Kidney Week

Abstract: SA-OR017

Development and Validation of a Mortality Prediction Model for Conservative Management (CM) vs. Dialysis in Advanced CKD: Analysis of Two National Cohorts

Session Information

• CKD: Advancing Epidemiology, Risk Factors, and Prevention
  November 08, 2025 | Location: Room 362A, Convention Center
  Abstract Time: 05:40 PM - 05:50 PM

Category: CKD (Non-Dialysis)

• 2301 CKD (Non-Dialysis): Epidemiology, Risk Factors, and Prevention

Authors

• Rhee, Connie, University of California Los Angeles, Los Angeles, California, United States

Connie Rhee, MD, MS

• Nguyen, Danh V., University of California Irvine, Irvine, California, United States

Danh V. Nguyen, PhD

• Narasaki, Yoko, University of California Los Angeles, Los Angeles, California, United States

Yoko Narasaki, MS, PhD

• You, Seungsook, University of California Los Angeles, Los Angeles, California, United States

Seungsook You, MS

• Crowley, Susan T., Yale University, New Haven, Connecticut, United States

Susan T. Crowley, MD, MBA, FASN

• Kovesdy, Csaba P., The University of Tennessee Health Science Center Department of Medicine, Memphis, Tennessee, United States

Csaba P. Kovesdy, MD, FASN

• Mukamel, Dana B., University of California Irvine, Irvine, California, United States

Dana B. Mukamel

• Zisman-Ilani, Yaara, Temple University, Philadelphia, Pennsylvania, United States

Yaara Zisman-Ilani, MA, PhD

• Kalantar-Zadeh, Kamyar, The Lundquist Institute, Torrance, California, United States

Kamyar Kalantar-Zadeh, MD, DrMed, PhD, MPH, FASN

Background

We developed and externally validated a prediction model for survival with CM vs. dialysis transition in two national advanced CKD cohorts.

Methods

We developed a mortality risk prediction tool in 61,118 Veterans with advanced CKD (≥2 eGFRs <25 separated by ≥90 days) treated with CM vs. dialysis (non-receipt vs. receipt of dialysis within 2-yrs of 1^st eGFR <25 [index eGFR]) over 2010-19. We externally validated this tool in 76,687 advanced CKD patients treated with CM vs. dialysis from the OptumLabs^® Data Warehouse, which contains de-identified administrative claims, including medical/pharmacy claims and enrollment records for commercial/Medicare Advantage enrollees, and EHR data. Prediction models for 1-yr mortality were developed on the basis of survival data up to 2-yrs after the index eGFR using Cox models.

Results

In the VA cohort (43,197 vs. 17,921 receiving CM vs. dialysis), characteristics associated with death included older age; higher index eGFR, UACR, and VA frailty index; faster eGFR decline; lower albumin and BMI; prior 1-yr hospitalization; heart disease; sepsis; and dialysis transition. Discrimination in the development and internal validation VA cohort yielded C-statistics of 0.70 (95%CI 0.70-0.71) and 0.69 (0.68-0.70). In the OptumLabs^® external validation cohort, the risk score exhibited a C-statistic of 0.69 (0.69-0.70). Risk score/calibration plots for internal vs. external validation cohorts are shown in Fig 1.

Conclusion

A new tool to predict individualized probability of survival with CM vs. dialysis transition was developed and externally validated with acceptable performance in two national advanced CKD cohorts to inform shared decision-making with clinicians and patients.

Funding

• NIDDK Support

Digital Object Identifier (DOI)

• doi: 10.1681/ASN.2025620kmnhx