ISPOR 2026 in Review

ISPOR 2026 provided a bird’s eye view of many factors that are driving the evolution of healthcare systems everywhere. Here are several themes that stood out to us.

Is it all about the price? US pricing policy

A session titled “Value-based pricing for CMS drug price negotiation – mission impossible?” showed us how the Inflation Reduction Act (IRA) is only loosely connected to value-based pricing. Ceiling prices and the mechanics of arriving at a Maximum Fair Price can be a barrier to negotiation outcomes reflecting “pure” value-based pricing derived from incremental benefit, patient outcomes, or societal value. Furthermore, what qualifies as “fair”, and who is it “fair” for – Centers for Medicare & Medicaid Services (CMS), patients, manufacturers, or the overall health care system?

Stakeholders may interpret fairness differently so it is important that manufacturers approach IRA negotiations with the most robust evidence package possible to maximize favorable negotiation outcomes. Demonstrating a deep understanding of clinical pathways, clear comparative clinical benefit, defensible modeling, patient-centered outcomes, and credible real-world evidence is important.

Continuing the focus on drug pricing, another session titled “Are drug prices rising too fast? Rethinking inflation benchmarks through the lens of value, access, and innovation” highlighted how waves of innovation increase average drug spending as newer therapies enter the market at higher price points. The discussion also emphasized that patient affordability is often driven more by benefit design – deductibles, co-insurance, and specialty-tier cost sharing – than by changes in drug prices alone. Speakers stressed that “inpatient and outpatient care spending in the US still far exceeds prescription drugs” which highlights how rising healthcare expenditure is a health system challenge and not purely a drug pricing one.

Global pricing and access

A session titled “MFN meets IRP: How US pricing reform reshapes global market interdependencies” focused our attention on global pricing and access. Despite Most Favored Nation (MFN) policy not yet being codified, it is clear the US will continue to move toward some sort of value-based pricing at a previously unseen level. Speakers noted that “on the surface, MFN is a pricing policy but operationally it is an entire industry redesign” as manufacturers need to converge traditionally siloed pricing and access strategies in the US and rest of world (RoW) markets. Ex-US market launch delays were cited as a likely consequence of MFN although this relied on the premise that in the absence of ex-US reference market price data, the US will allow prices to remain status quo. If that is not the case then manufacturers pursuing this strategy may be risking the opportunity cost of delayed launch ex-US revenue and lower US pricing. And if it is the case and delayed RoW launches are a viable strategy to maintain US pricing at current levels, then it raises the question – Is MFN really about drug pricing at all?

Despite the uncertainty around exactly how MFN will play out, it is clear that traditional approaches to pricing and launch strategy are no longer sufficient. Manufacturers need to be thinking about pricing strategy as a truly global effort (i.e. not US vs Global) with a focus on global net pricing and where patient access is at the forefront of decision making. The opportunity for innovation is high.

Broader discussions at ISPOR 2026 highlighted the growing importance of health economics and outcomes research (HEOR) and the responsibility to generate good evidence on the value of medicines to help drive pricing decisions and avoid a ‘race to the bottom’ spiral to the lowest price regardless of value.

Gene therapies: expensive but worth it?

Gene therapies represent a new frontier in treatment, costing $1–3 million per patient but often delivering curative outcomes that most people agree will justify this costly individual investment. However, the current financing model is not designed to accommodate the science behind these treatments, and practical challenges remain – many gene therapies require patients to be hospitalized for around 100 days. While expensive new drugs continue to enter the market, some cost pressure is being offset by patent expirations on older brands.

Several common myths distort the drug pricing conversation. Contrary to popular belief, new drug launches are not the primary driver of US prescription spending growth, and medicines can generate meaningful clinical and economic value. The benefits delivered per patient in terms of health outcomes and/or avoided costs, are increasing faster than general price levels. The 10.6% increase in drug spending from 2024–25 was largely driven by obesity drugs and other protected brands, rather than broad increases in prices for all drugs. Out-of-pocket costs for prescription drugs remain low for most Americans.

Measurement challenges complicate policy discussions. Andrew York, of the Maryland Prescription Drug Affordability Board, noted that insurers report drug costs comprise 20–30% of their spending, but this is based on gross figures. Net spending would be more accurate, but it is typically unknown – even insurers often do not know the true net price. The wholesale acquisition cost (WAC) or manufacturer’s list price, is “functionally meaningless” but remains the best available benchmark, and co-insurance is tied to this inflated figure.

For HEOR, the opportunity is significant. Inpatient and outpatient care present enormous need for evidence to inform value-based care. A US health technology assessment (HTA) framework is needed to guide value assessment, but to be effective it must incorporate real-world evidence (RWE) and patient-relevant outcomes.

HEOR: the compass to right decisions?

An interesting debate came up at a session titled “HEOR as a compass, not a roadblock: How and when development-stage biotech companies should leverage HEOR.” There is growing openness to using HEOR insights much earlier in development. Early analyses (e.g. preliminary models and RWE) can shape views on feasible price, key value drivers, comparator strategy, and addressable population. Yet HEOR remains underutilized upstream and the question of whether a therapy can achieve reimbursement at a commercially viable price is still often left too late.

The point is not that senior stakeholders need HEOR to learn the mechanics; they need it to reduce uncertainty around the decisions that drive value. Early HEOR should therefore lead with the implications for the choices that matter most – what comparator strategy will be credible, which populations are worth pursuing, what evidence will be required to secure access, and what pricing is realistically defendable. Presented in this way HEOR can function as a strategic compass that strengthens decision making – not as a technical gatekeeper that slows down progress.

AI’s versatility continues to surface

As we expected, a lot was said about Artificial Intelligence (AI) at ISPOR 2026. The session: “The GenAI paradox for qualitative evidence summarization: Exploring real-world use cases and validation frameworks for understudied but impactful use cases” showcased where Generative Artificial Intelligence (GenAI) can realistically add value in qualitative/patient experience evidence – while emphasizing that the risks are greatest when AI is used without strong input controls and human verification.

For us the key points were these:

• Input quality drives output quality: the value and reliability of AI summaries are only as good as the underlying qualitative data and how it is collected, structured, and documented
• AI can support patient experience and qualitative evidence work, particularly in collecting, consolidating, and summarizing information
• Review by humans is non-negotiable: validation and verification must be designed into the process from the beginning – not added as an afterthought at the end
• AI can help to surface patient-relevant experiences that are not well-described in published literature, positioning it as one tool among many rather than a replacement for primary qualitative research. Social media and other real-world digital sources can generate incremental insights, especially for understudied experiences not captured in traditional literature, and can be used to inform subsequent non-AI qualitative work – for e.g. interview guides, discussion guides, hypothesis generation
• Using AI chatbots to conduct in-trial patient interviews is a stretch too far though: AI can support preparation such as drafting questions, but it does not have the sensitivity needed for patient interviews. The potential for errors, omissions, or hallucinations is too high and could compromise the fidelity of the data and overall patient experience participating in an interview. The patient voice is too valuable to justify introducing that level of uncertainty

Never stop listening to the patient voice

Whether the population was adult or pediatric, evidence of the patient-centric skeleton of health care was everywhere at ISPOR 2026. Sessions were dedicated to rare and ultra-rare disease states with at least one considering “Is it worth building disease-specific patient-reported outcomes (DSPRO) value propositions for orphan and innovative therapies? Bridging economic models and patient lived experience.” Speakers here stressed both the return on investment for study sponsors and the “return on burden” concept for benefits for patients and caregivers: while they are willing to complete measures, including multiple or lengthy ones if the clinical and research benefits are clear, generic tools like EQ-5D, SF-36, or even widely used pediatric tools can be irrelevant or upsetting for parents when they ask about abilities a child cannot perform, while still being requested by HTA bodies. The panel converged on a practical decision logic: start with the construct(s) that matter most and what the therapy is expected to change; use an appropriate generic/cross-cutting measure (including PROMIS item banks) when it truly fits; avoid redundancies across measures; develop a DSPRO when existing instruments cannot credibly capture meaningful change in the target population.

At another panel titled: “Patient experience data as a non-negotiable: Can US payers afford to stay behind?” the clearest message was that patient experience data (PED) including patient-reported outcomes (PROs), quality of life, preferences, and patient-generated/digital data, has shifted from “nice to have” to an expected part of the evidence package. This is driven by regulatory evolution – notably the 21st Century Cures Act and the Food and Drug Administration (FDA) Patient-Focused Drug Development (PFDD) guidance – and increasing emphasis from European HTA bodies on patient-relevant outcomes in value/price discussions. Panelists argued the core issue is no longer whether PED matters, but how to make it credible, decision-useful, and actionable for payers who may be uncertain how to weigh PED, concerned about noise/subjectivity, and inconsistent in applying it. A recurring theme was “pull-through”: PED is often generated for regulatory purposes but not translated into payer-relevant narratives and endpoints (e.g. connecting fatigue improvements to healthcare utilization, productivity, caregiver burden). The panel emphasized early, frequent, cross-functional planning (regulatory, HEOR, access, clinical) and sustained patient involvement to ensure the right concepts are measured and communicated through the full product lifecycle. 

For manufacturers the PED strategy should not simply be added on late in the drug development process or confined to an appendix in the value dossier. Manufacturers will increasingly need an end-to-end PED plan that starts pre-Phase III, aligns evidence generation across regulatory and access needs, and explicitly maps PED endpoints to payer decision levers (utilization management, contracting, total cost of care). Operationally, that means investing earlier in fit for purpose Clinical Outcome Assessments (COAs). Meaningful quality of life (QoL) improvement requires strong tools, framing, and early alignment. Potential enablers include novel PROs (especially in rare disease) and possibly digital health technologies (DHTs)/wearables to support more comparable outcomes.

As these sessions demonstrated, PED is more than a module in a dossier and can be used to build disease context, endpoint strategy, and a coherent patient-centered value narrative. PED can help to define meaningful change thresholds that resonate beyond labeling and produce “payer-ready” PED evidence modules that translate patient impacts into economic and utilization consequences. There is also a stakeholder engagement opportunity: structured, iterative touchpoints with the FDA (and when feasible, payers/HTA) to de-risk endpoint strategy, while minimizing administrative burden on patients and avoiding perceptions of bias/conflict with patient advocacy groups.

Final thoughts

As this year’s sessions made clear, pricing – and the increasing interplay between US and RoW policy – continues to be a hot topic, with key themes including whether the IRA offers “true” value-based pricing and the global implications of the MFN policy. Evidence – in particular credible comparative evidence, patient-centered evidence, and fit for purpose RWE – is becoming the most important asset that manufacturers can still control. Tools like GenAI can speed up the pace of getting work done but still only if supervised and controlled by humans. We appreciate the thoughtful insights and robust discussions that emerged from this year’s meeting and look forward to continuing the conversation in our ongoing collaborations with key stakeholders to develop tightly integrated solutions that drive meaningful impact.