Serrano 2010.

Study characteristics
Methods	Single‐centre, prospective, randomised controlled trial Run‐in period: unspecified. Author stated 31 months. Registration date: unspecified, study published in 2010 Number of study centres and locations: single centre, authors from University of São Paulo, Brazil
Participants	A total of 81 participants were enrolled in this study. Mean age: 62.8 years Sex (female/male ratio): 0.52 Low‐risk patients (elective CABG) Inclusion criteria: adults with low‐risk stable CAD and preserved LVEF Exclusion criteria: age above 75 years, combined cardiac procedures, cardiac reoperations, left main stenosis of greater than 70%, LVEF lower than 0.50, inflammatory and consumptive diseases, acute infection, colagenoses, renal failure (Cr N 1.5 mg/dL), respiratory distress, emergency surgeries, acute coronary syndromes, use of vasoactive drugs 7 days before surgery date, coagulopathy, stroke, and ICU patients
Interventions	Off‐pump CPB
Outcomes	Complications of cardiac surgery. The aim of this study was to compare specific aspects of inflammation and myocardial injury between on‐pump and off‐pump coronary surgery. Plasma concentrations of CRP, IL‐6 and IL‐8, sP‐selectin, sICAM‐1, CK‐MB, and cTnI were determined for this purpose. In addition, associations between perioperative inflammation and myocardial damage and 1‐year all‐cause mortality were sought. Peripheral venous blood samples were obtained from patients preoperatively and 1 hour, 24 hours, and 5 days after surgery.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Randomization was performed by chance the day before surgery using a computerized randomization system."
Allocation concealment (selection bias)	Unclear risk	Authors do not describe how this was achieved.
Blinding of participants and personnel (performance bias) All outcomes	High risk	The authors state that "All incisions and closure techniques were the same for groups, limiting variability and maintaining blinding of group assignment for patients, family, and referring cardiologists", which would suggest that the participants were blinded. The blinding was not present for the operating theatre personnel, and it is unclear whether the person making the decisions to alter the bypass protocol was involved in the study analysis.
Blinding of outcome assessment (detection bias) All outcomes	High risk	As discussed previously, although the participants were allocated to an intervention group randomly, there was the option to change the allocation during surgery if the conditions did not allow for either on‐ or off‐pump CPB. Although this was appropriate to maintain patient safety, it is not clear whether the operating surgeon then had any involvement in data collection and analysis. Involvement in both the action of breaking the randomisation protocol and the analysis could introduce significant bias and could have been controlled by excluding the operating surgeons from the analysis. The authors do not clarify this point in the method sections.
Incomplete outcome data (attrition bias) All outcomes	High risk	Six participants were withdrawn from the intervention arm and switched to the control arm; while this was done for clinical reasons, this could have significantly impacted the results. Secondly, because of the study duration, only 80 participants completed the full follow‐up, which could have underpowered the study.
Selective reporting (reporting bias)	Low risk	Although no protocol was reported, the authors clearly state which biomarkers they wish to measure in the introduction and method sections, and these are reported in the results section.
Other bias	Low risk	Funding source declared as a nonprofit organisation. No apparent conflict of interests