DETERMINANTS OF RESPONSE TO BECLOMETHASONE AEROSOL AT VARIOUS DOSAGE LEVELS - MULTIPLE-REGRESSION ANALYSIS TO IDENTIFY CLINICALLY USEFUL PREDICTORS

[摘要] In a dose-response study of beclomethasone aerosol (BA) therapy considerable variation was noted in responsiveness of the 34 participating asthma patients. The data were analyzed by stepwise multiple regression to identify factors which favored or inhibited the response to the drug. Among the 33 variables tested, the significant determinant variables differed depending on the BA dosage. The range studied was 200, 400, 800 and 1600 .mu.g/day. The reliability of several sets of these criteria for predicting the results of a therapeutic trial of BA in a given patient were checked by a cross-validation method. This demonstrated 79.4% prognostic accuracy for a set of 4 clinical variables and 88.2% for a set of 6 clinical and laboratory variables. Those patients whose symptom improvement was poor, despite receiving 1600 .mu.g BA/day, differed from the responsive patients mainly in the greater degree of chronic obstructive lung disease (COLD) associated with asthma. The symptom response to this dosage of BA in this group tended to be more favorable in nonsmoking, female asthmatic patients with frequent asthma attacks but relatively little associated COLD and no recurrent mucopurulent bronchitis. A therapeutic trial of BA is likely to benefit patients with long-term asthma regardless of its pathogenetic classification or chronicity. They further suggest that a trial with high-dose BA rather than low dosage can give more accurately predictable results.