Abstract
OBJECTIVES: To compare physical information, such as age, sex, height, weight, body mass index (BMI) and pulmonary function test (PFT) results, between cough variant asthma (CVA) and chronic cough (CC) and establish a diagnostic model of CVA. DESIGN: A case-control study of patients with suspected CVA enrolled at The First Affiliated Hospital of Zhejiang Chinese Medical University. SETTING: One leader unit of the National Key Specialised Pulmonary Disease Cooperation Group in China. PARTICIPANTS: Enrolled 545 patients who underwent PFT and bronchial provocation tests. OUTCOME MEASURES: We obtained physical information and pulmonary test data and established the model using logistic regression analysis. The Hosmer-Lemeshow goodness-of-fit test, area under the receiver operating characteristic curve (AUC), calibration plot and decision curve analysis were used to evaluate this model. All data were analysed using SPSS V.27 and RStudio software. RESULTS: The CVA group had more female patients (%) (68.12% vs 51.48%, p value<0.001) and lower height (m) (1.61 (0.40) vs 1.65 (3.26), p value<0.001), weight (kg) (60 (56) vs 63 (85), p value<0.001) and BMI (kg/m(2)) (22.59 (17.91) vs 23.28 (21.81), p value=0.016) than the CC group. Differences between CVA and CC in forced vital capacity (FVC) in percent predicted values (FVC% pred)(94.4 (57.3) vs 91.60 (94.10), p value=0.006), forced expiratory volume in 1 s/FVC (FEV1/FVC) (%) (84.65±6.82 vs 86.91±6.71, p value<0.001), peak expiratory flow in per cent predicted values (PEF% pred) (93.00 (81.10) vs 98.00 (108.00), p value=0.005), maximal mid-expiratory flow in percent predicted values (MMEF% pred) (74.50 (100.60) vs 90.85 (170.30), p value<0.001), forced expiratory flow (FEF) at 50% of FVC in per cent predicted values (FEF(50%) pred) (78.9(113.50) vs 93.10(169.80), p value<0.001) and FEF at 75% of FVC in per cent predicted values (FEF(75%) pred) (69.70 (137.60) vs 85.60 (225.80), p value<0.001) were significant. Patients with CVA were more in number compared with patients with CC at a lower degree (<65%) of MMEF% pred (32.37% vs 14.50%, p value<0.001), FEF(50%) pred (26.09% vs 13.02%, p value<0.001) and FEF(75%) pred (39.13% vs 23.67%, p value<0.001). FVC% pred, FEV1/FVC, BMI and MMEF% pred aided in establishing a model with an AUC of 0.733 (95% CI: 0.6829 to 0.7831). The model was tested using internal and external data (p value=0.2865 and p value=0.3197, respectively). CONCLUSION: BMI, FVC% pred, FEV1/FVC (%) and MMEF% pred were used to establish the diagnostic model. Our model potentially indicates CVA. TRIAL REGISTRATION NUMBER: NCT06199830.