PERFORMANCE OF THE FIND-FH MACHINE LEARNING ALGORITHM FOR THE IDENTIFICATION OF INDIVIDUALS WITH SUSPECTED FAMILIAL HYPERCHOLESTEROLEMIA

Abstract

Therapeutic Area

CVD Prevention – Primary and Secondary

Background

Familial Hypercholesterolemia (FH) is an inherited disorder of cholesterol metabolism that is markedly underdiagnosed. This study evaluated the real-world performance of the FIND-FH score, a novel machine learning algorithm, in the identification of individuals with high likelihood of FH.

Methods

The FIND-FH model was applied to electronic health record (EHR) data from a large academic medical center. Manual chart review was performed to determine the diagnosis of FH by Simon Broome and Dutch Lipid Clinic Network (DCLN) criteria. Individual characteristics were compared across quintiles of the FIND-FH score. Individuals deemed suitable for clinical outreach for FH were identified using predetermined criteria.

Results

Of the 93,418 individuals with adequate EHR data, the FIND-FH algorithm identified 340 with high probability of FH, after appropriate exclusions. These individuals were mean age 49.8 years, 59% male, and mean highest LDL-C of 168.4 mg/dL (±51.9). A total of 20-32% met diagnostic criteria for at least possible FH based on available EHR data. When stratifying by FIND-FH score, several variables differed significantly by quintile, including Simon-Broome and DLCN probability. In the entire cohort, 191 (56%) had enough clinical suspicion for FH to warrant outreach. Among these, 101 (53%) had highest LDL-C <190 mg/dL and would be missed by traditional LDL-C based FH screening strategies.

Conclusions

In a large academic healthcare system EHR cohort, most individuals identified as higher risk for FH by the FIND-FH algorithm were deemed appropriate for further evaluation, even when EHR data alone could not confirm the clinical diagnosis of FH. This algorithm can be used as an adjunct to traditional LDL-C screening strategies to identify individuals with FH.