TY - JOUR T1 - 30 years of polio campaigns in Ethiopia, India and Nigeria: the impacts of campaign design on vaccine hesitancy and health worker motivation JF - BMJ Global Health JO - BMJ Global Health DO - 10.1136/bmjgh-2021-006002 VL - 6 IS - 8 SP - e006002 AU - Abigail H Neel AU - Svea Closser AU - Catherine Villanueva AU - Piyusha Majumdar AU - S D Gupta AU - Daniel Krugman AU - Oluwaseun Oladapo Akinyemi AU - Wakgari Deressa AU - Anna Kalbarczyk AU - Olakunle Alonge Y1 - 2021/08/01 UR - http://gh.bmj.com/content/6/8/e006002.abstract N2 - Introduction The debate over the impact of vertical programmes, including mass vaccination, on health systems is long-standing and often polarised. Studies have assessed the effects of a given vertical health programme on a health system separately from the goals of the vertical programme itself. Further, these health system effects are often categorised as either positive or negative. Yet health systems are in fact complex, dynamic and tightly linked. Relationships between elements of the system determine programme and system-level outcomes over time.Methods We constructed a causal loop diagram of the interactions between mass polio vaccination campaigns and government health systems in Ethiopia, India and Nigeria, working inductively from two qualitative datasets. The first dataset was 175 interviews conducted with policymakers, officials and frontline staff in these countries in 2011–2012. The second was 101 interviews conducted with similar groups in 2019, focusing on lessons learnt from polio eradication.Results Pursuing high coverage in polio campaigns, without considering the dynamic impacts of campaigns on health systems, cost campaign coverage gains over time in weaker health systems with many campaigns. Over time, the systems effects of frequent campaigns, delivered through parallel structures, led to a loss of frontline worker motivation, and an increase in vaccine hesitancy in recipient populations. Co-delivery of interventions helped to mitigate these negative effects. In stronger health systems with fewer campaigns, these issues did not arise.Conclusion It benefits vertical programmes to reduce the construction of parallel systems and pursue co-delivery of interventions where possible, and to consider the workflow of frontline staff. Ultimately, for health campaign designs to be effective, they must make sense for those delivering and receiving campaign interventions, and must take into account the complex, adaptive nature of the health systems in which they operate. Data are available upon request. Data will be made available upon reasonable request where sufficient deidentification of interview data transcripts can be made. Data requests can be directed to Svea Closser at sclosser@jhu.edu and Olakunle Alonge at oalonge1@jhu.edu. ER -