When It's Safe to Start Blood Thinners After a Bleed in the Brain
A retiree's subdural hematoma has stopped growing on serial scans, but three days of bed rest are stacking his clot risk in the other direction. The team has to decide how soon is actually safe.
F.A., a 71-year-old retired postal carrier, was up on a ladder clearing gutters when it shifted under him, and he struck his head against the patio stones on the way down. He arrived with an 8mm right subdural hematoma, alert and oriented, headache but no focal deficit — a bleed the team elected to manage nonoperatively from the start, and one that has now held stable across three separate scans, admission, six hours, and this morning's repeat at day three. That stability is the actual finding driving today's question, not a formality on the way to it: the risk that matters most in the first hours after a bleed, progressive hemorrhagic injury, has had three separate chances to show itself and hasn't.
What has been accumulating instead is the other side of the ledger. Three days of bed rest, mechanical prophylaxis alone, and a BMI of 34 stack real, measurable venous thromboembolism risk with every additional day pharmacologic prophylaxis is withheld. Phelan and colleagues' pilot trial of early anticoagulation after low-risk traumatic brain injury found no increase in hemorrhage progression when prophylactic heparin was started once repeat imaging confirmed stability, a finding that lines up with the broader pattern across observational TBI cohorts: the highest-risk window for rebleeding is the first 24 to 72 hours, precisely the window his three clean scans have already spanned. His aspirin, held since admission for the bleed itself, is a separate question — today's decision is only about starting prophylactic-dose low-molecular-weight heparin, not resuming his home antiplatelet, and conflating the two would overstate what's actually being reconsidered.
His wife, at the bedside for most of the admission, has already asked twice whether a clot in his leg is something they should be worried about — a question that reflects a real family awareness of the immobility risk stacking up in front of them, distinct from whatever the team decides today about the bleed side of the calculation. His known coronary disease, well-controlled for eight years prior to this fall, is not itself a reason to accelerate or delay pharmacologic prophylaxis; it bears on the later aspirin conversation, not on today's heparin decision.
Reading the third scan
Three clean scans across the window where progressive hemorrhagic injury actually happens — admission, six hours, twenty-four hours — plus a stable repeat this morning at day three. That's the evidence Phelan's own trial population looked like, and it found no increase in hemorrhage progression with prophylactic heparin started at that point. I'd start today. Every additional day we wait is a day of unopposed clot risk on a bed-bound, obese patient with nothing offsetting it.
I don't think the evidence is wrong, but subdurals specifically can occasionally show a delayed expansion pattern beyond the 72-hour window that epidural and contusion data don't capture as well — they're a different bleed mechanism, venous rather than arterial, and slower to declare themselves in a small subset of cases. I'd want one more scan before committing, given how little it costs to wait a single additional cycle.
If the concern is genuinely about a rare delayed pattern rather than the well-characterized early window, a graduated start — half-dose enoxaparin today, full prophylactic dose at twenty-four hours if the next scan is unchanged — gives us most of the anticoagulation benefit today without betting everything on a single reassuring scan. It's not what either of you actually proposed, but it directly addresses the specific worry on the table.
Reduced-dose enoxaparin started that afternoon; the day-4 scan was unchanged, and he was advanced to full prophylactic dosing the following morning without any evidence of hematoma expansion at either step.
The neurointensivist's caution was not disproven so much as never tested — the graduated approach meant nobody had to find out what a single-scan, full-dose start would have shown. Aspirin resumption was deferred to a separate conversation with cardiology once he was further out from the bleed.