Apixaban-associated traumatic brain injury with an expanding contusion and a three-hour road transfer
A 74-year-old woman on apixaban 5 mg twice daily for atrial fibrillation falls down three steps onto a tiled floor. She did not lose consciousness but has vomited twice. On arrival GCS is 14 (E4 V4 M6), pupils equal and reactive, blood pressure 168/92 mmHg, heart rate 88. There is a 4 cm occipital haematoma. Non-contrast CT shows a 9 mm acute subdural haematoma, a 15 mL frontal contusion and effaced sulci, with no midline shift. A repeat scan four hours later shows the subdural is unchanged but the contusion has grown to 30 mL with 4 mm of midline shift. The nearest neurosurgical unit is three hours away by road and the helicopter is grounded by weather. On site you have prothrombin complex concentrate and vitamin K, but no andexanet alfa and no apixaban-calibrated anti-Xa assay. The blood bank holds two units of red cells.
1. State your reversal strategy for this patient with doses, and justify what you would give and what you would withhold.
Read the worked answer
Give: four-factor prothrombin complex concentrate 50 IU/kg (approximately 3000 IU for a 60 kg woman) as a single intravenous dose. This is the best available agent when the specific antidote is unavailable; it bypasses the factor Xa inhibition and restores thrombin generation. Do not repeat it routinely - there is no assay to guide a second dose, and repeat dosing adds thrombotic risk in a patient with atrial fibrillation who will need anticoagulation again later.
Give: tranexamic acid 1 g intravenously if within three hours of injury. The evidence in traumatic brain injury is modest but the harm is minimal, and this patient has both a haemorrhagic contusion and an ongoing risk of expansion.
Withhold: vitamin K. It has no effect on direct oral factor Xa inhibitors, which do not deplete vitamin K-dependent factors. Giving it is not dangerous but it consumes time.
Withhold: fresh frozen plasma as primary reversal. It does not correct apixaban and adds volume that worsens cerebral oedema.
Withhold: platelet transfusion unless she is also on an antiplatelet agent or is thrombocytopenic - the platelets are functionally normal.
Withhold: activated charcoal - she is many hours from ingestion and has an unprotected, vomiting airway.
Simultaneously, and equally important: find out the time of the last dose; control systolic BP to below 140 mmHg while avoiding hypotension (a target of 110–140 systolic balances expansion against cerebral perfusion); head up 30 degrees; avoid hypoxia, hypercapnia, hyperthermia and hyponatraemia; hourly GCS and pupil checks by a named nurse; and repeat imaging before transfer. Reversal only buys time - the definitive treatment is neurosurgical assessment, and reversal should never delay the transfer decision.
2. The contusion has expanded with 4 mm of shift at four hours. Compare immediate transfer, local observation with interval scanning, and intubation before departure, and state what you would do.
Read the worked answer
Option 1 - transfer now. The expansion from 15 to 30 mL with new midline shift over four hours is a trajectory, not a stable picture. A contusion that has doubled in four hours can double again in the next four, and herringbone herniation can occur in a district hospital with no surgeon. Early transfer uses the current GCS 14 window: she is still protectable, still orientated enough to cooperate, and the ambulance is not yet a resuscitation room.
Option 2 - observe locally with repeat CT. Attractive because the transfer is long and carries its own risk, but it depends on hourly neurological observations that actually happen, a CT scanner available at 03:00, and a low threshold to move at the first sign of deterioration. It also means that if she deteriorates, she deteriorates somewhere without a surgeon, and the three-hour clock starts from a worse baseline.
Option 3 - intubate before departure. GCS 14 does not mandate intubation, and induction can cause hypotension, which is one of the two insults (with hypoxia) most strongly associated with worse outcome. But she has vomited twice, has an expanding lesion and faces three hours on a road with a paramedic who cannot manage an unprotected airway if she deteriorates. My judgement: transfer, with a physician or critical-care paramedic escort and full airway kit, and a low threshold to intubate at the first GCS fall, vomiting with reduced gag, or agitation - with ketamine or a careful titrated induction, avoiding hypotension.
What I would actually do: speak to the neurosurgeon at the receiving unit before departure, give the PCC and tranexamic acid, correct any coagulopathy and sodium, arrange a CT on arrival, and document a clear escalation plan. I would also ask her, while she can answer, what matters to her, because surgery is likely and she deserves a say.
3. The daughter asks when her mother can restart apixaban and whether the fall was caused by the drug. How do you respond, and what do you document?
Read the worked answer
Response. Apixaban did not cause the fall - it made the bleeding worse. Falls in older people are multifactorial (vision, gait, postural hypotension, sedatives, arrhythmia), and she should be assessed for those separately; but she is on an anticoagulant because she has atrial fibrillation and a meaningful stroke risk, and stopping it indefinitely carries a real chance of a devastating stroke.
On restarting. This is not an emergency-department decision to make alone. The competing risks are recurrent intracranial bleeding versus cardioembolic stroke, and the answer depends on the follow-up imaging, the size and location of the haematoma, her blood pressure control, her falls risk and her own priorities. Standard practice is a multidisciplinary decision (stroke physician, neurosurgeon, geriatrician, family doctor) at around four to eight weeks, with repeat imaging first. There is no role for bridging with low molecular weight heparin - that increases bleeding without protecting against stroke.
What I document. Time and dose of the last apixaban, the time of injury, the exact reversal given (agent, dose, time), the serial GCS and pupillary findings, the times of both CT scans, the blood pressure targets used, the conversation with the neurosurgical centre including the name of the clinician, and the conversation with the daughter - that the transfer is for assessment and possible surgery, not a guarantee of surgery. I would also notify the family doctor and flag the medication review, because the next clinician needs to know exactly what was reversed and when.