Log in to MyACC
CANHEART Population-Based Lab Prediction Models for ASCVD
Quick Takes
- ASCVD risk prediction models using age and lab results yield models with c-statistics of 0.77 in women and 0.71 in men for the internal cohort and 0.72 for both sexes in the external validation cohort.
- These sex-specific models were not statistically significantly different from those for the pooled cohort equations.
Study Questions:
Can sex-specific prediction models for atherosclerotic cardiovascular disease (ASCVD) using age and routine laboratory tests be developed and validated?
Methods:
The investigators used data from a population-based cohort study (CANHEART [Cardiovascular Health in Ambulatory Care Research Team] Lab Models) for the present analysis. Both the derivation and internal validation cohorts included adults aged 40–75 years without cardiovascular disease collected from April 2009–December 2015. The external validation cohort included primary care patients from January 2010–December 2014. Age and laboratory predictors collected in the outpatient setting included serum total cholesterol, high-density lipoprotein cholesterol, triglycerides, hemoglobin, mean corpuscular volume, platelets, leukocytes, estimated glomerular filtration rate, and glucose. The ASCVD outcomes were defined as myocardial infarction, stroke, and death from ischemic heart or cerebrovascular disease within 5 years.
Results:
Data from 2,160,497 women and 1,833,147 men were used to develop sex-specific models, which were then internally validated. Mean was age 64 years. Hypertension was the most common risk factor. Over a median follow-up of 7.8 years, there were 40,759 ASCVD events in women and 71,664 in men. There was a relative difference of <1% between mean predicted and observed risk for both sexes. The c-statistic was 0.77 in women and 0.71 in men. External validation in 31,697 primary care patients (mean age 54 years, mean follow-up 7.3 years) showed a relative difference of <14% and an absolute difference of <0.3 percentage points in mean predicted and observed risks for both sexes. Calibration plots showed that overestimation was greater for higher-risk patients but was well-calibrated overall for low-risk patients. The c-statistics for the laboratory models were 0.72 for both sexes and were not statistically significantly different from those for the pooled cohort equations (PCEs) in women (change in c-statistic, -0.01; 95% confidence interval [Cl], -0.03 to 0.01) or men (change in c-statistic, -0.01; 95% Cl, -0.04 to 0.02).
Conclusions:
The authors conclude that the CANHEART Lab Models predict ASCVD with similar accuracy to more complex models, such as the PCEs.
Perspective:
Current risk prediction models remain underutilized in clinical care. Models that depend solely on laboratory results and age may be easier to automate within an electronic medical record and thus more readily incorporated into clinical care for primary prevention.
Clinical Topics: Prevention
Keywords: Atherosclerosis, Risk Assessment
< Back to Listings
