Stanworth 2020.

Study characteristics
Methods	Design: RCT, 2‐arm parallel, multicentre international feasibility trial Setting: 12 sites (UK, Australia, New Zealand) Recruitment: 2015 to 2017 Maximum follow‐up: 3 months
Participants	38 participants: with diagnosis of MDS based on WHO criteria , ≥ 18 years of age and < 20% blasts on bone marrow aspirate, including non‐proliferative CMML and other MDS/MPN overlap diseases; transfusion dependent, defined as at least 1 red cell transfusion episode per month in the last 8 weeks; and with life expectancy > 6 months Liberal group: n = 18; M/F = 14/4; age = 80 (range 69 to 84) years Restrictive group: n = 20; M/F = 13/7; age = 79 (range 69 to 82) years
Interventions	Liberal strategy: to maintain Hb between 11.0 g/dL and 12.5 g/dL, transfused 2 units of RBCs when Hb ≤ 10.4 g/dL, and 1 unit of RBCs when Hb was between 10.5 g/dL to 110 g/dL (inclusive) Restrictive strategy: to maintain Hb between 8.5 g/dL and 10.0 g/dL, transfused 2 units of RBCs when Hb ≤ 7.9 g/dL, and 1 unit of RBCs when Hb was between 8.0 g/dL to 8.5 g/dL (inclusive)
Outcomes	Primary outcomes of this feasibility study, from day 0 to day 84, were: percentage of pre‐transfusion Hb below the target range of assigned red cell transfusion strategy (‘compliance to treatment threshold’); and achievement of ≥ 2.0 g/dL difference between mean pre‐transfusion Hb in liberal and restrictive strategy groups For the review primary outcome of mortality, 1 patient was reported to have died (restrictive arm), but the timing was not clear
Notes	Trial registration: ISRCTN26088319 Trial funding: "funding was provided by grants awarded by NHSBT R&D, the Australian and New Zealand Society of Blood Transfusion (ANZSBT) and the Wellington Division of the New Zealand Cancer Society" (Stanworth 2020 p 289) "The sponsor and funders had no role in the collection, analysis and interpretation of data, the writing of the report or the decision to submit" (p 283) COI statement by investigators: "there are no conflicts of interest" (Stanworth 2020 p 289)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated randomisation schedule, stratified block design (blocks of 2 and 4)
Allocation concealment (selection bias)	Low risk	Opaque sealed envelopes (prepared from the schedule by an independent member of the clinical trials unit) were opened in sequential order
Blinding of participants and personnel (performance bias) Objective outcomes	Low risk	Blinding of personnel for this intervention is not feasible, but in our view, for objective outcomes such as mortality (the primary outcome used within this review), we graded risk of bias as 'low' Patients were blinded to treatment arm and haemoglobin, but not to transfusions; investigators and clinicians were unblinded to treatment allocation
Blinding of outcome assessment (detection bias) Objective measures	Low risk	Blinding of mortality (the primary outcome used within this review) is not relevant, and we graded risk of bias as 'low'
Blinding of outcome assessment (detection bias) Subjective measures	High risk	Investigators and clinicians were unblinded to treatment allocation, and multiple quality of life measures were reported. Transfusion measures could be altered in response to perceived changes in quality of life measures
Incomplete outcome data (attrition bias) All outcomes	Low risk	All outcomes were reported; all participants were included in analyses
Selective reporting (reporting bias)	Low risk	Trial was prospectively registered, and all data on all outcomes are reported (ISRCTN26088319)
Other bias	Low risk	Some imbalances between arms were noted, but the trial was small. "Two patients were randomised in error as not red cell transfusion dependent" (Stanworth 2020 p 283); four patients were not transfused despite being a study of transfusion‐dependency (p 287)