A competitor changed a trial’s completion date. What does it actually mean?

What should you check before interpreting anything?

Three things, in order. First, which kind of date moved: a forecast (the registry labels it estimated) or a date already marked actual. They mean different things and carry different base rates, which is why the two cases get their own sections below.

Second, whether this is the first move. In our archive, 39% of trials that moved a forecast date moved it again, and 9,526 trials moved one three or more times. A third slip on the same trial is a different story from a first.

Third, what moved with it. A date change that arrives alone is routine more often than not. A date change in the same update as an enrollment revision or a status flip is a co-ordinated edit, and co-ordinated edits are how real schedule resets appear in registry data.

The date was still a forecast and it moved. What are the odds?

If the move was 30 days or more, it pushed the date later 93.5% of the time in our archive, with a median slip of 335 days. So the direction itself carries almost no information: slipping is what forecast dates do. What carries information is the size and the repetition.

A slip near the median, roughly a year, is the registry’s normal weather. A slip in the top decile, beyond 731 days, is not, and deserves a closer look at the record: enrollment, site count, status. And any second or third slip on the same trial belongs on a watchlist rather than in a quarterly review, because repetition is the best predictor the registry offers that a programme’s schedule is genuinely under pressure.

The rare opposite case is worth naming: 6.5% of moves bring the date earlier, median 154 days. A genuine pull-in on a competitor’s phase 3 is one of the most underrated signals in the registry, precisely because it is fifteen times rarer than a slip.

The date was marked actual and it still moved. Is that normal?

More normal than most people assume, and one of the least understood behaviours in the registry. Actual does not mean final. In our archive, sponsors changed a primary completion date that was already marked actual 5,052 times across 4,831 trials. The median restatement moved the date 243 days, and 1,796 of them moved it by more than a year.

A restatement is a different object from a slip. A slip says the future changed; a restatement says the sponsor’s account of the past changed, usually a correction, a re-baselining after a protocol amendment, or an administrative clean-up. If you are dating anything downstream from a competitor’s actual completion date, a readout estimate, a filing window, this is the base rate for how often that anchor itself moves.

What is a date change not telling you?

It is not telling you the trial failed, and it is not telling you the data are bad. Nothing in a date field speaks to efficacy, and the registry offers no way to distinguish a slip caused by slow enrollment from one caused by a manufacturing delay or an amended protocol. Treating a slip as evidence of clinical failure is the single most common over-read of registry data.

What it does tell you, reliably, is when your own assumptions expire. Every plan that referenced the old date now references a fiction. The value of catching the change on the day it happens is not the interpretation; it is that everyone downstream stops planning against a date the sponsor no longer believes.