The target trial framework and the HTA reckoning

Adigens Health helps sponsors design evidence strategies that meet the standard; from target trial specification to decision-ready confirmatory evidence packages. We will be at HTAi 2026 in Istanbul. Come and find us at our booth.Get in touch at info@adigenshealth.com.

The comparative effectiveness problem has long been HTA’s hardest question. One methodological framework is becoming its most compelling answer; for EU JCA, its timing couldn’t matter more.

Payers do not ask simply whether a new medicine works. They ask whether it works better; better than the treatments their health systems currently fund, better than the standard of care their patients currently receive. This comparative focus is the defining challenge of health technology assessment, and it is one that clinical trial evidence frequently fails to meet. Pivotal trials compare new medicines against placebo or long-superseded comparators; they recruit patients who bear little resemblance to those in routine clinical practice; they measure endpoints that regulators accept but payers reject. The gap between what trials demonstrate and what HTA bodies need to know can be substantial.

The target trial framework is, at its core, a disciplined answer to that gap. It reframes what real-world evidence can do for the HTA process; the framework asks researchers to specify, before any data are touched, the pragmatic target trial their study is meant to emulate: the comparators, the eligible patient population, the outcomes of interest, the follow-up period and the causal contrast under examination. The result is an analysis designed from the outset to answer the comparative effectiveness question that payers actually ask. That pre-specification is what separates defensible real-world evidence from data that merely looks like it.

What the target trial framework requires

Real-world data does not arrive organised around the questions that HTA bodies pose. Patients are not randomly assigned to treatments; clinicians make choices, and those choices correlate with outcomes. Confounding is endemic. Left unaddressed, it produces effect estimates that are unreliable and, in the hands of a sceptical assessor, indefensible.

The target trial framework addresses this not by eliminating confounding but by making the decisions that manage it visible and pre-specified. The protocol is written before the data are examined. Sensitivity analyses are defined in advance, not invented after the fact. The study can be audited against its own design. For an HTA reviewer tasked with judging whether an observational analysis has been conducted responsibly, this transparency is not incidental. It is the difference between evidence that can be trusted and evidence that cannot.

EU JCA and the future that doesn’t yet exist

Since January 2025, the EU Joint Clinical Assessment has been processing new medicines—oncology and advanced therapy medicinal products first, rare diseases to follow—through a single centralised evaluation. A methodology that satisfied one lenient national agency will no longer suffice. The joint assessment committee includes HTA bodies from across the EU; standards converge upward, not downward. But the deeper challenge posed by EU JCA is not methodological in the narrow sense. It is temporal.

JCA places real emphasis on early value assessment: structured consultation between companies and HTA bodies conducted during Phase II, sometimes earlier, to align the evidence being generated with the evidence that will eventually be assessed. The objective is clear and the difficulty is that it asks companies to envisage a future that does not yet exist. Which comparators will be clinically relevant at the time of submission? Which outcomes will matter to which member states? What will the standard of care look like in three or five years? No development team can answer those questions with certainty when Phase II data are still emerging.

This is precisely where the target trial framework earns its place, not just as a tool for analysing data, but as a planning instrument. Pre-specification does not require complete data. It requires a structured articulation of the comparative question: what trial would we need to run, if we could run one, to answer the HTA question we anticipate being asked? By committing that specification to paper early (i.e., during development planning, not at submission) companies create a living protocol that can be documented, updated transparently as the comparator landscape evolves, and presented to HTA bodies as evidence of methodological intent. The discipline of pre-specification does not lock a company into a single analytical choice. It creates an auditable record of how and why choices changed.

For early value assessment consultations, this matters in a specific and practical way. An HTA body asked to evaluate a company’s real-world evidence strategy at Phase II is not merely checking methodological boxes. It is asking if the company has thought carefully about the comparative effectiveness question it will eventually need to answer. A pre-specified target trial demonstrates that it has. It turns what might otherwise be a speculative conversation about future data into a structured dialogue about a defined evidentiary goal.

The PICO problem

The EU JCA process has invested heavily in PICO, the framework that specifies the Population, Intervention, Comparator and Outcome for each assessment. In principle, this is sound methodology. PICO forces assessors and companies to be explicit about the comparative question: for which patients, against which comparator, measured how. The extensive PICO scoping that precedes JCA submissions is a genuine advance over the informal and variable standards that characterised many national assessments.

The difficulty is that PICO is, in effect, a simplified version of the target trial framework. It captures the question; it does not specify how to answer it. The population maps to eligibility criteria; the intervention and comparator map to treatment strategies; the outcome maps to what is measured. These correspond to the first elements of a properly specified target trial. What PICO does not address is the harder part: the assignment mechanism, the time-zero definition, the follow-up rules, the causal contrast.  Neither does it address  the pre-specified analysis plan that determines how observational data will be used to emulate the trial the PICO describes.

This gap matters because the methodological pathologies that most commonly corrupt real-world evidence are invisible at the PICO level. Immortal time bias (i.e., where patients are credited with treatment exposure during a period in which, by definition, they have survived, producing artificially flattering effect estimates) cannot be identified from a PICO. Time-zero misalignment, confounding by indication, informative censoring and the conflation of intention-to-treat and per-protocol estimands are equally opaque. A submission can be fully PICO-compliant with the right population, the right comparator, the right outcome and still produce a treatment effect estimate that is systematically wrong. JCA’s focus on PICO gets a company to the right question. The target trial framework is what produces a trustworthy answer to it.

Rare diseases: the hardest case

The stakes are highest for rare diseases. Rare disease trials face well-documented challenges. Patient populations are small, natural history is poorly characterised, and comparators are often absent. Randomisation is sometimes impossible. Regulatory flexibility has accommodated these realities; external control arms, synthetic comparators and surrogate endpoints have all found a place in the rare disease development toolkit. JCA introduces the risk that evidence good enough for the EMA will not be good enough for the joint assessment committee.

The target trial framework does not resolve the data-scarcity problem. But it imposes structure on the data that does exist. External control arms and synthetic comparators, designed around a pre-specified target trial, carry considerably more evidential weight than those assembled after the fact. Natural history studies planned from the outset to populate a defined target trial form part of a comparative evidence strategy. The earlier that strategy is built, the more useful the data it generates.

The evidence bar has moved. The companies best placed to clear it are those that began pre-specifying their target trials before they had any data to analyse.