- Design
- Target trial emulation using two ICU databases
- Population
- 3,810 adults with septic shock on noradrenaline ≥0.25 µg/kg/min
- Primary outcome
- 28-day mortality
- Effect
- 48.1% vs 53.0%; risk difference −5.0 points (−6.8 to −3.2); HR 0.82 (0.78–0.85)
This study emulated a target trial using two large US ICU databases (MIMIC-IV and eICU). It included 3,810 adults with septic shock whose noradrenaline reached 0.25 µg/kg/min before vasopressin was started, and compared starting vasopressin within three hours of that threshold with starting it at three to six hours, using clone-censor-weight methods.
Weighted 28-day mortality was 48.1% with ultra-early vasopressin and 53.0% with early vasopressin, a difference of 5 percentage points (HR 0.82). Renal replacement therapy (HR 0.68) and treated arrhythmia were also less common with ultra-early initiation, though acute kidney injury was not.
Target trial emulation is a disciplined way to use observational data, but it cannot remove unmeasured confounding — sicker or better-staffed units may differ in timing. The finding fits the general logic of adding a second vasopressor early rather than pushing noradrenaline alone, and is a reason not to delay once the threshold is reached.
- Consider adding vasopressin promptly once noradrenaline reaches about 0.25 µg/kg/min in septic shock.
- Avoid escalating noradrenaline alone for hours before adding a second agent.
- Monitor for digital and mesenteric ischaemia on vasopressin.
- Reassess fluid status, source control and steroid need alongside vasopressor escalation.
Why it matters
It suggests timing, not just choice, of the second vasopressor may matter.
Don't overread it
This is observational data analysed as if it were a trial; only a randomised trial can confirm the mortality difference.
The statistics, in plain English
A 5-point absolute difference in mortality (95% CI 3.2 to 6.8) would be large if real. The narrow interval reflects the weighted statistical model; it does not account for confounding the data did not capture.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for emergency & critical care, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free