A retrospective cohort analysed 30,039 adult non-cardiac surgical cases at a South Korean academic centre from 2011 to 2020, stratified by median intraoperative pulse pressure: narrow (under 40 mmHg), reference (40 to 70) and wide (over 70). Entropy balancing adjusted for covariates including cumulative vasopressor load and duration of hypotension, and a two-by-two matrix crossed mean arterial pressure below or above 65 mmHg with pulse pressure category.
Acute kidney injury occurred in 6.7%. Narrow pulse pressure was independently associated with it (odds ratio 1.66, 95% CI 1.42 to 1.94). The matrix is the finding. Hypotension with a normal pulse pressure did not significantly raise risk (1.09, 0.86 to 1.38). Normotension with a narrow pulse pressure did (1.56, 1.27 to 1.89). Both together were worst (1.85, 1.21 to 2.81). The mechanistic substudy explains why: stroke volume index in the normotensive narrow group was 27 mL/m², against 47 in the hypotensive reference group. Risk became significant after 40 minutes.
A patient whose mean arterial pressure sits at 70 on a phenylephrine infusion with a pulse pressure of 32 is not a well-perfused patient — they are a low-flow patient whose pressure has been restored by squeezing. The practical change costs nothing: read the pulse pressure off the arterial trace you already have, and when it is narrow with a normal mean, ask whether the answer is more vasopressor or more volume and inotropy. This is observational and single-centre, so it identifies a signal to act on, not a proven target.
- Read pulse pressure, not only mean arterial pressure, off the trace you already have.
- A narrow pulse pressure with a normal mean on vasopressor suggests low stroke volume, not adequate perfusion.
- Forty minutes is roughly when the association became significant — treat sustained narrowing, not a single reading.
- Ask whether volume or inotropy, rather than more vasopressor, is what the number is telling you.
- Non-invasive cuff pulse pressure is less reliable; do not read this as finely without an arterial line.
The statistics, in plain English
The important comparison is not the headline odds ratio but the two-by-two matrix: hypotension alone carried an interval of 0.86 to 1.38, crossing 1.0 and meaning no detectable effect, while normotension with narrow pulse pressure did not cross it. That reversal is what makes the argument. Entropy balancing adjusts for measured confounders, including how much vasopressor was given, but cannot adjust for why an anaesthetist chose to give it.
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