Beyond the end of surgery: early postoperative laboratory trends improve prediction of acute kidney injury after noncardiac surgery

Acute kidney injury (AKI) is a common and consequential complication of noncardiac surgery, but most perioperative risk models are static, combining preoperative variables into a single estimate that does not update as intraoperative and postoperative data accumulate. We examined the association between perioperative hemodynamic and laboratory variables and postoperative AKI, and tested whether adding early postoperative laboratory trends improves risk discrimination beyond a preoperative/intraoperative model alone. We performed a retrospective cohort analysis of the publicly available VitalDB dataset, comprising perioperative clinical, intraoperative, and time-stamped laboratory data from 6,388 patients undergoing noncardiac surgery at a single tertiary center. AKI was defined by KDIGO serum creatinine criteria. A multivariable logistic regression model examined preoperative, intraoperative, and early postoperative (day-1) predictors of AKI. Discrimination of a preoperative–intraoperative model (variables available by the end of surgery) was compared with a perioperative model additionally incorporating day-1 maximum lactate, day-1 hemoglobin, and day-1 maximum potassium, using 5-fold cross-validated area under the receiver operating characteristic curve (AUC) and bootstrap resampling. Of 3,834 analyzable cases, 256 (6.7%) developed AKI. Independent predictors included ASA class ≥ III (OR 1.91), male sex (OR 1.42), longer surgery duration (OR 1.22 per hour), day-1 maximum lactate (OR 1.13 per mmol/L), and day-1 maximum potassium (OR 2.83 per mmol/L), while older age, higher preoperative hemoglobin, and higher preoperative albumin were associated with lower odds. The preoperative–intraoperative model discriminated AKI with an AUC of 0.75 (95% CI 0.72–0.78); adding early postoperative laboratory trends (perioperative model) improved this to 0.77 (95% CI 0.74–0.80; bootstrap difference 0.023, 95% CI 0.008–0.041, p = 0.002). Postoperative AKI after noncardiac surgery was associated with recognized preoperative and intraoperative risk factors, and incorporating early postoperative laboratory trends produced a modest but statistically significant improvement in discrimination. These findings support further development of dynamic, updatable AKI risk models that evolve across the perioperative period rather than relying on a single preoperative estimate.

Authors

Institutions

Publication Details

Journal
BMC Anesthesiology
Published
2026-11-12
DOI
https://doi.org/10.1186/s12871-026-04420-8
Primary Topic
Acute Kidney Injury Research
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Beyond the end of surgery: early postoperative laboratory trends improve prediction of acute kidney injury after noncardiac surgery

Zheng Tao, Xia Sun, Bin Liu, Yanfeng Ge
BMC Anesthesiology
Acute Kidney Injury Research
article

Beyond the end of surgery: early postoperative laboratory trends improve prediction of acute kidney injury after noncardiac surgery

Zheng Tao, Xia Sun, Bin Liu, Yanfeng Ge
article en

Abstract

Acute kidney injury (AKI) is a common and consequential complication of noncardiac surgery, but most perioperative risk models are static, combining preoperative variables into a single estimate that does not update as intraoperative and postoperative data accumulate. We examined the association between perioperative hemodynamic and laboratory variables and postoperative AKI, and tested whether adding early postoperative laboratory trends improves risk discrimination beyond a preoperative/intraoperative model alone. We performed a retrospective cohort analysis of the publicly available VitalDB dataset, comprising perioperative clinical, intraoperative, and time-stamped laboratory data from 6,388 patients undergoing noncardiac surgery at a single tertiary center. AKI was defined by KDIGO serum creatinine criteria. A multivariable logistic regression model examined preoperative, intraoperative, and early postoperative (day-1) predictors of AKI. Discrimination of a preoperative–intraoperative model (variables available by the end of surgery) was compared with a perioperative model additionally incorporating day-1 maximum lactate, day-1 hemoglobin, and day-1 maximum potassium, using 5-fold cross-validated area under the receiver operating characteristic curve (AUC) and bootstrap resampling. Of 3,834 analyzable cases, 256 (6.7%) developed AKI. Independent predictors included ASA class ≥ III (OR 1.91), male sex (OR 1.42), longer surgery duration (OR 1.22 per hour), day-1 maximum lactate (OR 1.13 per mmol/L), and day-1 maximum potassium (OR 2.83 per mmol/L), while older age, higher preoperative hemoglobin, and higher preoperative albumin were associated with lower odds. The preoperative–intraoperative model discriminated AKI with an AUC of 0.75 (95% CI 0.72–0.78); adding early postoperative laboratory trends (perioperative model) improved this to 0.77 (95% CI 0.74–0.80; bootstrap difference 0.023, 95% CI 0.008–0.041, p = 0.002). Postoperative AKI after noncardiac surgery was associated with recognized preoperative and intraoperative risk factors, and incorporating early postoperative laboratory trends produced a modest but statistically significant improvement in discrimination. These findings support further development of dynamic, updatable AKI risk models that evolve across the perioperative period rather than relying on a single preoperative estimate.

BMC Anesthesiology
Affiliated Hospital of Jiangsu University (CN)
Good health and well-being
Acute Kidney Injury Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon