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

Shi Zhao; Salihu S. Musa; Hao Fu; Daihai He; Jing Qin

We fit four different growth models to the LF confirmed cases and estimate the modelaverage reproduction number (R) after model selection. We show the growth model fitting results in Figure 3 and the model estimation and selection results in Table 1 . The models fit the epidemic data well, and most of them have AICs lower than the baseline AR(2) model. Most of the regions exhibit an epidemic turning point (Ï„) ranging from epidemiological week (EW) 4 to 10 in each year. Out of four outbreaks in the states of Bauchi and Plateau, there are three estimated Ï„s after EW 10 ( Table 1) , and this state also has the largest number of LF cases in the outbreaks of both 2017-18 (41.9% of all cases) and 2018-19 (36.0% of all cases). Hypothesized spatial heterogeneity in the R is tested by Cochran's Q test. The testing results for the Rs in the five states are significant (i.e., p-value < 0.05) for both the 2017-18 and 2018-19 LF epidemics. Thus, we report the existence of spatial heterogeneity in LF epidemics in Nigeria.