The Impact of Bundling on Clinical Practice
Jay B. Wish, MD
The implementation of a bundled payment system (“bundling”) for chronic dialysis in January 2011 was not a surprise and has not been associated with a radical change in clinical practice.
The legislative mandate for bundling was clearly spelled out in the Medicare Improvement for Patients and Providers Act of 2008, and most of the end-stage renal disease (ESRD) stakeholders were familiar with the specific provisions of bundling following the release by the Centers for Medicare and Medicaid Services (CMS) of the draft regulations in 2009 and the final regulations in 2010.
Moreover, their comfort with the bundling regulations led over 95% of providers to “opt-in” 100% to a bundled payment system in 2011, rather than choosing to phase in the bundled payment between 2011 and 2014.
The cost basis for the bundled payment in 2011 was actual dialysis provider costs in 2008. There had already been a secular trend toward lower erythropoiesis-stimulating agent (ESA) use and lower hemoglobin (Hb) targets in dialysis patients between 2008 and 2011 due to the results of randomized clinical trials of ESAs such as CHOIR and CREATE published in 2006, changes in the FDA labeling of ESAs that followed the publication of those studies, and the implementation by CMS in 2007 of a payment penalty for ESAs for dialysis patients with hemoglobin (Hb) levels greater than 13 g/dL for three consecutive months.
Many dialysis providers anticipated even greater ESA cost savings following the implementation of bundling in 2011 through the use of lower Hb targets, greater use of IV iron with higher targets for transferrin saturation (TSAT) and serum ferritin, placing a ceiling on ESA doses in ESA resistant patients, and the administration of ESAs via the subcutaneous (SC) rather than the intravenous (IV) route.
Quality incentive payment
The quality incentive payment (QIP) for 2012, based on provider data from 2010 (which is before bundling actually “kicked in”), rewards facilities for having the maximum percentage of patients with Hb in the 10-12 g/dL range. However, following the publication of the TREAT study in 2009, the FDA changed the ESA labeling in June 2011 and eliminated the target Hb range of 10-12 g/dL for patients receiving ESAs, the range upon which CMS based its QIP for 2012.
The current FDA recommendation for ESRD patients with anemia is to start ESAs when the Hb level is below 10 g/dL and to discontinue or decrease the ESA dose when the Hb level is above 11 g/dL. The FDA has removed the lower limit of the Hb target range for ESRD patients receiving ESA therapy, stating instead that the clinician should individualize therapy to use the lowest ESA dose required to avoid red blood cell transfusions. CMS followed the FDA ruling in July 2011 by eliminating the floor of the Hb target range for the QIP. For 2013 (based on 2011 data), dialysis facilities will be financially penalized for having more patients with Hb levels above 12 g/dL, but will not be penalized for having more patients with Hb levels below any particular value.
The point of the above introduction is that the changes in clinical practice in dialysis that have occurred over the past year of so since the onset of bundling may also be reactions to other events that occurred both before and after January 1, 2011.
Most facilities were aware of the particulars of the bundled payment regulations well before January 1, 2011 and were already testing or implementing new protocols for cost-containment. Medicare claims data for 2011, which will be the most comprehensive description of how clinical practice changed under bundling, will not be available for another 12-18 months and will not be reported in the U.S. Renal Data System (USRDS) Annual Data Report until 2013. Elab, the electronic laboratory data collection system which reports laboratory data on anemia, adequacy, nutrition, and bone/mineral metabolism for most prevalent dialysis patients during the fourth quarter of each calendar year will not complete its 2011 report until mid-2012.
This means that data for 2011 that are currently available are not national and are based on what has been released by individual provider organizations, surveys, and the Dialysis Outcomes and Practice Patterns Study (DOPPS) Practice Monitor.
Decreased monthly ESA use
In one poster presented at the American Society of Nephrology 2011 Renal Week, a group of investigators reported a 29% decrease in mean monthly ESA dose and 48% increase in mean monthly IV iron dose between January-April 2010 and January-April 2011 in two private nonprofit dialysis centers. Most other reports have not been this dramatic, since the large dialysis organizations, which care for around 70% of dialysis patients in the United States, had already been implementing anemia management protocols to increase IV iron use and decrease ESA use during 2010.
In an on-line survey of 303 nephrologists by BioTrends Research Group fielded Dec. 5-12, 2011, about one-quarter stated that their ESA use had not changed since implementation of bundling, but around 50% stated that it had decreased by 10% or more. In addition, 25% of these nephrologists stated that their IV iron used had increased by more than 10% since the onset of bundling, but 40% reported that IV iron use did not change.
The most common change in anemia management reported in this survey was the targeting of lower Hb levels since the onset of bundling, which includes initiating ESAs at lower Hb levels, holding ESAs at lower Hb levels, and down-titrating ESA doses at lower Hb levels. These nephrologists reported an increase in “some hemodialysis patients receiving ESAs via the SC route” from 12% in the fourth quarter of 2010 to 21% in the fourth quarter of 2011.
Thirty-seven percent of the nephrologists reported that they have observed an increase in blood transfusions in the previous six months compared with 13% who made the same observation in the survey fielded during the first quarter of 2011. Furthermore, 84% of respondents in the Dec. 2011 survey attributed the increase in blood transfusions to the FDA labeling change for ESAs and the change in the QIP which eliminated the target Hb floor, not to bundling per se.
|Indicator|| August 2010
| Mean hemoglobin (g/dL)
| Percentage of patients receiving ESAs
| Median weekly epoetin dose (1000 units)
| Upper 90th percentile weekly epoetin dose (1000 units)
| Percent of patients receiving IV iron in past month
| Median monthly IV iron dose prescribed (mg)
| Upper 90th percentile monthly IV iron dose (mg)
| Median serum ferritin (ng/mL)
| Upper 90th percentile serum ferritin (ng/mL)
| Median PTH (pg/mL)
| IV vitamin D use (percent of patients)
| Oral active vitamin D use (percent of patients)
| Cinacalcet use (percent of patients)
| 30-day annualized hospitalization rate
*ESAs = erythropoiesis stimulating agents; IV = intravenous; PTH = parathyroid hormone
Changes in Hb levels
DOPPS reports data on a stratified random sample of 145 U.S. dialysis facilities that are representative of a national sample by facility type (dialysis organization size, rural/urban, free standing/hospital based). The data are collected monthly and the DOPPS Practice Monitor (www.dopps.org/dpm), a website reporting data trends over the prior 12 months, is updated quarterly. The key findings from the DOPPS Practice Monitor are summarized on Table 1.
The most recent data are through August 2011 and show a decrease in the mean Hb only from 11.5 to 11.4 g/dL between August 2010 and July 2011, but a decrease to 11.2 g/dL in August 2011 following the change in the FDA product information on ESAs and the discontinuation of the target Hb floor for the QIP by CMS.
In addition, the mean Hb level fell from 11.2 to 10.9 g/dL between June and August 2011 among dialysis providers with 10 or more facilities, but only from 11.3 to 11.2 g/dL during the same period among dialysis providers with fewer than 10 facilities. This suggests that the larger dialysis organizations are more nimble in adapting to regulatory changes. There was a 15% decrease in the median dose of ESA prescribed among all the providers in the DOPPS sample between August 2011 and 2011, but the 90th percentile of ESA dose decreased by 23% during this period. It can be inferred that a ceiling on ESA dosing among patients with “ESA resistance” is driving much of the decrease in the mean dose.