Abstract
OBJECTIVE: To summarize EUS-FNA test performance in suspected pancreatic malignancy with meta-analysis. METHOD: Two reviewers searched MEDLINE (PubMed and Ovid from January 2002 to January 2012) database to identify relevant studies. The reference lists of the trials were manually searched. Included studies used histopathology or clinical and morphological (CT and MRI and US) follow-up as the "gold standard" and provided sufficient data to construct a diagnostic 2 × 2 table. A statistical program of Meta-Disc was used to calculate the pooled sensitivity, specificity, positive LR, negative LR, DOR, and the SROC curve. Subgroup analysis and meta-regression were calculated to evaluate potential sources of heterogeneity. RESULT: A total of 15 studies with 1860 patients were included for the analysis. The pooled sensitivity and specificity of EUS-FNA were 92 % (95 % CI = 91-93 %, p < 0.001, I (2) = 69.6 %) and 96 % (95 % CI = 93-98 %, p = 0.006, I (2) = 54.9 %), respectively. The positive LR and negative LR were 14.24 (95 % CI = 7.78-26.07) and 0.09 (95 % CI = 0.07-0.13), respectively. The area under the curve was 0.974. The subgroup analysis of six studies with rapid on-site evaluation (ROSE) showed a pooled sensitivity of 95 % (95 % CI = 93-96 %), with p value equal 0.622 and I (2) = 0. The sensitivity analysis of ten high-quality studies (a score of ≥4) showed a pooled sensitivity of 94 % (95 % CI = 93-96 %, p = 0.144, I (2) = 33.1 %), and the pooled specificity was 0.95 (95 % CI, 0.91-0.97). CONCLUSION: EUS-FNA had overall excellent specificity and sensitivity in accurately diagnosing solid pancreatic masses. ROSE could help to improve the accuracy of diagnostic test.