Diagnostic Accuracy of Ultrasound for Assessment of Synovial Abnormalities among Patients with Knee Pain: A Meta-Analysis
Ke Liu,Xiaoxiao Li,Qianlin Weng,Yilun Wang,Jie Wei,Chao Zeng,Guanghua Lei,Ting Jiang
DOI: https://doi.org/10.1002/acr.25205
2024-01-01
Arthritis Care & Research
Abstract:Objective Synovial abnormalities, which are modifiable treatment targets for knee pain, affect ~25% of adults. Ultrasound is a safe, inexpensive, and easily accessible imaging modality for assessing synovial abnormalities, but its diagnostic accuracy is still controversial. We conducted a meta‐analysis by comparing ultrasound with the “reference standard” method, ie, magnetic resonance imaging (MRI), in assessing synovial abnormalities among patients with knee pain. Methods PubMed, Embase, and Web of Science were searched from inception to January 7, 2022, to retrieve studies including patients with knee pain for evaluating 1) the diagnostic accuracy of ultrasound versus MRI for synovial abnormalities (synovitis and synovial effusion) and 2) the correlations of synovial abnormalities assessed by ultrasound and MRI. The summary of diagnostic accuracy was analyzed using the bivariate model, and the correlation coefficients were pooled using the random effects model. Results Fourteen studies were included, representing a total of 755 patients. The pooled sensitivity, specificity, and area under the curve were 0.88 (95% confidence interval [95% CI] 0.65–0.96), 0.70 (95% CI 0.51–0.84), and 0.81 (95% CI 0.77–0.84) for synovitis and 0.90 (95% CI 0.81–0.95), 0.86 (95% CI 0.77–0.92), and 0.94 (95% CI 0.91–0.96) for synovial effusion, respectively. Strong correlations between ultrasound‐ and MRI‐diagnosed synovitis (r = 0.64, 95% CI 0.56–0.71) and synovial effusion (r = 0.63, 95% CI 0.52–0.73) were observed. Conclusion Ultrasound demonstrated a promising accuracy in detecting synovial abnormalities among patients with knee pain. The use of ultrasound provides equivalent synovial information to MRI but is less expensive and more accessible. Therefore, it is recommended as an adjuvant for managing patients with knee pain during diagnostic strategy and individualized treatment decision‐making.