The accuracy of forecasting results of the Box-Jenkins method for time series analysis on the number of pneumonia patients

Abstract

Accurate forecasting in pneumonia cases is crucial for healthcare providers to effectively allocate resources and plan for patient care. This quantitative study employs a non-reactive approach, employing time series analysis through the autoregressive integrated moving average (ARIMA) Box-Jenkins method on secondary data. Monthly data simulations spanning 1, 3, and 6 months were conducted, divided into two groups: initialization data from 2016-2019 and actual data from 2020. The goal was to assess the forecasting accuracy using the Box-Jenkins method for pneumonia cases at Kamonji Public Health Center, Palu. The 1-monthly data simulation, with 48-time series as initialization data, yielded an appropriate forecasting model, ARIMA (1,1,1), predicting 289,166 pneumonia patients. However, the 3-monthly and 6-monthly simulations did not yield suitable forecasting models. In conclusion, none of the data simulations accurately predicted the number of pneumonia cases, often overestimating compared to actual data. Recommendations include using data spanning more than five years to enhance accuracy. For stakeholders, forecasting analysis in pneumonia cases aids in predicting cases, identifying risk factors, understanding prognosis, and tailoring treatment plans. Data-driven models empower healthcare providers to make informed decisions and optimize patient care. Accurate pneumonia forecasting remains essential in efficient healthcare resource allocation and planning.