Key clinical point: A mixed-effects regression model combined with a machine-learning algorithm predicted which patients are at higher risk of developing chronic obstructive pulmonary disease (COPD).
Major finding: The model showed strength in predicting FEV1 decline (root mean square error range: 0.18L-0.22L) and risk of airflow limitation (C statistic range: 0.86-0.87).
Study details: Retrospective analysis of spirometry assessments in 4,167 subjects and validation in separate cohorts of 2,075 and 12,913 subjects.
Disclosures: This study was funded by the Canadian Institutes of Health Research and Genome Canada: Genome British Columbia. The study authors report no relevant disclosures.
Chen W et al. Chest. 2019 Sep 19.