Combining machine learning and mathematical models of disease dynamics to guide development of novel malaria interventions

Abstract

AbstractWhile progress to reduce malaria burden appears to be stalling, substantial research is underway to develop next-generation interventions addressing current malaria control challenges. However, due to limited testing early in development, it is currently difficult to predefine optimal intervention specifications to achieve target health goals or prioritize investment in novel candidate interventions. Mathematical models of malaria transmission can provide quantitative evidence to guide this process. Models can simulate deployment and predict the potential impact of a new intervention before testing by considering operational, health-system, population, and disease characteristics. Here, we present a quantitative approach to guide intervention development. Our method relies on consultation with product development stakeholders to define the putative space of novel intervention specifications. Coupling malaria transmission dynamics models with machine learning enables searching this multi-dimensional space and efficiently identifying optimal properties of candidate interventions to reach a specified target health goal. We demonstrate the power of our approach by application to five malaria interventions currently under development. Aiming for health goals of malaria prevalence reduction, we quantify and rank intervention characteristics which are key determinants of health impact. Furthermore, we identify minimal requirements and tradeoffs between operational factors, intervention efficacy and duration to achieve different levels of impact and show how these vary across disease transmission settings. When single interventions cannot achieve significant impact, our method allows finding optimal combinations of interventions fulfilling the desired health goals. Enabling efficient use of disease transmission dynamics models, our approach supports decision-making and resource investment to develop new malaria interventions.Significance StatementNovel interventions are needed to reduce malaria burden and tackle insecticide and drug resistance. Developing these interventions entails optimizing their specifications against target health goals. Because clinical evidence is limited early in development, these specifications are defined with little quantitative considerations. Mathematical models simulating disease dynamics can quantify intervention impact; however, their complexity hinders exploring the high-dimensional space of putative intervention characteristics. We established a quantitative approach to guide intervention development through dialogue with product development stakeholders and leveraging models of malaria dynamics with machine learning. Our analysis identifies requirements of efficacy, coverage and duration of effect for five novel malaria interventions to achieve targeted reductions in malaria prevalence, highlighting the role of mathematical models to support intervention development.One Sentence SummaryDefining quantitative profiles of novel malaria interventions by combining machine learning with mathematical models of disease transmission