Finding missed cases of familial hypercholesterolemia in health systems using machine learning.

Familial hypercholesterolemia (FH) is an underdiagnosed dominant genetic condition affecting approximately 0.4% 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%. As part of the FH Foundation's FIND FH initiative, we developed a classifier to identify potential FH patients using electronic health record (EHR) data at Stanford Health Care. We trained a random forest classifier using data from known patients (n = 197) and matched non-cases (n = 6590). Our classifier obtained a positive predictive value (PPV) of 0.88 and sensitivity of 0.75 on a held-out test-set. We evaluated the accuracy of the classifier's predictions by chart review of 100 patients at risk of FH not included in the original dataset. The classifier correctly flagged 84% of patients at the highest probability threshold, with decreasing performance as the threshold lowers. In external validation on 466 FH patients (236 with genetically proven FH) and 5000 matched non-cases from the Geisinger Healthcare System our FH classifier achieved a PPV of 0.85. Our EHR-derived FH classifier is effective in finding candidate patients for further FH screening. Such machine learning guided strategies can lead to effective identification of the highest risk patients for enhanced management strategies.

Clinical and prognostic implications of the initial response to aspirin in patients with acute coronary syndrome.

Increased platelet reactivity and decreased response to antiplatelet drugs may result in recurrent ischemic events after acute coronary syndrome (ACS). We evaluated laboratory response to aspirin in patients with ACS before and after percutaneous coronary intervention (PCI) and assessed its effect on major adverse clinical events. Sixty-three consecutive patients with ACS were tested for response to aspirin by light transmittance aggregometry (LTA) and the IMPACT-R test (with arachidonic acid) before and 2 to 4 days after PCI and clopidogrel loading. Patients were followed for clinical events up to 15 months from PCI. Response to aspirin improved significantly after PCI and clopidogrel treatment (mean arachidonic acid-induced LTA decreased from 34.9 ± 3.35% before PCI to 15.2 ± 2.2% and surface coverage increased from 2.2 ± 0.27% to 6.2 ± 0.6%, p <0.0001 for the 2 methods). Improved response to aspirin after PCI correlated with response to clopidogrel (LTA and IMPACT-R, p <0.01). Patients with good laboratory response to aspirin before but not after PCI had a significantly lower major cardiovascular event rate during 15-month follow-up in multivariate analysis. In conclusion, laboratory response to aspirin is highly dynamic in patients with ACS. Improved response to aspirin after PCI may result from stabilization of coronary artery disease and/or clopidogrel treatment. Laboratory response to aspirin before PCI and clopidogrel loading is a sensitive marker for platelet reactivity that correlates with clinical outcome in patients with ACS.