Assessment of severity of illness on admission using SICK score and evaluation of the risk factors associated with mortality in children admitted in a paediatric urban tertiary care centre in south India

Abstract

Background: “Signs of inflammation that can kill” (SICK) score is one of the severities scoring systems used for predicting outcome of children at admission. The aim of the present study was to study the clinical and demographic profile of children admitted to Paediatric ward, to assess the usefulness of SICK score in predicting the mortality and evaluate the risk factors in predicting mortality. Methods: SICK scoring was done for 369 children on admission. The outcome was recorded as death or discharge. The associated factors were analysed using SPSS software package analysis. Receiver operating curve was used to arrive at the cut-off point of SICK score for predicting mortality. Quantitative data differences between children who died and children who were discharged from the hospital were analysed using student independent t test. Need for assisted ventilation, presence of shock, age less than 3 years, and SICK score>2 were studied to find their association with mortality. Statistical analysis was done using univariate analysis and those factors that were significantly associated with mortality were subjected multivariate logistic regression analysis.Results: The performance of SICK score was “excellent” in discriminating between death and survival with area under the receiver operating characteristics curve 0.94. Age<3-year presence of shock, need for mechanical ventilation and SICK score>2 showed statistically significant association with mortality as evidenced by multivariate logistic regression model. Conclusions: SICK score performed extremely well in predicting mortality on admission. Age<3 years, SICK score>2, Presence of Shock and need for assisted ventilation showed statistically significant association with mortality.