Development of a Scoring Model to Predict Severe Dengue in Children at Admission in the Emergency Care: An Observational Study.

Abstract

Objective: Dengue has emerged as the most widely spread mosquito-borne disease, hyperendemic in India. Although severe dengue occurs only in a small proportion of cases, delays in the recognition and management can result in significant morbidity and mortality. Risk stratification of children in the emergency care and identifying those at risk for worsening during hospital stay facilitates optimum utilization of health care resources. The objective of our study was to develop and validate a scoring model to predict the development of severe dengue in hospitalized children by identifying risk factors present in them at the time of admission in emergency department.

Methods: A retrospective study of consecutive children aged 1 month to 18 years admitted with serologically confirmed dengue from January 2019 through December 2021 was conducted. Analysis of clinical and laboratory parameters of children resulted in the development of a comprehensive predictive scoring model. This model was internally validated prospectively on 107 children presenting with nonsevere dengue in the emergency care.

Results: A total of 743 children with confirmed dengue were studied out of which 216 (29.1%) had severe dengue. Presence of third spacing (odds ratio [OR] 3.74, 95% confidence interval [CI] 1.088-7.42, P < 0.001), tender hepatomegaly (OR 1.62, 95% CI 1.04-2.52, P < 0.032), respiratory distress (OR 3.50, 95% CI 1.89-6.51, P < 0.001), and moderate (OR 3.51, 95% CI 1.82-6.74, P < 0.001) to severe (OR 4.76, 95% CI 2.59-8.76, P < 0.001) elevation of aspartate aminotransferase were independent risk factors found to be associated with development of severe dengue. A score ≥7 had a specificity of 87%, negative predictive value of 86%, and overall diagnostic accuracy of 78.5% for predicting severe dengue.

Conclusions: The dengue severity scoring model was found to have reasonable diagnostic accuracy in predicting severe disease prior to hospitalization. However, more studies are required to validate the score in different emergency care settings with varied pediatric populations.