Skip to main content
. 2023 Feb 7;2023(2):CD014909. doi: 10.1002/14651858.CD014909.pub2

Risk of bias for analysis 1.1 SSI.

Study Bias
Randomisation process Deviations from intended interventions Missing outcome data Measurement of the outcome Selection of the reported results Overall
Authors' judgement Support for judgement Authors' judgement Support for judgement Authors' judgement Support for judgement Authors' judgement Support for judgement Authors' judgement Support for judgement Authors' judgement Support for judgement
Lau 1988 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention Low risk of bias >95% of randomised patients were analysed. Low risk of bias The assessors were not blinded, but it is unlikely that knowledge of the intervention would have influenced the assessment, as SSIs are clearly defined clinically Some concerns No pre‐registered method (registry or protocol) available. Some concerns Some concerns due to the lack of information about a predefined analysis plan.
Takesue 2000 Some concerns There was no information on method of randomisation, but there was no baseline imbalance that would suggest a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention. Low risk of bias 83 % of the 100 randomised patients were analysed. The reasons for the exclusion of 17 patients are given and comprehensible. Low risk of bias The assessors were not blinded, but it is unlikely that knowledge of the intervention would have influenced the assessment, as SSIs are clearly defined clinically. Some concerns No pre‐registered method (registry or protocol) available. Some concerns Some concerns due to the lack of information about the randomisation method and about a predefined analysis plan.
Ishida 2001 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention Low risk of bias >95% of randomised patients were analysed. Low risk of bias The assessors were not blinded, but it is unlikely that knowledge of the intervention would have influenced the assessment, as the diagnosis was made using well‐defined criteria. Some concerns No pre‐registered method (registry or protocol) available. Some concerns Some concerns due to the lack of information about a predefined analysis plan.
Lewis 2002 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention. Low risk of bias All participants were included in the analysis Low risk of bias The asessors were not blinded, but it is unlikely that knowledge of the intervention would have influenced the assessment, as the diagnosis was made using well‐defined criteria. Some concerns No pre‐registered method (registry or protocol) available. Some concerns Some concerns due to the lack of information about a predefined analysis plan.
Espin‐Basany 2005 Some concerns There was no information on method of randomisation, there was no baseline imbalance that would suggest a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention Low risk of bias >95% of randomised patients were analysed. Low risk of bias The assessors were not blinded, but it is unlikely that knowledge of the intervention would have influenced the assessment, SSIs are clearly defined clinically Some concerns No pre‐registered method (registry or protocol) available. Some concerns Some concerns about the lack of information about the randomisation method and about a predefined analysis plan.
Kobayashi 2007 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention Low risk of bias >95% of randomised patients were analysed. Low risk of bias The asessors were not blinded, but it is unlikely that knowledge of the intervention would have influenced the rating, as the diagnosis was made using well‐defined criteria. Some concerns The existence of a study protocol was mentioned in the publication but was not available. Some concerns Some concerns due to the lack of information about a predefined analysis plan.
Horie 2007 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention Low risk of bias All participants were included in the analysis Low risk of bias The assessors were not blinded, but it is unlikely that knowledge of the intervention would have influenced the assessment, as the diagnosis was made using well‐defined criteria. Some concerns No pre‐registered method (registry or protocol) available. Some concerns Some concerns due to the lack of information about a predefined analysis plan.
Oshima 2013 Some concerns There was no information on the method of randomisation, but there was no baseline imbalance that would suggest a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention Low risk of bias 91 % of the randomised patients were analysed. 13 participants were excluded due to protocol violations (6 patients in group A and 7 patients in group B). Low risk of bias The asessors were blinded, but it is unlikely that knowledge of the intervention would have influenced the assessment, as the diagnosis was made using well‐defined criteria. Some concerns The existence of a study protocol was mentioned in the publication but was not available. Some concerns Some concerns due to the lack of information about the randomisation method and about a predefined analysis plan.
Sadahiro 2014 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention. Low risk of bias 95% of the randomised patients were analysed. 11 patients were excluded after randomisation due to the following reasons:

  • stoma creation (n=5)

  • no surgery performed (n=1)

  • intraoperative diagnosis of metastases / inoperability (n=4)

  • no tumour found (n=1)

 Low risk of bias The assessors were blinded to the allocation of patients. Low risk of bias The trial was registered as University Hospital Medical Information Network (UMIN) Clinical Trials Registry 000003435. Low risk of bias Low risk of bias in all domains.
Ikeda 2016 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention Low risk of bias >95% of randomised patients were analysed.
For our meta‐analysis, however, only patients who had MBP and oAB as intervention were eligible. The relevant data was provided to us by the study authors. Low risk of bias The infection control doctors and nurses who determined SSIs were blinded to the assignment. Low risk of bias This trial was registered with the UMIN Clinical Trials Registry (UMIN000019339). Low risk of bias Low risk of bias in all domains.
Hata 2016 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention Low risk of bias >95% of randomised patients were analysed. Low risk of bias The assessors were not blinded, but it is unlikely that knowledge of the intervention could have influenced the assessment, as the CDC criteria for diagnosing SSI were used in both groups. Low risk of bias This trial was registered with ClinicalTrials.gov, Number NCT00508690. Low risk of bias Low risk of bias in all domains.
Anjum 2017 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention. Low risk of bias >95% of randomised patients were analysed. Low risk of bias The assessors were blinded, but even without blinding it is unlikely that knowledge of the intervention would have influenced the rating, as the diagnosis is made using well‐defined criteria Some concerns No pre‐registered method (registry or protocol) available. Some concerns Some concerns due to the lack of information about a predefined analysis plan.
Uchino 2019 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention. Low risk of bias >95% of randomised patients were analysed. Low risk of bias The assessors were blinded to the allocation of patients. Low risk of bias The trail was registered with the University Hospital Medical Information Network Clinical Trials Registry (UMIN‐CTR 000013369). Low risk of bias Low risk of bias in all domains.
Rybakov 2021 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention. Low risk of bias 77 % of the 150 randomised patients were analysed. 34 participants were excluded due to unexpected changes in surgical procedure (e.g. inoperability or no restoration of bowel continuity via anastomosis; 18 in the oral + IV AP group and 16 in the IV AP group). None of the patients were lost to follow‐up. Low risk of bias The assessors were blinded to the allocation of patients. Low risk of bias The study protocol was registered on ClinicalTrials.gov (registrationnumber NCT03436719). Low risk of bias Low risk of bias in all domains.
Papp 2021 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention. Low risk of bias 89 % of the 597 randomised patients were analysed.
The reasons for exclusion were mostly protocol violations (n=71) or that no anastomosis was created during surgery (n= 47). 7 patients were excluded due to adverse events. Low risk of bias The assessors were blinded to the allocation of patients. Low risk of bias This trial was registered as EudraCT 2015‐005614‐27 Low risk of bias Low risk of bias in all domains.
Arezzo 2021 Low risk of bias The generation of the randomisation sequence and the allocation concealment were appropriate. There was no baseline imbalance that would indicate a problem with randomisation. Low risk of bias Both participants and those delivering the intervention were aware of intervention received. There were no deviations from intervention Low risk of bias All participants were included in the analysis. 
For our meta‐analysis, however, only patients who had MBP and oAB as intervention were eligible. The relevant data was provided to us by the study authors. Low risk of bias The assessors were not blinded, but it is unlikely that knowledge of the intervention has influenced the assessment, as the CDC criteria for diagnosing SSI were used in both groups. Low risk of bias The trial was registered with ClinicalTrials.gov (NCT: 04438655). Low risk of bias Low risk of bias in all domains.