Presentation (CTI)

OBJECTIVES: Pre-mortem analysis, a structured method from project management science, is used to identify failure risks before implementation. This study explored the application of AI-supported pre-mortem analysis to HEOR strategy development, with the aim of anticipating methodological, evidence, and HTA risks earlier in the decision lifecycle.
METHODS: A conceptual framework was developed integrating pre-mortem methodology with AI-supported scenario generation. HEOR strategies (e.g., evidence generation plans, modelling approaches, HTA submission strategies) were decomposed into key assumptions and decision points. An AI system was used to generate hypothetical failure scenarios based on historical HTA decisions, methodological guidance, and documented HEOR risks. Outputs were categorized into risk domains including evidence sufficiency, methodological alignment, stakeholder interpretation, and implementation feasibility. Expert review assessed plausibility and usefulness of identified risks.
RESULTS: The AI-supported pre-mortem analysis identified a broad range of potential failure modes, including misalignment between trial endpoints and HTA expectations, overreliance on immature evidence, structural weaknesses in modelling assumptions, and vulnerabilities in value narratives. Many risks aligned with issues observed retrospectively in failed or delayed HTA submissions. Experts reported that the approach helped surface implicit assumptions and encouraged earlier mitigation planning, though prioritization and contextual judgement remained human-led.
CONCLUSIONS: AI-supported pre-mortem analysis offers a structured approach to proactively identifying HEOR strategy risks before downstream investment and submission. When combined with expert interpretation, this method may enhance strategic robustness, reduce late-stage HTA failure, and support more resilient evidence generation planning.