Currently Viewing:
The American Journal of Managed Care June 2016
Development of a Tethered Personal Health Record Framework for Early End-of-Life Discussions
Seuli Bose-Brill, MD; Matthew Kretovics, MPH; Taylor Ballenger, BS; Gabriella Modan, PhD; Albert Lai, PhD; Lindsay Belanger, MPH; Stephen Koesters, MD; Taylor Pressler-Vydra, MS; and Celia Wills, PhD,
The Value of Decreasing Health Cost Volatility
Marc Herant, PhD, MD, and Alex J. Brown, MEng, MBA
Variations in Patient Response to Tiered Physician Networks
Anna D. Sinaiko, PhD
Primary Care Appointment Availability and Nonphysician Providers One Year After Medicaid Expansion
Renuka Tipirneni, MD, MSc; Karin V. Rhodes, MD, MS; Rodney A. Hayward, MD; Richard L. Lichtenstein, PhD; HwaJung Choi, PhD; Elyse N. Reamer, BS; and Matthew M. Davis, MD, MAPP
Impact of Type 2 Diabetes Medication Cost Sharing on Patient Outcomes and Health Plan Costs
Julia Thornton Snider, PhD; Seth Seabury, PhD; Janice Lopez, PharmD, MPH; Scott McKenzie, MD; Yanyu Wu, PhD; and Dana P. Goldman, PhD
Risk Contracting and Operational Capabilities in Large Medical Groups During National Healthcare Reform
Robert. E. Mechanic, MBA, and Darren Zinner, PhD
The Evolving Role of Subspecialties in Population Health Management and New Healthcare Delivery Models
Dhruv Khullar, MD, MPP; Sandhya K. Rao, MD; Sreekanth K. Chaguturu, MD; and Rahul Rajkumar, MD, JD
When Doctors Go to Business School: Career Choices of Physician-MBAs
Damir Ljuboja, BS, BA; Brian W. Powers, AB; Benjamin Robbins, MD, MBA; Robert Huckman, PhD; Krishna Yeshwant, MD, MBA; and Sachin H. Jain, MD, MBA
Review of Outcomes Associated With Restricted Access to Atypical Antipsychotics
Krithika Rajagopalan, PhD; Mariam Hassan, PhD; Kimberly Boswell, MD; Evelyn Sarnes, PharmD, MPH; Kellie Meyer, PharmD, MPH; and Fred Grossman, MD, PhD
Value of Improved Lipid Control in Patients at High Risk for Adverse Cardiac Events
Anupam B. Jena, MD, PhD; Daniel M. Blumenthal, MD, MBA; Warren Stevens, PhD; Jacquelyn W. Chou, MPP, MPL; Thanh G.N. Ton, PhD; and Dana P. Goldman, PhD
Currently Reading
Effects of Physician Payment Reform on Provision of Home Dialysis
Kevin F. Erickson, MD, MS; Wolfgang C. Winkelmayer, MD, ScD; Glenn M. Chertow, MD, MPH; and Jay Bhattacharya, MD, PhD

Effects of Physician Payment Reform on Provision of Home Dialysis

Kevin F. Erickson, MD, MS; Wolfgang C. Winkelmayer, MD, ScD; Glenn M. Chertow, MD, MPH; and Jay Bhattacharya, MD, PhD
The Medicare programís transition in 2004 to tiered fee-for-service physician reimbursement for dialysis care had the unintended consequence of reducing use of home dialysis.
ABSTRACT

Objectives:
Patients with end-stage renal disease can receive dialysis at home or in-center. In 2004, CMS reformed physician payment for in-center hemodialysis care from a capitated to a tiered fee-for-service model, augmenting physician payment for frequent in-center visits. We evaluated whether payment reform influenced dialysis modality assignment.

Study Design: Cohort study of patients starting dialysis in the United States in the 3 years before and the 3 years after payment reform.

Methods: We conducted difference-in-difference analyses comparing patients with traditional Medicare coverage (who were affected by the policy) to others with Medicare Advantage (who were unaffected by the policy). We also examined whether the policy had a more pronounced influence on dialysis modality assignment in areas with lower costs of traveling to dialysis facilities.

Results: Patients with traditional Medicare coverage experienced a 0.7% (95% CI, 0.2%-1.1%; P = .003) reduction in the absolute probability of home dialysis use following payment reform compared with patients with Medicare Advantage. Patients living in areas with larger dialysis facilities (where payment reform made in-center hemodialysis comparatively more lucrative for physicians) experienced a 0.9% (95% CI, 0.5%-1.4%; P <.001) reduction in home dialysis use following payment reform compared with patients living in areas with smaller facilities (where payment reform made in-center hemodialysis comparatively less lucrative for physicians).

Conclusions: The transition from a capitated to a tiered fee-for-service payment model for in-center hemodialysis care resulted in fewer patients receiving home dialysis. This area of policy failure highlights the importance of considering unintended consequences of future physician payment reform efforts.

Am J Manag Care. 2016;22(6):e215-e223
Take-Away Points
 
In 2004, CMS reformed physician payment for in-center hemodialysis care from a capitated to a tiered fee-for-service model, augmenting physician payment for frequent in-center visits. This policy may have influenced home dialysis use by making in-center dialysis more lucrative for some physicians. We compared home dialysis use among patients differentially affected by the policy. 
  • Patients most affected by the policy experienced nearly a 1% reduction in the absolute probability of home dialysis use following payment reform. 
  • Our findings indicate that transition to fee-for-service payment for in-center hemodialysis had the unintended consequence of reducing home dialysis use.
Pay-for-performance (P4P) initiatives tying payment to performance and the value of care have become a major component of recent healthcare reform efforts. Since the passage of the Affordable Care Act and, more recently, the repeal of Medicare’s Sustainable Growth Rate, P4P programs are increasingly targeting physician practices directly.1,2 Lessons from prior P4P initiatives can help inform the development of future policies that will apply to both managed care and fee-for-service settings.

More than 100,000 individuals develop end-stage renal disease (ESRD) every year in the United States.3 Due to a shortage of kidneys available for transplantation, the vast majority receive dialysis, which can be provided through 1 of 3 modalities. In-center hemodialysis is the most common dialysis modality and involves patients going to a dialysis facility 3 or 4 times per week to receive therapy; home-based dialysis therapies (which include peritoneal dialysis and home hemodialysis) are alternatives that offer more flexibility and lifestyle benefits for some patients.4-8 Ideally, dialysis modality is chosen after careful consideration of medical suitability, followed by shared decision making among patients, loved ones, and care providers.9 Evidence suggests that these discussions occur infrequently,10 leading many to conclude that home dialysis therapies are underutilized in the United States.11,12

It is uncertain whether physicians’ economic incentives influence dialysis modality choice. International comparisons indicate that the relative physician payment for patients on home versus in-center dialysis directly influences the fraction of patients on home dialysis.13 In the United States, higher Medicare payment to dialysis facilities for home therapies associated with the 2011 ESRD Prospective Payment System (“bundling”) coincided with a substantial increase in the use of peritoneal dialysis.3,14 However, surveys of nephrologists suggest that patient preferences and health, rather than economic factors, are the primary factors considered when recommending a dialysis modality.11,15

In 2004, in an effort to align economic incentives and encourage high-quality care, CMS transformed its payment to physicians caring for patients receiving in-center hemodialysis from a capitated model to a tiered fee-for-service model (eAppendix Table 1 [eAppendices are available at www.ajmc.com]).16 Under the new payment system, which continues to govern physician in-center hemodialysis reimbursement, physicians could increase professional fee revenues by conducting 4 or more visits per month to patients receiving in-center hemodialysis.

Although this policy was not focused on the delivery of home dialysis care, it may have influenced dialysis modality decisions by making in-center hemodialysis comparatively more lucrative for some physicians; physician payment for home dialysis therapy remained capitated and decreased slightly.16 In this study, we determined whether the transition to a tiered fee-for-service payment model influenced dialysis modality choices. We hypothesized that patients were less likely to receive home dialysis following payment reform, and that this decrease was more pronounced in places where physicians could increase in-center hemodialysis revenues at lower cost.

 

METHODS
Data and Patient Selection

We selected patients who started dialysis in the United States from January 1, 2001, through December 31, 2006—the 3 years prior to and the 3 years following physician payment reform. We excluded patients who received a kidney transplant within 60 days of ESRD onset. We obtained data on patients’ insurance coverage, home zip codes, and initial dialysis modality, as well as information about dialysis facilities from the United States Renal Data System, a national registry of patients with treated ESRD. We obtained data on patient comorbidities prior to ESRD from the CMS Medical Evidence Report (CMS-2728).17 Due to large numbers of missing values for Quételet’s (body mass) index, hemoglobin, and albumin, we used multiple imputations to estimate missing values.18-20 Information on population density came from Census-based rural-urban commuting area codes.21 Information on hospital referral region (HRR) came from the Dartmouth Atlas of Health Care.22

Outcomes and Study Design

The primary study outcome was the initial dialysis modality chosen, as reported by the nephrologist to CMS. We categorized dialysis modality as in-center hemodialysis or home dialysis, where home dialysis included home hemodialysis or peritoneal dialysis.

We used several difference-in-difference (DID) models to examine the effect of payment reform on dialysis modality. DID analysis is an econometric method commonly used to analyze policy,23 where patients are separated into treatment and control groups. The treatment group includes patients who were affected by the policy of interest and the control group includes patients who were not subject to the policy. Thus, any changes observed in the control group reflect changes in the population from measures not changed by the policy. The difference in the change of the outcome after implementation of the policy between the treatment and control groups characterizes the policy’s effect.

 

Comparison Groups


We formed comparison groups from 2 separate cohorts. In an Insurance Coverage cohort, we selected patients enrolled in either traditional Medicare as a primary payer or Medicare Advantage prior to start of dialysis. In this analysis, we only included patients 65 years or older at ESRD onset because patients are not permitted to enroll in Medicare Advantage if ESRD (rather than age) is their qualifying criterion; thus, most patients with ESRD with Medicare Advantage are 65 years or older. We conducted a DID analysis comparing the choice of dialysis modality among patients with traditional Medicare versus Medicare Advantage. We chose these groups because payment for services provided to patients with traditional Medicare was affected by payment reform, while payment for services provided to patients with Medicare Advantage was not.

In a “non–health maintenance organization (non-HMO) Medicare” cohort, we selected patients with traditional Medicare as a primary payer, or waiting for Medicare coverage, at the onset of dialysis. Because the majority of patients in the United States who develop ESRD qualify for Medicare within 90 days of ESRD onset, we assumed that patients documented as “waiting” for Medicare would soon receive it and that physicians would consider the financial implications of treating these patients as similar to treating patients already covered. In this cohort, we excluded patients with private insurance because they do not qualify for Medicare until 30 months have passed following the diagnosis of ESRD.

We previously demonstrated that the frequency of physician (or advanced practice provider) visits to patients receiving in-center hemodialysis was predominantly related to geographic and dialysis facility factors, rather than to patient clinical characteristics.24 Geographic measures—such as dialysis facility size and population density—that determine the costs physicians incur (in resources and time) traveling to visit patients at dialysis facilities have a substantial influence on visit frequency. All else being equal, it is more lucrative for physicians to see patients in larger dialysis facilities because physicians can collect revenue for more patient visits after incurring a fixed cost of traveling to a facility. Likewise, it is more lucrative for physicians to see patients in more densely populated areas due to lower travel costs to facilities.

Using the non-HMO Medicare cohort, we conducted 2 DID analyses to determine whether changes in the choice of dialysis modality following payment reform varied geographically, depending on how costly it was for physicians to see patients more frequently. Although the small decrease in physician payment for home dialysis was similar across all geographic regions, the change in physician payment for in-center hemodialysis after 2004 varied geographically. Physicians practicing in areas where the cost of more frequent visits was lower had an opportunity to increase their professional fee revenues after payment reform by assigning more patients to in-center hemodialysis. In contrast, physicians practicing in areas where it was too costly to visit patients 4 times per month would have experienced little or no increase in professional fee revenues by assigning patients to in-center hemodialysis. We used the 2 geographic characteristics previously found to be associated with visit frequency and, therefore, the relative gain in professional fee revenue from in-center hemodialysis—dialysis facility size and population density—to determine if changes in physician payment influenced dialysis modality choice.

We averaged dialysis facility size across the HRRs where patients lived. We calculated dialysis facility size from the average number of patients receiving in-center hemodialysis documented in annual facility surveys in the 3 years prior to payment reform. We divided HRRs into quintiles based on their average facility size and assessed the proportion of prevalent in-center patients seen 4 or more times per month, as well as associated changes in revenues, in the 3 years following payment reform within each quintile. We observed that the proportion of patients with 4 or more visits per month was smallest in the lowest mean facility size quintile. Consequently, we categorized HRRs in the lowest quintile of mean facility size as areas with “smaller facilities.”

We dichotomized population density into “small town/rural” and “non–small town/non-rural.” The differences in visit frequency across population density category were small relative to differences across dialysis facility size (eAppendix Table 2).

 

Statistical Methods


Due to large population size, we used a 10% standardized mean difference as a marker of heterogeneity when comparing differences in characteristics among treatment groups.25 In all DID analyses, we used logistic regression to estimate odds ratios and corresponding 95% CIs. We controlled for regional differences in population density and dialysis facility size, as well as in patient age, sex, race, ethnicity, and medical comorbidities (Table 1).26 We did not adjust for dialysis facility characteristics because the facility where a patient receives dialysis is often a consequence of dialysis modality choice. An interaction term between binary variables representing the start of dialysis before versus after payment reform, and whether patients were in the treatment or control group, estimated the effect of the policy on the odds of dialysis modality choice for each comparison.

 
Copyright AJMC 2006-2017 Clinical Care Targeted Communications Group, LLC. All Rights Reserved.
x
Welcome the the new and improved AJMC.com, the premier managed market network. Tell us about yourself so that we can serve you better.
Sign Up
×

Sign In

Not a member? Sign up now!