Comparing the Performance of Two Screening Questionnaires for Chronic Obstructive Pulmonary Disease in the Chinese General Population

Abstract

Purpose: Screening questionnaires can help identify individuals at a high risk of COPD. This study aimed to compare the performance of the COPD population screener (COPD-PS) and COPD screening questionnaire (COPD-SQ) on the general population as a full cohort and stratified by urbanization. Methods: We recruited subjects who underwent a health checkup at urban and rural community health centers in Beijing. All eligible subjects completed the COPD-PS and COPD-SQ, then spirometry. Spirometry-defined COPD was defined as a post-bronchodilator FEV1 /FVC<70%. Symptomatic COPD was defined as a post-bronchodilator FEV1 /FVC<70% and respiratory symptoms. Receiver operating characteristic (ROC) curve analysis compared the discriminatory power of the two questionnaires, and stratified by urbanization. Results: We identified 129 spirometry-defined and 92 symptomatic COPD cases out of 1350 enrolled subjects. The optimal cut-off score for the COPD-PS was 4 for spirometry-defined and 5 for symptomatic COPD. The optimum cut-off score for the COPD-SQ was 15 for both spirometry-defined and symptomatic COPD. The COPD-PS and COPD-SQ had similar AUC values for spirometry-defined (0.672 vs 0.702) and symptomatic COPD (0.734 vs 0.779). The AUC of the COPD-SQ tended to be higher in rural areas than that of the COPD-PS for spirometry-defined COPD (0.700 vs 0.653, P = 0.093). Conclusion: The COPD-PS and COPD-SQ had comparable discriminatory power for detecting COPD in the general population while the COPD-SQ performed better in rural areas. A pilot study for validating and comparing the diagnostic accuracy of different questionnaires is required when screening for COPD in a new environment.