A simple tool to predict end-stage renal disease within 1 year in elderly adults with advanced chronic kidney disease

Abstract

OBJECTIVES: To develop and validate a model to predict 1-year risk of end-stage renal disease (ESRD) in elderly subjects with advanced chronic kidney disease (CKD).

DESIGN: Retrospective.

SETTING: Veterans Affairs Medical Center.

PARTICIPANTS: Individuals aged 65 and older with CKD with an estimated glomerular filtration rate (eGFR) less than 30 mL/min per 1.73 m(2) .

MEASUREMENTS: The outcome was ESRD within 1 year of the index eGFR. Cox regression was used to develop a predictive model (Veterans Affairs (VA) risk score) that was validated in a separate cohort.

RESULTS: Of the 1,866 participants in the developmental cohort, 77 developed ESRD. Risk factors for ESRD in the final model were age, congestive heart failure, systolic blood pressure, eGFR, potassium, and albumin. In the validation cohort, the C index for the VA risk score was 0.823. The risk for developing ESRD at 1 year from lowest to highest tertile was 0.08%, 2.7%, and 11.3% (P < .001). The C-index for the recently published Tangri model in the validation cohort was 0.780.

CONCLUSION: A new model using commonly available clinical measures shows excellent ability to predict the onset of ESRD within the next year in elderly adults. The Tangri model also had good predictive ability. Individuals and physicians can use these risk models to inform decisions regarding preparation for renal replacement therapy in individuals with advanced CKD.