Impact of GLI-2012 Spirometric References and Lower Limit of Normal on Prevalence of COPD in Older Urban Thai Persons
Dejsomritrutai, W.; Chuaychoo, B.
Journal of the Medical Association of Thailand 99(3): 276-281
2016
ISSN/ISBN: 0125-2208 PMID: 27276738 Document Number: 16583
Previous analysis of survey data from a study in older urban Thai persons reported prevalence of chronic obstructive pulmonary disease (COPD) of 7.11% according to formerly recommended criteria that used fixed ratio (70%) of pre-bronchodilator FEV1/FVC as the cut-off value. In 2012, the Global Lung Initiative (GLI)-2012 equations for spirometric reference were established and the diagnostic criteria for COPD were changed. To reanalyze the data collected during the 2002 respiratory health survey to determine prevalence of COPD in older urban Thai persons according to the new GLI-2012 equations. Demographic and clinical data of 3,094 subjects aged ≥ 60 years from 124 urban communities were re-analyzed. Prevalence of COPD determined by post-bronchodilator FEV1/FVC and GLI-2012 equations using lower limit of normal by the Lambda-Mu-Sigma method (LMS-LLN) was compared to COPD prevalence findings based on criteria used in previous analysis. Prevalence of COPD in the study population using post-bronchodilator FEV1/FVC and LMS-LLN was 6.50% (95% CI: 5.63-7.3 7), compared with 7.82% (95% CI: 6.87-8.77) when fixed ratio (70%) of FEV1/FVC was used as cut-off. Diagnostic agreement of COPD between LMS-LLN and fixed ratio was high (kappa 0.88, p < 0.0001). However 45/242 (18.6%) subjects diagnosed as COPD by fixed ratio criteria were considered as "over-diagnosed" when LMS-LLN was used as the standard cut-off The same comparison applied to subjects aged 80 years revealed an increased rate of over-diagnosis to 7/22 (31.8%) subjects. Higher agreement was observed (kappa 0.95, p < 0.0001) when comparing between cut-offs for diagnosis of "at least GOLD stage II" COPD. Prevalence of COPD in older urban Thai persons was lower when LMS-LLN was substituted for fixed ratio (70%) of FEV1/FVC. Agreement in COPD diagnosis between both criteria was high, but a substantial proportion of subjects may be over-diagnosed
Document emailed within 1 workday
