Embedded pragmatic clinical trials (ePCTs) test interventions in routine healthcare settings using existing infrastructure, such as electronic health records (EHRs) and frontline clinicians. Unlike traditional trials with strict eligibility criteria and controlled conditions, ePCTs enroll broader, more representative populations to assess whether interventions work in everyday clinical practice. The FDA supports this approach through initiatives such as the Oncology Center of Excellence’s Project Pragmatica and CDER’s Streamlined Trials Embedded in Clinical Practice (STEP) demonstration project. By integrating research into routine care, sponsors can capture more diverse populations, potentially shorten timelines, and reduce patient burden. However, operational complexity can compromise trial execution even when the intervention itself is effective.
Pilot ePCTs test operational feasibility before sponsors commit substantial resources. Their purpose is to determine whether patient recruitment, clinical workflows, intervention delivery, and outcome collection can function reliably in real-world settings, not to establish treatment efficacy or effectiveness. This distinction has direct regulatory implications. An ePCT is a trial design approach, not a separate FDA approval pathway, and its pragmatic nature does not lower the evidentiary standard for approval. Explanatory trials assess treatment effects under controlled conditions, whereas pragmatic trials evaluate performance in routine practice. Sponsors seeking approval, new indications, or labeling changes must ensure their study design supports the intended regulatory claim.
For drugs and biologics, ePCTs may contribute to an investigational new drug (IND) program, a new drug application (NDA), a biologics license application (BLA), supplemental applications, or certain postmarketing requirements. The FDA’s real-world evidence (RWE) framework allows consideration of real-world data for appropriate regulatory decisions, including new indications for approved drugs. However, routine healthcare data are not automatically adequate evidence; their relevance, reliability, and ability to support valid conclusions must be established. For medical devices, real-world evidence may support applicable 510(k), De Novo, or premarket approval (PMA) submissions, depending on the device, intended use, risk profile, and regulatory question. Embedding a study in routine care does not exempt sponsors from applicable safety and effectiveness requirements.
Sponsors should establish their regulatory strategy before conducting the pilot, defining the intended claim, population, comparator, endpoints, follow-up, statistical analysis, and required evidence. They should assess whether randomization, blinding, or additional data collection is necessary to minimize bias. Early FDA engagement can help identify concerns with study design, endpoints, data sources, safety monitoring, and inspection readiness. CDER’s STEP demonstration project offers eligible sponsors an opportunity to discuss pragmatic trial approaches with FDA, subject to current program requirements. A successful pilot establishes feasibility, not regulatory sufficiency. It does not independently demonstrate efficacy or effectiveness or guarantee that a subsequent trial will support approval. The full-scale study must still generate reliable evidence appropriate for its intended regulatory purpose.
A workshop drawing on five years of experience and 23 pilot studies funded by the National Institute on Aging (NIA) IMPACT Collaboratory identified three essential areas for pilot ePCT design. First, sponsors should determine whether existing data sources, such as EHR diagnosis codes and pharmacy databases, can reliably identify eligible patients and support enrollment without extensive manual screening. They should also assess whether these data adequately capture patient characteristics, treatment exposure, and relevant clinical history. Second, sponsors should evaluate whether clinicians can integrate study procedures into routine workflows without disrupting care. The RE-AIM framework (Reach, Efficacy, Adoption, Implementation, Maintenance) can help assess reach, clinician adoption, and implementation, revealing barriers that could compromise protocol adherence. Third, sponsors must establish whether outcomes can be captured accurately through existing health records without increasing administrative burdens. EHR data may contain missing variables or incomplete entries compared with research-coordinator-collected data. Sponsors should assess completeness, validity, coding consistency, traceability, and missing-data risks, and address confounding and outcome ascertainment in the analysis plan.
A quantitative traffic-light system can guide decisions. Green means predefined feasibility thresholds are met and the study can proceed. Amber indicates that modifications are needed before proceeding. Red means feasibility endpoints have failed and the approach is not viable in its current form. Interviews with physicians, nurses, and patients complement quantitative findings by explaining why workflows succeed or fail. IRBs and regulators recognize ePCTs’ potential for greater external validity and systematic identification of operational risks. However, simplified procedures must not compromise patient privacy, blinding where required, randomization, safety reporting, or data integrity. Sponsors must address applicable informed-consent requirements, IRB review, and IND or investigational device exemption (IDE) obligations, as appropriate. Clear responsibilities for clinical care, research oversight, data access, adverse-event reporting, protocol deviations, and monitoring should be established before trial initiation.
Pilot ePCTs help identify operational bottlenecks before they become costly trial failures. They assess whether patients can be identified, clinicians can implement study procedures, and outcomes can be captured reliably. Experience from the NIA IMPACT Collaboratory highlights the value of early testing in protecting resources and strengthening healthcare-system partnerships. For sponsors pursuing FDA approval or labeling changes, operational feasibility must be paired with a clear regulatory strategy and appropriate evidentiary standards. Test feasibility before scaling, engage FDA early when regulatory use is intended, and design the full-scale trial around the evidence required for the specific regulatory decision. This approach reduces avoidable execution risks while preserving scientific rigor and the potential to improve real-world patient care.
