Clinical Report: Systematic Review and Meta-Analysis of Predictive Models for Acute Kidney Injury Linked to Sepsis
Overview
This systematic review and meta-analysis evaluates the performance of predictive models for sepsis-associated acute kidney injury (SA-AKI). The pooled C-statistic was 0.817, although significant heterogeneity exists among studies.
Background
Sepsis is a major cause of mortality in intensive care units, with acute kidney injury occurring in 40% to 50% of sepsis patients. Identifying patients at high risk for SA-AKI is critical for timely intervention. The development of reliable predictive models is essential to enhance early detection and management of this condition.
Data Highlights
Metric
Pooled Estimate
95% CI
C-statistic
0.817
0.781–0.847
Heterogeneity (I²)
92.8%
-
Models in Asia C-statistic
0.845
-
Models in North America C-statistic
0.777
-
Low risk of bias C-statistic
0.847
-
High risk of bias C-statistic
0.762
-
Key Findings
The pooled C-statistic for SA-AKI prediction models was 0.817.
Substantial heterogeneity was observed among studies (I² = 92.8%).
Models developed in Asian regions showed a higher pooled C-statistic compared to those from North America (0.845 vs. 0.777).
Studies with a low risk of bias demonstrated better predictive performance than those with a high risk of bias (0.847 vs. 0.762).
No significant publication bias was detected (P = 0.1891).
Clinical Implications
Clinicians should consider the variability in model performance based on regional and bias-related factors.
Conclusion
The study highlights the moderate-to-good performance of current SA-AKI prediction models, emphasizing the need for further validation.
