LAUSR

OBJECTIVE Despite the presence of highly reliable data, studies on pRBC:FFP ratio suffer from limited sample size and the presence of survivor bias. We sought to study the association between FFP:pRBC and early mortality in the hemorrhaging trauma patient. DESIGN Retrospective nationwide cohort. SETTING All TQIP participating hospitals (2013-2016). PATIENTS We included all trauma patients who were transfused ≥10 pRBCs and ≥1 FFP within 24 hours. We excluded transferred patients and those who died in the emergency department or had missing/inaccurate transfusion data. Patients were assigned to seven FFP:pRBC cohorts (range 1:1 to 1:6, and 1:6+) only if the ratio was similar at 4 and 24 hours and, to avoid survival bias, were excluded otherwise. MAIN OUTCOME MEASURE Multivariable analyses correcting for all available confounders [age, demographics, comorbidities, vital signs, injury severity scale (ISS) and mechanism, procedures performed] were derived to study the independent relationship between FFP:pRBC and 24-hour mortality. RESULTS Out of 1,002,595 patients, 4,427 patients were included. Mean age was 41 years, 79% were males, 61% had blunt trauma, and median ISS was 29. Most patients were transfused in a 1:1, 1:2, or 1:3 ratio (31%, 41%, and 11%, respectively); mortality ranged between 28% for 1:1 and 62% for 1:4. In multivariable analyses, the odds of mortality independently and incrementally increased to 1.23 [1.02-1.48] for a 1:2 ratio, 2.11 [1.42-3.13] for 1:4, and as high as 4.11 [2.31-7.31] for 1:5 (all p<0.05). CONCLUSIONS A 1:1 FFP:pRBC ratio is associated with the lowest mortality in the hemorrhaging trauma patient and mortality increases with decreasing ratios.