- Design
- Single-centre retrospective cohort with multivariable regression
- Population
- 4,358 adults having cardiac surgery with cardiopulmonary bypass, 2018 to 2024
- Primary outcome
- 30-day mortality by time-weighted indexed oxygen delivery
- Effect
- Adjusted OR 1.16 (95% CI 1.09 to 1.23) per 10 ml/min/m2 decrease; mortality 1.69% versus 0.74%
A single-centre retrospective cohort of 4,358 adults having cardiac surgery with cardiopulmonary bypass between 2018 and 2024 calculated time-weighted indexed oxygen delivery from pump flow, arterial saturation, haemoglobin and body surface area. Thirty-day mortality was 1.7% (75 patients).
Each 10 ml/min/m2 decrease in oxygen delivery was associated with higher odds of death (adjusted OR 1.16, 95% CI 1.09 to 1.23). Adjusted mortality was 1.69% at the 25th percentile of delivery (226.2 ml/min/m2) compared with 0.74% at the 75th (295.6). Lower delivery was also associated with acute kidney injury (adjusted OR 1.05 per 10 units), longer ICU and hospital stay, but not with stroke.
The authors say the findings give a rationale for mechanistic and interventional studies, which is the correct reading. With only 75 deaths, models are sensitive to chosen covariates, and confounding by patient condition is likely.
- Each 10 ml/min/m2 decrease in oxygen delivery went with adjusted odds of death of 1.16.
- Adjusted mortality was 1.69% versus 0.74% at the 25th versus 75th percentile.
- Acute kidney injury and longer stay were also linked.
- Stroke was not associated.
- Pump flow, haemoglobin and saturation are all levers that perfusion teams can see.
Why it matters
It strengthens the case for treating oxygen delivery, not only pump flow, as a perfusion target.
Don't overread it
A retrospective single-centre association cannot show that raising delivery lowers mortality.
The statistics, in plain English
An odds ratio of 1.16 per 10 units compounds across the range, so a difference of 70 units between the 25th and 75th percentile is substantial. With 75 deaths, the estimate is imprecise even though the p value is small.
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 anaesthesiology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free