Anticoagulation Through a Pacemaker Implant, Three Months After a TIA
A single patient whose device implant is routine and whose anticoagulation history is not, which turns a standard pre-procedure checklist into a genuine judgment call.
G.F., a 71-year-old man, spent forty years as a locksmith, a trade he only fully retired from last year after his hands started giving him trouble with the finer work. He has atrial fibrillation, diagnosed six years ago and managed on apixaban since, and a new problem that's brought him back to cardiology: recurrent episodes of lightheadedness severe enough that his daughter now insists on driving him anywhere his appointments take him, culminating in a diagnosis of sick sinus syndrome and a recommendation for a dual-chamber pacemaker.
His CHA2DS2-VASc score sits at 4 — hypertension, his age bracket, and a transient ischemic attack three months ago that resolved without lasting deficit but left the whole family, and his care team, more attentive to anything touching his anticoagulation since. He has no other significant cardiac history and no bleeding history of his own. The pacemaker itself is a routine procedure for the electrophysiology team; what isn't routine is deciding what to do with his apixaban on the day of implant, given how recently his stroke risk was demonstrated to be real rather than theoretical.
Apixaban's anticoagulant effect comes from direct, reversible inhibition of Factor Xa, with a relatively short half-life compared to warfarin — part of why the question of interrupting even a single dose carries real, calculable stakes rather than an abstract one; missing one dose measurably lowers his circulating anticoagulant effect within hours, not days. That short half-life cuts both ways in this debate: it's exactly why a brief interruption is a meaningfully smaller gap than the multi-day warfarin bridging protocols this trial data was originally designed to replace, and exactly why a recently symptomatic patient like him has less margin to spare even within that smaller gap.
Deciding what happens to his apixaban on implant day
I want him fully anticoagulated through the procedure, including his morning dose. The best trial data we have on anticoagulation strategy around device implantation found continued and interrupted DOAC produced statistically identical rates of clinically significant hematoma — 2.1% either way, close enough that the trial was stopped early for futility. There's no bleeding-safety argument for interrupting him, and given how recently he had a TIA, I don't want to create even a brief gap in his coverage for a theoretical bleeding benefit the data don't actually support.
I'd go further than “no reason to interrupt him” — three months out from a TIA is still inside the window where I'd consider his stroke risk elevated above his baseline CHA2DS2-VASc score alone. Any interruption, even a single skipped dose, is a real consideration I don't want treated as routine just because this is a common procedure. Uninterrupted anticoagulation is the right call, and I'd want that explicitly documented as a clinical decision, not a default.
I agree with proceeding uninterrupted, with one practical caveat worth naming rather than assuming: his implant is a dual-chamber device, which typically means a longer procedure with more instrumentation. The trial reported a pooled hematoma rate across a mixed population of pacemaker and defibrillator implants; it did not break that rate out by procedural complexity, so we can't say what it predicts for a two-lead case specifically. I'd want the implanting team to confirm directly that they're comfortable with a fully anticoagulated field for this procedure, not just defer to the general trial result.
Apixaban continued uninterrupted through the implant, including the morning-of dose, after the implanting team confirmed directly that they were prepared for a fully anticoagulated field given the dual-chamber device's longer procedure time.
His anticoagulation strategy is validated for future procedures without needing to revisit this decision.
The team's documented reasoning — an elevated-stroke-risk patient in a procedure whose complexity the trial did not separately report on — becomes directly relevant to how the event gets evaluated.
The neurologist's insistence on treating this as a real clinical decision rather than a routine default was the point of disagreement that mattered — not with the electrophysiologist's conclusion, which everyone reached, but with the risk that a common procedure gets waved through without the reasoning being made explicit for a patient whose baseline risk had just changed three months earlier.