Temporal association between hospital admissions for asthma in Birmingham and ambient levels of sulphur dioxide and smoke.

BACKGROUND A study was performed to determine whether daily and weekly variations in the levels of smoke and sulphur dioxide (SO2) in Birmingham are related to hospital admissions for asthma and acute respiratory diseases. METHODS Daily numbers of hospital admissions for asthma (ICD code 493) and acute respiratory conditions (ICD 466, 480-486, 490-496) for residents of Birmingham between 1988 and 1990 were obtained from West Midlands RHA Körner inpatient data. Average daily levels of sulphur dioxide and smoke were obtained from Birmingham City Council for the same period, together with daily meteorological summaries from the Department of Geography, University of Birmingham. With the exception of one day, all air pollution measurements remained within current EC guide levels. Data were divided into seasons and the relation between hospital admissions and pollutant levels were explored by stepwise least squares regression models. Meteorological variables (temperature, pressure, humidity) were entered into the model if they showed significant association with hospital admissions during the season in question. Analysis was undertaken for daily (same day and lagged by two days) and weekly pollutant levels. Admissions were lagged behind pollution levels to allow for delayed effects of pollutants. RESULTS The mean daily level of smoke was 12.7 micrograms/m3 and of SO2 was 39.1 micrograms/m3, with maxima of 188.3 micrograms/m3 and 126.3 micrograms/m3, respectively. Significant associations were found between hospital admissions for respiratory disease lagged by two days, and smoke and SO2 levels during winter. Associations between admissions for asthma and smoke and SO2 levels were significant at the 5% level. These were independent of temperature, pressure, and humidity. Stepwise regression including both pollutants showed that smoke, but not SO2, was a significant independent predictor of hospital admissions for both asthma and all respiratory conditions. During winter a rise of 100 micrograms/m3 smoke might result in five (95% CI 0.6 to 9) more asthma admissions and 21.5 (95% CI 10 to 33) more acute respiratory admissions each day in Birmingham. A 100 micrograms/m3 rise in SO2 might result in four (0 to 7) more asthma admissions and 15.5 (6 to 25) more respiratory admissions each day. Independent associations were also found between weekly mean smoke and SO2 levels and all respiratory admissions during autumn and winter. During summer, daily mean smoke and SO2 levels were significantly associated with non-lagged daily admissions for all respiratory diseases (p < 0.02). There was no association between air pollution and hospital admissions during spring. CONCLUSIONS Daily variations in smoke and SO2 levels are significantly associated with hospital admissions for asthma and respiratory disease during winter in Birmingham at levels of air pollutants within the EC guide levels. This association was independent of potential confounding effects of weather (temperature, pressure, humidity) and suggests that current levels of air pollution can still produce significant health effects.