“We acknowledge that there have been no randomized trials of CKD screening,” Mr. Roach noted, “but our recommendations are based on extensive analysis of observational trials, clinical trials of diabetes and antihypertensive drugs, and clinical experience.”
Critics question CKD criteria
A recent U.S. Preventive Services Task Force (USPSTF) report took a position similar to the ACP’s, stating there is currently insufficient evidence to recommend general population-based screening. That report noted that although early screening could affect outcomes for people already diagnosed with conditions such as diabetes or hypertension, no evidence existed to show the benefits of early treatment in people without these risk factors.
The ASN swiftly responded. "We sent [the USPSTF] a letter asking them to reconsider," said Dr. Molitoris. "That was within the last 6 months. They came back with a ‘high recommendation’ for collecting more data to analyze the cost effectiveness of screening."
In 2002, the NFK-sponsored Kidney Disease: Improving Global Outcomes (KDIGO) released landmark guidelines on CKD. In a 2012 guidelines update, KDIGO kept the criteria it established in 2002 for CKD, including diagnostic thresholds of a GFR less than 60 mL/min per 1.73 m2 and an albumin/creatinine ratio (ACR) of at least 30 mg/g. And it maintained its five-stage CKD classification system.
But it expanded the existing five-stage classification system to include two subcategories of grade 3 GFR categories – G3a, with a GFR of 45-59; and G3b, with a GFR of 30-44 – and three additional categories of albuminuria (Ann. Intern. Med. 2013;158:825-30).
Those CKD criteria raised eyebrows not just at the ACP, but also with some evidence-based medicine scholars. In an analysis published online earlier this year in BMJ, Dr. Roy Moynihan of the Centre for Research in Evidence-Based Practice, Robina, Australia, cited data that the GFR value used by KDIGO to determine early CKD falls within the normal range for men over age 60 and women over age 50.
By "labeling so many people at low risk of symptoms as having chronic kidney disease, the new definition axiomatically produces overdiagnosis," Dr. Moynihan and his colleagues cautioned (BMJ 2013;347:f4298).
Prior to the 2002 publication of the KDIGO criteria for CKD, according to Dr. Moynihan, the incidence rate for CKD in U.S. adults was estimated to be around 1.7%. After 2002, using the KDIGO criteria, it rose to 1 in 8 adults, or about 14% of the adult population.
"The controversial aspect around that is, at what point as your GFR decreases does it become abnormal?" cautioned Dr. Molitoris. "If you have a reduced GFR, and proteinuria, we know that’s abnormal. That’s why the screening for proteinuria is so important. If you just have a reduced GFR to a certain extent, it may be a normal aging process. That’s the European perspective, and I think they have a point there."
‘Different worldview’
While the ACP guidelines for CKD refer to the KDIGO criteria as their counterpoint, Dr. Cooke said she did not know what motivated KDIGO to expand the criteria. She suggested it might be the "different worldview" phenomenon that often occurs between subspecialists and generalists.
"It’s tempting to say, ‘Let’s just test everybody’s creatinine,’ " Dr. Cooke noted. "But it’s just like with prostate cancer screening: Once you start to factor in all the costs – and not just the money – of testing on that broad of a scale, it doesn’t [add up]."
But Dr. Molitoris countered that CKD screening is more akin to testing for cardiovascular disease.
"One could make the comparison to checking for cholesterol, only that’s invasive – that’s drawing blood," he said. "This is just taking urine, and yet, how many people do we screen for cholesterol? There is no risk in doing the proteinuria test. You collect the urine, put a stick into it, and read it. If it’s positive, then you will follow up with a more sophisticated chemical test." He also noted that false positives in proteinuria tests are rare.
The ACP guidelines offer physicians what it calls "high-value care advice": Skip early CKD screening, because in the absence of evidence that screening improves clinical outcomes, "testing will add costs, owing to both the screening test and additional follow-up tests (including those resulting from false-positive findings), increased medical visits, and costs of keeping or obtaining health insurance."
Dr. Molitoris disagreed, citing "the aggressive and silent nature" of kidney disease in all its stages.
"You have to think of the cost analysis as how many screenings you can do for the cost of one patient who goes on to need dialysis. That costs between $80,000 and $90,000 a year to maintain for every year [a patient is] on it," Dr. Molitoris explained. "You can do a lot of screening on that."