Kidney Week

Abstract: FR-PO1139

Performance of Three Creatinine-Based Equations in Predicting All-Cause Mortality and Cardiovascular Events in a Multiethnic Asian Population: The Singapore Epidemiology of Eye Diseases Study

Session Information

• CKD: Kidney Function and Extrarenal Complications
  October 25, 2024 | Location: Exhibit Hall, Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: CKD (Non-Dialysis)

• 2302 CKD (Non-Dialysis): Clinical, Outcomes, and Trials

Authors

• Liew, Zhong Hong, Singapore General Hospital, Singapore, Singapore

Zhong Hong Liew, MD

• Chee, Miao li, Singapore Eye Research Institute, Singapore, Singapore

Miao li Chee

• Lim, Cynthia Ciwei, Singapore General Hospital, Singapore, Singapore

Cynthia Ciwei Lim, MBBS

• Cheng, Ching-Yu, Singapore Eye Research Institute, Singapore, Singapore

Ching-Yu Cheng, MD, PhD, MPH

• Sabanayagam, Charumathi, Singapore Eye Research Institute, Singapore, Singapore

Charumathi Sabanayagam, MD, PhD

Background

The comparative prediction of novel creatinine-based equations for all-cause mortality and cardiovascular events is not well established. We aimed to compare risk prediction for all-cause mortality and cardiovascular events by estimated glomerular filtration rate (eGFR) using CKD-EPI creatinine equation 2009 (Creat-ASR 2009), CKD-EPI creatinine equation 2021 (Creat-AS 2021) and CKD-EPI creatinine-cystatin equation (Creat-cys 2021) in a multiethnic Asian population.

Methods

In this population-based cohort study, we included Chinese and Indian adults aged 40-80 years who attended baseline visit (2007-2011). Serum cystatin C was measured using particle-enhanced turbidimetric assay. Information on death and cardiovascular events was obtained by data linkage with National Registry of Diseases Office until 31 Mar 2021. Outcomes were all-cause mortality and incident cardiovascular events among those without prior cardiovascular disease. Using Cox proportional hazards model, we conducted multivariate regression analyses to evaluate association of CKD and the outcomes. Net reclassification improvement was performed to compare prediction performance of these 3 equations.

Results

During a mean follow-up of 11.3±2.2 years, all-cause mortality rate was 12.9% (743 of 5738 participants). Using Creat-ASR 2009, Creat-AS 2021, and Creat-cys 2021, mortality rates among those with CKD (eGFR < 60 ml/min/1.73 m²) were 6.67% vs 5.04% (P <0.001) vs 6.26% (P <0.001), respectively. After excluding those with existing cardiovascular disease, incident cardiovascular event rate was 9.9% (508 of 5120 participants).
Reclassification using Creat-AS 2021 compared with Creat-ASR 2009 had poorer prediction performance for all-cause mortality and cardiovascular event (Creat-AS 2021 vs Creat-ASR 2009: NRI= -0.040, p<0.001 and -0.050, p<0.001, respectively). Reclassification using Creat-cys 2021 compared with Creat-ASR 2009 allowed better prediction for all-cause mortality and cardiovascular event (Creat-cys 2021 vs Creat-ASR 2009: NRI = 0.077, p<0.001 and 0.091, p<0.001, respectively).

Conclusion

Creat-cys 2021 was associated with improved risk prediction of all-cause mortality and cardiovascular event, compared with Creat-ASR 2009 and Creat-AS 2021.