Among adult patients with liver disease, the ability to identify those most likely to have cirrhosis noninvasively is challenging.
To identify simple clinical indicators that can exclude or detect cirrhosis in adults with known or suspected liver disease.
We searched MEDLINE and EMBASE (1966 to December 2011) and reference lists from retrieved articles, previous reviews, and physical examination textbooks.
We retained 86 studies of adequate quality that evaluated the accuracy of clinical findings for identifying histologically proven cirrhosis.
Two authors independently abstracted data (sensitivity, specificity, and likelihood ratios [LRs]) and assessed methodological quality. Random-effects meta-analyses were used to calculate summary LRs across studies.
Among the 86 studies, 19 533 patients were included in this meta-analysis, among whom 4725 had biopsy-proven cirrhosis (prevalence rate, 24%; 95% CI, 20%-28%). Many physical examination and simple laboratory tests increase the likelihood of cirrhosis, though the presence of ascites (LR, 7.2; 95% CI, 2.9-12), a platelet count <160 × 103/μL (LR, 6.3; 95% CI, 4.3-8.3), spider nevi (LR, 4.3; 95% CI 2.4-6.2), or a combination of simple laboratory tests with the Bonacini cirrhosis discriminant score >7 (LR, 9.4; 95% CI, 2.6-37) are the most frequently studied, reliable, and informative results. For lowering the likelihood of cirrhosis, the most useful findings are a Lok index <0.2 (a score created from the platelet count, serum aspartate aminotransferase and alanine aminotransferase, and prothrombin international normalized ratio; LR, 0.09; 95% CI, 0.03-0.31); a platelet count ≥160 × 103/μL (LR, 0.29; 95% CI, 0.20-0.39); or the absence of hepatomegaly (LR, 0.37; 95% CI, 0.24-0.51). The overall impression of the clinician was not as informative as the individual findings or laboratory combinations.
For identifying cirrhosis, the presence of a variety of clinical findings or abnormalities in a combination of simple laboratory tests that reflect the underlying pathophysiology increase its likelihood. To exclude cirrhosis, combinations of normal laboratory findings are most useful.