Resilience outcomes and associated factors among workers in community-based HIV care centres during the Covid-19 pandemic: A multi-country analysis from the EPIC program

The study investigates resilience and associated mental health outcomes among community health workers (CHW) engaged in HIV community-based organizations during the COVID-19 pandemic. The context underscores widespread mental health burdens on healthcare workers (HW) globally, including high stress, anxiety, and post-traumatic symptoms, with limited evidence focused on CHW. CHW play crucial roles in HIV care and prevention, particularly for key populations facing stigma and barriers to services, and during COVID-19 they shouldered a dual burden: maintaining HIV services while responding to pandemic needs. Prior work suggests resilience may protect against anxiety and depression, but evidence among CHW is scarce. The objective is to assess resilience levels and identify factors associated with low resilience among CHW in Benin, Colombia, Guatemala, and Spain during 2021.

The paper synthesizes evidence showing substantial mental health impacts of COVID-19 on healthcare workers, with nurses often at higher risk for adverse outcomes. Studies from the US, Spain, and Indonesia report high stress, anxiety, and post-traumatic symptoms, and a negative correlation between anxiety and resilience. Literature on CHW highlights their essential frontline roles in low- and middle-income countries and during COVID-19, including community education and support, but also challenges such as inadequate PPE, financial strain, supply disruptions, and workload. The HIV response historically centers community leadership, and COVID-19 disrupted HIV/HCV services, prompting community-based organizations to implement innovations to sustain care. A study from Mali found high rates of depression, anxiety, and insomnia among CHW early in the pandemic. Evidence points to the importance of self-efficacy and coping strategies in mitigating psychological impacts, and to discrimination related to COVID-19 as a contributor to mental distress. This study addresses gaps by focusing on CHW in HIV community settings across multiple countries.

Design: Multi-country, cross-sectional, community-based quantitative survey within the EPIC program (Coalition PLUS), conducted March–December 2021 among CHW in Benin, Colombia, Guatemala, and Spain. Data collection was anonymous via online questionnaires (Voxco) disseminated by local coordinators; some sites used face-to-face or phone administration depending on local context. Sampling: Convenience sampling of CHW (employees or volunteers) from participating community-based organizations (CBOs) and partner networks. Instruments: Psychological well-being and resilience measured using validated scales: Brief Resilience Scale (BRS, 6 items; mean 1–5; categorized as low 1.00–2.99, normal 3.00–4.30, high 4.31–5.00), PHQ-9 for depression (0–27), and GAD-7 for anxiety (0–21). Scores were standardized to 0–100 for comparability; computed if at least half of items were completed. Explanatory variables: Sociodemographics (age, gender identity, education, residence), identification with key populations (e.g., MSM, PLHIV), financial change due to COVID-19; perceptions of COVID-19 risk and severity; organizational impact (functioning, activity cessation); work experiences (overall impact, workload, new professional skills, workplace protection, necessity of work, relations with colleagues, impact on personal life including separation from loved ones, training entourage in prevention); experiences of discrimination/rejection due to COVID-19 risk; frequency of exchanges with beneficiaries; time elapsed since pandemic start to survey completion. Analysis: Descriptive statistics for standardized scores (median, IQR). Associations between resilience and psychological outcomes assessed with Spearman correlation. For resilience categorization, normal and high were grouped. Group comparisons (low vs normal/high resilience) used Chi-square (categorical) and Mann–Whitney (continuous). Logistic regression identified factors associated with low resilience; variables with p < 0.20 in univariable analyses were candidates for multivariable modeling; backward selection via Likelihood Ratio tests retained variables with p < 0.05. Models adjusted for age, time since pandemic onset, and country. Psychological scores were excluded as covariates due to high correlation with resilience. Ethics: Local ethical approvals obtained in Benin, Colombia (two committees), Guatemala, and Spain. Informed consent obtained electronically.

Sample: 295 CHW responded: Benin n=56 (21.0%), Spain n=133 (49.8%), Guatemala n=19 (7.1%), Colombia n=59 (22.1%). Median age 34 years (IQR 28–44), 50.6% women, 85.1% higher education; 35.7% identified as MSM, 13.9% as PLHIV, 24.4% other key populations. Standardized scores: Resilience median 58.33 (IQR 50.0–75.0; n=267); Depression 18.52 (IQR 7.4–33.3; n=282); Anxiety 19.05 (IQR 4.8–33.3; n=274). Correlations: Resilience was negatively correlated with anxiety (rho = -0.49, p < 0.001, n = 266) and depression (rho = -0.44, p < 0.001, n = 267). Resilience distribution: 24.0% low resilience (64/267); 76.0% normal/high resilience (203/267). Univariable comparisons: No significant differences by age, gender identity, or education. Low-resilience CHW reported less workplace protection from COVID-19 (71.0% vs 83.9%, p = 0.024), more discrimination/rejection in public spaces (42.6% vs 23.0%, p = 0.011), and other trends (e.g., fewer new professional skills, p = 0.150). Multivariable model (n=257, adjusted for age, time since pandemic onset, and country): - Urban residence vs semi-urban/rural: aOR 2.29 (95% CI 1.11–4.72), p = 0.024. - No development of new professional skills since the COVID-19 crisis began: aOR 2.01 (1.03–3.93), p = 0.041. - Did not separate from loved ones to avoid exposing them to COVID-19: aOR 4.69 (1.69–13.00), p = 0.003. - Experienced discrimination/rejection due to COVID-19 risk from people in public spaces (agree vs disagree): aOR 2.62 (1.20–5.72), p = 0.015. Country variable was not significant in adjusted models.

Findings indicate that most CHW in HIV community-based settings demonstrated normal or high resilience during COVID-19, with resilience inversely associated with depression and anxiety, suggesting a protective effect on mental health. Urban residence was linked to greater odds of low resilience, potentially reflecting higher exposure to COVID-19 and increased work strain in urban contexts. Two self-efficacy-related elements emerged as protective: developing new professional skills during the crisis and taking protective actions (e.g., separating from loved ones). These may enhance perceived control and coping capacity, contributing to resilience. Conversely, experiencing COVID-19-related discrimination in public spaces was associated with lower resilience, aligning with literature on stigma’s detrimental mental health effects among health workers. Despite multinational data, resilience levels did not significantly vary by country in adjusted models, suggesting commonalities in CHW experiences across settings. The results underscore the importance of rapid training (e.g., COVID-19 prevention), organizational support, and anti-stigma measures to bolster CHW resilience and mental health during health emergencies, and highlight the need to consider system-level resilience to sustain individual coping over time.

This multi-country community-based study documents predominantly normal to high resilience among CHW working in HIV services during the COVID-19 crisis and confirms resilience’s protective association with depression and anxiety. Modifiable factors linked to lower resilience include lack of skill development, absence of protective coping actions (e.g., separation from loved ones), urban residence, and experiences of discrimination. Programs should prioritize rapid training to increase CHW skills and self-efficacy, ensure workplace protection, implement anti-discrimination initiatives, and integrate CHW support into broader public and private health system responses. Future research should: assess system-level resilience alongside individual factors; explore causal pathways between training/self-efficacy and resilience; evaluate targeted interventions to enhance CHW resilience; and include longitudinal designs with baseline measures to track changes over time.

• Convenience sampling and small country subsamples limit representativeness and preclude robust cross-country structural comparisons. - Lack of refusal data and potential biases due to computer literacy/availability may affect sample profile. - Many CHW were from Coalition PLUS network organizations that may have received support during the crisis, potentially elevating resilience. - No baseline (pre-pandemic) resilience measures; low resilience during the study may reflect pre-existing levels. - Flexible EPIC design with optional sections reduced available data for some analyses and limited inclusion of additional countries/themes. - Focus on individual-level resilience may under-emphasize structural/system-level determinants and constraints.