Likelihood Ratio, Skull Roentgenograms, and Clinical Decision-Making
1 New England Medical Center, Tufts University School of Medicine, Boston, MA 02111
In Reply.
We appreciate the thoughtful comments of Leventhal and Lembo and concur with their conclusion that the clinician needs to know "the probability of skull fracture in a patient with head trauma." Unfortunately, their proposed "clinical likelihood ratio" (CR) will not further that end because it compares the predictive value (or, more precisely, the posterior probability) of a skull fracture after a positive clinical finding to the posterior probability after a negative finding. After the patient has been examined, the patient does not have both findings; thus, the CR cannot apply to the individual patient.




