Assessment of the Heat-Health Relationship in Florida with Threshold Recommendations for Heat Warning/Advisories

BACKGROUND: As global temperatures increase, heat related morbidity/mortality (HRI) will be one of the challenges faced. The HRI burden will vary by population based on health status (e.g. acclimatization), resources available, and population/individual behavioral response to heat. Very little research on HRI has been conducted in the southeastern US where individuals may be accustomed to high heat and humidity. This study was conducted in Florida to assess the burden of occupational and non-occupational HRI.  To enhance prevention efforts, threshold criteria for heat warnings/advisories were also assessed.  

METHODS: A time series analysis was conducted in each of Florida’s seven National Weather Forecasting (NWF) areas during May-October, 2005-2009. Maximum daily temperature and heat-index were used. The Agency for Health Care Administration provided daily HRI (ICD-9-CM 992.0-992.9) hospitalizations and emergency visits (ED). Occupational cases were defined as individuals age ≥16 years where the expected payer was worker’s compensation or the record contained a work-related Ecode. Population data were obtained from the American Community Survey.  Analysis was restricted to Florida residents.  

RESULTS: Between 2005 and 2009, 17,843 Floridians were treated for non-occupational HRI (age-adjusted rate=19.7/100,000 Floridians; 95% CI=16.7,22.6) and 2168 Floridians were treated for occupational HRI (age-adjusted rate=3.8/100,000 workers; 95% CI=1.9,5.6). Eighty-three percent of all cases occurred in May-September; however, cases occurred throughout the year with 6% of HRI cases occurring in October and 4% in April. As temperatures increased, the relative rate among occupational cases increased faster than among non-occupational cases when compared to their respective baseline rates. In the Melbourne NWF area, for every 5°F increase in the heat index there was a 55% (95% CI=1.48,1.63) increase in the rate of non-occupational HRI and 89% (95% CI=1.59,2.24) increase in the rate of occupational HRI. The temperature/heat index at which the highest rate of non-occupational HRI occurred varied by NWF area (e.g. Melbourne: heat index=108°F–<110°F, incidence rate=7.6/1,000,000 person-days; 95% CI=2.4,24.5). 

CONCLUSIONS: This study provides one of the first examinations of the heat-health relationship in a southeastern state. Even in hot and humid climates, as temperature increases the rate of HRI increases. Non-urban populations were found to have a higher proportional burden of disease than urban populations. Additionally, HRI distribution differed between occupational and non-occupational populations.  When establishing heat advisory triggers in Florida, the combination of heat and humidity may be more important than temperature alone.  HRI is preventable and an understanding of HRI in all climates is essential in reducing the burden of disease.