Title Page

The Full Outline of UnResponsiveness (FOUR) Score and its use in Outcome
Prediction: A Scoping Review of the Pediatric Literature
Alysa Almojuela, BSc, MD,1 Mohammed Hasen, MD, MSc,1,2 Frederick A. Zeiler, BSc, MD, FRCSC,3,4

1. Section of Neurosurgery, Department of Surgery, Rady Faculty of Health Sciences, University of

Manitoba, Winnipeg, Canada
2. Imam Abdulrahman Bin Faisal University, Department of Neurosurgery, King Fahad University
Hospital, Saudi Arabia
3. Clinician Investigator Program, Rady Faculty of Health Sciences, University of Manitoba, Winnipeg,
Canada
4. Division of Anaesthesia, Department of Medicine, Addenbrooke’s Hospital, University of Cambridge,
Cambridge, UK

Corresponding Author

Alysa Almojuela BSc MD

Section of Neurosurgery

Department of Surgery

Rady Faculty of Health Sciences

University of Manitoba

GB-1 820 Sherbrook Street

Winnipeg, MB R3A 1R9

Tel: 204-787-7221

Fax: 204-787-3851

Email: [email protected]

ORCID ID: 0000-0002-3343-7546

Contributing Authors

Mohammed Hasen MD MSc

Section of Neurosurgery

Department of Surgery

Rady Faculty of Health Sciences

University of Manitoba

Email: [email protected]

Frederick A. Zeiler BSc MD FRCSC

Assistant Professor

Co-Director of Research – Neurosurgery

Department of Surgery

Rady Faculty of Health Sciences

University of Manitoba

Email: [email protected]

Word count (excluding abstract, key words, figures and tables): 3365

Abstract

2
The Full Outline of Unresponsiveness (FOUR) score is a neurological assessment score. Its

benefit over pre-existing scores is its evaluation of brainstem reflexes and respiratory pattern.

Our goal was to perform a scoping systematic review of the literature on the application of the

FOUR score within pediatric patients. 6 databases were searched and 2 reviewers

independently screened the results. The initial search yielded 1709 citations; ultimately, 6

studies comprised of 571 pediatric patients were used. 4 studies examined inter-observer

reliability of the FOUR score and found it to be good to excellent. All 6 studies demonstrated

equivalency of the FOUR score and Glasgow Coma Scale (GCS) in predicting outcome. The

existing literature suggests the FOUR score is equivalent to GCS in outcome prediction in

pediatric patients; its true superiority over the GCS has not yet been established. It displays

good to excellent inter-rater reliability among physicians and nurses.

Keywords

FOUR score, prognosis, outcome prediction

Introduction

3
Clinical assessment of neurological status is a vital element in decision making, outcome

prediction and information sharing among medical professionals. Traditionally, the Glasgow

Coma Scale (GCS) has been widely adopted to document and formally assess neurological

status1. This scale has been praised for its simplicity and ease of use among healthcare workers.

However, a number of shortcomings of the GCS have been identified as well, including the fact

that it does not attempt to assess brainstem reflexes or respiratory pattern which arguably are

important clinical factors reflecting severity of impairment. The GCS can also be inaccurate in

certain patient populations, including intubated patients, dysphasic patients and those

possessing other significant language deficits, as the verbal component of the GCS becomes

difficult to apply2. Raters will make adjustments to the GCS score to account for this but these

adjustments are not standardized across institutions.

Another important group of patients that are notoriously difficult to assess with the GCS is the

pediatric population, secondary to the varying definitions of “normal” functioning according to

patient age and developmental status. This makes the standardized neurological assessment of

pediatric patients as a group difficult to accomplish. Various coma scales over the years have

been developed to describe the neurological status of an infant or child, including the Starship

Infant Neurological Assessment Tool3 and the pediatric version of the GCS4, but currently there

is no gold standard tool.

4
In 2005, Wijdicks et al5 devised a new coma score, the Full Outline of UnResponsiveness (FOUR)

score, which aimed to address the pitfalls of the GCS. The benefit that the FOUR score has over

pre-existing systems is based on its inclusion of specific categories for eyelid movement, motor

exam, brainstem reflexes and respiratory pattern; in this way, the FOUR score unlike the GCS

provides a structured scoring system for aspects of brainstem function that can be assessed in

all patients, including those unable to fully verbally communicate. Although both the FOUR

score and the GCS require some language function in order to follow simple commands, by

bypassing a formal verbal score, the FOUR score may be easier to apply in pediatric patients

with varying levels of language development and capabilities; by including a brainstem and

respiratory component, it may also offer more useful information about severity of disease.

The FOUR score, along with the GCS and its pediatric modification (for children less than 2 years

old) are illustrated in Table 1.

Since its inception, the FOUR score has been studied in a variety of settings and patient

populations. Our goal was to perform a scoping systematic review of the existing literature on

the application of the FOUR score within pediatric patients specifically and its use in outcome

prediction.

Methods

A systematic review using the methodology outlined in the Cochrane Handbook for Systematic

Reviewers6 was conducted. Data was reported following the Preferred Reporting Items for

5
Systematic Reviews and Meta-Analyses (PRISMA) guidelines 7. The search strategy was decided

upon by the primary author (AA) and supervisor (FAZ).

Search Question, Population, Inclusion and Exclusion Criteria

We aimed to answer the broad question: What literature is available for the FOUR score and

outcome prediction in critically ill pediatric patients? The primary outcome of interest was

patient global outcome, as assessed by mortality or any other functional or neuropsychiatric

outcome. Studies documenting inter-observer variability were also included in order to provide

context to the reliability of the FOUR score system.

Inclusion criteria were: humans, adults, prospective randomized controlled trial, prospective

cohort, cohort/control, case series, prospective and retrospective studies. Non-English studies

and those involving animals were excluded. Ultimately, studies on pediatric patients only were

included.

Search Strategy

6 databases were searched from inception to September 2017: MEDLINE, BIOSIS, Scopus,

Cochrane Libraries, Globalhealth and Embase. Published meeting proceedings were included in

the search. Following study selection, reference sections of each paper were examined to

ensure relevant papers not captured by the initial search were included in the review. Appendix

A of the supplementary materials highlights the search strategy implemented for each

database.

6
Study Selection

A 2-step review was performed. Two reviewers independently screened each resulting title and

abstract from the initial search for inclusion. Full texts for citations passing this initial screen

were obtained. Inclusion and exclusion criteria were applied to each article to obtain final

articles for review. No conflict in article selection occurred between the two reviewers.

Data Collection

Data was extracted from the final list of articles and stored electronically. Data extracted

included study country, design, demographics (including number of patients and ages of

children involved), objectives, outcomes and conclusions made by the study authors.

Quality of Evidence Assessment

Each study was evaluated for quality of evidence using the RTI Item Bank on Risk of Bias and

Precision of Observational Studies8. This validated item bank is applicable to a variety of

observational study designs and evaluates the risk of bias and internal validity of studies using a

comprehensive list of itemized questions. Appendix B of the supplementary materials provides

the tabulated results of the bias assessment for each study included in this scoping review.

Statistical Analysis

A meta-analysis was not performed due to the heterogeneity of data and study design within

the studies, as well as the inclusion of only 6 papers relevant to the present systematic review.

7
Results

The initial search yielded 1709 citations. Of 55 articles selected for final review, 6 were based

on pediatric populations. 2 studies were done on patients presenting to the emergency

department (including 1 on traumatic brain injury patients and 1 on non-trauma patients) and 4

studies were done in an intensive care unit setting. Of these, 2 were performed in a

neurosciences intensive care unit, 1 in a combined intensive care unit, and 1 study included

general pediatric patients from both an intensive care unit and pediatric ward. All 6 articles

were performed prospectively. A total of 571 pediatric patients were studied. Figure 1 displays

the PRISMA flow diagram of the search results and filtering processes.

Quality of evidence

Quality of evidence was assessed using the RTI Item Bank on Risk of Bias and Precision of

Observational Studies8. Based on its itemized list of questions, there was an overall low risk of

bias in the studies included in this review.

Inter-observer reliability of the FOUR Score

4 of the 6 studies chose to examine inter-observer reliability of the FOUR score among

physician or nurse raters. Table 2 demonstrates this tabulated data. In general, a kappa value of

0.4 or less is considered poor, values of 0.4-0.6 are considered fair to moderate, values of 0.6-

0.8 are considered good, and values above 0.8 are considered to have excellent inter-rater

8
agreement. All values in the literature were good to excellent, with the highest score being a

weighted kappa value of 0.95 (95% CI 0.91-0.99) as demonstrated by Cohen et al 9, who also

demonstrated the weighted kappa value for GCS to be 0.74 (95% CI 0.59-0.87) in comparison.

While Czaikowski et al 10 used a slightly modified version of the FOUR score to allow for grading

of children and infants less than 2 years old, they also demonstrated an excellent kappa value

of 0.89 (95% CI 0.83-0.94) for the FOUR score versus GCS (k=0.89, 95% CI 0.84-0.94).

Khajeh et al11 documented the kappa values for each subcategory for the FOUR score; the

lowest kappa value was for the eye subscore (k=0.72, 95% CI 0.67-0.77), followed by the

brainstem subscore (k=0.74, 95% CI 0.69-0.80), the motor subscore (k=0.78, 95% CI 0.73-0.84)

and finally the respiration subscore (k=0.82, 95% CI 0.77-0.87). Rather than using weighted

kappa values to assess inter-observer reliability, Jamal et al 12 used interclass correlation

coefficient, which was also good for the FOUR score (ICC=0.98 for FOUR, 95% CI 0.964-0.992

versus ICC=0.96 for GCS, 95% CI 0.970-0.985). Thus, no studies that chose to examine inter-

observer reliability failed to demonstrate at least good reliability. Appendix C of the

supplementary materials provides the tabulated results from those studies assessing inter-

observer reliability.

Prognostic value of the FOUR Score compared to GCS

2 studies were conducted on patients in an emergency department setting; Table 3 outlines the

results of these studies. Buyukcam et al13 focused on traumatic brain injury patients only and

found that both the FOUR score and GCS were similar in predicting in-hospital mortality (FOUR

9
AUC=0.975, 95% CI 0.931-1.019 versus GCS AUC=0.965, 95% CI 0.909-1.020), hospitalization

more than three days (FOUR AUC=0.716, 95% CI 0.595-0.837 versus GCS AUC=0.726, 95% CI

0.607-0.845) and Glasgow Outcome Scale of 1-3 at discharge (FOUR AUC=0.870, 95% CI 0.746-

0.993 versus GCS AUC=0.884, 95% CI 0.783-0.986) and at 3 months (FOUR AUC=0.975, 95% CI

0.931-1.019 versus GCS AUC=0.965, 95% CI 0.909-1.021) post-injury. Jamal et al 12 focused on

non-traumatic injury patients instead, but also found similar prognostic values for the FOUR

score compared to GCS in predicting in-hospital mortality (FOUR AUC=0.80, 95% CI 0.62-0.87

versus GCS AUC=0.83, 95% CI 0.70-0.90), mortality at 3 months (FOUR AUC=0.74, 95% CI 0.62-

0.87 versus GCS AUC=0.78, 95% CI 0.67-0.90) and poor outcome (score of 4-5 on the Pediatric

Overall Performance Category Scale) at 3 months (FOUR AUC=0.79, 95% CI 0.68-0.90 versus

GCS AUC=0.82, 95% CI 0.72-0.93).

4 studies included pediatric patients in an intensive care unit; Table 4 outlines the results of

these studies. 2 of the 4 studies were conducted strictly in a neurosciences intensive care unit.

This included Cohen et al9, who found comparable prognostic values between the FOUR score

and GCS in predicting both in-hospital mortality (FOUR AUC=0.81, 95% CI 0.69-0.90 versus GCS

AUC=0.77, 95% CI 0.64-0.87) and poor outcome (Modified Rankin Scale of 3-6) at discharge

(FOUR AUC=0.78, 95% CI 0.65-0.88 versus GCS AUC=0.76, 95% CI 0.64-0.86); and Khajeh et al 11,

who demonstrated the odds ratio for the FOUR score in predicting in-hospital mortality to be

0.13 (95% CI 0.06-0.29, p<0.001) versus 2.49 (95% CI 1.44-4.32, p<0.001) for GCS - citing that

lower odds ratios are related to a positive predictive value for a higher chance of a positive

outcome.

10
Kochar et al14 examined use of the FOUR score in nontraumatic pediatric patients with

impairment of consciousness in either a pediatric intensive care unit or ward. The authors

found no significant difference in the ability of FOUR versus GCS to predict in-hospital mortality

(FOUR AUC=0.940, 95% CI 0.889-0.990 versus GCS AUC=0.916, 95% CI 0.855-0.978) and poor

outcome (Pediatric Overall Performance Category Scale score 3-6) at discharge (FOUR

AUC=0.745, 95% 0.584-0.908 versus GCS AUC=0.732, 95% CI 0.568-0.896).

Czaikowski et al10 was the only group in the literature to modify the FOUR score as originally

devised by Wijdicks et al5 in order to theoretically allow more meaningful application to a

pediatric population less than 2 years old, as they were the only study to include patients this

young. The modifications however were quite minor and consisted of changing the motor M4

score to include “age appropriate spontaneous movement without stimulation”, adding

“eyelids open and tracking” to the eye E4 score, and modifying the respiratory score to reflect

the wide variety of age-appropriate breathing patterns. With these modifications, the authors

too showed similar prognostic values for in-hospital mortality (FOUR AUC=0.93 versus GCS

AUC=0.91, p=0.2113) and poor outcome (Pediatric Cerebral Performance Category score 4-6) at

discharge (FOUR AUC=0.90 versus GCS AUC=0.91, p=0.9552) in an intensive care unit that

included neurological, cardiac, trauma and post-operative pediatric patients.

Discussion

11
We aimed to perform a scoping review of the FOUR score and its use in outcome prediction in

pediatric populations. The existing literature around the FOUR score in pediatric populations,

while scarce, does demonstrate that it potentially possesses similar prognostic value in

comparison to the GCS in predicting mortality and functional outcomes in a variety of pediatric

patients with depressed level of consciousness, including those suffering from both traumatic

and non-traumatic etiologies, and those from infancy to 18 years of age. The literature also

supports good to excellent inter-rater reliability for this score when applied to pediatric

populations by ICU-trained and pediatrics-trained physicians and nurses.

The accurate neurological assessment of pediatric patients using the GCS - while done routinely

at many institutions - is often inaccurate, due to difficulties assessing the verbal component as

well as the presence of variable age-related and developmentally-related milestones 15. In

patients with severe neurological injury, the occasionally necessary use of sedating medications

and endotracheal intubation compounds this problem.

The FOUR score bypasses this dilemma by not including a verbal score based on orientation or

the ability of a patient to communicate. Instead, its core subcategories of eye, motor,

brainstem and respiratory function can be accurately quantified in a great number of patients,

including those that are intubated and those with new or baseline language inabilities. In

comparison to the GCS, the FOUR score is also helpful in further subcategorizing patients with

severe neurological impairment based on their brainstem function and respiratory pattern,

which the GCS is unable to do. These advantages, combined with its good interrater reliability,

12
give the FOUR score the potential to replace conventional scoring systems in the future and

allow for precise and consistent neurological assessments among pediatric health care

providers.

Limitations

Despite the promising results surrounding the application of the FOUR score in pediatric

populations, there are some limitations which deserve highlighting.

First, a limitation of this review is the small number of studies done thus far. This fact,

combined with the assortment of patient populations and selected outcomes of interest across

studies, makes a meaningful meta-analysis of results is impossible to conduct. There is also no

gold standard tool for grading pediatric neuro-functional outcome. Thus, we are left with a

descriptive analysis of the available literature, consisting of only 6 studies. With that said, these

6 studies include over 500 pediatric patients and do provide some support for its use in clinical

grading and outcome prediction. However, the FOUR score was shown to only be similar - and

not superior - to the GCS in this regard. True superiority rather than equivalence only of the

FOUR score to the GCS may be demonstrated by studies done with larger sample sizes in the

future. In contrast to the 6 studies included here, there are 49 studies that examine the FOUR

score in the adult literature, as illustrated in Figure 1. A scoping systematic review of this

population is currently being undertaken by the authors. Preliminary results have

demonstrated the FOUR score’s usefulness as an outcome predictor in adult patients with

depressed level of consciousness and its good inter-rater reliability, but results are not reported

13
here and are awaiting publication at this time. Though derived from a substantially smaller

body of literature, results from the pediatric population appear to mirror these results, lending

consistency to our conclusions. It must be acknowledged, however, that despite the promising

results for the application of the FOUR score in the pediatric population displayed in this

review, there exists the need for further large prospectively conducted studies on the

application of this clinical grading system in various subpopulations of pediatric patients, based

on age and pathology.

Second, the FOUR score is limited by the fact that it requires a more detailed neurological

examination and the experience to confidently conduct such an examination. The saving grace

of more simplistic systems, such as the GCS, is they can be readily employed by various medical

and paramedical professionals with fairly consistent reliability. This flexibility of the GCS in the

face of varied training and backgrounds of the assessor is a major benefit of the system. The

FOUR score requires a slightly higher background knowledge of the nervous system, which may

limit its application in other settings, such as the pre-hospital environment. Despite this

theoretical disadvantage, good inter-rater reliability of the FOUR score was still demonstrated

when executed not only by ICU-trained nurses but resident pediatric doctors, both groups

which may not have the same experience as health professionals with specific neurology or

neurosurgical training.

Finally, the majority of the literature identified within this review focuses on larger groups of

pediatric patients. Conclusions regarding the FOUR score in specific sub-populations of

14
patients, such as those with acute hydrocephalus, is quite limited at this time. In addition, the

FOUR score as it stands would be difficult to apply to pediatric patients less than 2 years old, as

a full motor score requires the ability to follow commands. However, making a minor

adjustment to the score to include an age appropriate motor response - as illustrated by

Czaikowski et al10 - may allow this score to be used appropriately in this very young population,

similar to the pediatric GCS. More studies are required to compare the modified versions of the

FOUR score and GCS for children less than 2 years old, and further work in sub-populations in

general is required to accurately determine the potential role of the FOUR score.

Conclusions

The existing literature demonstrates that the FOUR score is equivalent to GCS in outcome

prediction in pediatric patients with depressed level of consciousness, although its true

superiority over the GCS has not yet been established. It displays good to excellent inter-rater

reliability among physicians and nurses.

Acknowledgements and Funding

FAZ has received salary support for dedicated research time, during which this manuscript was

completed.  Such salary support came from:  the Cambridge Commonwealth Trust Scholarship,

the Royal College of Surgeons of Canada – Harry S. Morton Travelling Fellowship in Surgery and

15
the University of Manitoba Clinician Investigator Program. FAZ’s research is also supported

through the Thorlakson Chair in Surgical Research Establishment Grant

Author contributions

AA contributed to the conception and design of this work, the acquisition, analysis and

interpretation of data, the initial drafting of the manuscript, the critical revision of the work and

the final approval of the manuscript to be published.

MH contributed to the interpretation of data, the critical revision of intellectual content and the

final approval of the manuscript to be published.

FAZ contributed to the conception and design of the work, the interpretation of data, the

critical revision of the work and the final approval of the manuscript to be published.

Declaration of conflicting interests

The authors have no conflicts of interest to disclose.

Ethical approval

No ethics approval was required for this study.

16
References

1. Teasdale G, Maas A, Lecky F, et al. The Glasgow Coma Scale at 40 years: standing the
test of time. Lancet Neurol. 2014;13(8):844-854.
2. Braine ME, Cook N. The Glasgow Coma Scale and evidence-informed practice: a critical
review of where we are and where we need to be. J Clin Nurs. 2017;26(1-2):280-293.
3. Birse J. Pilot Testing of the Starship Infant Neurological Assessment Tool. J Neurosci
Nurs. 2006;38(4):206-211.
4. Holmes JF, Palchak MJ, MacFarlane T, Kuppermann N. Performance of the pediatric
glasgow coma scale in children with blunt head trauma. Acad Emerg Med.
2005;12(9):814-819.
5. Wijdicks EF, Bamlet WR, Maramattom BV, et al. Validation of a new coma scale: The
FOUR score. Ann Neurol. 2005;58(4):585-593.
6. Higgins J, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version
5.1.0. 2011.
7. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic
reviews and meta-analyses of studies that evaluate health care interventions:
explanation and elaboration. PLoS Med. 2009;6(7):e1000100.
8. Viswanathan M, Berkman N. Development of the RTI Item Bank on Risk of Bias and
Precision of Observational Studies. Rockville, MD: Agency for Healthcare Research and
Quality;2011.
9. Cohen J. Interrater reliability and predictive validity of the FOUR score coma scale in a
pediatric population. J Neurosci Nurs. 2009;41(5):261-267.
10. Czaikowski BL, Liang H, Stewart CT. A pediatric four score coma scale: Interrater
reliability and predictive validity. J Neurosci Nurs. 2014;46(2):79-87.
11. Khajeh A, Fayyazi A, Miri-Aliabad G, et al. Comparison between the ability of glasgow
coma scale and full outline of unresponsiveness score to predict the mortality and
discharge rate of pediatric intensive care unit patients. Iran Journal of Pediatr.
2014;24(5):603-608.
12. Jamal A, Sankhyan N, Jayashree M, et al. Full Outline of Unresponsiveness score and the
Glasgow Coma Scale in prediction of pediatric coma. World J Emerg Med. 2017;8(1):55-
60.

17
13. Buyukcam F, Kaya U, Karakilic ME, et al. Predicting the outcome in children with head
trauma: Comparison of FOUR score and Glasgow Coma Scale. Ulusal Travma ve Acil
Cerrahi Dergisi. 2012;18(6):469-473.
14. Kochar GS, Gulati S, Lodha R, Pandey R. Full outline of unresponsiveness score versus
Glasgow Coma Scale in children with nontraumatic impairment of consciousness. J Child
Neurol. 2014;29(10):1299-1304.
15. Ghaffarpasand F, Razmkon A, Dehghankhalili M. Glasgow Coma Scale Score in Pediatric
Patients with Traumatic Brain Injury; Limitations and Reliability. Bull Emerg Trauma.
2013;1(4):135-136.
16. Almojuela A, Hasen M, Zeiler FA. The Full Outline of UnResponsiveness (FOUR) Score
and its Use in Outcome Prediction: A Scoping Systematic Review of the Adult Literature.
J Neurocrit Care (forthcoming).

18
Figure Legends

Figure 1: Diagram of Study Selection

19
Tables

Table 1: Neurological Grading Scales

Full Outline of UnResponsiveness Score Glasgow Coma Scale Pediatric Glasgow Coma Scale (age <2 years old)
Eye response Eye opening Eye opening
E4 Eyelids open or opened, tracking or E4 Spontaneous E4 Spontaneous
blinking to command
E3 Eyelids open but not tracking E3 To verbal command E3 To speech
E2 Eyelids closed but open to loud voice E2 To pain E2 To pain
E1 Eyelids closed but open to pain E1 None E1 None
E0 Eyelids remain closed with pain Verbal response Verbal response
Motor response V5 Oriented V5 Coos, babbles
M4 Thumbs-up, fist or peace sign V4 Confused V4 Irritable, cries
M3 Localizing to pain V3 Inappropriate words V3 Cries to pain
M2 Flexion response to pain V2 Incomprehensible sounds V2 Moans to pain
M1 Extension to pain V1 None V1 None
M0 No response to pain or generalized Motor response Motor response
myoclonus status
Brainstem reflexes M6 Follows commands M6 Normal spontaneous movement
B4 Pupil and corneal reflexes present M5 Localizes pain M5 Withdraws to touch
B3 One pupil wide and fixed M4 Withdraws from pain M4 Withdraws to pain
B2 Pupil or corneal reflexes absent M3 Flexion to pain M3 Abnormal flexion
B1 Pupil and corneal reflexes absent M2 Extension to pain M2 Abnormal extension
B0 Absent pupil, corneal and cough M1 None M1 None
reflex
Respiration
R4 Not intubated, regular breathing
pattern
R3 Not intubated, Cheyne-Stokes
breathing
R2 Not intubated, irregular breathing
R1 Breathes above ventilator rate
R0 Breathes at ventilator rate or apnea

20
Table 2: Studies examining inter-observer reliability of the FOUR score

Citation Design Demographics Raters Patient Pathology Inter-observer reliability assessment

Patients in the Emergency Department
Jamal 2017 Prospective Country: India Resident Neuro-infection (n=34) Interclass correlation coefficient for
observational pediatric Non-infectious (n=29) GCS was 0.96 (95% CI 0.970-0.985);
Full Outline of Sample size: doctors  3 leading non-infectious 0.98 for FOUR (95% CI 0.964-0.992)
Unresponsiveness 63 patients causes were epilepsy with
score and the seizure recurrence (n=7),
glasgow coma hepatic encephalopathy
Age: 5-12 (n=5), and
scale in years
prediction of intoxication/envenomation
pediatric coma (n=4)
Patients in a neurological sciences intensive care unit
Cohen 2009 Prospective Country: Pediatric Brain tumor (n=25) Interrater reliability for GCS was good
observational United States critical Hydrocephalus (n=7) (k=0.74, 95% CI 0.59-0.87); reliability
Interrater care Traumatic brain injury (n=5) for FOUR was excellent (k=0.95, 95%
Reliability and Sample size: nurses Spinal surgery (n=5) CI 0.91-0.99)
Predictive Validity 60 patients Seizure (n=4)
of the FOUR Arteriovenous malformation
Score Coma Scale (n=4) Cranial remodeling (n=3)
Age: 2-18 Moyamoya disease (n=2)
in a Pediatric years
Population Encephalitis (n=2)
Subdural hematoma (n=2)
Near drowning (n=1)
Khajeh 2014 Prospective Country: Iran Nurses Intracranial hemorrhage (n=36) Inter-rater reliability of FOUR was
observational Intracranial infection (n=31) good to excellent; eye k=0.72 (95% CI
Comparison Sample size: Hydrocephalus (n=29) 0.67-0.77), motor k =0.78 (95% CI
between the 200 patients Aneurysm (n=28) 0.73-0.84), brainstem k=0.74 (95% CI
Ability of Glasgow Seizure (n=27) 0.69-0.80), respiration k=0.82 (95% CI
Coma Scale and Brain tumor (n=22) 0.77-0.87)
Age: 2-12 Other (n=27)
Full Outline of years
Unresponsiveness
Score to Predict
the Mortality and
Discharge Rate of
Pediatric
Intensive Care
Unit Patients
Patients in a combined intensive care unit or with combined neuro and non-neuro pathology
Czaikowski 2014 Prospective Country: Pediatric Pathology not specified, Pediatric FOUR is excellent for
observational United States ICU however authors state that interrater reliability (k=0.89 for GCS,
A Pediatric FOUR nurses there were a variety of patients: 95% CI 0.84-0.94; k=0.89 for FOUR,
Score Coma Sample size: cardiac, neurological, 95% CI 0.83-0.94; k=0.67 for RASS,
Scale: Interrater 78 patients postoperative and trauma 95% CI 0.53-0.80)
Reliability and patients who were admitted to a
Predictive Validity multidisciplinary pediatric ICU.
Age: not
specified,
“infant” to 53.9% were in the “alert” group
“adolescent” 11.3% were in the “clouding of

21
 (13+ years) consciousness” group
23.5% were in the
“obtundation” group
11.3% were in the
“stupor/coma” group

Table 3: Studies examining patients in the emergency department

Citation Design Demographics Patient Pathology Objective Outcomes Conclusion

Studies on traumatic brain injury patients
Buyukcam Prospective Country: Turkey The majority of Investigate whether - In-hospital - AUC ROC for predicting in-hospital
2012 observational Sample size: patients (>45% of FOUR score is better mortality mortality was 0.965 for GCS (95%
100 patients patients) had a than GCS in - Hospitalization CI 0.909-1.020, p=0.0001) and
Predicting Age: 2-17 years minor traumatic predicting morbidity > 3 days 0.975 for FOUR (95% CI 0.931-
the brain injury (GCS and mortality in - GOS (Glasgow 1.019, p=0.0001),
outcome in 13-15) children with head Outcome - AUC ROC for predicting
children trauma. Scale) at hospitalization >3 days was 0.726
with head Approximately discharge and for GCS (95% CI .607-0.845) and
trauma: 40% of patients at 3 months 0.716 for FOUR (95% CI 0.595-
comparison had a moderate 0.837)
of FOUR traumatic brain - AUC ROC for predicting GOS 1-3 at
score and injury (GCS 9-12) discharge was 0.884 for GCS (95%
GCS CI 0.783-0.986) and 0.870 for
FOUR (95% CI 0.746-0.993); at 3
months AUC ROC was 0.965 for
GCS (95% CI 0.909-1.021) and
0.975 for FOUR (95% CI 0.931-
1.019)
- Predictive value for the above
were similar for GCS and FOUR
- FOUR score provides no significant
advantage over GCS in predicting
morbidity and mortality in
children with head trauma
Studies on non-traumatic injury patients
Jamal 2017 Prospective Country: India Neuro-infection Compare the FOUR - In-hospital - AUC ROC for in-hospital mortality
observational Sample size: 63 (n=34) score and GCS as mortality for GCS was 0.83 (95% CI 0.7-0.9);
Full Outline patients Non-infectious predictors of - Mortality at 3 FOUR was 0.8 (95% CI 0.7-0.9)
of Age: 5-12 years (n=29) outcome in children months - AUC ROC for 3 month mortality
Unresponsi 3 leading non- presenting to the ER - Poor outcome for GCS was 0.78 (95% CI 0.67-
veness infectious causes with non traumatic on pediatric 0.90); 0.74 for FOUR (95% CI 0.62-
score and were epilepsy impaired overall 0.87)
the with seizure consciousness of <7 performance - AUC ROC for poor functional
glasgow recurrence (n=7), days. category scale outcome at 3 months was 0.82 for
coma scale hepatic at 3 months (a GCS (95% CI 0.72-0.93); 0.79 for
in encephalopathy score of 4, 5 or FOUR (95% CI 0.68-0.90).
prediction (n=5), and death) - Interclass correlation coefficient
of pediatric intoxication/enve - Interrater for GCS was 0.96 (95% CI 0.970-
coma nomation (n=4) reliability of 0.985); 0.98 for FOUR (95% CI
GCS and FOUR 0.964-0.992)
- FOUR score was as good as GCS in
predicting in-hospital mortality
and mortality and functional
outcome at 3 months.
- FOUR score has a good inter-rater

22
 reliability

Table 4: Studies examining patients in an intensive care unit

Citation Design Demographics Patient Pathology Objective Outcomes Conclusion
Neurosciences intensive care unit only
Cohen 2009 Prospective Country: Brain tumor (n=25) Compare interrater - Interrater - Interrater reliability for GCS was
observational United States Hydrocephalus reliability and reliability of good (k=0.74, 95% CI 0.59-0.87);
Sample size: (n=7) predictive validity GCS and reliability for FOUR was excellent
Interrater
60 patients Traumatic brain of the FOUR score FOUR (k=0.95, 95% CI 0.91-0.99)
Reliability and
Age: 2-18 injury (n=5) and GCS in - In-hospital - For in-hospital mortality, AUC ROC
Predictive Validity
years Spinal surgery (n=5) pediatric patients mortality for FOUR was 0.81 (95% CI 0.69-
of the FOUR Score
Seizure (n=4) in a neuroscience - Patient 0.90); 0.77 for GCS (95% CI 0.64-
Coma Scale in a
Arteriovenous ICU. morbidity at 0.87)
Pediatric
malformation (n=4) discharge - For poor outcome at end of
Population
Cranial remodeling using hospitalization, AUC ROC was 0.78
(n=3) Modified for FOUR (95% CI 0.65-0.88); 0.76
Moyamoya disease Rankin Scale for GCS (95% CI 0.64-0.86)
(n=2) (mRS 3-6) - Results are consistent with adult
Encephalitis (n=2) studies which suggest that the
Subdural hematoma FOUR score is a reliable and valid
(n=2) tool for use in a wide variety of
Near drowning neuroscience patients
(n=1) - Interrater reliability using FOUR
score was better than with GCS
Khajeh 2014 Prospective Country: Iran Intracranial Determine the - In-hospital - Odds ratios for FOUR score are
observational Sample size: hemorrhage (n=36) ability to predict mortality somewhat lower than for GCS for
200 patients Intracranial mortality and - Inter-rater predicting in-hospital mortality
Comparison
Age: 2-12 infection (n=31) discharge rate of reliability of (FOUR OR=0.13, 95% CI 0.06-0.29,
between the
years Hydrocephalus patients with FOUR p<0.001 and GCS OR=2.49, 95% CI
Ability of Glasgow
(n=29) neurological or 1.44-4.32, p<0.001)
Coma Scale and
Aneurysm (n=28) neurosurgical - Inter-rater reliability of FOUR was
Full Outline of
Seizure (n=27) disorders in the good to excellent; eye k=0.72 (95%
Unresponsiveness
Brain tumor (n=22) ICU using the FOUR CI 0.67-0.77), motor k =0.78 (95% CI
Score to Predict
Other (n=27) score and compare 0.73-0.84), brainstem k=0.74 (95%
the Mortality and
results with those CI 0.69-0.80), respiration k=0.82
Discharge Rate of
of GCS. (95% CI 0.77-0.87)
Pediatric Intensive
- FOUR score is more capable than
Care Unit Patients
GCS in predicting the mortality and
discharge of patients admitted to
the pediatric ICU
Combined unit and/or combined neuro and non-neuro pathology
Czaikowski 2014 Prospective Country: Pathology not To determine - Interrater - Pediatric FOUR is excellent for
observational United States specified, however whether the agreement interrater reliability (k=0.89 for GCS,
Sample size: authors state that Pediatric FOUR - In hospital 95% CI 0.84-0.94; k=0.89 for PFSS,
A Pediatric FOUR
78 patients there were a variety Score Scale (PFSS) mortality 95% CI 0.83-0.94; k=0.67 for RASS,
Score Coma Scale:
Age: not of patients: cardiac, could enhance the - Poor 95% CI 0.53-0.80)
Interrater
specified, neurological, clinical outcome or - Excellent for prediction of poor
Reliability and
“infant” to postoperative and neurological PCPC outcome (AUC ROC for FOUR was
Predictive Validity
“adolescent” trauma patients assessment of (Pediatric 0.9043; 0.9054 for GCS; p=0.9552)
(13+ years) who were admitted pediatric intensive Cerebral - Excellent for prediction of in-
* included to a care patients, Performance hospital mortality (AUC ROC for
intubated and multidisciplinary including those Category) FOUR was 0.9296; 0.9095 for GCS;
sedated pediatric ICU. intubated/sedated. score 4-6 at p=0.2113)
patients *included neuro, discharge - No statistically significant difference
cardiac, between Pediatric FOUR and GCS
53.9% were in the
postoperative and - Pediatric FOUR is excellent for
“alert” group
trauma patients interrater reliability and for
11.3% were in the

23
 “clouding of * made prediction of poor outcome and in-
consciousness” adjustments to the hospital mortality in a pediatric
group adult FOUR score population, but the study failed to
23.5% were in the show that Pediatric FOUR is better
“obtundation” than GCS
group
11.3% were in the
“stupor/coma”
group
Kochar 2014 Prospective Country: India CNS infections Compare the FOUR - In-hospital - In predicting death, AUC ROC for
observational Sample size: (n=38) score with GCS as a mortality GCS was 0.916 (95% CI 0.855-
70 patients Noninfectious predictor of - Functional 0.978) and 0.940 for FOUR (95% CI
Full Outline of
Age: 5-18 causes (n=32) mortality and poor outcome at 0.889-0.990)
UnResponsiveness
years Hepatic failure (n=9) functional outcome discharge as - In predicting poor functional
Score Versus
Intracranial bleed at hospital assessed by outcome at discharge, AUC ROC for
Glasgow Coma
(n=8) discharge in POPC GCS was 0.732 (95% CI 0.568-0.896)
Scale in Children
Post cardiac surgery children with non- (Pediatric and 0.745 for FOUR (95% CI 0.584-
with
hypoxic-ischemic traumatic Overall 0.908)
Nontraumatic
encephalopathy impairment of Performance - Both GCS and FOUR score are good
Impairment of
(n=4) consciousness Category) predictors for in-hospital mortality
Consciousness
Other/unknown *included patients scale – poor and functional outcome, but there
(n=4) admitted to both a outcome is a is no significant difference between
Hypertensive pediatric ward or score of 3-6 the ability of the 2 scores to predict
encephalopathy ICU outcome
(n=3)
Uremic
encephalopathy
(n=3)
Diabetic
ketoacidosis (n=1)
FOUR = full outline of unresponsiveness, GCS = glasgow coma scale, ICU = intensive care unit, CI = confidence interval, k = weighted kappa
score, OR = odds ratio, AUC = area under the curve, ROC = receiver operating characteristic curve

24