Associations between mean systemic filling pressure and acute kidney injury: An observational cohort study following cardiac surgery.

Abstract

Venous congestion has been implied in cardiac surgery-associated acute kidney injury (CSA-AKI). The mean systemic filling pressure may provide a physiologically more accurate estimate of renal venous pressure and renal perfusion pressure but its association with CSA-AKI has not been reported. Patients admitted to ICU following cardiac surgery without pre-operative renal dysfunction were included with monitoring of mean arterial pressure (MAP) and central venous pressure (CVP) and cardiac output (CO) to calculate the mean systemic filling pressure analogue (Pmsa ). The AKI-KDIGO guidelines were used to define CSA-AKI. Logistic regression models including CO, heart rate, MAP, CVP and Pmsa were used to ascertain the association with CSA-AKI and reported by odds ratio (OR) with 95% confidence interval (95%CI) and area under the curve (AUROC). One hundred and thirty patients (out of 221 screened) were included of whom 66 (51%) developed CSA-AKI. Patients with CSA-AKI were older, with greater weight and increased stay in ICU while the proportion of comorbidities, type of surgical procedures, APACHE III scores and fluid volumes administered were similar to patients without AKI. The Pmsa , but not CVP, was associated with CSA-AKI (OR 1.2 95%CI [1.16-1.25]). Renal perfusion pressure was associated with CSA-AKI estimated as MAP-Pmsa (OR 0.81 [0.76-0.86]) and MAP-CVP (OR 0.89 [0.85-0.93]) with the former generating a higher AUROC (median difference 0.10 [0.07-0.12], P<.001) in the regression model. The Pmsa in post-operative cardiac surgery patients was associated with the development of CSA-AKI also when incorporated into estimates of renal perfusion pressure.