How does adopting single unit and restrictive transfusion strategies in autologous transplantation and intensive chemotherapy for hematological malignancies affect blood use? – The sustainable study

Abstract

Background: A restrictive red blood cell (RBC) transfusion strategy appears to be safe and effective in reducing RBC transfusions in surgical and critical care patients. Variability in transfusion practice however exists in patients receiving treatment for hematological malignancies. The study center took stepwise changes in the transfusion policy of patients treated with chemotherapy and autologous stem cell transplant. First, double‐unit was replaced by single‐unit RBC transfusion in Feb 2015. A more restrictive hemoglobin (Hb) trigger for transfusion, 7 g/dl in lieu of 8 g/dl, was adopted in Sep 2016. The guidelines for platelet transfusion or the use of growth factors remained the same. Aims: To compare RBC utilization and patients’ outcomes in three one‐year periods retrospectively to study the impact of the new transfusion practice. Methods: Adult patients with acute leukemia, lymphoma or myeloma receiving intensive chemotherapy or autologous transplant during Jan 2014–Dec 2014 (Period A: old practice), Apr 2015–Mar 2016 (Period B: single‐unit transfusion) and Oct 2016–Sep 2017 (Period C: single‐unit & restrictive transfusion) were recruited. Patients were excluded if not meeting the criteria of cytopenia (trough Hb ≤ 8 g/dl, ANc ≤ 0.5 × 109/L & platelet ≤ 20 × 109/L). One treatment cycle lasted from the first day of treatment until sustained recovery from cytopenia (≥3 days) without transfusion support. The primary outcomes were age and sex adjusted incidence rate ratios (IRRs) of RBC units transfused and number of transfusion episodes per day in a treatment cycle in different periods. Other study outcomes were adjusted IRRs of platelet use per day in a cycle and length of hospital stay, and odd ratios of 100‐day mortality in different periods. Results: The study included 121, 118 and 120 cycles of treatment given to 44, 50 and 45 patients in Periods A, B & C respectively. Patients’ demographics were comparable in the three periods. In Periods A, B & C, the median length of treatment cycle [IQR] was 27[20–31], 26[20–34] and 27[21–33] days (P = 0.35 in one way test); total number of RBC transfused was 556, 497 and 332 units, given in 308, 444 and 297 episodes; rates of single unit transfusion were 19%, 89% and 88%; rates of restrictive transfusion (Hb trigger ≤ 7 g/dl) were 26%, 25% and 89% respectively. Comparing with Period A, the number of RBC transfused per day in a treatment cycle reduced by 14% in Period B (IRR 0.86 [95% CI: 0.72–1.02], P = 0.02) and by 37% in Period C (IRR 0.63 [95% CI 0.54–0.73], P < 0.001). The frequency of RBC transfusions per day increased by 38% in Period B (IRR 1.38 [95% CI: 1.15–1.65], P < 0.001) but did not change in Period C (IRR 1.00 [95% CI: 0.86–1.18], P = 0.94). The median number of platelets transfused per cycle was 12[8–24], 16[8–28], 8[8–16] units in Periods A, B & C. Platelets transfused per day did not change in Period B (P = 0.22) but were 20% less in Period C (P = 0.04). The median length of hospital stay was 15[9–23], 17[10–25] and 15[12–21] days; 100‐day mortality rates were 4%, 6% and 6% in Periods A, B & C respectively. No differences were observed in the three study periods for these two parameters. Summary/Conclusions: Single‐unit and restrictive RBC transfusion strategies significantly reduced RBC utilization in patients receiving intensive chemotherapy or autologous transplant. The frequency of RBC transfusions did not increase when both strategies were applied. The new transfusion practice did not increase platelet use, length of hospital stay or mortality.