The Dynamic Relationship between Temperature and Morbidity
利用加州2005-2014年医院和急诊就诊高频数据，发现极端温度（低于40°F或高于80°F）对当日及后续数周的急诊就诊量有显著影响，并区分了生物和行为机制。
This paper examines the relationship between temperature and hospital usage with a focus on the role of behavioral responses to temperature. I use high-frequency data on the near universe of hospital and emergency department (ED) visits in California between 2005 and 2014 to estimate the effects of temperature on hospital usage patterns. I find that a day with mean temperature under 40°F leads to a 6.1% decrease in ED visits on the day of the event but that total net visits increase by approximately 11.0% above the daily mean after accounting for visits in the weeks that follow. Additionally, I find that a day over 80°F is associated with a same-day increase in ED visits of 3.5% and a total net increase of 5.1%. For both cold and hot temperatures, I provide evidence of the mechanisms—whether biological or behavioral—that explain these patterns.