The accuracy of point-of-care ultrasound to diagnose long bone fractures in the ED

Objectives: To determine the diagnostic accuracy of emergency physician performed point-of care ultrasound (POCUS) for detecting long bone fractures compared to standard radiography. Methods: This was a single-blinded, prospective observational study of patients presenting to two emergency departments (ED) with trauma to long bones. The study used a convenience sample of patients seen during the study investigators' scheduled clinical shifts. Patients presenting to the ED with complaints of long bone trauma were included in the study when a study investigator was available in the ED. POCUS examinations of injured long bones were performed using a standard protocol. The investigators documented their interpretation prior to radiographs being performed. After standard radiographs were performed, the final radiology reading by a radiology attending physician was obtained from the medical record. Results: One-hundred six patients were enrolled into the study, and 147 long bone POCUS examinations were performed. Forty-two fractures were present by radiographs and the prevalence of fractures was 29%. The sensitivity was 90.2% (4/41, 95% CI: 76.9-97.3) and specificity was 96.1% (4/102, 95% CI: 90.3-98.9). The positive likelihood ratio was 23.0 (95% CI: 8.8-60.5), and the negative likelihood ratio was 0.102 (95% CI: 0.040, 0.258). The positive predictive value was 90.2% (4/41, 95% CI: 76.9-97.3) and the negative predictive value was 96.1% (4/102, 95% CI: 90.3-98.9). Conclusions: Emergency physicians can accurately evaluate long bone fractures in the ED using POCUS. In particular, long bone fractures can be excluded with a high degree of confidence.