The potential role of an expert computer system to augment the opportunistic detection of individuals with familial hypercholesterolaemia from a community laboratory

Abstract

BackgroundFamilial hypercholesterolaemia (FH) is the most common monogenic cause of premature atherosclerotic cardiovascular disease (CVD). However, most individuals with FH remain undiagnosed. We sought to determine if an expert system (ES) at a community laboratory could identify information relevant for estimating an individual's likelihood of FH using the Dutch Lipid Clinic Network criteria (DLCNC). MethodsAn ES (RippleDown®) retrospectively analysed laboratory results and clinical details on the current and previous lipid requests from a community laboratory in Western Australia, over 12months. Results84,823 individuals had ≥1 LDL-cholesterol request with data available on 84,083 (99.1%). Clinical details were provided on 71,282 (84.8%) individuals' current or previous requests. History relevant to the DLCNC was present in 883 (1.1%) individuals, with premature CVD and non-cardiac vascular disease present in 177 and 64 individuals, respectively. Statin therapy was reported in 5118 individuals; 112 individuals with a current LDL-cholesterol of <6.5mmol/L had a previous LDL-cholesterol of ≥6.5mmol/L. ConclusionsThe ES was able to identify information that increased the likelihood of FH in 5471 cases. The ability to detect individuals with premature CVD and to classify them based on their highest LDL-cholesterol may augment FH detection, although further investigation is required to confirm this.