ICRAKI: intermittent haemodialysis first failed non-inferiority and was linked to more deaths than CRRT

Design

Multicentre, open-label, randomised non-inferiority trial; stopped early

Population

554 critically ill adults with AKI needing renal replacement therapy

Primary outcome

Major adverse kidney events by day 90

Effect

78.1% vs 71.6% (difference 6.3, 95% CI -1.2 to 13.8); death HR 1.30 (1.04-1.63)

ICRAKI, published in NEJM on 14 October 2026, was a multicentre open-label French trial in critically ill adults with acute kidney injury who needed renal replacement therapy. Patients were randomised to start with intermittent haemodialysis or continuous renal replacement therapy. It was stopped early after interim analyses showed a sustained signal of higher mortality with intermittent haemodialysis.

Of 535 patients analysed, major adverse kidney events by day 90 — death, dialysis dependence or creatinine more than 25% above baseline — occurred in 78.1% with intermittent and 71.6% with continuous therapy (adjusted difference 6.3 points, 95% CI -1.2 to 13.8), so non-inferiority was not shown. Day-90 mortality was 63.3% against 53.6% (HR 1.30, 1.04 to 1.63). Hypotension occurred in about two-thirds of both groups.

This challenges the view, based on earlier meta-analyses, that the two modalities are interchangeable in unstable patients. It is one trial, stopped early, which can exaggerate effects; but where both are available, starting with continuous therapy in a critically ill patient is now the better-supported choice.

• Where both are available, consider continuous renal replacement therapy as the first modality for critically ill patients with AKI.
• Reserve intermittent haemodialysis as initial therapy for more stable patients or where continuous therapy is not available.
• Review local ICU protocols that default to intermittent haemodialysis.
• Where continuous therapy is unavailable, intermittent or prolonged intermittent therapy remains necessary; this trial did not test prolonged intermittent therapy.
• Monitor for hypotension with either modality; rates were similar.

Why it matters

Intermittent and continuous therapy have long been treated as equivalent; this trial suggests the first choice may affect survival.

Don't overread it

Stopped early, which can overstate the effect; the primary composite difference had an interval crossing zero.

The statistics, in plain English

The primary outcome showed a 6-point excess with intermittent haemodialysis, but the interval (-1.2 to 13.8) was too wide to prove non-inferiority. The mortality hazard ratio of 1.30 excludes no effect, but early stopping makes the true size uncertain.

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.