Cardiovascular disease (CVD) clinical risk estimation models, such as the Framingham Risk Score and the Pooled Cohort Equations, estimate the risk of a first CVD event using routinely collected risk factors-age, blood pressure, diabetes, smoking status, and cholesterol levels. These risk estimates can be used to inform shared decision-making with patients regarding lifestyle improvements, risk factor control, and statin therapy. Because inherent uncertainty in predictions persists, even in well-calibrated and externally validated models, there is substantial interest in improving existing risk models by incorporating nontraditional risk factors and, ultimately, reducing CVD events and mortality. A coronary artery calcium score (CACS) is a radiologic marker of the degree of calcification in the coronary arteries; it has been proposed to improve current risk prediction. 1 In this issue of JAMA Internal Medicine, Dr Bell and colleagues 2 report the findings of a meta-analysis demonstrating that addition of CACS to traditional CVD risk models was associated with improvement in model discrimination, with a pooled gain in C statistic of 0.036 (95% CI, 0.020-0.052)-a small but statistically significant finding of unknown clinical significance. The authors based their literature search 2 on a prior systematic review that was used by the US Preventive Services Task Force for its 2018 guidelines, which concluded that there was insufficient evidence to recommend adding CACS to traditional risk models. 3 Six studies were found to meet the study criteria for inclusion in the meta-analysis by Dr Bell and colleagues. 2 Even without the addition of CACS, the clinical models they identified were found to have good discrimination, with C statistics ranging from 0.69 to 0.80.Although the addition of CACS to traditional models in this analysis modestly increased discrimination, 2 a measure of how often a model estimates higher risk for patients who experience an event compared with those who do not, it is only 1 measure of a model's performance. The authors did not include data on whether calibration-the ability of a model to assign accurate probabilities of an event occurring-was improved by the addition of CACS. The lack of reporting on calibration has been noted as a widespread problem in developing cardiovascular prediction models. 4 Furthermore, the addition of CACS to traditional models was not associated with a clear improvement in net reclassification in this study. Although the addition of CACS was associated with increased numbers of patients who