Balancing enrollment and mortality in hemorrhage control trials: A secondary analysis of the PROPPR trial.

Abstract

Designing clinical trials on hemorrhage control requires carefully balancing the need for high enrollment numbers with the need of focusing on the sickest patients. The Pragmatic Randomized Optimal Platelet and Plasma Ratios (PROPPR) trial enrolled patients within 2 hours of arrival to the emergency department for a trial of injured patients at risk for massive transfusion. We conducted a secondary analysis to determine how time-to-randomization affected patient outcomes and the balance between enrollment and mortality. Patients from the Pragmatic Randomized Optimal Platelet and Plasma Ratios trial were compared based on 30-minute time to randomization intervals. Outcomes included 24-hour and 30-day mortality, time to hemostasis, adverse events, and operative procedures. Additional analyses were conducted based on treatment arm allocation, mechanism of injury, and variation in start time (arrival vs. randomization). Randomization within 30 minutes of arrival was associated with higher injury severity (median Injury Severity Score, 29 vs. 26 overall; p < 0.01), lower systolic blood pressure (median, 91 vs. 102 mm Hg overall; p < 0.01), and increased penetrating mechanism (50% vs. 47% overall; p < 0.01). Faster time-to-randomization was associated with increased 24-hour (20% for 0- to 30 minute entry, 9% for 31-minute to 60-minute entry, 10% for 61-minute to 90-minute entry, 0% for 91-minute to 120-minute entry; p < 0.01) and 30-day mortality (p < 0.01). There were no significant associations between time-to-randomization and adverse event occurrence, operative interventions, or time to hemostasis. Increasing time to randomization in this large multicenter randomized trial was associated with increased survival. Fastest randomization (within 0-30 minutes) was associated with highest 24-hour and 30-day mortality, but only 57% of patients were enrolled within this timeframe. Only 3% of patients were enrolled within the last 30-minute window (91-120 minutes), with none of them dying within the first 24 hours. For a more optimal balance between enrollment and mortality, investigators should consider shortening the time to randomization when planning future clinical trials of hemorrhage control interventions. Prognostic and Epidemiologic; Level II.