Background: Ultrasound-based transient elastography (TE) is a non-invasive alternative to liver biopsy for the staging of hepatic fibrosis due to various chronic liver diseases. This meta-analysis aims to assess the diagnostic accuracy of TE for detecting liver cirrhosis (F4) and severe fibrosis (F3) in patients with chronic liver diseases, in comparison to the gold standard liver biopsy.

Methods: A systematic search was performed using PubMed search engine following Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) guidelines from inception to May 2021. The meta-analysis studies evaluating the diagnostic accuracy of TE for severe fibrosis and cirrhosis were identified. We conducted a meta-meta-analysis to generate pooled estimates of the sensitivity, specificity, and diagnostic odds ratios (ORs) for F3 and F4 fibrosis stage.

Results: We included five studies with a total of 124 sub-studies and 20,341 patients in our analysis. Three studies have reported the diagnostic accuracy of TE in detecting F3/severe fibrosis stage and found 81.9% pooled sensitivity (95% confidence interval (CI): 79.9-83.7%; P < 0.001) (I² = 0%), 84.7% pooled specificity (95% CI: 81.3-87.6%) (I² = 81%; P = 0.02). All five studies reported the diagnostic accuracy of TE in detecting F4/liver cirrhosis stage. We found 84.8% pooled sensitivity (95% CI: 81.4-87.7%) (I² = 86.4%; P < 0.001), 87.5% pooled specificity (95% CI: 85.4-89.3%) (I² = 90%; P < 0.001) and pooled diagnostic OR (41.8; 95% CI: 3.9 - 56.5) (I² = 87%; P < 0.001).

Conclusions: Ultrasound-based TE has excellent diagnostic accuracy for identifying cirrhosis and liver fibrosis stages 3. Future studies should focus on estimating the diagnostic accuracy of other fibrosis stages in chronic liver disease patients. This will eventually decrease the risk associated with invasive liver biopsy.