Meta-Analysis of Enteral Nutrition versus Total Parenteral Nutrition in Patients with Severe Acute Pancreatitis

OBJECTIVE: To compare the safety of enteral nutrition and total parenteral nutrition in nutrition support of patients with severe acute pancreatitis. DATA SOURCES: Medline, Embase, and manual search. STUDY SELECTION: 295 articles were screened for randomized controlled studies (RCTs) that compared enteral nutrition with total parenteral nutrition in patients with severe acute pancreatitis. Finally, six RCTs were identified and included in the meta-analysis. DATA EXTRACTION: six RCTs with 224 participants were analyzed. The main outcome were infections, artificial nutrition-related complications, pancreatitis-related complications, non-pancreatitis-related complications, organ failure and mortality. The meta-analysis was performed with the fixed effects model or random effects model. RESULTS: Compared with total parenteral nutrition, enteral nutrition was associated with a significantly lower risk of infections [odds ratio (OR) 0.236; 95% confidence interval (95% CI) 0.120-0.464, p<0.001], pancreatitis-related complications (0.456; 0.234-0.888, p=0.021), organ failure (0.334; 0.167-0.670, p=0.002), multiple organ dysfunction syndrome (0.306; 0.128-0.736, p=0.008), and mortality (0.251; 0.095-0.666, p=0.005). There were no significant differences in artificial nutrition-related complications (0.642; 0.354-1.162, p=0.143), and non-pancreatitis-related complications (0.716; 0.325-1.576, p=0.406) between the two groups. CONCLUSIONS: Enteral nutrition appears safer than total parenteral nutrition in nutrition support of patients with severe acute pancreatitis.