Emergency Department Visits for Heat Stroke in the United States, 2009 and 2010

Abstract

Background: The effect of extreme heat on health has become a growing public health concern due to climate change. We aimed to examine the epidemiological patterns of hospital-based emergency department (ED) visits for heat stroke in the United States.

Findings: We analyzed data from the 2009 and 2010 Nationwide Emergency Department Sample, the largest ED data system sponsored by the Agency for Healthcare Research and Quality. ED visits for heat stroke were identified by screening the recorded diagnoses using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code 992.0. Annual incidence rates of ED visits for heat stroke were computed according to demographic characteristics and geographic regions. In 2009 and 2010, there were an estimated 8,251 ED visits for heat stroke in the United States, yielding an annual incidence rate of 1.34 visits per 100,000 population (95% Confidence Interval [CI] = 1.23-1.45). Significantly higher incidence rates were found in males (1.99 per 100,000; 95% CI = 1.82-2.16), adults aged ≥ 80 years (4.45 per 100,000; 95% CI = 3.73-5.18), and residents living in the southern region (1.61 per 100,000; 95% CI = 1.43-1.79). The majority (63.1%) of ED visits for heat stroke occurred during the summer months of June, July and August. Over one-half (54.6%) of the ED visits for heat stroke required hospitalization and 3.5% of the patients died in the ED or hospital.

Conclusions: Heat stroke results in approximately 4,100 ED visits each year in the United States, with the majority occurring in the summer months and requiring admission to the hospital. Men, the elderly, and people living in the south region are at heightened risk.

Keywords: Emergency medical service; Epidemiology; Global warming; Heat stroke; Public health.

Figure 1

Estimated two-year frequency of emergency department visits for heat stroke by season, United States, 2009 and 2010. Note: Monthly and yearly data are not presented as the report of data for any subgroup with 10 or fewer subjects is prohibited by the HCUP NEDS data use agreement. There were 1408 missing values for admission month, 93.7% of which are from hospitals in the southern states.