ClinicalTrials.gov Results Reporting: What the FDA's 2026 Enforcement Push Means for Medical Researchers

In March 2026, the U.S. Food and Drug Administration sent compliance reminder letters to more than 2,200 clinical trial sponsors and individual researchers, flagging over 3,000 registered studies with no results posted to ClinicalTrials.gov as required by federal law. The agency's own internal analysis found that approximately 30 percent of studies meeting the legal definition of “applicable clinical trial” under the Food and Drug Administration Amendments Act had never submitted results at all. For researchers who assumed that ClinicalTrials.gov was primarily a logistical registration tool managed by a hospital compliance office, this is worth more than a passing read.

The compliance obligation in question has existed since 2017, when the FDA finalized its implementing rules under FDAAA 801, the statute that Congress passed in 2007. Its persistence as a large structural gap almost a decade after the Final Rule took effect signals something more than administrative inertia. Industry-funded trials have gradually improved their reporting rates over that period. Academic and government-funded trials have not kept pace. And because journals increasingly compare published outcomes against the registry records that the ICMJE has required to be posted since 2005, the downstream consequences have moved from regulatory risk into editorial risk.

These two risks, the FDA's reporting requirements and the journal's registered-outcome expectations, are not parallel tracks that researchers can manage separately. What a researcher posts to ClinicalTrials.gov becomes part of the permanent public record against which journal editors and peer reviewers compare the eventual manuscript. Getting the relationship between these two obligations wrong creates problems that appear when the paper is already in review, at the worst possible moment.

What FDAAA 801 Actually Requires

Most clinical researchers know that registration on ClinicalTrials.gov is required before the first participant is enrolled. The results reporting requirement is less widely understood in its specifics. Under the 2017 FDAAA Final Rule, “applicable clinical trials” must submit summary results to ClinicalTrials.gov within 12 months of the primary completion date. The primary completion date is the date when the last participant was examined or received an intervention for the purposes of the study's primary outcome measure, not the date when follow-up ended or the paper was written.

The definition of “applicable clinical trial” is narrower than some researchers expect but broader than many hope. It covers interventional studies involving FDA-regulated drugs, biologics, or devices that have at least one U.S. site or that are conducted under an Investigational New Drug application or Investigational Device Exemption. Phase 1 studies and small device feasibility studies are exempt. Phase 2, 3, and 4 studies are not. A trial of an already-approved drug for a new indication, a Phase 3 comparative device study, or a comparative effectiveness trial involving a licensed therapeutic can fall squarely inside the mandate even when the investigator thinks of the work as purely academic research with no commercial application.

The required summary results are not a narrative writeup. They are structured data fields in the ClinicalTrials.gov Protocol Registration and Results System: participant flow information, baseline characteristics of all enrolled participants, point estimates and measures of variability for every primary and secondary outcome measure listed in the registration, adverse event data in a standardized table format, and a set of statements about agreements between the sponsor and data owners. Academic investigators who have only posted protocol information often do not realize that the results entry is a separate, substantial data submission task. Institutions that completed the registration years before may have no remaining staff member who remembers how to log in.

Why Academic Institutions Keep Falling Short

A 2015 analysis published in the New England Journal of Medicine established the scale of the academic compliance gap in an early data snapshot. At five years after trial completion, only 27.7 percent of trials funded by academic institutions or other government sources had posted results to ClinicalTrials.gov, compared to 41.5 percent for industry-funded trials. Both numbers were poor at that time. The structural gap pointed to something that has not meaningfully changed: industry sponsors have dedicated regulatory affairs departments and internal compliance systems built around FDA interactions. Academic investigators are usually principal investigators who move from trial to trial without a standing infrastructure for post-trial regulatory obligations.

Academic medical centers have improved their formal policies since 2018. A 2023 survey revisited in a 2025 publication found that the share of academic organizations with written registration policies had grown from 43 to 74 percent, and policies for results reporting had grown from 35 to 68 percent. Having a written policy is not the same as achieving timely compliance, and the FDA's own numbers suggest the gap between policy and practice remains wide.

The FDA's Office of Compliance 2025 Annual Report, released in April 2026, documented 42 Preliminary Notices of Noncompliance and two formal Notices of Noncompliance issued during 2025, the largest single-year count since enforcement began under FDAAA 801. Formal Notices of Noncompliance can precede civil monetary penalty proceedings. Whether the March 2026 reminder campaign translates into more formal enforcement letters over the following months remains to be seen, but the trajectory of the preceding two years indicates this will not be a one-cycle administrative effort.

The Journal-Side Problem: Registered Outcomes versus Published Outcomes

When a clinical trial is registered before enrollment begins, the registry entry becomes a public commitment about what the trial will measure and how it will measure it. If the published paper later reports different primary or secondary outcomes from those registered, readers cannot evaluate whether the favorable results reflect what was genuinely planned or what happened to turn out statistically significant after the data came in. This is the core of the outcome switching problem, and it remains persistent in published clinical research.

The COMPare project, a systematic audit conducted by researchers at the Centre for Evidence-Based Medicine at the University of Oxford, provides the most rigorous published baseline. Monitoring every randomized controlled trial published in five major medical journals (NEJM, JAMA, The Lancet, Annals of Internal Medicine, and BMJ) over a four-month period in 2015 and 2016, the team compared preregistered outcomes against those actually reported. Of 67 trials assessed, only 9 correctly reported all their prespecified outcomes without adding or dropping any. Across the group, 301 preregistered outcomes had been quietly omitted from the published papers, and 357 outcomes that had not appeared in the original registration had been silently added. For each discrepancy, the team submitted a letter to the relevant journal.

The COMPare audit was conducted a decade ago. Its findings have since been replicated in multiple fields. A cross-sectional analysis published in JAMA Internal Medicine found that when primary outcomes changed between registration and publication, the studies overestimated the intervention effect size by approximately 16 percent relative to trials with stable primary outcomes. That is not a small margin. It is large enough to move a result across the significance threshold in a field where effect sizes are often modest. The problem extends beyond randomized controlled trials. A meta-epidemiological study published in BMJ Open in June 2026, examining intervention cohort studies registered on ClinicalTrials.gov between 2014 and 2016 and subsequently published by 2024, found substantial rates of outcome discrepancy in this category of studies as well, which has historically attracted less scrutiny than traditional randomized trials.

How Journals Now Check Outcome Consistency

The ICMJE has required that manuscripts reporting clinical trials include the trial registration number in the abstract since 2005. That requirement, now adopted as standard practice far beyond ICMJE member journals, means any peer reviewer or editor with internet access can look up the registration entry and compare it against the outcomes as described in the paper. Not all editors do this as a routine step during initial screening. Reviewers who specialize in clinical trial methodology do, and statistical reviewers assigned to assess randomized trials almost always check.

The journals that publish the most clinical trials have made this more explicit. BMJ's instructions to reviewers specify that the trial registration record should be consulted as part of the review. JAMA's research integrity process includes outcome consistency as a review criterion. The New England Journal of Medicine has published multiple editorials over the past five years addressing its expectations on this point. At most of these journals, a discrepancy between registered and published outcomes triggers a formal query to the corresponding author, not a quiet rejection. Authors who receive a query asking them to explain why the primary outcome changed between registration and publication, and who do not have a documented protocol amendment predating the change, are in a difficult position that could have been avoided.

The right explanation for any outcome discrepancy requires showing two things: that the change was made before the data were unblinded or analyzed, and that there was a legitimate scientific reason for it that did not depend on knowing the results. A dated protocol amendment, an ethics committee approval for the amended protocol, or a registration update with a date that predates the completion of enrollment are all evidence that can support a legitimate explanation. An updated registry entry dated after the trial's completion date is not.

The Registration-to-Publication Pipeline

The frame that helps most is to treat ClinicalTrials.gov not as a regulatory checkbox completed before enrollment and then forgotten, but as a running ledger that should remain consistent with the eventual manuscript. The pipeline has distinct steps that need to go in a particular order, and deviations from that order create problems that compound over time.

Before the first participant is enrolled, the study is registered with primary and secondary outcomes defined with enough specificity that a reader could understand what a positive or negative result would look like. If the protocol changes materially, including any change to primary or secondary outcomes, the registration is updated with a dated amendment before the change takes effect. After the primary completion date passes, results are submitted to ClinicalTrials.gov within the 12-month window for applicable trials. The journal manuscript then describes the same outcomes that were registered, reports results for each of them, and explains in the methods section any deviations that occurred, when they occurred, and why.

Where this breaks down in practice: registration happens after enrollment has already begun, because investigators treat the registration as a submission formality rather than a prior commitment. Outcomes are adjusted post-hoc without protocol amendments. Results are never posted to ClinicalTrials.gov because no one takes responsibility for the task after the trial ends. The manuscript is written by selecting the outcomes that performed best rather than the outcomes that were prespecified. Each of these failure modes is common, each is detectable by editors and reviewers, and each carries consequences that have grown more serious as journals have tightened their checking procedures.

A Pre-Submission Self-Audit for Clinical Researchers

If you are preparing a clinical trial manuscript, the following review should happen before you draft the cover letter, not during revision after an editor asks for it.

None of these questions requires specialized regulatory expertise. They do require treating ClinicalTrials.gov as a live document, not an archived filing. The version of the registry entry that an editor sees when your paper is under review may be different from what you submitted years ago, particularly if registration records were updated without attention to how those updates affect the outcome consistency question.

What the FDA's 2026 Enforcement Push Signals for the Rest of the Year

The FDA did not accompany its March 2026 outreach campaign with new regulations or increased penalty amounts. The existing tools, civil monetary penalties for persistent noncompliance following a formal Notice of Noncompliance, have been available since 2017 and used sparingly. What changed in 2026 is the scale and visibility of the outreach effort. Sending letters to more than 2,200 responsible parties simultaneously, coordinated with NIH, and publicizing the campaign through FDA press channels, is a different kind of enforcement signal than the quiet letter-by-letter approach that characterized earlier years.

The follow-on training program reinforces the signal. The FDA and NIH announced joint virtual training sessions for academic medical center staff, scheduled for July 14, July 22, and July 30, 2026. These sessions address FDAAA 801 obligations specifically, the Protocol Registration and Results System interface, and frequently asked questions from academic institutions. The FDA also offered a pre-recorded video module that can be completed before attending the live sessions. Questions can be submitted in advance. This level of instructional infrastructure being built specifically for academic medical centers suggests that the agencies view academic noncompliance as a solvable operational problem, one they are willing to put training resources toward before escalating to penalties.

For individual investigators, the practical takeaway is not that enforcement actions are imminent for every overdue filing. It is that the period of informal tolerance for academic-sector ClinicalTrials.gov noncompliance appears to be ending. Academic medical center compliance offices are responding to the 2026 outreach by reviewing their trial portfolios for overdue results. Investigators who have open applicable clinical trials with primary completion dates more than 12 months in the past should contact their institution's clinical trials registration office to determine whether results submissions are outstanding.

The intersection of FDA compliance and journal editorial review makes this more than a regulatory concern. A researcher who ignores ClinicalTrials.gov results reporting is also building a paper trail that will eventually be visible to the editors evaluating their manuscript. The compliance gap and the outcome reporting problem share the same structural cause: treating ClinicalTrials.gov as a one-time task rather than a durable commitment. Getting both sides right, in the right order, before submitting to a journal, is not procedural busywork. It is the baseline standard for published clinical research in 2026.

Further Reading