A Meta-analysis of Association Between Short-term Ambient Ozone Exposure and Respiratory Hospital Admissions

Abstract

PP-29-054 Background/Aims: Tropospheric ozone is an air pollutant common in urban areas, associated with many adverse human health impacts, including increased risk of respiratory symptoms. Many studies have found significant positive associations between ozone and risk of respiratory hospitalizations; however, other studies found contradictory results. We conducted a meta-analysis of ozone and respiratory hospital admissions, which combines results from multiple studies accounting for their statistical uncertainty. Methods: A systematic review of the literature based on our protocol identified 136 estimates from 96 time-series and case-crossover studies. We stratified main estimates extracted from previous literature according disease categories (eg, total respiratory disease, chronic obstructive pulmonary disease), hospitalization types (eg, emergency hospital admission, emergency room visit), and age groups (eg, all-ages, children). We investigated how the concentration-response relationship was affected the ozone metric used (eg, 1-hour maximum), age of study population, season, and region of study. Results were compared with those from multicity studies. In addition, we examined publication bias and calculated a summary effect adjusted for publication bias. Results: Results indicate significant associations between ozone exposure and hospital admissions for various disease categories. The estimates were elevated by publication bias, but some associations remain statistically significant after adjustment of this bias. For instance, results indicate that risk of emergency hospital admissions for total respiratory disease increased 4.47% (95% interval 2.48%, 6.50%) per 10 ppb increment of 24 hours average ozone among the elderly. After adjusting for publication bias, the estimate becomes 2.97% (1.05%, 4.94%). Results reveal that different ratios to convert among ozone metrics did not greatly change results. The relationship between ozone and hospital admissions was affected by different study characteristics (eg, lag selection, age group). Conclusion: This meta-analysis can provide a better understanding of ozone exposure's influence on human health and guide future studies. Findings can also provide quantitative evidence for environmental policy decision making.