Machine Learning vs Traditional Approaches to Predict All-Cause Mortality for Acute Coronary Syndrome: A Systematic
Aashray K Gupta1, Cecil Mustafiz2, Daud Mutahar3
1Discipline of Surgery, University of Adelaide, Adelaide, Australia.
Insights
Machine learning models show superior accuracy in predicting mortality for acute coronary syndrome (ACS) patients compared to traditional risk scores. Further validation is needed for widespread clinical use of these advanced models.
Area of Science:
- Cardiology
- Medical Informatics
- Data Science
Background:
- Acute coronary syndrome (ACS) is a leading global cause of mortality.
- Accurate risk prediction is crucial for effective ACS patient management and treatment.
- Existing risk stratification models (e.g., GRACE, TIMI) have limitations in predictive accuracy.
Purpose of the Study:
- To compare the performance of machine learning (ML) models against traditional risk scores for predicting all-cause mortality in ACS patients.
- To evaluate the discriminative value of ML models in ACS prognosis.
- To synthesize evidence on the comparative effectiveness of ML in ACS risk stratification.
Main Methods:
- Systematic review and meta-analysis of studies comparing ML models and traditional methods for ACS mortality prediction.
- Searched multiple databases (PubMed, Embase, Web of Science, etc.) up to October 30, 2024.
- Primary outcome: comparative discrimination measured by C-statistics.
Main Results:
- Twelve studies involving 250,510 patients were analyzed.
- ML models achieved a summary C-statistic of 0.88, outperforming traditional methods (0.82).
- ML models demonstrated significantly superior discrimination for all-cause mortality in ACS patients.
Conclusions:
- Machine learning models offer improved risk prediction for ACS mortality compared to established tools.
- Despite superior performance, clinical application of ML models requires further validation due to identified risks of bias.
- Future research should focus on robust validation of ML models for reliable clinical implementation in ACS care.
Background:
Acute coronary syndrome (ACS) remains one of the leading causes of death globally. Accurate and reliable mortality risk prediction of ACS patients is essential for developing targeted treatment strategies and improve prognostication. Traditional models for risk stratification such as the GRACE and TIMI risk scores offer moderate discriminative value, and do not incorporate contemporary predictors of ACS prognosis. Machine learning (ML) models have emerged as an alternate method that may offer improved risk assessment. This review compares ML models with traditional risk scores for predicting all-cause mortality in patients with ACS.
Methods:
PubMed, Embase, Web of Science, Cochrane, CINAHL, Scopus, and IEEE XPlore databases were searched through October 30, 2024, as well as Google Scholar and manual screening of reference lists from included studies and the grey literature for studies comparing ML models with traditional statistical methods for event prediction of ACS patients. The primary outcome was comparative discrimination measured by C-statistics with 95% confidence intervals (CIs) in estimating risk of all-cause mortality.
Results:
Twelve studies were included (250,510 patients). The summary C-statistic of best-performing ML models across all end points was 0.88 (95% CI 0.86-0.91), compared with 0.82 (95% CI 0.80-0.85) for traditional methods. The difference in C-statistic between ML models and traditional methods was 0.06 (P < 0.0007). Five studies undertook external validation. The PROBAST tool demonstrated high risk of bias for all studies. Common sources of bias included reporting bias and selection bias. Best-performing ML models demonstrated superior discrimination of all-cause mortality for ACS patients compared with traditional risk scores.
Conclusions:
Despite outperforming well established prognostic tools such as the GRACE and TIMI scores, current clinical applications of ML approaches remain uncertain, particularly in view of the need for greater model validation.
