Advancing Mortality Prediction in Pulmonary Embolism with Machine Learning Algorithms

Strategic Insights -

Researchers from the Hamadan University of Medical Sciences in Iran have conducted a systematic review and meta-analysis to evaluate the performance of machine learning (ML) models in predicting mortality among pulmonary embolism (PE) patients. The study, published in Pulmonary Circulation, synthesized data from 17 studies encompassing 844,071 cases, finding that ML models outperform traditional risk stratification tools in predicting PE mortality. The results suggest that advanced ML models demonstrate high sensitivity and specificity, offering robust potential for clinical decision-making.

Key Takeaways:

  • The systematic review and meta-analysis evaluated the performance of machine learning (ML) models in predicting mortality among pulmonary embolism (PE) patients, synthesizing data from 17 studies encompassing 844,071 cases.
  • Logistic Regression was the most commonly used algorithm, followed by advanced models like Random Forests, Support Vector Machines, XGBoost, and Neural Networks.
  • The pooled performance metrics from 12 studies demonstrated a sensitivity of 0.88 (95% CI: 0.78-0.94, I 2 = 90.43%), specificity of 0.79 (95% CI: 0.62-0.89, I 2 = 99.53%), positive likelihood ratio of 4.1 (95% CI: 2.2-7.7), negative likelihood ratio of 0.16 (95% CI: 0.08-0.29), diagnostic odds ratio of 26 (95% CI: 10-71), and an AUROC of 0.91 (95% CI: 0.88-0.93), indicating excellent discriminative ability.
  • Subgroup analyses revealed higher sensitivity in advanced ML models (89.7%) and non-USA studies (97.2%), with advanced ML showing lower specificity heterogeneity (I 2 = 0%).
  • Significant heterogeneity was observed, particularly in specificity (I 2 = 99%), driven by traditional ML and USA-based studies.
  • Minimal publication bias was noted for sensitivity (Egger's p = 0.942), but specificity showed potential bias (Egger's p = 0.038 after outlier exclusion).

Statistics:

  • 844,071 cases were included in the 17 studies synthesized in the meta-analysis.
  • Sensitivity was 0.88 (95% CI: 0.78-0.94, I 2 = 90.43%).
  • Specificity was 0.79 (95% CI: 0.62-0.89, I 2 = 99.53%).
  • Positive likelihood ratio was 4.1 (95% CI: 2.2-7.7).
  • Negative likelihood ratio was 0.16 (95% CI: 0.08-0.29).
  • Diagnostic odds ratio was 26 (95% CI: 10-71).
  • Area under the receiver operating characteristic curve (AUROC) was 0.91 (95% CI: 0.88-0.93).

Sources:

  • "Advancing Mortality Prediction In Pulmonary Embolism Using Machine Learning Algorithms-systematic Review and Meta-analysis." Pulmonary Circulation, 2025;15(3).
  • Research authors: Pooya Eini, Peyman Eini, Homa Serpoush, Mohammad Rezayee, Jason Tremblay.
  • Institutional affiliation: Hamadan University of Medical Sciences, Infectious Diseases Research Center, Hamadan, Iran.