Evaluating the sensitivity of the ACC/AHA pooled cohort risk calculator to predict atherosclerotic cardiovascular events within 10 years: how many events are we failing to predict?

Abstract

Abstract Background The ACC/AHA Pooled Cohort Equation (PCE) for atherosclerotic cardiovascular disease (ASCVD) has been recommended as the initial step in cardiovascular risk assessment. The sensitivity of this tool to detect those who will develop ASCVD within 10-years, while considering age and sex groups, has not been extensively studied. Methods Using the Rochester Epidemiology Project (REP) we evaluated a community-based cohort of consecutive patients that sought primary care in Olmsted County, MN, between the years 1998–2000 and were followed up through March 1st 2016. Inclusion criteria were ages 40–79 and complete data to calculate the PCE. We excluded those with known ASCVD, atrial fibrillation or heart failure. Criteria were similar to those used to derive the PCE. Events were validated in duplicate and included fatal and non-fatal myocardial infarction and ischemic stroke. Patient information was ascertained using the record linkage system of the REP. Follow-up was truncated at 10 years. We assessed the ASCVD predicted risk (categorized as low <5%, intermediate 5–9.9%, high 10–19.9%, and very high ≥20% risk) at baseline, in subjects having an ASCVD event within 10-years in the community across age (<65 years) and sex categories. We also categorized ideal cardiovascular health as ≥4 metrics [non-smoker, body mass index <25 kg/m2, and not having of elevated blood pressure (≥130/80 mmHg), LDL cholesterol (>100 mg/dL), or fasting blood glucose (>100 mg/dL), in the absence of a medical diagnosis or treatment]. Results We included 30,042 adults, mean ± SD age 48.5±12.2 years, 54% women, with a median follow-up of 16.5±5.3 years. There were 1,555 ASCVD events (5.2%) at 10 years of follow-up. The performance of the PCE was similar to what was described in the original report (0.78 vs 0.79). Overall, among those who suffered an ASCVD, 54% of women and 41% of men were not high risk as predicted by PCE (Figure 1A). Most women (73%) <65 years of age would had been considered low risk within 10-years before the event, and only 10% would have been considered to be high risks (Figure 1B). Nonetheless, women <65 years who had an ASCVD event and low 10-year predicted ASCVD risk by PCE were less likely to have ideal cardiovascular health [55 (0.40%) vs 3884 (28.39%), p-value<0.0001], when compared to women in the low risk category without an event. Conclusion The PCE fails to identify most women who will develop an ASCVD event, particularly women <65 years of age. These results underscore the importance of using additional information when estimating ASCVD risk among women and the need for better cardiovascular risk prediction tools. Figure 1 Funding Acknowledgement Type of funding source: Foundation. Main funding source(s): Mayo Clinic