Risk of respiratory and cardiovascular hospitalisation with exposure to bushfire particulates: new evidence from Darwin, Australia.

The risk of hospitalisation from bushfire exposure events in Darwin, Australia, is examined. Several local studies have found evidence for the effects of exposure to bushfire particulates on respiratory and cardiovascular hospital admissions. They have characterised the risk of admission from seasonal exposures to biomass air pollution. A new, unanalysed data set presented an additional chance to examine unique exposure effects, as there are no anthropogenic sources of particulates in the vicinity of the exposure monitor. The incidence of daily counts of hospital admissions for respiratory and cardiovascular diagnoses was calculated with respect to exposures of particulate matter (PM(10)), course particulate matter, fine particulate matter (FPM) and black carbon composition. A Poisson model was used to calculate unadjusted (crude) measures of effect and then adjusted for known risk factors and confounders. The final model adjusted for the effects of minimum temperature, relative humidity, a smoothed spline for seasonal effects, 'date' for a linear effect over time, day of the week and public and school holidays. A subset analysis adjusted for an influenza epidemic in a particular year. The main findings suggest that respiratory admissions were associated with exposure to PM(10) with a lag of 1 day when adjusted for flu and other confounders (RR = 1.025, 95 % CI 1.000-1.051, p < 0.05). This effect is strongest for exposure to FPM concentrations (RR = 1.091, 95 % CI 1.023-1.163, p < 0.01) when adjusted for flu. Respiratory admissions were also associated with black carbon concentrations recorded the previous day (RR = 1.0004, 95 % CI 1.000-1.0008, p < 0.05), which did not change strength when adjusted for flu. Cardiovascular admissions had the strongest association with exposure to same-day PM and highest RR for exposure to FPM when adjusted for confounders (RR = 1.044, 95 % CI 0.989-1.102). Consistent risks were also found with exposure to black carbon with lags of 0-3 days.