Impulse oscillometry-derived equation for prediction of abnormal FEV1/FVC ratio for COPD screening in Chinese population: a multicenter cross-sectional study.

The diagnosis of chronic obstructive pulmonary disease (COPD) is based on spirometry that requires a forced expiratory manoeuvre, which is laborious and difficult for mass screening. Impulse oscillometry (IOS) is easier than spirometry and performed with tidal breathing. We sought to develop an equation for predicting forced expiratory volume in 1 s (FEV1)/forced vital capacity (FVC) and screening COPD using IOS parameters. Data from patients who simultaneously underwent spirometry and IOS were obtained from databases at five tertiary hospitals in China. Multivariable linear regression analysis was used to develop a predictive model for pre-bronchodilator (BD) FEV1/FVC. Model performance was analyzed against spirometric criteria of airflow obstruction (AO, defined as pre-BD FEV1/FVC < 0.7) and COPD (post-BD FEV1/FVC < 0.7). Using 15,113 patients and externally validated with 9586 patients, the model estimated FEV1/FVC ratio could identified AO and spirometry-defined COPD in internal (AUC = 0.822 and 0.849, respectively) and external (AUC = 0.790 and 0.828, respectively) validation. A clinical algorithm was constructed to classify patients into three different groups: estimated FEV1/FVC < 0.7: likely COPD; estimated FEV1/FVC ≥ 0.7 and ≤0.73: suspicious for COPD; estimated FEV1/FVC > 0.73: unlikely COPD. The sensitivity and specificity for detecting spirometry-defined COPD were 88.0% and 77.0%, respectively, while the negative predictive value ranged from 93.7% to 98.6% and positive predictive value ranged from 26.5% to 62.1% across different COPD prevalence groups in the Chinese population. This equation could be useful to screen for COPD particularly in community and primary care settings. The Financial Budgeting Project of Beijing Institute of Respiratory Medicine.