STOP 2 2005.

Study characteristics
Methods	Multicentre RCT, extension to the STOP trial conducted in 23 centres (including the 12 centres in STOP) in the U.S. and Canada to determine whether regular blood transfusions for the prevention of stroke could be stopped in children and youth 5 ‐ 20 years of age with SCD. Children were monitored by transcranial Doppler examinations after transfusions were halted and by resuming transfusions if the examination indicated a high risk of stroke. This trial was an extension of the previous STOP trial, in which children with abnormal velocities on TCD ultrasonographic examination were administered transfusions to prevent a first stroke. The trial was meant to be a 54‐month trial involving 50 participants in each group, with 60 of the participants enrolled during the first 12 months and 40 during the next 24 months; after recruitment ended, there were 18 months of follow‐up. Four interim analyses and one final analysis were planned for the composite end point. The trial was stopped on the advice of the data safety and monitoring committee because of concern about safety at the fourth interim analysis with 79 participants enrolled.
Participants	Inclusion criteria: children whose Doppler studies normalized after 30 or more months of transfusion were eligible for the present trial. In addition, children who had not participated in the previous STOP trial whose condition met the criteria for eligibility and treatment were also eligible for the present trial. Adequate participation in a transfusion program (≥ 24 transfusions in 30 months and Hb S < 30% in at least 20 of the 30 months); 2 Normal TCD examinations at least 2 weeks apart while receiving transfusions within 4 months of randomisation; age, 5 to 20 years; consent to participate in trial. Exclusion criteria: prior stroke; Indication for chronic transfusion; contraindication for chronic transfusion; moderate‐to‐severe intracranial arterial disease on MRA. Participant flow: screened: not reported; eligible: not reported; randomised: N = 79. Transfusion continued: N = 38 Sex: male: 20 (53%) Age: mean (SD): 12.5 (3.3) HbS % mean (SD): 21.0 (8.6) HbF% mean (SD): 2.4 (1.8) Alpha thalassaemia: not reported TCD velocity: mean (SD): 139 (16) cm/sec Lesions on initial MRI: 10 (26%) Phenotypes:Not reported (trial included only HbSS or HbSβº) Transfusion halted: N = 41 Sex: male: 13 (32%) Age: mean (SD): 12.05 (3.1) HbS% mean (SD): 19.0 (11) HbF% mean (SD): 2.3 (1.5) Alpha thalassaemia: not reported TCD velocity: mean (SD): 143 (18) cm/sec Lesions on initial MRI: 11 (27%) Phenotypes: not reported (trial included only HbSS or HbSβº)
Interventions	Transfusion continued: n = 38. Transfusion could be simple, manual exchange or automated exchange. antigens. Red cell component: leucocyte‐depleted, negative for haemoglobin S. Red cell matching: ABO, Rh and Kell antigens. Iron chelation: chelation therapy with the use of deferoxamine was recommended if serum ferritin levels exceeded 2500 ng per millilitre. Transfusion halted: n = 41. Participants in the transfusion‐halted group could receive transfusions to treat complications of sickle cell disease. Initiation of hydroxyurea therapy or regular transfusion was designated as a cross‐over and data was censored on the patient as of the date of treatment.
Outcomes	Primary outcome: composite end point was a stroke (cerebral infarction or intracranial haemorrhage) or reversion to abnormal velocity on transcranial Doppler ultrasonography, defined as 2 consecutive studies with abnormal velocities, 3 consecutive studies with an average velocity of 200 cm per second or more, or 3 consecutive inadequate studies plus evidence of severe stenosis on MRA. Secondary outcomes: also reports deaths, acute chest syndrome and transfusion adverse events.
Notes	Funding: supported by grants (U01 HL 052193 and U01 HL 052016) from the National Heart, Lung, and Blood Institute. Declarations of interest: no potential conflict of interest relevant to this article was reported. Trial registration: no registration found. Mean length of follow‐up: the median time from randomisation to an end‐point event was 3.2 months (range, 2.1 to 10.1), and the mean (SD) was 4.5 (2.6) months. Power calculation: for a 54‐month trial involving 50 participants in each group, with 60 of the participants enrolled during the first 12 months and 40 during the next 24 months; after recruitment ended, there were 18 months of follow‐up. Analysis: the trial was stopped by the National Heart, Lung, and Blood Institute on the advice of the data safety and monitoring committee because of concern about safety at the fourth interim analysis.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were stratified at randomisation according to the presence or absence of ischaemic lesions on MRI; random, permuted blocks of 4 or 6 participants were used within each group as defined by MRI. Institutional balancing with a tolerance of two participants per site was imposed to maintain an approximate balance in treatment assignments at each site. Eligible participants underwent randomisation with equal probability of continuing or halting transfusion.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants and personnel were unblinded.
Blinding of outcome assessment (detection bias) Stroke and TIA	Low risk	"Suspected strokes were adjudicated by experts unaware of the treatment assignment using clinical data and all available imaging data. Stroke was defined as persistent neurologic abnormalities or transient symptoms accompanied by a new cerebral lesion appropriate to the patient’s clinical presentation." "The Doppler studies were transmitted to central readers who were unaware of the treatment assignments. All results were recorded as the time‐averaged mean of the maximum velocity in the middle cerebral or internal carotid artery and were classified as normal (all mean velocities of <170 cm per second), conditional (at least one mean velocity of 170 to 199 cm per second but none ≥200 cm per second), abnormal (at least one mean velocity of at least 200 cm per second), or inadequate (no information available on one or both middle cerebral arteries)".
Blinding of outcome assessment (detection bias) All outcomes apart from stroke or all‐cause mortality	High risk	Unblinded trial.
Incomplete outcome data (attrition bias) All outcomes	High risk	It was not stated whether intent‐to‐treat analysis was used. Data on 9 participants assigned to no continued transfusion who did not have a primary end‐point event were censored: 5 of these participants resumed chronic transfusion and four started treatment with hydroxyurea. Of 38 participants assigned to continued transfusion, 5 discontinued participation in the trial. N = 14 (17%) of participants discontinued or censored.
Selective reporting (reporting bias)	Unclear risk	Primary outcomes reported. Did not state secondary outcomes. Not clear if all adverse events reported and also censored data not contributing to outcome reporting.
Other bias	Unclear risk	The trial was stopped by the National Heart, Lung, and Blood Institute on the advice of the data safety and monitoring committee because of concern about safety at the fourth interim analysis.