The CV Risk Calculator and the Roles of Statins, Revascularizations, and Underascertainment in the Women's Health Study
There is no doubt that the release of the new Pooled Cohort Equations (PCE), included in the updated 2013 AHA/ACC risk assessment guidelines, sparked significant controversy and disagreement in the field of preventive cardiology. In many respects this controversy caused a rift that has split the field into two opposing camps; those who believe the new PCE risk-score is inaccurate and should be revised ("con"),1,2 or those who believe the PCE is acceptable ("pro," either because they think it is accurate or because they think that the issue of overestimation is not clinically important in the application of guideline recommendations3-5). Over the past year, the opposing sides have released a number of scientific papers that have, perhaps unsurprisingly, backed up their own positions (either "pro" or "con"). It is an immutable fact of life that we all, as physicians and investigators, have our own internal biases (often subconscious).
Unfortunately, the recent scientific papers addressing the issue of the PCE are no exception to this rule. For example, papers in favor of the PCE have often included the guideline authors themselves3 and should be interpreted very cautiously with this in mind.
In this context, along comes another paper by Ridker and Cook, who have been very much in the "con" camp. As expected, the science is rigorous and the results are compelling. This current report is comprehensive and extends the original non-peer reviewed report, published in the Lancet; a report that threatened to overshadow and undermine the new guidelines in their very infancy. The current report is comprehensive in that it finds that overestimation is extremely robust to a number of conservative sensitivity analyses that attempt to account for statin therapy and revascularizations. The fact is, according to these authors, that there would need to be an increase of ~60% in events in order for the PCE to perform well. The "pro" camp will not be happy with these findings. Thus, on the basis of their findings, it appears to be indisputable that the PCE over estimates risk in their cohort. If this was to apply to the population in general, and particularly those in whom the treatment to start statins is unclear, then we should all, as a clinical community, be very concerned.
However, that is the crux of the matter: do these results apply to the general population? In a thoughtful commentary on this paper by Jennifer Robinson, a member of the risk assessment guideline group (see my comments on internal bias above), argues that the results of Ridker and Cook are difficult to generalize. Specifically, these data are derived from Caucasian nurses and health care professionals who a priori would be expected to be at lower risk for atherosclerotic cardiovascular (ASCVD) events than the general population. Indeed, this issue alone may account for the entire underestimation seen in the Women's Health Study. Dr. Robinson makes a great point. However, it is worth mentioning that there are other data from population-based cohort studies (albeit in older Europeans) demonstrating that the PCE also over-estimates in more generalizable samples.6 Thus, niggling uncertainties about the PCE will persist. Accordingly, there is now a pressing need for more calibration and discrimination analyses evaluating the performance of the PCE in modern multi-ethnic studies, such as the Multi-Ethnic Study of Atherosclerosis. These analyses should also try to address the question about whether the "risk" associated with risk factors themselves is changing (e.g., the harmful contents in cigarettes have changed over the decades) or whether our modern health care system (statins/revascularization/aspirin/blood pressure treatment) is impacting the observed event rates seen in modern populations.
So, what do we know? Unfortunately, the answers are complex.
- The PCE risk estimator is a major advance in how we practice preventive cardiology in that it adds stroke as an endpoint of interest and that it allows specific calculation of gender and race (white/black) estimates. Future estimators should incorporate other ethnicities if feasible.
- The PCE may remain under fire as long as there are concerns about risk overestimation with this tool. Studies from generalizable, multi-ethnic population based studies are now needed more than ever.
- The PCE may stand the test of time if the outstanding question of under-ascertainment of events can be answered definitively or if overestimation is found to occur only in high or low risk groups (groups in which there is little uncertainty about statin therapy anyway).3 Similarly, given there is evidence for benefit of statin therapy in persons with as low as a 5% ten-year risk of ASCVD events, the concerns about overestimation may have little real-world clinical significance and may, in the future, be relegated to the realm of abstract academic discourse.
Only time will tell.
For now, the practicing physician would do well to remember that risk scores are merely estimates of the average risk that would be expected from a hypothetical group of individuals with similar traditional risk factors to the actual patient in front of same physician (and that this patient may be an "outlier" in this hypothetical group and have an actual risk far different than the estimated risk produced by the PCE).7 Thus, clinical acumen should take priority, considering both non-traditional risk factors (family history, obesity, pre-diabetes) and adjunctive testing (coronary artery calcium or high-sensitivity C-reactive protein) when the patient or physician has concerns that the PCE may not suffice in order to make the required management decision.
- Ridker PM, Cook NR. Statins: New american guidelines for prevention of cardiovascular disease. Lancet 2013;382:1762-1765.
- Nissen SE. Prevention guidelines: Bad process, bad outcome. JAMA Intern Med 2014;174:1972-1973.
- Muntner P, Colantonio LD, Cushman M, Goff DC, Jr., Howard G, Howard VJ, Kissela B, Levitan EB, Lloyd-Jones DM, Safford MM. Validation of the atherosclerotic cardiovascular disease pooled cohort risk equations. JAMA 2014;311:1406-1415.
- Muntner P, Safford MM, Cushman M, Howard G. Comment on the reports of over-estimation of ascvd risk using the 2013 aha/acc risk equation. Circulation 2014;129:266-267.
- Lloyd-Jones DM, Goff D, Stone NJ. Statins, risk assessment, and the new american prevention guidelines. Lancet 2014;383:600-602.
- Kavousi M, Leening MJ, Nanchen D, Greenland P, Graham IM, Steyerberg EW, Ikram MA, Stricker BH, Hofman A, Franco OH. Comparison of application of the acc/aha guidelines, adult treatment panel iii guidelines, and european society of cardiology guidelines for cardiovascular disease prevention in a european cohort. JAMA 2014;311:1416-1423.
- McEvoy JW, Diamond GA, Detrano RC, Kaul S, Blaha MJ, Blumenthal RS, Jones SR. Risk and the physics of clinical prediction. Am J Cardiol 2014;113:1429-1435.
Keywords: African Americans, African Continental Ancestry Group, Aspirin, Atherosclerosis, Calibration, Cohort Studies, Coronary Vessels, Diabetes Mellitus, Obesity, Risk, Risk Factors, Risk Assessment, Stroke, Uncertainty, Women's Health
< Back to Listings