(Open access funding provided by Semmelweis University)
Acute kidney injury (AKI) develops in 20-50% of patients undergoing cardiac surgery
(CS). We aimed to assess the predictive value of urinary biomarkers (UBs) for predicting
CS-associated AKI. We also aimed to investigate the accuracy of the combination of
UB measurements and their incorporation in predictive models to guide physicians in
identifying patients developing CS-associated AKI.All clinical studies reporting on
the diagnostic accuracy of individual or combined UBs were eligible for inclusion.
We searched three databases (MEDLINE, EMBASE, and CENTRAL) without any filters or
restrictions on the 11th of November, 2022 and reperformed our search on the 3rd of
November 2024. Random and mixed effects models were used for meta-analysis. The main
effect measure was the area under the Receiver Operating Characteristics curve (AUC).
Our primary outcome was the predictive values of each individual UB at different time
point measurements to identify patients developing acute kidney injury (KDIGO). As
a secondary outcome, we calculated the performance of combinations of UBs and clinical
models enhanced by UBs.We screened 13,908 records and included 95 articles (both randomised
and non-randomised studies) in the analysis. The predictive value of UBs measured
in the intraoperative and early postoperative period was at maximum acceptable, with
the highest AUCs of 0.74 [95% CI 0.68, 0.81], 0.73 [0.65, 0.82] and 0.74 [0.72, 0.77]
for predicting severe CS-AKI, respectively. To predict all stages of CS-AKI, UBs measured
in the intraoperative and early postoperative period yielded AUCs of 0.75 [0.67, 0.82]
and 0.73 [0.54, 0.92]. To identify all and severe cases of acute kidney injury, combinations
of UB measurements had AUCs of 0.82 [0.75, 0.88] and 0.85 [0.79, 0.91], respectively.The
combination of urinary biomarkers measurements leads to good accuracy.
