Abstract Body (Do not enter title and authors here): Introduction: PREVENT is a new prediction tool used for cardiovascular risk management. The purpose of this study was to investigate whether the PREVENT risk score is more accurate than the traditional Pooled Cohort Equations (PCE) in predicting atherosclerotic cardiovascular disease (ASCVD).
Hypothesis: It is hypothesized that the novel PREVENT equations will outperform the PCE equations by including kidney function parameters and different risk coefficients in the calculation.
Methods: We used the All of Us dataset which is a national dataset of diverse patients aimed to advance precision medicine, and it has more than 840,000 patients. We identified participants who were not taking statins, did not have a history of cardiovascular events and had at least 10 years of follow-up unless they experienced an event before the 10-year mark. We calculated both the PREVENT and PCE scores for ASCVD which was defined as myocardial infarction, stroke (hemorrhagic or ischemic) or heart failure. Participants were then followed to assess the occurrence of any adverse outcomes. The strength of the model was calculated using standard C-statistics and DeLong’s t-test was applied to evaluate statistical significance. The difference in proportion of adverse events was calculated using Chi-squared test.
Results: A total of 29,660 participants met out inclusion criteria. The average age was 57.4 and 53.1% were women. White participants represented 66% of the cohort, while black participants represented16.5%. Hispanic participants represented 11.8% the cohort. A total of 29.8% of the participants experienced an event. The PREVENT model demonstrated significantly better discriminative performance compared to the PCE model (C-statistic 0.721 vs 0.706; p= 0.0012). The PREVENT model performed numerically better in all subgroups, although this difference was not consistently statistically significant, table 1.
Conclusions: In a national cohort with up to 10-year follow-up, the PREVENT calculator demonstrated superior accuracy compared to the PCE models for predicting cardiovascular risk.