A study’s registration record is not fixed. Sponsors update it over the life of a trial, and the version showing today may not match what was filed before the trial started. If you want to know whether a study kept its original plan or changed course along the way, the answer is on the public record itself: every registry entry has a version history, and the plan a trial started with can be compared side by side with the plan it ended with.
Where the Registration Change Log Actually Lives
On ClinicalTrials.gov, the current, most recent version of a study record is what displays by default. The full change history sits in a separate view: the record’s Record History tab (sometimes reached through a “History of Changes” link), which lists every version of that record in order by submission date. From there, two versions can be selected and compared directly, showing what was added, removed, or reworded between them.
This history is a normal, expected part of the system. Sponsors are required to keep records current throughout a trial’s life cycle, so version changes on their own are not evidence of anything improper.
Dates: What to Line Up Before Comparing Anything
A registration change log is only useful if you anchor it to the right dates. Four dates matter most:
- Study start date — when enrollment or the trial itself began.
- Primary completion date — when the last participant was assessed for the primary outcome.
- The date a specific field was changed — visible in the version history, not the overall record.
- Last Update Posted — the most recent date the public record was refreshed, which tells you how current the displayed version is.
The single most important question to ask of any change is: did this happen before or after the study’s primary completion date? A change made while a trial is still enrolling is a routine protocol adjustment. A change made after data collection has effectively finished carries more weight, because at that point the people making the change may already know how the results are trending.
Outcomes: The Field Worth Checking First
Not every field in a registration record carries equal risk when it changes. Eligibility criteria, contact information, and recruitment status are updated routinely and rarely raise questions. The primary outcome measure is different — it is the pre-specified yardstick the whole trial is designed around, and changing it after the fact can shift what “success” means for the study.
This pattern, known as outcome switching, has been studied directly. One cross-sectional analysis of 389 trials found that 130 of them — about a third — had at least one primary outcome change between registration and publication, and that trials with a changed primary outcome tended to report a larger treatment effect than trials whose outcome stayed the same. That does not mean a changed outcome is disqualifying on its own. It means a changed primary outcome is the single strongest reason to open the version history and read the “before” and “after” wording side by side, rather than taking the published outcome at face value.
What the Registry Does — and Does Not — Tell You
The version history shows what changed and when. It generally does not explain why. A sponsor is not required to justify a change in the public record, so a shift from one primary outcome to another may reflect a legitimate protocol amendment, a correction of an error in the original filing, or a change made for less transparent reasons — and the registry entry alone usually cannot tell you which.
This is the honest limit of what registration data can answer. Reading the change log tells a reader where to look more closely. It does not, by itself, tell a reader whether a given change was justified.
Using SPIRIT and CONSORT as a Second Check
The SPIRIT and CONSORT reporting guidelines exist to make this kind of comparison possible in the first place. SPIRIT sets out a 34-item checklist of what a trial protocol should specify before a study begins, including how the primary outcome should be defined. CONSORT sets out a parallel 30-item checklist for reporting a trial’s results once it is finished. Read together, they establish what “should have been fixed from the start” for a well-reported trial — which gives a reader a second, independent standard to check a registration change against, beyond just comparing two versions of the same registry entry.
A Practical Way to Read Any Trial’s History Tab
- Open the Record History tab and note how many versions exist and over what span of time.
- Identify the study’s primary completion date first, before looking at what changed.
- Compare the earliest version’s primary outcome wording against the current version’s wording.
- Note the date of that specific change relative to the primary completion date.
- Treat a pre-completion change as routine and a post-completion change as worth a closer look — not as a verdict either way.
- Check whether the published results (if available) match the current registered outcome, the original one, or neither.
What Remains Unresolved
Version comparison tells a reader what changed and when. It cannot independently confirm intent, and it cannot substitute for reading the full published study when one exists. For any trial tied to a specific health condition or treatment decision, the registration history is a starting point for evaluating how reliable a study’s reporting is — not a stand-in for medical guidance, and not a basis for a treatment decision on its own.
Next Step
For background on how a study record is structured before diving into its history, see ClinicalTrials.gov’s own guide to reading a study record. For the standards trial protocols and reports are measured against, see the SPIRIT and CONSORT statements. Related reading on this site is available in Study Design, and background on this publication’s approach is on the About page.