Applying an LDL-C threshold-based approach to identify individuals with familial hypercholesterolemia

Abstract

Familial hypercholesterolemia (FH) remains underdiagnosed and undertreated. The optimal electronic health record (EHR) screening strategy for FH is unclear. To evaluate an LDL-C threshold-based approach of identifying patients with FH from the EHR to determine the optimal LDL-C range for FH consideration. Individuals from UT Southwestern Medical Center with an LDL-C level ≥190mg/dL at any time were enrolled in an FH registry. These 5,786 patients were divided into four categories of LDL-C (190- 219, 220 - 249, 250 - 299, and ≥ 300mg/dL) with 100 individuals randomly selected for manual chart review in each category. Chart review included 1) the presence of secondary causes of dyslipidemia, 2) diagnosis of possible/definite FH by modified Simon Broome criteria, and 3) probable/definite FH by modified Dutch Lipid Clinic Network (DLCN) criteria. Of the 400 individuals with an LDL-C level ≥190mg/dL (mean age 52 years ± 14), the presence of secondary causes increased across each LDL-C category (p < 0.001) with the greatest prevalence in those ≥ 300mg/dL (52%). The prevalence of possible/probable or definite FH also varied by LDL-C category, with the highest prevalence of FH by Simon Broome criteria in the 220 - 249mg/dL category (52%) and by DLCN criteria in the 250 - 299mg/dL category (46%). Among those with LDL-C ≥ 190mg/dL, the prevalence of secondary causes increased markedly with higher LDL-C, while the diagnosis of FH has a parabolic relationship. Patients with intermediate LDL-C (220 - 299mg/dL) may be the optimal group to prioritize for FH screening.