The topic in depth
A blow to the head

Chronic subdural haematoma

It is the slow bleed that collects between the brain and its coverings over weeks, almost always in older people and often with no blow anyone remembers. It has an effective treatment. This page explains when it is operated on and when it is watched, what the operation involves, what embolisation adds, and what results to expect, even at ninety.

Chronic subdural haematoma9 min read21 references

Before you read on

This page sets out what the published research shows. It is not advice for your case: those decisions are made in consultation, with your scans and your history in front of us. It is here so you arrive knowing what to ask.

Why it is a disease of old age

In a United States population, 17.3 cases are diagnosed per 100,000 people per year, but the figure rises to 354.8 per 100,000 per year in men aged 85 or over1. In Finland the incidence doubled between 1990 and 2015, from 8.2 to 17.6 per 100,000 per year, and in those over 80 it almost tripled, from 46.9 to 129.52. Over that same period the median age at diagnosis went from 73 to 79 years, and the proportion of patients taking anticoagulants or antiplatelet drugs —medicines that make the blood harder to clot— almost doubled, from 27 % to 49 %2.

It is not an old clot. Around the collected blood the body forms new membranes —neomembranes— with fragile vessels that bleed a little, and bleed again; that, and not the blow, is what keeps the haematoma growing over weeks3.

That is why the blow is usually a minor trigger, and often no blow at all: in the Finnish series only 59 % of patients had a documented injury, and among those aged 80 or over the mechanism was a fall from their own height in 54 % of cases2.

When the decision to operate is taken, and what happens if it is not

There is no universal threshold. The decision combines the symptoms the haematoma can account for —confusion, persistent headache, weakness down one side, unsteadiness when walking, falls— with the size of the collection and how far it pushes the brain across.

What has been measured is what happens when there is no operation. In an analysis of 35 studies and 2,095 patients with a chronic subdural haematoma causing few symptoms and managed without surgery, 19.8 % ended up needing rescue surgery4. Waiting and watching is a legitimate option in someone with few symptoms, but one in five is operated on anyway.

It is almost never an emergency of minutes. In a series of 179 patients, the mean time between the CT scan that made the diagnosis and the operation was 76 hours, and that delay was not associated with higher mortality, worse functional outcome, longer stays in hospital or more reoperations5.

What the operation is

We operate through burr holes —one or two small openings in the bone—. The collection is washed out and a drain is left in place, a fine tube that clears what remains.

Leaving that drain is one of the few surgical steps with a mortality reduction demonstrated in a randomised trial. In 215 patients operated on, recurrence —the haematoma forming again— was 9.3 % with a drain against 24 % without it, and mortality at 6 months was 8.6 % against 18.1 %6. The trial was stopped early because of the benefit.

How long to leave it also has an answer. A Danish trial randomised 347 patients to 6, 12 or 24 hours of drainage: symptomatic recurrences at 3 months were 27 %, 20 % and 10 %7. That is why the drain comes out at around 24 hours.

Many of these patients are operated on awake, under local anaesthesia with sedation, with the area numbed and the patient breathing unaided. Across 18 studies and 4,367 patients, that technique did not change recurrence, mortality or the need for reoperation compared with general anaesthesia, but general anaesthesia was associated with 2.4 times more complications8. These are observational studies, not trials: frailer patients tend to be given a general anaesthetic, and that may explain part of the difference.

Even so, the haematoma comes back in about 12 % of cases9. When it does, it tends to be early: in a series of 283 haematomas, the mean time to reoperation was 47 days10.

Middle meningeal artery embolisation

The outer membrane of the haematoma is fed by branches of the middle meningeal artery, and it is from that fragile network that the small repeated bleeds making it grow come3. Embolisation aims to cut off that supply: through a puncture in an artery in the arm or the groin, a very fine tube is taken up to those branches and they are blocked off. It does not empty the haematoma; it attacks the mechanism that feeds it.

Five trials have tested it, and they did not all say the same thing. In the table, treatment failure gathers what each trial measured: the haematoma coming back, a significant remnant being left, or further surgery being needed. The first figure is for the patients who were embolised, the second for those who were not.

TrialPatientsTreatment failure
EMBOLISE4004.1 % versus 11.3 % at 90 days11
STEM31016 % versus 36 % at 180 days12
MAGIC-MT7226.7 % versus 9.9 % at 90 days; no significant difference13
EMPROTECT34214.8 % versus 21.0 % at 6 months; no significant difference14
MEMBRANE37611.6 % versus 22.1 % at 6 months15

Pooled, the five trials show less treatment failure, with no difference in mortality or in stroke16. Put another way: it reduces recurrences, it does not save lives.

Nor is the benefit shared out evenly. When the first three trials are pooled, the overall effect did not reach statistical significance; split by group, the reduction appeared in the patients who were not operated on and not in those who were17.

It is not a harmless procedure. In EMBOLISE, 2.0 % of the embolised patients had a serious adverse event related to the procedure within 30 days, including two disabling strokes11.

And there is the cost. An analysis of four trials, carried out in the British public health system, confirmed the reduction in recurrence requiring surgery —relative risk 0.40— but concluded that embolisation is not cost-effective as a treatment for everyone, and suggested reserving it for patients at high risk of the haematoma coming back18.

Steroids, a question with an answer

They were used for decades. A British trial of 748 patients put them to the test: a favourable outcome at 6 months was reached by 83.9 % with dexamethasone against 90.3 % with placebo19. Dexamethasone did reduce reoperation for recurrence —1.7 % against 7.1 %— but it worsened the overall functional outcome19. Less recurrence did not mean being better off.

The companion question also has an answer: can steroids be given instead of an operation? A Dutch trial of 252 patients comparing dexamethasone with surgery was stopped early on safety grounds. Complications were 59 % with dexamethasone against 32 % with surgery, and 55 % of the dexamethasone group were operated on anyway, against 6 % of the surgical group20.

Results in the very old patient

Extreme age on its own is not a contraindication to surgery. In a cohort of 1,312 patients operated on for chronic subdural haematoma, those in their tenth decade of life —90 years or over— achieved clinical and radiological results similar to those of younger patients21.

That does not make the overall prognosis good, and the context is worth having. In the United States cohort, the median age of patients with chronic subdural haematoma was 76 years, and only 36.3 % were functionally intact before the episode1. Mortality was 9.4 % at 30 days and 32.9 % at one year1.

That one-year figure frightens people more than it should. Early death was attributed to the haematoma in 48.4 % of cases, and late death in only 16.1 %; the commonest causes of late death were neurodegenerative disease, at 27.2 %, and cardiovascular disease, at 28.4 %1. The operation resolves the haematoma in the great majority. What it cannot do is give back the health the person no longer had beforehand.

If you already have a date for surgery, the care for the days beforehand and for the recovery is in care before and after the operation.