O-22 From bite to insight: clinical, demographic and laboratory predictors of severe outcomes in adult dengue fever patients

Methods: This retrospective observational study included 289 adults (≥18 years) with laboratory-confirmed dengue presenting to the Accident and Emergency Department, GPHC between July 2023 and January 2025. Outcomes were classified as mild/moderate versus severe/death (including dengue hemorrhagic fever, dengue shock syndrome, ICU admission, or mortality). Data were analyzed using SPSS version 27. Univariate analyses were performed, and variables with p<0.25 were considered, to avoid excluding potentially important predictors, for multivariable logistic regression. Model performance was assessed using the omnibus test, Nagelkerke R², classification accuracy, and the Hosmer-Lemeshow goodness-of-fit test.

Results: The mean age was 39.0±17.6 years. Severe dengue was more frequent in females (25.0%, p=0.039) than males (15.2%) and increased with age, from 15.3% in patients aged 18–40 years (p=0.035) to 32.1% among those aged ≥66 years. Severe clinical manifestations (abdominal pain, vomiting, mucosal bleeding, p=0.021) and symptom duration >8 days (35.9%; p=0.038) were associated with poor outcomes. Vital signs and comorbidities were not significant predictors. Platelet count was the only laboratory parameter associated with severity (p=0.002), with thrombocytopenia (≤150 × 10⁹/L) accounting for 27.3% of severe cases. The multivariate model was significant (χ²(16) = 60.45, p < 0.001), explained 30.7% of variance (Nagelkerke R²=0.307), achieved 82.0% accuracy, and showed good fit (Hosmer-Lemeshow p=0.063). Lower platelet count was the strongest independent predictor (AOR = 0.993; 95% CI: 0.989–0.997; p<0.001).

Conclusion: Older age, prolonged symptom duration, severe clinical features, and thrombocytopenia independently predict severe dengue at GPHC. Platelet count is a practical marker for early risk stratification and triage optimization in Guyana.