be carried out in R-Statistical Package. The descriptive summary statistics of study participants' characteristics (age, gender, ethnicity, religion, health status, Shona Symptom Score, etc.) will comprise mean or proportion-based categories depending on the variable type and a 95% Confidence Interval. All missing information will be properly coded to distinguish between missing data or survey error. The frequency table of missing values by each variable will also be added in the analysis to interpret the sensitivity of questions (e.g., regarding questions in which the participants were less engaged or responsive).

The evaluation of research question one (the program effect of the income-generation activities on mental health outcomes) will be performed based on a DiD regression model.

Y = β0 + β1Treated * After + β2 Treated + β3 After + u

The β1, in the above model, will explain the changes of Shona Symptom Score of the treatment group after receiving the income-generation activities compared to the group who did not. That value will explain the magnitude of changes in the Shona Symptom Score due to the program impact.

For the second hypothesis, the sample participants from the treatment stratum will be applied, because the purpose of the analysis aims to identify the moderating variables (beneficiaries’ characteristics) that influence the most or least influences on outcomes. The application of the test model will be varied based on the type of independent variable. The sample analysis plan is explained in the following table:

Table 2: Statistical Analysis Plan

 

Dependent Variable

 

Independent Variable(s)

 

Statistical Test1

Shona Symptom Score

(continuous variable)

Beneficiaries Age

(continuous variable)

Correlation, or

Simple linear regression

Gender

(Categorical variable)

Chi-square test

Ethnicity

(Categorical variable)

Chi-square test

Religion

(Categorical variable)

Chi-square test

Health status

(Categorical variable)

Chi-square test

1 https://stats.idre.ucla.edu/other/mult-pkg/whatstat/


Author Biography

Armando Garcia (he/him & they/them) graduated in 2021 with a Master of Arts in Social Work and a Certificate in Health Administration & Policy with a Concentration in Global Health from the Crown Family School of Social Work, Policy, and Practice. Their clinical and research interests include HIV prevention, global mental health, contextual behavioral practices, Latinx health, LGB, trans & GNC health, immigrant health, and human rights.

Nicholus Tint Zaw (he/him) graduated in June of 2022 with a Master of Public Policy from the University of Chicago Harris School of Public Policy with a Concentration in International Development and Data Analysis for Public Policy. His policy interests include the intersection of social protection, global health, and financial inclusion. 

Gracia J. Lee (she/her) graduated in 2021 from the Master of Arts Program in the Social Sciences. Her research interests are in economic and cultural sociology.

Adrian Hernandez earned his Master of Arts degree from the Crown Family School of Social Work, Policy, and Practice in 2021.