Opportunities in the New Bundled Payment Initiative

The following is a white paper reprinted with permission from Singletrack Analytics.

On Aug. 22, CMS release the details of the "Bundled Payments for Care Improvement Initiative" program, which is the lesser known of the two pilot programs implemented by the Patient Protection and Affordable Care Act health reform legislation. (The other pilot program, of course, is accountable care organizations.) The requirements for participation in this program, which were issued as a "request for application" rather than draft regulations, are remarkably clear, generally unambiguous and provide wide flexibility to providers interested in participating in this project. This is in contrast to the ACO draft regulations, which some observers believe are so complex, confusing and onerous that few participants are expected in that program. The application specifies four general models under which hospitals can participate, but within each category providers are allowed significant flexibility to propose a structure that will work well for them while meeting CMS’s goals. Many of the initial concerns raised about this program after passage of PPACA (including those of this author [1]) were alleviated by the level of flexibility allowed to participants.

Participation Models

The initiative provides for four different models in which an organization can participate. These are:

•    Model 1 — Retrospective payment for acute hospital stay only
•    Model 2 — Retrospective payment for physician and hospital services during an acute stay, including post-acute services
•    Model 3 — Retrospective payment for post-acute services, not including the acute episode, and
•    Model 4 — Prospective payment for physician and hospital services for the acute stay.

A summary of the participation model characteristic is shown below.


Patients and services covered
Several characteristics are common to all models. First, each model is defined by the characteristics of patients who will be participating. In Model 1, all inpatients must participate, while in the remaining models the proposing organization can specify selected MS-DRG that will be included, while patients in other MS-DRGs will not be included. 

Models are also defined by the services that are included in the bundled payment. In Model 1, only hospital party payments are included, while the other models include physician payments and, for Models 2 and 3, payments for post-acute services.

All patients involved in covered episodes at a participating hospital must be included in the bundled payment. There is no opportunity for a patient to "opt out," or for the hospital or physician to exclude a particular patient in a bundled MS-DRG. This prevents "cream-skimming" that would occur if potentially high-cost patients could be excluded from the bundled payment. 

Discounts and provider payments
Applicants for the program must provide CMS with a discount from the average aggregate payment amounts that would otherwise be paid for these patients on a fee-for-service basis. The amount of the discount is left to the applicant but obviously will be considered by CMS in selection of participants for the program. This discounted amount will become the fixed bundled payment that will be reconciled with FFS payments paid to providers under Models 1-3 and the amount paid to the contracting entity under Model 4.

The discount given is the "savings" that CMS recognizes from the initiative. There’s no further "shared savings" as occurs in ACOs and other similar programs. Instead, the providers' compensation is the fixed bundled payment amount, regardless of what FFS payments would be for each patient. If the bundled payment exceeds the FFS amount, the providers retain the difference. If the FFS amount would exceed the bundled amount the providers lose the difference. This reduces the incentive for providers to provide unnecessary services [2] and may allow providers to reduce their costs without reducing their revenue.

In previous demonstration projects such as the Acute Care Episodes project, the bundled payment was made to the contracting organization, which was required to have the infrastructure to distribute payments to each of the providers who provided services to the covered patients. This required these organizations to implement a significant claims payment infrastructure, which was burdensome to many providers, particularly to those without a significant proportion of employed physicians. 

This requirement has been eliminated for Models 1-3, in which all providers will continue to be paid by CMS on a FFS basis. Periodic reconciliations will be performed to determine whether those payments were higher or lower than the bundled payment amount, and adjustment payments will be made by the organization or CMS as required to make the aggregate payments to the contracting organization equal the bundled amount. In Model 4 the bundled payment will be made to the organization, which is responsible for distributing those payments to the providers, who will not be paid by CMS for services to the included patients. CMS notes in several places in the application document that the contracting organization will be responsible for any payments in excess of the bundled amount, even if those payments were made to an unrelated provider (for example a non-participating physician or a readmission to a different hospital).


In general, gainsharing payments to physicians who have created cost savings are allowed, since CMS is waiving several related regulations that cover hospital payments to physicians. Participation of physicians in gainsharing programs must be voluntary, and the organization must document the involvement of physicians in these cost saving processes as well as the payment methodology and the ways in which gainsharing will support care redesign to improve quality and create cost savings. Physicians must meet quality requirements before gain sharing payments can be made.  Gainsharing payments to physicians cannot exceed 50 percent of the FFS-equivalent payments that they would have received absent the bundled payment program, which is an increase from the 25 percent maximum that was used in the ACE demonstration. CMS apparently believes that greater sharing amounts are appropriate if the physicians significantly participate in the savings.

Quality measures

Maintenance and improvement of quality is a key characteristic of this program, and organizations are required to submit various quality metrics. Participating hospitals must have received their full IPPS and OPPS annual payment updates for reporting quality measures since at least FY 2009, and both hospitals and physicians must continue to participate in all quality reporting initiatives.  However, applicants are encouraged to propose new quality metrics in a wide variety of areas.  

Participation in multiple CMS initiatives
The PPACA law specifies that organizations cannot participate in multiple "shared savings" programs of CMS. Therefore, there was some question as to whether a health system that was forming an accountable care organization could participate in the bundled payment pilot. CMS clarifies this issue in the application by stating that the bundled payment initiative program does not constitute a shared savings program and that organizations who form ACOs can still participate.

Episode Definition Models

The initiative defines the following four participation models:

Model 1
This model is designed primarily for hospitals to create internal cost savings that can be shared to physicians as part of gain sharing payments. It includes all inpatient admissions regardless of MS-DRGs but does not include physician services or post-acute.  

Model 2
This is the model that was prescribed in the PPACA law. It includes the hospital and physician services and also acute and post-acute care. Applicants have two options to specify the post-acute episode; under the first option the episode it extends 30 to 89 days, while under the second option it extends 90 days or longer after the hospital discharge. Organizations selecting the second option are given flexibility to propose its definition. CMS’s objective from this option is to explore ways in which patients can be transitioned back into the community, presumably from better discharge planning and postoperative care.

These rules clarify a major question that arose from the PPACA law by allowing the organization to select the MS-DRG to be included. Since the law was nonspecific, many readers were concerned that an organization would be required to participate in a specified list of MS-DRGs, including those for which their organizations not well-equipped to manage on a bundled basis. By allowing the organization to specify the MS-DRGs to be included, they can play to their strengths and include areas in which their physician relationships are strong, and in which significant cost savings are possible.

In addition, CMS is allowing organizations to propose further definitions of the episode; for example to exclude costs associated with unrelated readmissions.  However, cost of readmissions, including physician services, must be included in the bundled payment.

Model 3
This model includes post-acute services only, with services provided at a SNF, inpatient rehab facility, LTAC or HHA for at least 30 days following an inpatient acute discharge. It also includes Part A and Part B payments for related readmissions. CMS will give preference to organizations proposing a longer period because of their interest in redesigning care to facilitate a patient’s transition back into the community.

Model 4
This model is generally a continuation of the Acute Care Episodes demonstration project with several exceptions. First, it allows the contracting organization to specify the MS-DRGs to be included (the ACE demo only included certain cardiac and orthopedic DRGs). It also will not limit participants to a specific demographic area (served by one Medicare Administrative Contractor) as the ACE demo did. Finally, CMS has eliminated the beneficiary cost savings component in which beneficiaries who utilized ACE-participating hospitals were given a credit on their Part B premiums. This feature was intended to create steerage to the ACE hospitals, but was apparently too burdensome to administer and was also confusing to the beneficiaries.

Similar to the ACE demo but different from the other models, the bundled payment will be paid directly to the contracting organization rather than to the individual providers. The contracting organization will then pay the providers based on whatever payment methodology has been developed. This avoids the need for periodic reconciliations with CMS but requires the organization to implement some form of claims payment infrastructure. This model may be attractive to current ACE participants who wish to continue with their existing infrastructure but does not appear to be as attractive to other participants who don’t have this type of infrastructure in place.

The Bundled Payment Business Model

The effects of a bundled payment model are significantly different for those providers inside and outside of the contracting group. Those inside of the group have agreed to accept a fixed payment for a specific episode of care. Their compensation will be unrelated to the services that they provide to the patients. While they must maintain quality, they can reduce or eliminate services that are not necessary to proper patient care without any commensurate reduction in revenue as would occur under a FFS payment system. 

Therefore, one of the incentives under a bundled payment model is for providers who are part of the contracting group to reduce their service level to those required for proper patient care. This will generally lower the providers’ internal costs and/or make their time and resources available for other patients. This concept also extends to reducing readmissions and thereby the associated hospital and physician costs of treating those patients.

Another incentive is to reduce the utilization of providers outside of the contracting group. For example, in the CMS Heart Bypass demonstration project cardiac surgeons reduced the number of cardiology consults by performing those services themselves. This commensurately reduced FFS payments to the cardiologists, which left more of the bundled payment to be distributed among the providers in the contracting organization.  (The cardiologists were understandably not pleased with this result.)

The third incentive is for hospitals and physicians to cooperate to reduce internal hospital costs (supplies, drugs, etc.) because the gainsharing features of the initiative allow those savings to be shared with the physicians who help create them. This process was quite successful for some hospitals in the ACE demonstration and resulted in significant savings for the hospitals, which were then shared with the physicians who helped create them.

Counterbalancing these positive incentives are two negative factors for participants. First, CMS requires contracting organizations to discount the bundled payment amount from the historical payments made for the proposed episodes. This means that the contracting providers must achieve cost reductions of several percent to break even. In addition, the contracting organization and its providers won’t get paid more for complex cases, as they would under FFS payment. Not only will the contracting providers be a fixed amount to provide more services, but they will also be liable for any services rendered by providers outside of the contracting organization who will continue to be paid on a FFS basis.  Therefore, the contracting organization’s providers can have some significant risk exposure to high-cost cases, including outliers and readmissions.

This business model needs to be clearly understood by all participants since it’s significantly different from the FFS models under which physicians generally operate. It creates additional risks but also the opportunity for significant rewards, for the participating physicians.

Episode and pricing analytics

Critical to the success of these organizations will be their ability to select, define and price the episodes of care that they propose. If carefully defined, the historical FFS provider payments for an episode should reasonably predictable and stable, affected by factors that are identifiable and and controllable, and provide an opportunity for cost reductions and quality improvements. Depending on the model, these payments can include physician payments, outlier payments to the hospital, payments to hospitals and physicians for related readmissions, post-acute providers, durable medical equipment and others.

To facilitate these analyses, CMS will allow organizations to request a "limited dataset" containing the relevant claims for the proposed model for 2008 and 2009. Organizations must complete a "Data Use Agreement" to allow this information to be released to them. The data will be supplied as up to seven multiple “Standard Analytical Files” containing multiple fields (the claims data contains 180 data fields) that must be merged together to consolidate all relevant claims. Once that database is built, episodes of care must then be constructed according to the definitions to be used by the proposing organization. The organization can then analyze this data to develop the episodes and pricing.

Hospital cost accounting data will also play a significant role in designing the proposal. Understanding the fixed and variable costs in each DRG will be important in determining the level of discount to propose for those DRGs. In addition, the components of all cost elements in each DRG form the basis of identifying opportunities for cost reductions in that DRG.

The analytics team
The team tasked with developing the CMS proposal and implementing the initiative if accepted by CMS, will require a wide range of skills in finance, clinical operations, medical decision-making, along with the analytical skills necessary to translate the information needs of the group into discrete analyses and reports. Physicians who specify patient treatment, use of ancillary services, consultations ordered and other similar cost-related decisions must work alongside the clinical managers of their departments who are knowledgeable about supply costs, staffing and department logistics. Finance staff members can work with cost accounting data to identify high-cost medical supplies or other cost drivers, particularly when their use varies based on the physician who orders them. The opportunity for gainsharing provides a unique framework for a multidisciplinary team to attack a common problem and create win-win solutions.

Opportunities to reduce hospital costs
This is the most attractive opportunity for several reasons. First, all of the parties benefit (other than the hospital’s vendors) from cost reductions. Hospitals benefit directly, and physicians who create the savings can benefit through gainsharing. These savings also do not create reductions in payments to providers. To identify the possibilities of creating these savings, the analytics team should analyze the cost components of each DRG under consideration from the hospital cost accounting data. For example, significant cost savings have been recognized in ACE demonstration hospitals when orthopedists agreed to reduce the number of hip prostheses from 7 to 2. This reduced the number of vendors of prostheses and forced the remaining vendors to aggressively cut prices to maintain the hospitals business.  This created significant savings which was shared with the orthopedists who participated in the process.

Opportunities to reduce payment variations
Some DRGs may not be judged suitable for inclusion in the initiative because of a high amount of variation in provider payments across admissions. These variations may be caused by differences in length of stay (which may affect physician payments even though it doesn't affect DRG payments), outlier payments or readmissions, or varying use of post-acute services in Models 2 or 3.  Even if the proposed bundled payment rate is established at the average payment amount, the variation between admissions within that DRG may be so high as to cause and unacceptable level of risk.  The analytics team should review these variations, assess their causes and determine if they can be reduced.  For example, the rate of readmissions may be reduced by careful post-discharge planning, which would reduce the variability caused by those payments.  Reducing the variation generally has the effect of lowering the costs.

Opportunities to reduce provider payments through utilization reductions
Payments to all providers represent a charge against the bundled payment budget.  If unnecessary services can be reduced, more funds are available to the contracting organization for distribution to participating providers.  Therefore, the analytics team should review the charges for each episode to identify areas in which utilization can be affected, using tools such as the graph in Figure 1 below.  Since these reductions will primarily be driven by physician behavior, participation by the physician members of the team will be critical to its success.


Assessment of provider network completeness

The potential for success of a bundled payment initiative is directly related to the physician participation in managing each episode. Therefore, the MS-DRGs proposed should be those in which physicians in the related specialty are on-board with the contracting organization, understand and agree with the objectives of the initiative, and are willing to participate in its success. 

Payment amounts vs. hospital costs
The hospital may also wish to consider the internal costs and profitability of an MS-DRG in deciding whether to include it in the proposal or in determining the discount rate to be proposed.  DRGs that are more profitable to a hospital may be candidates for larger discounts, while those with a lower margin may not be as attractive for discounting.

Analytical tools
The analytics team will require a robust tool set to sort through all of the data, drill down into the details, build simulation models of selected episodes at various levels of utilization, and create a comprehensive proposal for CMS that has a high probability of acceptance and successful implementation. The tools must include a back-end database that integrates the different CMS-provided files together to provide a composite view of each admission that includes all providers. Ideally that data can be integrated with the cost accounting data to provide a single view of the admission, integrating hospital costs with the provider payments. Such a system would provide a single, coordinated source of data about the admission.


The Bundled Payment Initiative provides hospitals with a wide variety of opportunities to experiment with coordinated patient care that offers significant benefits to hospitals and physicians alike. Many hospitals may find this alternative to be more attractive than the “all-in” approach of accountable care organizations.  

[1] The Medicare Bundled Payment Program – Participation Considerations – Healthcare Financial Management, September 2010
[2] Unfortunately it also reduces the incentive for providers to provide ANY services, which is why the quality measurement component of the initiative is important.

More Articles on Bundled Care:

HHS Announces New Bundled Payment Initiative

Copyright © 2024 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.


Featured Whitepapers

Featured Webinars