The objective of clinical prediction rules (decision rules) is to reduce the uncertainty inherent in medical practice by defining how to use clinical findings to make predictions.1 A decade ago, we assessed the quality of clinical prediction rules published from 1981 through 1984 in four major medical journals. Since then the availability of powerful computers and easy-to-use software programs has increased dramatically. This environment has stimulated the production of clinical prediction rules. During the last calendar year, JAMA published more clinical prediction rules than it did during the 4-year period covered by our original review.
In this issue of THE JOURNAL, Stiell et al2 demonstrate a method for targeting use of knee radiographs in acute knee injuries by applying five clinical predictors of knee fractures. When the prevalence of fracture is 6%, the probability of a clinically important fracture is estimated to be less than 0.4%, if these