Two Antithrombotic Clocks, Two Different Amounts of Patience
Both patients are on drugs that make bleeding worse and stopping them dangerous. What actually divides these two cases isn't the drug class — it's how much time each patient's underlying disease is willing to give before the risk of stopping outweighs the risk of continuing.
Harold B., a 74-year-old man, has kept the same Sunday-morning bridge group going for eleven years, missing only twice — once for a hip replacement, once for the week his wife was in the hospital. He has permanent atrial fibrillation and has taken apixaban 5mg twice daily for six years without a bleeding event or a stroke. His CHA₂DS₂-VASc adds to 4 without a stroke ever having happened: one point for being 65 to 74, one for hypertension, one for type 2 diabetes, and one for vascular disease — aortic atherosclerotic plaque reported incidentally on the same CT that found the stone. A CT for unrelated back pain incidentally found a 9mm mid-ureteral stone, and over the past three weeks he's had two emergency-department visits for colic severe enough that his bridge partners have started asking if he's alright mid-game. He needs ureteroscopy, and the actual disagreement in the room isn't about doing the procedure — ureteroscopy on a DOAC is manageable, tamponadable, endoscopically visualized bleeding, nothing like the retroperitoneal risk of a percutaneous tract. The disagreement is about what to do with the apixaban itself in the days around it.
The honest answer is that the evidence here is thinner than it looks, and it points in two directions at once. The AUA/International Consultation on Urological Diseases consensus states plainly that ureteroscopy can be performed with oral anticoagulation continued — the one recommendation written for this procedure rather than borrowed. Ingimarsson and colleagues' Mayo Clinic series was assembled specifically to test that statement, and found significant bleeding in a third of the handful of patients continued on a DOAC against roughly one in seventeen on warfarin, a striking split on its face but drawn from nine patients and seventeen, too few for the authors to call it significant. The EAU's periprocedural DOAC guidance recommends stopping one to three days beforehand, but that guidance was written for uro-oncologic procedures — TURBT, prostate biopsy, major reconstructive cases — not for ureteroscopy, which sits at a different point on the bleeding-risk spectrum entirely. Harold's six uneventful years on apixaban are the one piece of evidence in the room that is actually about him, and neither body of guidance knows he exists.
Clinic, weighing a chronic risk
I know the AUA/ICUD consensus says ureteroscopy can be done on continued anticoagulation. Ingimarsson's Mayo series is the study that went looking at whether that holds, and it found significant bleeding in a third of the DOAC patients continued through the procedure against a fraction of that on warfarin. It's underpowered — nine patients on a DOAC, the authors say so themselves — but it's the only direct evidence on the actual question in front of us, and it doesn't point the way the consensus statement does.
I'd hold apixaban for one to two doses before the case. Ureteroscopy isn't urgent enough today to justify skipping that.
The bleeding subgroup you're citing is real, but it's a handful of patients in a retrospective series that couldn't reach statistical significance — that's a signal worth noting, not a finding to build a stroke-risk decision around.
His CHA₂DS₂-VASc is 4. That's a continuously accruing stroke risk, and interruption-related thromboembolic events are one of the best-measured harms in the entire anticoagulation-bridging literature — unlike the ureteroscopy bleeding data, which is thin and uncertain in both directions. If we're going to accept a real, known risk to avoid an unconfirmed one, I want that trade named explicitly, not assumed.
You're both extrapolating from the wrong reference point, in opposite directions. The EAU's 1-3 day interruption guidance exists for uro-oncologic procedures — TURBT, major reconstructive surgery, prostate biopsy — and it's written around THEIR bleeding profile, not ureteroscopy's. Ureteroscopic bleeding is directly visualized and endoscopically tamponadable in a way none of those procedures are.
The closest thing we have to a real standardized framework for this exact tension — interruption interval balanced against thromboembolic risk, without bridging — is PAUSE, Douketis's three-thousand-patient cohort. Its logic was to set the interval by procedural bleeding-risk category rather than by a fixed multi-day default: one day off before a low-bleed-risk procedure, two before a high-risk one. Worth being exact about two things, because both get misremembered. PAUSE adjusted for renal function in the dabigatran arm only, so Harold's CrCl of 68 doesn't enter this calculation at all. And its low-risk arm omits more than the single dose I'm about to propose — so what I'm suggesting is shorter than PAUSE, not endorsed by it.
Most published continue-through-ureteroscopy series use exactly that kind of middle path: hold only the morning-of dose, resume that evening. It doesn't fully answer the bleeding subgroup data, and it doesn't fully answer the stroke-risk argument either — it's a genuine compromise, not a resolution of who's actually right.
Agreed: hold apixaban the morning of the procedure only, resume that evening. This was explicitly a compromise, not a resolution — the Urologist would still have preferred a longer hold given the bleeding subgroup data, and the Cardiologist would still have preferred no interruption at all given his stroke-risk profile. Neither position was overruled; the middle path was adopted because it was the one every voice could accept, not because the group agreed on which risk mattered more.
Dominic P., a 61-year-old man, had a drug-eluting stent placed six weeks ago after an anterior STEMI, and has been rebuilding his stamina one flight of stairs at a time since — he made it to the third floor of his apartment building without stopping for the first time last week, a milestone he mentioned twice during his urology visit. He's on dual antiplatelet therapy, aspirin and ticagrelor, and is now febrile, with flank pain and a CT showing a 2.5cm obstructing stone with a partial staghorn component and hydronephrosis — an infected, obstructed collecting system that needs decompression, not a stone that can wait for a more convenient antiplatelet window.
He has no prior stone history and no chronic kidney disease — this staghorn-component stone is new, discovered only because the infection it caused finally announced itself. His cardiac history is otherwise unremarkable apart from the index MI: no prior events, no heart failure, and an ejection fraction of 48% on his most recent echocardiogram, reduced but not steeply, which is the one number on his chart pointing the right way and the reason the next several months are worth protecting rather than gambling. The guideline default after an acute coronary syndrome treated with a drug-eluting stent is twelve months of dual antiplatelet therapy, with shortened courses reserved for patients whose bleeding risk is high enough to buy that trade. Dominic is six weeks in. Nothing anyone does tonight moves him meaningfully closer to that number.
What makes Dominic's case categorically harder than Harold's isn't the drug class — it's the clock. Harold's stroke risk from AF is chronic and continuously accruing; interrupting his anticoagulation for a day costs him a small, roughly proportional increment of that ongoing risk. A drug-eluting stent six weeks old has not finished endothelializing, and stent thrombosis in that window carries a mortality rate that dwarfs almost anything else on this list — a risk that isn't proportional to how long you interrupt DAPT, it's concentrated specifically in the early months regardless of duration. And PCNL itself, unlike ureteroscopy, is a percutaneous renal-parenchymal tract — bleeding there isn't endoscopically visualized or easily tamponaded the way ureteroscopic bleeding is. He is febrile at 38.6°C with a white count of 14.2 above an obstructed, infected kidney, which means the room does not get to wait for either clock to run down. It has tonight.
Admitting service, deciding tonight
Six weeks after a drug-eluting stent is still well inside the window where the vessel hasn't finished endothelializing. Stent thrombosis in that window carries a mortality rate high enough that I don't think almost anything on the urologic side outweighs it electively — and definitive PCNL, on full DAPT interruption, is exactly the kind of elective timing decision that risk should govern.
He is six weeks out from a stent placed for a STEMI, and the guideline-recommended DAPT course after an acute coronary syndrome is measured in months, not weeks. Waiting doesn't get him to safe ground; it only gets him further from the worst of it.
An infected, obstructed kidney isn't the same category as an elective stone I can schedule around a cardiac calendar. Untreated, this is a real, immediate urosepsis risk — his white count and fever are already telling us that. I don't have the luxury of treating this as fully deferrable the way a asymptomatic stone would be.
I hear the stent-thrombosis mortality data, and I'm not dismissing it. But I need this system decompressed today, not in three weeks.
I don't think this is actually a choice between waiting three weeks and doing definitive PCNL today on full DAPT — both of those are the wrong frame.
A temporizing ureteral stent, placed under continued aspirin with the P2Y12 inhibitor held only briefly, is a much lower-bleeding-risk intervention than a percutaneous tract, and it decompresses the infected system today — addressing the Urologist's actual urgency without accepting the stent-thrombosis risk the Interventional Cardiologist is right to name.
What it doesn't resolve is how soon after that definitive PCNL should follow, or exactly how long the P2Y12 inhibitor can safely stay held. That's a real open question, not a detail.
Agreed: temporizing ureteral stent placed the same day, aspirin continued throughout, ticagrelor held only for that shorter procedure and resumed immediately after per cardiology's guidance. This decompresses the infected system without accepting the stent-thrombosis risk of a full percutaneous tract on interrupted DAPT.
Not agreed, and carried forward explicitly rather than smoothed over: how soon after this definitive PCNL should follow. The Interventional Cardiologist would prefer putting as much distance as possible between the stent and any full DAPT interruption, without a specific date to point to; the Urologist is concerned that a staghorn-component stone left in place that long risks recurrent infection around the temporizing stent itself. Both positions are being tracked at his two-week follow-up rather than settled today.