Coma: the Glasgow Coma Scale improves agreement

Clinical bottom line (level 1a)

  1. Clinical staff who use the Glasgow Coma scale to measure the degree of consciousness in comatose patients are more likely to agree than if they use simple descriptions.
Teasdale et al: Journal of Neurology, Neurosurgery, and Psychiatry 1978; 41: 603-610
Expires October 2003

The study

Setting: 3 acute hospitals, UK, Holland, USA

42 patients in a coma. Observers (doctors in training, general surgeons, neurosurgeons and nurses) graded the patients' degree of unconsciousness using
  • commonly-used terms
  • value judgements for eye opening, motor response, speech
  • Glasgow coma scale


  • The study was divided into three parts
    • 27 observers studying 16 patients
    • 18 observers studying 12 patients over 2 successive days
    • multiple observers viewing a film of 14 patients

The evidence

disagreement rate for 'conscious' v. 'unconscious': 0.19
disagreement rate for 'conscious'/ 'semi-coma'/ 'light'/ 'moderate'/ 'deep': 0.19
disagreement rate for eye opening judgements: 0.14
disagreement rate for speech judgement: 0.054
disagreement rate for motor judgements: 0.11
disagreement rate for GCS motor scale: 0.081
disagreement rate for GCS verbal response: 0.091
disagreement rate for GCS eye opening: 0.089
disagreement rate for overall GCS: 0.074
  • Disagreement rates for the Glasgow Coma Scale were not found to change over 2 successive days.

Comments

  1. Disagreement rate: (the sum of number of observations disagreeing from the consensus times the difference from the consensus) / (number of observations times the larger of (consensus minus 1, or top value minus consensus)) It has a range of 0 (best) to 0.5 (worst)
  2. This study is a validation of a clinical prediction rule.

Citation

  1. Teasdale G, Knill-Jones R, van der Sande J: observer variability in assessing impaired consciousness and coma. Journal of Neurology, Neurosurgery, and Psychiatry 1978; 41: 603-610
Contributor: Chris Ball and Clare Wotton, October 1999
Reviewer: Malcolm Daniel

Clinical Question.
Patient
Intervention or Exposure
Outcome