The problem with the K/DOQI stages

In 2002 at the Spring Clinical Meeting of the National Kidney Foundation, K/DOQI released the Clinical Practice Guidelines for Chronic Kidney Disease Evaluation, Classification and Stratification.

These guidelines have become the dogma of CKD and all of my residents can accurately determine the CKD stage of their patients. The classifications have allowed epidemiologists to measure the burden of CKD. The crux of the guideline is that the severity of kidney disease is solely determined by the GFR.  This is helpful in determining where the patient has been but it is not good at determining where patients are headed.

In some ways, it is a negative prognostic tool, people with worse stages of CKD actually have better outcomes and vice versa.

To understand how this works one needs to understand how we calculate the GFR. The accepted equation was created by Levey et al using the MDRD data base. Levey AS, Greene T, Kusek JW, Beck GJ: A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 11:A0828, 2000 (abstract)

GFR=186 x sCr -1.154 x Age -0.203 x (0.742 if female) x (1.212 African-American)

African Americans, for the same creatinine, are given a 21% increase in their GFR and women lose 26%. If GFR provided prognostic information one would think that African Americans were protected from chronic kidney disease and women were at higher risk. Yet that is not the case. African Americans have the highest rates of ESRD, 998 per million compared to 273 for white Americans: (USRDS 2009 Annual Data Reports, NIH, NIDDK, Bethesda, MD, 2009.)

The lifetime risk of ESRD is 1 in 50 for woman and 1 in 40 for men (Kiberd BA, Clase CM. JASN 13: 1635-44, 2002).

The negative exponent on the age variable means that as one ages their GFR falls. The problem is that the risk for ESRD appears to fall, likely because of the competing endpoint, death. This was best shown in a 10 year study by Erikson and Ingebretsen. They showed that as patients aged they were more and more likely to die and less likely to develop ESRD.
I do not doubt that the MDRD eGFR is a good measure of renal function I am frustrated by the way it is used to frame management decisions.
Its like the nephrologists in the K/DOQI work group looked at the framework established in oncology and adopted a staging system but did not consider the importance of the grading system. Not all CKD stage threes are alike, some patients have more aggressive disease than others and this needs to be addressed in our clinical practice guidelines.