TEMPORAL MODELLING OF DENGUE CASES IN CÔTE D’IVOIRE FROM 2017 TO 2023: APPLICATION OF THE ARMA MODELS

Background and objectives. Dengue fever is a persistent public health problem in Côte d’Ivoire, with a recent upsurge in cases leading to instability in the health system. The aim of this study was to train the ARMA and SARMA models to analyze time series on the weekly prevalence of confirmed and suspected dengue cases in Côte d’Ivoire and to use these models to forecast dengue prevalence for the next 52 weeks. Methods. This secondary study, based on weekly data on confirmed and suspected cases from 2017 to 2023, uses the ARMA (autoregressive moving average) and SARMA (seasonal autoregressive moving average) models to forecast dengue prevalence in Côte d’Ivoire. The data comes from the Institut National de l’Hygiène Publique. Results. The ARMA and SARMA models were identified as the best performing models, with respective MAE and RMSE of (MAE): 4.41 and 8.68 for the ARMA model and (MAE): 4, 3 and RMSE: 8.66 for the SARMA model. Conclusion. These models provide valuable information for healthcare planning and resource allocation, even though external factors and complex interactions need to be taken into account.