The validity and reliability of the navicular fat stripe as a screening test for detection of navicular fractures

Abstract

The purpose of this study was to determine if the navicular fat stripe (NFS) is a valid and reliable screening tool in assessing potential navicular fractures. Retrospective analysis. Minor trauma section of the Los Angeles County-University of Southern California Medical Center emergency department. Study population included patients seen for wrist injury between January 1987 and December 1988. The individuals who interpreted the radiographs included eight observers with varying degrees of radiographic experience. None The positive predictive value averaged 12% or 15% depending on how we counted a straight NFS, and the negative predictive value averaged 93% or 95%, again depending on how we counted a straight NFS. A straight NFS was ten times as likely to occur among normal wrists than among those with navicular fractures. Interobserver reliability was found to be high with a Hoyt reliability coefficient of .79. An abnormal NFS is a poor predictor of a navicular fracture while both a straight and a normal NFS are good predictors of no navicular fracture. A normal NFS may allow less stringent treatment by the emergency physician. Extremes of radiographic experience were found to have no significant effect on NFS interpretation.