Gomez‐Villagran 1984.

Methods	Parallel RCT (conducted from January 1981 to June 1982 ). Single centre. Country Spain
Participants	Inclusion criteria: Adults and children with acute myeloid leukaemia (AML) undergoing induction chemotherapy Exclusion criteria: had to have no evidence of infection or fever Total randomised: N = 35 Total analysed: N = 35 Arm 1 (Granulocyte transfusions): N = 19; Acute myeloid leukaemia = 19 Arm 2 (Control): N = 16, Acute myeloid leukaemia = 16
Interventions	Comparison between standard treatment and prophylactic granulocyte transfusions Granulocyte dose: 1.24 x 1010 (0.55 to 4.2) Granulocyte method of collection: intermittent flow centrifugation Donor premedication: dexamethasone Initiation of granulocyte transfusions: Neutrophil count < 0.5 x 109/L Frequency of granulocyte transfusions: Daily Termination of granulocyte transfusions: Until The neutrophil count was greater than 0.5 x 109/L, or 21 days of granulocyte transfusion had been given. There were no donors available or There were "other signs of bone marrow recovery"
Outcomes	Primary Outcome: Not reported Other Outcomes: Mortality/survival, fever (days or episodes), infection (days or episodes), antibiotic use, adverse events
Definition of infection	Febrile day: Not reported Febrile episode: Temperature ≥ 38oC in the absence of a recent blood product transfusion Proven infection: Not reported
Definition of neutropenia	Not reported but the 'trigger' neutrophil count was 0.5 x 109/L
Co‐interventions	Prophylactic antibiotics: oral non‐absorbable drugs neomycin, colimycin and nystatin Therapeutic antibiotics: cephalosporin, tobramycin and carbenicillin as empirical treatment of infection Therapeutic granulocyte transfusions: participants in the control group were not eligible for granulocyte transfusions during the on‐study period (21 days), even if a documented infection developed.
Notes	Funding Sources: Not reported Conflict of Interests: Not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of sequence generation was not reported
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	It was not reported whether participants were blinded to the intervention. "The patients under prophylactic transfusions did not require isolated platelet transfusions during pancytopenic episodes because of the high contaminating level of platelets in the granulocyte concentrate" Therefore medical staff would have been aware of whether patients were in the intervention arm.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	It was not reported whether outcome assessors were blinded to the intervention
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants either died or remission status was assessed after the chemotherapy
Selective reporting (reporting bias)	Unclear risk	The protocol was not available to assess whether any pre‐specified outcomes were not reported or outcomes were reported that were not pre‐specified
Other bias	Unclear risk	The median age of participants was higher in the control group 35 years versus 27.5 years. Three of the participants in the control group had AML M3, and none of the participants in the intervention group had AML M3. Participants were receiving induction chemotherapy and therefore participants with AML M3 would have had a higher risk or early death due to bleeding or disseminated intravascular coagulation. There were also more AML M1 cases in the transfused arm (14/19) when compared with controls (6/16).