Objectives: Previous studies have shown that, in general, severity scores are predictive of mortality and outcome, but are inconsistently applied. The objective of this study is to investigate the usefulness of AVPU (Alert-responsive to Verbal stimulation-responsive to Pain-Unresponsive) in predicting mortality and Glasgow Outcome Score (GOS) at discharge. Hypothesis: AVPU can satisfactorily predict pediatric TBI outcome at acute care discharge. Participants: The population studied was 768 children and adolescents 0 to 19 years of age who were admitted to or died in a hospital with TBI in Minnesota during 1998. Methods: Cases were identified from the population-based Minnesota TBI Registry and death certificates. Expanding on the 1993 pediatric TBI study, the initial EMS/ambulance AVPU, the initial emergency department (ED) AVPU, AVPU at admission, and outcome measures were abstracted. Logistic regressions were run for AVPU on mortality and dichotomized GOS. Result: For mortality, AVPU-ED had an R² of 0.58 and an odds ratio of 21.7. For both fatal and nonfatal outcomes as measured by GOS, AVPU-ED had an R² of 0.51 and an odds ratio of 4.4. AVPU-ED was missing for only 3% of cases. Conclusions: AVPU is nearly universally obtained and is predictive of outcome. TBI data systems should collect AVPU for pediatric cases. TBI data systems should collect AVPU for pediatric cases.

REFERENCE

1. Reid S, Roesler J, Gaichas A, Tsai A. The epidemiology of pediatric traumatic brain injury in Minnesota. Arch Pediatr Adolesc Med. 2001;155:784-789.