Datasets for "Risk Factors Associated with Successive Waves of COVID-19 in North Kivu Province, Democratic Republic of the Congo, March 2020–September 2023"

We used the independent variables related to the epidemiological, socio-political, socio-economic, demographic, and health-related characteristics to cluster the 34 health zones of North Kivu province. We did this through a hierarchical classification of principal components (HCPC) based on a multiple correspondence analysis (MCA). We performed the MCA with the "FactoMineR" and "factoextra" packages in R software version 4.2.0. The differences in the number of recorded cases of COVD-19 between the resulting clusters were tested using the ANOVA test. The association between COVID-19 cases and the independent variables was assessed using multivariable negative binomial regression analyses, with the 2020-2023 average population of each health zone included as an offset variable. All the independent variables were included in the full model. To obtain a reduced model through iterative multivariable fitting, we performed model stepwise selection using the “stepAIC” function from the MASS package. This function selects the best model based on the Akaike Information Criterion (AIC), which assesses the quality of a model by considering the number of parameters and the quality of fit. A lower AIC is indicative of a better model. The reduced model was then chosen for the principle of parsimony. A statistically significant association was determined using the model’s risk ratio (RR) value reported with its 95% confidence interval and P-value less than 0.05. Additionally, multi-collinearity was assessed between the independent variables using the variance inflation factor (VIF), and this was less than 5 for all the predictors of the reduced multivariable model. The negative binomial regression statistical analyses were conducted using R software version 4.2.0.