Risk scores overestimated risk for CVD in newly diagnosed type 2 diabetes
Available risk scores are in need of recalibration but still serve as useful decision-making tools, according to an ACP Journal Club commentary.
Six different scores for predicting cardiovascular disease (CVD) risk (the QRISK2 score and five diabetes-specific scores) were compared in a recent cohort study of patients with newly diagnosed type 2 diabetes. The study included 181,399 Scottish adults, and it found that none of the scores accurately predicted risk of hospital admission or death from myocardial infarction; stroke; unstable angina; transient ischemic attack; peripheral vascular disease; or coronary, carotid, or major amputation procedures (all c-statistics <0.68).
The study was published in the September Diabetes Care. The following commentary by Michael Tanner, MD, FACP, was published in the ACP Journal Club section of the Nov. 20 Annals of Internal Medicine.
The CVD risk score advocated by the UK's National Institute of Health and Clinical Excellence—QRISK2—overestimated the 5-year risk for incident CVD (24%) in patients with newly diagnosed type 2 diabetes compared with the real-world risk (9.7%) observed in the Scottish National Diabetes Registry. This finding is analogous to the American experience with pooled cohort equations developed to estimate 10-year risk for new atherosclerotic CVD (ASCVD), which are widely used to make decisions about blood pressure targets and identify patients who should receive aspirin and statins. Subsequent calibration studies, however, found that the pooled cohort equations overestimate 10-year risk for incident ASCVD by 60% to 90%, mainly due to lack of generalizability from the highly selected cohorts and use of effective therapies that reduce CVD rates. Read and colleagues found that 5 of the 6 CVD prediction scores overestimated the proportion of patients with ≥10% risk for CVD, and state that “The poor performance of QRISK2 among people with type 2 diabetes could lead to the overtreatment of people at low risk.”
But just how bad is that? “Overtreatment” of CVD risk with lower blood pressure targets and more aggressive use of statins probably prolongs life expectancy. In fact, cholesterol guidelines published in 2013 by 2 prominent American cardiology societies recommend “overtreatment”—that is, starting statin therapy at 40 years of age for patients with diabetes and LDL cholesterol >70 mg/dL (1.8 mmol/L), even if their 10-year risk for ASCVD is very low.
Risk scores should be accurate. For the time being, however, despite a tendency to overestimate risk, the UK's QRISK2 and the US's 10-year ASCVD risk calculator are useful decision-making tools in need of recalibration.