3120 — Retrospective Case Series Analysis of Process Variability in Red Blood Cell (RBC) Transfusion Practice – Associations with Acuity of Patient Care
Yomtovian R (Cleveland VAMC; Case Western Reserve University School of Medicine), Stockman DL
(Case Western Reserve University School of Medicine), Guadalupe K
(Cleveland VAMC; Case Western University School of Nursing), Fitch T
(Cleveland VAMC), Hoover D
(Cleveland VAMC; Case Western Reserve University School of Medicine ), Aron D
(Cleveland VAMC; Case Western Reserve University School of Medicine)
Although RBC transfusion is the most common in-patient procedure, the Joint Commission has identified it as particularly error prone. Little is known about RBC transfusion process variation and how acuity of patient care impacts this. Our aim was to elucidate RBC transfusion process variation.
Process mapping identified five key steps: 1) transfusion order, 2) patient unit assignment, 3) unit release for transfusion, 4) transfusion initiation, and 5) transfusion completion. We assessed retrospectively timeliness and variability at each step in consecutive transfusion episodes involving patients in two settings: general medical ward (GMW) (n = 50) and medical intensive care unit (MICU) (n = 50). The first RBC transfusion in each episode was studied. Statistical process control methods were used for data analysis. For this study, we assumed higher acuity of illness in the MICU versus GMW.
The mean interval in minutes from order to completion of transfusion was greater for GMW compared to MICU (459 +/- 1 SD 236 and 314 +/- 195, respectively, p = 0.002, t test). The shorter process time for MICU resulted from reduction in duration of the interval from step 3 to 4 (22 +/- 27 versus 14 +/- 16 for GMW and MICU, respectively, p = 0.071), and for the interval from step 4 to 5 (145 +/- 44 versus 115 +/- 48 for GMW and MICU, p = 0.002). Intervals from step 1 to 2 and 2 to 3 were not significantly different. Analysis of SPC control charts revealed a uniformly greater number of special cause variation in GMW versus MICU in all intervals. The most prominent difference in variance was in steps 3 to 4 (p = 0.022, Levene’s Test), with GMW showing significantly greater variance than MICU.
MICU patients receive RBC transfusions significantly faster with less process variation than GMW patients. However, all steps in the process exhibit need for improvement in both timeliness and reduction in variability.
There is high variation in the transfusion therapy process, and improvement is necessary to achieve improved safety and timeliness of care. Analyzing interval means and variability in patients with differing acuity of care will help optimize the development of standardized transfusion guidelines.