The Predictive Value of Baseline Spirometry, Age, and Gender for Airway Hyper-Responsiveness in Adults With Suspected Asthma

Background: Bronchial provocation test (BPT) plays a crucial role in diagnosing airway hyper-responsiveness (AHR) among patients with asthma-like symptoms. Given that BPT may induce severe bronchospasm, baseline spirometric parameters are expected to help predict a positive outcome, guiding the decision of referring a patient to BPT. Methods: Baseline spirometry and BPT conducted on suspected asthmatics between July 2011 and 2013 at Department of Pulmonary Medicine, Zhongshan Hospital, were retrospectively collected. Baseline characteristics from positive and negative BPT groups were compared. The predictive accuracy of single parameter for AHR was evaluated by plotting receiver operating characteristic curve, and a composite model was developed to improve the accuracy by the logistic regression. The relationships of FEV₁/FVC, FEV₁ %predicted, and forced expiratory flow at 50% of FVC exhaled (FEF₅₀ %predicted) with the provocation dose causing a 20% fall in FEV₁ (PD₂₀-FEV₁) were examined. Results: The positive BPT group exhibited reduced FEV₁ %predicted, FEV₁/FVC, and small airway function parameters compared with the negative group. Among these parameters, FEF₅₀ %predicted, forced expiratory flow at 75% of FVC exhaled (FEF₇₅ %predicted), and forced expiratory flow between 25% and 75% (FEF_25-75 %predicted) demonstrated significant accuracy. Notably, females and younger subjects were more prone to a positive outcome. A predictive model that combined FEV₁ %predicted (cutoff 98.65%, area under the curve 0.714, sensitivity 39.10%, specificity 88.10%), FEV₁/FVC (cutoff 79.22%, area under the curve 0.690, sensitivity 32.80%, specificity 90.50%), and FEF₅₀ %predicted (cutoff 74.45%, area under the curve 0.761, sensitivity 50.50%, specificity 84.60%) with age and gender was found to enhance the accuracy for a positive BPT (area under the curve 0.786, 95% CI 0.758-0.814, sensitivity 52.27%, specificity 86.14%). Moreover, FEV₁/FVC, FEF₅₀ %predicted, and FEF_25-75 %predicted showed differences among groups with varying AHR levels. The significant correlations between these 3 parameters and PD₂₀-FEV₁ were exclusively demonstrated in the severe AHR group. Conclusions: This study revealed that FEV₁ %predicted, FEV₁/FVC, and FEF₅₀ %predicted along with age and gender were predictors of AHR in subjects with suspected asthma. Their combination improved the predictive accuracy over using FEF₅₀ %predicted alone, thus offering a complement for clinical decision-making regarding referrals to BPT.