Expression of bronchodilator response: comparison of four indices

Several indices have been suggested to express the bronchodilator response but there is no agreement on what is the ideal expression. The forced expiratory volume in I s (FEV1) was measured in 190 patients of bronchial asthma before and after inhalation of 200 micrograms salbutamol. Four different indices of bronchodilator response were calculated. The change in FEV1 was expressed as absolute value (delta FEV1), as a percentage of initial FEV1 (delta FEV1% initial), as a percentage of the predicted normal value (delta FEV1% predicted) and as a percentage of maximum achievable reversibility, i.e. the difference between predicted and initial values (delta FEV1% [predicted-initial]). A significant negative correlation was observed between delta FEV1 initial and initial FEV1% predicted (r = -0.33, P < 0.0001) but not delta FEV1 and initial FEV1% predicted, or between delta FEV1% predicted and initial FEV1% predicted. The values of delta FEV1% [predicted-initial] tended to reach towards infinity as the initial FEV1% predicted approached 100%. delta FEV1% predicted had similar sensitivity but greater specificity than delta FEV1% initial in detecting a true bronchodilator response. It was concluded that being dependent on the initial value, delta FEV1% initial was not an appropriate method to express the bronchodilator response. Although all other methods had limitations, delta FEV1 and delta FEV1% predicted appeared to be better indices.