Cholesterol and hypertension treatment but not time-dependent covariates or competing risks improve coronary risk prediction: the REGICOR study

Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Carlos III Health Institute Agency for Management of University and Research Grants

Cardiovascular (CV) risk functions are the recommended tool to identify high-risk individuals. However, discrimination of CV risk functions is not optimal. While the effect of biomarkers in CV risk prediction has been extensively studied, there is no data of CV risk functions including time-dependent covariates, competing risks and treatments.

To examine the effect of including time-dependent covariates, competing risks and treatments in CV risk prediction.

Participants from the REGICOR population cohorts (North-Eastern Spain) aged 35-74 years without previous history of cardiovascular disease were included (n=8,470). Coronary and stroke events, and mortality due to other CV causes or to cancer were recorded during the follow-up (median=12.6 years). A multi-state Markov model was constructed to include competing risks and time-dependent classical risk factors and treatments (2 measurements). This model was compared to Cox models with the basal measurement of classical risk factors, treatments or competing risks. Models were cross-validated and compared by their discrimination (area under the ROC curve), calibration (Hosmer-Lemeshow test) and reclassification (categorical net reclassification index).

Cancer mortality was the event with the highest cumulative incidence. In coronary event prediction, cholesterol and hypertension treatment addition to classical risk factors, improved significantly discrimination by 2% and reclassification by 7-9%. In stroke event prediction, inclusion of time-dependent covariates decreased significantly discrimination by 3-5%.

Coronary risk prediction improves when cholesterol and hypertension treatment are included in risk functions. Coronary/stroke prediction does not improve with 2 measurements of covariates or with competing risks.