Using artificial Intelligence and health-encounter data from 933 people with familial hypercholesterolemia and 83,000 without it, researchers say they have developed a model for more precisely screening for the condition that leads to high LDL cholesterol levels and elevated risk for heart disease. By some estimates,1.3 million Americans have undiagnosed familial hypercholesterolemia, an inherited genetic condition.
“It’s fair to say that we think based on all of this data, precision screening for FH is now a reality,” said Daniel J. Rader, MD, one of the study authors and chair of the department of genetics and chief of the division of translational medicine and human genetics at the University of Pennsylvania Perelman School of Medicine, said during a presentation this week.
Rader and his colleagues reported their results in the Lancet Digital Health this week and also presented them at an FH Foundation meeting. The foundation and Amgen, Sanofi, and Regeneron paid for the study.
When the researchers applied their model to a national database of 170 million patients, it flagged 1.3 million as possibly having familial hypercholesterolemia. When experts reviewed a small subset of those flagged people, they found that 87% were candidates for further evaluation and possible diagnosis and treatment.
The researchers used the same process to sift through a smaller database of 173,000 patients at a number of health care systems. The experts said 77% of a small subset of those flagged patients would be in the "should be evaluated" group.