Petersen 1987.

Methods	Parallel RCT (2:1 ratio) (conducted from February 1981 to March 1984). Multicentre (2 sites). Country USA
Participants	Inclusion criteria: people with a haematological malignancy admitted to the Fred Hutchinson Cancer Research Center or the Swedish Hospital Medical Center for marrow transplantation from HLA‐matched sibling donors Exclusion criteria: people on broad‐spectrum antibiotics and those with a "documented major infection" Total randomised: N = 134 Total analysed: N = 112 Arm 1 (Granulocyte transfusions): randomised = 87, analysed = 67. Acute myeloid leukaemia = 32, Acute lymphocytic leukaemia = 14, Chronic myeloid leukaemia = 19, Myelodysplasia = 2 Arm 2 (Control): randomised = 47 , analysed = 45, Acute myeloid leukaemia = 15, Acute lymphocytic leukaemia = 12, Chronic myeloid leukaemia = 17, Myelodysplasia = 1
Interventions	Comparison between prophylactic broad‐spectrum antibiotics and prophylactic granulocyte transfusions Granulocyte dose: Not reported Granulocyte method of collection: continuous flow centrifugation Donor premedication: not reported Initiation of granulocyte transfusions: Neutrophil count < 0.2 x 109/L Frequency of granulocyte transfusions: Daily Termination of granulocyte transfusions: Neutrophil count "self‐sustaining" > 0.2 x 109/L
Outcomes	Primary Outcome: Major infectious complications occurring in patients treated in conventional rooms Other Outcomes: Mortality/survival, infection (days or episodes)
Definition of infection	Febrile day ‐ Not reported Febrile episode ‐ Fever was defined as an oral temperature >38.3oC. Proven infection: Septicaemia: single positive blood culture associated with signs and symptoms of infection (documented local site and/or fever) or two consecutive positive blood cultures of the same organism Major localised infection: potentially life‐threatening infection requiring systemic antibiotic therapy and/or surgical intervention
Definition of neutropenia	Not reported but the 'trigger' neutrophil count was 0.2 x 109/L
Co‐interventions	Prophylactic antibiotics: broad‐spectrum antibiotics given to the study arm not receiving prophylactic granulocyte transfusions. Therapeutic antibiotics: no restrictions were imposed on the attending physician. Therapeutic granulocyte transfusions: were given if a patient in the prophylactic broad‐spectrum antibiotic arm had a major infectious episode with deterioration of the clinical condition despite appropriate antibiotics and a granulocyte donor was available.
Notes	Funding Sources: PHS Grant Numbers CA 15704, CA 18029, CA 18221, and CA 18579, awarded by the National Cancer Institute, DHHS. Dr. Petersen is the recipient of a grant (12‐5057) from the Danish Medical Research Council. Dr. Thomas is the recipient of a Research Career Award (AI 02425) from the National Institute of Allergy and Infectious Diseases. 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	Unclear risk	It was not reported whether participants were blinded to the intervention. It was not reported whether clinicians or investigators were blinded to the intervention.
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	High risk	87 participants were randomised to receive granulocyte transfusions and 47 to receive prophylactic broad‐spectrum antibiotics. Twenty participants in the granulocyte transfusions group did not receive granulocyte transfusions and were excluded from the study because of problems related to using the proposed granulocyte donor or because the granulocyte donor was found to have antibodies to cytomegalovirus (CMV) while the recipient was negative. Two participants in the prophylactic broad‐spectrum antibiotics group were excluded because of delay of the transplant and subsequent placement into laminar air flow isolation.
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	High risk	Only 42 (63%) participants received granulocyte transfusions according to the study's protocol (as reported in final report of study, initial protocol not available. Granulocyte transfusions were stopped in 25 (37%) participants, 19 permanently and six temporarily. Ten (40%) of the 25 protocol violations were caused by complications in the donor, and in four of these cases the transfusions were resumed after one to three days. Eight (32%) of the protocol violations were due to complications in the patient, five of which were transfusion reactions involving pulmonary symptoms in four. All five transfusion reactions occurred in participants receiving granulocyte transfusions from a parent (three from the mother and two from the father), and none were observed following transfusions from HLA‐matched donors. Seven (28%) protocol violations were caused by the donor being needed for platelet transfusion.