Role of osmolality of contrast media in the development of post-ERCP pancreatitis: a metanalysis.

The role of osmolality of contrast media (CM) in the development of post-ERCP pancreatitis (PEP) is debated. We therefore performed a metanalysis to determine whether osmolality affects the incidence of PEP. A literature search of English-language studies was performed using computerized databases and manual searching of abstracts and article bibliographies. Randomized controlled trials comparing the incidence of PEP associated with high- and low-osmolality contrast media (HOCM, LOCM) were considered. The outcome assessed was clinical pancreatitis as evidenced by both elevation of pancreatic enzymes and pain. Data were analyzed using logistic regression with terms for study and osmolality. Fisher's exact test was done to compare PEP rates. Homogeneity between studies was indicated by the nonsignificance of the study effect in the logistic regression model. Logistic regression also indicated no difference in PEP rates between LOCM and HOCM (P = 0.399). Comparison of PEP rates in both groups using Fisher's exact test did not indicate a difference in any individual study (all P values > 0.05). Due to the large variation of study sample sizes, we repeated the analysis by creating three study groups. The effect of osmolality was invariant to how the data were combined. The results of this metanalysis indicate that there is no significant difference between HOCM and LOCM with respect to clinical PEP.