Association Between Intraoperative Oliguria and Postoperative Acute Kidney Injury in Patients Undergoing...

Background/Objectives: Intraoperative oliguria has long been considered a marker of impaired renal perfusion, but its prognostic value for postoperative acute kidney injury (AKI) remains controversial, particularly within Enhanced Recovery After Surgery (ERAS) protocols. We investigated the association between intraoperative oliguria and postoperative AKI in patients undergoing major hepatobiliary and pancreatic surgery within an ERAS protocol. Methods: Patients who underwent major hepatobiliary and pancreatic surgery within an institutional ERAS protocol were retrospectively analyzed. Intraoperative oliguria was defined as urine output < 0.3 mL kg^-1 h^-1. Postoperative AKI was defined according to the KDIGO serum creatinine criterion within 7 days after surgery. The association was assessed using multivariable logistic regression and sensitivity analyses were performed using an alternative oliguria threshold of <0.5 mL kg^-1 h^-1 and incorporating additional surgical covariates. Results: Among 816 patients, intraoperative oliguria occurred in 51 (6.3%), and postoperative AKI developed in 60 (7.4%). AKI incidence did not differ between the oliguria and non-oliguria groups (11.8% vs. 7.1%, p = 0.332), and median intraoperative urine output was comparable between the AKI and non-AKI groups (0.7 [0.5-1.1] vs. 0.8 [0.6-1.4] mL kg^-1 h^-1, p = 0.069). In multivariable analysis, intraoperative oliguria was not independently associated with AKI (OR 1.68, 95% CI 0.60-4.01; p = 0.276). Oral carbohydrate loading, thoracic epidural analgesia, and total intraoperative fluid volume were not associated with AKI. Results were consistent across both sensitivity analyses. Conclusions: In patients undergoing hepatobiliary and pancreatic surgery within the ERAS protocol, intraoperative oliguria was not associated with postoperative AKI, although modest association cannot be excluded given the limited number of AKI and oliguria events. These findings suggest that intraoperative urine output alone may not be a reliable indication for additional fluid administration, and larger prospective studies are needed to confirm this.