Abstract: SA-OR016
Performance of the PREVENT Risk Score in Predicting Mortality Among Adults with CKD: Analysis of NHANES, 2005-2018
Session Information
- CKD: Advancing Epidemiology, Risk Factors, and Prevention
November 08, 2025 | Location: Room 362A, Convention Center
Abstract Time: 05:30 PM - 05:40 PM
Category: CKD (Non-Dialysis)
- 2301 CKD (Non-Dialysis): Epidemiology, Risk Factors, and Prevention
Authors
- Ali, Sundus, University Hospitals Health System, Cleveland, Ohio, United States
- Makhlouf, Mohamed, University Hospitals Health System, Cleveland, Ohio, United States
- Al-Kindi, Sadeer, Houston Methodist, Houston, Texas, United States
- Dobre, Mirela A., University Hospitals Health System, Cleveland, Ohio, United States
Background
The Predicting Risk of Events via Estimated Net Trajectories (PREVENT) score, developed in general populations, estimates long-term mortality risk using clinical and laboratory parameters. Its performance in individuals with chronic kidney disease (CKD), particularly in more advanced stages, remains uncertain.
Methods
We evaluated the predictive utility of the PREVENT score for all-cause and cardiovascular (CVD) mortality among adults with CKD stages 2–5 using NHANES data from 2005–2018, with mortality follow-up through 2019 via the National Death Index. The PREVENT score was calculated per its published algorithm. All-cause mortality and CVD mortality were defined using ICD-10 codes. Discrimination was evaluated using area under the receiver operating characteristic curve (AUC) in the overall cohort and stratified by race (non-Hispanic White vs. non-Hispanic Black), sex, age (<60 vs. ≥60 years), and CKD stage. Individuals with missing mortality follow-up were assumed to be alive.
Results
Among 16,403 study participants, there were 3,151 deaths, including 674 attributed to CVD. The PREVENT score showed modest discrimination overall (AUC 0.60 for all-cause and 0.59 for CVD mortality, Figure). Predictive performance for all-cause mortality declined progressively with worsening kidney function: Stage 2 (AUC 0.67), Stage 3a (0.61), Stage 3b (0.57), Stage 4 (0.57), and Stage 5 (0.52). Subgroup AUCs for all-cause mortality were: White 0.58 vs. Black 0.60; Male 0.59 vs. Female 0.58; Age <60 years 0.61 vs. ≥60 years 0.51.
Conclusion
In a nationally representative cohort of adults with CKD, the PREVENT score demonstrated modest predictive ability for all-cause and CVD mortality, with performance decreasing as CKD severity increased. Discrimination was highest in younger adults, and was consistent across race and sex. These findings suggest the PREVENT score may aid in mortality risk stratification in early CKD but highlight the need for improved models in advanced CKD.