---
title: "Cross-System ML Meta-Analysis Identifies Value-Specific Risk Drivers for Acute Kidney Injury"
id: "medrxiv-7-cross-system-meta-analysis-of-machine-learning-predictors-identifies-value"
canonical_url: "https://medichelpline.com/clinical-feed/medrxiv-7-cross-system-meta-analysis-of-machine-learning-predictors-identifies-value"
content_type: "clinical_feed_article"
specialty: "Nephrology"
source_name: "medRxiv (Clinical Preprints)"
source_url: "https://www.medrxiv.org/content/10.64898/2026.08.31.26361849v1?rss=1"
published_at: "2026-09-02T12:00:00.000Z"
evidence_level: "Verified Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# Cross-System ML Meta-Analysis Identifies Value-Specific Risk Drivers for Acute Kidney Injury
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/medrxiv-7-cross-system-meta-analysis-of-machine-learning-predictors-identifies-value
- **Specialty:** [Nephrology](https://medichelpline.com/clinical-feed/nephrology.md)
- **Primary Source:** medRxiv (Clinical Preprints)
- **Source URL:** [Original Journal Publication](https://www.medrxiv.org/content/10.64898/2026.08.31.26361849v1?rss=1)
- **Published At:** 2026-09-02T12:00:00.000Z
- **Evidence Rating:** Verified Feed
## Executive GIST (TL;DR)
- This study pooled interpretable machine learning results from nine U.S. academic medical centers using electronic health records from 785,497 adult inpatients (2010–2019) to identify generalizable risk patterns for **acute kidney injury (AKI)**. - Site-level models used interpretable gradient boosting machines to quantify predictor–outcome associations; results were integrated using meta-regression to characterize nonlinear value–risk relationships and bivariate interactions. - Key continuous biomarker findings included a 1.46-fold higher AKI risk when glucose increased from 100 mg/dL to 140 mg/dL, a 1.14-fold risk increase for anion gap across 4–12 mmol/L, and a 1.28-fold risk increase for chloride across 96–100 mEq/L. - Electrolytes such as potassium, calcium, and sodium exhibited quadratic (U-shaped) associations with AKI risk, indicating elevated risk at both low and high values. - Bivariate meta-regression identified interactions between important predictors, showing that combinations of biomarkers jointly modulate AKI susceptibility rather than acting independently. - The analysis translates machine learning-derived associations into clinically interpretable, value-specific thresholds that could inform risk stratification and personalized prevention in hospital care. - Data were de-identified EHRs governed by the Greater Plains Collaborative and PaTH networks and are not publicly available; access requires data use agreements and coordinating center approval. - The paper discloses several unrelated industry financial relationships for some authors and notes IRB waivers or approvals as reported by participating institutions.
## Clinical Analysis & Structured Key Points
Cross-System Meta-Analysis of Machine Learning Predictors Identifies Value-Specific Risk Drivers and Interactions Underlying Acute Kidney Injury | medRxiv Skip to main content Cross-System Meta-Analysis of Machine Learning Predictors Identifies Value-Specific Risk Drivers and Interactions Underlying Acute Kidney Injury Ho Yin Chan , View ORCID Profile Deyi Li , Alan S.L. Yu , John A. Kellum , Dana Y. Fuhrman , Qi Xu , Elizabeth A Chrischilles , Lindsay G. Cowell , Sravani Chandaka , Alfred Jerrod Anzalone , Jacob Kean , Kathleen M. McTigue , Abu Saleh Mohammad Mosa , Bradley Taylor , Mahanaz Syed , Lemuel R. Waitman , Yong Hu , Mei Liu doi: https://doi.org/10.64898/2026.08.31.26361849 Ho Yin Chan 1 Department of Health Outcomes and Biomedical Informatics, College of Medicine, University of Florida, Gainesville, Florida, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Deyi Li 1 Department of Health Outcomes and Biomedical Informatics, College of Medicine, University of Florida, Gainesville, Florida, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Deyi Li Alan S.L. Yu 2 Division of Nephrology and Hypertension and the Kidney Institute, University of Kansas Medical Center, Kansas City, Kansas, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site John A. Kellum 3 Department of Critical Care Medicine, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Dana Y. Fuhrman 3 Department of Critical Care Medicine, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Qi Xu 1 Department of Health Outcomes and Biomedical Informatics, College of Medicine, University of Florida, Gainesville, Florida, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Elizabeth A Chrischilles 4 Department of Epidemiology, College of Public Health, The University of Iowa, Iowa City, Iowa, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Lindsay G. Cowell 5 Department of Health Data Science and Biostatistics, Peter O Donnell Jr. School of Public Health, UT Southwestern Medical Center, Dallas, Texas, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sravani Chandaka 6 Department of Population Health, University of Kansas Medical Center, Kansas City, Kansas, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Alfred Jerrod Anzalone 7 Department of Biostatistics, College of Public Health, University of Nebraska Medical Center, Omaha, Nebraska, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jacob Kean 8 Department of Internal Medicine, Spencer Fox Eccles School of Medicine, University of Utah, Salt Lake City, Utah, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Kathleen M. McTigue 9 Department of Medicine, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Abu Saleh Mohammad Mosa 10 Department of Biomedical Informatics and Data Science, Heersink School of Medicine, University of Alabama at Birmingham, Birmingham, Alabama, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Bradley Taylor 11 CTSI Center for Biomedical Informatics, Medical College of Wisconsin, Milwaukee, Wisconsin, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mahanaz Syed 12 Department of Population Health Sciences, University of Texas Health Science Center at San Antonio, San Antonio, Texas, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Lemuel R. Waitman 13 Department of Biomedical and Health Informatics, University of Missouri-Kansas City, Kansas City, Missouri, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Yong Hu 14 Big Data Decision Institute, Jinan University, Guangzhou, Guangdong, China Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mei Liu 1 Department of Health Outcomes and Biomedical Informatics, College of Medicine, University of Florida, Gainesville, Florida, USA; Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: mei.liu{at}ufl.edu Abstract Info/History Metrics Supplementary material Data/Code Preview PDF Abstract Background: Current understanding of acute kidney injury (AKI) risk factors remains largely descriptive, offering limited precision into how specific biomarker values or physiologic thresholds influence susceptibility. We aimed to synthesize knowledge from machine learning models trained across multiple health systems to identify generalizable, value-specific risk drivers and biomarker interactions contributing to AKI risk. Methods: We analyzed electronic health records (EHRs) from 785,497 adult inpatients between 2010 and 2019 across nine U.S. academic medical centers within PCORnet. Interpretable gradient boosting machine models were independently developed at each health system to quantify predictor-outcome associations. Meta-regression was applied to integrate these site-level results, characterize nonlinear value-risk relationships, and identify bivariate interactions between predictors. Results: Meta-analysis revealed consistent, value-specific risk drivers across health systems. An increase in glucose from 100 mg/dL to 140 mg/dL was associated with a 1.46-fold higher risk of AKI. Chloride and anion gap also demonstrated elevated AKI risk with risk increases overlapping portions of their reference ranges, with anion gap showing a 1.14-fold increase across 4-12 mmol/L and chloride a 1.28-fold increase across 96-100 mEq/L. Electrolytes including potassium, calcium, and sodium showed quadratic associations with AKI risk. Bivariate meta-regression identified interactions between key predictors, highlighting pathways that jointly modulate AKI risk. Conclusion: This cross-system meta-analysis synthesizes machine learning-derived evidence into clinically interpretable knowledge, revealing how specific biomarker ranges and interactions modulate AKI risk. By moving beyond surface-level associations to quantitative, generalizable physiologic thresholds, these findings provide actionable insights to enhance risk stratification and personalized prevention in hospital care. Competing Interest Statement JAK discloses paid consulting from BioMerieux, AstraZeneca, Bayer, Chugai Pharma, Mitsubishi Tanabe, and Novartis, and is a full-time employee of Spectral Medical, all unrelated to the present work. KMM has led and participated in research studies funded by Pfizer, Amgen, and Eli Lilly, all unrelated to the present work. AY has served as consultant or on advisory boards for Regulus, Calico, Janssen and Sarepta, none of which are related to the current work. Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The Institutional Review Boards of the University of Florida, the University of Pittsburgh, and the University of Missouri waived ethical approval for this work. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes Data Availability The data used in this study were de-identified EHRs from the Greater Plains Collaborative (GPC) and PaTH clinical research networks, that is not publicly available due to institutional and regulatory restrictions. Access to the data is governed by GPC and PaTH coordinating centers and requires institutional data use agreements and approval to protect patient privacy and compliance with ethical and legal considerations. Researchers interested in accessing similar data may contact GPC and PaTH front doors to inquire about potential data access. Funder Information Declared NIDDK , R01DK116986 NSF Smart and Connected Health , 2444044 Copyright The copyright holder for this preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. It is made available under a CC-BY-NC-ND 4.0 International license . Back to top Previous Next Posted September 02, 2026. Download PDF Supplementary Material Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Cross-System Meta-Analysis of Machine Learning Predictors Identifies Value-Specific Risk Drivers and Interactions Underlying Acute Kidney Injury Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Cross-System Meta-Analysis of Machine Learning Predictors Identifies Value-Specific Risk Drivers and Interactions Underlying Acute Kidney Injury Ho Yin Chan , Deyi Li , Alan S.L. Yu , John A. Kellum , Dana Y. Fuhrman , Qi Xu , Elizabeth A Chrischilles , Lindsay G. Cowell , Sravani Chandaka , Alfred Jerrod Anzalone , Jacob Kean , Kathleen M. McTigue , Abu Saleh Mohammad Mosa , Bradley Taylor , Mahanaz Syed , Lemuel R. Waitman , Yong Hu , Mei Liu medRxiv 2026.08.31.26361849; doi: https://doi.org/10.64898/2026.08.31.26361849 Share This Article: Copy Citation Tools Cross-System Meta-Analysis of Machine Learning Predictors Identifies Value-Specific Risk Drivers and Interactions Underlying Acute Kidney Injury Ho Yin Chan , Deyi Li , Alan S.L. Yu , John A. Kellum , Dana Y. Fuhrman , Qi Xu , Elizabeth A Chrischilles , Lindsay G. Cowell , Sravani Chandaka , Alfred Jerrod Anzalone , Jacob Kean , Kathleen M. McTigue , Abu Saleh Mohammad Mosa , Bradley Taylor , Mahanaz Syed , Lemuel R. Waitman , Yong Hu , Mei Liu medRxiv 2026.08.31.26361849; doi: https://doi.org/10.64898/2026.08.31.26361849 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Areas All Articles Addiction Medicine (614) Allergy and Immunology (896) Anesthesia (326) Cardiovascular Medicine (4770) Dentistry and Oral Medicine (474) Dermatology (407) Emergency Medicine (643) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1614) Epidemiology (15835) Forensic Medicine (32) Gastroenterology (1194) Genetic and Genomic Medicine (6983) Geriatric Medicine (727) Health Economics (1057) Health Informatics (4982) Health Policy (1426) Health Systems and Quality Improvement (1736) Hematology (581) HIV/AIDS (1336) Infectious Diseases (except HIV/AIDS) (16244) Intensive Care and Critical Care Medicine (1161) Medical Education (663) Medical Ethics (151) Nephrology (720) Neurology (7194) Nursing (363) Nutrition (1074) Obstetrics and Gynecology (1231) Occupational and Environmental Health (1001) Oncology (3573) Ophthalmology (1036) Orthopedics (393) Otolaryngology (446) Pain Medicine (466) Palliative Medicine (139) Pathology (704) Pediatrics (1772) Pharmacology and Therapeutics (733) Primary Care Research (759) Psychiatry and Clinical Psychology (5832) Public and Global Health (9691) Radiology and Imaging (2387) Rehabilitation Medicine and Physical Therapy (1432) Respiratory Medicine (1241) Rheumatology (633) Sexual and Reproductive Health (766) Sports Medicine (576) Surgery (766) Toxicology (107) Transplantation (303) Urology (286)
## Related Clinical Research

- [Long-term Renal Outcomes After Neonatal and Pediatric Renal Vein Thrombosis](https://medichelpline.com/clinical-feed/medrxiv-15-renal-outcomes-in-survivors-of-neonatal-and-pediatric-renal-vein-thrombosis.md)
- [Nintedanib-Linked Glomerular Endothelial Injury and Secondary Collapsing FSGS: SGLT2 Inhibitor as](https://medichelpline.com/clinical-feed/pubmed-42674550.md) (DOI: 10.1111/nep.70274)
- [Infection Does Not Impair Decongestion or Short-Term Kidney Outcomes in Cardiorenal Syndrome Type 1](https://medichelpline.com/clinical-feed/plos-one-11-impact-of-infection-on-decongestion-and-kidney-outcomes-in-patients-with.md)
- [Iron-related Parameters and Sepsis-associated Acute Kidney Injury — Source Content Not Available](https://medichelpline.com/clinical-feed/frontiers-in-immunology-19-association-of-iron-related-parameters-with-the-development-of-sepsis.md)
- [Consensus recommendations to prevent and manage immunoglobulin therapy adverse reactions in primar](https://medichelpline.com/clinical-feed/frontiers-in-immunology-4-expert-consensus-recommendations-for-the-prevention-and-management-of.md)

## Navigation
- [← Back to Nephrology Feed](https://medichelpline.com/clinical-feed/nephrology.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.