Abstract

To the Editor Drs Ridker and Wilson 1 proposed that statin guidelines formulate treatment recommendations through consideration of “… patient populations for whom clinical trials have demonstrated benefit.” They failed to provide a clear definition of benefit. The authors cited approximately 20 randomized clinical trials; however, the measures of benefit varied widely between some of these trials. In addition, the vascular events terminology is insufficiently specific. Efficacy measures in randomized clinical trials of statins often include outcomes that differ in their clinical importance, ranging from death to clinician-driven end points such as revascularization procedures and hospitalizations. For example, in the Scandinavian Simvastatin Survival Study (4S), 2 in which participants with a history of coronary heart disease were enrolled, the primary efficacy measure was total mortality. In the Subcutaneous Heparin and Angioplasty Restenosis Prevention (SHARP) trial, 3 in which participants with chronic kidney disease were enrolled, the modified efficacy outcome was a broad composite of events (ie, nonfatal myocardial infarction, coronary death, nonhemorrhagic stroke, or any arterial revascularization procedure). In fact, a reduction in revascularization procedures contributed most to the statistically significant result in SHARP. 4 Even though total mortality was reduced in the 4S trial (relative risk, 0.70 [95% CI, 0.58-0.85]; P < .001), this was not the case in SHARP (rate ratio, 1.02 [95% CI, 0.941.11]; P =. 63). Despite the qualitative differences in the benefit demonstrated in these 2 clinical settings, Ridker and Wilson 1 endorsed statin therapy equally for patients with chronic kidney disease as for patients with a history of coronary heart disease. The authors suggested incorporation of the number needed to treat metric into clinical decision making, but the number needed to treat for which outcome? Guidelines that use oversimplified and ambiguous references to benefit or lack of benefit (without specification of the efficacy measures tested in the relevant trials) are unlikely to be usable by clinicians and patients. In our view, before guideline panels engage in interpretation and application of evidence, they should first specify for guideline users a hierar

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call