Appropriate Use of the Glasgow Coma Scale in

Intubated Patients: A Linear Regression

Prediction of the Glasgow Verbal Score from the
Glasgow Eye and Motor Scores
Rutledge Robert MD FACS; Lentz, Christopher W. MD;
Fakhry, Samir MD, FACS; Hunt, John MD

The Journal of Trauma: Injury, Infection, and Critical CareJournal of

Trauma: Injury, Infection & Critical Care.
41:p 514-522, September 1996.

Author Information

From the Department of Surgery, University of North Carolina at Chapel

Hill, Chapel Hill, North Carolina.

Address for reprints: Robert Rutledge, MD, FACS, Associate Professor of

Surgery, Department of Surgery, University of North Carolina at Chapel
Hill, Campus Box 7210, Burnett-Womack, Chapel Hill, NC 27599-7210.

Abstract

The Glasgow Coma Scale (GCS) has been shown to be a valuable tool in
assessing the neurologic and physiologic status of critically ill patients.
Unfortunately, the GCS requires assessment of the verbal response of the
patient and this can be blocked by intubation. The purpose of this study was
to assess the ability of a regression model based upon the eye and motor
components of the GCS to accurately predict the verbal response of the
GCS. The primary hypothesis was that the verbal response could be derived
from the motor and eye responses of the GCS.

Data were collected prospectively in an intensive care unit computer data

base. Patients were divided into training and test data sets. Linear regression
was used to derive a model of verbal score from the motor and eye scores of
the GCS in the training data set. Correlation between the actual and the
predicted verbal scores was calculated.

A total of 2,521 GCS assessments were available for analysis. The second
order multiple regression model was an accurate predictor of the verbal
score (Pearson's Correlation r =0.9, R2 = 0.8, p = 0.0001) in 1,463
observations in the training data set.

Estimated GCS Verbal = (2.3976) + [GCS Motor X (-0.9253)] + [GCS Eye

X (-0.9214)] + [(GCS Motor) sup 2 X (0.2208)] + [(GCS Eye)2 X (0.2318)]
where r = 0.91, R2 = 0.83, and p = 0.0001.

The accuracy of this model was confirmed by comparing the predicted

verbal score to the actual verbal score in the test data set (n = 736, r = 0.92,
R2 = 0.85, p = 0.0001)

The GCS is a useful tool in the intensive care unit and a critical part of the
APACHE II assessment of patient acuity. GCS has been shown to be a
useful tool in its own right as a predictor of outcome in the critically ill. Its
use is limited with intubation. (See Segatore M, Way C: Heart Lung 21:548,
1992; and Lieh-Lai MW, Theodorou AA, Sarnaik AP, et al: J Pediatr
120:195,1992.) The present study demonstrates that a relatively simple
regression model can use the eye and motor components of the GCS to
predict the expected verbal component of the GCS, thus allowing the
calculation of the GCS sum score in intubated patients.

The Glasgow Coma Scale (GCS) has become a versatile tool in the care of
a spectrum of patients. It was originally developed by Teasdale and Jennett
as a method of grading various levels of coma and consciousness. [1] The
scale uses three gross indicators of central nervous system function,
including eye opening (E), verbal performance (V), and motor
responsiveness (M), and could be easily and consistently applied by all
health care personnel. Although initially designed as a practical scoring
system to assess the level of consciousness in patients with head injuries, its
role has expanded beyond that of a descriptor. Application of the GCS
provided the framework that facilitated the comparison of patients with
neurologic dysfunction within and among various institutions. [2,3] The
GCS has also been used to compare the effectiveness of treatment, to serve
as a prognostic indicator [4-8] and to facilitate patient triage, thus
improving resource utilization.

The GCS scoring system has been demonstrated to be a significant

predictor of outcomes in patient populations both with and without head
injuries. It has been observed that the presence of coma significantly affects
the outcome of intensive care unit patients. [9] Champion et al. incorporated
the GCS into the original Trauma Score [10] as well as the Revised Trauma
Score, [11] both of which have been shown to be accurate predictors of
outcome for all trauma victims, including those with and those without
serious head injuries. [12] In the development of the acute physiology and
chronic health evaluation (APACHE II) score, Knaus et al. included
neurologic function as measured by the GCS. [13] Subsequent studies have
shown that the GCS is the most powerful predictive factor in the APACHE
II score, inasmuch as the contribution of GCS to the predictive model is
larger than any other variable measured. [14,15]

Despite its usefulness, the GCS has some limitations restricting its
application in certain patient populations. [16-18] The most frequent
problem with the use of the GCS is the inability to assess the verbal
response because of endotracheal intubation. [3,17] In a review of the
reporting of GCS in the neurosurgical literature, Starmark et al. found 44 of
56 head-injury studies did not report how untestable features of the GCS
were handled. [19] Some series of reports excluded between 4 to 18% of
patients because of the absence of testable components of the GCS. [20-23]
Gale et al. found that 38% of 451 admissions with head injury had
incomplete GCS scores. [24] Several methods have been used to overcome
this problem of missing information, including assigning the lowest
possible score to untestable components [3] or pseudoscoring of missing
values based on the measurements of testable features. [25] Unfortunately,
these compensations for untestable features are not applied universally and
can account for the disparity in the mortality reported from different
centers. Therefore, there is a need for establishing some criteria to
accurately assign values to the untestable verbal component in intubated
patients.
The purpose of this study was to attempt to assess the association of the
GCS eye and motor components with the GCS verbal score. Furthermore, it
was our goal to attempt to explore this relationship and to attempt to create
a mathematical model of this relationship that might allow for the
appropriate use of the GCS in intubated patients. It was our primary
hypothesis that a mathematical relationship could be derived between the
eye, motor, and verbal components of the GCS. That is to say, it was our
hypothesis that there was a strong correlation between the GCS verbal
component and the eye and motor components of the GCS. Our goal was to
test this hypothesis and to derive a variety of candidate models for
calculating the GCS verbal score from available GCS motor and eye scores.

Back to Top

MATERIALS AND METHODS

Back to Top

Study Population

The study population for this analysis was derived from the surgical
intensive care unit (SICU) data base at the University of North Carolina
Hospitals. This data base, which has been in place since 1987, is a
concurrent data collection system that tracks a variety of patient-specific
information on a daily basis. Included in this daily acquisition of data is the
GCS component and sum score, as well as a number of other physiologic
variables. This data base was the source of the information used in this
study. Nonintubated patients with a complete GCS sum score were
considered for analysis.

Back to Top

Inclusion/Exclusion Criteria

All patients admitted to the SICU are monitored in the SICU data base.
Each patient receives an admission note and a daily note. These notes are
derived from the SICU data base system that prompts the staff to enter a
GCS value for each patient for each day in the SICU. All admitted patients
during the time of the study were analyzed. Patients were selected for the
study who had both the outcome of the SICU stay and the GCS available
for analysis. Patients included were a heterogenous group. Some patients
were receiving sedatives and hypnotics; some patients were recovering
from general anesthesia. Unfortunately, these data were not available for
stratification, but Glasgow is routinely used in these situations. It is
assumed that controlling for these factors might improve the predictions of
the regression model. Thus, if the eye and motor scores could predict the
verbal score in this situation without this information, it might do even
better with further stratification based upon other patient factors. Both the
admission and daily GCS sum scores were used in the study. Each GCS
measure was analyzed independently.

Back to Top

Linear Regression Modeling of the GCS Verbal Score

After creation of the study population data set, the analysis plan included an
initial exploratory analysis that plotted the association between the two
independent variables (GCS eye and motor scores) and the dependent
variable (GCS verbal score). In addition, the plots allowed assessment of
possible outliers for exclusion from the regression modeling process. Based
on these graphs, a number of different variable transformations were
included in the regression modeling. Subsequently, the statistical analysis
system was used to create a linear regression model of the GCS verbal score
from the GCS eye and motor scores. To decrease problems with overfitting
of the data, the study population was divided into two subgroups. The first
two thirds of the data base was used arbitrarily as the training data set. The
last third was used to test the various regression models that were generated
during the training phase of the analysis. In the test data base, various
predictive models generated from the training data were used to calculate
predicted values of the GCS verbal score based upon the eye and motor
scores. These predictive values were then tested for correlation with the
actual values of the verbal score in the test data set.
Back to Top

RESULTS

Back to Top

Demographics

In the study population, there were a total of 665 nonintubated patients (287
women and 378 men) with a total of 2,521 complete GCS sum score
measurements. The average age was 53.6 +/- 19.4 and the mean APACHE
II score was 12.1 +/- 6.7. These patients were admitted to general surgery
and surgical subspecialties as depicted in Table 1.

Admitting service of nonintubated patients admitted to the SICU.

Back to Top

Exploratory Analyses
The first step in the project plan was an exploratory graphical analysis of
the relationships of the eye and motor components of the GCS with the
verbal score. The frequency distribution of the study GCS values are shown
in Figure 1 and Figure 2. There was a positive correlation between both the
GCS motor and eye score with the verbal score. This is shown in Figure 3
and Figure 4. The frequency distribution of the observed verbal score at
each eye and motor score is shown in Table 2 and Table 3 and is plotted in
Figure 2. There was a significant number of GCS scores with a motor
component of 6 and a verbal component of 1, as well as a significant
number of eye scores of 4 with a verbal component of 1. This may have
represented patients with tracheotomies, not requiring mechanical
ventilation, with an arbitrary assignment of 1 for the verbal component of
the GCS sum score. Of particular concern are the cells with GCS verbal
scores of 1 and eye scores of 3 or 4. Figure 3 shows the mean verbal score
for each eye score and demonstrates a nearly linear relationship. These
inconsistencies are a demonstration of the fact that this data set is a real
world system in daily use for patient care and not a reasearch project with
specially funded collection of the GCS. This was viewed as both a strength
and a weakness of the project. Table 2 and Figure 2, Figure 3, Figure 4 and
Figure 5 show the relationships of the eye, motor, and the verbal scores. As
the table and the figures show, there is some linearity in the relationship
between the scores, but there are also what seem to be outliers with motor
scores of 6 and verbal scores of 1. Figure 5 and Table 4 are three-way plots
of the motor and eye scores versus mean verbal scores. Again, this table and
graph demonstrate that the relation among the scores is not perfectly linear
with what appear to be outliers at motor scores of 2 and 6. There were other
outliers observed that represented unlikely score combinations. Patients
with motor scores of 2 or 3 (decerebrate or decorticate posturing) could not
have verbal scores greater than 1. Rather than trying to overfit our
mathematical model to explain these outliers, they were removed from the
analysis. A total of 322 (13%) GCS scores were excluded because the score
components were deemed to be outliers. A separate analysis was also
performed including the outliers, and the results, although not presented
here, showed very similar predictive power to the analyses presented below.
Distribution of GCS in the study population.

The frequency distribution of the GCS motor and verbal scores.

GCS eye score and mean GCS verbal score.

GCS motor score and mean GCS verbal score.

The frequency distribution of the GCS eye score and GCS verbal scores.

The frequency distribution of the GCS motor score and GCS verbal scores.
GCS eye and motor scores and mean verbal score.

GCS eye and motor scores and mean verbal score.

Back to Top

Statistical Analysis

Back to Top
Linear Regression Analysis

Because this is the first presentation of this approach to basing the GCS
verbal score on the GCS eye and motor scores, a detailed description of the
statistical analyses performed is presented below. The first step in the linear
regression analysis was to begin modeling the dependent variable, the GCS
verbal score, by using the available independent variables, the GCS eye and
motor scores, in the training data set of 1,463 measurements. The multiple
linear regression model was initially limited to a first order (linear)
equation. Successive mathematical models were given access to the square
and the cube of the GCS eye and motor scores. Thus, in other modeling
runs with the linear regression process, the model was given access to the
square and the cube of the GCS eye and motor scores. The first regression,
Model 1, is shown in Table 5. The total R2 for the model was 0.76. The
predictive equation for the model is shown below: Equation 1

Parameter estimates, R2 and p values for Model 1.

Because calculation of this equation is somewhat complex, a table of the

calculated predicted values for the GCS verbal score from Model 1 is
shown in Table 6 and displayed graphically in Figure 6. Inspection of the
table and the figure demonstrates the linear nature of the model, which is to
be expected because this is a first order equation.

Calculated predicted GCS verbal score from eye and motor scores using linear regression Model 1.

Calculated verbal score using Model 1.

The second regression model was based upon the GCS eye and motor
scores as well as the squares of both scores. The results of the forward
selection process are shown in Table 7. The total R2 for Model 2, in which
the linear regression procedure was given access to the GCS eye and motor
scores and to the square of the GCS eye and motor scores, had an R sup 2
of 0.83. This was an improvement of 0.07 over Model 1. The equation for
calculating the estimated GCS verbal score is shown below: Equation 2

Parameter estimates, R2 and p values for Model 2.

A table of the calculated predicted values for the GCS verbal score from
Model 2 is shown in Table 8 and displayed graphically in Figure 7.
Calculated predicted GCS verbal score from eye and motor scores using linear regression Model 2.

Calculated verbal score using Model 2.

The total R2 for Model 3, in which the linear regression procedure was
given access to the GCS eye and motor scores and to the square and the
cube of the GCS eye and motor scores, had an R2 of 0.832 (Table 9). This
was an improvement of 0.07 over the model, but of no difference from the
model with access to only the square of the eye and motor scores. The
equation for calculating the estimated GCS verbal score is shown below:
Equation 3
Parameter estimates, R2 and p values for Model 3.

(Table 10) shows the calculated values of the verbal score based upon the
equation from the third regression model. This is displayed graphically in
Figure 8.
Predicted GCS verbal score using linear regression Model 3.

Calculated verbal score using Model 3.

Back to Top

Analysis of the Association of the Predicted Verbal Scores with Actual

GCS Verbal Scores in the Test Data Set

After generating a number of possible candidate mathematical models from

the training data set, these models were tested as predictors of the GCS
verbal score in the test data set of 736 measurements. Calculated verbal
GCS scores were obtained by substituting the GCS motor and eye scores
from the test data set into each of the mathematical models. The correlation
of the calculated predicted verbal score with the actual verbal score in the
test data set is shown in Table 11. There was a statistically significant
correlation between the derived verbal score from the first order multiple
regression model and the actual observed verbal score (r = 0.87, p =
0.0001). The correlation using the second and third order mathematical
models Equation 2 and Equation 3 was much improved over the linear
model (r = 0.9179, p = 0.0001; r = 0.9194, p = 0.0001).

Correlation analysis in the test data set.

Back to Top

DISCUSSION

The GCS has provided clinicians with a simple, yet reliable, method of
quickly quantifying a patient's level of consiciousness. Since its
introduction in 1974, it has evolved into much more than a neurologic
assessment device. In the absence of expensive or invasive measurements,
the GCS makes it possible to grade the severity of head injury based on
observations obtained at the bedside. It has also been demonstrated that this
grading serves as a useful predictor of outcome in patients with cerebral
trauma, [4-8,26-30] as well as other non-traumatic causes of coma. [31-39]

The GCS has been useful in predicting the outcome of patients in the
intensive care unit. [38,40,41] Teres et al. [9] found that the presence of
coma, from either metabolic or neurologic causes, was associated with a
significant increase in `mortality of intensive care unit patients. Data
collected for the Finnish National Intensive Care Study demonstrated that
the GCS had a positive predictive value of 76.5% in predicting mortality of
604 patients admitted to intensive care after cardiac arrest. [14] They also
noted that the GCS was responsible for a significant proportion of the
predictive power of the APACHE II in this patient population. Using the
data base for the APACHE III study, [42] Bastos et al. [15] evaluated the
ability of the GCS to predict hospital mortality rates for nontrauma
admissions to intensive care. They reported that varying degrees of
alterations in consciousness, measured by the GCS, were able to predict
mortality rates in nontrauma patients requiring critical care. The
discriminating ability of the GCS to predict outcome with intermediate sum
scores was poor, with relatively constant death rates at scores between 7 and
11. These studies emphasize the potentially important role of the GCS in
stratifying mortality risk in critically ill patients with and without head
injury.

Calculation of the GCS has other important functions in the management of

critically ill patients. Several authors have advocated the use of the GCS in
determining treatment options in different patient populations. Based on
100 consecutive gunshot wounds to the head, the division of neurological
surgery at the Barrow Neurologic Institute has used the postresuscitation
GCS to decide whether aggressive surgical therapy should be offered. [25]
Also, the GCS has been used as a factor in determining prolonged
intubation in trauma victims. [43,44] These studies concluded that a low
GCS score [7,10,11] after 48 to 72 hours in the intensive care unit
correlated positively with greater than 14 days of mechanical ventilation,
supporting early tracheotomy in these patients.

Although the calculation of the GCS sum score in critically ill patients has
multifaceted applications, the presence of untestable components of the
GCS limits its usefulness. [17,18] The most common untestable feature of
the GCS is the verbal score, which is usually due to the presence of
endotracheal intubation. [3] The rapid treatment of patients with a severe
head injury often includes prehospital intubation, but such measures
compromise the ability to obtain an accurate GCS score in the emergency
department. A variety of techniques have been proposed in different centers
for designating a verbal score in intubated patients. These include assigning
an arbitrary score of 1 in patients on mechanical ventilation. Others have
suggested pseudoscoring techniques that assign missing components with
an average value of the testable scores. [24] Another method assigns the
verbal score for intubated patients as follows: 5, if patient seems able to
talk; 3, if there is a questionable ability to talk; and 1, if patient is generally
unresponsive. In a recent survey of major head-injury centers in the United
States, most neurosurgical personnel used a nonnumerical designation of
"T" for the verbal score in intubated patients. This, however, prevents the
calculation of the GCS sum score and thereby reduces the utility of this
assessment device. Unfortunately, these compensations for untestable
features are not universally applied, have not been validated, and may
account for disparity in the mortality reported from different centers.
Therefore, there is a clear need for establishing some criteria to accurately
assign values to the untestable verbal component in patients on mechanical
ventilation.

The goal of this study was to devise a method of determining an accurate

GCS verbal score in patients that are endotracheally intubated. This study
supports the hypothesis that a mathematical relationship exists between the
individual components of the GCS. There is a strong and consistent positive
correlation between the GCS eye and motor scores and the verbal score.
Furthermore, this relationship is easily modeled by using first, second, and
third order multiple linear regression equations. These models are very
accurate predictors of the verbal score when applied to the test data set. All
models functioned well; however, the more complex second and third order
equations were more accurate predictors of the observed verbal scores.
Because all these equations are cumbersome to apply clinically, the tables
facilitate the calculation of a predicted verbal score given the GCS eye and
motor scores. Given the well-documented importance of the GCS and the
large number of patients in whom the GCS in its present form cannot be
calculated, it may be that the present approach is the best methodology for
assessing the GCS in intubated patients. Both of the higher order models do
an excellent job of modeling the verbal score and could be recommended.
The continued use of the "T" designation for the GCS is highly limiting for
comparison purposes and should be dropped.
There are a number of factors in this study that raise questions about the
results. First, the study was performed in a population of SICU patients and
the majority of these patients had relatively high GCS values. Furthermore,
the sample population came only from one hospital and the overall sample
size could be larger. All of these issues need to be addressed by others
attempting to confirm our work. A major concern is the presence of what
we elected to classify as outliers. These were GCS measurements that were
felt to be recorded in error. Other studies have clearly documented a relation
between experience with the GCS and its accurate assessment. [45-47] All
of our measures are performed by SICU residents during the SICU rotation.
We recognize that this group of individuals can and does have limited
expertise with the GCS before their rotation with our team. Thus, a more
accurate data set of training data could lead to a better model.

Another concern was the elimination of 322 GCS measurements from the
population before the regression analysis. These patients were eliminated
on the basis of having a verbal score of 1 with a motor score of 6 and an eye
score of 4. There are few clinical scenarios in this population of
nonintubated patients that could satisfy this score. Because there were a
significant number of patients admitted to the intensive care unit from the
otolaryngology service, it was believed that they accounted for this
disparity. These patients are uniformly admitted to the intensive care unit
with tracheotomies after a major head and neck dissection. It is recognized
that eliminating these patients may be responsible for the high correlation
coefficient of the linear regression models. However, overfitting these
regression models to explain these outliers could not be justified.

The authors would like to include a cautionary note again emphasizing the
skewed nature of the patient population from which the data were derived.
The vast majority of patients had high GCSs. A smaller group had GCS
score of 3 and only a very small minority had GCS values in the midrange
of the GCS where prediction is the most difficult. Although it is true that
this is what one would expect in any other intensive care unit population,
the "good" relation between eye, motor, and verbal scores may not be as
clear in the subgroup of intermediate values of the GCS. Just as the
APACHE score is useful in predicting individual outcome at its extremes of
values, it performs much less effectively at the intermediate zones. The
readers should recognize that the intermediate values could be of different
reliability than those at the extremes.

Back to Top

CONCLUSION

The GCS is a critically important tool in a number of areas of medicine and

surgery. It is of value in head injury, in trauma and in critically ill patients
with CNS dysfunction, shock or other factors leading to depressed levels of
consciousness. Because of its great utility, the limitations of the GCS are all
that much more problematic. It is very important to be able to include
patients who are intubated in the group of patients who can be assessed with
the GCS. The present study demonstrates that there is a strong correlation
between the eye and motor components and the verbal component. The
verbal score is easily and accurately modeled, and based on two simple
tables, the predicted GCS can easily be calculated at the bedside in
intubated patients.

Back to Top

REFERENCES

1. Teasdale G, Jennett B: Assessment of coma and impaired consciousness:

A practical scale. Lancet 2:81, 1974
Cited Here...
2. Langfitt TW: Measuring the outcome from head injuries. J Neurosurg
48:673, 1983
Cited Here...
3. Marshall LF, Becker DP, Bowers SA, et al: The national traumatic coma
data bank. Part I: Design, purpose, goals, and results. J Neurosurg 59:276,
1983
Cited Here...
4. Young B, Rapp RP, Norton JA, et al: Early prediction of outcome in
head-injured patients. J Neurosurg 54:300, 1981
Cited Here...
5. Williams JM, Gomes F, Drudge OW, et al: Predicting outcome from
closed head injury by early assesment of trauma severity. J Neurosurg
61:581, 1984
Cited Here...
6. Gensemer IB, Smith JL, Walker JC, et al: Psychological consequences of
blunt head truma and relation to their indicies of severity of injury. Ann
Emerg Med 18:9, 1989
Cited Here...
7. Pal J, Brown R, Fleiszer D: The value of the Glasgow Coma Scale and
Injury Severity Score: Predicting outcome in multiple trauma patients with
head injury. J Trauma 29:746, 1990
Cited Here...
8. Rocca B, Martin C, Viviand X, et al: Comparison of four severity scores
in patients with head trauma. J Trauma 29:299, 1990
Cited Here...
9. Teres D, Brown RB, Lemeshow S: Predicting mortality of intensive care
patients: The importance of coma. Crit Care Med 10:86, 1982
Cited Here...
10. Champion HR, Sacco WJ, Carnazzo AJ, et al: The trauma score. Crit
Care Med 9:672, 1981
Cited Here...
11. Champion HR, Sacco WJ, Copes WS, et al: A revision of the trauma
score. J Trauma 29:623, 1990
Cited Here...
12. Schwartz ML, Sharkey PW, Andersen JA: Quality assurance for
patients with head injuries admitted to a regional trauma unit. J Trauma
31:962, 1991
Cited Here...
13. Knaus WA, Draper EA, Wagner DP, et al: APACHE II: A severity of
disease classification system. Crit Care Med 13:818, 1985
Cited Here...
14. Niskanen M, Kari A, Nikki P, et al: Acute physiology and chronic
health evaluation (APACHE II) and Glasgow coma scores as predictors of
outcome from intensive care after cardiac arrest. Crit Care Med 19:1465,
1991
Cited Here...
15. Bastos PG, Sun X, Wagner DP, et al: Glasgow Coma Scale score in the
evaluation of outcome in the intensive care unit: Findings from the Acute
Physiology and Chronic Health Evaluation III study. Crit Care Med
21:1459, 1993
Cited Here...
16. Ellis A, Cavanagh SJ: Aspects of neurosurgical assessment using the
Glasgow Coma Scale. Intensive Crit Care Nurs 8:4, 1992
Cited Here...
17. Segatore M, Way C: The Glasgow Coma Scale: Time for change. Heart
Lung 21:548, 1992
Cited Here...
18. Lieh-Lai MW, Theodorou AA, Sarnaik AP, et al: Limitations of the
Glasgow Coma Scale in predicting outcome in children with traumatic
brain injury. J Pediatr 120:195, 1992
Cited Here...
19. Starmark JE, Holmgren E, Stalhammar D: Current reporting of
responsiveness in acute cerebral disorders: A survey of the neurosurgical
literature. J Neurosurg 69:692, 1988
Cited Here...
20. Jagger J, Fife D, Vrnberg K, et al: Effect of alcohol intoxication on the
diagnosis and apparent severity of brain injury. Neurosurgery 15:303, 1984
Cited Here...
21. Klauber MR, Toutant SM, Marshal LF: A model for predicting delayed
intracranial hypertension following severe head injury. J Neurosurg 61:695,
1984
Cited Here...
22. Turazzi S, Bricolo A, Pasut ML: Review of 1,000 consecutive cases of
severe head injury treated before the advent of CT scanning. Acta Nerochir
72:167, 1984
Cited Here...
23. vanDongen KJ, Braakman R, Gelpke GJ: The prognostic value of
computerized tomography in comatose head-injured patients. J Neurosurg
59:951, 1983
Cited Here...
24. Gale JL, Dikmen S, Wyler A, et al: Head injury in the Pacific
Northwest. Neurosurgery 12:487, 1983
Cited Here...
25. Grahm TW, Williams FC, Harrington T, et al: Civilian gunshot wounds
to the head: A prospective study. Neurosurgery 27:696, 1990
Cited Here...
26. Margulies DR, Hiatt JR, Vinson D Jr, et al: Relationship of
hyperglycemia and severity of illness to neurologic outcome in head injury
patients. Am Surg 60:387, 1994
Cited Here...
27. Katz DI, Alexander MP: Traumatic brain injury. Predicting course of
recovery and outcome for patients admitted to rehabilitation. Arch Neurol
51:661, 1994
Cited Here...
28. Livingston DH, Loder PA, Hunt CD: Minimal head injury: Is admission
necessary? Am Surg 57:14, 1991
Cited Here...
29. Metz CA, Livingston DH, Smith JS, et al: Impact of multiple risk
factors and ranitidine prophylaxis on the development of stress-related
upper gastrointestinal bleeding: A prospective, multicenter, double-blind,
randomized trial. The Ranitidine Head Injury Study Group. Crit Care Med
21:1844, 1993
Cited Here...
30. O'Sullivan MG, Statham PF, Jones PA, et al: Role of intracranial
pressure monitoring in severely head-injured patients without signs of
intracranial hypertension on initial computerized tomography [see
comments]. J Neurosurg 80:46, 1994
Cited Here...
31. Tuhrim S, Dambrosia JM, Price TR, et al: Intracerebral hemorrhage:
External validation and extension of a model for prediction of 30-day
survival. Ann Neurol 29:658, 1991
Cited Here...
32. Sano K: Grading, and timing of surgery for aneurysmal subarachnoid
haemorrhage. Neurol Res 16:23, 1994
Cited Here...
33. Chan B, Gaudry P, Grattan-Smith TM, et al: The use of Glasgow Coma
Scale in poisoning. J Emerg Med 11:579, 1993
Cited Here...
34. Gott PS, Rabinowicz AL, DeGiorgio CM: P300 auditory event-related
potentials in nontraumatic coma: Association with Glasgow Coma Score
and awakening. Arch Neurol 48:1267, 1991
Cited Here...
35. Giacino JT, Kezmarsky MA, DeLuca J, et al: Monitoring rate of
recovery to predict outcome in minimally responsive patients. Arch Phys
Med Rehabil 72:897, 1991
Cited Here...
36. Wolf-Gould C, Osei L, Commey JO, et al: Pediatric cerebral malaria in
Accra, Ghana. J Trop Pediatr 38:290, 1991
Cited Here...
37. Benesch CG, McDaniel KD, Cox C, et al: End-stage Alzheimer's
disease: Glasgow Coma Scale and the neurologic examination. Arch Neurol
50:1309, 1993
Cited Here...
38. Karkela J, Pasanen M, Kaukinen S, et al: Evaluation of hypoxic brain
injury with spinal fluid enzymes, lactate, and pyruvate. Crit Care Med
20:378, 1992
Cited Here...
39. Bratton SL, Jardine DS, Morray JP: Serial neurologic examinations
after near drowning and outcome. Arch Pediatr Adolesc Med 148:167, 1994
Cited Here...
40. Antonelli M, Moro ML, Capelli O, et al: Risk factors for early onset
pneumonia in trauma patients. Chest 105:224, 1994
Cited Here...
41. Lee H, Hawker FH, Selby W, et al: Intensive care treatment of patients
with bleeding esophageal varices: Results, predictors of mortality, and
predictors of the adult respiratory distress syndrome. Crit Care Med
20:1555, 1992
Cited Here...
42. Zimmerman JE (ed): The APACHE III study design: Analytic plan for
evaluation of severity of outcome. Crit Care Med 17(suppl):S169, 1989
Cited Here...
43. Johnson SB, Kearney PA, Barker DE: Early criteria predictive of
prolonged mechanical ventilation. J Trauma 33:95, 1992
Cited Here...
44. Lanza DC, Koltai PJ, Parnes SM, et al: Predictive value of the Glasgow
Coma Scale for tracheotomy in head-injured patients. Ann Otol Rhinol
Laryngol 99:38, 1990
Cited Here...
45. Marion DW, Carlier PM: Problems with initial Glasgow Coma Scale
assessment caused by prehospital treatment of patients with head injuries:
Results of a national survey. J Trauma 36:89, 1994
Cited Here...
46. Menegazzi JJ, Davis EA, Sucov AN, et al: Reliability of the Glasgow
Coma Scale when used by emergency physicians and paramedics. J Trauma
34:46, 1993
Cited Here...
47. Rowley G, Fielding K: Reliability and accuracy of the Glasgow Coma
Scale with experienced and inexperienced users [see comments]. Lancet
337:535, 1991
Cited Here...

© Williams & Wilkins 1996. All Rights Reserved.