HealthDay News — The FIND FH model can successfully identify individuals with familial hypercholesterolemia (FH) from a scan of large, diverse health care encounter databases, according to a study published in The Lancet Digital Health.
Kelly D. Myers, from the Familial Hypercholesterolemia Foundation in Pasadena, California, and colleagues trained the FIND FH machine learning model using deidentified health care encounter data (e.g., procedure and diagnostic codes, prescriptions, and laboratory findings) from 939 clinically diagnosed individuals with FH (395 of whom had a molecular diagnosis) and 83,136 individuals presumed free of FH, sampled from four US institutions. Model validation occurred with application to a national health care encounter database (170 million individuals) and an integrated health care delivery system dataset (174,000 individuals).
The researchers found that applying the model with a measured precision (positive predictive value) of 0.85, recall (sensitivity) of 0.45, area under the precision-recall curve of 0.55, and area under the receiver operating characteristic curve of 0.89, the model identified 1,331,759 (of 170,416,201) patients in the national database and 866 (of 173,733) individuals in the health care delivery system dataset as likely to have FH. A review of the flagged individuals by FH experts categorized 87% in the national database and 77% in the health care delivery system dataset as having a high enough clinical suspicion of FH to warrant guideline-based clinical evaluation and treatment.
“We no longer need to screen everyone to find individuals who are at genetic risk for heart attacks and strokes,” a coauthor said in a statement.
Several authors disclosed financial ties to the pharmaceutical industry, which funded the study.