Community Health Worker Role Expansion from HIV to Maternal, Newborn, And Child Health In Iringa, Tanzania: A Mixed Methods Case Study

Background Community health workers (CHWs) include many types of community-based health cadres. There is growing evidence that CHW interventions can improve a variety of health outcomes, and expansion of CHW programs is an important strategy for achieving Universal Health Coverage. With widespread shortages in the skilled health workforce, CHWs are being tasked with more complex and varied responsibilities, but whether CHWs should focus on one health area (“vertical” approach) or multiple health areas (“horizontal” approach) is still being debated. Central to this debate is appropriate CHW workload, with overload being a key risk. This dissertation explores the expansion of a single role CHW program focused on HIV to a dual role program of HIV and maternal, newborn, and child health (MNCH) in Tanzania. This research was designed to help guide implementation decisions by Tanzania’s Ministry of Health as they develop a paid, national CHW cadre that will integrate work across multiple health areas. Methods A mixed methods case study examined an existing cadre of HIV-focused CHWs in two districts of Iringa Region: Iringa Rural and Kilolo. CHWs received additional MNCH training and supervision along with MNCH data reporting requirements. In Manuscript 1, interrupted time series analyses used routine monthly data on HIV and MNCH visit workload to assess whether CHWs could effectively provide additional MNCH services without compromising their HIV workload. Interrupted time series methods were again utilized in Manuscript 2 to evaluate whether MNCH health promotion by dual role CHWs increased the number of deliveries occurring at health facilities. In Manuscript 3, qualitative data from in-depth interviews with CHWs, supervisors, and program managers explored how CHWs managed additional MNCH responsibilities and why the integrated program worked (or not). Field observations documented implementation differences between HIV and MNCH roles among CHWs. The workload balance of HIV and MNCH household visits was examined, along with task prioritization, and drivers of feasibility, acceptability, and adoption of the integrated CHW model. Results Manuscript 1 presents quantitative evidence that HIV-MNCH integration by CHWs is feasible and confirmed that CHWs began conducting MNCH-related household visits shortly after training. However, in the immediate month of the MNCH intervention, an initial 6 to 9% drop in the average number of monthly HIV household visits among dual role CHWs was evident, relative to what would have been expected in the absence of the MNCH intervention (Iringa Rural: aIRR=0.94, p=.10; Kilolo: aIRR=0.91, p=.04). There was no significant difference between single and dual role CHWs in the trajectories of monthly HIV household visits before and after adding MNCH duties. Manuscript 2 reports no significant change from baseline in the average number of facility deliveries observed at intervention health centers and dispensaries, relative to the expected change in the absence of the MNCH intervention. At the hospital level, there was a significant 16% increase in monthly deliveries at each district hospital, moving from an average of 202 to 234 deliveries in Iringa Rural and from 167 to 194 deliveries in Kilolo during the pre-intervention and intervention periods, respectively. Total facility deliveries were relatively stable over time at the district level, increasing about 1%, yet the relative change in the proportion of hospital deliveries out of total facility deliveries significantly increased by approximately 17% in Iringa Rural and 15% in Kilolo (p