Letters Section Editor: Robert M.
Golub, MD, Senior Editor.
In Reply: Ms Yap and Dr Yap raise an important point: overreliance on the summary results of clinical trials leads to performance measures that can harmfully mismeasure quality.
If clinical trials do not evaluate individual variations in patient risks and benefits, then the performance measures based on the average results of these trials may frequently mandate discretionary and even contraindicated care for many patients.
Although well-designed performance measures may be the strongest mechanism for promoting more efficient and higher-quality care when performed prudently, simplistic, all-or-nothing, “evidence-based”
measures can harmfully prioritize low-value care.1- 3 Unfortunately,
examples of such poorly designed measures are common. Related to one of the illustrations in our Commentary, 2 new NCQA diabetes performance measures mandate tight control for blood pressure and glucose for all patients with diabetes (glycated hemoglobin <7% and blood pressure <130/80 mm Hg),4 although risk-based analyses have shown that the expected benefits of achieving these tight goals are often minimal and that these rigid tight-control measures systematically encourage and reward the use of treatments that are speculative, expensive, burdensome, and of unproven long-term safety when used in low-risk patients or in those who are only modestly above goal on standard therapies.1- 3 More complete risk-based analysis of clinical trials, as advocated in our Commentary,
is a necessary (although insufficient) step toward the design of performance measures that can promote the improvement of high-priority care.
Hayward R, Kent DM. Using Clinical Trial Summary Results to Establish Quality Measures—Reply. JAMA. 2007;298(23):2740-2741. doi:10.1001/jama.298.23.2741-a