When dealing with complicated patients and complex medical decision-making, it is often easiest to be a “lumper” (vs a "splitter") and classify patients into distinct risk categories in order to simplify the process. The CHA2DS2-VASc score—used to predict risk of stroke and eligibility for anticoagulation in patients with atrial fibrillation (AF)—is a great example of such a practice. The truth, however, is that there can be wide variability among individuals and groups of individuals with the same CHA2DS2-Vasc score in their risk for stroke and the benefit they derive from anticoagulation. In fact, in published observational studies, stroke rates vary significantly across individuals with nonvalvular AF who are not on oral anticoagulation. Eventually, such variability can dramatically affect the net clinical benefit of our interventions (such as oral anticoagulation).
In a recent thought-provoking research article published in the Annals of Internal Medicine, the authors sought to determine the effect of the known variation in published stroke rates on the net clinical benefit of anticoagulation in a large (n=33,434) group of community-dwelling adults with incident AF. The majority of the patients (81.3%) had a CHA2DS2-VASc ≥2.
Depending on the cohort studied, there was wide variability in the benefit of oral anticoagulation. Outcomes were measured in quality-adjusted life-years (QALYs). For example, the population benefit of warfarin in the ATRIA (AnTicoagulation and Risk Factors In Atrial Fibrillation) study was 6290 QALYs vs 24,110 QALYs in the Danish National Patient Registry—a nearly 4-fold variation in the benefit of the intervention. Similarly, the CHA2DS2-VASc threshold for anticoagulation was ≥3 in ATRIA, ≥2 in the Swedish Atrial Fibrillation Cohort study, ≥3 in the SPORTIF (Stroke Prevention using ORal Thrombin Inhibitor in atrial Fibrillation) program, and 0 in the Danish National Patient Registry. These differences were true even when accounting for the better safety profile of the direct oral anticoagulants.
It's likely that these dramatic differences reflect underlying differences in the populations studied. However, the study was truly eye-opening in its demonstration of how the variability in published stroke rates is translated into variability of net clinical benefit with oral anticoagulation.
Clinical guidelines on management of thromboembolic risk seem to suggest a black and white case, whereas results of this analysis suggest the balance between risk and benefit of oral anticoagulation is actually in the gray zone.
Source: Shah SJ, Eckman MH, Aspberg S, Go AS, Singer DE. Effect of variation in published stroke rates on net clinical benefit of anticoagulation for atrial fibrillation. Ann Intern Med. [Epub ahead of print ] doi: 10.7326/M17-2762.