Skip to main content
. 2020 Nov 18;4(5):bjgpopen20X101114. doi: 10.3399/bjgpopen20X101114

Table 2. Model discrimination in the external validation cohort for identifying familial hypercholesterolaemia in general practice (n = 1 030 183).

Models AUC(c-statistic) Standard error a 95% CI
FAMCAT 0.844 0.005 0.834 to 0.854
Simon Broome criteriab 0.730 0.006 0.719 to 0.741
Dutch Lipid Clinic Network criteriac 0.766 0.006 0.755 to 0.778
Cholesterol >99th centiled 0.579 0.005 0.571 to 0.588

aJack-knife procedure to estimate standard errors.20 bTotal cholesterol >7.5 mmol/l or LDL-cholesterol >4.9 mmol/l+ family history of premature myocardial infarction.2 cScore based on LDL-cholesterol, family history, clinical history, and physical examination.9 dThe UK National Institute for Health and Care Excellence recommendation of screening for FH for cholesterol >99th centile. That is, total cholesterol >9.0 mmol/l or LDL-cholesterol >6.6 mmol/l if aged >30 years; total cholesterol >7.5 mmol/l or LDL-cholesterol >4.9 mmol/l if aged ≤30 years.2