Allen 2007.

Study characteristics
Methods	Single‐centre, prospective, randomised trial Run‐in period: not specified Number of centres: 1 Location: Ireland
Participants	Thirty‐seven OPCAB surgery patients were randomly allocated to control (retransfusion of unwashed shed blood) and treatment (retransfusion of washed shed blood or discarding of unwashed blood) groups.
Interventions	Intervention: treatment group n = 18 (retransfusion of washed shed blood or discarding of unwashed blood) and control group n = 19 (retransfusion of unwashed shed blood) Control: in the control group, all cardiotomy suction blood was collected in a cardiotomy suction reservoir and returned to the patient after completion of the last graft. In the treatment group, this shed blood was either discarded (if < 500 mL total volume) or processed in a Dideco (Mirandola, Italy) compact cell saver and then returned to the patient.
Outcomes	Circulating blood and urinary samples were obtained. In addition, in the treatment group, blood was withdrawn from the shed mediastinal blood before and after it was processed in the cell saver. Plasma samples were assayed for TNF‐α, IL‐8, and IL‐6, as well as the anti‐inflammatory cytokines, IL‐10, IL‐1ra, and TNF soluble receptor‐2 (TNFsr‐2). Samples were also assayed for serum creatinine. Shed mediastinal blood was assayed for TNF‐α, IL‐6, IL‐8, TNFsr‐2, fHB, and full blood count. Urinary samples were assayed for IL‐1ra, TNFsr‐2, creatinine,α‐1‐microglobulin, albumin, and β‐NAG. All cytokine kits were from RD Systems (Minneapolis, MN) and were sandwich enzyme‐linked immunosorbent assay kits.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were randomly allocated to control (retransfusion of unwashed shed blood) or treatment (retransfusion of washed shed blood or discarding of unwashed blood) using an a priori random number generation software (GraphPad StatMate; San Diego, CA).
Allocation concealment (selection bias)	Unclear risk	No description about allocation concealment, although it states that blood was processed in line with treatment allocation, suggesting that personnel were not blinded to the allocations. This is unlikely to have had a significant impact on the outcomes reported if they were aware of allocations, though.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	No information about blinding is provided. Information provided suggested no blinding occurred.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Insufficient information. No information about blinding is provided. Information provided suggested no blinding occurred.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No missing data are identified, and outcomes for all participants were reported. No participants were withdrawn from analysis.
Selective reporting (reporting bias)	Unclear risk	Expected outcomes from methodology were reported, as were outcomes one would expect from this study design. The study does note that it was not powered to detect clinical differences but that the treatment group had an increased requirement for Plt transfusion. The need for this and the transfusion in itself could introduce significant variation in the outcomes.
Other bias	Low risk	None noted