Seeking common ground on CKD screening

ACP's 2013 guideline on screening, monitoring, and treatment of stage 1 to 3 chronic kidney disease (CKD) did more than provide recommendations based on review of the evidence. It also raised an immediate response from leaders of the American Society of Nephrology (ASN), which felt the College missed the big picture of CKD prevention when developing the guideline.

Cordial subsequent talks have diminished the emotional tenor of the conversation and emphasized areas of common ground. However, there are still some disagreements, from how to view the potential harms of screening—false-positives, increased costs and testing, and labeling—to what constitutes a useful guideline and how primary care physicians should interpret it for their daily practice.

“It's not good guys versus bad guys. It's just 2 different perspectives,” said Molly Cooke, MD, MACP, immediate past president of the College and professor of medicine and director of education for global health sciences at the University of California, San Francisco, School of Medicine. Both ACP and ASN have the best interests of patients at heart, she emphasized.

The guideline (“Screening, Monitoring, and Treatment of Stage 1 to 3 Chronic Kidney Disease: A Clinical Practice Guideline from the American College of Physicians”), published online Oct. 22, 2013, and in the Dec. 17, 2013, print edition of Annals of Internal Medicine, was sponsored by the Agency for Healthcare Research and Quality. It made 4 recommendations (see sidebar, “ACP's recommendations on screening, monitoring, and treatment for chronic kidney disease, stages 1 to 3”).

While there was general agreement about 3 of the 4 recommendations, the first recommendation led to the most controversy.

That recommendation stated that “ACP recommends against screening for chronic kidney disease in asymptomatic adults without risk factors for chronic kidney disease (Grade: weak recommendation, low-quality evidence).”

The ASN issued a press release immediately after the guideline was published calling the position “irresponsible,” a serious accusation in the world of guidelines and especially so for the College, which prides itself on its rigorous guidelines process and reputation (see sidebar, “What's in a guideline?”).

ASN president Sharon M. Moe, MD, FACP, said the heated response stemmed from concern among nephrologists about the recommendation's wording, whether internists would read the entire guideline thoroughly enough to understand that there is agreement about testing patients with a risk factor for CKD, and the fact that kidney disease is often asymptomatic.

Instead of “Don't screen,” she said better wording would have been “Screen patients with diabetes, hypertension, a family history of CKD, obesity, and other risk factors.” This would emphasize how common CKD is, she noted.

But that wasn't the point of the guideline, said Dr. Cooke, who is on ACP's clinical guidelines committee. “We were thinking about screening ... [as] something you do for the well population who you don't really suspect has the disease,” she said.

Different views

As a nephrologist and a member of ASN who was also an ACP president, Virginia Hood, MBBS, MPH, MACP, relates to ASN's concern that CKD affects a large population—about 22 million people, according to the ACP guideline—and that physicians need to identify those at risk for progression or other adverse consequences of early-stage disease. She acknowledged that kidney specialists are keyed into prevention because they know it might have helped some of the patients they see who have end-stage renal disease and are on dialysis.

She also believes that despite the rhetoric, ACP and ASN have the same concerns for minimizing the impact of CKD.

“ACP was not saying we shouldn't look for people with [kidney disease] or try to stop its progression,” said Dr. Hood, a professor of medicine at the University of Vermont in Burlington. “I agree with Dr. Moe that the emphasis for clinicians should be on screening for those with risk factors, and this is stated in the guideline.”

However, she noted, the guideline was addressing population screening in those not at risk. Concerns about prevention don't change the published evidence that was examined to create the guideline.

Unfortunately, she said, the only evidence for or against the recommendation was weak, as the guideline clearly stated. The controversy is not about the evidence, but what clinicians should do when evidence is lacking, she said.

Dr. Cooke said she wasn't surprised by the intensity of the response. The guideline may have rankled the nephrologists, she said, because it comes on the heels of a 10-year effort to make internists sensitive to the spectrum of impairment in the kidney and any potential benefit in finding people at risk before they need dialysis.

Amir Qaseem, MD, PhD, MHA, FACP, director of clinical policy for ACP, attributes the response's intensity to confusion between screening and case finding.

Everyone agrees that diabetes, hypertension, and cardiovascular disease are known risk factors. “Screening identifies people with no sign or symptoms of the disease in whom early intervention will lead to improved clinical outcomes, compared to deferring treatment until symptomatic disease has developed,” he said.

But case finding is a different situation. “Case finding is in patients who already have signs and symptoms,” Dr. Qaseem explained.

Because of the large proportion of Americans with at least one risk factor for CKD, most patients seen in a primary care setting will have a risk factor. However, physicians should not wait for symptoms, said Dr. Moe, the Stuart A. Kleit professor of medicine and director of the division of nephrology at the Indiana University School of Medicine in Indianapolis. Dr. Moe said ASN's response reflects the daily experiences of its members and their mission to stop CKD in its tracks.

“The vast majority of patients seen in their primary care physician's office would benefit from a creatinine test because they have at least one risk factor for CKD,” she said.

The College stands by its rigorous guideline process, the guideline, and its perspective, Dr. Qaseem emphasized.

“There's no disagreement that prevention is better than treatment,” he said. “But the burden of proof for value or benefit of mass screening in an asymptomatic population is higher than it is for testing symptomatic patients because screening is associated with harms. I'm still waiting for someone to come back and—not based on what they do in their practice—point me to the good evidence for screening an asymptomatic population.”

Debate details

Here is an in-depth look at the debate.

Guideline grading. Dr. Moe said she is concerned that internists won't fully understand that the College's recommendation against CKD screening in asymptomatic adults without risk factors is graded as “weak,” meaning that the guidelines committee found no evidence supporting screening in people without risk factors but believes that future research might change this.

“The guideline looked for an answer to a specific question: If an otherwise healthy individual happens to be in their primary care physician's office with no risk factors, should they get a creatinine? Such a study has just not been done, in part because most patients seen in clinics do have a risk factor,” said Dr. Moe. “The lack of evidence should not be mistaken for evidence demonstrating that it should not be done.”

The College's CKD guideline committee decided that physicians should know before doing something whether or not there is evidence for that practice, even if the recommendation is weak.

“Weak grading of the recommendation does not mean no evidence. We have evidence showing harms,” Dr. Qaseem contended. “However, there is no evidence for benefit, so why adopt screening as a default?”

False-positives. Screening can lead to false-positives and thus subsequent additional costs, tests, and treatment, said Dr. Qaseem. Widespread testing, he said, leads to inappropriate diagnoses and treatments with no benefit to the patients. A particular concern is that normal values for estimated glomerular filtration rate (GFR) are not adjusted for age, despite the fact that there is an expected age-related decrement in GFR that does not reflect or lead to CKD.

There are false-positive rates of 13% and 18% for GFR estimation with the CKD Epidemiology Collaboration equation and the Modification of Diet in Renal Disease calculator, Dr. Qaseem noted. “Think about all the unnecessary testing and ... unnecessary treatment.”

Costs versus value. ASN's Dr. Moe pointed out that the initial testing for CKD is inexpensive—depending upon the health care system, less than $10 for a serum creatinine level and less than $20 for a urine albumin level.

But focusing on costs misses the point of population screening, according to Dr. Qaseem. “Cheap is a relative term. Stop looking at just the cost. Start looking at the value based on benefits, harms, and costs,” he said. “Even just the $10 per person for unnecessary screening for CKD in everyone in the U.S. will add to almost $2.5 billion.”

For example, an expensive intervention can provide high value where the benefits clearly outweigh the harms. But a low-cost test doesn't if it doesn't do the same, he explained.

Dr. Cooke understands ASN's thinking: The creatinine level is a good, inexpensive test. “I'm glad it doesn't cost $1,500 and produce horrible side effects,” she said. “But that doesn't affect the conclusion that it isn't a useful practice.”

Labeling. Citing a July 30, 2013, BMJ article (available online) , Dr. Hood said increased labeling of patients with CKD is a function of expanding definitions for the disease, but with potential negative consequences. Those include additional costly and potentially harmful tests and psychological stress for the patient.

Beckie Michael, DO, disagrees. In a routine situation, patients get a urinalysis and a creatinine test, said the nephrologist from Marlton Nephrology and Hypertension in Marlton, N.J. “So we don't cause harm [by ordering creatinine for CKD screening],” she said.

But there is psychological stress, said ACP Member Mark M. Ryan, MD, MBA. “I don't think most patients grasp that abnormal creatinine values do not always signal chronic kidney disease. One abnormal test can lead to a repeat metabolic panel, urinalysis, urine studies, an ultrasound, and a consultation with a nephrologist for ultimately no reason at all,” said Dr. Ryan, a general internist at Woodlands Medical Specialists in Pensacola, Fla.

Dr. Moe emphasized that one creatinine test does not label a patient as having CKD. Before that happens, a repeat creatinine is done, not a huge array of other tests unless the creatinine is quite elevated, she said. If CKD is then diagnosed, there is a chance to prevent disease progression by dictating the choice of antihypertension medications and allowing tighter glycemic control, dietary changes, and other appropriate responses.

A diagnosis also empowers patients, who then know to avoid toxins, such as contrast agents and over-the-counter pain medications, that might put them and their kidney health at risk, she noted. “If I were a patient, I would want to know,” Dr. Moe said.

Clinical practice impact

Although Dr. Cooke said ACP and ASN are now on the same page thanks to a face-to-face meeting, the organizations have agreed to disagree on certain points. The most important issue is how ACP's guideline should impact your practice.

Dr. Moe said that the first question primary care physicians should ask is “Who should I screen?” Examples include a patient with a long-standing history of hypertension, a newly diagnosed diabetic patient, and an asymptomatic patient with a family history of CKD.

Other risk factors include diabetes, cardiovascular disease, older age, use of high-dose over-the-counter NSAIDs, and obesity. For those patients, primary care physicians should measure serum creatinine and determine an estimated GFR. If slightly abnormal, a repeat creatinine is warranted. Then the next test should be to check for albumin or protein in the urine, Dr. Moe advised. Dr. Hood added that it is it is important to know the kidney function of patients undergoing certain diagnostic tests or treatments.

“Ultrasounds are rarely warranted,” Dr. Moe said. She pointed out that they were a candidate for ASN's final list of “Five Things Physicians and Patients Should Question,” which was included in the ABIM Foundation's “Choosing Wisely” campaign. “So the real problem with overdiagnosis and cost containment lies in an inappropriate or premature response to an elevated creatinine, not in the creatinine test itself,” she said.

Since there's no evidence to support having a 44-year-old healthy new patient get a creatinine test just to see what it is, Dr. Cooke recommends taking it off the general panel. Instead, know which of your patients are at risk for CKD and then order the test, she said.

However, with so many good reasons to order a basic metabolic panel that includes kidney function, it makes sense to do so, said Dr. Michael, chair of ASN's practicing nephrologists advisory group. “Patients need to know and doctors need to know if there's a problem with kidney function,” she said. “It affects how we prescribe medications ... and treatment options for other illnesses.”

The guideline has already changed Dr. Ryan's practice. Since it was published, he said, he no longer includes a metabolic panel order for new patients unless the patient has a risk factor.

For existing patients who may have routinely had the test previously with normal results, he notes that their test wasn't abnormal before and asks whether they're on new medications or have new symptoms. If the answers are no, he tells them the test isn't necessary and thus insurance won't cover it.

He's finding that some patients ask for the test if he doesn't suggest it. “I say, ‘You don't need to have a $15 test done’ and they look at me like I'm crazy,” he said.

But if they want to pay for the test—and he said most do—he will interpret it for them.

Ultimately, internists should remember that guidelines are just that, not rules, often because there isn't a straightforward answer, Dr. Hood said.

“All we can do is work with the best thinking we have at the time and do things in the best interests of our patients to help and not harm them, and always look for new information,” she said.