Perioperative intravenous fluid for cardiothoracic surgery and acute kidney injury in Korea between 2007 and 2021
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background: Although fluid therapy in cardiothoracic surgery has been widely studied, how restrictions on synthetic colloids have influenced practice and kidney outcomes remain unclear. Objective: To describe temporal changes in perioperative intravenous fluid use and to examine associations with acute kidney injury (AKI) and hemodialysis in patients undergoing cardiac or thoracic aorta surgery. Design: Retrospective nationwide cohort study. Setting: Korean National Health Insurance Service (NHIS) database. Patients: Adult patients undergoing cardiac or thoracic aorta surgery in Korea between 2007 and 2021. Interventions: None; perioperative intravenous fluid exposure was assessed using national insurance claims data. Main outcome measures: The primary outcomes were AKI and newly required hemodialysis. Secondary outcomes included in-hospital and 1-year mortality. Results: A total of 112,928 patients were analyzed. AKI occurred in 2.4%, and 1.1% required hemodialysis. From 2007 to 2021, the use of synthetic colloids decreased substantially, whereas acetate-buffered balanced crystalloid use increased to near-universal adoption. Albumin administration and red blood cell transfusion were independently associated with higher odds of AKI and hemodialysis after adjustment for covariates. Over time, the incidence of AKI increased, while the incidence of hemodialysis remained stable. Conclusion: Between 2007 and 2021, the incidence of perioperative AKI increased from 1.4% to 3.1%, whereas the incidence of hemodialysis remained stable at approximately 1%. During the same period, perioperative fluid management changed substantially, with decreased use of synthetic colloids and near-universal adoption of acetate-buffered balanced crystalloids. These findings represent population-level temporal trends and should be interpreted as descriptive rather than causal.