Scaling up deworming programs: implementation science approaches to guide the transition from school-based to community-wide mass drug administration for soil-transmitted helminths

Thesis (Ph.D.)--University of Washington, 2022Soil-transmitted helminths (STH) are a group of intestinal parasites that affect 1.5 billion people globally. Pre-school and school-age children are at most-risk of various nutritional, cognitive and physical morbidities associated with chronic or high intensity STH infections. Global guidelines recommend controlling STH-associated morbidity amongst pre-school age and school age children via school-based mass drug administration (MDA) programs. However, recent clinical trial evidence suggests that an expanded community-wide MDA strategy may lead to more substantial reductions in intensity and prevalence than school-based MDA, while mathematical models suggest that community-wide MDA may achieve STH interruption transmission, thus reducing the risk of community reinfection. The primary objective of this research is to utilize implementation science research approaches to guide the potential policy transition from school-based to community-wide MDA for STH. We conducted a series of primary analyses using data from the DeWorm3 Project, a hybrid Type I cluster randomized trial in Benin, India, and Malawi assessing the feasibility of interrupting STH transmission via community-wide MDA. We applied three rigorous analytical methods guided by DeWorm3 implementation science research activities that aim to develop a scalable and sustainable community-wide MDA delivery model. In the first study, we utilized social network analysis (SNA) to map all stakeholders and to systematically identify the most influential stakeholders within school-based and community-wide MDA networks. We also sought to define how network dynamics may impact implementation and scale-up across these delivery platforms. Analyses of stakeholder maps revealed an expansive network of hundreds of stakeholders in both delivery platforms, a majority of which held a positive attitude towards community-wide MDA. We found that stakeholders who were primarily responsible for program delivery exerted a high degree of influence on intervention scale-up (i.e., implementation at-scale), while stakeholders who were in charge of policymaking & program leadership often controlled resource flow. Across networks for both community and school MDA programs, we observed a high number of connections; however, these connections were not concentrated on one single stakeholder, as indicated by low network centralization scores was shared equally across stakeholders. Additionally, low network density scores suggested these networks had poor overall connectedness across administrative levels. Understanding these network dynamics provides important decision-making evidence critical for launching and scaling MDA programs, such as identifying program champions to support policy change and drive MDA uptake, as well as securing sufficient program resources for optimized MDA scale-up, and mitigating implementation bottlenecks to drive effective intervention delivery. In our second study, we applied coincidence analysis (CNA), a novel cross-case analytical method, to identify any necessary and/or sufficient combinations of intervention delivery activities related to drug supply chain, implementer training, community sensitization, intervention duration, and implementation context that resulted in high coverage of community-wide MDA. Using pooled implementation data from three sites and across six intervention rounds, we found that efficient duration of MDA delivery (within 10 days) uniquely emerged as a fundamental component for achieving high MDA coverage when combined with other influential activities, including a conducive implementation context, early drug arrival before planned MDA onset, or a flexible community sensitization strategy. No individual activity proved sufficient by itself for producing high MDA coverage. Findings from this analysis demonstrate how effective MDA delivery can be achieved with flexible implementation strategies that incorporate various combinations of influential intervention components. Our third study employed Mokken scaling, a nonparametric item response theory (IRT) method to assess the measurement properties of the Structural Readiness for Scale-up Survey, an instrument aimed to assess health system readiness to transition from school-based to community-wide MDA delivery. The 36-item instrument included five hypothesized domains of readiness – change commitment (n=8 items), change efficacy (n=13 items), capacity (n=5 items), organizational structure (n= 6 items), and flexibility (n=4 items) – as identified in several organizational readiness and change management theories. Findings from our Mokken analysis revealed the final survey structure including eight items within two unidimensional subscales – change efficacy (n=5 items) and organizational structure (n=3 items). Our findings providing a potential starting point for assessing health system readiness to scale-up mass deworming programs. We discuss the challenges of using organizationally-oriented tools to measuring readiness within complex, multi-organizational health systems in low-and middle-income (LMIC) countries and conclude that additional qualitative validation is necessary to further refine the survey items and develop applicable readiness instruments for use in LMIC settings. This dissertation contributes to the growing global health implementation science literature by evaluating key factors influencing successful delivery and scale-up of community-based programs such as MDA, including strong stakeholder networks, valid pre-implementation readiness assessments, and implementation plans that are context-specific. Study findings provide useful insights for policy makers and implementers in STH-endemic countries aiming to interrupt STH transmission and reduce infection burden by transitioning from school-based to community-wide MDA