Modification of Glasgow Coma Scale criteria for injured elders

Acad Emerg Med. 2011 Oct;18(10):1014-21. doi: 10.1111/j.1553-2712.2011.01164.x. Epub 2011 Sep 26.


Objectives: An abnormal field Glasgow Coma Scale (GCS) score of ≤13 has been used in our emergency medical services (EMS) system to prompt transport to a trauma center. For elders, Ohio has recently adopted a GCS of ≤14 to prompt EMS transport to a trauma center, as older patients respond differently to trauma and may benefit from a different GCS threshold. This study sought to determine if a field GCS of 14 is an appropriate cutoff to initiate transport to a trauma center among injured elders.

Methods: This was a retrospective, observational statewide analysis of injured patients ≥16 years old captured by the Ohio Trauma Registry from 2002 to 2007. Outcomes studied included mortality, traumatic brain injury (TBI), neurosurgical intervention, and endotracheal intubation (ETI). Multiple imputation was performed to account for missing data. Age-stratified sensitivity and specificity for proposed GCS cutoffs of 13 and 14 were calculated. A series of multivariate logistic regression models was then constructed using each outcome as a dependent variable. Independent variables included age, GCS score, sex, blood pressure, injury type, nontrauma center, race, ethnicity, and Injury Severity Score (ISS). Two separate analyses were performed. For each age group, odds ratios (ORs) of each outcome were calculated both for the decrease in GCS from 15 to 14 and for the decrease from 14 to 13. The group of elders with GCS 14 was then compared to adults with GCS 13.

Results: A total of 52,412 study patients were identified. For a GCS cutoff of 13, sensitivity among elders for each outcome was >20% less than sensitivity for adults, and specificity was 5% to 10% greater. Increasing the GCS cutoff for elders to 14 resulted in improved sensitivity for all outcomes (approximately 10%), with a decline in specificity to values near that of adults with GCS 13. In the multivariate models for elders, mortality increased with a decrease in GCS both from 15 to 14 (OR = 1.40, 95% confidence interval [CI] = 1.07 to 1.83) and from GCS 14 to 13 (OR = 2.34, 95% CI = 1.57 to 3.52). In adults, mortality did not increase with the drop from GCS 15 to 14 (OR = 1.22, 95% CI = 0.88 to 1.71) or from GCS 14 to 13 (OR = 1.45, 95% CI = 0.91 to 2.30). When comparing elders with GCS 14 to adults with GCS 13, elders had greater odds of mortality (OR = 4.68, 95% CI = 2.90 to 7.54) and TBI (OR = 1.84, 95% CI = 1.45 to 2.34).

Conclusions: Changing the EMS trauma triage cutoff for elders from GCS 13 to GCS 14 results in improved sensitivity for clinically relevant outcomes. In injured elders, the decline in GCS from 15 to 14 is associated with increased mortality, a finding not observed in younger adults. Elders with GCS 14 have greater odds of mortality and TBI than adults with GCS 13. These results support recent changes in EMS trauma triage guidelines for elders adopted in Ohio.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Aged, 80 and over
  • Female
  • Glasgow Coma Scale*
  • Humans
  • Injury Severity Score
  • Likelihood Functions
  • Logistic Models
  • Male
  • Ohio
  • Registries
  • Retrospective Studies
  • Sensitivity and Specificity
  • Triage / methods*
  • Wounds and Injuries / classification*