Five studies were included in the review (n=574). Study quality was excellent: four studies scored 14 out of 14 and two scored 13 out of 14 on the QUADAS assessment.
Prediction of fibrosis (four studies): Prevalence of significant fibrosis was 51% . The summary diagnostic odds ratio was 7.8 (95% CI 5.1 to 11.9; four studies) for prediction of significant fibrosis across all tests. There was no evidence of heterogeneity (p=0.68). Meta-regression found no significant association between any of the variables investigated and estimates of test accuracy. The summary diagnostic odds ratio for the APRI test was 6.8 (95% CI 3.8 to 12.1; three studies). There was no evidence of heterogeneity (p=0.79). At a threshold of 0.5, summary sensitivity and specificity for the APRI were 86% (95% CI 73% to 93%) and 41% (95% CI 34% to 48%). At the higher threshold of 1.5, summary sensitivity was 50% (95% CI 43% to 56%) and specificity was 92% (95% CI 87% to 95%).
Prevalence of cirrhosis: Prevalence of cirrhosis was 16%. Summary diagnostic odds ratio was 11.0 (95% CI 4.6 to 26.2; four studies) for cirrhosis. There was no evidence of statistical heterogeneity (p>0.2).