Predicting Heart Events in Older Patients Using Machine Learning
Author Information
Author(s): Tang Wen, Wang Xuedong, Yang Xuebing, Sun Ying
Primary Institution: Beijing Friendship Hospital, Capital Medical University
Hypothesis
Can machine learning methods effectively predict in-hospital major adverse cardiac and cerebrovascular events (MACCE) in older patients with coronary heart disease (CHD)?
Conclusion
Machine learning models can effectively predict in-hospital MACCE events in older patients with CHD.
Supporting Evidence
- The XGBoost model achieved an accuracy of 93.56% and an area under the curve of 0.8736.
- 106 patients experienced MACCE during the study.
- The top three predictive features were troponin T, the FRAIL score, and the NYHA functional classification.
Takeaway
This study used computers to help doctors figure out which older patients with heart problems might have serious health issues while in the hospital.
Methodology
The study created prediction models using various machine learning techniques and evaluated their performance with five-fold cross-validation.
Participant Demographics
Older inpatients with coronary heart disease, average age 74.68 years, 58.3% male.
Statistical Information
Confidence Interval
95% CI: 0.756–0.967
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website