AI, RWE, Thresholds & Orphan Drugs

MarcusAccess Brief. I'm Marcus, with Sara. Today we're looking at AI integration in HTA decision frameworks — the methodological black holes where commercial interests masquerade as evidence. RWE integration in African markets — where data infrastructure gaps are creating two-tiered access realities. Cost-effectiveness thresholds post-pandemic — how economic shockwaves are recalibrating value narratives. And orphan drug RWE requirements — the tension between regulatory flexibility and evidentiary rigor. Sara, where do you want to start?

SaraAI in HTA is the most overhyped conversation in evidence right now.

MarcusThe EMA released its draft guidance on AI in HTA last week. It's a substantial document that essentially acknowledges the potential while establishing guardrails. The core tension is between algorithmic transparency and the "black box" problem that HTA bodies cannot tolerate.

SaraBut Marcus, that's framing it wrong. The real issue isn't transparency — it's the lack of validation frameworks. HTA bodies don't have the expertise to assess AI models, and they're being asked to make decisions on technologies they fundamentally don't understand.

MarcusI don't think that's right. The problem isn't a lack of expertise; it's a lack of standardized validation protocols. The NICE AI guidance from last year established clear principles that could be adopted across jurisdictions.

SaraThat's the obvious read but you're missing the point here. The NICE guidance is aspirational, not operational. When was the last time you saw an HTA body actually implement those principles in a real-world assessment?

MarcusThe German IQWiG used AI in their assessment of a digital therapeutic for chronic pain last quarter. They applied a modified version of the NICE framework and published their methodology transparently.

SaraOne isolated case doesn't create a precedent. The real challenge is that AI models evolve faster than HTA processes. By the time an assessment is complete, the technology may have changed, making the assessment irrelevant.

MarcusThat's a valid concern, but it's not unique to AI. All evidence becomes outdated over time. The key is establishing dynamic assessment frameworks that allow for continuous evaluation rather than one-off appraisals.

SaraDynamic assessment sounds great in theory but creates operational chaos for HTA bodies already stretched thin. We need practical solutions, not theoretical frameworks that increase burden without clear value.

MarcusThe African Union's new RWE framework for oncology drugs was finalized last month. It's a significant departure from Western approaches, focusing on pragmatic trial designs and local epidemiology rather than head-to-head superiority against global standards.

SaraFinally some pragmatism in HTA. For too long, African markets have been forced to use evidence that doesn't reflect their populations or healthcare systems. This framework could be a game-changer for access.

MarcusI'm skeptical. The framework lacks specific methodological guidance on how to handle confounding factors in real-world settings that are more pronounced in African healthcare systems. It's well-intentioned but methodologically weak.

SaraThat's too generous to the document. It's not just weak — it's vague to the point of being unusable. There are no clear standards for data quality, no guidance on handling missing data, no benchmarks for what constitutes acceptable evidence.

MarcusThose are precisely the points that need to be filled in through implementation. Frameworks are always high-level; the devil is in the operational details that will emerge through application.

SaraThe devil is also in the commercial incentives. Pharma companies will interpret this vague framework in ways that serve their commercial interests, not necessarily what's best for patients or health systems.

MarcusThat contradicts what we saw with the malaria RWE initiative in West Africa — where manufacturers collaborated on standardized data collection protocols despite competing interests. Evidence there showed that transparency can coexist with commercial objectives.

SaraOne disease area with a long history of collaboration doesn't prove this can work across all therapeutic areas. Oncology is far more complex and commercially significant than infectious diseases in this context.

MarcusNICE published their updated threshold consultation document last week. They're proposing a revised threshold range, significantly higher than previous levels. The justification is increased productivity losses from premature mortality.

SaraThis is exactly what I've been predicting. The pandemic fundamentally changed how we value health outcomes, especially for working-age populations. The old thresholds no longer reflect the economic reality.

MarcusThat's the obvious read but you're missing the point here. The productivity argument is fundamentally flawed because it double-counts economic benefits that are already captured in other healthcare economic evaluations.

SaraDouble-counting? Marcus, we're talking about a global health crisis that caused unprecedented economic disruption. The traditional cost-effectiveness framework never accounted for societal-level economic shocks of this magnitude.

MarcusThat's precisely the problem. HTA should not be reactive to short-term economic shocks. The threshold increase lacks a solid methodological foundation and risks creating inconsistent value judgments across indications.

SaraMethodological purity is less important than relevance in a post-pandemic world. If the framework doesn't reflect current economic realities, it becomes irrelevant to decision-makers.

MarcusRelevance without rigor leads to arbitrary decisions. The threshold increase should be justified by systematic evidence, not just the perception of changed economic conditions.

MarcusThe FDA released new guidance on RWE for orphan drug development last month. It's a significant shift toward accepting real-world data as primary endpoints in rare disease trials when traditional endpoints are impractical.

SaraThis could transform development pathways for ultra-rare diseases. For conditions with fewer than 100 patients globally, traditional trial designs simply don't work. This recognition of RWE is long overdue.

MarcusI don't think that's right. The guidance creates more questions than answers. It doesn't establish clear standards for what constitutes acceptable RWE in this context, leaving manufacturers to navigate uncharted territory.

SaraUncharted territory is better than no territory at all. For decades, ultra-rare diseases have been stuck in regulatory limbo. This flexibility could bring treatments to patients who have waited far too long.

MarcusThat's the obvious read but you're missing the point here. The lack of clear standards creates regulatory uncertainty that could ultimately delay approvals as manufacturers struggle to design studies that will satisfy reviewers.

SaraUncertainty is better than certainty that leads to no treatments at all. The alternative is maintaining a status quo that denies patients access to potentially life-changing therapies.

MarcusThe real issue is that this guidance doesn't address the fundamental challenge of generating meaningful RWE in populations smaller than most clinical trials. It's a step forward but doesn't solve the core methodological problem.

MarcusWe're entering an era where HTA must balance methodological rigor with practical reality, especially as evidence generation becomes more complex and diverse.

SaraThe future of market access will be defined by how well we adapt our evidence frameworks to new technologies and diverse healthcare systems, not how rigidly we maintain traditional approaches.