Statistical analysis plan for the PlAtelet Transfusion in Cerebral Haemorrhage (PATCH) trial: a multicentre randomised controlled trial

BACKGROUND: Use of antiplatelet therapy shortly before stroke due to spontaneous primary intracerebral haemorrhage (ICH) is associated with higher case fatality in comparison to ICH without prior antithrombotic drug use. The PlAtelet Transfusion in Cerebral Haemorrhage (PATCH) trial aimed to assess the effect of platelet transfusion in patients presenting with ICH while using antiplatelet therapy. The main hypothesis of PATCH was that platelet transfusion would reduce death or dependence by reducing ICH growth. METHODS/DESIGN: PATCH was a multicentre prospective, randomised, open, blinded endpoint (PROBE) parallel group trial, conducted at 60 hospitals in The Netherlands, Scotland and France. Forty-one sites enrolled 190 patients with spontaneous supratentorial ICH aged ≥18 years, who had used antiplatelet therapy for ≥7 days preceding ICH, if Glasgow Coma Scale was ≥8. Participants were randomised (1:1, with a secure web-based system using permuted blocks, stratified by study centre and type of antiplatelet therapy pre-ICH) to receive either platelet transfusion within 6 hours of symptom onset and 90 minutes of diagnostic brain imaging, or standard care without platelet transfusion. The primary outcome was modified Rankin Scale (mRS) score assessed blind to treatment allocation at 3 months after ICH. Planned secondary outcomes included ICH growth on brain imaging performed approximately 24 hours after randomisation, survival at 3 months, disability at 3 months scored using the Amsterdam Medical Centre linear disability score, heterogeneity of treatment effect on mRS and ICH growth according to presence of the computed tomography angiography spot sign, causes of poor outcome, and cost-effectiveness. Safety outcomes were transfusion reactions, thromboembolic complications, and serious adverse events occurring during hospitalisation. This statistical analysis plan was written without knowledge of the unblinded data. TRIAL REGISTRATION: The trial was registered with the Netherlands Trial Register on 29 April 2008 (NTR1303).