Researchmlvenous thromboembolismrisk predictionexternal validation
Machine Learning Predicts Venous Thromboembolism Effectively
7.0
A systematic review and meta-analysis published in Journal of Medical Internet Research (2025) reviewed 27 studies up to March 26, 2025, covering 596,092 patients to evaluate machine learning models for venous thromboembolism (VTE) prediction. Pooled metrics showed sensitivity 0.79, specificity 0.82, and C-index 0.84, but 67% of studies had high risk of bias, prompting calls for external validation and improved methodology.
Key Points
- 1Reports pooled performance: sensitivity 0.79 and specificity 0.82 across 27 studies.
- 2Highlights methodological concern: 67% of studies had high risk of bias, limiting reliability.
- 3Urges external validation and better missing-data handling to enable clinical translation.
Scoring Rationale
Comprehensive meta-analysis with strong pooled performance; limited by high methodological bias and need for external validation.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
