Factors associated with 90-day mortality in Vietnamese stroke patients: Prospective findings compared with explainable machine learning, multicenter study

Fri Sep 20 14:00:00 UTC 2024: ## Vietnamese Stroke Patients: Mortality Rates and Machine Learning Insights

**Hanoi, Vietnam** – A groundbreaking multicenter study in Vietnam has shed light on mortality rates and key predictors of death in patients suffering from ischemic stroke and intracerebral hemorrhage. The research, published in PLOS ONE, involved 2,209 patients from ten specialized stroke centers across the country.

**Key findings:**

* **Mortality:** The study found a 90-day mortality rate of 8.2% for ischemic stroke and a significantly higher rate of 20.5% for intracerebral hemorrhage.
* **Risk Factors:**
* **Ischemic stroke:** Factors associated with higher mortality included a history of atrial fibrillation, higher National Institutes of Health Stroke Scale (NIHSS) scores, lower Alberta Stroke Program Early CT-Scan (ASPECTS) scores, and specific stroke etiologies (large artery atherosclerosis and cardioembolic stroke).
* **Intracerebral hemorrhage:** Factors associated with higher mortality included higher NIHSS scores, the presence of subarachnoid hemorrhage, and intraventricular hemorrhage.
* **Machine Learning:** An explainable machine learning model, using extreme gradient boosting and SHapley Additive exPlanations (SHAP), predicted 90-day mortality with 91% accuracy. This model identified key features influencing mortality, including consciousness level, NIHSS score, heart rate, and age.

**Implications:**

* The study highlights the significant burden of stroke in Vietnam, emphasizing the need for improved stroke care and prevention strategies.
* The identification of key risk factors will enable clinicians to better identify high-risk patients and tailor interventions to improve survival.
* The application of machine learning holds promise for enhancing stroke prediction and treatment, particularly in resource-limited settings like Vietnam.

**Future Directions:**

* Continued research is needed to further explore the use of machine learning in stroke care and its potential impact on patient outcomes.
* Additional research focusing on rehabilitation and recovery factors that occur after hospital discharge could provide further insights into long-term mortality trends.
* The study’s findings underscore the importance of continued efforts to improve stroke awareness, prevention, and treatment in Vietnam and other developing countries.

Read More