Panel statements are revisited on a schedule, and a new version replaces the old one in public. The record below is an illustrative example of a revisited statement.
A revisit is not a correction. The earlier version was the best reading of the evidence available when it was issued, and it stays visible with its own date beside the new one. What the record adds is the reason for the change: the study, replication or retraction that moved the conclusion, and which panel members changed position because of it. A reader comparing the two versions can see whether the field shifted or the panel simply reconsidered.
Revisit dates are set when the first statement is issued, not decided later. The panel judges how fast the underlying evidence is likely to move and commits to a date on that basis, which is why a statement on an active outbreak carries a short interval and one on a settled measurement carries a long one. If something significant lands before the date, the panel can be reconvened early, and the record shows that it was brought forward and why. A statement that passes its revisit date without being looked at is marked as overdue rather than left looking current.
Yes. Stop heavy outdoor work at 31 °C wet-bulb, high certainty. Lowered from 32 °C after a large 2026 field study found kidney strain starting a degree earlier.