Inter-hospital variation in transfusion practices for severe trauma.

Abstract

Background: Ideal blood transfusion practices have evolved over the last decade, with updated recommendations for the plasma:red blood cell (RBC) ratio. A ≥ 1:1 ratio of plasma:RBC has been associated with improved survival. The objective of the current study was to evaluate interhospital variation in plasma:RBC ratio and the associated inpatient mortality.

Methods: All adult patients (≥18 years) with severe injuries undergoing transfusion within 4 hours of admission were identified in the 2020-2021 Trauma Quality Improvement Program database. Transfusion was considered balanced when whole blood or a ≥ 1:1 ratio of plasma:RBC units was administered. Multilevel mixed-effects models were utilized to generate empirical Bayesian estimates of random intercepts for risk-adjusted plasma:RBC ratio at each center, with centers in the highest quartile labeled High-Ratio Centers (HRC). Multivariable logistic regression was constructed to identify factors independently associated with mortality.

Results: Of 35,215 patients receiving care across 424 facilities, 38.0% were admitted to HRC. An estimated 17% of plasma:RBC variation was attributable to hospital effects (intraclass correlation coefficient = 0.17). Following risk-adjustment, HRC (Adjusted Odds Ratio [AOR] 0.81, 95% Confidence Interval [CI] 0.76-0.86) and balanced transfusion (AOR 0.92, 95%CI 0.86-0.98) were associated with reduced odds of mortality. The association of HRC with lower odds of mortality persisted when examining only unbalanced transfusions (n = 28,280, AOR 0.84, 0.78-0.90 95%CI).

Discussion: Care at centers with high plasma:RBC ratios was linked to reduced mortality, even among unbalanced transfusion. Our findings demonstrate the utility of this value as a hospital quality metric.