An 84-year-old on warfarin who keeps falling - and has just hit his head

A persisting headache two days after a head strike in an 84-year-old taking warfarin is not reassuring. My priority is to exclude a delayed intracranial haemorrhage rather than to treat the headache symptomatically.

• Arrange same-day urgent CT head, including the posterior fossa and a bone window. Most guidance advises imaging after any head injury in an anticoagulated patient, irrespective of whether the Canadian CT Head Rule criteria are met, because the rule was not designed to rule out bleeding in this group and the consequences of a missed bleed are severe.
• Check INR today and record GCS, pupils, focal neurology and blood pressure. If the INR is above the therapeutic range, or there is any neurological deterioration, discuss reversal (vitamin K ± prothrombin complex concentrate) with haematology or the emergency department before discharge.
• If CT is normal, neurological observations and a written safety-net plan are still needed, because delayed bleeding can occur. I would not send him home alone; the daughter should be present overnight and know the specific symptoms that require an emergency call.
• Examine the scalp haematoma and look for a skull fracture or cervical tenderness. Document the discussion with both patient and daughter.

I would acknowledge the concern directly, then separate two questions that are often fused: does warfarin increase bleeding, and does that outweigh its benefit in a man who falls?

• With a CHA2DS2-VASc score of 5, his untreated annual stroke risk is in the region of 6–7%, and AF-related strokes are often severe or fatal. The absolute increase in intracranial haemorrhage from warfarin is much smaller, of the order of 0.3–0.5% per year, and falls alone are not a recognised reason to withhold anticoagulation. Stopping it would probably cause more harm than it prevents.
• The falls are a reason to modify risk, not to abandon treatment. Diclofenac is the clearest target: an NSAID plus warfarin substantially increases gastrointestinal bleeding and also raises blood pressure. Stopping it and using topical NSAID or paracetamol for the knee is a genuine gain.
• Other modifiable factors: review INR stability, consider switching to a DOAC if there is no mechanical valve or severe mitral stenosis and renal function permits, address vision, footwear, home hazards, knee pain and any sedating medication, and check bone health.
• Explore what "everything possible" means to the family. Document a shared decision and offer advance care planning, because the daughter's framing suggests unspoken goals of care that should be discussed while he can participate.

The aim is to reduce falls and bleeding without removing treatments that provide net benefit.

• Stop diclofenac. It worsens hypertension and renal risk, causes fluid retention, and combines with warfarin to increase gastrointestinal bleeding. Replace with paracetamol and topical NSAID, and consider referral for knee strengthening or intra-articular injection.
• Review amlodipine against lying and standing blood pressures. If there is symptomatic postural hypotension, a lower dose may reduce falls; if not, leave it alone rather than changing things reflexively.
• Continue warfarin unless CT or INR control demands otherwise. If INR control is poor and he is reliable with medication, a DOAC is a reasonable discussion, but it is not automatically safer in a faller and cost and monitoring matter in Hong Kong.
• Check vitamin D and bone density, since a fall plus osteoporosis is the real fracture risk; treat if indicated.
• Non-drug measures: home safety review, vision check, footwear, and a falls programme. His MMSE of 22 raises adherence and supervision issues.
• Ask the daughter to supervise NSAID avoidance and to bring the INR card to every visit. Reassess in four weeks rather than leaving it to the next routine appointment.