Accuracy of emergency department bedside ultrasonography

To determine which focused ultrasound examinations can be interpreted accurately by emergency physicians who have limited training and experience. To determine whether image quality and/or the operator's level of confidence in the findings correlates with accurate scan interpretation.

A prospective sample of consenting adult emergency department patients with the conditions was selected for study. Scans were performed by emergency physicians who had attended a 3-day focused ultrasound examinations instruction course. All scans were videotaped and subsequently reviewed by a radiologist. Accuracy was determined by comparing the emergency physicians scan interpretation with preselected gold standards. Chi-squared tests were employed to determine if the individual performing the scan, the type of scan, patient's body habitus, image quality and/or operator confidence were reliable predictors of accuracy.

Between September 1997 and January 1999, 221 scans were studied. Accuracy varied widely depending on the type of scan performed: aortic scans were 100% accurate whereas renal scans had 68% accuracy. On bivariate analyses, there was little variation in the various operators' levels of proficiency and accuracy of interpretation was not associated with patient body habitus, image quality or operator confidence.

Neophytes can accurately perform and interpret aortic scans; additional training and/or experience appear to be necessary to achieve proficiency in conducting most of the other scans studied. Inexperienced operators are unable to discern whether their scan interpretations will prove accurate.