A double-blind, randomized study of three antimicrobial regimens in the prevention of infections after elective colorectal surgery.

The objective of this study was to assess the prophylactic efficacy of cefoxitin, ceftizoxime, and metronidazole-gentamicin in colorectal surgery. A double-blind, randomized prospective clinical trial design was used in a Canadian tertiary care teaching hospital. Patients were randomized to one of three treatment groups and received three doses of a study drug (30 min preoperative and 2 postoperative doses at 12 and 24 h). Cefoxitin and ceftizoxime were given as 1000-mg doses. Metronidazole-gentamicin was given as 500 mg of metronidazole plus 120 mg of gentamicin in a minibag. High-risk patients (bowel ischemia, diabetic, current steroid use, etc.) received 10 postoperative doses. Patients with infections, prior antibiotics, or study drug allergies were excluded. Over 30 months, 153 patients were enrolled. Thirty-one patients were excluded for protocol violations. Of the 122 evaluable patients (38 ceftizoxime, 45 metronidazole-gentamicin, 39 cefoxitin), there was no difference across groups regarding sex, age, weight, preoperative Apache II score, and prior history of bowel surgery. Groups were equivalent regarding surgeon, nursing unit, high-risk status (six ceftizoxime, seven metronidazole-gentamicin, seven cefoxitin), bowel preparation, and procedure (including blood loss, drains, organ injury, intraoperative complications). Clinically significant infection requiring systemic antibiotics (7-day hospital and 30-day follow-up) was identified in 0% of ceftizoxime, 15% of metronidazole-gentamicin, and 26% of cefoxitin receiving patients (p = 0.005). Mean ASEPSIS scores for each group were 2.3 (range 0-15) for ceftizoxime, 9.2 (range 0-45) for metronidazole-gentamicin, and 10.4 (range 0-75) for cefoxitin (p = 0.01). Ceftizoxime patients tended to have a shorter total hospital stay (12.2 days versus 19.7 days for cefoxitin versus 13.9 days for metronidazole-gentamicin; p = 0.04), although the procedure to discharge interval was not significantly different (p = 0.09). There was no difference in clinical outcome according to risk status. Anaerobic bacteria were observed more commonly in the ceftizoxime and cefoxitin groups, whereas enteric Gram-negative aerobes were observed most often in the metronidazole-gentamicin group. The study regimens were generally well tolerated. Drug costs were equivalent between ceftizoxime and cefoxitin and lowest with the metronidazole-gentamicin regimen. Ceftizoxime appears to be more effective for the prevention of infection in colorectal surgery than either cefoxitin or metronidazole-gentamicin in the dosage regimens studied.