DETERMINE: an individualised blood pressure target during thrombectomy changed nothing

Design

Multicentre, open-label, blinded-endpoint randomised trial at eight French comprehensive stroke centres; Class II evidence

Population

433 adults with acute ischaemic stroke from anterior large vessel occlusion undergoing mechanical thrombectomy, median age 69.3 years

Primary outcome

Modified Rankin Scale 0 to 2 at 90 days

Effect

44.2% individualised vs 48.8% standard; adjusted OR 0.82 (95% CI 0.54 to 1.24, p = 0.34)

Eight French comprehensive stroke centres randomised 433 adults with anterior large vessel occlusion undergoing mechanical thrombectomy, open-label with a blinded endpoint. One arm held mean arterial pressure within 10% of the first value measured before the procedure — the patient's own pressure as the target. The other used the conventional band, systolic 140 to 180 mmHg. Favourable outcome, modified Rankin 0 to 2 at 90 days, was reached by 44.2% with the individualised target and 48.8% with the standard one, adjusted odds ratio 0.82 (95% CI 0.54 to 1.24, p = 0.34). Mortality at 90 days was 18.8% against 16.4%, and symptomatic intracranial haemorrhage did not differ.

One detail decides how much this means. The trial reports no significant difference in mean pressures or in pressure variability between the groups during the procedure. So the two protocols largely produced the same haemodynamics, and the null result is at least partly a null contrast rather than a null effect. That is a common failure mode in periprocedural blood pressure trials and it should temper the conclusion in both directions.

What it does support is leaving current practice alone. There is now Class II randomised evidence that individualising the target adds nothing detectable, so a unit running the standard band has no reason to change, and one considering a bespoke target has no evidence to justify the extra complexity. Effort during thrombectomy is better spent on avoiding the extremes — hypotension on induction, and uncontrolled hypertension after reperfusion — than on tuning a personalised number.

• Keep the conventional intraprocedural target; individualising to the patient's own baseline has no demonstrated advantage.
• Guard against hypotension at induction, which this trial was not designed to address and which still harms.
• Record the pre-procedure mean arterial pressure anyway — it remains useful context after reperfusion.
• Treat the null result cautiously: the two arms achieved similar pressures, so the strategies were barely separated.
• Nothing here changes post-reperfusion targets, which are governed by different trials.

The statistics, in plain English

An adjusted odds ratio of 0.82 with an interval from 0.54 to 1.24 straddles 1.0, so the trial cannot separate benefit from harm; the point estimate happens to favour standard care but the interval is compatible with a moderate advantage either way. With 433 patients this is a trial powered to detect a large difference, not a small one. The more important limitation is not statistical: because achieved pressures and their variability were similar in both arms, the comparison was between two protocols that ended up doing much the same thing. A trial that fails to separate its arms answers a narrower question than the one it asked.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.