Falls, Bruising, and the Decision to Keep Anticoagulation
A 74-year-old woman with permanent atrial fibrillation (CHA2DS2-VASc 5: prior TIA, hypertension, diabetes, age 65-74, female) has taken apixaban 5 mg twice daily for three years. Weight 62 kg, creatinine 1.1 mg/dL, eGFR 55. Over four months she has fallen twice at home - once rising at night, once in the bathroom - without fracture. Orthostatic vitals: supine 138/74, standing 104/62 with presyncope. Medications include hydrochlorothiazide 25 mg, terazosin 2 mg at bedtime, metformin, atorvastatin, and naproxen two or three times weekly for knee pain. Hemoglobin is 10.6 g/dL (baseline 11.5) with ferritin 12 ng/mL. She has extensive bruising on both forearms and asks, "Can we just stop the blood thinner? I'm going to crack my head open."
1. How do you counsel her about stopping apixaban? Quantify the fall-versus-stroke trade-off rather than giving a vague reassurance.
Read the worked answer
Recommendation: continue anticoagulation; treat the falls, not the prescription.
Reasoning:
- At CHA2DS2-VASc 5, her untreated annual stroke risk is roughly 6-7% per year, and anticoagulation reduces that by about two-thirds. That is a large absolute benefit - on the order of one stroke prevented per 30-40 patient-years.
- The fear driving her question is fall-related intracranial hemorrhage. Modeling studies put the threshold at roughly 300 falls per year before the bleed risk from falls outweighs the stroke benefit of warfarin-type therapy. Two falls in four months is nowhere near that.
- The key reframe: falls are a marker of frailty and of reversible problems - orthostasis, sedatives, poor lighting. They are a reason to intensify fall prevention, not to withdraw a therapy with a much larger expected benefit.
- Exceptions worth naming: hospice/limited life expectancy, or a patient whose stroke risk is genuinely low. Neither applies.
Say plainly: "Stopping the thinner would not make you safer overall - it would trade a small bleeding risk for a much larger stroke risk." Then pivot to the plan that lowers both risks.
2. List the specific medication and risk-factor changes you would make today, and justify whether 5 mg twice daily is the correct apixaban dose for her.
Read the worked answer
Dose check - 5 mg BID is correct. Apixaban is reduced to 2.5 mg BID only when at least two of three criteria are met: age ≥80, weight ≤60 kg, creatinine ≥1.5 mg/dL. She is 74, weighs 62 kg, and has creatinine 1.1 - she meets none. Dose-reducing without criteria would under-treat her.
Modifiable bleeding risks:
- Stop naproxen. NSAIDs plus an anticoagulant plus anemia is the highest-yield single intervention here.
- Work up the anemia: ferritin 12 with falling hemoglobin means iron deficiency until proven otherwise - check stool for occult blood, replace iron, and consider whether GI evaluation is warranted; add a PPI if there is any GI bleeding history.
- Limit alcohol; review over-the-counter supplements (fish oil, high-dose vitamin E, NSAID-containing cold remedies).
Modifiable fall risks:
- Deprescribe or reduce hydrochlorothiazide and terazosin - the classic orthostatic combination; the terazosin has no strong indication in this history.
- Recheck orthostatic vitals after the change; review vision, footwear, night lighting, home hazards, and refer to physical therapy for gait and balance training.
- Avoid benzodiazepines, Z-drugs, and antihistamines for sleep.
Repeat CBC after these changes and document the shared decision.
3. Six weeks later she returns after a witnessed fall in which she struck the back of her head on a doorframe. She is alert, oriented, has no headache, vomiting, or focal deficit, and asks to go home. What is your plan?
Read the worked answer
She gets a noncontrast head CT before anyone discusses discharge.
Reasoning:
- A patient on a direct oral anticoagulant with head trauma is a different risk category from an anticoagulant-naive patient. The validated decision rules (Canadian CT Head, New Orleans) were derived and validated in general populations and are not safely applied to anticoagulated patients.
- Intracranial hemorrhage after a fall can present with a lucid interval of hours. A normal exam in the emergency department, and even a negative initial scan, does not close the case.
- Practical sequence: noncontrast CT now; if negative, observe for at least 6-24 hours with serial neurologic checks and clear return precautions (worsening headache, vomiting, drowsiness, weakness, unequal pupils). Many centers admit or arrange a repeat scan at 6-12 hours, particularly with any anticoagulant on board, age >65, or a fall from standing or higher.
- Do not reverse anticoagulation for a negative scan. Reversal agents (andexanet alfa or four-factor PCC) are reserved for confirmed or strongly suspected major bleeding.
- Close the loop on the cause of the fall: reassess orthostatics, medications, and whether the earlier deprescribing actually happened.