Renal Near-Infrared Spectroscopy for Assessment of Renal Oxygenation in Adults Undergoing Cardiac Surgery: A Method Validation Study

Abstract

To investigate the correlation between invasively measured renal venous oxygen saturation (SrvO2) and tissue oxygenation (rSO2) measured with near-infrared spectroscopy (NIRS) in adult patients undergoing cardiac surgery. Prospective observational study. Single cardiac surgery center at a university hospital. Thirteen adult patients with skin- to- kidney distance ≤4 cm undergoing open cardiac surgery with cardiopulmonary bypass (CPB). All patients received renal vein catheters for invasive measurement of SrvO2, and NIRS electrodes for assessment of renal rSO2 were placed over the kidney using ultrasound guidance. Measurements were made before CPB, during CPB at 3 different flow rates (2.4, 2.7, and 3.0 L/min/m2), and after CPB. Repeated- measures correlation analyses and Bland-Altman plots were used to study the correlation and agreement between rSO2 and SrvO2. For all measurement points, renal rSO2 correlated with SrvO2 (rrm = 0.61, p < 0.001), and the mean difference (bias) between rSO2 and SrvO2 was -2.71 ± 7.22 (p = 0.002), with an error of 17.6%. When measurements during CPB and before and after CPB were studied separately, rSO2 and SrvO2 were correlated (rrm = 0.51, p < 0.007 and rrm = 0.73, p < 0.001, respectively). During CPB, renal rSO2 predicted SrvO2 with a bias of -3.41 ± 7.76 (p = 0.009) and an error of 18.8%. Before and after CPB, the mean difference was -1.93 ± 6.60 (p = 0.092), with an error of 16.2%. Renal rSO2 is correlated to and predicts SrvO2 with a small bias and acceptable agreement. Further studies are needed before renal NIRS can be used as a surrogate marker of renal oxygenation in clinical practice.