Antibiotic Prophylaxis Choices and SSI Risk in Colorectal Surgery
Key Takeaways
- Broad-spectrum penicillins showed the most consistent association with lower surgical site infection risk, and third-generation cephalosporins and metronidazole-based cephalosporin or aminoglycoside combinations were also associated with lower risk.
- Reduced 30-day mortality was reported with broad-spectrum penicillins and with fluoroquinolone-plus-penicillin regimens, although the latter signal was low certainty.
- No significant differences were observed across regimens for adverse events or length of stay, and certainty across comparisons ranged from very low to moderate.
The analysis included 105 randomized clinical trials with 18,273 patients and compared 32 antibiotic nodes defined by class or class combinations. It evaluated prophylaxis administered within 24 hours before surgery rather than route alone. Thirty-day mortality, adverse events, and length of hospital stay were also assessed across the network.
A systematic review and frequentist random-effects network meta-analysis published in JAMA Network Open included randomized clinical trials of adults undergoing elective colorectal procedures and required reporting of 30-day surgical site infection as the primary outcome. Secondary outcomes were 30-day mortality, adverse events, and length of hospital stay, and oral antibiotic regimens started more than 24 hours before surgery were excluded. Searches covered MEDLINE, Embase, Cochrane CENTRAL, CINAHL, and Scopus through July 17, 2025, supporting class-level comparisons rather than route-only comparisons.
Beyond the lead finding, third-generation cephalosporins were associated with lower SSI risk versus placebo or no prophylaxis, RR 0.27 (95% CI 0.16-0.45). Metronidazole plus second-generation cephalosporins showed a similar association, RR 0.27 (95% CI 0.17-0.44), and tetracyclines were also associated with lower risk, RR 0.32 (95% CI 0.20-0.53). Metronidazole plus aminoglycosides was also associated with lower SSI risk, RR 0.42 (95% CI 0.25-0.70).
For mortality, broad-spectrum penicillins were associated with lower 30-day risk, RR 0.21 (95% CI 0.05-0.90), and fluoroquinolones plus penicillins showed RR 0.14 (95% CI 0.03-0.79). Investigators rated the fluoroquinolone combination as low certainty and identified broad-spectrum penicillins and cephalosporin-based combinations as the strongest SSI signals.
No significant differences were observed among regimens for adverse events or length of hospital stay overall. Adverse events included diarrhea, nausea or vomiting, urticaria, rash, candidiasis, hypotension, phlebitis, and erythema. One comparison favored metronidazole plus broad-spectrum penicillins over metronidazole plus third-generation cephalosporin for adverse events, RR 0.07 (95% CI 0.01-0.81). Inconsistent reporting and nonstandardized adverse-event definitions limited interpretation, and no regimen was associated with shorter hospitalization than placebo or other active regimens.
Certainty across the network ranged from very low to moderate, and many comparisons were informed by few or older trials. Most trials were at high risk of bias in at least one domain, with concerns often involving allocation concealment or blinding. Most prespecified subgroup analyses could not be undertaken because event counts were low and data were sparse. The feasible subgroup analysis by overall risk of bias did not show evidence of a subgroup effect. Baseline comorbidity, cancer status, bowel-preparation reporting, and outcome definitions varied across trials, while class selection appeared more informative than route alone in this elective colorectal surgery synthesis.