Association between injury indices and outcome after severe traumatic brain injury in children

Abstract

Objectives: (1) To determine whether indices of traumatic brain injury (TBI) in children are associated with outcome at hospital discharge and 5 to 7 years later; (2) to describe persisting disabilities. Design: Retrospective, uncontrolled study of a cohort of children with severe, nonpenetrating TBI. Setting: Consecutive admissions to a level 1 trauma center over 2 years. Subjects: Seventy-five children younger than 17 years old were previously studied to identify predictors of disability at acute care discharge. Thirty-three of the 50 survivors (66%) were enrolled. Main Outcome Measures: A database of variables abstracted from medical records was available from the previous study. Subjects were surveyed about premorbid problems, school, employment, and current function, and school records were reviewed. Using all information, a Glasgow Outcome Scale (GOS) score was assigned 5 to 7 years after TBI. Associations between database variables and GOS score at discharge and follow-up were examined using nonparametric analyses. The odds ratio for good recovery was calculated for all significant associations. Results: Late GOS was good recovery for 27%, moderate disability for 55%, and severe disability for 18%. Discharge GOS scores were related ( p ≤ .01) to the head Abbreviated Injury Scale score, Injury Severity Scale score, Glsagow Coma Scale (GCS) score measured in the field and at 6, 24, and 72 hours, the length of coma, and initial discharge site. Late GOS scores were related ( p ≤ .01) to the same variables except the field and 6-hour GCS scores, as well as pupillary responses in the field and the discharge GOS. At follow-up, 64% were independent in mobility, 70% in self-care, and 24% in cognitive items on the Functional Independence Measure. Seventy percent of children received special education services. Employment histories were poor. Most subjects were not receiving neurological or rehabilitation follow-up. Conclusions: Early and late outcome after TBI are related to variables measured at and after injury. Subjects had long-term educational and vocational problems but often did not utilize the medical model of neurorehabilitation.