Decreased mean perfusion pressure as an independent predictor of acute kidney injury after cardiac surgery
Raymond Hu, Yasmean Kalam, Jeremy Broad, Timothy Ho, Frank Parker, Matthew Lee, Rinaldo Bellomo
Abstract
Acute kidney injury after cardiac surgery (AKICS) is common. Previous studies examining the role that mean arterial pressure (MAP) during cardiopulmonary bypass (CPB) may have on AKICS have not taken into account how baseline central venous pressure (CVP) and mean perfusion pressure (MPP) (i.e. MAP - CVP) can influence its evolution. To assess whether the change in MPP to the kidneys (i.e. delta MPP or DMPP) during CPB compared to baseline is an independent predictor of AKICS. After ethical approval, a retrospective observational study was performed on all patients undergoing CPB between October 2013 and June 2015 at a university-affiliated hospital. Known risk factors for the development of AKICS were recorded, as were the MPP values at baseline and during CPB. From this, statistical modelling was performed to identify predictors of postoperative AKICS. 664 patients were identified. Analysis was performed on 513 patients after exclusion. On logistic regression, significant and independent predictors of AKICS included: d20DMPP (cumulative duration of MPP values during CPB that were 20% below baseline and exceeded three consecutive minutes) (P = 0.010); baseline CVP; age; pre-operative creatinine level; and left ventricular (LV) dysfunction (ejection fraction (EF) < 45%). On alternative modelling, the cumulative number of MPP values during CPB that were 10% below baseline was also independently associated with AKICS (P = 0.003). Modelling without taking into account CVP also supported this association. The duration of differences in perfusion pressure to the kidneys during CPB compared to baseline is an independent predictor of AKICS.