Validation of spirometry prediction equations in western Kenya

Abstract

SETTING:Community of Eldoret, Kenya.OBJECTIVE:To test the performance of three common-ly used spirometry prediction equations in a healthy Kenyan population.DESIGN:Cross-sectional assessment of healthy adults in Eldoret.RESULTS:Of the 331 subjects enrolled in the study, 282subjects aged 18–85 years (45% males, 55% females)produced high-quality spirograms. Lung function pre-dictions were made using the Global Lung Initiative2012 (GLI 2012) prediction equations for African Americans, the National Health and Nutrition Exami-nation Survey III (NHANES III) prediction equations for African Americans, and the Crapo prediction equation.Bland-Altman analyses were performed to measure the agreement between observed and predicted spirometry parameters. Overall, the GLI 2012 and NHANES equations for African Americans performed similarly for forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1), significantly overestimating FVC while accurately predicting observed FEV1values.CONCLUSION:The study brings into question the utility of three major spirometry prediction equations in a Kenyan population. The significant overestimation of FVC by the best-performing equations despite accurate prediction of FEV1suggests poor performance of these equations in our population.