Optimal hemodynamic targets for preventing acute kidney injury (AKI) have remained elusive. We hypothesized that lower tissue perfusion pressure (TPP), a novel perfusion index representing the difference between mean arterial pressure and the critical closing pressure, is predictive of AKI after cardiac surgery.

Individual patient TPP waveforms were constructed from continuous hemodynamic data in 1224 patients after cardiac surgery. The relationship of TPP and AKI was determined and stratified by vasoactive inotrope score. Logistic regression was performed to identify the optimal TPP threshold for predicting AKI. Unsupervised machine learning was used to explore different hemodynamic phenotypes and their association with AKI.

AKI occurred in 17.6% of patients and was associated with higher rates of mortality (15.8% vs 2.0%; P < .001) and major morbidity (45.1% vs 14.2%; P < .001) and significantly lower average TPP (37.6 mm Hg [33.7-41.0 mm Hg] vs 39.0 mm Hg [34.6-42.6 mm Hg], P < .001). A threshold TPP <38 mm Hg effectively stratified patients by AKI risk (odds ratio, 1.75; 95% CI, 1.30-2.35). For patients requiring vasoactive medications, average TPP <38 mm Hg indicated higher risk of AKI independent of average mean arterial pressure (adjusted odds ratio, 1.69; 95% CI, 1.17-2.45). K-means clustering identified a high-risk phenotype with lower average TPP (35.6 mm Hg [30.9-39.8 mm Hg] vs 40.4 mm Hg [35.9-44.5 mm Hg]; P < .001), higher average vasoactive inotrope score (2.8 [0.3-6.8] vs 0.5 [0.0-2.5], P < .001), and greater incidence of AKI (29.8% vs 10.1%; P < .001).

Lower TPP is associated with greater risk of AKI after cardiac surgery. TPP could serve as an adjunct to traditional hemodynamic measures to guide hemodynamic management, especially in patients with higher vasopressor requirements.