Children's Hospital of Philadelphia Philadelphia, Pennsylvania, United States
Abstract: Introduction/
Objective: Excessive red blood cell (RBC) transfusion is associated with prolonged intensive care unit (ICU) length of stay, volume overload, and hospital acquired infections in critically ill patients. Restrictive transfusion thresholds are widely accepted in adult and, increasingly, in pediatric critical care. Evidence-based guidelines informing transfusion practices in the pediatric cardiac ICU are scant and practices are variable.
Methods: Our institution organized a multidisciplinary workgroup aimed at reducing unnecessary RBC transfusions. Transfusion thresholds were developed for stable patients in six physiologic groups (Figure 1) based on published guidelines and workgroup consensus. Patients deemed hemodynamically unstable were excluded. An electronic medical record (EMR) best practice alert (BPA) was developed and implemented in the pediatric ICU and pediatric cardiac ICU. The BPA notified ordering providers when the patient met clinical stability and the ordered transfusion exceeded the recommended transfusion threshold. Patients with active bleeding, prematurity, severe ARDS, hemolytic anemia, ventricular assist device, or ECMO and procedural blood product orders were excluded.
Results: The cardiac ICU BPA alerted ordering providers for 65 transfusions between 5/2 and 7/1; 52 (80%) transfusions had clear clinical indication and 13 (20%) transfusions in 11 patients lacked clear indications. Of these 11 patients, three had single ventricle physiology (transfused for mean Hb of 13 mg/dL, range 12.4-14.1), four had superior cavopulmonary or Fontan physiology (mean Hb of 10 mg/dL, range 9.4-10.9), three were recovering from orthotopic heart transplant (mean Hb 9.4 mg/dL, range 8.4-9.9), one had pulmonary hypertension (Hb 10 mg/dL), and one had repaired congenital heart disease with significant residual disease (Hb 10.2 mg/dL). The most common transfusion indication was “relative anemia”. By comparison, 64 alerts fired over the same period in the pediatric ICU, and of these, only four were deemed inappropriate.
Conclusions: We report on RBC transfusion guidelines and BPA development and implementation to alert providers ordering transfusions for patients meeting stability criteria and exceeding guideline-based thresholds. In the first two months of its use, the BPA alerted providers 65 times and 20% of transfusions that triggered the alert were potentially avoidable. In the pediatric ICU, providers less often deviated from guidelines because restrictive transfusion practices may be more widely established or because of less patient physiologic variability. During this implementation phase, the BPA’s influence on provider behavior and center-wide blood product usage and costs are unknown.
