Associations Of Temperature With Daily Ambulance Dispatches And Emergency Room Visits In Kunshan, China

Background:The World Health Organization (WHO) has stated that global climate warming is the biggest public health challenge in the21st century. As temperature and frequency of extreme weather events increase, the health effects of temperature and extreme weather conditions arouse growing concerns. Numerous epidemiological studies investigated the associations between daily temperature variations and health events, most of which were based on mortality outcomes. These studies had been blamed for only explaining for the severest health effects. Emergency room visits were believed to be a good indicator of health effects. However, Emergency room visits might be influenced by hospitalization bias and selection bias. Recently, some studies have attempted to use emergency ambulance dispatches as indicators of temperature-related health effects.Objective:Time-series approach was used to investigate the exposure-response relationship between daily temperature and ambulance dispatches and emergency room visits, and then evaluate the application of ambulance dispatches as a health indicator in environmental epidemiology.Methods:We collected daily counts of ambulance dispatches in the period of Jan1,2008to Dec31st,2009from the Ambulance Center of Kunshan City, Jiangsu Province, daily visits to emergency room of Kunshan Chinese Traditional Medicine Hospital, and weather data from Kunshan Meteorological Bureau. For the time series regression, we fitted the Poisson generalized additive model, after adjustment for long-term and seasonal trends using natural cubic smooth splines, day of week, and relative humidity. For the smooth splines, we applied8degrees of freedom (df) per year to seasonal control and3df for adjusting for relative humidity. We introduced linear or nonlinear terms of temperature in regression models after plotting the exposure response relationship curves. We selected the average of current day and previous day’s temperature (lag01days) as the main lag in the analysis. In sensitivity analysis, we also estimated single-day lag’s effects of0to7days, in addition to varying df of smoothness of time from4to8df per year. The time-series models were fitted using mgcv package in R2.14.1.Results:Daily mean temperature has similar exposure-response relationship with ambulance dispatches and emergency room visits, characterized by a significant break point at20℃. When the daily mean temperature was higher than20℃, a1℃increase in the average of current-day and previous day’s temperature was associated an increase of3.01%(95%CI:1.49%,4.53%) in ambulance service and2.82%(95%CI:1.11%,4.52%) in emergency room visits, respectively. When the daily mean temperature was lower than20℃, lagged exposure of0to3days to temperature was not significantly with ambulance dispatches and emergency room visits;4or more lag days’temperature could have significant impact on daily emergency room visits. Sensitivity analyses indicated that our time-series models were generally robust to model specifications.Significance:The exposure-response relationship between temperature and ambulance dispatches was similar with that between temperature and emergency room visits, suggesting ambulance service might be as a good health indicator as emergency room visit. Our findings might have important significance for local health authorities to regulate and assign medical resources of emergency treatment. Considering more potential high temperature events in the context of global climate warming, in combinations with aging Chinese populations, we should have every reason to enforce investigations and improve population’s capability of adaption to climate change by monitoring, precaution and prevention.Conclusions:This time series analysis in Kunshan, Jiangsu Province found high temperature (greater than20℃) could increase the ambulance dispatches and emergency room visits, and low temperature had significant but lagged impact on daily emergency room visits. This is the first study to report a sensitive health indicator of emergency ambulance in environmental epidemiology. Nevertheless, our findings need to be replicated in more locations. Further studies, such as spatial epidemiology and modification factor investigations might be important.