The ability of statewide prehospital pediatric blood transfusion protocols to predict early in-hospital blood product administration: A National Trauma Data Bank analysis.

backgroundThe ability of statewide prehospital pediatric blood transfusion protocols to identify patients who receive early in-hospital blood transfusion is unknown. We aimed to characterize these protocols and compare their test characteristics.

methodsThe National Trauma Data Bank was used for this study. Pediatric (1-15 years of age) patients enrolled in the National Trauma Data Bank from 2017 to 2020 and transported by emergency medical services were analyzed. Components of available statewide transfusion protocols were abstracted by a single investigator using a standardized template. Test characteristics and associated 95% confidence intervals were calculated for each protocol using early in-hospital transfusion as the reference standard. We defined early transfusion as blood product administration within 4 hours of hospital arrival. Prehospital heart rate, systolic blood pressure, and Glasgow Coma Scale values were available in the data set. Shock index was calculated as heart rate/systolic blood pressure. Altered mental status (AMS) was defined as a Glasgow Coma Scale score of <15. Head injury was defined as an Abbreviated Injury Scale head score of >1. Intoxication was defined as a blood ethanol level of >80 mg/dL.

resultsOf the 78,430 patients analyzed, 2,125 (2.7%) received early transfusion. Four statewide prehospital transfusion protocols were included (Maryland [age-defined hypotension plus age-defined tachycardia or AMS without head injury or intoxication], West Virginia [at least two: systolic blood pressure < 70+ (2(age)); AMS without head injury; heart rate >130 beats per minute; shock index >1], Delaware [systolic blood pressure <70 mm Hg or elevated age-adjusted pediatric shock index], and Pennsylvania [age-defined hypotension or shock index >1 or AMS without head injury]). Test characteristics varied widely across protocols (Maryland: sensitivity of 8.0% [6.7-9.4%], specificity of 99.7% [99.6-99.7%]; West Virginia: sensitivity of 44.0% [41.1-46.9%], specificity of 86.9% [86.6-87.2%]; Delaware: sensitivity of 53.6% [51.2-55.9%], specificity of 81.2% [80.9-81.5%]; Pennsylvania: sensitivity of 58.7% [56.2-61.1%], specificity of 72.3% [72.0-72.7%]).

conclusionFew statewide prehospital pediatric transfusion protocols exist. Existing protocols have variable inclusion criteria with suboptimal and wide-ranging sensitivity and specificity for the outcome of early in-hospital transfusion. LEVEL OF EVIDENCE: Prognostic and Epidemiological; Level III.