Objective: To assess the epidemiology and outcomes associated with RBC transfusion in patients with severe acute kidney injury requiring continuous renal replacement therapy.
Design: Post hoc analysis of data from a multicenter, randomized, controlled trial.
Setting: Thirty-five ICUs in Australia and New Zealand.
Patients: Cohort of 1,465 patients enrolled in the Randomized Evaluation of Normal versus Augmented Level replacement therapy study.
Interventions: Daily information on morning hemoglobin level and amount of RBC transfused were prospectively collected in the Randomized Evaluation of Normal versus Augmented Level study. We analyzed the epidemiology of such transfusions and their association with clinical outcomes.
Measurements and main results: Overall, 977 patients(66.7%) received a total of 1,192 RBC units. By day 5, 785 of 977 transfused patients (80.4%) had received at least one RBC transfusion. Hemoglobin at randomization was lower in transfused than in nontransfused patients (94 vs 111 g/L; p < 0.001). Mean daily hemoglobin was 88 ± 7 and 99 ± 12 g/L in transfused and nontransfused patients. Among transfused patients, 228 (46.7%) had died by day 90 when compared with 426 (43.6%) of nontransfused patients (p = 0.27). Survivors received on average 316 ± 261 mL of RBC, whereas nonsurvivors received 302 ± 362 mL (p = 0.42). On multivariate Cox regression analysis, RBC transfusion was independently associated with lower 90-day mortality (hazard ratio, 0.55; 95% CI, 0.38-0.79). However, we found no independent association between RBC transfusions and mortality when the analyses were restricted to patients surviving at least 5 days (hazard ratio, 1.29; 95% CI, 0.90-1.85). We found no independent association between RBC transfusion and renal replacement therapy-free days, mechanical ventilator-free days, or length of stay in ICU or hospital.
Conclusions: In patients with severe acute kidney injury treated with continuous renal replacement therapy, we found no association of RBC transfusion with 90-day mortality or other patient-centered outcomes. The optimal hemoglobin threshold for RBC transfusion in such patients needs to be determined in future randomized controlled trials.