Journal of Neuroscience Nursing:
A Review of the Predictive Ability of Glasgow Coma Scale Scores in Head‐Injured Patients
Questions or comments about this article may be directed to Molly McNett, MSN RN, at email@example.com. She is a doctoral student at Kent State University, Kent, OH, and research coordinator in the department of nursing at MetroHealth Medical Center, Cleveland, OH.
According to 1999 data from the Centers for Disease Control and Prevention, traumatic brain injuries (TBI) caused by motor vehicle accidents, firearms, and falls are recorded as a leading cause of death and lifelong disability for young adults in the United States. Researchers have investigated if correlations exist between variables in the acute stage of injury and outcome measures in TBI patients. The Glasgow Coma Scale (GCS) score is one variable that was extensively studied for its ability to predict outcome in TBI patients. However, the use of different designs and methodologies in these studies makes the interpretation of the cumulative findings difficult. Therefore the purpose of this review was to provide a summary of the research findings on the ability of the GCS scores to predict outcome in TBI patients. A search was done on MEDLINE ® and CINAHL® to identify studies that investigated the predictive ability of the GCS score. Studies that used the GCS as a variable in predicting outcome with adult patients who had sustained some type of head injury were included. GCS scores are most accurate at predicting outcome in head‐injured patients when they are combined with patient age and pupillary response and when broad outcome categories are used. The motor component of the GCS yields similar prediction rates as the summed GCS score, and better prediction occurs with very high or very low GCS scores. Information about the cumulative research findings on the predictive ability of GCS scores aids nurses in providing support and education to family members during the acute stage of injury, and in coordinating the services of members of the healthcare team, which could result in improved outcomes for both patient and family.
Approximately 1.4 million Americans every year survive traumatic brain injury (TBI) with varying degrees of lifelong disability (CDC, 2004). Many of these patients with long‐term or lifelong disability are essentially in the prime of life when the injury occurs; the average age at the time of injury is 32 years old (CDC, 1999). Of the estimated $37.8 billion that is spent on treating TBI annually, $20.6 billion is allocated toward injury‐related work loss and disability (CDC, 1999).
Since its introduction into the literature in 1974, the Glasgow Coma Scale (GCS) has been widely used as a measure to record the level of consciousness in a number of patient populations, and particularly in patients who have sustained a TBI. Because the GCS score is one of the first things assessed at the time of injury and continues to be assessed throughout the course of treatment in TBI patients, many researchers have examined if a relationship exists between these GCS scores and certain outcome categories in TBI patients.
Identifying the predictive ability of the GCS score has implications for TBI patients, their family members, and for members of the healthcare team. Such information is particularly useful for nurses caring for this patient population. Nurses often provide bedside support and education to family members of TBI patients. Knowledge about correlations between GCS scores and potential patient outcomes would aid nurses in their understanding and explanation of possible outcomes. In addition, nurses are frequently involved in coordinating care services for TBI patients and their families with various healthcare professionals, such as social workers, case managers, chaplains, and physical therapists. Understanding the potential outcomes associated with GCS scores places the nurse in a key position to facilitate earlier consultation of these services, which could influence outcomes for the patient and his or her family.
This article reviews development of the GCS in assessing level of consciousness in TBI patients and synthesizes research findings on the predictive ability of the GCS in these patients. Implications of this information for neuroscience nurses and areas of future research are discussed.
Methods of Retrieval
In order to identify research that investigated the predictive ability of the GCS in patients who have sustained a traumatic brain injury, a search was done on MEDLINE® and CINAHL® for articles published from 1974, when the GCS was introduced, until 2005, using combinations of the following key words: Glasgow Coma Scale, GCS, traumatic brain injury, head injury, predictor, and outcome. Studies that used the GCS as a variable in predicting outcome with adult patients who had sustained some type of head injury were included. Many of the studies described the participants as having sustained a head injury due to a traumatic event, such as a motor vehicle crash, a fall, or an assault. Although these studies did not specifically label their participants as having sustained a traumatic brain injury, they were included in this review because the mechanism and type of injury were consistent with traumatic brain injury patients. For the purposes of this review, they will be referred to as head injury patients in order to be consistent with the research findings.
Research studies that examined outcomes only in pediatric patients were excluded from this review, based on the fact that the use of the GCS, patient outcomes, and prescribed therapies are often different in the pediatric population (Wellons & Tubbs, 2003). The reference lists of the included studies were reviewed and additional articles that met the above criteria were obtained and included. The findings from these research articles are presented in the following sections.
Development of the GCS
The GCS was developed by Teasdale and Jennett (1974) as a standardized method for healthcare practitioners to evaluate and describe the degree of altered consciousness or coma in patients who had sustained head injuries. The original GCS, introduced in 1974, was later revised by its creators (Teasdale & Jennett, 1976; Jennett & Teasdale, 1977). The resulting GCS (Jennett & Teasdale) is comprised of three separate components: eye opening, verbal response, and motor response. Within each category, there is a range of observations that serve to define the specific components. Each of these observations is associated with a score that is recorded by the healthcare provider as a means of measuring or classifying the degree of consciousness of the patient. The eye opening category ranges from 1 (no response) to 4 (spontaneous); the motor component ranges from 1 (no movement) to 6 (obeys commands); and the verbal component ranges from 1 (no response) to 5 (appropriate/oriented conversation). The score from each category (eye, motor, and verbal) is summed, to provide a total GCS score.
The development of the GCS by Teasdale and Jennett (1974) incorporated the theoretical model of level of consciousness that had been proposed by Plum and Posner (1972). Plum and Posner's model described level of consciousness as a continuum of different degrees of human responsiveness. At one end of the continuum is normal consciousness, where human responsiveness is characterized by maximum degrees of awareness and arousal. As the continuum progresses downward, the levels of awareness and arousal decrease, representing the states of clouding consciousness, delirium, and stupor. Finally, at the opposite end of the continuum is coma, which possesses no measurable degree of awareness or arousal (Plum, 1971; Plum & Posner; Posner, 1975; Segatore & Way, 1992).
In their research evaluating levels of consciousness, Plum (1971) and Posner (1975) assert that the evaluation of five clinical parameters can be used to identify the cause and anatomic location of the injury. These parameters include state of consciousness, respiratory status, pupillary size and reactivity to light, eye movements and ocular reflexes, and motor response. Jennett and Teasdale (1977) acknowledged these findings in their development of the GCS, but asserted that the cause and location of the injury in patients with head injury is usually known. They therefore proposed a simplified model, the GCS, which would measure only key aspects of level of consciousness in patients who had sustained head injuries. In their assessment, this simplified model would be easy to use by various members of the healthcare team, and could also be used repeatedly at different time periods to measure changes in level of consciousness in the headinjured patient (Teasdale & Jennett, 1974; Jennett & Teasdale). At the time the GCS was introduced into the literature, no other standardized measure for evaluating level of consciousness existed. The GCS provided a mechanism for quickly evaluating the severity of a head injury and a means to communicate this information effectively with other members of the healthcare team.
The GCS gained popularity because it was easy to use and provided a standardized way to evaluate level of consciousness. It was also due to these reasons that the scale began to be used widely with other patient populations when assessing their responsiveness. Although the scale became the standard for reporting level of consciousness, no research was done initially to investigate if it was a reliable tool or a valid measure for assessing severity of head injury or level of consciousness. This fact might account for the modest correlations reported in later research examining the relationship between GCS scores and patient outcomes.
Research on Predictive Ability of GCS
In reviewing the cumulative research on the predictive ability of the GCS, it was found that the individual research findings could be grouped into the following categories: the GCS score when combined with other clinical variables, the use of specific GCS scores, the individual components of GCS scores, and the GCS score in relation to certain outcome categories. These categories will be used to present the research findings on the relationships between GCS scores and outcomes in head‐injured patients.
GCS and Other Predictor Variables
In the years immediately following the introduction of the GCS, many researchers examined the GCS score on admission along with other variables present in the acute stages of head injury and correlated these scores with different outcome measures at specified time intervals postinjury. Admission GCS scores were collected in conjunction with numerous variables, such as age, eye movements and pupillary responses (Choi, Ward, & Becker, 1983; Choi, Narayan, Anderson, & Ward, 1988; Lokkeberg & Grimes, 1984; Narayan et al., 1981), computed tomography (CT) data and type of injury (Choi et al., 1983; Choi et al., 1988; Choi et al., 1991; Narayan et al.), multimodality evoked potentials and intracranial pressure (Narayan et al.), injury severity score (Pal, Brown, & Fleizser, 1989) periods of posttraumatic amnesia (Bishara, Partridge, Godfrey, & Knight, 1992), cerebral perfusion pressure (Changaris et al., 1987), head Abbreviated Injury score and mechanism of injury (Demetriades et al., 2004), and hyperglycemia and leukocytosis (Rovlias & Kotsou, 2004).
Logistic regression analyses in these studies reveal that the admission GCS score, when combined with age and pupillary or oculocephalic response, was 80%‐84% accurate in predicting patient outcome into one of two categories, good outcome (as defined by no disability or moderate disability), or poor outcome (defined by severe disability or persistent vegetative state; Choi et al., 1983; Narayan et al., 1981). Age, GCS score, and pupillary reaction were found to be the most significant variables in predicting patient outcome (p<.02, p<.0002, p<.0002, respectively; Narayan et al.; Rovlias & Kotsou, 2004). Similar studies reported that age, best motor response, and pupillary reaction were the best predictors of outcome using stepwise logistic regression with mortality as the main outcome measure (Mamelak, Pitts, & Damron, 1996). Combining the variables of age, summed GCS score, and pupillary response with the patient's best motor score in a separate study yielded a similar accuracy prediction rate of 78% (Choi et al., 1988). However, the outcome categories in this study were not as broad. Instead, accuracy of prediction was based on how well these variables correctly placed patient outcome into one of five categories: good recovery/no disability, moderate disability, severe disability, persistent vegetative state, or death. These categories collectively are known as the Glasgow Outcome Scale (GOS; Jennett & Bond, 1975) and will be discussed further in the following sections of this paper.
Only one study reported the amount of variance accounted for by the GCS score. In a study of 93 head‐injured patients, Lokkeberg & Grimes (1984) found that admission GCS score, when combined with patient age, accounted for 36% of the variance in patient outcome at 12 months, with outcome being measured by the GOS. Even when all other variables were included in the analyses, the total variance explained only rose to 40% (Lokkeberg & Grimes). Thus, 60% of the variance in predicting outcome in these patients remains unexplained by the variables in the study, which included demographic information, CT scan abnormalities, surgical interventions, and skull fractures.
The majority of studies that examined the GCS score in conjunction with CT data failed to report that the data from the CT scans substantially influenced the outcome prediction rate in headinjured patients (Choi et al., 1983; Choi et al., 1988; Narayan et al., 1981; Young et al., 1981). One study found that adding CT scan data to six clinical variables (age, GCS score, pupillary reaction, eye movements, surgical decompression, and motor posturing) improved the confidence level in the percentage of accurate outcome prediction. The six clinical variables alone yielded an 82% accurate prediction rate (with 43% above the 90% confidence level), whereas adding CT scan data changed the prediction rate to 77% (but was 52% above the 90% confidence level; Narayan et al.).
A separate study reported that the presence of an intracerebral lesion on CT scan influenced the prediction accuracy in patients who had sustained head injuries (Choi et al., 1991). However, this prediction rate was based on a number of other variables in addition to CT scan data, and only included the separate GCS scores of eye opening, verbal response, and motor response. The summed GCS score was not included in the analysis. The authors report that combining the following variables accurately predicted outcome in 78% of head‐injured patients: (1) the presence of intracerebral lesion on CT scan, (2) patient age, (3) best motor response, and (4) pupillary response (Choi et al., 1991). The GOS categories described above were used here again as the outcome categories. Lastly, Young and associates (1981) reported that combining data from CT scan (i.e., the presence of midline shift) with initial GCS scores did not prove to significantly improve the prediction of outcome in head‐injured patients.
One study investigated the length of posttraumatic amnesia (PTA) as a variable along with the GCS score in examining outcomes in head‐injured patients at 6 and 12 months (Bishara et al., 1992). Not only was a strong correlation found between these two variables (r = −.64, p < .0001), but both duration of PTA (r = −.50, p < .0001 at 6 months, and r = −.59, p < .0001 at 12 months) and GCS scores (r = .45, p < .0001 at 6 months, and r = .46, p < .0001 at 12 months) were shown to be statistically significant when examining patient outcome at both 6 months and 12 months (Bishara et al.). The outcome categories in this study were again those measured by the GOS.
Few studies investigated the GCS as the sole variable in predicting outcome in patients with head injury (Diringer & Edwards, 1997; Zafonte et al., 1996; Balestreri et al., 2004). Zafonte and associates examined the relationship between the GCS score and a number of outcome variables, such as the Disability Rating Scale (DRS), the Rancho Los Amigos Levels of Cognitive Function Scale (LCFS), and the cognitive and motor components of the functional independence measure (FIM‐COG and FIM‐M, respectively), which are all standardized instruments that are commonly used in assessing level of disability in head injury patients. Correlational analysis between GCS scores and these different outcome measures yielded only modest results with correlation coefficients ranging from −.28 (DRS) to .37 (FIM‐COG; Zafonte et al.). Similarly, Diringer and Edwards reported a modest ability of the GCS score to accurately predict level of independence (71%) in patients who had sustained head injuries.
Healey and associates (2003) examined the GCS score as a single predictor of outcome, but used a general trauma population in their sample rather than only head‐injured trauma patients. Findings from this study revealed that the motor component of the GCS was more effective in predicting outcome in trauma patients. This finding will be discussed further in the following section addressing the individual components of the GCS score.
Lastly, Balestreri and associates (2004) investigated correlations between admission GCS scores and GOS scores at 6 months postinjury in patients admitted to their facility over a period of 10 years. While significant correlations were found between admission GCS scores and GOS scores for the first 5 years of the study (r = .41, p < .00001, n = 183), the correlations for the remainder of the study period were not as strong (r = .091, p = 226, n = 175).
More recent research on predicting outcome in patients who have sustained head injuries includes the GCS score as part of a larger group of predictor variables labeled as injury severity variables (Bush et al., 2003; Novack, Bush, Meythaler, & Canupp, 2001). In an attempt to establish a path model for outcome prediction in this patient population, path analyses in these studies reveal that injury severity variables (one being the GCS score) exhibited a significant causal relationship with functional status (.54 and .35) and cognitive status (.46 and −.48; Bush et al., 2003; Novack et al., 2001). Thus, grouping the GCS scores with other clinical variables again influences its ability to predict outcome in this patient population.
Individual Components of GCS
One criticism of the GCS is that the summed score does not always provide an accurate depiction of a patient's condition. For example, the verbal score, particularly in trauma patients, may be poor due to the fact that the patient requires endotracheal intubation, and might result in the patient being assigned a much lower verbal score, which would give a falsely low summed GCS score. Much of the earlier research on the predictive ability of the GCS score used the summed GCS score, rather than the scores of the individual components of the GCS (eye, motor, verbal; Changaris et al., 1987; Choi et al., 1991; Choi et al., 1994; Lokkeberg & Grimes, 1984; Narayan et al., 1981; Young et al., 1981). However, one of the earlier studies described in the preceding section not only included the summed GCS score, but also investigated if the separate components of the scale could be used to predict outcome (Choi et al., 1988). The authors concluded that while the GCS, when combined with age, was an acceptable predictor of outcome in the study, using only the motor score from the GCS and patient age yielded similar accuracy rates for prediction (Choi et al., 1988).
More recent research on the predictive ability of the GCS has compared individual components of the GCS score with the GCS total score (Gill, Windemuth, Steele, & Green, 2005; Healey et al., 2003; Ross, Liepold, Terregino, & O'Malley, 1998). Ross and associates reported that the motor component of the GCS score accurately predicted outcome in head injury patients with nearly the same accuracy as the total GCS score (motor score sensitivity 91%, specificity 85%; total GCS score sensitivity 92%, specificity 85%). When comparing motor scores and total GCS scores using Receiver Operating Characteristic (ROC) curves, two separate studies demonstrated similar results. Healey and associates (2003) demonstrated that the motor component of the GCS occupied nearly the same area under an ROC curve (ROC = .87) as did the total GCS score (ROC = .89), arguing that there was only a 2% difference between the two scores in their ability to predict mortality in a sample of 200,000 general trauma patients. Gill and associates (2005) also found that the motor score, when compared to the total GCS score occupied similar areas under an ROC curve (ROC = .894 for motor, ROC = .906 for total GCS) in predicting mortality in patients who had sustained head injuries. Lastly, a study by Meredith and associates (1995) examined the motor component of the GCS and reported 59% sensitivity and 97% specificity rates in the ability of the motor score to predict outcome; however, these rates were not compared with the total GCS score. The cumulative findings from these studies lend support to the argument that the individual components of the GCS, particularly the motor score, might prove as useful as the total GCS score in predicting outcome in head‐injured patients.
Speci. c GCS Scores
A number of the studies included in this review used descriptive analyses in reporting the relationship between specific GCS scores and certain outcome categories (Bishara et al., 1992; Changaris et al., 1987; Lieberman et al., 2003; Pal et al., 1989; Young et al., 1981). Most studies reported positive outcomes in patients who presented with high initial GCS scores. Young and associates found that 72 of the 76 patients with initial GCS scores of 8‐15 had a good recovery or moderate disability, as measured by the GOS. Similarly, 99% of patients with GCS scores of 13‐15 had good recovery (Pal et al.), while 89%‐96% of patients with a GCS score of 8 or above (Bishara et al.) and 75% of patients with GCS score 6 and above (Changaris et al.) also had good recovery or moderate disability outcomes.
Similar patterns were noted among the studies describing the relationships between low GCS scores and patient outcomes (Bishara et al., 1992; Changaris et al., 1987; Lieberman et al., 2003; Young et al., 1981). In one study, only 14% of patients with GCS scores of 3‐5 were found to have good recovery, while 85% were reported as having moderate disability (Bishara et al.). However, the sample size for patients with these values was small (n = 7), which should be considered when attempting to generalize from these findings. In studies with slightly larger sample sizes, patients with GCS scores of 3‐4 had primarily poor outcomes. Changaris and associates reported that 50 of the 51 patients with GCS scores of 3‐4 died, while in a separate study 20 of the 21 patients with GCS scores of 3‐4 were either in a persistent vegetative state or had died (Young et al.). All of the patients in one study who presented to the emergency department with a GCS of 3 and had fixed and dilated pupils died, while 67% of those who presented with GCS scores of 3 but did not have fixed and dilated pupils also died within 72 hours (Lieberman et al.).
The research findings reviewed here support the obvious assumption that patients with high GCS scores will tend to have better outcomes, while patients with low GCS scores will most likely have poor outcomes. Studies examining GCS scores of 4‐9 describe various patient outcomes (Pal et al., 1989; Young et al., 1981). In one study, patients with GCS scores of less than 9 were classified as having either good outcome (35%) or death (41%; Pal et al.). In a separate study, 36 of 73 patients with GCS scores 5‐7 had good recovery or moderate disability (Young et al.). Because of the range of outcomes for head injury patients with these GCS scores, outcome prediction in this patient population by using GCS scores remains difficult.
Several of the reviewed studies used the GOS as the primary outcome measure (Bishara et al., 1992; Changaris et al., 1987; Choi et al., 1988; Choi et al., 1991; Lokkeberg et al., 1984; Pal et al., 1989). The GOS was developed by Jennett and Bond (1975) to provide a classification system to describe the various types of outcome that occur in patients with head injuries. As previously mentioned, the GOS is comprised of five categories with corresponding scores. These categories range from good recovery (GOS 5) to death (GOS 1). In studies examining the predictive ability of the GCS score, many researchers have used the GOS as the primary outcome measure because it allows for a general description of patient outcome.
Of the reviewed studies that used the GOS as the primary outcome measure, only two reported results of correlational analysis between GCS scores and GOS in attempting to predict outcome in head‐injured patients. Bishara and associates (1992) reported statistically significant relationships (r = .45; p < .0001) between admission GCS scores and outcome, as measured by the GOS at 6 and 12 months postinjury, while Changaris and associates (1987) reported significant (p < .001) positive correlations between GCS scores and GOS categories. The remaining studies that used the GOS as the primary outcome measure did not supply information pertaining to the statistical correlations between GCS score and GOS scores.
Some researchers in the reviewed studies chose to further combine the categories of the GOS in order to create broader outcome categories when examining the predictive ability of the GCS score. Choi and associates (1983), Narayan and associates (1981), and Young and associates (1981) created the outcome categories of good or favorable outcome and poor or unfavorable outcome. The category of good outcome was composed of patients who had good recovery or moderate disability, while the poor outcome category included patients who had severe disability, persistent vegetative state, or death. By creating these broad outcome categories, the researchers were able to demonstrate better accuracy in their outcome predictions. Choi and associates (1983) and Narayan and associates reported accurate outcome predictions as high as 80% when using early GCS scores and these outcome categories. Poon, Zhu, Ng, and Wong (2005) reported a 71% accuracy prediction rate of the GCS using similar outcome categories of moderate/ severe disability and good recovery. Thus, the use of broad outcome categories increased the ability of the GCS score to predict patient outcomes.
In contrast to the above studies using the GOSbased categories as outcome measures, a few researchers chose to include outcome measures that more specifically reflected the level of patient disability (Bush et al., 2003; Diringer & Edwards, 1997; Novack et al., 2001; Zafonte et al., 1996). Instruments such as the disability rating scale (DRS), the functional independence measure (FIM), the Rancho Los Amigos levels of cognitive functioning scale (LCFS), and the community integration questionnaire (CIQ) were used in these studies and are commonly employed in rehabilitation settings to measure the level of disability in the patient who has sustained a head injury. Correlations between GCS scores and these outcome measures were reported to be only modest (−.28 to .37; Zafonte et al.), or not significant (.10; Novack et al.).
Lastly, other researchers based outcome measures on length of stay (Shah & Muncer, 2003; Lieberman et al., 2003), patient mortality (Healey et al., 2003; Lieberman et al.; Mamelak et al., 1996), or other investigator‐developed categories (Diringer & Edwards, 1997). Although the study by Lieberman and associates did not examine the appropriateness of using the GCS to predict length of stay, Shah and Muncer reported that the GCS, in fact, was not an effective predictor of length of stay. The remaining studies using patient mortality and an investigator‐developed category did not report correlational analyses between the GCS score and these outcome measures.
Reviewing the outcome categories used in these studies examining the predictive ability of the GCS score reveals that the use of broad outcome categories, such as good recovery or poor recovery, allows for the best prediction rates. As outcome measures become more specific, the predictive accuracy using the GCS score decreases, as evidenced in the studies using instruments to measure functional ability. However, the use of these specific categories allows for a more accurate description of patient outcome, rather than simply classifying the patient as having had a good or poor outcome.
Research has shown that using the GCS score along with other clinical data, such as age and pupillary response, improves its predictive ability.
The above review of the research literature highlights what is known about the ability of the GCS score to predict outcomes in patients who have sustained a head injury. Research has shown that using the GCS score along with other clinical data, such as age and pupillary response, improves its predictive ability. Use of the motor component of the GCS score, rather than the summed GCS score, has been shown to yield similar or improved outcome prediction rates in head‐injured patients. Research has illustrated that a higher correlation exists between certain GCS scores (high GCS scores or low GCS scores) and different measures of outcomes. Lastly, the use of broad outcome categories improves the ability of the GCS score to predict patient outcomes.
This information is important for nurses who care for head‐injured patients when they are coordinating resources for patients and providing support to family members. Nurses are frequently involved in consultation with other disciplines that play a crucial role in the rehabilitation process of the head‐injured patient. Physical, speech, and occupational therapists; social workers; and case managers are often involved in caring for the headinjured patient from the acute stage of treatment to rehabilitation, and even in the home setting. By having information about the potential outcomes in head‐injured patients, nurses may facilitate earlier consultation of these services, which could improve outcomes for the head‐injured patient. Earlier consultation of these services could also aid in more efficiently aligning the needed resources for family members. Counseling family members regarding the anticipated physical and cognitive changes as well as addressing expected home‐care needs of the patient both can be started prior to the patient's discharge from the acute care facility. This could facilitate family members' emotional adjustments to the patient's injury and allow them to prepare for the needs of the patient following the rehabilitation phase of treatment.
Nurses are in a pivotal position to provide support and education to family members of headinjured patients throughout the course of treatment. Knowledge about potential outcomes for these patients during the acute stage of injury has been cited as one of the most important needs of family members (Duff, 2002; Hauber & Testani‐Dufour, 2000; Man, 2002). It is therefore beneficial for nurses to have information about the possible outcomes for these patients in order to address the often difficult questions that may be posed by family members.
Areas for Future Research
A review of the research findings on the predictive ability of the GCS provides for identification of future areas of research. Based on the above review, one such area of needed research is to identify other variables that might account for outcome in patients who had sustained head injury. The fact that approximately 60% of the variance in reporting outcome in this population remains unaccounted for suggests that other factors may play a crucial role in influencing how a patient recovers from head injury. A second area of research is to explore the validity of the GCS. The fact that the reviewed studies found the motor score to perform similarly to the summed GCS score indicates that the total GCS score may not be a valid measure in recording level of consciousness. Lastly, the review of the outcome categories presented here illustrates that outcome prediction is most effective when using broad outcome measures. Although these measures give a general indication of outcome, they are not useful in actually describing level of disability in head‐injured patients. Thus, additional research on predicting outcome into more specific categories is needed.
The identification of variables in the acute stage of head injury that could provide insight into patient outcomes is of great interest to healthcare providers and researchers. The GCS score is one such variable. Numerous researchers have investigated the use of the GCS score in predicting outcomes in head injury patients. The cumulative findings from these studies reveal that GCS scores are most accurate at predicting outcome in head‐injured patients when they are combined with patient age and pupillary response and when broad outcome categories are used. In addition, the motor component of the GCS yields similar prediction rates as the summed GCS score, and better prediction occurs with very high or very low GCS scores. Knowledge about the potential outcomes for head‐injured patients based on these early GCS scores can prove to be invaluable to the families, as well as to nurses and other healthcare professionals who care for and support these patients and their loved ones throughout the recovery process.
Balestreri, M., Czosnyka, M., Chatfield, D., Steiner, L., Schmidt, E., Smielewski, P., et al. (2004). Predictive value of Glasgow Coma Scale after brain trauma: Change in trend over the past ten years. Journal of Neurology, Neurosurgery, & Psychiatry, 75
Bishara, S. N., Partridge, F. M., Godfrey, H. P., & Knight, R. G. (1992). Post-traumatic amnesia and Glasgow Coma Scale related to outcome in survivors in a consecutive series of patients with severe closed-head injury. Brain Injury, 6
Bush, B. A., Novack, T. A., Malec, J. F., Stringer, A. Y., Millis, S. R., & Madan, A. (2003). Validation of a model for evaluating outcome after traumatic brain injury. Archives of Physical Medicine and Rehabilitation, 84
Centers for Disease Control and Prevention, Division of Injury and Disability Outcomes and Progress. (2004). Traumatic brain injury in the United States: Emergency department visits, hospitalizations, and deaths. Retrieved March 10, 2006, from www.cdc.gov/ncipc/pub-res/TBI_in_US_04/TBI-USA_Book-Oct1.pdf
Changaris, D. G., McGraw, C. P., Richardson, J. D., Garretson, H. D., Arpin, E. J., & Shields, C. B. (1987). Correlation of cerebral perfusion pressure and Glasgow Coma Scale to outcome. Journal of Trauma, 27
Choi, S. C., Barnes, T. Y., Bullock, R., Germanson, T. A., Marmarou, A., & Young, H. F. (1994). Temporal profile of outcomes in severe head injury. Journal of Neurosurgery, 81
Choi, S. C., Muizelaar, J. P., Barnes, T. Y., Marmarou, A., Brooks, D. M., & Young, H. F. (1991). Prediction tree for severely head-injured patients. Journal of Neurosurgery, 75
Choi, S. C., Narayan, R. K., Anderson, R. L., & Ward, J. D. (1988). Enhanced specificity of prognosis in severe head injury. Journal of Neurosurgery, 69
Choi, S. C., Ward, J. D., & Becker, D. P. (1983). Chart for outcome prediction in severe head injury. Journal of Neurosurgery, 59
Demetriades, D., Kuncir, E., Murray, J., Velmahos, G., Rhee, P., Chan, L. (2004). Mortality prediction of head abbreviated injury score and Glasgow Coma Scale: Analysis of 7,764 head injuries. Journal of the American College of Surgeons, 199
Diringer, M. N., & Edwards, D. F. (1997). Does modification of the Innsbruck and the Glasgow Coma Scales improve their ability to predict functional outcome? Archives of Neurology, 54
Duff, D. (2002). Family concerns and responses following a severe traumatic brain injury: A grounded theory study. Axon, 24
Gill, M., Windemuth, R., Steele, R., & Green, S. M. (2005). A comparison of the Glasgow Coma Scale score to simplified alternative scores for the prediction of traumatic brain injury outcomes. Annals of Emergency Medicine, 45
Hauber, R., & Testani-Dufour, L. (2000). Living in limbo: The lowlevel brain-injured patient and the patient's family. Journal of Neuroscience Nursing, 32
Healey, C., Osler, T. M., Rogers, F. B., Healey, M. A., Glance, L. G., Kilgo, P. D., et al. (2003). Improving the Glasgow Coma Scale score: Motor score alone is a better predictor. Journal of Trauma, 54
Jennett, B., & Bond, M. (1975). Assessment of outcome after severe brain damage. Lancet, 1
Jennett, B., & Teasdale, G. (1977). Aspects of coma after severe head injury. Lancet, 1
Lokkeberg, A. R., & Grimes, R. M. (1984). Assessing the influence of non-treatment variables in a study of outcome from severe head injuries. Journal of Neurosurgery, 61
Lieberman, J. D., Pasquale, M. D., Garcia, R., Cipolle, M. D., Mark Li, P., & Wasser, T. E. (2003). Use of admission Glasgow Coma Score, pupil size, and pupil reactivity to determine outcome for trauma patients. Journal of Trauma, 55
(3), 437-442; discussion 442-433.
Mamelak, A., Pitts, L., & Damron, S. (1996). Predicting survival from head trauma 24 hours after injury: A practical method with therapeutic implications. Journal of Trauma: Injury, Infection, and Critical Care, 41
Man, D. (2002). Family caregivers' reactions and coping for persons with brain injury. Brain Injury, 16
Meredith, W., Rutledge, R., Hansen, A., Oller, D., Thomason, M., Cunningham, P., et al. (1995). Field triage of trauma patients based upon the ability to follow commands: A study in 29,573 patients. Journal of Trauma, 38,
Narayan, R. K., Greenberg, R. P., Miller, J. D., Enas, G. G., Choi, S. C., Kishore, P. R., et al. (1981). Improved confidence of outcome prediction in severe head injury. A comparative analysis of the clinical examination, multimodality evoked potentials, CT scanning, and intracranial pressure. Journal of Neurosurgery, 54
Novack, T. A., Bush, B. A., Meythaler, J. M., & Canupp, K. (2001). Outcome after traumatic brain injury: Pathway analysis of contributions from premorbid, injury severity, and recovery variables. Archives of Physical Medicine and Rehabilitation, 82
Pal, J., Brown, R., & Fleiszer, D. (1989). The value of the Glasgow Coma Scale and Injury Severity Score: Predicting outcome in multiple trauma patients with head injury. Journal of Trauma, 29
Plum, F. (1971). Clinical aspects of coma. Clinical Neurosurgery, 18,
Plum, F., & Posner, J. (1972). The diagnosis of stupor and coma.
Philadelphia: FA Davis.
Poon, W. S., Zhu, X. L., Ng, S. C., & Wong, G. K. (2005). Predicting one year clinical outcome in traumatic brain injury at the beginning of rehabilitation. Acta Neurochirurgica, 93 (
Posner, J. (1975). Clinical evaluation of the unconscious patient. Clinical Neurosurgery, 22,
Ross, S., Leipold, C., Terregino, C., & O'Malley, K. (1998). Efficacy of the motor component of the Glascow Coma Scale in trauma triage. Journal of Trauma, 45,
Rovlias, A., & Kotsou, S. (2004). Classification and regression tree for prediction of outcome after severe head injury using simple clinical and laboratory variables. Journal of Neurotrauma, 21
Segatore, M., & Way, C. (1992). The Glasgow Coma Scale: Time for change. Heart and Lung, 21
Shah, S., & Muncer, S. (2003). A comparison of rehabilitation outcome measures for traumatic brain injury. OTJR: Occupation, Participation and Health, 23
Teasdale, G., & Jennett, B. (1974). Assessment of coma and impaired consciousness: A practical scale. Lancet, 2
Teasdale, G., & Jennett, B. (1976). Assessment and prognosis of coma after head injury. Acta Neurochirurgica, 34,
Wellons, J., & Tubbs, R. (2003). The management of pediatric traumatic brain injury. Seminars in Neurosurgery, 14
Young, B., Rapp, R. P., Norton, J. A., Haack, D., Tibbs, P. A., & Bean, J. R. (1981). Early prediction of outcome in head-injured patients. Journal of Neurosurgery, 54
Zafonte, R. D., Hammond, F. M., Mann, N. R., Wood, D. L., Black, K. L., & Millis, S. R. (1996). Relationship between Glasgow Coma Scale and functional outcome. American Journal of Physical Medicine & Rehabilitation, 75
© 2007 American Association of Neuroscience Nurses