Randomized Comparison of Incision Types

Data analysis will be performed using primary analysis (ITT), followed by secondary analysis (as-treated principles). The ITT set includes all patients who were randomised regardless of whether they received each incision, and the “per-protocol” analysis set includes patients who were treated according to protocol, excluding major protocol violations. Given our expectation, very few patients will crossover between incision types. All baseline and outcome data will be presented using frequencies with proportions for categorical variables and means with standard deviations for continuous data (or medians with interquartile ranges, whichever is more appropriate). The patients will be compared based on their baseline characteristics, including age, sex, body mass index, American Society of Anaesthesiologists classification, preoperative treatment (none, radiotherapy, or chemotherapy), comorbidities, medications, history of previous abdominal surgery, location of the tumour (right or left), and baseline short form-12 health survey questionnaire to determine the balance between the two groups. The primary outcome (the incidence of incisional hernia) will be analysed using the chi-square test. The secondary outcomes will be analysed using the chi-square test for categorical variables (i.e. type of surgery, resection, and anastomosis, incidence of 30-day postoperative complications, histologic type, and depth of the tumour), and Student’s t-test or Mann–Whitney U test for quantitative variables (i.e. operative time, length of hospital stay, blood loss, postoperative pain scores, size of the tumour, and the number of lymph nodes) as appropriate. In accordance with symptoms, each incision type will be evaluated during follow-up after surgery. However, it is not expected that there will be missing data relating to the primary outcomes. For other possible missing data, multiple imputations will be made, based on the assumption that the data are missing at random.
Subgroup analyses will be conducted for each randomisation stratum. The results will be evaluated at a significance threshold of p < 0.05 (two-sided). All of the statistical analyses will be performed using R statistical software, version 3.4.3 (R Foundation for Statistical Computing).