Design: updated systematic review and meta-analysis of randomised controlled trials, RoB-2 assessed
Population: 1,889 adults with confirmed chronic subdural haematoma across 7 trials
Primary outcome: haematoma recurrence or persistence
Effect: recurrence relative risk 0.63 (95% CI 0.46 to 0.85); reoperation 0.39 (95% CI 0.28 to 0.56); serious adverse events 0.87 (95% CI 0.72 to 1.06)

Chronic subdural haematoma is a geriatric disease: an ageing population on antithrombotics, a minor head injury weeks earlier, and a burr-hole drainage that works and then recurs in a substantial minority. Seven randomised controlled trials covering 1,889 adults were pooled to assess middle meningeal artery embolisation added to standard treatment, with risk of bias assessed by RoB-2.

Recurrent or residual haematoma fell with embolisation, relative risk 0.63 (95% CI 0.46 to 0.85), and reoperation fell further, relative risk 0.39 (95% CI 0.28 to 0.56). Serious adverse events did not increase, relative risk 0.87 (95% CI 0.72 to 1.06), nor did neurological death, all-cause mortality or poor functional outcome. Haematoma resorption itself did not differ significantly. Subgroup and sensitivity analyses held across age, timing of the procedure and follow-up duration.

The reoperation figure is the one that changes a conversation with a family. A second craniotomy or burr-hole in an 84-year-old on an anticoagulant is not a minor event: it means another anaesthetic, another week of immobility, and the hospital-associated disability that follows. Cutting that by nearly two-thirds is a functional benefit even though the analysis found none in the formal functional outcome measures.

The timing result is practically useful in a resource-constrained service. If benefit does not depend on whether embolisation is done before, with or after drainage, then a unit without immediate interventional neuroradiology access does not have to choose between drainage now and embolisation now. The limits are that resorption did not improve — so this prevents recurrence rather than clearing blood faster — and that no trial here has long follow-up.

Raise embolisation at the neurosurgical referral for a chronic subdural haematoma, particularly in a patient who will need anticoagulation resumed
Use the reoperation figure, not the recurrence figure, when discussing benefit with families — it is the event that costs function
Do not expect faster haematoma clearance; resorption did not differ
Take the timing independence as operational permission to sequence around local availability
Keep the usual geriatric work alongside it: falls assessment, antithrombotic review, delirium prevention, early mobilisation

Why it matters

It removes the second operation, which is where an older patient with a chronic subdural loses function.

The statistics, in plain English

A relative risk of 0.39 for reoperation with an interval from 0.28 to 0.56 is both large and precise, which is unusual and reflects a consistent effect across seven trials. The serious adverse event interval, 0.72 to 1.06, includes 1.0 and sits mostly below it — meaning no signal of harm, though it cannot exclude a small increase. The absence of any difference in functional outcome or mortality is worth reading honestly: this reduces a procedure, not yet demonstrably a death or a disability. Subgroup robustness across timing is reassuring but subgroup analyses are always less powered than the main comparison.