Elsevier

Annals of Emergency Medicine

Volume 72, Issue 3, September 2018, Pages 259-269
Annals of Emergency Medicine

Emergency medical services/original research
A Two-Center Validation of “Patient Does Not Follow Commands” and Three Other Simplified Measures to Replace the Glasgow Coma Scale for Field Trauma Triage

https://doi.org/10.1016/j.annemergmed.2018.03.038Get rights and content

Study objective

Out-of-hospital personnel worldwide calculate the 13-point Glasgow Coma Scale (GCS) score as a routine part of field trauma triage. We wish to independently validate a simpler binary assessment to replace the GCS for this task.

Methods

We analyzed trauma center registries from Loma Linda University Health (2003 to 2015) and Denver Health Medical Center (2009 to 2015) to compare the binary assessment “patient does not follow commands” (ie, GCS motor score <6) with GCS score less than or equal to 13 for the prediction of 5 trauma outcomes: emergency intubation, clinically significant brain injury, need for neurosurgical intervention, Injury Severity Score greater than 15, and mortality. As a secondary analysis, we similarly evaluated 3 other measures simpler than the GCS: GCS motor score less than 5, Simplified Motor Score, and the “alert, voice, pain, unresponsive” scale.

Results

In this analysis of 47,973 trauma patients, we found that the binary assessment “patient does not follow commands” was essentially identical to GCS score less than or equal to 13 for the prediction of all 5 trauma outcomes, with slightly superior positive likelihood ratios (eg, those for mortality 2.37 versus 2.13) offsetting slightly inferior negative ones (eg, those for mortality 0.25 versus 0.24) and its graphic depiction of sensitivity versus specificity superimposing the GCS prediction curve. We found similar results for the 3 other simplified measures.

Conclusion

In this 2-center external validation, we confirmed that a simple binary assessment—“patient does not follow commands”—could effectively replace the more complicated GCS for field trauma triage.

Introduction

Out-of-hospital personnel worldwide are taught to calculate the Glasgow Coma Scale (GCS) score as a routine part of their assessment of injured patients. This 13-point (counting from 3 to 15) tool is a core component of the Centers for Disease Control and Prevention National Field Triage Guidelines,1 used to evaluate risk, guide treatment, and help determine the most appropriate hospital for care (eg, trauma center). The ubiquitous acceptance of the GCS presumes that it is an optimal tool for this purpose; however, it has multiple potent limitations.2, 3, 4

Editor’s Capsule Summary

What is already known on this topic

The Glasgow Coma Scale (GCS) is widely used in out-of-hospital trauma care despite little evidence that it is an optimal scale.

What question this study addressed

Can alternative scales or individual items meet or surpass the performance of the GCS?

What this study adds to our knowledge

This 2-site retrospective study of 47,973 trauma patients showed that several simpler scores performed as well as the GCS. In particular, the single item “patient does not follow commands” was identical to GCS score less than or equal to 13.

How this is relevant to clinical practice

This study provides sufficient justification for out-of-hospital systems to experiment with omitting the GCS for simpler measures.

First, the GCS is unreliable. It includes multiple subjective elements and has demonstrated surprisingly low interrater reliability in a wide variety of settings.5, 6, 7, 8, 9, 10, 11, 12, 13, 14 In a study of independent paired assessments by attending emergency physicians, for example, GCS scores were the same in just 38% and were 2 or more points apart in 33%.5

Second, the GCS is widely perceived as complicated2, 3, 13, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38 and takes more than just a few seconds to evaluate. Feldman et al39 found that out-of-hospital providers could calculate the GCS score correctly only 40% of the time, underscoring the inability of this scale to be consistently remembered and applied. Riechers et al37 noted that only 15% of military physicians could correctly calculate the GCS score, despite all of them being familiar with the scale and most having completed the Advanced Trauma Life Support course. Bassi et al13 observed that less than half (48%) of clinicians correctly scored the GCS in a written clinical scenario, with neurosurgeons correct just 56% of the time. Other studies have demonstrated a similarly low level of accuracy.17, 18

Third, summing the 3 different component scales into its 120 different scoring combinations was never intended by the GCS creators,40 and is statistically unsound in that it inappropriately assumes that each gradation of each subscale exhibits an identical magnitude of clinical importance—a presumption that is neither intuitive nor supported by existing evidence.2, 4, 29, 32, 41, 42

Fourth, the GCS is unnecessarily complex in that its subscales and other simpler derivatives have been reported to demonstrate similar predictive capability.19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 38, 43, 44 The GCS predicts mortality well at its extremes and poorly in its midrange,32 and thus most of its predictive capacity is anchored by the endpoints. Accordingly, only some of the GCS elements appear truly predictive, whereas the others are either redundant or simply “noise.” Simplified derivatives that have been previously reported to demonstrate similar accuracy are the binary assessments of motor GCS score less than 6,26, 30 motor GCS score less than 5,43 and the 3-element Simplified Motor Score (motor GCS score of 6, 5, or <5, corresponding to a Simplified Motor Score of 2, 1, or 0, respectively) (Figure 1).19, 20, 21, 36, 38, 44

Despite the above evidence, in 2011 the Centers for Disease Control and Prevention National Expert Panel on Field Triage declined to replace the GCS with a simplified tool, citing “…the lack of confirmatory evidence, the long standing use of GCSt [total GCS] and its familiarity among current EMS [emergency medical services] practitioners, the inclusion of the motor score within the GCSt, and complications because of the difficulty of comparing scoring systems.”1 They advocate the field triage of patients with total GCS scores of less than or equal to 13 to trauma centers of the highest available level.1

Most recently, Kupas et al33 tested motor GCS score less than 6—or “patient does not follow commands”—in a statewide database of 393,877 trauma patients and noted similar performance in the prediction of trauma outcomes. A subsequent systematic review44 noted “slightly greater” areas under receiver operating characteristic (ROC) curves for the total GCS compared with simplified scales, but acknowledged the differences as “likely to be clinically unimportant.” The differences in ROC area were noted to be at the extremes rather than the central curve area, where clinical decisionmaking occurs, and thus do not appear to be relevant or important.4

If a condensed assessment were similarly predictive, the replacement of the GCS with one would save more than a minimal amount of time on every out-of-hospital encounter, would simplify out-of-hospital and other emergency care provider training, and might ultimately improve care of injured patients by streamlining field trauma triage.

We performed an external validation of the motor GCS score less than 6 “patient does not follow commands” assessment by Kupas et al33 while similarly testing of the relative accuracy of 3 other simplified out-of-hospital instruments: motor GCS score less than 5, the Simplified Motor Score, and the “alert, voice, pain, unresponsive” assessment advocated by the Advanced Trauma Life Support course.45

Section snippets

Study Design and Setting

We performed a secondary analysis of prospectively collected and systematically maintained trauma registries from Loma Linda University Health (Loma Linda, CA) and Denver Health Medical Center (Denver, CO), both large urban Level I trauma centers. This study was approved by the institutional review boards of both institutions.

Selection of Participants

We included all adult and pediatric patients in the trauma registries at Loma Linda University Health from January 1, 2003, to December 31, 2015, and from Denver Health

Characteristics of Study Subjects

Demographics and the frequency of the various outcome measures are shown in Table 1. At least one out-of-hospital GCS subscale assessment was missing in approximately one third of entries at both hospitals and, as planned, were multiply imputed according to demographic and other available variables. Our outcomes based on imputation were essentially identical to those obtained through a complete case sensitivity analysis (Tables E1 to E5, Figures E1 to E5, available online at //www.annemergmed.com

Limitations

The principal limitation to this study was that GCS score data were missing in approximately one third of trauma registry entries. Inconsistent GCS score documentation is a common problem with out-of-hospital research, and may be overlooked by ambulance personnel because of a perception of limited contribution, the burden of its calculation, or both. To compensate for such missing data elements, we performed multiple imputation, and our sensitivity analyses found similar comparative results,

Discussion

We report a large, 2-center, external validation of the decision rule of Kupas et al,33 who found in a statewide 393,877-patient trauma registry that a simple, binary motor GCS score less than 6 (ie, “patient does not follow commands”) predicted trauma outcomes as effectively as the existing standard of GCS score less than or equal to 13. Our data corroborate this important finding and confirm that out-of-hospital trauma triage could be simplified by replacing the more complicated GCS with this

References (48)

  • G. Teasdale et al.

    Glasgow Coma Scale: to sum or not to sum

    Lancet

    (1983)
  • F.C.M. Reith et al.

    Differential effects of the Glasgow Coma Scale score and its components: an analysis of 54,069 patients with traumatic brain injury

    Injury

    (2017)
  • W.G. Baxt et al.

    The Trauma Triage Rule: a new, resource-based approach to the prehospital identification of major trauma victims

    Ann Emerg Med

    (1990)
  • R.L. Wears

    Lessons from the Glasgow Coma Scale

    Ann Emerg Med

    (2012)
  • C.L. Horton et al.

    Trauma airway management

    J Emerg Med

    (2014)
  • S.M. Sasser et al.

    Guidelines for field triage of injured patients: 2011 report

    Morb Mortal Wkly Rep

    (2012)
  • M. Zuercher et al.

    The use of Glasgow Coma Scale in injury assessment: a critical review

    Brain Inj

    (2009)
  • M. Gill et al.

    Interrater reliability of 3 simplified neurologic scales applied to adults presenting to the emergency department with altered levels of consciousness

    Ann Emerg Med

    (2007)
  • G. Teasdale et al.

    Observer variability in assessing impaired consciousness and coma

    J Neurol Neurosurg Psychiatry

    (1978)
  • M. Fischer et al.

    Inter-rater reliability of the Full Outline of Unresponsiveness score and the Glasgow Coma Scale in critically ill patients: a prospective observational study

    Crit Care

    (2010)
  • A. Holdgate et al.

    Variability in agreement between physicians and nurses when measuring the Glasgow Coma Scale in the emergency department limits its clinical usefulness

    Emerg Med Australas

    (2006)
  • V.J. Juarez et al.

    Interrater reliability of the Glasgow Coma Scale

    J Neurosci Nurs

    (1995)
  • J. Tesseris et al.

    A comparative study of the Reaction Level Scale (RLS 85) with Glasgow Coma Scale (GCS) and Edinburgh-2 Coma Scale (Modified) (E2CS(M))

    Acta Neurochir (Wien)

    (1991)
  • S. Bassi et al.

    Glasgow Coma Scale: a help or a hindrance?

    Br J Neurosurg

    (1999)
  • Cited by (9)

    • Prehospital quick sequential organ failure assessment score to predict in-hospital mortality among patients with trauma

      2019, American Journal of Emergency Medicine
      Citation Excerpt :

      The proposed cutoff values were nearly identical to those set by the qSOFA score. Furthermore, a recent study suggested that the GCS could be replaced with simpler measurements, such as “patient does not follow commands,” in a prehospital trauma triage setting [16]. Therefore, we do not consider the lack of the prehospital GCS to be a major flaw.

    • Those who speak survive: the value of the verbal component of GCS in trauma

      2023, European Journal of Trauma and Emergency Surgery
    View all citing articles on Scopus

    Please see page 260 for the Editor’s Capsule Summary of this article.

    Supervising editor: David L. Schriger, MD, MPH

    Author contributions: SMG and JSH designed the study. EH and MK prepared and organized the data. EH and JSH analyzed the data. SMG drafted the article. EH, MK, and JSH revised the article critically for important intellectual content. All authors approved the final version and are accountable for all aspects of the work. SMG takes responsibility for the paper as a whole.

    All authors attest to meeting the four ICMJE.org authorship criteria: (1) Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND (2) Drafting the work or revising it critically for important intellectual content; AND (3) Final approval of the version to be published; AND (4) Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

    Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article as per ICMJE conflict of interest guidelines (see www.icmje.org). The authors have stated that no such relationships exist.

    Dr. Schriger was the supervising editor on this article. Dr. Green did not participate in the editorial review or decision to publish this article.

    Readers: click on the link to go directly to a survey in which you can provide feedback to Annals on this particular article.

    A podcast for this article is available at www.annemergmed.com.

    View full text