All you need to read in the other general journals

In search of better ways to predict, and prevent, end stage kidney disease

Only some patients with chronic kidney disease will progress to kidney failure or experience cardiovascular events. Early identification of these people could help prevent more serious disease. Currently, prediction relies on estimated glomerular filtration rate (eGFR), which is based on creatinine, and possibly albuminuria, but these do not provide enough information for accurate identification of those at highest risk. Two new studies help fill the gap.

In one prospective study of 26 643 US adults, researchers used three markers of chronic kidney disease: eGFR of <60 mL/min/1.73 m², based on creatinine or cystatin C, and albuminuria defined as urine albumin:creatinine ratio of 30 mg/g or higher. Only a quarter of participants without chronic kidney disease according to eGFR based on creatinine were confirmed not to have the disease when both remaining markers were also used. Compared with creatinine alone, the odds of predicting death were three times higher when one more marker was positive, and 5.6 times higher when all three markers indicated chronic kidney disease. Using all three markers improved classification by 13.3% for death and 6.4% for kidney failure.

Another study used a different approach, searching for predictors of kidney failure in a development cohort of 3449 patients, of whom 11% had kidney failure, and validating the model in a separate cohort of 4942 patients (24% had kidney failure). The most accurate model included age; sex; eGFR based on creatinine; albuminuria; as well as serum concentrations of calcium, phosphate, bicarbonate, and albumin. Compared with a model including only age, sex, eGFR based on creatinine, and albuminuria, this model improved the classification of 8% (95% CI 2.1% to 13.9%) of patients with stage 3 chronic kidney disease and 4.1% (−0.5% to 8.8%) with stage 4 disease.

The editorial (doi:10.1001/jama.2011.502) explains why these …