Oral GLP-1s, AI Dominance, and Digital Health Parity

Episode 10 — 27 May 2026 EMA approves first oral GLP-1 for weight management while AI rises to top HEOR trend. NICE expands technology appraisals to digital health technologies and FDA raises real-world data standards. Hosted by Marcus and Sara. Access Brief — Daily briefings on HEOR, HTA strategy and the evidence access landscape. Full transcript and sources at outcomes-analytica.no/podcast

Transcript

SPEAKER_00 0:00

Welcome to Access Brief, the daily AI podcast on HEOR, HTA, and market access. I'm Marcus with Sarah. Today, EMA's first oral GLP1 approval for weight management, AI dominating HEOR trends, and NICE putting digital health technologies on legal par with medicines. Let's get into it.

SPEAKER_01 0:18

Starting with EMA's CHMP meeting from May, eight medicines recommended for approval, but the standout is WIGOV tablets. Semaglutide is the first oral GLP1 receptor agonist for weight management. Alternative to weekly subcutaneous injections.

SPEAKER_00 0:33

The evidence bar for oral formulations is fascinating here. Same indication as the injectable. Obesity or overweight with weight-related comorbidities, plus diet and physical activity. But the HEOR implications are substantial. Adherence profiles, administration costs, patient preference utilities.

SPEAKER_01 0:52

I'm skeptical this moves the needle as much as everyone thinks. Yes, oral is convenient, but we're still talking about the same patient population, same clinical endpoints. The real question is whether payers will see differentiated value or just formulary complexity.

SPEAKER_00 1:07

But you're missing the adherence multiplier effect. Oral dosing fundamentally changes the cost-effectiveness calculation. Better adherence drives better outcomes, which drives better ICERS. That's not formulary complexity, that's legitimate clinical differentiation.

SPEAKER_01 1:22

Fair point on adherence, but I'd want to see head-to-head persistence data before assuming oral automatically equals better outcomes. Convenience doesn't always translate to compliance, especially with GLP1 side effect profiles.

SPEAKER_00 1:35

Moving to ISPOR's 2026 to 2027 trends report, AI jumped from number three to the top spot. The expansion is comprehensive. Systematic literature reviews, data set structuring, accelerated analysis.

SPEAKER_01 1:48

This isn't just about efficiency anymore. AI performing systematic reviews in a fraction of the time fundamentally changes how we build evidence dossiers. Question is whether HTA bodies are ready for AI-generated evidence packages.

SPEAKER_00 2:02

The speed advantage is undeniable, but I'm concerned about the quality control mechanisms. Systematic reviews are the foundation of evidence synthesis. If AI introduces systematic bias or misses critical studies, we're building on flawed foundations.

SPEAKER_01 2:17

But that assumes human systematic reviews are error-free, which they're not. At least AI bias is potentially identifiable and correctable. The bigger issue is whether regulatory bodies will accept AI-derived evidence at face value or require human validation layers.

SPEAKER_00 2:33

That's exactly my point. If we need human validation anyway, where's the efficiency gain? We might be trading speed for credibility without realizing it.

SPEAKER_01 2:42

NICE announced a significant expansion. Technology appraisals now cover digital health technologies, putting them on legal par with medicines from April 2026.

SPEAKER_00 2:52

This is overdue. Digital health technologies have been in regulatory limbo, clinical evidence requirements unclear, cost effectiveness thresholds undefined. Legal parity should clarify the evidence bar.

SPEAKER_01 3:04

Should, but will it? Digital health technologies don't fit traditional QLI models cleanly. How do you measure cost effectiveness for a diabetes app versus Metformin? The clinical endpoints are fundamentally different.

SPEAKER_00 3:17

The methodology challenges are real, but that's not an argument against standardization. It's an argument for better frameworks. NICE has been working on this for years. Putting digital technologies through the same rigorous process should improve evidence quality.

SPEAKER_01 3:33

I disagree. One size fits all technology appraisals might work for small molecules, but digital technologies iterate constantly.

SPEAKER_00 3:46

Finally, FDA issued complementary real-world data guidance between December 2025 and March 2026. One for medical devices, one for drugs and biologics, the philosophy is clear. Lower privacy barriers, higher requirements for data structure and provenance.

SPEAKER_01 4:02

This is the regulatory maturation we needed. Real-world evidence has been the Wild West. Inconsistent standards, questionable data quality. Raising the bar should improve evidence credibility across the board.

SPEAKER_00 4:15

That's the theory. But in practice, we're creating a two-tiered system. Companies with sophisticated data infrastructure will thrive. Smaller innovators might be locked out entirely.

SPEAKER_01 4:25

That's market consolidation, not regulatory failure. If you can't meet basic data standards, should you be making regulatory claims based on real-world evidence?

SPEAKER_00 4:34

Clear standards, but the implementation timeline matters. If FDA doesn't provide adequate transition periods, we risk losing valuable innovation from companies that have good science but limited data infrastructure capabilities.

SPEAKER_01 4:51

Whether that's progress or gatekeeping depends on where you sit.

SPEAKER_00 4:55

And whether the evidence quality improvements justify the access barriers we're creating. Back tomorrow on Access Brief. Show notes at outcomes dash analytica.