RESUMO
Objective: To develop and validate a bedside dengue severity score in children less than 12 years for predicting severe dengue disease. Methods: We carried out an analysis of data on the clinical and laboratory parameters of patients with confirmed dengue, hospitalized in October, 2019 at our center. A comprehensive patient’s score was developed. Predictive models for severity were built using a forward step-wise method. This model was validated on the data of 312 children with dengue admitted during September- October, 2021. Results: Severe dengue was predicted by the dengue severity score with a sensitivity of 86.75% (95% CI 77.52%-93.19%), specificity of 98.25% (95% CI 95.56-99.52%), a positive predictive value of 95.34% (95% CI 92.18%-97.26%) and a negative predictive value of 94.74% (95% CI 87.16%-97.95%). The overall predictive accuracy was 95.2% (95% CI 92.19%- 97.28%). Conclusion: The proposed bedside dengue severity scoring system was found to have good validity. Validating the score in different settings and patient populations is suggested.