Screening for undiagnosed COPD; accuracy of different strategies among primary care patients in China: Breathe Well study

Aim: To identify the most effective screening strategy for identifying undiagnosed COPD amongst primary care patients in China. Methods: Cross-sectional study in eight community hospitals from four cities, with participants completing all index tests (screening questionnaires [CDQ, CAPTURE, Symptom-based questionnaire, COPD-SQ], microspirometer [COPD-6] and peak flow [USPE]) and the reference test (ndd Easy On-PC spirometer). Test performance of all screening tests and strategies were compared against the reference test. Cases were defined as those below the lower limit of normal (LLN-GLI) on the reference test. Results: 2439 participants completed the index and reference test. Participants had a mean age of 59.8 (SD 9.6) years, 39.0% (n=950) were male, 69.0% (n=1682) were never-smokers and 3.4% (n=84) had an existing COPD diagnosis. 13.7% (n=335) of participants had spirometry-confirmed airflow obstruction; amongst those with no previous COPD diagnosis, the prevalence was 10.7%. The microspirometer had a sensitivity of 52.7% (95% CI 47.2%-58.2%) and specificity 94.3% (95% CI 93.2%- 95.2%) using a cut-point of <0.75. The peak flow meter had a sensitivity of 65.9% (95%CI 60.5%-70.9%) and specificity of 82.5% (95%CI 80.8%-84.1%) using a cut-point of <350 litres/min for men and <250 litres/min for women. The least accurate screening questionnaire was the CDQ with a sensitivity of 46.0% (95%CI 40.5%-51.5%) and a specificity of 21.3% (95%CI 19.6%-23.1%), while the Chinese symptom-based questionnaire had the best accuracy with a sensitivity of 61.8% (95%CI 56.4%-67.0%) and a specificity of 74.2% (95%CI 72.3%-76.1%). Conclusion: The prevalence of spirometry-confirmed COPD was 13.7% in the overall sample, and 10.7% amongst those with no previous diagnosis. In our sample, the peak flow meter had slightly better test performance than the microspirometer at the cut-points chosen, and the Chinese symptom-based questionnaire was the most accurate screening questionnaire. Alternative cut-points and combinations of tests need to be further evaluated. Implementation Science/Service Development Research Ideas on Respiratory Conditions and Tobacco Dependency Abstract Declaration of Interest The authors declare no conflicts of interest. The NIHR Global Health Research (GHR) Programme is funded through UK Official Development Assistance (ODA) via the Department of Health and Social Care (DHSC). The study is registered at http://www.isrctn.com (ISRCTN13357135). References and Clinical Trial Registry Information The views expressed in this publication are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. We gratefully acknowledge IPCRG for introducing us to the primary care networks involved in this study and for its continued facilitation of clinical engagement