Forecasting Bronchopneumonia Disease Using the SARIMA Method: A Case Study at Hospital X

Abstract

Pneumonia remains the leading cause of death among children globally. In Indonesia, the prevalence of pneumonia in 2023 was particularly concerning, ranking first among diseases commonly affecting children under five, with a staggering 31.4% of cases. East Java Province reported the highest incidence, accounting for 32.03% or 45,041 affected children. To tackle this pressing issue, advancements in statistical methods, specifically the time series approach, offer valuable tools for prediction. This study employs the Seasonal Auto-Regressive Integrated Moving Average (SARIMA) forecasting method. Conducted as a non-reactive study utilizing secondary data, this research follows a retrospective cohort design, collecting data over a specific timeframe. The SARIMA (0,1,0)(1,1,0, 12) model is suitable for predicting bronchopneumonia cases, as it contains significant parameters and satisfies all necessary assumptions. The predictions generated from this model will extend over the following 12 periods.

Keywords: pneumonia, children, ARIMA, SARIMA