Can Four Score Replace GCS For Assessing Neurological Status Of Critically Ill Patients –An Indian Study.

http://doi.org/10.33698/NRF0154 – Anita Mercy S, Ramesh Thakur, Sandhya Yaddanpudi, Hemant Bhagat

Abstract -Since its development in 1974, the Glasgow Coma Scale(GCS) has been the gold standard for assessing the patients’ Level Of Consciousness(LOC) and acute changes in neurological status. Despite its widespread use, the GCS has many limitations that are well documented. The Full Outline of UnResponsiveness(FOUR) score is a new coma scale that was recently validated as a proposed alternative to GCS. The main purpose of the study was to assess the reliability of the FOUR score in critically ill patients of main ICU of PGIMER, Chandgiarh. In this study, scoring of GCS and FOUR score were performed by anesthetists and nurses on 21 patients. In a total of 100 pair-wise ratings, the inter-rater agreement for both the FOUR score (k = 0.65) and the GCS (k = 0.66) was good among nurse – anesthetist pair. The internal consistency for both the FOUR score (á = 0.97) and the GCS scale (á = 0.94) was excellent. A good correlation was found between the FOUR score and the GCS (ñ = 0.94, p = 0.001). The predictive validity of the FOUR score [Area under the receiver operating curve(AUROC)= 0.8] is slightly higher than the GCS (AUROC = 0.76) in this population. But the results could not be concluded with the availability of very limited number of patients. Majority of the raters strongly agreed that the FOUR score can be used as an alternative to the GCS scale, though there is no significant difference in the raters’ opinion about both the scales. The study suggests that the new FOUR score has precise clinical acumen in detecting subtle changes in neurological status as the GCS scale. Inclusion of assessment of brainstem function along with the assessment of cortex function adds value to the new tool.

Keywords

Glasgow Coma Scale (GCS), Full Outline of UnResponsiveness score (FOUR), Unconsciousness

Correspondance at

 Anita Mercy

MSc. Nursing 2nd yrs student (Oncology Nursing) National Institute of Nursing Education (NINE), PGIMER, Chandigarh

Introduction

Consciousness is a state of general awareness of oneself and the environment and includes the ability to orient towards new stimuli.1 Despite advances in technology, a thorough clinical assessment is still the key to identify subtle changes in a patient’s neurological status and is fundamental to the management of neuroscience patients. To provide quality patient care, the bedside nurse must therefore be able to accurately and consistently assess and communicate these changes.2

Scales have been constructed to improve communication among health care personnel.3 The most commonly used scale is the GCS scale.3 GCS scale misses key essential elements of a comprehensive neurological examination for comatose patients.4 Failure to assess the verbal score in intubated patients and inability to test brain stem reflexes are shortcomings of GCS.5, 3, 6 The ability to assess the GCS motor score is often impacted by the administration of sedatives or neuromuscular blocking agents and the presence of confounders such as spinal cord injury. The GCS is skewed toward motor assessment, with a maximum of 6 points. This affects the ability to assign an accurate GCS to patients who are receiving medications or who have injuries that interfere with motor assessment.2 Subsequent investigations of the GCS had revealed disagreement among the raters, especially between experienced and inexperienced users.5, 7, 8, 9 Prior efforts to modify or replace this scale have been unsuccessful because no scale could improve on its simplicity and practical usefulness.

To address the many limitations to the GCS, Wijdicks et al, at the Mayo Clinic designed the FOUR score as a proposed alternative in 2005. 7, 10, 11 The FOUR score has been developed to assess the depth of coma in a more detailed manner than the GCS scale.11 The FOUR score assigns a value of 0 to 4 to each of four functional categories: eye response(E), motor response(M), brainstem reflexes(B), and respiration(R). In each of these categories, a score of 0 indicates nonfunctioning status, and a score of 4 represents normal functioning. The number of components and the maximal grade in each of the categories is four (E4, M4, B4, R4).3 In contrast to the GCS, the FOUR score eye response category assesses eye tracking in addition to eye opening, which allows it to differentiate vegetative state from minimally conscious state (MCS) patients.2, 7, 12 The FOUR scale also more accurately and expeditiously diagnose the locked in syndrome by specifically assessing voluntary eye movements.7, 12 The motor assessment includes response to pain, ability to follow simple commands, and the presence of generalized myoclonus status epilepticus, a poor prognostic sign in comatose survivors after cardiac resuscitation.2, 3, 13 Brainstem reflex category was created to assess the function of the mesencephalon, pons, and medulla, which allows diagnosis of uncal herniation.2, 7 Lower brainstem function is evaluated using the respiration category to identify irregular breathing patterns, including Cheyne-Stokes respirations.2

The FOUR score has been well received in and outside the United States and has been implemented at the Mayo Clinic Saint Mary’s Hospital. The nursing staff too have enthusiastically embraced the new coma scale in the United States. Studies on its usefulness outside the boundaries of the Neurological- Neurosurgical ICU are under way.11 Wijdicks et al (2005) and other investigators, have conducted validation of FOUR score in different clinical settings like medical ICU, neurology ICU and emergency department of Mayo Clinic involving different group of raters including experienced and inexperienced nurses and neuro physicians. Validation of FOUR score has also been done in different group of patients including stroke patients,15 pediatric population,2 critically ill patients11, 14 and neuroscience patients6, 11, 16 in different settings by various groups of raters in other places outside the Mayo Clinic. No studies related to FOUR score have been done in India. It is very important to have a simple and reliable clinical scale system to determine the level of consciousness in our clinical settings with which both the nurses and doctors are comfortable.

In this preview, the present study was planned to assess the reliability and feasibility of the FOUR score in critically ill patients admitted in main Intensive Care Unit.

Materials and Methods

This study was conducted at main ICU of Nehru hospital, PGIMER, Chandigarh, a premier institution of medical education and research in India which has been functioning as a tertiary care hospital. The main ICU is a 12 bedded multi-specialty ICU under the Department of Anesthesia and Intensive Care, which caters to medical, neurological, surgical, gynecological, and other intensive and critical care conditions. The target population of the study was critically ill patients and the raters (nurses and anesthetists) and the accessible population was critically ill patients admitted in main ICU of PGIMER, Chandigarh and the raters (nurses and anesthetists) of the same unit. Patients were selected by purposive sampling i.e.) all critically ill patients of four different categories (alert, drowsy, stuporous and comatose) admitted in main ICU, patients with age >12 yrs during 16th July till 31st august 2011 and 21 patients were included in the study. All patients who were on neuromuscular blocking agents were excluded from the study. Raters were selected by purposive sampling i.e.) all nurses and anesthetists who were on duty were included in the study and 32 raters were included in the study. All raters who were not willing to participate were excluded from the study. The sample size was determined before the study and a total of 100 pair-wise ratings of FOUR score and GCS were included in the study. The tools used for data collection were demographic proforma of the patients and the raters, the FOUR score, the GCS scale and the raters’ opinionnaire.

GCS scale & four score were standerised tools and other tools were prepared by thorough literature review and experts opinion and validated by different experts of the medical and nursing field. Ethical approval for the study was obtained from the Ethics Committee of the institution. A written permission was obtained from the HOD, Department of Anesthesia and main ICU, PGIMER. Sister-in-charge of main ICU was informed about the study. The anonymity and confidentiality of the participants (raters) in relation to findings was protected while repor ting the study. The period of data collection was from 16th July to 31st August 2011. The techniques of data collection followed were observation and records. Scoring on each patient was performed at the first possible admittance by the researcher for predictive validity analysis. Raters were oriented to the FOUR score by a teaching session using PowerPoint presentation on FOUR score along with demonstration on actual patients. Return demonstration was also taken from the raters. Posters on FOUR score was put up on walls of each cubical of main ICU. Written instructions about the FOUR score were given to each evaluator at the time of rating. Raters’ pair were nurse-nurse, anesthetist–nurse and anesthetist– anesthetist. Paired observations were made by the raters on each patient and each patient was observed by different pair of raters at different timings. Within the same hour, each evaluator in the pair recorded a FOUR score and a GCS score for the patient. At the end of data collection, raters were given opinionnaire on GCS and FOUR score and the responses were taken by 5-point Likert scale. The data was analyzed using both descriptive and inferential statistics. Calculation was carried out manually using a calculator and with the help of Microsoft excels (2007) and Statistical Package for Social Science (SPSS) programme version 16.

Results

During the study period of 6 weeks, 100 observations were done on 21 pateints.

Among 100 observations, 29 were done by nurse – nurse pair, 41 by nurse – anesthetist pair and 30 anesthetist – anesthetist pair. In total, 31 raters were enrolled in the study.

Among 21 patients, 9 (42.9%) belong to the middle adulthood (41 – 65 Yrs) group and the mean age of the patients was 44.4 ±1.9 years (range 17- 78 years). Among the 21 patients, 14 (66.7%) were males and 7 (33.3%) were females, 11 (52.4%) were intubated using orotracheal tube and 10 (47.6%) were intubated using tracheostomy and 8 (38.1%) died in ICU and 13 (61.9%) were alive and got transferred from ICU. Among 21 patients, 6(28.6%) were admitted with head injury. Among 32 raters, 12(37.5%) were holding Diploma in Nursing (GNM), 13(40.6%) were B. Sc Nursing (Nsg) holders, 6(18.8%) had   professional   qualification as 3 B. B. S, currently undergoing M.D and 1(3.1%) had M. D as qualification. The mean professional experience of the raters was 4.6 ± 5.5 years (range 2 months – 23 years). The mean experience of the raters in ICU was 2.7 ±2 years (range 2 months – 6 years & 8 months).

Internal consistency of the FOUR score and the GCS

Internal consistency of the FOUR score and the GCS was analyzed using Cronbach’s á and spearman’s correlation coefficient. á value of 0.5 or less is considered unacceptable, values between 0.5 and 0. 6 are considered poor, values between 0. 6 and 0.7 are considered questionable, values between 0.7 and 0.8 are  considered acceptable, values between 0. 8 and 0.9 are considered good and values above 0.9 are considered excellent internal consistency.6,18 á values for both the FOUR score and the GCS are shown in table. 1. Cronbach’s á for the FOUR score (0.95) and the GCS (0. 94) indicate excellent internal consistency for both the scores.

Table. 1: Internal consistency of the FOUR score and the GCS                                          N = 100

Variable Cronbach’ s alpha (á) p value
FOUR score 0.97 <0.001**
GCS 0.94 <0.001**

**Internal consistency is significant at the 0.01 level

Spearman correlation coefficient was used to analyze the correlation of the FOUR score and the GCS for 200 observations. A highly positive correlation of ñ = 0.94 between the FOUR score and the GCS score (Figure. 1)

Inter-rater agreement of the FOUR score and the GCS

Inter-rater agreement of the FOUR score and the GCS was analyzed using Cohen’s kappa. A kappa value of 0.4 or less is considered poor, values between 0.4 and 0.6 are considered fair to moderate, values between 0.6 and 0.8 are considered good and values above 0.8 are considered  excellent agreement.5, 2, 6

Kappa statistics of the FOUR score is shown in table-2 and kappa statistics of the GCS is shown in table-3. In 29 pair-wise ratings, the two nurses agreed exactly in 51.7% of the observations of the total FOUR score [k = 0.47(fair to moderate observer agreement)] and in 68.9% of the observations of the total GCS score [(k = 0.65(good observer agreement)], in 41 pair – wise ratings the nurse agreed exactly with the

2.00                 4.00                 6.00                 8.00               10.00               12.00

GCSscore

 Figure 1. Correlation between the GCS score and the FOUR score anesthetist in 68. 3% of the observations of the total FOUR score (k = 0.65) and in 71% of the observations of the total GCS score (k = 0.66) which shows that the degree of agreement between Nurse – Anesthetist pair is higher than that of other pairs (Nurse – Nurse pair and Anesthetist – Anesthetist pair), for FOUR score, k = 0.65(good observer agreement) and for GCS, k = 0.66 (good observer agreement), in 30 pair – wise ratings the two anesthetists agreed exactly in 53. 3% of the observations of the total FOUR score (k = 0. 48) and in 53.3% of the observations of the total GCS score [k = 0.46(fair to moderate agreement)] Overall among 100 pair – wise ratings the two group of raters agreed exactly in 59% of the observations of the total

FOUR score (k = 0. 55) and in 65% of the observations of the total GCS score (k = 0.6) which shows that there is fair to moderate agreement among all pairs.

Table. 2: Kappa values for the inter-rater agreement of the FOUR score                            N = 100

 

Rater Pair

  

n

 Kappa value (k)
Total

Score Score

 Eye

response Score

 Motor

response Score

 Brainste

reflexes

 Respiration

Score
Nurse – Nurse 29 0.47 0.58 0.74 0.74 0.90
Nurse – Doctor 41 0.65 0.70 0.64 0.86 0.78
Doctor- Doctor 30 0.48 0.69 0.59 0.78 0.32       `
Overall 100 0.55 0.67 0.66 0.80 0.69
C Curve            Motor response

Score
Kappa value (k)
 

Rater Pair

Table. 3: Kappa values for the inter-rater agreement of the GCS score                            N = 100

 

ICU mortality status prediction using the FOUR score was analyzed using ROC curvePredictive value of the FOUR score and the GCS

0.0

0.0                 0.2                   0.4                 0.6                  0.8                   1.0

1 – Specificity

 Diagonal segments are produced by ties.

Figure 2. shows the predictive value of the GCS (AUROC is 0.76, 95% confidence interval 0.55 – 0.96, p = 0.55) and the predictive value of the FOUR score (AUROC is 0.8, 95% CI: 0.6 – 1, p = 0.25). Out of 8 patients who died in ICU, 1 (12.5%) got the minimum GCS score of 3 and 1 (12.5%) got the minimum FOUR score of 2.

Practical feasibility of the FOUR score The descriptive statistics of the raters’ opinion about the practical feasibility of the GCS is shown in table. 4. Among 31 raters, 18(58.1%) agreed to the statement that GCS is reliable, 20(64.5%) agreed to the statement that GCS is easy to use, 19(29%) agreed to the statement that GCS is reproducible, 16(51.6%) agreed to the statement that GCS gives more clinical information and 16(51.6%) agreed to the statement that GCS takes less time to perform.

Table. 4: Raters’ opinion on practical feasibility of the GCS                                                  N = 31

Statements regarding

GCS

 Strongly

Disagree

 Disagree No opinion Agree Strongly

agree
n % n % n % N % n %
GCS is reliable — — 1 03.2 7 22.6 18 58.1 5 16.1
GCS is easy to use — — — — 1 03.2 20 64.5 10 32.3
GCS is reproducible — — 2 06.5 16 51.6 19 29.0 4 12.9
GCS gives more clinical

information

 1 3.2 5 16.1 8 25.8 16 51.6 1 03.2
GCS takes less time to perform 3 9.7 1 03.2 2 06.5 16 51.6 9 29.0

The descriptive statistics of the raters’ opinion about the practical feasibility of the FOUR score is shown in table. 5. Among 31 raters, 20(64.5%) agreed to the statement that FOUR score is reliable, 20(64.5%) agreed to the statement that FOUR score is easy to use, 13(41.9%) agreed to the statement that FOUR score is reproducible, 20(64.5%)agreed to the statement that FOUR score gives more clinical information, 16(51.6%) agreed to the statement that FOUR score takes less time to perform & 13(41.9%) agreed & 10(32.3%) strongly agreed to the statement that FOUR score can be used as an alternative to GCS.

Very few among both the raters’ group were disagreeing that FOUR score can be used as an alternative to GCS. Most of the raters expressed that they find difficulty in eliciting withdrawal response to pain and often face confusion between withdrawal response to pain and flexion response to pain in GCS. Some of them verbalized difficulty in assessing the motor component of GCS in quadriplegic patients. Raters felt that the addition of brain stem reflexes in FOUR score would give them more clinical information.

Table. 5: Raters’ opinion on practical feasibility of the FOUR score

N = 31

 

Statements regarding FOUR score Strongly Disagree Disagree No opinion Agree Strongly agree
n      % n       % n        % n       % n      %
FOUR score is reliable — — — — 2 6.5 20 64.5 — —
FOUR score is easy to use — — 1 3.2 4 12.9 20 64.5 6 19.4
FOUR score is reproducible — — 3.2 12 38.7 13 41.9 5 16.1
FOUR score gives more clinical information — — — — 3 3 20 64.5 8 25.8
FOUR takes less time to

perform

 3 9.7 4 12.9 5 16.1 16 51.6 3 9.7
FOUR score can be used as an alternative to GCS 1 3.2 1 3.2 6 19.4 13 41.9 10 32.3

 

Discussion

Neurological disturbances pose a greater challenge in the critically ill patients and patients with neurological disorders. The GCS has been the gold standard for assessing the LOC in patients with significant brain injury since it was developed in 1974.2 The GCS is widely used and accepted but gives relatively limited information about brainstem function, eye opening and tracking, and respiratory patterns.13 Since its introduction in 2005, FOUR score has been refined in clinical use, and its usefulness has been confirmed by hundreds of neurosurgical patients and dozens of doctors.7 FOUR score maintains simplicity and provides far better information, particularly for intubated patients. The FOUR score is a good predictor of the prognosis of critically ill patients and has important advantages over the GCS in the ICU setting.14 The FOUR score has been developed to assess the depth of coma in a more detailed manner than the GCS.11

Studies have been conducted among different group of patients (stroke patients,15 pediatric population,2 critically ill patients11,14 and neuroscience patients6, 11, 16). This is the first study to evaluate this newly validated FOUR score in India for its application in critically ill patients. Wijdicks et al (2005) and other investigators, have conducted validation of FOUR score in different clinical settings like medical ICU, neurology ICU and emergency department of Mayo Clinic involving different group of raters including experienced and inexperienced nurses and neuro physicians.2,3,5,6 This study involved nurses and anesthetists of diverse experience and different qualification.

Study results demonstrated that the predictive value of the FOUR score for ICU mortality is slightly on the higher side than the GCS as Cohen et al (2009) found a higher predictive value for the FOUR score than the GCS.2 The study also demonstrated an excellent internal consistency for both GCS and FOUR score and good correlation between GCS and FOUR score. High degree of internal consistency for both GCS and FOUR score and good correlation between GCS and FOUR score were also elicited in 2 studies conducted by Wijdicks et al (2005)3 and Iyer VN et al (2009).14

The study conducted by Wolf et al (2007) found that the inter-rater agreement among experienced and inexperienced neuroscience ICU nurses was good to excellent with the FOUR score and the GCS6 but in this present study only fair to moderate agreement was found for the FOUR score and GCS among all 100pairs of raters and at the same time, good inter-rater agreement existed between Nurse – Anesthetist pair for the FOUR score. In other pairs (Nurse-nurse pair and anesthetist-anesthetist pair) the inter-rater agreement was fair to moderate.

Majority of the raters of this present study agreed and 10 of them strongly agreed that the FOUR score can be used as an alternative to the GCS. The FOUR score and the GCS were almost equally able to predict mortality in this population. The excellent internal consistency and good level of inter-rater agreement among Nurse – Anesthetist pair suggests that the new scale is consistent and reliable and that nurses with differing levels of experience and expertise are more likely to correctly assess the patient and assign the same score using the FOUR score. The FOUR score has the potential to become an important measure in prospective clinical studies. However, the ease of use and global acceptance raises the potential of GCS scoring in the critical care neurological assessment, provided one keeps in mind its limitations.

The raters agreed that Four score is reliable, easy to use, reproducible gives more clinical information and takes less time to perform and can be used as alternate to GCS. The study findings recommend that the Four score can replace the GCS scale since the inter-rater agreement between nurse-nurse pair and anesthetist-anesthetist pair was fair to moderate and the inter-rater agreement between nurse-anesthetist pair was good in total FOUR score. It is also recommended that the health care team need to be trained and made proficient in using the FOUR score before its implementation in the clinical settings and a similar study can be replicated with large sample size involving more number of patients of different population for a longer period of time to confirm the usability and reliability of the FOUR score.

References:

  1. Hickey The clinical practice of neurological and neurosurgical nursing. 4th ed. Philadelphia: Lippincott; 1997.134-7
  2. Cohen Interrater reliability and predictive validity of the FOUR score coma scale in a pediatric population. Journal of neuroscience nursing 2009 Oct; 41(5): 261-7
  3. Wijdicks EF, Bamlet WR, Maramattom BV, Manno EM, McClelland Validation of a new coma scale: The FOUR score. Annals of neurology

2005; 58: 585- 93

  1. Murthy A new score to validate coma in emergency department – FOUR score. Indian journal of neurotrauma. 2009; 6(1):59-62
  2. Fischer M, Rüegg S, Czaplinski A, Strohmeier M, Lehmann A, Tschan F, et Inter- rater reliability of the Full Outline of UnResponsiveness score and the Glasgow Coma Scale in critically ill patients: a prospective observational study. Critical care 2010;14(2): R64
  3. Wolf CA, Wijdicks EFM, Bamlet WR, McClelland Fur ther validation of the FOUR score coma scale by intensive care nurses. Mayo clinic proceedings. 2007 Apr; 82(4):435-8
  4. Ankavipat Endorsement of the FOUR score for consciousness assessment in neurosurgical patients. Neurolgia Medico- Chirurgica 2009 Dec; 49(12): 565-71
  5. Balestreri M et al. Predictive value of Glasgow coma scale after brain trauma: change in trend over the past ten Journal of Neurology Neurosurgery Psychiatry (2004); 75: 161, 162
  6. Rowley G, Fielding Reliability and accuracy of the Glasgow Coma Scale with experienced and inexperienced users. Lancet. 1991; 337:535 – 38
  7. SoRelle Ruth. FOUR score challenges Glasgow Coma Emergency Medicine News [online] August 2006; 28(8): 50-55. Available from: URL: http://journals.lww.com/ em-news/Fulltext/2006/08000/ FOUR_Score_Challenges_ Glasgow_Coma_Scale.2.aspx [accessed January 2012]
  8. s EFM, Bamlet WR, Maramattom BV, Manno EM, McClelland RL. Does the JFK Revised Coma Recovery Complement the FOUR Score? Annals of neurology 2006 Dec; 60(6): 745
  9. Schnakers C, Giacino J, Kalmar K, Piret S, Lopez E, Boly M, et Does the Four Score Correctly Diagnose the Vegetative and Minimally Conscious States? Annals of neurology 2006 Dec; 60(6): 746
  10. Wijdicks EFM, Kokmen E, O’Brien PC. Measurement of impaired consciousness in the neurological intensive care unit: a new test. Journal of neurology neurosurgery psychiatry. 1998; 64:117 – 19
  11. Iyer VN, Mandrekar NJ, Danielson RD, Zubkov AY, Elmer JL, Wijdicks Validity of the FOUR score coma scale in the medical intensive care. Mayo clinic proceedings 2009 August; 84(8): 694-701
  12. Idrovo L, Fuentes B, Medina J, Gabaldón L, Ruiz-Ares G, Abenza MJ, et Validation of the FOUR score (Spanish Version) in acute stroke: An interobserver variability Study. European Neurology 2010; 63: 364-69
  13. Ledoux D, Bruno M, Jonlet S, Choi P, Schnakers C, Damas F, et al. Full Outline of Unresponsiveness compared with Glasgow coma scale assessment and outcome prediction in Critical Care. 2009;13(1): 107