• The Glasgow Coma Scale (GCS) is the best available measure of the impact of disease or injury on neurological function.
• The GCS score is such an important determinant of patient outcome that it is used in most contemporary severity scoring systems.
• When combined with abnormal brainstem responses, the GCS score identifies individuals at high risk of death or a vegetative state.
The Glasgow Coma Scale (GCS) was described in 1974 as a practical method for assessing coma. Impaired consciousness was classified according to three dimensions of responsiveness: best ocular, best motor, and best verbal responses. Over the past 20 years, the GCS has become a universal language for reliably describing patients with impaired consciousness in a reproducible way. In addition, a numerical score which quantifies the extent of abnormality for the patient's best ocular, motor, and verbal responses permits calculation of a 13-point GCS score ranging from 3 to 15 ( Table 1). In providing a summary measure of neurological function, the GCS score assumes that the patient is normotensive, normoxic, and has received no paralytic, narcotic, or other medications that artificially depress neurological status.
new Optn ftfH»lifw&mJv 4
IMfYHHd IKfNfl t
fdHAfkilt Jttnuiul Htnim 1
imiMi DunHHdnuuiMiuk «
Table 1 The GCS score
Because it can be used to describe impaired consciousness in a wide variety of medical and surgical diseases, the GCS score is probably the most widely used and best known severity scoring system. The ocular, motor, and verbal responses included in the GCS and the associated score have been used independently or in combination with other neurological findings to reflect the severity of neurological damage for patients with head injury, cardiac arrest, intracerebral hemorrhage, cerebral infarction, sepsis, and other causes of non-traumatic coma. The GCS has also been included as part of most contemporary general severity scoring systems including the Mortality Probability Model (MPM II), the Simplified Acute Physiology Score (SAPS II), the Pediatric Risk of Mortality (PRISM), and Acute Physiology and Chronic Health Evaluation II and III (APACHE II and III).
The GCS has been used for group risk stratification in clinical outcome studies, to evaluate intensive care resource use, and to evaluate prognosis for individual patients. In clinical research, the GCS score is the most frequently used measure of severity of head injury. Data from the United States National Traumatic Coma Bank demonstrated a clear association between a decrease in mortality and an increase in the GCS score. For patients in this study with GCS scores below 8, three treatable conditions (hypoxia, shock, and increased intracranial pressure) were found to be significantly correlated with death or severe disability. This information has been used to develop evidence-based guidelines for the management of severe head injury ( Chesnut 1995).
The GCS score has also been used to generate computer-based predictions of outcome for individuals with severe head injury and to measure the impact of these estimates on patient management (Murrayeia/ 1993). There was increased use of intensive monitoring and treatment for patients predicted to have a less than 0.4
probability of a poor outcome (dead or vegetative), and a 39 per cent reduction in the use of the same interventions for patients predicted to have a greater than 0.8 probability of a poor outcome.
The GCS score has been used to assess prognosis for a variety of medical and surgical diseases treated in intensive care units (ICUs). As shown in Fig 1, a decrease in GCS score is associated with a consistent but non-linear increase in hospital mortality among both non-operative and postoperative patients. The results from both risk stratification and multivariate analysis clearly showed that the GCS score is able to identify patients likely to survive and those likely to die ( B.a,s.t.o.s e.t..,a/ 1993). However, discriminating ability within intermediate levels of prognosis is poor, with death rates relatively constant between GCS scores of 7 and 11. It is for patients in this intermediate range where the value of factors such as other acute physiological abnormalities, age, comorbidity, and diagnosis are of great prognostic importance.
Fig. 1 Relationship between first ICU day GCS score and hospital mortality rate for postoperative patients ( n = 6786) (gray bar) and non-operative patients (n = 9187) (green bar). (Reproduced with permission from Bastosefa/ (1993))
For 596 patients with non-traumatic coma, most commonly due to cardiac arrest (36 per cent) and cerebral infarction or intracerebral hemorrhage (36 per cent), components of the GCS combined with abnormal brainstem response identified those at high risk of predicted death or severe disability ( Ham.®.! §LaL 1995). For patients with one or more abnormal brainstem responses (absent pupillary or corneal response, absent or dysconjugate roving eye movements) and no motor response to pain, the rate of death or severe disability at 2 months was 96 per cent. These results are similar to those reported by other investigators among patients with intracerebral hemorrhage and comatose survivors of cardiac arrest. For patients with intracerebral hemorrhage, a predictive model using the initial GCS score, hemorrhage size, and pulse pressure provided a valid and easy-to-use model for predicting probability of survival ( TuMm eL§L 1995).
Despite its worldwide acceptance and prognostic usefulness, the GCS score has several important limitations. First, initial scoring may be impossible for patients with severe head injury. This is because highly trained emergency medical personnel are able to intubate, sedate, or paralyze these patients before transport to a trauma center. As a result, it has been impossible to make an accurate determination of a true GCS score for almost 50 per cent of comatose head-injured patients in the emergency department. Second, current management for patients with severe head injury often includes sedation, narcotics, and paralysis for control of elevated intracranial pressure. Thus it may also be impossible to determine an accurate daily GCS score for these patients while in the ICU. Third, periorbital swelling hypotension hypoxia and intubation can interfere with scoring. We have encountered similar problems in measuring the GCS score for medical and surgical ICU patients who were included in the APACHE III data collection. Thus, for many critically ill patients, these limitations can interfere with the basic assumption that the GCS score reflects the severity of neurological injury or disease.
Recommendations for overcoming these problems have included the following.
1. Obtain an initial GCS score within 1 to 2 h after injury.
2. Avoid scoring until hypotension or hypoxia have been stabilized.
3. Use an ocular score of 1 for patients with severe periorbital swelling.
4. Carefully adhere to the definitions provided in the original description of the GCS.
5. Delay scoring until 10 to 20 min after the estimated half-life of medications which cause sedation or paralysis.
6. Record a GCS score of 15 if there is no prior accurate determination and sedation, or paralysis cannot be reduced.
At present there is no better measure that summarizes the impact of disease or injury on neurological function. Thus, whether used alone or as an important component of APACHE III or another prognostic system, the GCS score is an important predictor of patient outcome. This is why everything possible should be done to obtain a GCS score on all intensive care admissions. When direct measurement of the GCS score is impossible, a consistent policy of either recording a normal score or using other patient data to impute or estimate the score are the two alternatives. While this means that risk may be underestimated, treating missing data in a systematic manner within a clinical trial will minimize the impact for subsequent analyses.
Was this article helpful?