Predictability of infectious disease outbreak severity: Chikungunya as a case study

Abstract

A single pathogen can cause outbreaks of varying size and duration in different populations. Anticipating severe outbreaks would facilitate public health preparedness, but the extent to which this is possible is unclear. We conducted a data-driven investigation into the predictability of outbreak severity, using chikungunya virus (CHIKV) as a case study. For mosquito-transmitted viruses like CHIKV, the potential for severe outbreaks is often assessed using climate-based estimates of the basic reproduction number, $R_0$ . We derived a large set of $R_0$ estimates for CHIKV by fitting a mechanistic model to data from 86 chikungunya outbreaks. These $R_0$ estimates were weakly predicted by climatic and other factors. Among deterministic drivers of outbreak severity, the contribution of $R_0$ was comparable to that of generation interval length, transmission distance, and population network structure. While aspects of chikungunya outbreak severity are predictable, innovative approaches are needed that look beyond the impacts of climate on $R_0$.