Kidney Week

Abstract: SA-PO754

Validation of Predictive Model for Progression of CKD to ESRF in Singapore

Session Information

• CKD: Epidemiology, Risk Factors, Prevention - III
  October 27, 2018 | Location: Exhibit Hall, San Diego Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: CKD (Non-Dialysis)

• 1901 CKD (Non-Dialysis): Epidemiology, Risk Factors, and Prevention

Authors

• Pang, Jolyn Hui qing, National University of Singapore, Singapore, Singapore

Jolyn Hui qing Pang,

• Kwek, Jia Liang, Singapore General Hospital, Singapore, Singapore

Jia Liang Kwek,

• Li, Huihua, Singapore General Hospital, Singapore, Singapore

Huihua Li,

• Lim wei wei, Lydia, Singapore General Hospital, Singapore, Singapore

Lydia Lim wei wei,

• Choo Chon Jun, Jason, Singapore General Hospital, Singapore, Singapore

Jason Choo Chon Jun,

• Choong, Lina, Singapore General Hospital, Singapore, Singapore

Lina Choong,

• Chan, Choong Meng, Singapore General Hospital, Singapore, Singapore

Choong Meng Chan,

• Foo, Marjorie Wai Yin, Singapore General Hospital, Singapore, Singapore

Marjorie Wai Yin Foo,

Background

Risk stratification of chronic kidney disease (CKD) patients allows clinicians to individualize treatment and improves resource allocation on the national level. Tangri et al (JAMA. 2011 Apr 20;305(15):1553-1559) proposed and validated a predictive model for progression of chronic kidney disease to kidney failure in a Canadian population and subsequently in a meta-analysis of multinational cohorts. As accuracy of the prediction model may vary among different populations, we aim to validate this prediction model in a Singapore CKD cohort and to determine if a calibration factor is required.

Methods

The study population was derived from newly referred patients to Department of Renal Medicine at Singapore General Hospital (SGH), Singapore in 2009. Eight variables based on the most accurate model in the Tangri et al study were obtained within 90 days of the initial visit. Primary outcome measure is kidney failure in 5 years and time to kidney failure was defined from initial renal medicine visit to kidney failure.

Results

Out of 2216 patients reviewed, 795 were included in the analysis. The mean age is 65.8 years old, mean eGFR is 34.5ml/min/1.73m². Majority are Chinese (74.6%) with 15.5% Malay, 6.2% Indian, 1.1% Eurasian and 0.6% others (2 Thais, 1 Filipino, 1 Indonesian and 1 Arab). Six hundred and forty (80.4%) are on angiotensin converting enzyme inhibitor (ACEi) or angiotensin receptor blocker (ARB). Four hundred and ninety-seven (62.4%) have diabetes mellitus. Two hundred and twelve (26.7%) reached kidney failure. Mean time to renal failure is 1288 days (3.5 years).

Both the 8-factor (age, gender, estimated glomerular filtration rate (CKD-EPI) (eGFR), albuminuria (ACR), serum albumin, serum phosphate, serum bicarbonate and serum calcium) and 4-factor (age, gender, eGFR and ACR) model are accurate (C-index 0.863 and 0.865 respectively) in predicting risk of kidney failure with the 8-factor model outperforming the 4-factor model (adequacy index of 98.8% vs 97.2%)

Conclusion

The predictive model developed by Tangri et al is accurate in predicting the progression of CKD to ESRF in 5 years in the Singapore CKD population.