The inter-rater reliability and prognostic value of coma scales in Nepali children with acute encephalitis syndrome. (2nd January 2018)

• Record Type: Journal Article
• Authors: Ray, Stephen; Rayamajhi, Ajit; Bonnett, Laura J.; Solomon, Tom; Kneen, Rachel; Griffiths, Michael J.
• Is Part Of: Paediatrics and international child health. Volume 38:Number 1(2018)
• Page Start: 60
• Page End: 65
• Publication Date: 2018-01-02
• Subjects: Acute encephalitis syndrome -- coma scales -- inter-rater reliability -- prognostic value; AES, acute encephalitis syndrome -- ACS, Adelaide coma scale -- AVPU, alert, verbal, pain, unresponsive -- BCS, Blantyre coma scale -- ETAT, emergency triage assessment and treatment -- LOS, Liverpool outcome score -- NTBI, non-traumatic brain injury -- PIM, paediatric risk of mortality -- PRISM, paediatric risk of mortality score -- RPS, resource-poor setting -- TBI, traumatic brain injury; Pediatrics -- Developing countries -- Periodicals; Children -- Health and hygiene -- Developing countries -- Periodicals; Children -- Diseases -- Developing countries -- Periodicals; 618.920009172405
• DOI: 10.1080/20469047.2017.1398503
• Languages: English
• ISSNs: 2046-9047

Abstract

Abstract: Background: Acute encephalitis syndrome (AES) is a common cause of coma in Nepali children. The Glasgow coma scale (GCS) is used to assess the level of coma in these patients and predict outcome. Alternative coma scales may have better inter-rater reliability and prognostic value in encephalitis in Nepali children, but this has not been studied. The Adelaide coma scale (ACS), Blantyre coma scale (BCS) and the Alert, Verbal, Pain, Unresponsive scale (AVPU) are alternatives to the GCS which can be used. Methods: Children aged 1–14 years who presented to Kanti Children's Hospital, Kathmandu with AES between September 2010 and November 2011 were recruited. All four coma scales (GCS, ACS, BCS and AVPU) were applied on admission, 48 h later and on discharge. Inter-rater reliability (unweighted kappa) was measured for each. Correlation and agreement between total coma score and outcome (Liverpool outcome score) was measured by Spearman's rank and Bland–Altman plot. The prognostic value of coma scales alone and in combination with physiological variables was investigated in a subgroup ( n = 22). A multivariable logistic regression model was fitted by backward stepwise. Results: Fifty children were recruited. Inter-rater reliability using the variables scales was fair to moderate. However, the scales poorly predicted clinical outcome. Combining the scales with physiological parameters such as systolic blood pressure improved outcome prediction. Conclusion: This is the first study to compare four coma scales in Nepali children with AES. The scales exhibited fair to moderate inter-rater reliability. However, the study is inadequately powered to answer the question on the relationship between coma scales and outcome. Further larger studies are required.