Beyond the Anamnesis: Big Data Reveals Rash, Not Bleeding, as the Critical Predictor of Dengue Shock Syndrome

Abstract

Dengue shock syndrome (DSS) is a critical, life-threatening complication; however, early prediction remains highly challenging. Clinicians historically rely on classic warning signs obtained through subjective anamnesis, such as persistent fever and spontaneous bleeding, to triage patients despite the potential for observational bias. To address this clinical gap, this study aimed to evaluate the objective predictive value of isolated, patient-reported symptoms on the incidence of shock using a large-scale population dataset. A retrospective observational design utilized big data analytics on routine surveillance records from the Banyumas Regional Health Agency, spanning from 2024 to 2025. The analyzed dataset comprised 1,381 patients, and statistical models, including bivariate analyses and logistic regression, were constructed to quantify the association between individual clinical symptoms and shock outcomes. The findings reveal a striking clinical anomaly: classically feared symptoms, including bleeding and fever, lack significant predictive power for shock deterioration. Conversely, the presence of a rash emerges as a critical clinical indicator; patients presenting with a rash demonstrate a nearly twofold increased risk of progressing to shock (OR = 1.95). Relying exclusively on subjective anamnesis and universal symptom profiles proves unpredictable and insufficient for anticipating severe dengue. Therefore, a rash serves as a crucial early red flag for underlying plasma leakage, mandating healthcare providers to move beyond rigid textbook paradigms and adopt a holistic, adaptive diagnostic approach to optimize patient safety.