Researchers associated with a foundation focused on familial hypercholesterolemia (FH) and Stanford University School of Medicine have developed a machine learning algorithm to guide the identification of probable FH individuals within a healthcare system.
FH is an underdiagnosed dominant genetic condition affecting approximately 0.4 percent of the population and has up to a 20-fold increased risk of coronary artery disease if untreated. Simple screening strategies have false positive rates greater than 95 percent. Less than 10 percent of all FH cases are diagnosed.
New research from the FH Foundation’s FIND (Flag, Identify, Network, Deliver) FH initiative leverages machine learning and big data to identify individuals who should be evaluated by clinicians for diagnosis. The study, published in npj Digital Medicine, applied EHR-based screening and demonstrated for the first time that the screening algorithm correctly identified, 84 percent of the time, individuals with the highest probability of having FH.
In addition to validation within Stanford Health Care, the EHR algorithm was independently validated with EHR data from Geisinger Health.
The paper notes that “compared to the implementation of universal genetic testing or clinical criteria-based screening, the economics of EHR-based detection of FH through machine-learning are extremely favorable and can massively improve the ability of a health system to find patients at risk. We believe the use of supervised learning to build a classifier that finds undiagnosed cases of FH is a compelling example of machine learning that matters. As a next step, we are working on deploying the model in a clinical setting, at Stanford Healthcare and at additional sites in partnership with the FH Foundation.”
"It's imperative we change the paradigm for how we identify individuals with FH within healthcare systems because they are at high risk for cardiovascular disease," said Daniel J. Rader, M.D., chair of the department of Genetics in the Perelman School of Medicine at the University of Pennsylvania, Chief Scientific Advisor of The FH Foundation and co-author, in a prepared statement. "Applying advanced technology such as this novel FIND FH algorithm is particularly relevant for FH, because once flagged and diagnosed, individuals can immediately benefit from treatment with readily-available therapies."
