SAN DIEGO—On the heels of a new dialysis reimbursement system and updated treatment recommendations, data demonstrating changes in clinical practice continue to accumulate, but how long-lasting these trends will be and what effects they'll have on patient outcomes remain to be seen.
Here at the American Society of Nephrology (ASN) Kidney Week 2012, the United States Renal Data System (USRDS) added to that evidence base with analyses from 2009–2011 Parts A and B Medicare claims.
“It is too early to determine exactly how many aspects of patient care are being impacted,” said Allan J. Collins, MD, Professor of Medicine at the University of Minnesota and Director of the USRDS, in an e-mail to Nephrology Times. “The transfusion increase is clear, but the patient-related impact needs to be carefully considered.”
The factors influencing shifts in dialysis practice are multiple. On Jan. 1, 2011, the End-Stage Renal Disease (ESRD) Prospective Payment System (PPS) went into effect, bringing under a bundled payment dialysis services that previously had been separately billable, including injectable drugs like the antianemia erythropoiesis-stimulating agents (ESAs). Then in June 2011, the Food and Drug Administration (FDA)-approved label for ESAs was updated to convey more conservative dosing recommendations. In response, the Centers for Medicare & Medicaid Services (CMS) proposed dropping the payment penalty for average hemoglobin (Hb) levels below 10 g/dL from the ESRD Quality Incentive Program (QIP), later making the change final.
Transfusion rates were flat in 2009 and stayed that way until October 2010, when they started to rise, noted David T. Gilbertson, PhD, Director of Epidemiology and Biostatistics for the USRDS, in his presentation at the ASN meeting. This increase continued in 2011, and the pattern stayed consistent across ages and races.
“Part of the reason for the increase may be that providers are potentially ‘finding their way’ in terms of recognizing not only the financial challenge of using a larger amount of ESAs but also the medical risks associated with higher dose of ESA or targeting higher hemoglobins,” said Rajnish Mehrotra, MD, MS, Professor of Medicine at the University of Washington. “It is possible that providers are in a flux with regard to the protocols they use for administration of ESAs that may be defined as we move forward.
“It is something that we need to monitor very closely because transfusion is something that we want to avoid, as the primary indication for giving ESAs says, and we want to make sure that patients are getting enough ESAs that we are avoiding transfusions.”
The analysis from the USRDS is in line with previously presented information from the data system, as well as reports from the Dialysis Outcomes and Practice Patterns (DOPPS) Practice Monitor (DPM) and CMS. The data are early, however, representing practice only a year after the move to the bundled payment.
“Whether or not this is a sustained trend in increase in transfusion, I would reserve my judgment at this time,” Dr. Mehrotra said. “It would be prudent to wait and watch and see how this evolves.”
A major concern surrounding transfusions in patients with kidney disease is the potential for sensitization, making it much harder to find a compatible organ donor.
“Most people would say that you don't want to be transfusing dialysis patients,” said Suzanne Watnick, MD, Medical Director of the Dialysis Unit at the Portland VA Medical Center and Training Program Director for Nephrology at Oregon Health & Science University. “When you think about the implications regarding kidney transplantation, which is probably the best way to treat a patient with end-stage kidney disease, how will an increased rate of transfusion impact our capacity to transplant patients in the future?
“These are all of the questions that you need to consider as you start to make broad policy changes that are going to affect a very both chronically ill and probably socioeconomically disadvantaged population of patients for whom we have an obligation to give the best quality care possible.”
Changes in ESA Use
As expected, a decline was observed in the proportion of patients receiving any erythropoietin. In 2009, just under 90% received the agent. The proportion began to decrease in 2010 and then remained relatively flat between January 2011 and June 2011 before declining more dramatically, with about 80% of patients receiving some erythropoietin by November 2011. Between early 2010 and late 2011, the proportion of patients getting the agent went down by about 7%.
The number of erythropoietin administrations per month for patients receiving the agent decreased in 2011 and has remained essentially stable through the first three months of 2012.
The monthly dose of erythropoietin went up in 2009 and down in 2010. It declined more dramatically in 2011, with an approximately 14,000-unit drop.
Hemoglobin levels, in turn, have also decreased. The decline began in 2010 and was more marked in 2011, gaining additional momentum in midsummer of that year. Mean levels decreased by about 0.2 g/dL in 2010 and about 0.5 g/dL in 2011. Hemoglobin remained flat at the end of 2011, with perhaps a slight decrease in early 2012.
The pattern in hemoglobin levels was relatively similar for patients taking darbepoetin compared with erythropoietin, as well as across races.
“I think a decrease in hemoglobin level was long overdue,” Dr. Mehrotra said, noting the higher percentage of patients who had hemoglobin levels above 12 g/dL at the time when the QIP was introduced. “That proportion has substantially come down, and it needed to come down recognizing the safety risk.
“I am less concerned about the decrease in average hemoglobin, as long as the pendulum doesn't swing too much to the other side. What is unclear, in reducing risk for transfusion, is what the floor for hemoglobin should be.”
The USRDS analyzed the anemia data by dialysis provider.
“The fall in Hb levels and the rise in transfusions were also important in that there is considerable provider group variation,” Dr. Collins said.
“Providers appear to be adapting to the new payment system and the new FDA ESA label. Most providers may not have been aware of the increasing use of blood transfusions, as these events are not usually reported to the dialysis centers.
“Now that these data are in the public domain, many providers will further adapt their protocols, which should become clear as more data become available.”
All-cause hospital admission rates were similar or lower in 2011 compared with 2010, and cardiovascular hospital admission rates showed a dramatic decrease during this time frame, noted USRDS Deputy Director Robert N. Foley, MB, MSc, during his Kidney Week presentation. The data were taken from CMS Standard Analytical Files (SAF) from 2009–2010 and quarterly files for 2011.
“When you see something unexpected for the first time, you have to start to muddle through the data to see what associations are found,” Dr. Watnick said. “You can't talk about causation until you really get a good prospective, randomized, controlled trial to start testing some of the hypotheses that are generated from the data.”
The reason behind the observed decrease in hospitalizations for cardiovascular disease (CVD) is currently unclear, Dr. Collins agreed.
“The decline in CVD hospitalization rates appears to be an improvement, but these data need to be compared with a full-year data set where all hospitalizations are reported,” he said.
“It is important to realize, hospitals were and are under a performance system assessing their rehospitalization rates, so some of these changes may not be related to the changing payment system.
“Alternatively, there is a clear incentive to keep patients in the dialysis unit and reduce short-stay hospitalizations. From this perspective, this reduces risk to patients.
“It needs to be carefully assessed to make sure there was no rise in outpatient deaths, which the USRDS intends to address as the complete data sets are available.”
Quality and Cost
Separately billable oral medications such as phosphate binders and calcimimetics are to be brought into the bundled payment in 2014.
“The bundle will change incentives to use less of these medications unless there is medical justification based on some outcome,” Dr. Collins said. “Unfortunately, there is little controlled clinical trial data to help guide these choices.
“The potential for reductions in care may need to be monitored through more direct patient consequences, such as fracture rates, versus simple surrogates or process measures.”
Continuing to track the inevitable evolution of nephrology practice will remain essential as clinicians are called to keep both quality and costs in mind.
“If you look at how this country is dealing with health care with an aging population, with a clear knowledge that we can't continue to have health care costs rise as rapidly as they have been over the last many years, you can't help but recognize that we need to think about value, which means not only thinking about quality but also thinking about cost as we move forward,” Dr. Watnick said.
“Initiating the Prospective Payment System is one way that the US government is trying to accomplish that. I think with the ACA [Affordable Care Act] and a lot of those changes that are going to be implemented in 2014 and then 2015, you're going to see a lot of the rest of the country following the suit of how we are dealing with costs in the ESRD care population.
“I think it's going to be successful in reducing costs, but, most importantly, we have to make sure that we're being successful in at least providing the type of quality that we have in the past—hopefully more so. I think the only way to do that is going to be by actually making sure you have very robust data-collection systems.”