Large-scale Lassa fever outbreaks in Nigeria: quantifying the association between disease reproduction number and local rainfall

Background Lassa fever (LF) is increasingly recognized as an important rodent-borne viral hemorrhagic fever presenting a severe public health threat to sub-Saharan West Africa. In 2018, LF caused an unprecedented outbreak in Nigeria, and the situation was worse in 2019. This work aims to study the epidemiological features of outbreaks in different Nigerian regions and quantify the association between reproduction number ( R ) and local rainfall by using modeling analysis. Methods We quantify the infectivity of LF by the reproduction numbers estimated from four different growth models: the Richards, three-parameter logistic, Gompertz, and Weibull growth models. LF surveillance data are used to fit the growth models and estimate the R s and epidemic turning points ( τ ) in different regions at different time periods. Cochran’s Q test is further applied to test the spatial heterogeneity of the LF epidemics. A linear random-effect regression model is adopted to quantify the association between R and local rainfall with various lag terms. Findings Our estimated R s for 2017-18 (1.33 with 95% CI: [1.29, 1.37]) and 2018-19 (1.29 with 95% CI: [1.27, 1.32]) are significantly higher than those for 2016-17 (1.23 with 95% CI: [1.22, 1.24]). We report spatial heterogeneity in the R s for outbreaks in different Nigerian regions. For the association between rainfall and R , we find that a one unit (mm) increase in average rainfall over the past 7 months could cause a 0.62% (95% CI: [0.20%, 1.05%]) rise in R . Conclusion There is significant spatial heterogeneity in the LF epidemics in different Nigerian regions. We report clear evidence of rainfall impacts on LF outbreaks in Nigeria and quantify the impact.