Background: Case reports have indicated that a tropical cyclone may increase Emergency Department (ED) visits significantly. Study Objectives: To examine emergency health care demands across a series of tropical cyclones, and to build a predictive model to analyze a cyclone's potential effect. Methods: This was an observational non-concurrent prospective study performed in Taiwan. Twenty hospitals were included. The number of daily ED visits in each hospital was our primary end point, and data were retrieved from the database provided by the National Health Insurance Research Database. Our study examined the period from 2000 to 2008. A total of 22 tropical cyclones (typhoons) that had passed over eastern Taiwan and covered the area under study were included. Multiple linear regression time-series models were employed to estimate the effects of "days since typhoon landfall" and various characteristics of the typhoons on the end point of daily ED visits to each hospital. Results: The final multiple linear regression time-series model showed that the number of daily ED visits increased in areas where a strong typhoon had landed directly, with the increase being evident during the first 2 days since landfall. Our model also indicated that the three most important variables to predict a change in the pattern of daily ED visits were intensity of typhoon, simultaneous heavy rain, and direct landfall. Conclusions: During tropical cyclones, emergency services were under increased demand in selected time periods and areas. Health care authorities should collect information to build local models to optimize their resources allocation in preparation.
ASJC Scopus subject areas