Speakers - 2026

Title: Beyond High-Risk Classification: AI-Informed Analysis Reveals Hidden Risk Patterns in Postpartum Hemorrhage

Abstract

This study aimed to evaluate the distribution of PPH cases across predefined risk categories and to explore the contribution of artificial intelligence (AI)-informed analysis in identifying hidden risk patterns using real-world clinical data.

A pilot retrospective cohort study was conducted using data from more than 3,000 deliveries in 2024 from a high-volume public maternity network in Brazil. AI-informed analytical approaches, including machine learning techniques and AI-assisted processing of unstructured clinical data, were applied to enhance pattern recognition and support epidemiological interpretation.

The overall prevalence of PPH was 6.8%. Notably, 48.1% of cases occurred in patients not classified as high risk prior to delivery, including 22.9% in the medium-risk and 25.2% in the low-risk groups. These findings demonstrate that a substantial proportion of PPH cases are not captured by traditional risk-based models.

Current risk-based prevention strategies for PPH may be insufficient in real-world clinical settings. AI-informed approaches represent a promising pathway toward more dynamic and accurate risk assessment, with potential to improve early identification and support clinical decision-making.

What will the audience take away?

• Nearly half of PPH cases occur outside the high-risk group, challenging current prevention models
• Evidence of limitations in traditional risk stratification in real-world settings
• Importance of continuous and dynamic intrapartum risk assessment
• Role of AI-informed analysis in uncovering complex clinical patterns
• Practical implications for improving maternal outcomes