Objective: To assess the relationship between Modified Glasgow Coma Scale (MGCS), its components and survival in children with acute coma. Design: Prospective observational study. Setting: Tertiary care referral hospital. Patients: Consecutive children (n = 270) with acute nontraumatic coma between 2 months to 12 years. Methodology: MGCS and brainstem reflexes were assessed at 6 hourly intervals for 72 hours from the time of admission. The lowest score of the MGCS and worst brain stem reflexes were used for the analysis. Main outcome measure: Survival. Results: Total score (Spearman rank correlation coefficient IRI = O.577, ocular response (IRI = O.641), motor response (IRI = O.729), verbal response (lRI = 0.608), brain stem response (lRI = O.843) were all found to be associated with adverse outcome. Multivariate regression analysis revealed that ocular response and motor response were individually predictive of short-term outcome. Conclusion: A score incorporating the brain stem reflexes, ocular response and motor response in the assessment and prognostication of comatose patients needs to be evaluated.

Key words: Coma, Glasgow coma scale, Outcome.

Acute nontraumatic coma is a common problem in pediatric practice accounting for 10 -15% of all hospital admissions(1) and is associated with significant mortality. Assessment of the severity of coma is essential to comment on the likelihood of survival in comatose children. In the last three decades various scores have been used to assess the severity of coma and to predict its outcome. These include the Glasgow Coma Scale (GCS)(2), James Adaptation of Glasgow Coma Scale(3), the Simpson and Reilly Scale(4), Children’s Coma Scale(5), the Children’s Orthopedic Hospital and Medical Center Scale(6), the Jacobi’s Scale(7), etc. Among these, the modified Glasgow Coma Scale (MGCS)(8), in spite of its various drawbacks, has been widely used for assessing pediatric coma, though, only very few studies are available to support its use in pediatric coma as a whole(9). This study was conducted to assess the relationship between MGCS, its components and survival in children with acute coma.

This prospective cross sectional study was conducted in a tertiary care referral hospital in the Union Territory of Pondicherry, India from October 1998 to March 2000. All children between 3 months and 12 years of age admitted with coma of less than 7 days duration were included in this study. Coma was defined as the ‘unintentional failure of the patients to open their eyes spontaneously or in response to noise, inability to obey commands or localize painful stimulus with or without the ability to express comprehensible words or age appropriate language responses’(10). Children with neurodevelopmental delay, any other pre-existing neurological illness and those in whom the coma was secondary to trauma were excluded from the study. Standard case definitions were used to identify the etiology of coma (Table I). The subjects underwent focused neurological examination (MGCS and brain stem reflexes -oculocephalic, oculovestibular, pupillary reactivity) at 6 hourly intervals from the time of admission. The children were assessed for 72 hours after admission and the findings were recorded in a proforma. The lowest score of MGCS and worst brain stem reflexes score (based on the number of reflexes present) were used for analysis. In case of children who were intubated or who developed respiratory failure secondary to neurological etiology, the worst score before intubation respiratory failure was taken. Preformed standardized protocols based on current guidelines were used in the management of these children.

Pyogenic meningitis(19)	Acute febrile encephalopathy with culture of compatible microorganisms
			from the CSF or the presence of 2 or more of the following abnormalities 
			in CSF – (i) polymorphonuclear leukocytosis (ii) glucose ~ 40 mg/dL or 
			50% of blood sugar (iii) micro-organisms seen by gram staining 

Tuberculous meningitis	Based on the criteria by Ahuja, et al.(20)

Viral meningoencephalitis	Acute febrile encephalopathy with CSF pleocytosis with lymphocyte 
			predominance (>5 cells/mm3) and absence of bacteria on direct 
			microscopy culture with no other alternative diagnosis identifiable.

Enteric encephalopathy	Encephalopathy with definitive investigational evidence (positive blood 
			or stool culture for Salmonella or Shigella or positive serology for 
			Salmonella) of infection with Salmonella or Shigella.

Hypertensive encephalopathy	Acute coma in association with blood pressure more than 95th percentile 
			for age and sex with or without retinal changes.

Hypoxic-ischemic encephalopathy Encephalopathy following hypoxic cerebral injury such as near drowning, 
			   accidental or homicidal hanging.

Toxic encephalopathy		Encephalopathy following ingestion of toxin containing substances (neem 
			oil, kerosene, organophosphorus).

Hepatic encephalopathy(21)	Spectrum of potentially reversible neuropsychiatric abnormalities seen in 
			patients with liver dysfunction after exclusion of unrelated neurologic 
			and metabolic abnormality.

Dyselectrolytemia		Acute afebrile encephalopathy with abnormal serum electrolyte levels 
			(sodium) with normal CSF examination with no other alternative 
			diagnosis identifiable.

Data from standardized study forms were entered into a computer database for analysis using SPSS software. Spearman rank analysis was used to form the correlation matrix between the various clinical parameters to identify the significant relationships among them. Multivariate linear regression analysis was then performed to identify the various factors, which individually affected the immediate outcome.

Two hundred and seventy children (mean age: 5 yr 3 mo) were admitted with acute nontraumatic coma during the study period. The characteristics of the study population are given in Table II. Intracranial Infection (viral encephalitis, pyogenic meningitis and tuber-culous meningitis) was the most common cause of coma in this study (n = 218, 80.7%) with viral encephalitis (n = 115, 41.5%) forming the largest group. It is interesting to note that 12 children with tuberculous meningitis presented with acute coma. Metabolic disturbances (hepatic encephalo-pathy, dyselectrolytemia) contri-buted to 9.3% (n = 25) of comatose children. Rest of the cases included encephalopathies: i.e., enteric (n = 10), toxin induced (n = 8), hypoxic-ischemic (n = 5) and hypertensive (n = 4). The mortality was highest in the metabolic group (48%). The mortality in intracranial infections and encephalopathies was 36.6% and 30.5%, respectively.

Total number of children

270

  Male

142

  Female 

128

  Male 

50

  Female 

50

  3 -36 months 

100

  37 -72 months 

74 

  73-108 months 

54 

  109 -144 months 

42

  Mild (13 -15) 

15

  Moderate (9 -12) 

107 

  Severe (<8) 

148

Age, sex, etiology, total MGCS score, score of the individual components and brain stem reflexes, which were thought to be having an association with survival in acute coma were included in the statistical analysis. Spearman rank analysis was performed on these parameters to identify the various significant correlations. Of the possible 72 (9 × 8) correlations between the variables, 30 Spearman rank correlation IRI values were significant. Among these, the correlation coefficients relevant to the present study are given in Table III. It can be noted that total MGCS score, its individual components and the brain stem reflexes had a significant correlation with immediate outcome. Odds ratio was calculated by choosing cutoff levels, which maximized the ratios for each of the factors that were shown to be significant by Spearman’s rank analysis Table IV.

Parameter

Cutoff value

Fatal : Nonfatal

Odds ratio (95% CI)

Realtive risk

Total score

<8

≥ 8

5:137

78.9
(27.9 - 240.2)

27.1

Ocular response

≤ 2

>2

7:117

29.3
(12.1 - 74.4)

11.3

Motor response

1

>

65:169

91.0
(13 - 1822.8)

3.5

Verbal response

≤ 2

> 2

4:92

28.3
(9.5 - 9.5 - 94.9)

13.2

Brain stem responses

≤ 1

>1

17:167

271.8
(71.7 - 1214)

10.5

In order to identify the factors that independently predicted the outcome, multi-variate linear regression analysis was per-formed on those significant factors setting outcome as the dependent variable and unranked significant factors as the inde-pendent variables. Lower ocular response (P = 0.003) and motor response scores (P = 0.001) and the absence of brain stem reflexes (P = 0.001) were found to be associated with adverse outcome (death) in acute coma.

It is a well-known fact that the prognosis in coma depends on its severity. Assessing the severity of coma by subjective, poorly defined terms such as stupor, semi coma, and deep coma was ineffective in predicting the outcome and there was a great deal of inconsistency when different observers carried out the assessment(11). The Glasgow Coma Scale is a standardized system developed initially in traumatic coma to assess the degree of coma and to identify the seriousness of brain injury in relation to outcome(2). It has gained widespread use as it is highly reproducible, can be quickly performed at the bedside and provides useful information on the progress and prognosis of a comatose individual(12,13). One limitation of the Glasgow Coma Scale is that only a few studies are available regarding its useful-ness in nontraumatic pediatric coma as a whole(14). Another drawback of GCS score is loss of information, which occurs due to the summation of the individual scores(15). Further the importance of the individual components of Glasgow Coma Scale in assessing nontraumatic pediatric coma has not been well studied.

In our study, a low total MGCS score was found to be associated with adverse short-term outcome i.e., death by Spearman rank analysis. The likelihood of death in patients with MGCS less than 8 was much higher than when the MGCS was >8 (odds ratio 78.9, relative risk 27.1). Studies in both traumatic and non-traumatic coma have indicated that mortality is high when the GCS is less than 8(14,16). However, they have not studied the association of individual components of GCS and outcome. In our study, we have noted that mortality is significantly higher when the ocular and verbal response score is <2, motor response 1 and brainstem score <1.

However, multivariate analysis did not substantiate this finding and revealed that it is not the total score but the ocular and motor response score which are relevant in short term outcome prediction. It has been pointed out by Teasdale(15), who had developed GCS that the information conveyed by the coma score is less than that contained in the individual three responses and summation of the score can lead to loss of information. Sacco, et al.(17) has noted that motor response alone a significant independent predictor of outcome than the total score. In our study we have also found that verbal response score does not predict poor outcome individually and other authors have not reported on this aspect of GCS.

Another interesting finding noted in our study is the relationship between brain stem reflexes and poor outcome. Absence of one or more brain stem reflexes predicted adverse short-term outcome in our study population. Various other authors also have noted that use of brain stem reflexes in addition to GCS has improved the outcome prediction in acute coma(17,18).

In our study it has been noted that etiology did not affect the short-term outcome. This is in contrast to other authors who have concluded that outcome of coma was dependent on the etiology(11,17,18). It is important to realize that studies on prognosis of coma are affected by certain issues like self-fulfilling nature of the underlying disease, host response and treatment strategies and these are likely to have a significant effect on the outcome(11). Further death in coma often results not from the failure of the primary neurological mechanisms but from other secondary non-neurological causes. Also, the distribution of etiology of coma in our study population was uneven with infection accounting for more than three-fourths of comatose patients. These issues could have affected the assessment of the effect of etiology on outcome in our study.

We conclude that ocular, motor response scores and brainstem reflexes are more predictive of the short-term outcome than the total MGCS score. A score incorporating ocular response, motor response and brain stem reflexes should be evaluated to assess the outcome in nontraumatic coma in the pediatric population.

Contributors: PCNP designed the study, collected and interpreted the data. PN designed the study, critically evaluated the data and the paper and she will act as the guarantor of the paper. VTS drafted the article, analyzed the data and prepared the manuscript.

Funding: None.

Competing interest: None stated.

1. Tasker RC, Cole GF. Acute encephalopathy of childhood and intensive care. In: Brett EM, editor. Pediatric Neurology, 3rd edn. Edin-burgh: Churchill Livingstone, 1996. p 691-729.

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

3. Tatman A, Warren A, Williams A, Powell JE, Whitehouse W. Development of a modified pediatric coma scale in intensive care clinical practice. Arch Dis Child 1997, 77: 519-521. Erratum in Arch Dis Child 1998; 78: 289.

4. Simpson D, Reilly P. Pediatric coma scale. Lancet 1982; 2: 450.

5. Raimondi AJ, Hirschauer J. Head injury in the infant and toddler: coma scoring and outcome scale. Childs Brain 1984; 11: 12-35.

6. Morray JP, TyierDC, Jones TK, Stuntz JT, Lenire RJ. Coma scale for use in brain injured children. Crit Care Med 1984; 12: 1018-1020.

7. Gordon NS, Fois A, Jacobi G, Minns RA, Seshia SS. Consensus statement: The manage-ment of the comatose child. Neuropediatrics 1983; 14: 3-5.

8. Reilly P, Simpson O, Sprod R, Thomas L. Assessing the conscious level in infants and young children: A pediatric version of Glasgow Coma Scale. Child Nerv Syst 1988; 4: 30-33.

9. Levy DE, Caronna JJ, Singer Bh, Lapinski RH, Frydman H, Plum F. Predicting outcome from hypoxic ischemic coma. JAMA 1985; 253: 1420-1426.

10. Awasthi S, Moin S, Iyer SM, Rehman H. Modified Glasgow Coma Scale to predict mortality in children with acute infections of the central nervous system. Nat Med J Ind 1997; 10: 214-216.

11. Bates D. Defining prognosis in medical coma. J Neurol Neurosurg Psychiatry 1981; 44: 552- 554.

12. Menegazzi JJ, Davis EA, Sucor AN, Paris PM. Reliability of the Glasgow Coma Scale either used by emergency physicians and para-medics. J Trauma 1993; 34: 46-48.

13. Prasad K. The Glasgow Coma Scale: A critical appraisal of its clinimetric properties. J Clin EpidemioI1996; 49: 755-763.

14. Chaturvedi P, Kishore M. Modified Glasgow Coma Scale to predict mortality in febrile unconscious children. Indian J Pediatr 2001; 68: 314 -318.

15. Teasdale G, Jennett B, Murray L, Murray G. Glasgow Coma scale: to sum or not to sum? Lancet 1983; 2: 678.

16. Palmer S, Bader MK, Qureshi A, Palmer J, Shaver T, Borzatta M, The impact on outcomes in a community hospital setting of using the AANS traumatic brain injury guidelines. Americans Associations for Neurologic Surgeons. J Trauma 2001; 50: 657 -664.

17. Sacco RL, Van Gool R, Mohr JP, Hauser WA. Nontraumatic coma. Glasgow Coma Score and coma etiology as prediction of 2 week outcome. Arch Neurol 1994; 47: 1181-1184.

18. Levy DE, Bates D, Caronna JJ, Cartilidge NE, Knill-Jones RP, Lapinski RH, et al. Prognosis in nontraumatic coma. Ann Intern Med 1981; 94: 293- 301.

19. Quizi SA, Shan MA, Mughal N, Ahmad M, Joomro B, Sakata Y, et al. Dexamethasone and bacterial meningitis in Pakistan. Arch Dis Child 1996; 75: 482-488.

20. Ahuja GK, Mohan KK, Prasad K, Behan M. Diagnostic criteria for tuberculous meningitis and their validation. Tuber Lung Dis 1994; 75: 149-152

21. Ferenci P, Lockwood A, Mullen K, Tarter R, Weissenborn K, Blei AT. Hepatic encephalo-pathy – definition, nomenclature, diagnosis and quantification: Final report of the working party at the 11th World Congress of Gastroenterology, Vienna 1998. Hepatology 2002; 35: 716-721.