Interhospital Variability in Perioperative Red Blood Cell Ordering Patterns in United States Pediatric Surgical Patients

Objective To evaluate perioperative red blood cell (RBC) ordering and interhospital variability patterns in pediatric patients undergoing surgical interventions at US children's hospitals. Study design This is a multicenter cross-sectional study of children aged <19 years admitted to 38 pediatric tertiary care hospitals participating in the Pediatric Health Information System in 2009-2014. Only cases performed at all represented hospitals were included in the study, to limit case mix variability. Orders for blood type and crossmatch were included when done on the day before or the day of the surgical procedure. The RBC transfusions included were those given on the day of or the day after surgery. The type and crossmatch-to-transfusion ratio (TCTR) was calculated for each surgical procedure. An adjusted model for interhospital variability was created to account for variation in patient population by age, sex, race/ethnicity, payer type, and presence/number of complex chronic conditions (CCCs) per patient. Results A total of 357 007 surgical interventions were identified across all participating hospitals. Blood type and crossmatch was performed 55 632 times, and 13 736 transfusions were provided, for a TCTR of 4:1. There was an association between increasing age and TCTR (R-2 = 0.43). Patients with multiple CCCs had lower TCTRs, with a stronger relationship (R-2 = 0.77). There was broad variability in adjusted TCTRs among hospitals (range, 2.5-25). Conclusions The average TCTR in US children's hospitals was double that of adult surgical data, and was associated with wide interhospital variability. Age and the presence of CCCs markedly influenced this ratio. Studies to evaluate optimal preoperative RBC ordering and standardization of practices could potentially decrease unnecessary costs and wasted blood.