Diagnostic accuracy of salivary gland ultrasound in Sjögren's syndrome: A systematic review and meta-analysis.

Background To systematically review the diagnostic accuracy of salivary gland ultrasound in primary Sjogren's syndrome (pSS). Methods PubMed, Embase, CINAHL, Cochrane Central and Scopus and ClinicalTrials.gov were searched to identify diagnostic or validation studies in patients with pSS meeting the diagnostic criteria. A diagnostic test meta-analysis was performed using a bivariate model to calculate the pooled sensitivity, specificity, positive/negative likelihood ratios, and the diagnostic odds ratio. Meta-regression analyses were done for several pSS covariates. Results Sixty-five studies met our criteria for the qualitative review. Fifty-four studies with a total of 6087 patients were included in the meta-analysis. Pooled sensitivity for salivary gland ultrasound was 80% [95% confidence interval (CI): 77-83%; I2 = 78%], and specificity was 90% (95% CI: 87-92%; I2 = 76%). The pooled positive and negative likelihood ratios were 8 (95% CI: 6.4-10) and 0.22 (95% CI: 0.19-0.25), respectively. The corresponding pooled diagnostic odds ratio (DOR) was 37 (95% CI: 28-48). Separate meta-regression models resulted in similar diagnostic estimates: (a) adjusted for mean age: sensitivity 81% (95% CI:77-84%; I2 = 99%) and specificity 90% (95% CI: 87-93%; I2 = 99%); (b) adjusted for mean disease duration, sensitivity 79% (95% CI:72-84%; I2 = 99%), and specificity 90% (89-94%; I2 = 99%). The diagnostic estimates were robust to sensitivity analyses by quality criteria, pSS diagnostic criteria and ultrasound scoring systems. Conclusion Salivary gland ultrasound is a valuable modality for the diagnosis of Sjogren's syndrome. It is plausible that salivary gland ultrasound can be used as an important criterion for the diagnosis of pSS.