Road traffic crashes are a leading cause of death and disability globally and are a significant contributor to emergency department (ED) referrals, which places a considerable burden on ED resources and staff. As a major tourist destination and the largest city in northern Thailand, Chiang Mai experiences a surge in visitors, particularly during the New Year and Songkran festivals, which are referred to as the “Seven Dangerous Days” periods. The objective of the present study was to predict the number of crash-related ED referrals at Maharaj Nakorn Chiang Mai Hospital using time-series analysis. Data from 19,164 patients referred to the ED in Maharaj Nakorn Chiang Mai Hospital between May 2007 and December 2022 were used to develop ARIMA and SARIMA models. The predicted number of referrals from these models was then compared with the observed number of referrals collected from 2023 to 2024. Furthermore, this study also examined patient referrals during the “Seven Dangerous Days” periods. The sensitivity analyses for investigation of effect of structural changes were performed by excluding the data in trained dataset before fully Trauma Referral Audit (TRA) implementation and/or after COVID-19 pandemic. Motorcycle crashes were the most common cause of injury (49.8%). The SARIMA(0,1,2)(1,0,1) 12 model provided the best fit for motorcycle and car crashes-related referrals, while the SARIMA(1,1,1)(1,0,1) 12 model was selected for combined referrals. The results of sensitivity analysis model excluded train dataset before fully TRA complementation show the higher prediction performance than baseline model. For the “Seven Dangerous Days” periods, the ARIMA(0,1,0) and ARIMA(2,1,1) models were the most suitable for the New Year and Songkran periods, respectively. The time-series based patient referral forecasting model that considered for the structural changes during study period could provide useful information for several simulations of preparation for proactive hospital management. Utilizing these models after adjusting for these major events could assist hospitals and relevant agencies in high-risk areas or during festival periods by providing predictions of traffic accident-related referrals that would enable the proactive allocation of necessary staff and resources.
Srikummoon et al. (Sat,) studied this question.