Factors predictive of epidemic waves of COVID-19 in Africa during the first 2 years of the pandemic

Abstract

Objectives

The objective was to study the epidemic wave curves, according to the characteristics of the countries, to identify the differences and the predictive factors of evolution.

Methods

We have carried out modeling of the COVID-19 epidemic data from validated databases for 53 African countries.

Results

All countries recorded at least four waves. The duration of the waves had decreased over time (P <0.001) and extended with the rainy season (P = 0.03). The incidence rates were higher for countries with the best development indicators (P <0.001). Positive spatial autocorrelation was significant for all wave characteristics, except for relative amplitude at the end of the wave. The time-adjusted multivariate analysis identified seasons for duration (P = 0.017) and human development index for peak incidence rate (P <0.001) and relative amplitude at the end of the wave (P = 0.041) as predictors of wave characteristics.

Conclusions

The duration of the waves was influenced by the seasons and the study periods, the incidences by the economic development, and health indicators. The appearance of new variants seemed associated with the start of the waves. None of the factors studied is associated with an inflection and a decrease in the curve.