Today's change
This week's change is not measured for this date
What moved across 149 tracked conditions, from the public trial registries. Compared with the snapshot from seven days earlier, 26 September 2026.
Collected . This page is dated 3 October 2026 because that is the UTC date the overnight sync ran under, which is not always the same calendar day it finished in Sydney. The time above is when this page's figures were actually computed.
Phase mix of all tracked trials
Hover or tap a phase to see what it means.
What the phases mean
- Phase 1
- The first tests in people. Phase 1 tests safety in a small group.
- Phase 2
- A larger group. Phase 2 tests whether the treatment actually works.
- Phase 3
- Phase 3 compares the new treatment to the current standard of care at scale, and is the last hurdle before a regulator will consider approval.
- Phase 4
- Phase 4 comes after approval, monitoring a treatment already in real-world use.
Of 112,174 trials we track that record a phase, whether or not they are still recruiting. A further 135,761 — 54.8% of the 247,935 trials we track, so more than are counted here — record no phase at all. Those are mostly observational studies, medical devices and registries, which never sit on the drug-approval phase pathway. They are excluded from this calculation, so these percentages describe trials with a recorded phase and not all research. Early Phase 1 is counted within Phase 1. How phases work →
Trials that changed phase this week
No trial we track had its registered phase changed in the 7 days to 3 October 2026. That is a real reading of the registries for this window, not a missing figure. Phase changes are rare. Over the 33 weekly windows to 12 August 2026 the registries changed the phase of between 0 and 6 trials a week, out of more than 200,000 we track, so a week with none is an ordinary week rather than a missing figure.
| Phase | Changed to | Changed from | Net | Total now | Search |
|---|---|---|---|---|---|
| Phase 1 | 0 | 0 | 0 | 23,228 | Search these → |
| Phase 2 | 0 | 0 | 0 | 49,363 | Search these → |
| Phase 3 | 0 | 0 | 0 | 25,295 | Search these → |
| Phase 4 | 0 | 0 | 0 | 14,288 | Search these → |
Counted over the 7 days to 3 October 2026, over every trial we track, whether or not it is recruiting. A trial that finishes, is terminated or stops recruiting does not appear here — it is still a trial we track and still at the same phase, and its status change is what the boards below report. Total now is how many trials we currently hold at that phase, whether or not they are recruiting — a level, not a change. It is not last week's total plus “Net”: a trial also joins or leaves a phase count when we first index it or when the registry re-tags it, and neither of those is a phase change, so neither is counted in the three columns beside it. The four levels come to 112,174; a further 135,761 trials record no phase at all, and the two together are the 247,935 trials we track. Phase changes are rare. Over the 33 weekly windows to 12 August 2026 the registries changed the phase of between 0 and 6 trials a week, out of more than 200,000 we track, so a week with none is an ordinary week rather than a missing figure.
How these dates work. These changes are dated by when our nightly sync saw them, not by when the registry recorded them. We have since checked every one of them against ClinicalTrials.gov’s own version history: we saw them a median of six days late, and in one case twenty-nine. A change therefore falls in a later week here than the week it happened in — so a recent week can count changes that belong to an earlier one and read high, while the week they belong to reads low. A trial we have only just begun tracking has no earlier phase to compare against, so its first recorded phase is never counted as a change.
Trials that entered recruitment this week
Not measured for this window. We hold snapshots at both ends, but this board now counts only trials whose status changed at ClinicalTrials.gov inside the window, and those change dates have not been resolved for this date. The earlier version of this board counted our own catch-up as registry activity, so nothing is shown rather than that number.
Longest without new activity
- 94+ days
- 94+ days
- 94+ days
- 94+ days
- 94+ days
A quiet week doesn't reflect a condition's severity or research importance, only recent registry activity.
Our record of this begins on 1 July 2026, so a figure marked “+” means we have seen no increase for the whole time we have been recording — the true silence may be far longer. These numbers understate quietness; they cannot overstate it.
See all 149 conditions ranked →
About these figures
Sourced from public government trial registries. Voxsanity is editorially independent and this page is not medical advice. Talk to your doctor about anything you read here.
Australian coverage is understated. Trials registered only with ANZCTR and not with ClinicalTrials.gov are not counted, so the Australian figure is a floor, not a total.
This week's change is not shown, and that is deliberate. A week-over-week difference in our own totals is mostly us catching up rather than the registries changing. In one week we examined in full, 89 trial records moved on a single condition and 3 of them had genuinely changed at the registry inside that week — the rest were late corrections landing. This runs in both directions: a condition can appear to gain, or to fall, in a week when nothing happened at the registry, and can appear flat in a week when something did. So this page withholds the figure for dates where we cannot yet date the changes to the registry's own clock, rather than publishing the other number.
Percentages can mislead on small numbers. A condition with very few trials can show a large percentage change from a single new trial, which is why every row above leads with the absolute count and shows the percentage only as context.