Table 2.

Meta-analyses investigating the role of MBP and OABs in elective colorectal surgery

Author (year)	Sample size	Design	Groups	Outcomes	Results	Quality assessment/risk of bias
Rollins (2019) [8]	Total: 69,517 patients; 28 RCTs with 6437 patients; 12 cohort studies with 63,080 patients	Meta-analysis of RCTs and non-randomized studies	1. MBP + OABs vs MBP 2. MBP + OABs vs OABs 3. MBP + OABs vs no preparation 4. OABs alone versus no preparation 5. OABs vs MBP	SSI, anastomotic leak, 30-day mortality, overall morbidity, development of ileus, reoperation and Cdiff infection	1.—SSIs reduced in the combined and in the RCT analysis - AL, mortality, morbidity, ileus, reduced in the combined but NS in the RCT analysis - Cdiff NS 2.—30 days mortality, ileus reduced in the combined but NS in the RCT analysis - other outcomes NS in the combined and RCT analysis 3. RCTs unavailable 4. RCTs unavailable 5. RCTs unavailable	Variable risk of bias in the RCTs, poor documentation on randomization methods, allocation concealment, and blinding Heterogeneity due to the different antibiotic regimens
Toh (2018) [9]	38 RCTs with 8458 patients	Network meta-analysis of RCTs	1. MBP + OABs vs MBP 2. MBP + OABs vs OABs 3. MBP + OABs vs no preparation (indirect comparison) 4. OABs alone versus no preparation (indirect comparison) 5. MBP vs no preparation 6. OAB vs MBP	Primary: total, incisional, and organ/space SSI Secondary: anastomotic leak, mortality, readmissions/reoperations, UTI, pulmonary complications	Primary: 1. Reduced total, incisional and organ/space SSIs 2. NS 3. Reduced total SSIs 4. Reduced organ/space SSIs 5. Reduced organ/space SSIs 6. SSIs NS Secondary: NS for every outcome and every comparison	Variable risk of bias, heterogeneity +  +  + for the different resection sites and antibiotic treatments, no RCTs in every category led to the use of indirect comparisons