Effect of evidence updates on key determinants of measles vaccination impact: a DynaMICE modelling study in ten high-burden countries

Abstract Background Model-based estimates of measles burden and the impact of measles-containing vaccine (MCV) are crucial for global health priority setting. Recently, evidence from systematic reviews and database analyses have improved our understanding of key determinants of measles vaccine impact. We explore how updated representations of these determinants affect model-based estimation of MCV impact in ten countries with highest measles burden. Methods Using Dynamic Measles Immunisation Calculation Engine (DynaMICE), an age-structured compartmental model of measles transmission and vaccination, we evaluated the effect of evidence updates for five determinants of MCV impact: case fatality risk, contact patterns, age-dependent vaccine efficacy, the potential of supplementary immunisation activities (SIAs) to reach zero-dose children, and the basic reproduction number. We also evaluated the incremental impact of the first dose (MCV1), second dose (MCV2), and SIA dose of measles vaccines, based on country-specific coverage estimates from the World Health Organization. The MCV impact was assessed by cumulative vaccine-averted cases, deaths, and disability-adjusted life years over 2000–2050. Results Incorporated with the updated data sources, DynaMICE projected 252 million measles cases, 3.7 million deaths and 230 million disability-adjusted life years incurred over 2000–2050 in the ten high-burden countries when MCV1, MCV2, and SIA doses were implemented. Compared to no vaccination, the administration of MCV1 contributed to 66% reduction in cumulative measles burden, while MCV2 and SIAs reduced this further to 89%. With routine and supplementary vaccination, India and countries with high vaccination coverage could maintain measles incidence below 1 per million. Among the updated determinants, shifting from fixed to linearly-varying vaccine efficacy by age and from static to time-varying case fatality risks had the biggest effect on the model projections of MCV impact. While varying the basic reproduction number showed a limited effect on vaccine-averted burden, updates on the other four determinants together led to an overall reduction of MCV impact by 0.87–26.7%. Conclusions High coverage of measles vaccine through both routine and SIA delivery platforms are essential for achieving and maintaining low incidence in high-measles burden settings. Incorporating updated evidence particularly on vaccine efficacy and case fatality risk reduces estimates of the impact of vaccination slightly, but its overall impact remains considerable.