Loading...
Like many of you, we find statistical representation of results rarely applicable at the bedside and spare you most of them when summarizing studies. But there are exceptions.
Research studies of diagnostic tests that include a piece of the patient’s history, a physical exam maneuver, a blood test, or an imaging study often report statistical characteristics as sensitivity and specificity, which can be hard to translate when reasoning through a case. Enter likelihood ratios (LRs). While LRs are not always reported in diagnosis research studies, they can be derived from sensitivities and specificities and, in our experience, they are much more useful in clinical reasoning.
LRs help clinicians quickly estimate how the probability of a diagnosis…