Evaluation of the Relationship between Ambient Air Pollution and Hospitalization for Acute Exacerbation of Chronic Obstructive Pulmonary Disease at Temple University Hospital

Abstract

Background: Air pollution has been associated with adverse health effects for all-cause and specific respiratory morbidity and mortality outcomes. Acute exacerbations of COPD (AE-COPD) accelerate the decline in pulmonary function and are associated with greater mortality, morbidity, health care utilization, and reduced quality of life. Since the 1970 Clean Air Act was implemented, important reductions in air pollution have been achieved, but no safe threshold has been identified. Objectives: The study was planned to evaluate associations between exposure to ambient concentrations of five criteria air pollutants (CO, SO2/, NO2/, ozone, PM2.5/) in Philadelphia, Pennsylvania, and visits to Temple University Hospital for AE-COPD, from January 1, 2005 through March 31, 2007. To identify subgroups with greater susceptibility to air pollution, associations were examined according to age, gender, race, residence, and antibiotic prescription. Methods: Average daily air pollutant concentrations were obtained from the EPA's Air Quality Services Data Mart. Air pollutant exposures were evaluated for the day of the visit (lag0), one and two days preceding the visit (lag1 and lag2), and the average concentration over three days (lag012). Poisson regression provided rate ratios (RRs) to estimate associations between air pollution exposures and AE-COPD hospital visits. Results: Of 1546 hospital visits for AE-COPD, 43% were from persons 65 years or older, 50% of each gender, and 90% from Philadelphia. In single pollutant models, increased RRs were present at all lags for NO2/ (e.g., RR = 2.27 [95%CI: 1.52, 3.38] at lag012) and SO2/ (e.g., RR = 1.70 [95%CI: 1.38, 2.08] at lag012). For PM2.5/, the direct effect was present only during the winter at lag1, lag2, and lag012 (RR = 1.79 [95%CI: 1.08, 2.96]). Inverse associations were present for ozone at all lags (e.g., RR = 0.64 [95%CI: 0.53, 0.76] at lag012). Compared to the cohort as a whole, those ≥ 65 years of age were at greater risk of an AE-COPD hospital visit associated with PM2.5/ and CO at lag012, with NO2/ and SO2/ at lag0 and lag012, but there was no difference in ozone effect. Conclusions: Primary gaseous air pollution exposures (SO2/, CO, NO2/) were associated with increased AE-COPD hospital visits among COPD patients at Temple University Hospital. The effects of SO2/, CO, NO2/, and PM2.5 were greater for the subgroup ≥ 65 years of age compared to the cohort as a whole. Inverse associations with ozone were consistent across subgroups. These results suggest that air quality during the study period was insufficient to protect the health of COPD patients, especially those ≥ 65 years old. Further study is needed to understand generalizability to other populations and to evaluate lower ranges of exposure from current levels of air pollution.