Investigating the Variability in Mild Traumatic Brain Injury Definitions: A Prospective Cohort Study

Arch Phys Med Rehabil. 2018 Jul;99(7):1360-1369. doi: 10.1016/j.apmr.2017.12.026. Epub 2018 Jan 31.

Abstract

Objective: To prospectively compare the proportion of traumatic brain injuries (TBIs) that would be classified as mild by applying different published definitions of mild TBI to a large prospectively collected dataset, and to examine the variability in the proportions included by various definitions.

Design: Prospective observational study.

Setting: Hospital emergency departments.

Participants: Children (N=11,907) aged 3 to 16 years (mean age, 8.2±3.9y). Of the participants, 3868 (32.5%) were girls, and 7374 (61.9%) of the TBIs were the result of a fall. Median Glasgow Coma Scale score was 15.

Main outcome measures: We applied 17 different definitions of mild TBI, identified through a published systematic review, to children aged 3 to 16 years. Adjustments and clarifications were made to some definitions. The number and percentage identified for each definition is presented.

Results: Adjustments had to be made to the 17 definitions to apply to the dataset: none in 7, minor to substantial in 10. The percentage classified as mild TBI across definitions varied from 7.1% (n=841) to 98.7% (n=11,756) and varied by age group.

Conclusions: When applying the 17 definitions of mild TBI to a large prospective multicenter dataset of TBI, there was wide variability in the number of cases classified. Clinicians and researchers need to be aware of this variability when examining literature concerning children with mild TBI.

Keywords: Brain injuries, traumatic; Child; Rehabilitation.

Publication types

  • Multicenter Study
  • Observational Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Brain Concussion / classification*
  • Child
  • Child, Preschool
  • Datasets as Topic
  • Emergency Service, Hospital / statistics & numerical data*
  • Female
  • Glasgow Coma Scale / statistics & numerical data*
  • Humans
  • Male
  • Prospective Studies