Preoperative combined mechanical and oral antibiotic bowel preparation for preventing complications in elective colorectal surgery

Abstract

Background: The success of elective colorectal surgery is mainly influenced by the surgical procedure and postoperative complications. The most serious complications include anastomotic leakages and surgical site infections (SSI)s, which can lead to prolonged recovery with impaired long-term health. 
Compared with other abdominal procedures, colorectal resections have an increased risk of adverse events due to the physiological bacterial colonisation of the large bowel. Preoperative bowel preparation is used to remove faeces from the bowel lumen and reduce bacterial colonisation. This bowel preparation can be performed mechanically and/or with oral antibiotics. While mechanical bowel preparation alone is not beneficial, the benefits and harms of combined mechanical and oral antibiotic bowel preparation is still unclear. Objectives: To assess the evidence for the use of combined mechanical and oral antibiotic bowel preparation for preventing complications in elective colorectal surgery. Search methods: We searched MEDLINE, Embase, CENTRAL and trial registries on 15 December 2021. 
In addition, we searched reference lists and contacted colorectal surgery organisations. Selection criteria: We included randomised controlled trials (RCTs) of adult participants undergoing elective colorectal surgery comparing combined mechanical and oral antibiotic bowel preparation (MBP+oAB) with either MBP alone, oAB alone, or no bowel preparation (nBP). We excluded studies in which no perioperative intravenous antibiotic prophylaxis was given. Data collection and analysis: We used standard methodological procedures as recommended by Cochrane. Pooled results were reported as mean difference (MD) or risk ratio (RR) and 95 % confidence intervals (CIs) using the Mantel-Haenszel method. The certainty of the evidence was assessed with GRADE. Main results: We included 21 RCTs analysing 5264 participants who underwent elective colorectal surgery. None of the included studies had a high risk of bias, but two-thirds of the included studies raised some concerns. This was mainly due to the lack of a predefined analysis plan or missing information about the randomisation process. Most included studies investigated both colon and rectal resections due to malignant and benign surgical indications. For MBP as well as oAB, the included studies used different regimens in terms of agent(s), dosage and timing. 
Data for all predefined outcomes could be extracted from the included studies. However, only four studies reported on side effects of bowel preparation, and none recorded the occurrence of adverse effects such as dehydration, electrolyte imbalances or the need to discontinue the intervention due to side effects. Seventeen trials compared MBP+oAB with sole MBP.
The incidence of SSI could be reduced through MBP+oAB by 44% (RR 0.56, 95% CI 0.42 to 0.74; 3917 participants from 16 studies; moderate-certainty evidence) and the risk of anastomotic leakage could be reduced by 40% (RR 0.60, 95% CI 0.36 to 0.99; 2356 participants from 10 studies; moderate-certainty evidence). No difference between the two comparison groups was found with regard to mortality (RR 0.87, 95% CI 0.27 to 2.82; 639 participants from 3 studies; moderate-certainty evidence), the incidence of postoperative ileus (RR 0.89, 95% CI 0.59 to 1.32; 2013 participants from 6 studies, low-certainty of evidence) and length of hospital stay (MD -0.19, 95% CI -1.81 to 1.44; 621 participants from 3 studies; moderate-certainty evidence). Three trials compared MBP+oAB with sole oAB.
No difference was demonstrated between the two treatment alternatives in terms of SSI (RR 0.87, 95% CI 0.34 to 2.21; 960 participants from 3 studies; very low-certainty evidence), anastomotic leakage (RR 0.84, 95% CI 0.21 to 3.45; 960 participants from 3 studies; low-certainty evidence), mortality (RR 1.02, 95% CI 0.30 to 3.50; 709 participants from 2 studies; low-certainty evidence), incidence of postoperative ileus (RR 1.25, 95% CI 0.68 to 2.33; 709 participants from 2 studies; low-certainty evidence) or length of hospital stay (MD 0.1 respectively 0.2, 95% CI -0.68 to 1.08; data from 2 studies; moderate-certainty evidence). One trial (396 participants) compared MBP+oAB versus nBP. The evidence is uncertain about the effect of MBP+oAB on the incidence of SSI as well as mortality (RR 0.63, 95% CI 0.33 to 1.23 respectively RR 0.20, 95% CI 0.01 to 4.22; low-certainty evidence), while no effect on the risk of anastomotic leakages (RR 0.89, 95% CI 0.33 to 2.42; low-certainty evidence), the incidence of postoperative ileus (RR 1.18, 95% CI 0.77 to 1.81; low-certainty evidence) or the length of hospital stay (MD 0.1, 95% CI -0.8 to 1; low-certainty evidence) could be demonstrated. Authors' conclusions: Based on moderate-certainty evidence, our results suggest that MBP+oAB is probably more effective than MBP alone in preventing postoperative complications. In particular, with respect to our primary outcomes, SSI and anastomotic leakage, a lower incidence was demonstrated using MBP+oAB. Whether oAB alone is actually equivalent to MBP+oAB, or leads to a reduction or increase in the risk of postoperative complications, cannot be clarified in light of the low- to very low-certainty evidence. Similarly, it remains unclear whether omitting preoperative bowel preparation leads to an increase in the risk of postoperative complications due to limited evidence. Additional RCTs, particularly on the comparisons of MBP+oAB versus oAB alone or nBP, are needed to assess the impact of oAB alone or nBP compared with MBP+oAB on postoperative complications and to improve confidence in the estimated effect. In addition, RCTs focusing on subgroups (e.g. in relation to type and location of colon resections) or reporting side effects of the intervention are needed to determine the most effective approach of preoperative bowel preparation.