A useful new coma scale in acute stroke patients: four score.
Abstract: Assessment of the severity of unconsciousness in patients with impaired consciousness, prediction of mortality and prognosis are currently the most studied subjects in intensive care. The aim of this study was to investigate the usefulness of the Full Outline of UnResponsiveness (FOUR) score in intensive care unit patients with stroke and the associations of FOUR score with the clinical outcome and with other coma scales (Glasgow [GCS] and Acute Physiology and Chronic Health Evaluation II). One hundred acute stroke patients (44 male, 56 female), who were followed in a neurology intensive care unit, were included in this prospective study. The mean age of the patients was 70.49 [+ or -] 12.42 years. Lesion types were determined as haemorrhagic in 30 and ischaemic in 70 patients. FOUR scores on the day of admission and the first, third and 10th days of patients who died within 15 days were lower when compared to scores of patients who survived (P=0.005, P=0.000, P=0.000 and P=0.000 respectively). Receiver operating characteristic curve analysis showed significant trending with both FOUR score and GCS for prognosis; the area under curve ranged from 0.675 (95% confidence interval 0.565 to 0.786) when measurements had been made on day 3 to 0.922 (95% confidence interval 0.867 to 0.977) and 0.981 (95% confidence interval 0.947 to 1.015) for day 10. We suggest that FOUR score is a useful scale for evaluation of acute stroke patients in the intensive care unit as a homogeneous group, with respect to the outcome estimation.

Key Words: FOUR score, coma, acute stroke
Article Type: Report
Subject: Coma (Physiological aspects)
Stroke patients (Prognosis)
Stroke patients (Care and treatment)
Authors: Kocak, Y.
Ozturk, S.
Ege, F.
Ekmekci, A.H.
Pub Date: 01/01/2012
Publication: Name: Anaesthesia and Intensive Care Publisher: Australian Society of Anaesthetists Audience: Academic Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2012 Australian Society of Anaesthetists ISSN: 0310-057X
Issue: Date: Jan, 2012 Source Volume: 40 Source Issue: 1
Geographic: Geographic Scope: United Kingdom Geographic Code: 4EUUK United Kingdom
Accession Number: 280719628
Full Text: Cerebrovascular disorders are the second most frequent cause of mortality worldwide; thus, estimation of prognosis by grading the clinical symptoms utilising particular scales is increasingly gaining in importance (1,2). The aim of utilising prognostic scales is to use current diagnostic and management protocols more efficiently. Clinicians may refer to different terms for determining the grade of neurological dysfunction, particularly of unconsciousness. As a result, dependable coma scales with regard to their sensitivity are required. Coma and other impairments in consciousness frequently arise in cerebrovascular diseases and can be evaluated with a variety of scales.

The search for a reliable coma scale was first initiated with development of the Glasgow Coma Scale (GCS) in 1974 and to date, various scales have been developed and used for this purpose (3,4). However, even the GCS, which is the most frequently utilised, has some limitations and disadvantages; thus new scoring systems are being developed. The Full Outline of UnResponsiveness (FOUR) score was created by Wijdicks et al in 2005 and is widely utilised currently for grading consciousness impairment (5). The FOUR score was actually developed considering the limitations of GCS, and it is claimed that it is superior due to several of its features.

FOUR score was formerly studied in comatose states caused by several aetiologies (6-9). The current study was designed to investigate the association of FOUR score with clinical symptoms and other coma scales related to prognosis in patients with cerebrovascular events who were followed in an intensive care unit.


One hundred patients with stroke (44 male, 56 female) who were followed in the Neurology-1 intensive care unit of Ankara Numune Education and Research Hospital between March 2008 and June 2009 were included in the study. Patients receiving neuromuscular blocking agents and sedatives were excluded.

Stroke was diagnosed according to the definition of the World Health Organization: neurological deficit of cerebrovascular cause that persists beyond 24 hours or is interrupted by death within 24 hours (10).

Previous cerebrovascular events, systemic disorders (diabetes mellitus, hypertension, cardiac diseases), smoking, alcohol consumption, drug abuse and family background regarding vascular diseases were recorded. Physical and neurological examinations were performed and blood pressure, pulse and body temperature were measured.

Brain computerised tomography scans of all patients were evaluated and lesion types (ischaemic, haemorrhagic), dimensions and localisations (frontal, parietal, occipital, basal ganglia, thalamus, brainstem, cerebellum) were determined. For calculating the lesion size, the cross-section of the largest part of the lesion was taken into consideration, and lesion size was calculated as length x width x number of cross-sections occupied. Accompanying oedema, mass effect, periventricular hypodensity and atrophy were also recorded.

Blood samples were obtained on the day of admission to ascertain fasting blood glucose, urea, creatinine, alanine aminotransferase, aspartate aminotransferase, total protein, albumin, alkaline phosphatase, lipid profile, blood count, haemostatic profile and sedimentation rate.

Level of consciousness of patients, cranial nerve paresis, lateralisation of paresis and degree of paresis were recorded.

FOUR score, GCS, Acute Physiological Assessment and Chronic Health Evaluation (APACHE) II, Glasgow Outcome Scale and Rankin scales were performed on the day of admission and on the first, third and 10th days by the same neurologist (5,11-13).

SPSS version 15.0 software was used in the statistical analyses. Student's t-test, Mann-Whitney U test, chi-square test and Pearson correlation test were utilised for statistical evaluation of data. Receiver operating characteristic (ROC) curve analysis was used to compare FOUR and GCS scores. Values of P <0.05 were established as indicating significance threshold.


One hundred consecutive patients were included in the study. The mean age of the 44 male and 56 female patients was 70.49 [+ or -] 12.42 years. Demographic features of the patients are shown in Table 1.

The lesion type was haemorrhagic in 30 and ischaemic in 70 patients. Consciousness was unimpaired in 14% of the patients; somnolence was determined in 46%, stupor in 26% and coma in 14%. Lesion localisations were as follows: four frontal, three temporal, 55 parietal, one occipital, seven basal ganglia, eight thalamic, 11 brainstem, nine cerebellum and two unlocalised. In 52 patients lesions were accompanied by mass effect and in 75 patients by oedema. Additionally, periventricular hypoperfusion was determined in 81 patients and generalised atrophy in 76 patients.

Mortality was considered regarding the prognostic evaluation. Coma and stupor were significantly associated with poor prognosis (P=0.046). Degree of upper extremity paresis was positively correlated with poor prognosis (P=0.000), as was severity of lower extremity paresis (P=0.000).

Lesion type influenced the outcome. Mortality was significantly higher in the ischaemic stroke group as compared to haemorrhagic stroke (P=0.000). Presence of oedema, mass effect and periventricular hypoperfusion were also associated with poor outcome (P=0.000). FOUR scores on the day of admission and on the first, third and 10th days were lower in patients who died within 15 days (9.6286, 8.5652, 6.8136, 5.2353 respectively) when compared to scores of survivors (11.9000, 12.9333, 13.500, 15.3571 respectively; P=0.005, P=0.000, P=0.000 and P=0.000 respectively). GCS values were also significantly lower in patients who died (8.8571, 8.0435, 6.3559, 5.1765 respectively) when compared to scores of survivors (10.4667, 11.3333, 12.0000, 14.0714 respectively; P=0.021, P=0.000, P=0.000, P=0.000 respectively). Again, APACHE II scores on admission were significantly lower in patients who died compared to survivors (13.9667 and 18.1714 respectively; P=0.000) (Table 2, Figures 1 and 2). Additionally, to evaluate two scales (GCS and FOUR scale) for trend for prognosis between survivors and non-survivors, ROC curve analysis was applied for days on the day of admission and on the first, third and 10th days. With the ROC analysis for FOUR scores on admission, third and 10th days, the area under curve ranged from 0.675 (95% confidence interval [CI] 0.565 to 0.786) when measurements had been made on day 3 to 0.922 (95% CI 0.867 to 0.977) and 0.981 (95% CI 0.947 to 1.015) for day 10. Similarly with the ROC analysis for GCS scores on admission, third and 10th days, the area under curve ranged from 0.624 (95% CI 0.510 to 0.738) when measurements had been made on day 3 to 0.925 (95% CI 0.873 to 0.977) and 0.982 (95% CI 0.951 to 1.013) for day 10 (Figure 3).





It is crucial to evaluate and estimate a prognosis for a patient with unconsciousness during follow-up. The prognosis of such a patient depends on aetiological factors, the severity and duration of unconsciousness and any accompanying systemic symptoms. The degree of unconsciousness is the major factor affecting the prognosis. While the survival rate of ischaemic stroke patients with no impairment in consciousness is 96%, it is below 30% in comatose patients (14). Another study revealed the mortality rate in comatose patients as 71%, in stupor 45%, in somnolence 27%, and in patients without consciousness impairment as 11% (15).

The frequency of consciousness impairment in intensive care units varies, but is generally seen in neurological intensive care, with rates ranging between 46 to 80%. Wijdicks et al found the frequency as 70%, while Gremec et al found it as 80% (6,16). The frequency of consciousness impairment in our study was 86%, which is consistent with the results of the trials mentioned above.

A wide-ranging neurological evaluation including observation, neurological examination, laboratory tests, neuroimaging and electrophysiological investigation should be performed in unconscious patients. Furthermore, the follow-up of those patients necessitates teamwork and common language. Brief but informative systems that lead to rational follow-up and allow health care providers to estimate prognosis are required (17).

An ideal coma scale should expose the status of consciousness and its alterations over time so that anticipation of prognosis can be made (17).

The GCS, which was developed by Teasdale and Jennett in 1974, is still being utilised not only by neurologists and brain surgeons but also by other physicians as a standardised coma scale (3,17). Although the GCS is easy to use and efficient, it has some limitations, such as motor responses dominate while verbal responses of intubated patients cannot be evaluated (18,19). Application problems due to each parameter's having separate total points were also reported (19).

Aside from GCS, there are other coma scales that are still used on a widespread basis. One of the frequently used scales is APACHE II, which was developed by Knaus et al in 1981 (11,20). In addition to the GCS score, it comprises 12 different parameters concerning physiological variables, age and chronic diseases (11,20). By means of many trials, APACHE II has been shown to be fairly accurate in estimating mortality in coma. Rordorf et al indicated that both GCS and APACHE II exhibit dependable estimation rates about mortality within 30 days (21). Separate examinations of these two scales revealed higher APACHE II scores and lower GCS scores in patients who died when compared to survivors (22-24). Nevertheless, APACHE II possesses some limitations, in that the parameters depend on laboratory results, which leads to calibration and calculation difficulties; furthermore, there is no convention on the manner of application between raters (21).

Wijdicks provided a comprehensive definition of an ideal coma scale, and emphasised that it should be an instrument capable of measuring different aspects of coma and be reliable, valid, applicable, easily recalled and informative about the course of a patient. It is also preferred that it be internally consistent (i.e. if the value of one parameter changes, values of others should change), easily applicable by all health care providers, and with comparable results between raters. An ideal coma scale is neither overly detailed, causing difficulties in application, nor too simple, leading physicians to overlook alterations in the status of a patient. It should also be sensitive and repeatable for prospective trials (17).

Taking into account these criteria, Wijdicks et al developed a new coma scale in 2005 (5). This new scale, termed FOUR, evaluates coma through four parameters: eye response, motor response, brainstem reflexes and breathing (17).

Verbal response is not included in the scoring since aphasic and intubated patients cannot be evaluated, as was denoted and discussed in earlier studies (5,17).

For each parameter, five different values can be chosen (0 to 4). The category of motor response includes myoclonic status epilepticus, indicating poor prognosis after cardiac resuscitation (25).

Decorticate response and flexion are combined, as their differentiation is difficult. Hand position tests are used to evaluate both praxis and alertness (26).

Three distinct brainstem reflexes are used in several combinations for examining the mesencephalon, pons and medulla oblongata functions. Cough reflex, corneal reflex and pupillary reaction are the major functions evaluated. Breathing patterns are also graded. Cheyne-Stokes and irregular breathing indicate bi-hemispheric or caudal brain stem dysfunction. These examinations are crucial for evaluating comatose patients with stroke involving the brainstem (5,27).

The first evaluative study about FOUR score was performed in a neurology intensive care unit with a total of 120 patients with stroke (24%), head trauma, surgery and central nervous system infection. As a consequence of that study, FOUR score was found to be as effective as GCS; moreover, to be superior to it in some aspects (5). Evaluation of FOUR score in general intensive care units with fewer stroke patients revealed similar results (6,7). Sacco et al suggested that FOUR score is a reliable instrument to assess coma in stroke patients (28).

Studies performed in emergency services are valuable for evaluating the easy and rapid application of FOUR score. A trial in an emergency service, in which stroke patients constituted 22% of the total subjects, revealed similar results between the FOUR and GCS score (8). Eken et al found no difference between the effectiveness of GCS and FOUR score in an emergency service; however, the high correlation of results between FOUR raters was observed (9). Aetiological heterogeneity of the patients might have influenced the test results in their study.

To our best knowledge, our study is the first performed to assess the usefulness of FOUR score in acute stroke patients in an intensive care unit. The initial scores of our patients were obtained in the emergency service, and FOUR and GCS were found to be similarly effective for estimating prognosis. Only stroke patients were included in our study, and our results regarding the prognosis were in line with the first study, in which FOUR scores of patients who died within 15 days were significantly low on admission, and on the first, third and 10th days (P=0.005, P=0.000, P=0.000 and P=0.000 respectively). We also found FOUR score to be consistent with GCS by correlative tests concerning prognosis estimation. Additionally, ROC curve analysis showed that both GCS and FOUR score were valuable to predict prognosis between survivor and non-survivor patients. Interestingly, in ROC curve analysis, the value of these tests is increasing over the following days.

It was also found that FOUR score is easily applicable by assistant health care providers (29). Since the raters in our study were exclusively neurologists, we are unable to comment about application of the scale by assistant health care providers.

The GCS, FOUR score and APACHE II were found to be equally effective for prognosis estimation in our study, and there was also a correlation between them. However, APACHE II can be considered disadvantageous in light of the difficulties in its application and the long application time. While GCS and FOUR score depend on clinical evaluation and can be applied easily within a short duration, APACHE II depends on laboratory tests and thus is time-consuming for scoring. GCS and FOUR score necessitate no equipment and thus can be applied in ambulances and outside the hospital. APACHE II, on the other hand, requires some equipment and a hospital setting, and therefore has poor practicability compared to the other assessment scales. FOUR score is favourable in aphasic stroke patients in that it does not evaluate the verbal response.

FOUR score was compared not only with GCS but also with APACHE II in our study, and test results revealed a high correlation between the two scales. Former studies have not compared FOUR score with APACHE II. Yet many trials have investigated the correlation between GCS and APACHE II and have found results of the two highly concordant (16).

We suggest that FOUR score will be helpful in estimating prognosis in stroke patients in particular since it does not depend on verbal responses, provides information about brainstem reflexes and includes the evaluation of myoclonic epilepsy.

As our study was designed to investigate the utility of the FOUR score scale in patients with acute cerebrovascular events in particular, conditions that may influence consciousness through different mechanisms have been excluded; thus, information about a homogeneous patient group has been obtained. In conclusion, we consider that FOUR score is a beneficial and reliable scale for estimating prognosis in stroke patients in intensive care units.


(1.) Feigin VL. Stroke epidemiology in the developing world. Lancet 2005; 365:2160-2161.

(2.) World Health Organization. Neurological disorders; a public health approach. In: Neurological disorders. A Public Health Challenge. Switzerland: WHO Press 2006. p. 151-163.

(3.) Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet 1974; 2:81-84.

(4.) Wijdicks EF, Kokmen E, O'Brien PC. Measurement of impaired consciousness in the neurological intensive care unit: a new test. J Neurol Neurosurg Psychiatry 1998; 64:117-119.

(5.) Wijdicks EFM, Bamlet WR, Maramattom BV, Manno EM, McClelland RL. Validation of a new coma scale: The FOUR score. Ann Neurol 2005; 58:585-593.

(6.) Iyer VN, Mandrekar JN, Danielson RD, Zubkov AY, Elmer JL, Wijdicks EFM. Validity of the FOUR score coma scale in the medical intensive care unit. Mayo Clin Proc 2009; 84:694-701.

(7.) Orken DN, Kocaman Sagduyu A, Sirin H, Togay Isikra C, Gokce M, Sutlas N et al. Reliability of the Turkish version of a new coma scale: FOUR Score. Trakya Univ Tip Fak Dergisi 2010; 27:28-31.

(8.) Stead LG, Wijdicks EFM, Bhagra A, Kashyap R, Bellolio MF, Nash DL et al. Validation of a new coma scale, the FOUR score, in the emergency department. Neurocrit Care 2009; 10:50-54.

(9.) Eken C, Kartal M, Bacanli A, Eray O. Comparison of the Full Outline of Unresponsiveness Score Coma Scale and the Glasgow Coma Scale in an emergency setting population. Eur J Emerg Med 2009; 16:29-36.

(10.) Stroke--1989. Recommendations on stroke prevention, diagnosis, and therapy. Report of the WHO Task Force on Stroke and other Cerebrovascular Disorders. Stroke 1989; 20: 1407-1431.

(11.) Knaus WA, Zimmerman JE, Wagner DP, Draper EA, Lawrence DE. APACHE-acute physiology and chronic health evaluation: a physiologically based classification system. Crit Care Med 1981; 9:591-597.

(12.) Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet 1975; 1:480-484.

(13.) Rankin J. Cerebral vascular accidents in patients over the age of 60. II. Prognosis. Scott Med J 1957; 2:200-215.

(14.) Weimar C, Ziegler A, Konig IR, Diener Hans-C. Predicting functional outcome and survival after acute ischemic stroke. J Neurol 2002; 249:888-895.

(15.) Kassell NF, Torner JC, Haley EC Jr, Jane JA, Adams HP, Kongable GL. The International Cooperative Study on the Timing of Aneurysm Surgery. Part 1: Overall management results. J Neurosurg 1990; 73:18-36.

(16.) Grmec S, Gasparovic V. Comparison of APACHE II, MEES and Glasgow Coma Scale in patients with nontraumatic coma for prediction of mortality. Acute Physiology and Chronic Health Evaluation. Mainz Emergency Evaluation System. Crit Care 2001; 5:19-23.

(17.) Wijdicks EFM. Clinical scales for comatose patients: the Glasgow Coma Scale in historical context and the new FOUR Score. Rev Neurol Dis 2006; 3:109-117.

(18.) Meredith W, Rutledge R, Fakhry SM, Emery S, Kromhout-Schiro S. The conundrum of the Glasgow Coma Scale in intubated patients: a linear regression prediction of the Glasgow verbal score from the Glasgow eye and motor scores. J Trauma 1998; 44:839-844; discussion 844-845.

(19.) Healey C, Osler TM, Rogers FB, Healey MA, Glance LG, Kilgo PD et al. Improving the Glasgow Coma Scale score: motor score alone is a better predictor. J Trauma 2003; 54:671-678; discussion 678-680.

(20.) Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985; 13:818-829.

(21.) Rordorf G, Koroshetz W, Efird JT, Cramer SC. Predictors of mortality in stroke patients admitted to an intensive care unit. Crit Care Med 2000; 28:1301-1305.

(22.) Weingarten S, Bolus R, Riedinger MS, Maldonado L, Stein S, Ellrodt AG. The principle of parsimony: Glasgow Coma Scale score predicts mortality as well as the APACHE II score for stroke patients. Stroke 1990; 21:1280-1282.

(23.) Portenoy RK, Lipton RB, Berger AR, Lesser ML, Lantos G. Intracerebral haemorrhage: a model for the prediction of outcome. J Neurol Neurosurg Psychiatry 1987; 50:976-979.

(24.) Tuhrim S, Dambrosia JM, Price TR, Mohr JP, Wolf PA, Heyman A et al. Prediction of intracerebral hemorrhage survival. Ann Neurol 1988; 24:258-263.

(25.) Wijdicks EF, Parisi JE, Sharbrough FW. Prognostic value of myoclonus status in comatose survivors of cardiac arrest. Ann Neurol 1994; 35:239-243.

(26.) Wijdicks EF, Kokmen E, O'Brien PC. Measurement of impaired consciousness in the neurological intensive care unit: a new test. J Neurol Neurosurg Psychiatry 1998; 64:117-119.

(27.) Bordini AL, Luiz TF, Fernandes M, Arruda WO, Teive HAG. Coma scales: a historical review. Arq Neuropsiquiatr 2010; 68:930-937.

(28.) Sacco S, Carolei A. The FOUR Score: a reliable instrument to assess the comatose stroke patient. Eur Neurol 2010; 63:370-371.

(29.) Wolf CA, Wijdicks EFM, Bamlet WR, McClelland RL. Further validation of the FOUR score coma scale by intensive care nurses. Mayo Clin Proc 2007; 82:435-438.

Y. KOCAK *, S. OZTURK ([dagger]), F. EGE ([dagger]), A. H. EKMEKCI ([section])

Department of Neurology, Selcuk University, Selcuklu Medical Faculty, Konya, Turkey

* MD, Neurologist, Department of Neurology, Ankara Numune Education and Research Hospital.

([dagger]) MD, Professor of Neurology.

([double dagger]) MD, Neurologist, Department of Neurology, Ankara Numune Education and Research Hospital.

([section]) MD, Associate Professor of Neurology, Selcuklu Medical Faculty, Selcuk University.

Address for correspondence: Dr S. Ozturk. Email: serefnur.ozturk@ noroloji.org.tr

Accepted for publication on September 28, 2011.
Table 1
Demographic features of the patients

                                   n     %

Gender                    Male     44    44
                          Female   56    56

Heart disease             Yes      36    36
                          No       64    64

Diabetes mellitus         Yes      32    32
                          No       68    68

Hypertension              Yes      74    74
                          No       26    26

Smoking                   Yes      12    12
                          No       88    88

Alcohol                   Yes      2     2
                          No       98    98

History of previous CVA   Yes      20    20
                          No       80    80

Total                              100   100

CVA=cerebrovascular accident.

Table 2
FOUR scores, GCS, APACHE II and GOS results on the day of
admission and on the first, third and 10th days in patients
who died within 15 days and in survivors

Prognostic scale         Prognosis      n    Mean      SD        P

FOUR (on admission)      Survivor       30   11.9000   2.56434   0.005
                         Non-survivor   70   9.6286    3.96423

FOUR on the 1st day      Survivor       30   12.9333   2.47656   0.000
                         Non-survivor   69   8.5652    4.31645

FOUR on the 3rd day      Survivor       30   13.5000   2.40330   0.000
                         Non-survivor   59   6.8136    4.22013

FOUR on the 10th day     Survivor       28   15.3571   1.98540   0.000
                         Non-survivor   34   5.2353    4.18554

APACHE II on admission   Survivor       30   13.9667   4.42160   0.000
                         Non-survivor   70   18.1714   5.16390

GCS on admission         Survivor       30   10.4667   2.52891   0.021
                         Non-survivor   70   8.8571    3.35911

GCS on the 1st day       Survivor       30   11.3333   2.44009   0.000
                         Non-survivor   69   8.0435    3.39305

GCS on the 3rd day       Survivor       30   12.0000   2.44949   0.000
                         Non-survivor   59   6.3559    2.92293

GCS on the 10th day      Survivor       28   14.0714   2.20989   0.000
                         Non-survivor   34   5.1765    2.72442

GOS on admission         Survivor       30   3.2000    0.40684   0.057
                         Non-survivor   70   3.0429    0.35857

GOS on the 1st day       Survivor       30   3.2000    0.40684   0.017
                         Non-survivor   69   3.0145    0.32051

GOS on the 3rd day       Survivor       30   3.3000    0.46609   0.000
                         Non-survivor   59   3.0169    0.22679

GOS on the 10th day      Survivor       28   3.7500    0.51819   0.000
                         Non-survivor   34   2.9706    0.38810

FOUR=Full Outline of UnResponsiveness, GCS=Glasgow Coma Scale,
APACHE II Acute Physiology and Chronic Health Evaluation II,
GOS=Glasgow Outcome Scale.
Gale Copyright: Copyright 2012 Gale, Cengage Learning. All rights reserved.