Model-informed optimal allocation of limited resources to mitigate infectious disease outbreaks in societies at war

Abstract

Infectious diseases thrive in war-torn societies. The recent sharp increase in human conflict and war thus requires the development of disease mitigation tools that account for the specifics of war, such as scarcity of important public health resources. Differential equation-based compartmental models constitute the standard tool for forecasting disease dynamics and evaluating intervention strategies. We developed a compartmental disease model that considers key social, war, and disease mechanisms, such as gender homophily and the replacement of soldiers. This model enables the identification of optimal allocation strategies that, given limited resources required for treating infected individuals, minimize disease burden, assessed by total mortality and final epidemic size. A comprehensive model analysis reveals that the level of resource scarcity fundamentally affects the optimal allocation. Desynchronization of the epidemic peaks among several population subgroups emerges as a desirable principle since it reduces disease spread between different sub-groups. Further, the level of preferential mixing among people of the same gender, gender homophily, proves to strongly affect disease dynamics and optimal treatment allocation strategies, highlighting the importance of accurately accounting for heterogeneous mixing patterns. Altogether, the findings help answer a timely question: how can infectious diseases be best controlled in societies at war? The developed model can be easily extended to specific diseases, countries, and interventions.Significance statementSocieties at war are particularly affected by infectious disease outbreaks, necessitating the development of mathematical models tailored to the intricacies of war and disease dynamics as valuable tools for policy-makers. The frequently limited availability of public health resources, such as drugs or medical personnel, yields a fundamental optimal allocation problem. This study frames this problem in a generic, modifiable context and proposes model-informed solutions by identifying allocation strategies that minimize disease burden, measured by total deaths or infections. The desynchronization of epidemic peaks among a heterogeneous population emerges as a general disease mitigation strategy. Moreover, the level of contact heterogeneity proves to substantially affect disease spread and optimal control.