To the Editor.
—It seems that every couple of months an article is published in JAMA demonstrating the predictive capability of a simple clinical maneuver. The study of the 6-minute walk in patients with congestive heart failure by Bittner et al1 is the most recent example. Studies of the ankle/arm blood pressure index2,3 have also appeared recently. The articles tend to conclude that such strongly predictive maneuvers should be incorporated into clinical practice to help determine treatment of patients. I would suggest that clinicians consider three criteria before doing so.First, the amount of risk prediction. Both the relative risks (RRs) and absolute risks should be considered in light of clinical utility. Both measures can be deceiving. For example, an RR of 3, which would be considered important in etiologic epidemiologic studies, may not be clinically usefully if the absolute risks do not differentiate patients who might be
Mark DH. The Predictive Capabilities of Clinical Tests: The 6-Minute Walk. JAMA. 1994;271(9):661-662. doi:10.1001/jama.1994.03510330039025