Kaplan-Meier plots of time from platelet transfusion to first ≥ grade 2 bleeding for each platelet characteristic. (A) Time from first platelet transfusion to first grade 2 or higher bleeding by platelet dose group. Time to bleeding was censored at the first date that any of the following occurred: missing data on whether ≥ grade 2 bleeding occurred; or end of study. Divergence of the curves after 24 days is probably the result of the small number of patients still at risk by that time. Platelet dose group was not a significant predictor of time to ≥ grade 2 bleeding (P = .77). (B) Time from first platelet transfusion to first grade 2 or higher bleeding by source of platelets. Time to bleeding was censored at the first date that any of the following occurred: transfusion of a platelet unit with a different source from the patient's initial platelet transfusion; missing data on whether ≥ grade 2 bleeding occurred; or end of study. Platelet source was not a significant predictor of time to ≥ grade 2 bleeding (P = .44). (C) Time from first platelet transfusion to first grade 2 or higher bleeding by ABO matching status. Time to bleeding was censored at the first date that any of the following occurred: transfusion of a platelet dose with a different ABO matching status from the patient's initial platelet transfusion or missing data on ABO matching status; missing data on whether ≥ grade 2 bleeding occurred; or end of study. Divergence of the curves after 15 days is probably the result of the small number of patients still at risk by that time. ABO matching status was not a significant predictor of time to ≥ grade 2 bleeding (P = .33). (D) Time from first platelet transfusion to first grade 2 or higher bleeding, by platelet storage duration, among patients whose first platelet transfusion was stored for 0 to 5 days. Time to bleeding was censored at the first date that any of the following occurred: transfusion of a platelet unit with a different storage duration from the patient's initial platelet transfusion; transfusion of a platelet unit with missing data on storage duration; transfusion of pooled WBP platelets of different storage durations; missing data on whether ≥ grade 2 bleeding occurred; or end of study. Divergence of the curves after 10 days is probably the result of the small number of patients still at risk by that time. Duration of platelet storage was not a significant predictor of time to ≥ grade 2 bleeding (P = .87).