Table 4—

Logistic regression with stepwise selection procedure for each reversibility criterion

Variables#OR (95% CI)p-value
ΔFEV1 ≥12% and ≥200 mL
 Sex0.610 (0.538–0.691)<0.0001
 Age yrs0.978 (0.971–0.984)<0.0001
 Smoking history pack-yrs1.003 (1.001–1.005)00.0006
 SGRQ0.985 (0.982–0.989)<0.0001
 FEV10.982 (0.978–0.987)<0.0001
ΔFEV1 ≥15%
 Sex1.371 (1.191–1.578)<0.0001
 Smoking history pack-yrs1.003 (1.001–1.005)0.0135
 SGRQ0.984 (0.981–0.988)<0.0001
 FEV10.926 (0.921–0.931)<0.0001
ΔFEV1 ≥10% pred
 Sex1.314 (1.159–1.490)<0.0001
 Smoking history pack-yrs1.003 (1.001–1.005)0.0063
 SGRQ0.986 (0.983–0.990)<0.0001
 FEV10.988 (0.983–0.992)<0.0001
  • An event in the definition of the odds ratio (OR) is a responsive patient according to one of the reversibility criteria. The ORs for the continuous variables (age, percentage predicted forced expiratory volume in one second (FEV1), St George’s Respiratory Questionnaire (SGRQ) total score and smoking history) are calculated according to a one-unit increase in these variables. CI: confidence interval; Δ: change; % pred: % predicted. #: selected from age, sex (female versus male), smoking status, smoking history, chronic obstructive pulmonary disease duration (in years), baseline percentage predicted FEV1 before bronchodilators and SGRQ total score. Variables with a p-value of <0.05 were kept in the model.