Introduction Transient elastography is a non-invasive method developed as an alternative to liver biopsy to assess fibrosis severity. In this meta-analysis, we assessed the performance of elastography in diagnosing fibrosis using liver biopsy as the reference standard.

Methods Electronic search of MEDLINE, EMBASE, Science citation index and Cochrane Library; and hand-search of major conference abstracts and article references were performed by two authors independently. Included studies used biopsy as reference standard, reported on data necessary to calculate the true-positive, false-positive, true-negative, and false-negative diagnostic results of elastography for a fibrosis stage and had a 3-month maximum interval between tests. METAVIR F2 and cirrhosis were analysed. Study quality was assessed using the QUADAS tool. Data were combined using the hierarchical summary receiver operator characteristics, and bivariate normal random-effects analysis of sensitivity and specificity methods, using the METANDI module.

Results We identified 40 eligible studies (32 full papers). No article was assessed as adequate for all the questions in QUADAS. Only nine studies had both an acceptable reference and index test quality. Overall, the summary sensitivity and specificity was 0.79 (95% CI 0.74 to 0.82) and 0.78 (95% CI 0.72 to 0.83) for F2 and 0.83 (95% CI 0.79 to 0.86) and 0.89 (95% CI 0.87 to 0.91) for cirrhosis. With an elastography result at or over the threshold value as determined in each study, for F2 or cirrhosis (“positive” result), the corresponding post-test probability for presence of these stages (if pre-test probability was 50%) was 78%, and 88%, respectively, while values falling below those thresholds (“negative” result), the post-test probability was 21% and 16%, respectively. No optimal stiffness cut-offs for specific fibrosis stages were validated in independent cohorts. There was a wide range and overlap of cut-offs within and between different fibrosis stages.

Conclusion Elastography is a good screening test for cirrhosis, with an 88% disease probability following a “positive” measurement, and a slightly worse tool in F=2. However, a “negative” measurement is less accurate, with disease being present in 15–20% of cases depending on fibrosis stage. Validation of liver stiffness cut-offs is required before elastography can be considered as sufficiently accurate for non-invasive diagnosis of fibrosis stages.