Paperonium

Neurology · Neurology · 2025

Summary: restarting blood thinners later goes with clearly fewer brain bleeds. Its range includes no change.

The summarycalled clear: 4% lower, could beno changeThe results section10% lower, rules out no changeRedone from the 9 studies’ data,exactly the same0.911 to 1.015×0.896×0.810.9920.8×0.85×0.9×0.95×1×1.05×no changelower oddshigher odds
The summarycalled clear:4% lower, could beno changeThe results section10% lower, rules outno changeRedone fromthe 9 studies’ data,exactly the same0.911 to 1.015×0.896×0.810.9920.8×0.9×1×1.1×no changelower oddshigher odds

Odds of another brain bleed, multiplied per extra day before restarting blood thinners

The paper says

The paper combines 9 studies of 435 people with mechanical heart valves (which need blood thinners against clots) who had a brain bleed. Its summary, at the top: each extra day before restarting goes with 4% lower odds of another bleed, a “significant” (clear) link.

Its own data show

The summary’s range (the values the data can’t rule out) is 9% lower to 1.5% higher: it includes no change. The results section reports 10% lower; its margin of error gives 19% to 0.8% lower. Redoing the math on the 9 studies gives exactly that; no subset gives both the summary’s figure and its range.

A link is “significant” (clear) only if its whole range sits on one side of no change; the summary’s range doesn’t. The paper’s suggested 10–15 day wait before restarting rests on the results section’s figure, which the data reproduce. The summary, the part most readers see, misstates that central finding.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

Timing and Safety of Anticoagulation Reinitiation After Intracranial Hemorrhage in Patients With Mechanical Valves · Neurology (2025) · doi:10.1212/WNL.0000000000214184