Related Experiment Videos
Development and Evaluation of a Simple Prognostic Score to Predict Mortality in Patients Hospitalized for Heatstroke
Yves M K Kantagba1,2, David Lankoande3, Seydou G Barro1,2,4
1Université NAZI BONI, Bobo Dioulasso, Burkina-Faso.
None:
Heatstroke is a serious medical emergency with a high mortality rate. Early identification of high-risk patients is crucial for optimizing care. Currently, no simple and validated prognostic score exists for this condition in sub-saharan Africa. This study aimed to develop and validate a simple prognostic score to predict mortality in patients hospitalized for heatstroke at the University Hospital of Bogodogo, Burkina Faso. This was a retrospective cross-sectional study including 167 patients hospitalized for heatstroke from March 1 to May 31, 2024. Univariate analysis was performed to identify predictive factors (p < 0.05). Multivariate logistic regression with LASSO regularization was used to select independent predictors. A simple score was developed based on five readily available clinical criteria. Validation was performed using bootstrapping (1000 iterations). Performance was assessed using the area under the receiver operating characteristic curve. Of 167 patients (mean age 72.8 ± 10.9 years), 86 (51.5%) died. Five independent risk factors were identified: Glasgow Coma Scale score < 10 (3 points), age ≥ 60 years (2 points), temperature ≥ 40°C (1 point), ≥ 2 comorbidities (1 point), and respiratory distress (1 point). The prognostic score (0-8 points) demonstrated good discriminatory performance (AUC-ROC = 0.854; 95% CI = 0.796-0.911). At the optimal threshold of 5 points, the sensitivity was 81.4% and the specificity 81.5%. Stratification identifies three groups: low risk (0-2 points, mortality 20%), moderate risk (3-5 points, mortality 27%) and high risk (6-8 points, mortality 86%). The resulting score is a simple, rapid, and effective tool for predicting mortality and stratifying risk in patients suffering from heat stroke. It could facilitate triage or reference and guide the intensity of care. External validation with large dataset is still needed before implementation in clinical practice.
