We sought to measure the proportion of patients on our clinical service who presented with clinical problems for which research evidence was available to inform estimates of pretest probability. We also aimed to discern whether any of this evidence was of sufficient quality that we would want to use it for clinical decision making.
