A time series analysis of trending dengue cases in Sri Lanka.

Loading...
Thumbnail Image
Date
2025
Authors
Kurukulasuriya Perera, Ruvani
Journal Title
Journal ISSN
Volume Title
Publisher
Lethbridge, Alta. : University of Lethbridge, Faculty of Health Sciences
Abstract
The study aimed to predict dengue case numbers in Sri Lanka from January 2024 to December 2025. The prediction will assist the National Dengue Control Unit of Sri Lanka in assessing the potential dengue case numbers before a seasonal dengue crisis. This allows the Ministry of Health of Sri Lanka to plan effective healthcare mobilization and manage its resources during dengue seasons. Secondary data on all island dengue cases was obtained from the National Dengue Control Unit's national surveillance system from 2015 to 2023. A seasonal ARIMA(0,1,1)(0,0,2)[12] model was generated in R software by the forecast package’s time series function based on the Box-Jenkins method. The ARIMA model was validated as a good fit for prediction with the Ljung-Box (p-value >0.05), Shapiro-Wilk (p-value >0.05), and ADF (p-value <0.05) tests. The prediction’s MAPE was estimated as accurate for forecasting (4.46). The seasonal ARIMA model demonstrated the ability to make a short-term prediction in univariate analyses.
Description
Keywords
Dengue , Seasonal crisis , Healthcare mobilization , Short-term prediction
Citation