Performance of platelet count/spleen diameter ratio for diagnosis of esophageal varices in cirrhosis: a meta-analysis

Abstract

Background: Platelet count/spleen diameter ratio (PSR) is a non-invasive method for the assessment of esophageal varices (EV), developed as an alternative to endoscopy.

Aim: To assess the performance of PSR for diagnosis of EV using meta-analysis.

Methods: PubMed, EMBASE, the Cochrane Library, ISI web of Knowledge, China National Knowledge Infrastructure, and article references were searched. We included studies using endoscopy as a reference standard, with the data necessary to calculate the true and false positive, true and false negative diagnostic results of PSR for EV. The quality of the studies was rated with the QUADAS tool. The hierarchical summary receiver operating characteristic (HSROC) was used to examine the PSR accuracy for the diagnosis of EV. Heterogeneity was explored using meta-regression. Clinical utility of PSR for EV was evaluated by a Fagan plot.

Results: In 20 studies (n = 3,063), the HSROC of the PSR for EV was 0.95 at various thresholds. At the threshold of 909, the summary sensitivities and specificities were 0.92 (95% CI, 0.79-0.97) and 0.87 (95% CI, 0.76-0.93), respectively. The HSROC was also 0.95 at the threshold of 909. If PSR was below 909 for EV ("positive" result), the post-test probability (if pre-test probability was 50%) was 87%, while if PSR was at or over 909 ("negative" result), the post-test probability was only 9%. PSR also had a high accuracy in diagnosis of EV in patients with compensated cirrhosis.

Conclusions: PSR can identify EV in cirrhosis with a high accuracy. Application of this index may decrease the need for endoscopy among cirrhotic patients.