Forecasting asthmatic wheezing using temperature velocity

A model was constructed to predict pediatric asthmatic wheezing visits to the emergency department. All wheezing visits to the Children's Hospital of Philadelphia Emergency Department were analyzed for 1982 and 1983, for ages two to 18. Nine thousand four hundred twenty-five visits fit the study requirements, 27% of the total number of emergency department visits for all causes. The a priori hypothesis used to construct the model was that temperature change, not absolute temperature, would be a good predictor of emergency department visits. Although not attempting to prove a cause-and-effect relationship, by studying temperature velocity (the rate of change of temperature) and the direction of change of temperature (a seasonal variable reflecting rising or falling temperature), a model was created that was highly significant (P less than 0.0001) and could explain 35% of the daily variation in asthmatic emergency department visits (R2 = 35%, r = 0.59). When weekly emergency department visits were analyzed, the model could explain 49% of the variation in the number of emergency department visits (R2 = 49%, r = 0.70). Carbon monoxide, barometric pressure, and relative humidity were also statistically significant predictors but were clinically insignificant, explaining only a few percentage points of the total variation. By taking advantage of the seasonal pattern of wheezing through the use of temperature velocity, predictive models for asthmatic wheezing can be greatly improved. They may also aid in planning emergency department staffing, and even help prevent emergency department visits by premedication or lifestyle change during high-risk periods.