Objective: To determine the role of antibiotics in patients with acute pancreatitis.
Design: Literature review.
Method: A Medline search for randomised clinical trials (RCTs) during the period 1990-1999 was carried out using the key words 'pancreatitis', 'acute' and 'antibiotics'. Outcome measures were pancreas-related infection, sepsis and death.
Results: Four RCTs were identified. Control groups were not given antibiotics. Based on the pooled data, the differences in risk in favour of the groups treated with various antibiotics were as follows: -14% (95% CI: -26(-)-1; p = 0.04) for pancreas infections, -25% (95% CI: -39(-)-12; p = 0.0002) for sepsis and -13% (95%-BI: -22(-)-3; p = 0.007) for death.
Conclusion: Administration of antibiotics to patients with severe acute (necrotising) pancreatitis reduces the risk of serious complications.