Anticoagulation After a Gastrointestinal Bleed in Atrial Fibrillation
A 71-year-old woman with permanent atrial fibrillation, hypertension, type 2 diabetes, and heart failure with an ejection fraction of 40% comes to clinic two months after a lower gastrointestinal bleed from colonic angiodysplasia. She required 3 units of packed red cells, hemostasis was achieved endoscopically, and she has had no bleeding since. Current medications are aspirin 81 mg daily, metoprolol, furosemide, empagliflozin, and atorvastatin. She stopped her apixaban on her own after the bleed and asks whether she can simply stay off blood thinners because "the bleeding was worse than a stroke would be." Blood pressure is 152/88 mmHg and she has no focal neurologic deficits. Laboratory values: hemoglobin 10.2 g/dL, platelets 210,000/µL, creatinine 1.6 mg/dL; weight 58 kg and estimated creatinine clearance 32 mL/min. Her CHA2DS2-VASc score is 5.
1. How do you weigh stroke risk against bleeding risk, and what do you tell her about remaining off anticoagulation?
Read the worked answer
The decision is not "anticoagulate versus protect her from bleeding" - it is "which risk is larger, and which parts of each risk can be modified." Her CHA2DS2-VASc is 5 (age 65-74, female sex, hypertension, diabetes, heart failure), corresponding to roughly 6-7% annual stroke risk, much of it disabling or fatal. Her HAS-BLED score is elevated mainly for prior bleeding and abnormal renal function, but a high HAS-BLED is a prompt to correct modifiable factors, not a reason to withhold anticoagulation. Her bleed was angiodysplasia, treated with endoscopic hemostasis, with no recurrence in two months.
Staying off anticoagulation is an active choice with a defined harm, and the conversation should say so plainly: her stroke risk without anticoagulation is substantial, long-term anticoagulation lowers it by roughly two-thirds, and her residual bleeding risk is manageable with monitoring. Modifiable factors include stopping aspirin, which adds bleeding risk without benefit here, controlling blood pressure to target, avoiding NSAIDs, treating anemia, and monitoring hemoglobin and renal function. She should be told to report melena or hematochezia and that a recurrent bleed would be managed endoscopically rather than by permanently abandoning anticoagulation.
2. Which anticoagulant would you choose, and at what dose? Justify the dose with the specific criteria.
Read the worked answer
Apixaban 2.5 mg twice daily. The reduced apixaban dose is not determined by creatinine clearance alone; it requires at least two of three criteria: age 80 years or older, weight 60 kg or less, and serum creatinine 1.5 mg/dL or higher. She is 71 and therefore does not meet the age criterion, but she weighs 58 kg and has a creatinine of 1.6 mg/dL, giving her two of three. Full-dose apixaban 5 mg twice daily would overexpose her.
Apixaban is a good choice because it has the lowest gastrointestinal bleeding signal among the direct oral anticoagulants in trials, is usable at a creatinine clearance near 32 mL/min, and needs no INR monitoring. Warfarin is an alternative but causes more intracranial hemorrhage and is harder to control in someone with recent bleeding. Dabigatran is a poor choice at this creatinine clearance and after a lower gastrointestinal bleed. The common errors to avoid are reducing the apixaban dose because of renal function alone and continuing aspirin alongside the anticoagulant.
3. List the additional management steps you would take at this visit.
Read the worked answer
Stop aspirin. She has no coronary stent and no recent acute coronary syndrome, so aspirin contributes bleeding risk without a clear benefit. Restart anticoagulation now; she is two months past hemostasis and well beyond the window in which early resumption would be hazardous - most guidance supports resuming 7-14 days after endoscopic hemostasis once bleeding has stopped. Control blood pressure toward 130/80 mmHg or lower. Advise against NSAIDs and suggest acetaminophen or topical therapy for osteoarthritis pain. Check hemoglobin and creatinine at intervals and consider iron replacement for the anemia.
A proton pump inhibitor is reasonable given her bleeding history, though it is not mandatory in the absence of other risk factors. Document a follow-up plan to reassess adherence, bleeding, and renal function, and explain that the reduced dose depends on both the weight and creatinine criteria persisting. Also confirm she understands that the DOAC must not be crushed or doubled after a missed dose, and that any invasive procedure requires a specific interruption plan rather than self-discontinuation.