COPD and disease-specific health status in a working population

BACKGROUND: It has been debated whether treatment should be started early in subjects with mild to moderate COPD. An impaired health status score was associated with a higher probability of being diagnosed with COPD as compared with undiagnosed COPD. PURPOSE: To investigate the health status in a healthy working population, to determine reference scores for healthy non-smoking subjects, and to investigate the relationship between their health status and airflow limitation. METHODS: A total of 1333 healthy industrial workers aged ≥40 years performed spirometry and completed the St. George’s Respiratory Questionnaire (SGRQ) and the COPD Assessment Test (CAT). RESULTS: The prevalence of COPD defined by the fixed ratio of the forced expiratory volume in one second (FEV(1))/forced vital capacity (FVC) was 10.9%, and the prevalence defined by the Lower Limit of Normal was 5.0%. All SGRQ and CAT scores were skewed to the milder end. In 512 non-smoking subjects with normal spirometry, the mean SGRQ score was 5.7, and the mean CAT score was 5.8. In 145 people with COPD defined by the fixed ratio, the mean SGRQ score was 7.9, with a zero score in 6.9% of the subjects. Using the CAT, the mean score was 7.3, with 7.6% of the scores being zero. The scores in patients identified using the Lower Limit of Normal approach were: SGRQ 8.4 (13.4% had a score of zero) and CAT 7.4 (13.4% had a score of zero). Although the 95th percentiles of the Total, Symptoms, Activity, and Impact scores of the SGRQ and CAT sores were 13.8, 34.0, 23.4, 7.2 and 13.6 in the 512 healthy non-smoking subjects, respectively, they were also distributed under their upper limits in over 80% of the COPD subjects. CONCLUSION: The COPD-specific health status scores in a working population were good, even in those with spirometrically diagnosed COPD. All scores were widely distributed in both healthy non-smoking subjects and in subjects with COPD, and the score distribution overlapped remarkably between these two groups. This suggests that symptom-based methods are not suitable screening tools in a healthy general population.