The clinical research ecosystem has spent more than a decade saying the right things about patient engagement. Health authorities, patient groups, sponsors, academic institutions, and a host of life science companies have all committed — publicly and repeatedly — to making it a priority. And yet the collective efforts across these stakeholders have remained fragmented.
Most stakeholders agree on its importance and have similar views of what good looks like. However, this has spawned a series of well-intentioned initiatives aimed at improving specific aspects of the value chain while missing the bigger picture. We have been tending to a handful of trees and ignoring the health of the forest.
When done well, real patient engagement begins at clinical development planning, shapes trial design from the start, and extends beyond the end of the study. That standard has proven harder to meet than to articulate. Cycle times have not improved, costs have continued to climb, and patients with chronic disease are still being asked to organize their lives around trial schedules rather than the other way around.
Two recent regulatory developments create a meaningful opening to do better. ICH E6(R3), the updated Good Clinical Practice guideline, codifies a quality-by-design (QbD) approach that includes a structured expectation for real stakeholder engagement, including patients. Alongside it, ICH E19 creates a path to fit-for-purpose selective safety data collection, which has the potential to lower burden across the clinical research enterprise.
Neither will fix the problem by mandate alone, but together, they remove some of the most durable excuses for not doing things differently and create an opportunity to collaborate with regulators right now, before local and regional health authorities finish defining their own implementation requirements.
Patient advisory boards have proliferated, advocacy partnerships have become standard practice, and the language of patient-centricity appears in nearly every development strategy. What has been slower to change is the underlying incentive structure that determines when patient input enters the process and whether it is really used.
The incentive problem
Speed-based milestones reward getting to first site initiation and first patient enrolled, but offer no real incentive for developing a protocol with representative eligibility criteria, fit-for-purpose data collection, or a design that reflects how patients flow through a healthcare system. Input from patients, investigators, and site staff that could meaningfully improve trial feasibility — and therefore its success — needs to carry appropriate weight before key decisions are made. This type of stakeholder engagement is essential in order to achieve a true QbD approach.
Site selection compounds the problem. Community-based practices with genuine access to underrepresented patient populations are often excluded from consideration because they lack the performance history required to get on the list. This creates a circular barrier, reinforced by high investigator turnover and a shortage of established research sites.
E6(R3): Making engagement accountable
As health authorities implement E6(R3), sponsors will need a structured mechanism to demonstrate how they engaged key stakeholders, including healthcare providers and patients, and whether that input was incorporated.
For a patient managing a chronic condition, a trial that requires taking time off work or missing family events is not a mere inconvenience. It is a reason to decline participation when the alternative is simply being treated off-study. Reducing that friction requires understanding it first, and that understanding has to come from patients before the key decisions are made.
E19: A more deliberate approach to data collection
E19 encourages sponsors to approach data collection more deliberately and with greater discipline, which runs counter to decades of instinct in clinical research. The guidance provides a framework for determining when selective safety data collection is appropriate, based on where an asset sits in development and the regulatory context around it.
The primary disincentive has always been perceived regulatory risk. Sponsors have been reluctant to collect less when uncertain whether regulators will accept it. What E19 offers is greater clarity in separating real regulatory risk from perceived regulatory risk.
In July 2025, the FDA's Center for Drug Evaluation and Research (CDER) released a white paper through its Center for Clinical Trial Innovation (C3TI) on selective safety data collection, and launched a Selective Safety Data Collection (SSDC) Demonstration Project inviting sponsors to bring forward late-stage trials where safety data collection could be appropriately streamlined. Participants have the opportunity to engage directly with CDER subject matter experts on study design and a fit-for-purpose inspection approach, in exchange for sharing lessons learned along the way. As those conversations mature and health authorities engage more with sponsors on what the true risks are, the space for responsible simplification grows.
Pragmatic trials: Moving past all or none
Decentralized and pragmatic trial approaches have been in serious discussion long enough that early enthusiasm has given way to a more measured assessment of what scales. One early misstep was framing these approaches as binary: a fully decentralized trial or a conventional one, with little space for anything in between.
What the evidence is beginning to show is that adoption of pragmatic elements can deliver meaningful benefits without requiring a wholesale redesign of trial infrastructure. The RECOVERY trial, for example, used broad eligibility criteria, a one-page consent process, and outcomes drawn from routine NHS records to show within months that dexamethasone reduced COVID-19 mortality. It went on to resolve questions about several other therapies using the same streamlined approach.
The Pragmatica-Lung trial is applying the same thinking in oncology, stripping out most of the eligibility criteria typical of registration studies and limiting required data to overall survival, basic safety, and treatment information. Both efforts highlight what becomes possible when stakeholders collaborate in different ways, showing that broad eligibility, focused data collection, and integration into routine care can help answer important questions more quickly and in a more representative way. While neither was a perfect execution, both generated lessons worth building on.
Data acquisition remains an unresolved challenge. Using point-of-care data to answer regulatory questions depends on interoperability standards and harmonized privacy frameworks that do not yet exist consistently across regions. Addressing this gap will require coordinated action from technology companies, health authorities, sponsors, and government agencies. That collaboration is complex, but it is now beginning to take shape.
What a successful 2026 looks like
Progress in 2026 will be visible in specific places: QbD approach to protocol development under E6(R3) that can point to documented patient input, studies launched with explicit selective safety data collection under E19, expanded use of demonstration programs run through CT3I, and continued development of supplemental and virtual control arms that reduce unnecessary randomization to placebo.
These are important signals, and they matter because the alternative is increasingly difficult to defend. The risks of continuing as we are — slow cycle times, rising costs, trials that exclude the populations they most need to serve — are real and well documented. The real question for 2026 is whether the organizations that shape clinical research have the courage to act on what is already clear. Patients have been willing to participate. The burden of proof now falls on the industry to meet them where they are.
