Table of Content - Volume 19 Issue 3 - September 2021
Value of coma scale in prediction of mortality in non-traumatic coma in 1-12 years old children
Stalin Ramprakash1, Bhavani M H2*
{1Assistant Professor, Department of Paediatrics} {2Associate Professor, Department of Obstetrics and Gynaecology} Shridevi Institute of Medical Sciences and Research hospital, NH-4, Bypass Road, Tumkur- 572106, INDIA. Email: hrbs2006@yahoo.co.in
Abstract Background: Acute non-traumatic coma is a common problem in pediatric practice accounting for 10-15% of all hospital admissions 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. Glasgow Coma Scale (MGCS) in spite of its various drawbacks has been widely used for assessing pediatric coma, though only few studies are available to support its use in pediatric coma as a whole. Hence the study was undertaken to evaluate the usefulness of coma scale in prediction of mortality in children between 1-12 years old admitted with coma of non-traumatic origin. Methods: A prospective study was conducted at Inpatient services of intensive medical care unit and general medical wards of Institute of child Health and Hospital for children, Egmore, Chennai, between October 2018 to November 2019. A Study Population consisting of children in the age group 1-12 years admitted in intensive care unit and medical wards with alteration in sensorium as one of the predominant complaints and admitted within 7 days of onset of coma were included. Sample size was 148. Chi-square test was used to study the association between the scores at Various times points and outcome. Results: It was observed that when the scores were between 3‐ 5 mortality rate was 67.0%, it was 17% when the minimal observed score was between 6-8. The group in which the minimal score never decreased be1ow 9 had no mortality. As the score increases the proportion of death decreases and the proportion having good recovery increases and vice versa. This type of association has been observed at different time points (Initial, 24 hrs, 48 hrs and 72 hrs). The significance of this association increases with the progression of time since admission. P value for the degree of association at various time points are: At presentation P < 0.00001, 24hrs - P < 0.000001, 48 hrs - P < 0.000001 and 72 hrs P < 0.00000001. Conclusion: There is highly statistically significant association between the GSC score levels and outcome. As the score increases the proportion of death decreases and the proportion having good recovery increases and vice versa. Key Words: Glasgow coma scale, Mortality rate, Non traumatic coma.
INTRODUCTION Acute non-traumatic coma is a common problem in pediatric practice accounting for 10-15% of all hospital admissions and is associated with significant mortality. Assessment of the severity of coma is essential to comment on the likelihood of survival in comatose children.1In the last three decades, various scores have been used to assess the severity of coma and to predict its outcome. These includes the Glasgow coma scale, the James adaptation of Glasgow coma scale, the Simpson and Reilly scale, the Children’s coma scale, and the Jacob’s scale. Among these, the modified Glasgow Coma Scale (MGCS) in spite of its various drawbacks has been widely used for assessing pediatric coma, though only few studies are available to support its use in pediatric coma as a whole. Coma scales have been devised to determine the depth of unconsciousness in children and some have included brainstem signs.2,3Because of these limitations in assessing the prognosis, we need to device a system so that the prognosis can be easily assessed. It should be simple so that it can be practiced universally. It should be cheap, reproducible and reliable. Interpretation should be easier and should not require sophisticated training. Glasgow coma scale is one such system, the role of which in traumatic coma is well established. Initially it was used in evaluation of coma in adults due to head injury, later in children. It is not widely used in non-traumatic coma as a Prognostic indicator. As far as the use of Glasgow coma scale in children is concerned it cannot be directly applied to all age groups. So modification of the scoring system is necessary to suit the younger children especially those younger than. Hence the study was undertaken to evaluate the usefulness of coma scale in prediction of mortality in children between 1-12 years old admitted with coma of non-traumatic origin. MATERIALS AND METHODS A prospective study was conducted at Inpatient services of intensive medical care unit and general medical wards of Institute of child Health and Hospital for children, Egmore, Chennai, between October 1998 toNovember 1999. A Study Population consisting of children in the age group 1-12 years admitted in intensive care unit and medical wards with alteration in sensorium as one of the predominant complaints and admitted within 7 days of onset of coma were included. Coma recovering within 24 hrs, coma following trauma, poisoning, cardiac or respiratory illness as the main underlying problem were excluded. Sample size was 148. Severity of coma is assessed using Glasgow coma scale in children above 5 years and Adelidescore for children below 5 years. Serial scoring was done at 8 hr interval in the first 24 hrs and then at 24 hrs intervals till 5th day of admission. Statistical analysis: The sensitivity and Specificity of Various Scores regarding the prediction of mortality were arrived at and the optimal cut Off point was arrived by construction of Receiver Operative Characteristic Curve (ROC). Chi-square test was used to study the association between the scores at Various times points and outcome. RESULTS Table 1: Age and sex distribution, N=148
Table 2: Etiology and mortality.
In our study we have observed that the mortality is higher in hepatic encephalopathy (61.9%) and hypoxic ischemic encephalopathy (85.7%) most often following revival from cardiac arrest. Table 3: Association between the score and mortality
Chi-Square = 57.31, P – < 0.000001 It was observed that when the scores were between 3‐ 5 mortality rate was 67.0%, it was 17% when the minimal observed score was between 6-8. The group in which the minimal score never decreased be1ow 9 had no mortality. Table 4: Association between the score at presentation and mortality.
Chi-Square = 55.45, p = 0.00001 When the relationship between the scores at presentation and final outcome were analysed it was found that those with a score of 5 or less had a mortality of 69% whereas the mortality was only 16.9% in those who had the presentation score as 6 and above. There is significant association between the scores at presentation and final outcome (p < 0.00001)
Table 5: Association between the score after 24 hours and mortality.
Chi-Square = 60.63, p < 0.00001 Those who had scores less than or equal to 5 had a mortality rate of 70.3%. Whereas those with a score above or equal to 6 had only 5.6% mortality.
Table 6:Association between coma score after 48 hrs of admission and mortality
Chi-Square = 63.1, p < 0.00001 From our observations it can be made out that those with a score 5 or less at 48 hrs had 72%mortality whereas mortality for the group with score of 6 or more at 48 hrs was only 7.5%.
Table 7: Association between the coma scores at 72 hrs and mortality
Chi-Square = 72.1, p < 0.00000001 Those who had scores less than or equal to 5 at 72hrs had a mortality of 66.7% but those with 6 and above at 72 hrs had a mortality of about 7.1%. Analysis shows that there is highly significant association between the scores at 72 hrs and mortality (p < 0.00000001). Table 8: Association between coma score trends and mortality
Chi-square = 38.8, P <0.000001 When the coma score remained static or showed deterioration the mortality was 55.7% and it was 8.9% when the scores improved by atleast one. When there was no change or deterioration in coma score over 48 hrs the mortality was 57.1% and it was 7.4 % when there was improvement in the score by atleast one. When coma score remained static or showed deterioration by 72 hrs the mortality was 59.3% and when there was improvement by atleast one during the same period the mortality was 7.4%. There was high degree of association between the change in scores over 72 hrs and the mortality. Among those children in whom score deteriorated or did not improve over first five day period or till death, mortality was 80.0% and it was 7.7% for the group which showed an improvement in score by atleast one.
Table 9: Sensitivity and Specificity of the different scores at presentation as predictors of mortality
Coma score of 5 Predicts the mortality with optimum sensitivity (79.0%) and specificity (74.4%). By using receiver operating characteristic (ROC) curve the optimum score that has the maximum sensitivity and specificity can be arrived. Score 5 is the optimum cut off point for the prediction of mortality at presentation. Score 5 or less at presentation has increased mortality. Table 10: Sensitivity and Specificity of different scores at 24 hrs as predictors of mortality
The optimum score which predicts the mortality at 24 hrs with high sensivity 91.8% and specificity 79.06% is 5 or less. A score of 5 or less at 24 hrs is associated with poor outcome.
Table 11: Sensitivity and Specificity of different scores at 48 hrs as predictors of mortality
At 48 hrs score 5 or less predicts the mortality with high sensitivity 83.8% and specificity 86%. Score 5 or less at 48 hrs can be taken as optimum cutoff point below which the prognosis is poor. Table 12: Sensitivity and Specificity of different scores at 72 hrs as predictors of mortality
The optimum score with high sensitivity and specificity is score 6 (86.3% and 80.27% respectively). Score 6 or less of 72 predicts the mortality with high sensitivity and specificity.
DISCUSSION Coma scale is a simple clinical tool which can be applied by any trained person. But inter observer variations are likely to occur. When uniform training is given to all those who would be applying the scale the variations can be minimized to a large event. The association between the coma score at various time points, change in scores with time and the mortality were studied using chi-square test. In our study we found that there is highly statistically significant association between the GSC score levels and outcome. As the score increases the proportion of death decreases and the proportion having good recovery increases and vice versa. This type of association has been observed at different time points (Initial, 24 hrs, 48 hrs and 72 hrs). The significance of this association increases with the progression of time since admission. P value for the degree of association at various time points are: At presentation P < 0.00001, 24hrs - P < 0.000001, 48 hrs - P < 0.000001 and 72 hrs P < 0.00000001. This high degree of association is in accordance with the study conducted by Soustiel et al.4 who reported that among the clinical parameters GCS provided the most accurate prognosis (in 80%). Study done by Grewal et al.5 who found that GCS trend or brain stem reflexes used alone were significantly correlated with outcome. Their study is with comatose children following head injury where as we have conducted the study in non- traumatic coma in children. We have studied the ability of different scores to predict the final outcome using their sensitivity and specificity at various time points viz. initial, 24hrs, 48hrs and72 hrs. We have arrived at the optimum scores that form a cutoff point in prediction of mortality and good recovery at various time points. Optimum scores that predict mortality, at presentation- Score 5 and below predicts increased mortality. Score 5 predicts mortality with sensitivity 79% and specificity 74.4%. At 24 hrs of admission - Score 5 and below predicts increased mortality. Score 5 predicts mortality with sensitivity 91.8% and specificity 79.06%. At 48 hrs of admission - Score 5 and below predicts increased mortality. Score 5 predicts mortality with sensitivity 83.8% and specificity 86.0%. At 72 hrs of admission - Score 6 and below predicts increased mortality. Score 6 predicts mortality with sensitivity 86.3% and specificity 80.2%. It can be seen that the optimum scores that predict mortality increase with time since admission. Longer a patient remains in a lower score poorer is the prognosis. The finding that the score of 5 or less within 24 hrs of hospitalization if associated with poor prognosis is in accordance with the study done by Awasthi et al. 6 from Lucknow. Rober H AHaslam7 in Nelson textbook of pediatrics also reports that score 5 or less is associated with grave prognosis. In the study conducted by Reddy AM et al.8 they concluded that a low total Glagow coma scale score was found to be associated with adverse short-term outcome. The likelihood of death in patients with GCS less than 8 was much higher than when the GCS was >8 i.e., 6 (15%) cases expired out of 38 cases with GCS ≤8 and 2 (4%) cases expired out of 42 cases with GCS >8. In a study by Cicero MX et al. 9on pediatric trauma victims they concluded that the Glasgow coma scale is predictive of mortality and injury outcomes, as well as both Emergency Department and hospital length of stay, and has excellent prognostic accuracy. The Glasgow motor component has predictive value for injury and mortality that is nearly equivalent to the full Glagow coma scale.Though many studies have analysed the scores at presentation, cutoff points for different time intervals after admission were not analysed.
CONCLUSION In our study we found that there is highly statistically significant association between the GSC score levels and outcome. As the score increases the proportion of death decreasesand the proportion having good recovery increases and vice versa. This type of association has been observed at different time points (Initial, 24 hrs, 48 hrs and 72 hrs). The significance of this association increases with the progression of time since admission.
REFERENCES
Policy for Articles with Open Access
|
|