Predictive Model for CKD Progression Shows Promise

Megan Brooks

May 24, 2011

May 24, 2011 (Vancouver, British Columbia) — Researchers say they've developed and validated a simple tool to predict short-term risk for progression to kidney failure in patients with stage 3 to 5 chronic kidney disease (CKD). The model uses 8 variables: age, sex, estimated glomerular filtration rate (eGFR), albuminuria, serum calcium, serum phosphate, serum bicarbonate, and serum albumin.

The research was published in the April 20 issue of the Journal of the American Medical Association, to coincide with its presentation here at the World Congress of Nephrology 2011.

Only a "small fraction" of patients with CKD develop kidney failure, first author and conference presenter Navdeep Tangri, MD, FRCPC, from Tufts Medical Center in Boston, Massachusetts, noted in an interview with Medscape Medical News.

"Although we are aware of some risk factors for progression, there has never been a predictive model or risk equation that integrates them and actually provides a numeric probability of a patient's true risk," he noted.

Development and Validation

To develop and validate such a model, Dr. Tangri's team used demographic, clinical, and laboratory data from 2 independent Canadian cohorts of patients with stage 3 to 5 CKD (eGFR, 10 to 59 mL/min per 1.73 m2) who were referred to nephrologists between April 2001 and December 2008. The primary outcome measured was kidney failure, defined as the need for dialysis or preemptive kidney transplantation.

The development cohort comprised 3449 patients (386 [11%] with kidney failure) and the validation cohort comprised 4942 patients (1177 [24%] with kidney failure).

The 8-variable model proved "highly accurate," the researchers report, with a C statistic of 0.917 (95% confidence interval [CI], 0.901 to 0.933) in the development cohort and 0.841 (95% CI, 0.825 to 0.857) in the validation cohort. This model was more accurate than a 4-variable model that looked at only age, sex, eGFR, and albuminuria.

As an illustration, the researchers describe a 70-year-old man with an eGFR of 30 mL/min per 1.73 m2, a urine albumin-to-creatinine ratio (ACR) of 200 mg/g, and other laboratory parameters suggestive of an adverse outcome. This man would have a 5-year risk for kidney failure of 19.8%, based on age, sex, and eGFR; a 16.3% risk based on age, sex, eGFR, and urine ACR; and a 26% risk when all 8 parameters are considered.

Not Ready for Prime Time?

Marcello Tonelli, MD, SM, FRCPC, from the University of Alberta, Edmonton, Canada, and Braden Manns, MD, MSc, FRCPC, from the University of Calgary, Alberta, comment on the research in an accompanying editorial. "Although technically feasible in many settings," they write, "immediate implementation of this promising strategy would be premature."

They point out that all patients in the analysis were already under the care of a nephrologist. "Because the major value of routinely reporting a risk prediction score would be to identify patients who would benefit the most from referral, external validation in the general population is a logical next step," Dr. Tonelli and Dr. Manns conclude.

Dr. Tangri told Medscape Medical News that further validation studies are underway in several different patient populations, including patients who have not yet been referred to a nephrologist and those in the earlier and later stages of kidney disease.

JAMA. 2011;305:1553-1559, 1593-1595. Abstract, Editorial


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.