Gu 1999.

Study characteristics
Methods	Prospective, randomised study Accepted for publication: 13 July 1998 Number of study centres and location: single centre, Departments of Cardiothoracic Surgery and Anaesthesiology, University Hospital Groningen, Groningen, the Netherlands
Participants	Forty cardiac surgical patients undergoing CPB were allocated randomly to a leucocyte depletion group (n = 20) and a control group (n = 20). Mean age: 65.5 years Sex (female/male ratio): 32.5% No high‐risk patients Inclusion criteria: undergoing CABG, heart valve replacement, or a combined procedure Exclusion criteria: allergy, infection, or preoperative pulmonary dysfunction
Interventions	Intervention group: leucocyte depletion was achieved with 2 filter sets (Pall Duplex filter sets (J1647G; Pall Biomedical, Portsmouth, UK)) located between the venous drainage and the venous reservoir and was driven by a separate roller pump of the heart‐lung machine. Leucocyte filtration was commenced after the start of rewarming but before the release of the aortic cross‐clamp. Control group: no leucocyte depletion filter
Outcomes	Mean Plt and leucocyte removal rate, inflammatory markers (IL‐8, elastase, L‐selectine), and clinical observations (postoperative PaO2 or pulmonary haemodynamics, postoperative intubation time and blood loss, urine output, length of stay in the ICU and in the hospital) Blood samples from participants were taken from the radial arterial catheter before operation, at several points during CPB, at the end of CPB, at the end of operation during skin closure, 1 hour and 3 hours after transfer to the ICU, and at 6 a.m. the next morning in the ICU.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No randomisation method description: "The patients were allocated randomly to a leukocyte depletion group (n = 20) and a control group (n = 20)"
Allocation concealment (selection bias)	Unclear risk	No allocation concealment information
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Not specified; unlikely because of the nature of the intervention
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not specified
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Levels of inflammatory mediators were determined mainly during the postoperative course in exclusively 10 participants in each group, but unclear criteria of selection for the subgroup.
Selective reporting (reporting bias)	Unclear risk	All expected outcomes were reported properly.
Other bias	High risk	Sponsorship provided by Pall Biomedical UK