Heat-related emergency hospitalizations for respiratory diseases in the Medicare population

Abstract

Rationale: The heat-related risk of hospitalization for respiratory diseases among the elderly has not been quantified in the United States on a national scale. With climate change predictions of more frequent and more intense heat waves, it is of paramount importance to quantify the health risks related to heat, especially for the most vulnerable.

Objectives: To estimate the risk of hospitalization for respiratory diseases associated with outdoor heat in the U.S. elderly.

Methods: An observational study of approximately 12.5 million Medicare beneficiaries in 213 United States counties, January 1, 1999 to December 31, 2008. We estimate a national average relative risk of hospitalization for each 10°F (5.6°C) increase in daily outdoor temperature using Bayesian hierarchical models.

Measurements and main results: We obtained daily county-level rates of Medicare emergency respiratory hospitalizations (International Classification of Diseases, Ninth Revision, 464-466, 480-487, 490-492) in 213 U.S. counties from 1999 through 2008. Overall, each 10°F increase in daily temperature was associated with a 4.3% increase in same-day emergency hospitalizations for respiratory diseases (95% posterior interval, 3.8, 4.8%). Counties' relative risks were significantly higher in counties with cooler average summer temperatures.

Conclusions: We found strong evidence of an association between outdoor heat and respiratory hospitalizations in the largest population of elderly studied to date. Given projections of increasing temperatures from climate change and the increasing global prevalence of chronic pulmonary disease, the relationship between heat and respiratory morbidity is a growing concern.

Figure 1.

Two hundred thirteen U.S. study counties and their average summer temperatures. Colors correspond to the county’s daily mean temperatures, averaged from May to September (1999–2008).

Figure 2.

Percent increase in respiratory hospitalizations for each 10°F daily outdoor heat increase, 1999 to 2008 (lag 0). Estimates are pooled across all 213 study counties; outdoor heat is measured as daily mean temperature, May to September. Horizontal lines show 95% posterior intervals. COPD = chronic obstructive pulmonary disease; RTI = respiratory tract infections.

Figure 3.

Relative risk of heat-related respiratory hospitalization with and without adjustment for air pollution, 1999 to 2008 (lag 0). Analysis for ozone (O₃) (n = 200), particulate matter with aerodynamic diameter less than or equal to 10 μm (PM₁₀) (n = 165), and particulate matter with aerodynamic diameter less than or equal to 2.5 μm (PM_2.5) (n = 196). Circles show community-level estimates. Larger circles indicate greater certainty. Black diamond shows estimate across all communities.

Figure 4.

Modification of county-level relative risk of heat-related respiratory hospitalization by a county’s average summer temperature (lag 0). Average county summer temperature for May to September, 1999 to 2008. Circles show county-level estimates. Larger circles indicate greater certainty. Solid line shows the predicted pooled estimate by average summer temperature (shaded region shows 95% posterior interval). T_mean = mean daily temperature.