Related Experiment Video
Updated: Jul 30, 2025

A Large Animal Model for Acute Kidney Injury by Temporary Bilateral Renal Artery Occlusion
Published on: February 2, 2021
Characterization of Risk Prediction Models for Acute Kidney Injury: A Systematic Review and Meta-analysis
Yunlin Feng1,2, Amanda Y Wang2,3,4, Min Jun2
1Department of Nephrology, Sichuan Provincial People's Hospital, University of Electronic Science and Technology of China, Chengdu, China.
This systematic review found that while acute kidney injury (AKI) prediction models show good discrimination, high heterogeneity and risk of bias limit their clinical utility. Standardized development and validation are needed for better patient outcomes.
Area of Science:
- Nephrology and Clinical Epidemiology
- Development and validation of predictive models for acute kidney injury (AKI)
Background:
- Despite numerous published prediction models for AKI, their real-world application and impact on patient outcomes remain under-evaluated.
- A systematic review is crucial to assess the current landscape of AKI prediction models across diverse clinical settings.
Approach:
- A comprehensive literature search of MEDLINE and Embase databases was conducted from January 1946 to April 2021.
- Studies developing AKI prediction models with at least two predictive variables were included, irrespective of population or design.
- Data extraction and analysis followed PRISMA guidelines, with pooled C statistics used to measure model discrimination.
Key Points:
- 150 studies involving 14.4 million participants met the inclusion criteria, revealing wide variations in study design, population, and AKI definitions.
- The overall pooled C statistic for AKI prediction models was 0.80, indicating good discrimination.
- High between-study heterogeneity was observed across all clinical settings, and 84.4% of models had a high risk of bias.
Conclusions:
- Published AKI prediction models demonstrate good discriminatory ability but exhibit significant heterogeneity and a high risk of bias, limiting their clinical utility.
- There is an urgent need for standardized methodologies in the development and validation of AKI prediction models to enhance their real-world applicability and improve patient care.
Related Concept Videos
Acute Kidney Injury I: Introduction
Acute Kidney Injury IV: Diagnostic Studies and Prevention
Acute Kidney Injury II: Pathophysiology
Acute Kidney Injury V: Interprofessional Care
Acute Kidney Injury III: Clinical Manifestations
Acute Kidney Injury VI: Nursing Management

