Mechanical bowel preparation versus no bowel preparation for oesophageal cancer surgery: a multicentre, randomized,...

To compare post-operation influence between mechanical bowel preparation (MBP) and no MBP (NBP) before surgery for patients with oesophageal cancer.

In this multicentre, parallel, open-labelled trial, eligible patients who were prepared to undergo oesophagectomy in 3 hospitals in China were randomly assigned (1:1) to an MBP or NBP group. Patients allocated to the MBP group prepared their bowel by drinking 3-4 l of polyethylene glycol dissolved in water the day before surgery. Patients in the NBP group did not receive MBP. The primary outcome was the incidence of major complications during or within 30 days after surgery.

Between 20 September 2022 and 30 April 2023, 652 patients were recruited (326 patients were randomized to MBP and 326 patients were randomized to NBP), with 612 patients included in the final analyses (311 patients for MBP and 301 patients for NBP). Risk difference (RD) was -2.6% between the 2 groups (29.9% in NBP vs 32.5% in MBP), and the relative risk (RR) was 0.92 [97.5% confidence interval (CI) 0.727-1.165]. The upper limit of the 1-sided 97.5% CI for RD was 4.73%, and the upper limit for RR was 1.165, both lower than the predefined non-inferiority margins (Δ = 10% for RD; threshold = 1.308 for RR). No significant differences of CCI were observed; CCI was similar between the 2 groups (15.6 in MBP vs 15.1 in NBP). Ten items in EORTC QLQ-C30 were better in the MBP group than in the NBP group in the 7 days after the operation (P < 0.05). No statistical differences were found in the EORTC OES-18 results between the 2 groups.

This study confirms that in preoperative preparation for oesophageal cancer surgery, the non-inferiority of NBP is established both by the RD and the RR. These results collectively support the safety of omitting MBP and provide robust evidence for simplifying preoperative protocols.