[Predicting major bleeding events in patients with peripheral arterial disease: the OAC³-PAD risk score]

Area of Science:

• Cardiovascular Medicine
• Medical Informatics
• Health Services Research

Context:

• Patients with peripheral arterial disease (PAD) face high bleeding risks due to comorbidities.
• Existing bleeding risk prediction tools are not validated for PAD patients.
• Exclusion of high-risk bleeding patients in trials limits treatment generalizability.

Purpose:

• To develop and validate a pragmatic prediction model for major bleeding events in PAD patients.
• To identify key predictors of major bleeding using routine health insurance claims data and machine learning.
• To facilitate tailored patient-centered risk-benefit assessments for antithrombotic therapy.

Summary:

• A novel risk score, OAC3-PAD, was developed using machine learning on health insurance claims data.
• The OAC3-PAD score identified eight variables predictive of major bleeding within one year of inpatient PAD treatment.
• This score enables a more precise assessment of bleeding risk in PAD patients.

Impact:

• The OAC3-PAD risk score can guide clinical decisions regarding antithrombotic therapy in PAD.
• It allows for optimized, individualized risk-benefit evaluations, maximizing treatment potential.
• Improves patient care by addressing the critical need for bleeding risk stratification in PAD.