Methods
A total of 413 consecutive patients—188 (45.5%) with suspected CD, 156 (37.8%) with UC, and 69 (16.7%) with indeterminate colitis—underwent abdominal surgery at the Massachusetts General Hospital between January 1993 and June 2007. One hundred one (24.5%) had received preoperative IFX ≤ 12 weeks before surgery. These patients were compared to those who did not receive IFX with respect to demographics, comorbidities, presence of preoperative infections, steroid use, and nutritional status. We then compared the cumulative rate of complications for each group, which included deaths, anastomotic leak, infection, thrombotic complications, prolonged ileus/small bowel obstruction, cardiac, and hepatorenal complications. Potential risk factors for infectious complications including preexisting infection, pathological diagnosis, and steroid or IFX exposure were further evaluated using logistic regression analysis.
Results
Patients were similar with respect to gender (IFX = 40.6% men vs. non-IFX = 51.9%, p = 0.06), age (36.1 years vs.37.8, p = 0.43), Charlson Comorbidity Index (5.3 vs. 5.7, p = 0.25), concomitant steroids (75.3% vs. 76.9%, p = 0.79), preoperative albumin level (3.3 vs. 3.2, p = 0.36), and rate of emergent surgery (3.0% vs. 3.5%, p = 1.00). IFX patients had higher rates of CD (56.4% vs. 41.9%, p = 0.02), concomitant azathioprine/6-mercaptopurine use (34.6% vs. 16.6%, p < 0.0001), and lower rates of intra-abdominal abscess (3.9% vs. 11%, p < 0.05). After surgery, the two groups had similar rates of death (2% vs. 0.3% p = 0.09), anastomotic leak (3.0% vs. 2.9%, p = 0.97), cumulative infections (5.97% vs. 10.1%, p = 1), thrombotic complications (3.6% vs. 3.0%, p = 0.06), prolonged ileus/small bowel obstructions (3.9 vs. 2.8, p = 0.59), cardiac complications (1% vs. 0.6%, p = 0.42), and hepatic or renal complications (1.0 vs. 0.6% p = 0.72). A logistic regression model was then created to assess the impact of IFX, as well as other potential risk factors, on the rates of cumulative postoperative infections. We found that steroids (odds ratio [OR] = 1.2, p = 0.74), IFX (OR 2.5, p = 0.14), preoperative diagnosis of CD (OR = 0.7, p = 0.63) or UC (OR = 0.6, p = 0.48), and preoperative infection (OR = 1.2, p = 0.76) did not affect rates of clinically important postoperative infections.
