Ultrafiltration Rate Reporting Could Lead to Longer Dialysis Times

The Centers for Medicare & Medicaid Services (CMS) has proposed introducing an ultrafiltration rate quality measure into its End-Stage Renal Disease Quality Incentive Program (QIP). A study of competing models of this quality measure found that meeting the standard is likely to require lengthening dialysis treatment times by durations that could prove challenging to dialysis unit operations.

CMS has been considering adding this measure for some time because fast ultrafiltration rates are associated with adverse outcomes, although data on direct links is far from definitive.

After CMS first proposed its model for the standard, the Kidney Quality Care Alliance (KQCA) responded with a proposal of its own. Both proposals use a benchmark of 13 milliliters per hour per kilogram of body weight as the upper acceptable limit, but there are two major differences in the plans. First, the CMS model relies on data from a single treatment, whereas KQCA uses the mean of three treatments in a week. Second, the KQCA proposal gives a facility credit for dialysis sessions that last four hours or more, regardless of the actual ultrafiltration rate.

A study in the August Clinical Journal of the American Society of Nephrology drew on the database of a large dialysis organization to analyze how some 150,000 patients would have fared against each measure. It found that 21–23% of patients would have exceeded the 13 mL benchmark under the CMS rules, and about 16% would have exceeded it under the KQCA rules. Although limiting fluid gain through diet is in many ways the better option, the most likely way facilities will lower ultrafiltration rates is by extending treatment times—and the researchers calculated that a 100-patient facility would need to add 33 treatment hours per week to get all its patients below 13 mL—and that’s using a treatment duration cap of 4 hours (per the KQCA measure).

“That number tripled—up to 98 hours per week—when we removed the 4-hour treatment cap,” said lead study author Jennifer Flythe, MD, MPH, assistant professor of medicine at the University of North Carolina. “So there are some very interesting patient and facility implications that need to be thought through to ensure that there are no unintended consequences from implementation of the ultrafiltration rate measure.”

The study identified some other pitfalls of applying a uniform standard to all patient groups. For example, ultrafiltration rates rose in winter and fell in summer, probably because patients’ hydration levels vary during colder and hotter seasons.

“Certain patient groups tended to have had higher ultrafiltration rates, including patients that were younger, women, nonblack, of Hispanic ethnicity, and smaller in body weight,” Flythe said. Higher rates for smaller patients are not surprising, considering that the measure is by definition indexed to body weight, but the implication of higher rates among these patients is an open question.

Why an ultrafiltration standard?

Flythe said that the physiological underpinnings for the desirability of lower ultrafiltration rates are sound: “The thought is that the faster you pull the fluid off during dialysis, the more cardiac and other organ stress you may be causing.” And the negative consequences are backed up by observational data, but the evidence base is not strong.

“We have an absolute lack of clinical trial data looking at the effects of ultrafiltration rates,” said Daniel Weiner, MD, MS, a nephrologist and associate professor of medicine at Tufts University School of Medicine, and chair of the ASN Quality Metrics Task Force. “Just like pretty much every other metric in dialysis, there are not good randomized trial data comparing different interventions and looking at important clinical outcomes”

Just the same, Weiner believes that volume control “is the next big thing” in dialysis management: “It is much more important in my opinion than anemia management and hypercalcemia.”

He said that when KCQA surveyed the dialysis community about developing new quality measures, fluid management was the top priority.

But he acknowledges the uncertainty of any standard at this point. The 13 mL threshold was endorsed by the National Quality Foundation, but “for some people, 8 mL may be too high, and for some people, an ultrafiltration rate of 18 mL may be OK. But you had to start somewhere. Importantly though, we need a continuing iterative process by which we reassess what may be the optimal filtration rate.”

Eduardo Lacson Jr., MD, MPH, also from Tufts, is not as enthusiastic about the prospect of an ultrafiltration rate measuring standard. He noted that the most direct evidence has come from studies using technologies to prevent intravascular volume depletion—and these studies found that the avoidance of depletion had no effect on outcomes.

Weiner and Lacson co-wrote an editorial that accompanied Flythe’s study and concluded that proposals like this one epitomize the balancing act between the two aphorisms, “the perfect should not be the enemy of the good” and “first do no harm.” But they also note that the QIP already exists, so the kidney community must be ready for new measures being added regularly.

CMS says it is coming

And CMS’ latest draft updates to policies and payment rates for end stage renal disease make it clear that this one is coming. Released on June 24, 2016, the updates propose incorporating an ultrafiltration rate reporting measure into the QIP in payment year 2020. That timing is actually a year’s postponement—the 2015 updates proposed beginning the program in 2019.

The 2016 CMS proposal moves toward parts of the KCQA model, including reliance on a week of testing instead of a single measure. Although this requirement actually increases a facility’s data reporting requirements, Weiner and Lacson said that one reason the KCQA included it was not only to get a more accurate picture of a patient’s status, but also to prevent dialysis facilities from gaming the system. As long as patients are scheduled for three sessions in a seven-day week, there are going to be shorter and longer periods between dialysis sessions, with more fluid building up during the longer breaks. “Patients who come in after the 72-hour gap are probably going to have more fluid taken off than those who come in after the 48-hour gap,” Weiner said. “So if you are a dialysis facility that draws its labs on a Wednesday or Thursday, you are going to look better under the [original] CMS measure than if you draw your labs on a Monday or Tuesday.” In addition, Lacson said that a facility could tailor a single treatment to meet the reporting requirement, temporarily leave the patient slightly fluid overloaded, and remove excess fluid at the next session.

Flythe’s study did in fact find that ultrafiltration rates varied according to the time between treatments, “with greater ultrafiltration rates occurring after the long interdialytic break.”

The updated CMS version did not include an exemption for dialysis sessions of four hours or more. This exemption in the KCQA model was part of the reason more facilities met the guideline compared with the CMS proposal. Weiner said that a session of this duration shows that “the dialysis facility is doing what they can to minimize the ultrafiltration rate. You don’t want dialysis units to limit the amount of fluid they take off someone, leaving them fluid overloaded, just to hit a measure.”

Lacson noted that not only would longer treatment times pose challenges for the operation of a dialysis facility—such as longer hours and juggling patient schedules when treatment durations are unknown—but many patients will not happily greet the prospect of longer treatment.

Implementation plan

CMS plans to score facilities on whether they successfully report the required data in a timely fashion, not on the values reported. Weiner said that CMS often phases in a measuring standard in this fashion to make sure that the data capture is feasible and reliable. He said the ASN Quality Metrics Task Force will be submitting comments on the latest CMS proposal.