An individualized model for predicting lung function trajectory and risk for airflow limitation was recently validated for the general population as well as for identifying individuals at a higher risk for chronic obstructive pulmonary disease, according to study results published in the journal CHEST.
Spirometry data from patients aged ≥20 years in the Framingham Offspring Cohort (n=4167) were used to train a machine learning algorithm to determine essential predictors of lung function trajectory and risk for airflow limitation. This model was then validated in 2 large, independent multicenter cohorts: Atherosclerosis Risk in Communities (ARIC) and Coronary Artery Risk Development in Young Adults (CARDIA).
The final model included 20 machine-selected predictors of lung function trajectory and risk for airflow limitation. In the 2 validation cohorts (CARDIA; n=2075 and ARIC; n=12,913, respectively), the model explained 88.4% to 90.6% of the variance in follow-up forced expiratory volume in 1 second values. In predicting future risk for airflow limitation, the model had a Brier score of 0.03 to 0.06 in the validation cohorts, suggesting good overall performance in discrimination and calibration.
“We developed, validated, and implemented a personalized risk model to predict the 20-year lung function trajectory and risk of airflow limitation for a given individual from the general population,” the researchers wrote. “Clinical prediction tools are critical enablers of personalization of care, and this study fills an important gap in objective prediction of lung function in the general population.”
Chen W, Sin DD, FitzGerald JM, Safari A, Adibi A, Sadatsafavi M. An individualized prediction model for long-term lung function trajectory and risk of COPD in the general population. CHEST. 2020;157(3):547-557.