A prospective observational study to assess the efficacy of “SICK” score in pediatrics

Abstract

Background: There is a need for pediatric scoring systems to monitor the progress in clinical outcomes, which are noninvasive, reliable, and easily measurable. Subjects and Methods: This prospective observational study evaluated the utility of a prevalidated “SICK” score for the purpose of assessing the score's efficacy in hospitalized children of age 1 month to 12 years. Results: Of the 531 cases, 145 (27.3%) were admitted to the pediatric intensive care unit (PICU) and 386 (72.7%) were admitted to the pediatric ward. The probability of PICU admission was 11.42 (95% confidence interval [CI] 6.3–20.7) when the “SICK” score was ≥2. Children with a score ≥2 had a significantly longer duration of stay (mean difference = 45.58 h, 95% CI [30.65–60.51], P = 0.001). The score performed with an area under curve of 0.691 for a cut off score. The positive predictive value of the score is 73.53% (95% CI 62.66–82.13) and the negative predictive value is 79.48% (95% CI 77.45–81.37). Conclusions: The score performed reasonably well in our center for a profile of moderately sick children and can be recommended to be used as an effective triaging tool in a similar setting.