Predicting 10-year risk of recurrent cardiovascular events and cardiovascular interventions in patients with established cardiovascular disease: results from UCC-SMART and REACH

Abstract Background Existing cardiovascular risk scores for patients with established cardiovascular disease (CVD) estimate residual risk of recurrent major cardiovascular events (MACE). The aim of the current study is to develop and externally validate a prediction model to estimate the 10-year combined risk of recurrent MACE and cardiovascular interventions (MACE+) in patients with established CVD. Methods Data of patients with established CVD from the UCC-SMART cohort (N = 8421) were used for model development, and patient data from REACH Western Europe (N = 14,528) and REACH North America (N = 19,495) for model validation. Predictors were selected based on the existing SMART risk score. A Fine and Gray competing risk-adjusted 10-year risk model was developed for the combined outcome MACE+. The model was validated in all patients and in strata of coronary heart disease (CHD), cerebrovascular disease (CeVD), peripheral artery disease (PAD). Results External calibration for 2-year risk in REACH Western Europe and REACH North America was good, c-statistics were moderate: 0.60 and 0.58, respectively. In strata of CVD at baseline good external calibration was observed in patients with CHD and CeVD, however, poor calibration was seen in patients with PAD. C-statistics for patients with CHD were 0.60 and 0.57, for patients with CeVD 0.62 and 0.61, and for patients with PAD 0.53 and 0.54 in REACH Western Europe and REACH North America, respectively. Conclusions The 10-year combined risk of recurrent MACE and cardiovascular interventions can be estimated in patients with established CHD or CeVD. However, cardiovascular interventions in patients with PAD could not be predicted reliably.