ARIMA-BASED TIME SERIES ANALYSIS AND FORECASTING OF WEEKLY LASSA FEVER CASES IN NIGERIA (JANUARY 2021–SEPTEMBER 2025)

Lassa fever (LF), also known as Lassa hemorrhagic fever, is an acute viral hemorrhagic infection that is a threat to public health. The virus that causes LF is a single-stranded RNA from the family Arenaviridae. This study used quantitative data from the Nigerian center for Decease Control (NCDC), and the data were analyzed from week 1, 2021, to week 36, 2025, which included confirmed Lassa fever cases and associated deaths. The Box–Jenkins ARIMA methodology model was used for the analysis and R was used for the computation. This ARIMA matched the fundamental Box–Jenkins model adequacy conditions, which are stationary and invertible, whose residuals do not have significant linear autocorrelation. It has the lowest AIC and BIC of all the tested models. Although one major setback is that a non-seasonal ARIMA (3, 0, 1) model does not incorporate the strong periodicity week, it was the best-fitting model. It yielded an in-sample cases per week, and a forecast for the twelve-week-ahead period showed an increase from about 15 to 29 cases per week by the end of November 2025, which is in line with the historical onset of the seasonal epidemic period. The report discusses model adequacy and implications for surveillance of Lassa fever and early-warning planning in Nigeria