BACKGROUND: Despite extensive studies on dengue fever, there is still limited knowledge about factors associated with poor outcomes in cases of dengue fever. The shock index (SI) is a bedside tool previously used in the adult population, adopted as a marker for poor outcomes in many shock states. There are limited studies applying the SI in children. There are also no known local studies applying an age-adjusted version as a marker or predictor of poor outcomes in severe acute illness, such as dengue.

OBJECTIVES: To determine the diagnostic ability of the age-adjusted pediatric shock index in predicting outcomes in cases of dengue admitted at a tertiary children's hospital.

METHODS: This is a prospective cohort study performed in a pediatric tertiary hospital over a period of 30 days. Admitting heart rate (HR) and systolic blood pressure (SBP) were taken to determine their shock index. This was then grouped according to age groups based on known literature and corresponding acceptable age-adjusted shock indices (ASI), and compared with outcomes such as final dengue classification (non-severe vs severe), use of inotropes, and mortality.

RESULTS: A total of 90 patients were identified for the study. Three were excluded due to exclusion criteria. 87 cases were followed up after admission from the ER. Unadjusted Shock Index (USI) was found not to be associated with both final dengue classification (as severe dengue) and use of inotropic support. In contrast, ASI was associated with both final dengue classification (p < 0.001) and use of inotropes (p < 0.039). The ASI had a fairly accurate capability of predicting poor outcomes for both final dengue classifications, with an area under the ROC curve of 0.7122, and eventual use of inotropes, with an area under the ROC curve of 0.6435.

CONCLUSIONS AND RECOMMENDATIONS: SI was found to be a helpful tool in predicting poor outcomes, but only when the Age-adjusted Shock Index (ASI) was used. A longer data collection period is recommended to be able to include mortality as an outcome. The predictive value of the tool can be tested against various other markers of poor outcome to widen the application of this non-invasive measure of hemodynamic status.