About the Author:
Jeff Kozloff brings more than 25 years of leadership experience across pharmaceutical, clinical, and commercial software. He has scaled multiple investor-backed companies through critical growth phases, including serving as CEO of Verilogue and TrialScope. He also currently serves as Chairman of the Board at ProofPilot.
On April 28, 2026, the FDA announced something that, if the trajectory holds, will create a meaningful divide between research organizations over the next decade: a real-time clinical trial model in which structured data flows continuously from research sites to regulators, rather than arriving in periodic batches months after the fact.
Two proof-of-concept trials are already running under this model. AstraZeneca’s and Amgen’s trials are transmitting safety signals and endpoint data to FDA reviewers in real time, through a pilot infrastructure the FDA selected for the initiative.
The operational question this raises, one the industry has not yet answered in any systematic way, is what continuous, structured data transmission actually requires of a research site’s infrastructure. For research sites, that makes the question fairly immediate. If the FDA’s vision became the standard tomorrow, could the infrastructure they have today support this kind of continuous data exchange?
Looking at what this model requires, then looking at how most sites operate today, reveals several areas that deserve attention now. The organizations best positioned for what comes next will be the ones that start building for continuous data exchange today, before it becomes an expectation.
What Real-Time Clinical Trials (RTCT) Could Require from Sites
The RTCT model changes more than the speed of reporting. It changes how data moves through a clinical trial. Under the current model, data moves in discrete batches: from site to sponsor, then sponsor to FDA, with review cycles often measured in months. With RTCT, structured data is captured at the source, evaluated algorithmically against FDA-defined criteria, and transmitted continuously. The data must be usable as it is generated, rather than cleaned up, reconciled, or assembled for submission later.
That continuous visibility is central to the FDA’s vision. Then-FDA Commissioner Marty Makary described RTCT as a model where, “We are boldly advancing a modern approach whereby FDA scientists can view safety signals and endpoints in real time as a trial progresses.” FDA Chief AI Officer Jeremy Walsh suggested the approach could reduce overall clinical trial time by as much as 40 percent. While that figure is still a projection, it gives a sense of the scale of what the agency believes could be possible.
The FDA hasn’t yet spelled out the technical requirements for participating in the pilot. The RFI asked the industry for input, and detailed protocols for the proof-of-concept trials haven’t been made public.
What we do know is that the model works. The FDA has confirmed it received and validated real-time safety signals from AstraZeneca’s trial, demonstrating that continuous signal sharing between a research site and a federal regulator is technically feasible today, not just theoretically possible.
Neither the FDA nor the pilot participants have published a technical specification for future RTCT participants. But the RFI itself, combined with what a continuous exchange requires by definition, points to a consistent set of capabilities a site’s infrastructure would need:
- Electronic source data capture. Trial data needs to originate digitally, whether it is captured directly in eSource, comes from an EHR, or originates in another electronic clinical system.
- Research-ready data. Data coming from an EHR or another source still needs to be identified, mapped, normalized, and validated so the relevant data points can move reliably into the research workflow.
- Continuous data availability. Once captured and prepared for research use, data needs to be available as part of an ongoing flow rather than dependent on periodic exports and manual preparation.
- Interoperable, structured outputs. Data has to move reliably between systems. Standards such as CDISC ODM and FHIR provide mechanisms for structured clinical data exchange, although the FDA has not specified which standards a future pilot will require.
- Traceability from source to transmission. Data needs a clear, auditable path from its initial capture through downstream systems and regulatory reporting.
- Minimal manual intervention. A continuous model depends on infrastructure that can move data from capture through evaluation and transmission without repeated manual preparation, reconciliation, or transfer.
Taken together, these capabilities put new importance on what happens at the beginning of the research data flow. Whether data originates in an EHR, is captured directly in eSource, or comes from another clinical system, it has to be structured, research-ready, and traceable before it can move reliably downstream. That is where eSource becomes particularly relevant.
eSource doesn’t create a real-time clinical trial on its own. But it does address an important part of the equation: capturing trial data electronically and in a structured format from the beginning. That reduces reliance on paper, manual transcription, and the work required to assemble usable data later.
Continuous vs. Assembly-Dependent eClinical Infrastructure
It is worth pausing on what continuous data transmission means in practice at the infrastructure level, because it describes a fundamentally different kind of system than most research organizations currently operate.
In an assembly-dependent model, information is captured across systems and brought together when it is needed. That can mean manual transcription between platforms, scheduled data pulls, reconciliation, and coordinator effort to prepare information for downstream review. The process can be fully compliant and produce high-quality data, but movement between systems still depends heavily on people.
A continuous model changes that expectation. Defined data needs to be structured, traceable, and available for downstream use much closer to the time it is generated. Instead of repeatedly preparing information to move to the next system or stakeholder, the infrastructure has to support that movement as part of the normal data flow.
That distinction matters for RTCT. The FDA isn’t simply exploring whether existing reporting processes can happen faster. It is testing what becomes possible when regulators can receive defined clinical trial signals while a study is underway.
Much of today’s eClinical infrastructure wasn’t designed with that kind of continuous exchange in mind. It was built for a regulatory environment centered on periodic review, reconciliation, and submission. RTCT is an early signal that the assumptions behind that infrastructure may be starting to change.
This kind of shift has precedent within FDA itself. In 2004, the agency’s Process Analytical Technology guidance pushed drug manufacturers away from a “quality by testing” model, pulling a sample and testing it in a lab after a batch was made, toward continuous, in-line monitoring during production. The premise was that quality had to be built into the process as it happened, not confirmed afterward. That transition took manufacturers years and significant capital, since it required new physical instrumentation built into production lines themselves.
RTCT points research sites toward the same principle, but a considerably lighter lift. The infrastructure it requires, structured, connected, continuously available data, is largely a software question rather than a hardware one. For sites already running a connected CTMS and eSource platform like RealTime’s, much of that foundation already exists, and extending it to support continuous data flow looks more like configuration than construction.
3 Infrastructure Questions for FDA Real-Time Clinical Trial Readiness
The FDA has not published a technical checklist for RTCT participation, so there is no definitive test of whether a site’s infrastructure is “ready.” But the proof-of-concept trials and the RFI give research organizations enough information to start examining how their current data environment would perform under a more continuous model.
Three questions stand out.
1. Is source data electronic when it is generated?
Continuous exchange starts with data that is available electronically. When trial data is first captured on paper and entered into an electronic system later, a manual step already exists between the clinical event and the digital record. That makes continuous downstream use more difficult. For sites still dependent on paper-first workflows, source data capture is the logical place to begin evaluating readiness.
2. Can source data move into sponsor EDC systems without being recreated?
Capturing source data electronically is only the first step. If coordinators still have to re-enter data into EDC, prepare files for transfer, or reconcile records across platforms, the workflow remains dependent on manual intervention.
Structured data exchange changes that. When eSource data can be mapped and transferred into sponsor EDC systems using established data standards, sites spend less time re-entering and reconciling the same information across systems. For sponsors, data becomes available for downstream use with fewer manual handoffs, while reviewers can spend less time comparing records across systems and more time investigating exceptions, safety signals, and data that requires clinical judgment. That distinction becomes especially important in a continuous model.
3. Is data available when it is needed, or assembled when someone asks for it?
In many research environments, the data exists, but someone still has to export it, reconcile it across systems, prepare a file, or pull information together before a sponsor or reviewer can use it.
That’s the practical difference between continuous and reconciliation-dependent infrastructure. In a continuous model, defined data moves as part of the workflow. In a reconciliation-dependent model, that movement depends on someone manually preparing, transferring, or reconciling it first, and everything takes longer because of it.
None of these questions constitutes an FDA eligibility requirement. Rather, they offer a way to assess whether today’s site infrastructure can support the agency’s direction: structured, traceable data moving from source to downstream review with substantially less delay and manual intervention.
This shift also has beneficiaries beyond the Regulator, provided sites use it to replace manual work rather than add to it.
- Patients enrolled in a trial benefit when a safety signal is visible while the study is underway, since it can prompt a protocol amendment or dose adjustment in time to matter, not months after the fact.
- Investigators benefit from cleaner, more current data when making clinical judgment calls about the patients in front of them.
- Site coordinators, who today absorb much of the manual work of preparing data for downstream use, stand to gain the most directly: less time re-entering the same information into multiple systems, chasing paper source, and responding to queries that surface weeks after a visit.
The caveat: a continuous model only delivers that relief if it replaces manual reconciliation rather than simply layering a steady stream of monitoring flags on top of it.
Final Thought: What RTCT Ultimately Changes
Moving data faster only works if the data is reliable when it begins moving.
Today’s clinical trial workflows leave time between data capture and regulatory review for monitoring, queries, reconciliation, and correction. RTCT shortens that window. When defined safety and efficacy signals can be evaluated while a trial is underway, the quality and timeliness of source data become more consequential.
That is why RTCT is not simply a regulatory reporting development. It reaches all the way back to how data is captured at the site, how easily it can move into sponsor systems, and how much human effort is required along the way.
No one knows yet exactly how far or how quickly the FDA will take this model. But the direction is worth paying attention to. Clinical trial data is moving toward being more structured, more connected, and available closer to the time it is generated.
For research organizations, that makes the work happening today around eSource and interoperability more than an efficiency exercise. It is part of preparing for a clinical trial environment in which data may increasingly need to move as the research happens, in real-time.