A Cochrane review (abstract , review [Abstract]) included 7 studies with a total of 404 patients. Antibacterial prophylaxis was associated with a decrease in mortality (8.4% versus controls 14.4%), but this was not statistically significant (RR 0.60, 95% CI 0.34 to 1.05; 7 studies, n=404). Infected pancreatic necrosis was not significantly different between treatment and placebo groups (antibiotics 19.7% versus controls 24.4%; RR 0.85, 95% CI 0.57 to 1.26; 7 studies, n=404). There were less episodes of non-pancreatic infections (antibiotic 23.7% versus control 36%; RR 0.62, 95% CI 0.36 to 1.06; 5 studies, n=318) and all-sites infections (antibiotic 37.5% versus control 51.9%, RR 0.69, 95% CI 0.44 to 1.09; 5 studies, n=318) in antibiotic treated participants but these differences were not statistically significant. Operative treatment (antibiotic 22.6% versus controls 24%; RR 0.90, 95% CI 0.62 to 1.31; 6 studies, n=378) and fungal infections (antibiotic 3.9% versus controls 5%; RR 1.06, 95% CI 0.41 to 2.70; 7 studies, n=404) were not significantly different. Insufficient data were provided concerning antibiotic resistance.
With beta-lactam antibiotic prophylaxis there was less mortality (9.4% treatment, 15% controls), and less infected pancreatic necrosis (16.8% treatment group, 24.2% controls) but this was not statistically significant (for mortality RR 0.72, 95% CI 0.37 to 1.40 and for infected pancreatic necrosis RR 0.69, 95% CI 0.40 to 1.19; 5 studies, n=302). The incidence of non-pancreatic infections was non-significantly different (21% versus 32.5%, RR 0.64, 95% CI 0.28 to 1.47; 4 studies, n=242), as was the incidence of overall infections (34.4% versus 52.8%; RR 0.63; 95% CI 0.35 to 1.13; 4 studies, n=242), and operative treatment rates. No significant differences were seen with quinolone plus imidazole in any of the end points measured. Imipenem on its own showed no difference in the incidence of mortality, but there was a significant reduction in the rate of pancreatic infection (RR 0.34, 95% CI 0.13 to 0.84; 3 studies, n=160).
Comment: The quality of evidence is downgraded by study quality (lack of blinding), by inconsistency (variability in results across studies), and by imprecise results (few patients and wide confidence intervals).
The following decision support rules contain links to this evidence summary:
Primary/Secondary Keywords