Air pollution from bushfires and their association with hospital admissions in Sydney, Newcastle and Wollongong, Australia 1994-2007

Objective: We examined the association between validated bushfire smoke pollution events and hospital admissions in three eastern Australian cities from 1994 to 2007. Methods: Smoke events were defined as days on which bushfire smoke caused the 24-hour citywide average concentration of airborne particles to exceed the 99^th percentile of the daily distribution for the study period. We used a time-stratifed case-crossover design to assess the association between smoke events and hospital admissions. Odds ratios (OR) and 95% confidence intervals (CI) were estimated for cardiovascular and respiratory conditions on event days compared with non-event days. Models were adjusted for daily meteorology, influenza epidemics and holidays. Results: Smoke events occurred on 58 days in Sydney (population: 3,862,000), 33 days in Wollongong (population: 406,000) and 50 days in Newcastle (population: 278,000). In Sydney, events were associated with a 6% (OR=1.06, 95%CI=1.02-1.09) same day increase in respiratory hospital admissions. Same day chronic obstructive pulmonary disease admissions increased 13% (OR=1.13, 95%CI=1.05-1.22) and asthma admissions by 12% (OR=1.12, 95%CI=1.05-1.19). Events were also associated with increased admissions for respiratory conditions in Newcastle and Wollongong. Conclusions: Smoke events were associated with increased hospital admissions for respiratory but not cardiovascular conditions. Large populations are needed to assess the impacts of brief exposures. Implications: Public health impacts from bushfire pollution events are likely to increase in association with a warming climate and more frequent severe fire weather.