- Design
- Multicentre observational cohort with joint modelling of longitudinal inotrope exposure
- Population
- 440 adults with cardiogenic shock on VA-ECMO, six Dutch hospitals, 2018 to 2024
- Primary outcome
- ECMO weaning failure at 30 days (death, ongoing ECMO or heart replacement therapy)
- Effect
- Weaning failure in 53.4%; cumulative inotrope exposure HR 1.04 (95% CI 1.00 to 1.07) for failure and 1.04 (1.01 to 1.08) for 30-day death
This six-hospital Dutch observational cohort followed 440 adults with severe cardiogenic shock supported on venoarterial ECMO between 2018 and 2024. Daily inotrope exposure was summarised as the area under a standardised inotrope score, and joint models related it to weaning failure within 30 days, a composite of death, continued need for ECMO or need for heart replacement therapy.
Weaning failed in 235 patients (53.4%). Inotrope support stayed largely unchanged in the days after ECMO started. Higher cumulative exposure was associated with weaning failure (HR 1.04, 95% CI 1.00 to 1.07), 30-day mortality (1.04, 1.01 to 1.08), arrhythmias (1.05, 1.01 to 1.08) and slower lactate clearance (HR 0.96, 0.92 to 0.99).
These are small hazard ratios, and sicker patients plausibly receive more inotrope, so the association may reflect severity rather than harm. The authors call for prospective studies of tapering or stopping inotropes during ECMO. It should not be read as a reason to withdraw support from a patient who needs it.
- Review the need for each inotrope daily once a patient is on VA-ECMO, rather than continuing by default.
- Check for arrhythmia and lactate trends when inotropes are continued or increased.
- Do not stop inotropes on this evidence alone; discuss with the ECMO team and the cardiology team.
- Record inotrope doses on the daily ECMO chart so exposure is visible.
Why it matters
It questions the habit of leaving inotropes unchanged once the circuit is carrying the circulation.
Don't overread it
This was observational, and sicker patients may need and receive more inotrope; it does not show that reducing inotropes improves recovery.
The statistics, in plain English
A hazard ratio of 1.04 describes the change in risk for each unit increase in the cumulative score, so the effect builds over days of exposure. The lower confidence limit for weaning failure sits at 1.00, so that estimate is borderline. Joint models account for exposure changing over time, but they cannot remove confounding by indication.
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