Date of Award
Spring 2024
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Public Health
First Advisor
Davis, J. Lucian
Abstract
Background: Community health workers (CHWs) play a key role in expanding access to health services. Providing CHWs with high-quality and relevant training is critical to facilitate their performance. Communities of Practice (CoPs) are a low-cost implementation strategy to promote continuous learning. CoPs are groups of professionals with a common goal who meet regularly, support each other, share knowledge, and innovate solutions. CoPs have been used within the health field to promote the delivery of evidence-based care by healthcare workers; however, there are a few gaps in our understanding of their utility. For example, few studies have assessed the effectiveness of CoPs in improving clinical outcomes among the target patient population. Furthermore, the majority of CoPs in the healthcare setting have included physicians and nurses in high-income countries; additional work on how CoPs can be used for CHWs working in low-resource settings is warranted. There is also a need to determine how CoPs function to improve CHW performance, including antecedents that are required for CoPs to be established and mechanisms of action, in order to optimize their implementation. This dissertation aims to address these gaps by examining how CHW CoPs focused on improving delivery of contact tracing for tuberculosis (TB) function in Uganda. Methods: The aim of Chapter 1 was to develop a conceptual model for how CoPs function in preparation for an upcoming stepped-wedge, cluster-randomized trial. Our research group implemented a CoP including CHWs focused on providing home-based contact tracing services for TB in Kampala, Uganda. We conducted semi-structured interviews with all participants of the CoP to understand their experiences. I analyzed these interviews using abductive analysis, a qualitative technique for merging social science theory with empirical data. In order to test this conceptual model of how CHW CoPs function, I culturally adapted and validated relevant scales to measure these constructs during the stepped-wedge, cluster randomized trial of CHW CoPs in Chapter 2. Specifically, we adapted the 10-item General Self-Efficacy Scale through cross-cultural discussions within our multidisciplinary research team, translation from English into Luganda and back-translation into English, and six cognitive interviews with CHWs. We administered the adapted scale in a staged, two-part cross-sectional study, including a total of 147 CHWs. We evaluated the psychometric properties of the scale by conducting exploratory factor analysis, assessing convergent and discriminant validity, and calculating Cronbach’s alpha. In Chapter 3, I conducted a convergent mixed methods study of a stepped-wedge, cluster-randomized trial of CHW CoPs to test self-efficacy and social support as mechanisms of action of CHW CoPs, as well as explore antecedents and additional mechanisms qualitatively. We used the validated scales from Chapter 2 to collect quantitative data on self-efficacy and social support before and after the CoPs were implemented. We used the product of coefficients approach to evaluate if self-efficacy and/or social support mediated the effects of the CoPs on the count of contacts completing TB evaluation. We also conducted focus groups four months after the CoP began and analyzed them using thematic analysis to explore antecedents and additional mechanisms of action. Results: In Chapter 1, all eight CHWs in the CoP agreed to participate in the interviews. We found that the CoP functioned to enhance self-efficacy and social support among CHWs. For example, CHWs described that their weekly CoP meetings provided an opportunity to learn from their peers’ successes and failures and receive support from each other, which enhanced their confidence dealing with difficult situations in the field. In Chapter 2, we aimed to validate the General Self-Efficacy Scale to test self-efficacy as a mediator in the stepped-wedge, cluster-randomized trial. Exploratory factor analysis yielded a three-factor solution, which had good model fit (standardized root mean square residual=0.07) and explained 53.4% of the variance. We found evidence of convergent validity, as scores for the total scale were positively correlated with years of experience (r=0.48; p<.001) and perceived social support (r=0.39, p<.001). Scores were also higher among those with higher educational attainment in one-way analysis of variance and Bonferroni-corrected post-hoc tests (F[2,72]=9.16, p<.001). We also found evidence of discriminant validity, as scores for the total scale were not correlated with age (r=-0.07, p=.55), in agreement with literature showing that general self-efficacy is an age-independent construct. The internal consistency of the adapted scale was within the acceptable range for a pilot study (Cronbach’s α=0.61). In Chapter 3, all 27 CHWs from the stepped-wedge, cluster-randomized trial agreed to participate in the surveys and focus groups. In the focus group discussions, we found that scheduling availability, external motivators (e.g., refreshments at the meetings), a champion, and a shared tool for defining their practice (i.e., feedback reports) were important antecedents of CHW CoPs. In the mediation analysis, neither CHW-reported self-efficacy nor social support were mediators of the relationship between the CoP and TB contact tracing completion (estimate=0.01, 95%CI: -0.06, 0.09, p=.81; estimate=0.04, 95%CI: -0.03, 0.14, p=.32; respectively). However, in the focus groups, CHWs did report perceiving improved collective efficacy, social support, and social accountability from the CoP, which enhanced their abilities and motivation to deliver home-based TB contact tracing. Thus, our findings suggested that CoPs may function by improving collective efficacy, rather than self-efficacy, and the quality, instead of the quantity, of social support. Conclusions: This dissertation produces knowledge about how to design, implement, and maintain effective CHW CoPs in low-resource settings. Other programs may use the results from this dissertation to inform the implementation of CHW CoPs in other similar settings. For example, these programs may wish to integrate a shared tool for defining the CoP’s practice (e.g., feedback reports) and innovate ways for CHWs to easily convene in the field, such as through virtual meetings. CHW CoPs may also benefit from organizing opportunities to support each other in the clinic as well as the field, e.g., by creating a mobile group chat on a messaging platform, and establishing accountability structures for performance, e.g., through group audit-and-feedback.
Recommended Citation
Hennein, Rachel, "Communities of Practice to Improve the Delivery of Contact Tracing for Tuberculosis by Community Health Workers: Antecedents and Mechanisms of Action" (2024). Yale Graduate School of Arts and Sciences Dissertations. 1504.
https://elischolar.library.yale.edu/gsas_dissertations/1504
