Since it started bundling payments for cardiac surgery, three-hospital Geisinger reports that all of its patients are receiving “best care,” based on 40 best-practice steps derived from the American Heart Association and the American College of Cardiology guidelines for cardiac surgery.View PDF of chart on Geisinger Health System's bundled-payment model
As a payer there's an unpredictability of how much something is going to cost, says Alfred Casale, associate chief medical officer and chairman of cardiothoracic surgery at 172-bed Geisinger Wyoming Valley Medical Center, Wilkes-Barre, Pa. “You wind up with huge variation in payment” that frustrates payers, caregivers and patients alike, he says.
In establishing bundled payments, Geisinger was trying to introduce some predictability for a specific episode of care, in addition to removing variation among these patients in terms of resource usage. In addition, it was transferring the financial risk of complications and readmissions from the payer and handing that responsibility to the providers to affect patient outcomes.
In essence, “You're throwing the ball into the court of the clinicians and hospitals, and they decide how you divvy up the goodies providing care,” Casale says.
Meanwhile, 490-bed Hillcrest Medical Center, Tulsa, Okla., part of Ardent Health Services, reports a similar success story with bundling.
One of the five sites participating in the CMS' Acute Care Episode demonstration project, Hillcrest estimates it has saved nearly $730,000 since it began implementing the ACE demonstration in May 2009 (See related story, April 6, 2009, p. 32)
Bundling is a “win, win, win” for the hospital, doctor and patient, says Steve Landgarten, CMO of eight-hospital Ardent, headquartered in Nashville.
It allowed Hillcrest “as a community hospital to address patients' and physicians' financial needs, so doctors could work with us to maintain high-quality outcomes and reduce the cost of those outcomes. Everybody gets a piece of reduced costs to continue high-quality outcomes.”
The concept of “bundling,” which essentially means paying providers a fixed amount per month or year for all covered services, has specifically been tagged by the Obama administration as a cost-saving measure. Last year, the new administration estimated that bundling Medicare payments covering hospitals and post-acute care could save as much as $17 billion over the next 10 years.
In addition to the CMS' ACE demonstration, the federal government plans to do more testing in the future, possibly through a national, voluntary, five-year pilot on bundling payments that was authorized by the new health law and would get under way in 2013.
A number of policy developments have gone into the promotion of bundling payments for episodes of care, says Peter Hussey, a policy researcher at RAND Corp.
“There's bipartisan agreement into this idea that fee-for-service is a major problem in terms of healthcare costs,” Hussey says. “Medicare currently pays hospital care by admission, but there are strong advocates who say the federal insurance program needs to think even more broadly” when designing payments for episodes of care. The federal government and the private sector are trying to get in front of this, and enthusiasm among the provider groups and employers is building to pay by value rather than volume, Hussey says.
At least in the private sector, the train driving payment bundling seems to be moving along at a steady clip.
In addition to Geisinger, other private-sector organizations, including the Brookings Institution and the Integrated Healthcare Association in California, are testing variations of the payment model. At the state level, Minnesota has a 2008 law implementing “baskets of care.” Early this year the state began bundling payments based on episodes of care, Hussey says.
“The private sector is right not to wait around” for the federal government to take the lead on bundling, says Deirdre Baggot, administrator for cardiac and vascular services at 361-bed Exempla St. Joseph Hospital in Denver. “They want to explore this as a viable payment strategy.”
Two-hospital Exempla Healthcare is another hospital system participating in the three-year ACE demonstration. The others in addition to Hillcrest are 1,275-bed Baptist Health System, with four campuses in San Antonio; 78-bed Oklahoma Heart Hospital in Oklahoma City; and Lovelace Health System, which has four hospitals in Albuquerque.
Bundling has progressed more slowly at the federal level because of a series of snags, even though the concept has been on HHS' agenda for decades.
Medicare in the 1990s conducted a successful demonstration on payment bundling for heart bypass procedures, but various explanations have been given for why the project hasn't become an actual payment model. “What we were told by the CMS is the project was very successful, but that they lost focus” trying to get ready for any Y2K problems, Baggot says.
Hussey offers a more political reason for the demonstration losing its momentum.
Hospitals and doctors at the time were strongly resisting the idea, especially physicians, since they were being rewarded based on volume, and hospitals were trying to decrease the cost per admission, Hussey says.
Bundling may have gained more support from the provider community since that experiment with heart bypass surgery, but doctors continue to wax pessimistic about it, he says.
“Most physicians don't want to give any more autonomy to the hospitals. They hate the idea of going to the hospital for their payments.” For that reason, negotiations between hospitals and physicians are going to be a major obstacle in advancing the concept of bundled payments, Hussey says.
Not necessarily, according to proponents of bundling, including those already seeing some success with the concept.
The ACE demonstration requires that all participating hospitals form a physician-hospital organization, or have one under development. The bundled payment is based on a computation of historical Medicare Part A and Part B payment data with a “discount,” or cut in reimbursement, off what the hospital would anticipate receiving as a traditional Medicare payment.
Hospitals are responsible for divvying up the payment through a voluntary agreement reached by the hospitals and physicians—with the goal of sharing savings achieved through improved quality and increased efficiencies, says Cynthia Mason, a project manager with the CMS' Medicare Demonstrations Group.
Ardent's Landgarten says this alliance between hospitals and physicians in the ACE demonstration has led to “gain-sharing payments,” where the hospital shares any savings achieved through the bundled payment with physicians.
Much of the savings accrued from the demonstration have come from “doctors coming to our side of the table and negotiating with devicemakers,” Landgarten adds. Instead of vendors lobbying physicians independently about device purchases, suddenly the hospitals and doctors are negotiating together on supply costs. Secondly, hospitals and doctors are working on operational efficiencies, “creating economic incentives to improve efficiency and reduce costs.”
Upfront, the hospital commits to a lower payment from Medicare, while the physicians get a full Medicare payment, plus as much as a 25% increase if they meet certain thresholds on compliance with quality metrics set by the hospital and approved by the CMS.
In addition, patients also share in the savings, Landgarten says. “They also have the possibility of getting a rebate of almost $1,000 if these outcomes are successful.” Patient incentives, paid directly to the patient by Medicare, to date have amounted to $170,377, according to Hillcrest. The end result is the hospital benefits from operational efficiencies and improved quality of care, and it attracts more physicians and patients.
At least at Exempla, the doctors don't feel as if the hospital is running the show, says Lisa Kettering, vice president of medical affairs and CMO at Exempla St. Joseph.
The physicians aren't wary of the bundling project, Kettering says; if anything “we've been very much involved in the leadership (of this project) in driving both quality and cost efficiencies.”
As with Hillcrest, physicians involved in the project will get 100% Medicare reimbursement and share in the cost savings if certain quality metrics are achieved.
Exempla has yet to start the bundling demonstration because Medicare contractor issues held things up for the hospital system, but the expected start date is in October, Kettering says.
Geisinger, likewise, was able to get around physician-hospital tensions because “the structure for unification was already in place” before the bundling effort was implemented.
The health system's staff-employed physicians have chosen to work as part of an integrated system, which runs under one governance structure “with all of the oars running in the same direction,” Casale says. For these reasons, bundling and establishing an electronic health record was able to work “because we were able to leverage our connectivity on best practices,” he says.
To date the ACE demonstration has focused primarily on acute-care hospitalization, but plans are in the works to expand the episode of care to include post-acute services. Other geographic areas “have always been under discussion,” the CMS' Mason says.
“Right now we're focusing on getting those five sites implemented for acute-care hospitalization,” Mason says. It's little too early to speculate which direction the ACE demonstration will proceed, she says.
RAND's Hussey expects the pilot outlined in the health reform law will end up being a “straightforward extension of the ACE demo,” but will have more flexibility than ACE based on the mere fact that it's a pilot, not a demonstration project.
Demonstration projects are limited in scope, Hussey says. They involve some sort of experiment where experts look at outcomes and conclusions, and then congressional action is required to put the demonstration into place, he said.
“The idea with a pilot is HHS has the discretion to expand it if it looks like it's working, so it's easier to take it to scale,” Hussey says.
It certainly signals that the pilot will extend bundling of payments into post-acute care, Hussey says. But whether bundling will be successful on such a broad scale—expansion beyond acute care—is still unclear, says Hussey, who explored these questions last year in a RAND study.
Once you involve post-acute care it gets even more complicated because you're involving more providers, Hussey says. Bundling is a simple process if there's only one hospital or doctor group participating, “but if there are multiple groups—a hospital, a post-acute-care facility or other ambulatory follow-up care, technically it gets more complex,” he says.
In analyzing claims information from a large group of Medicare beneficiaries, he and other researchers from RAND identified a number of issues that should be considered to determine how to identify episodes of care, and which provider is accountable for an episode.
Defining what provider is responsible for managing treatment for different conditions is a complicated task, because patients are frequently treated by a wide variety of providers and in numerous settings for different problems, the researchers found.
As an example, many of the patients studied had multiple chronic conditions such as high blood pressure and elevated cholesterol levels. “Encouraging a single-condition focus through an episode-based payment plan may not be optimal for these patients,” according to the study.
Landgarten concedes that attempting to bundle payments is more challenging for some conditions than others. Packaging payments for medical codes such as diabetes, pneumonia and congestive heart failure is more difficult than for a surgical procedure such as gall bladder surgery “because medical codes are far more diverse.”
Other challenges await for bundling beyond the acute-care setting, Kettering acknowledges.
“Providers with different tax IDs and settings organized around a medical condition” could make it difficult to establish a bundled payment in post-acute care. In addition, current health IT systems “are not set up to manage bundled payments, so those systems will need to be evaluated,” resulting in some capital investments to accommodate this new payment model, Kettering says.
Geisinger already has begun testing a post-acute “chronic care” bundled model, specifically in the areas of diabetes and adult preventive care, although its financial arrangement is somewhat different from the health system's work in the acute-care arena.
Specifically, the major difference between chronic- and acute-care bundling is the approach. In areas such as coronary artery disease, congestive heart failure, chronic kidney disease and diabetes, patients are managed through the entire continuum of the disease—in the hospital, in physicians' offices, at home or in other settings, Casale says.
For the diabetes management model, diabetes-care and primary-care doctors must meet a “bundle” of eight or nine elements of care for diabetic patients. In return health plans pay an increment to the physician based on how well they perform on these measures. In return, the diabetic patients get the full scope of care they need, Casale adds.
Geisinger in the meantime is interested in playing a role in the health law's proposed pilot project on bundling, he says.
“We certainly aspire to be part of those projects. I'm not saying (our model) can be taken out of the box and implemented in other places, but the concepts and lessons we learned could be applied more broadly,” Casale says.