Acute knee trauma: role of ultrasound

The purpose of this study was to determine the diagnostic accuracy of high spatial resolution ultrasonography (US) in the detection of lipohemarthrosis of the knee and to evaluate this sign as a criterion of intra-articular fracture. Forty-eight patients with clinical suspicion of knee fracture were prospectively examined by conventional radiography, sonography examination, and computed tomography (CT) within 48 h after trauma in order to depict direct (fracture line) and indirect (lipohemarthrosis) signs of intra-articular fracture. Lipohemarthrosis was defined as a multilayered collection in the subquadricipital recess. CT was considered as the gold standard for both direct and indirect fracture criteria. CT imaging showed direct signs of intra-articular fracture in 31 patients (65%). Among these patients, 30 (97%) had a lipohemarthrosis. Conventional radiographs showed intra-articular fracture in 26 patients (54%). Among these, 18 (69%) had a lipohemarthrosis. Sonographic examinations could not depict any direct sign of intra-articular fracture but showed a lipohemarthrosis in 29 (93%) of patients with proven fracture via CT. This allowed the depiction of four out of five occult knee fractures. The sensitivity, specificity, positive predictive value, and negative predictive value of sonography for the diagnosis of lipohemarthrosis were 97%, 100%, 100%, and 94%, respectively, as compared with 55%, 100%, 100%, and 55% with conventional radiographs. Using lipohemarthrosis as a criterion of fracture, the sensitivity, specificity, positive predictive value, and negative predictive value of sonography for early detection of intra-articular knee fractures were 94%, 94%, 97%, and 89%, respectively, as compared with 84%, 88%, 93%, and 75% with conventional radiographs. We concluded that, by showing lipohemarthrosis in the subquadricipital recess, high-resolution sonography is a reliable and accurate technique for the evaluation of intra-articular knee fractures.