Toward Accountable Payment

Designing effective gainsharing models

by Rob Lazerow

Learn how gainsharing through bundled payment programs differs from traditional models, review gainsharing ground rules, and explore how active bundled payment programs design effective gainsharing models to reward performance improvement.

As organizations continue to evaluate CMMI’s Bundled Payments for Care Improvement Initiative (BPCII), we have received a range of questions about gainsharing models. One of the key benefits of participating in the BPCII is gaining newfound ability to gainshare with physicians—especially independent physicians—underscoring the need to develop effective incentive models.

Although use of gainsharing is typically limited because of fraud and abuse regulations, CMMI can use its waiver authority to loosen these restrictions. To receive this clearance, however, CMMI requires a thorough explanation of providers’ proposed gainsharing methodology in the BPCII application.

To help Medicare Payment Innovation Project participants develop gainsharing models, this blog post addresses three of the most common questions we receive about gainsharing:

1. How does gainsharing through a bundled payment program differ from a traditional gainsharing program?

2. What rules or restrictions has CMMI placed on gainsharing for the BPCII?

3. What are key design considerations for developing effective gainsharing models within bundled payment programs?

Differentiating gainsharing within bundling programs

Hospital and health system leaders often ask why gainsharing within a bundled payment program is any different than traditional gainsharing programs, which faced several challenges in the past. First, programs needed to navigate the legal complexities of gainsharing, ensuring they could pass review by the Office of the Inspector General (OIG). Even after passing the legal hurdles, programs may have ultimately floundered. Past gainsharing programs often enjoyed a few years of initial success but then stalled once performance improvement began to plateau. As a result, physician bonuses diminished or dried up altogether, effectively ending the gainsharing programs.

Two key differences distinguish gainsharing within bundled payment programs from historical attempts to gainshare. First, bundling programs typically take an expansive approach to calculating cost savings. Although historical gainsharing programs focused narrowly on device costs, bundling programs can share cost savings generated across the entire episode of care. Depending on the scope of the bundle, hospitals can incorporate a range of cost savings targets in the gainsharing bonus pool—potentially including length of stay, ICU use, basic supplies, ancillary utilization, readmissions, and post-acute care utilization—in addition to device cost savings. Using an expansive approach to calculating cost savings reduces the likelihood that bundling programs exhaust their cost reduction opportunities, ultimately increasing program longevity.

Second, when gainsharing within a bundling program, organizations typically use a stable cost baseline—the benchmark used to calculate cost savings—throughout their gainsharing programs. In the past, gainsharing programs reset their cost baseline annually, rewarding physicians only for incremental improvement above and beyond the previous year’s gains. This annual reset was a key contributor to the plateau effect and subsequent decline in physician bonuses.

Within active bundling programs, such as Medicare’s ACE Demo, organizations do not reset the cost baseline annually, but instead use the same cost baseline for the duration of their multi-year bundled payment contracts. In essence, as long as reimbursement levels remain steady or increase across the contract term, the programs are willing to measure savings against the initial cost benchmark. As a result, physicians who improve performance quickly and maintain that success will continue to receive rewards for their efforts. Using a stable cost baseline re-positions gainsharing from a reward for incremental improvement to an annuity that pays out across the contract period—assuming physicians maintain cost savings performance.

Gainsharing ground rules

Now that we have outlined how gainsharing through a bundled payment program is different than historical gainsharing efforts, we can turn to the details of gainsharing design. The BPCII Request for Applications (RFA) includes relatively few restrictions on gainsharing design, offering providers ample opportunity to design innovative gainsharing methodologies. That said, CMMI set forth five ground rules to keep in mind: 

1. Provider eligibility and participation: Gainsharing bonuses may be paid to physicians and non-physician practitioners; applications must specify which providers will be eligible to receive gainsharing bonuses. Participation by eligible providers must be voluntary, and there cannot be negative consequences for providers who decline to participate.

2. Bonus size: Gainsharing bonuses cannot exceed 50% of providers’ non-discounted Medicare fee schedule payment rates. As a result, the maximum provider payment is 150% of standard fee schedule rates.

3. Quality standards: To qualify for gainsharing bonuses, eligible providers must meet minimum quality thresholds and actively engage in quality improvement efforts. Programs must propose these quality thresholds and their method of performance monitoring in their BPCII applications.

4. Patient care standards: Gainsharing models cannot encourage providers to reduce or limit medically necessary care.

5. Basis of bonuses: Gainsharing bonuses must be based on demonstrated cost savings. Further, bonus payments cannot reflect the value of participating physicians’ referrals.

Designing effective gainsharing models

With our gainsharing ground rules in place, we can now explore gainsharing methodologies. In our research of active bundled payment and standalone gainsharing programs, we have identified three main levers within gainsharing design: qualifying criteria, primacy of savings payouts, and timing of bonuses.

As the graphic below displays, the ways in which a bundling program pulls each of these levers is ultimately based on the strength of the underlying hospital-physician relationship. If a hospital needs to convince physician skeptics to join a bundling program, leaders will likely select design considerations that favor physicians (shown on the left side of the graphic). Conversely, a hospital that begins with the benefit of having strong functional integration can structure the gainsharing model to minimize hospital losses (shown on the right side of the graphic).


The following 15 questions can help bundled payment programs determine how best to use each of these levers given local dynamics:

Eligibility criteria

  • Which providers will be eligible to receive bonuses?
  • What quality and efficiency measures will the model include?
  • What performance thresholds must providers meet to qualify for bonuses?
  • Will the program measure and reward individual or group performance—or both?
  • Will the program measure and reward absolute success or improvement against historic performance—or both?

Primacy of savings payout

  • Should the program recoup the reimbursement discount before offering bonus payouts (e.g., retain first-dollar savings until the hospital breaks even)?
  • Should the program offer first-dollar bonuses to encourage voluntary provider participation?
  • Will the hospital share in savings distribution from the outset (e.g., 50/50 distribution from first-dollar savings)?
  • Will the model use a constant savings distribution rate (e.g., 50/50 distribution regardless of total performance) or develop a tiered model where the distribution rate varies based on total cost savings?
  • How will the program ensure that bonuses do not exceed the maximum allowed?

Timing of bonuses

  • Should the program pay bonuses frequently to reinforce the link between performance improvement and financial success?
  • What is an appropriate timeframe for measuring performance given the metrics selected (e.g., month, quarter, or year)?
  • As the measurement period increases in duration, what is the incremental risk of a limited number of high-cost cases eroding providers’ bonuses?
  • How much time will the program need to finalize performance measurement and calculate bonuses?
  • Should the program wait to pay bonuses until it collects reimbursement from payers?

Additional gainsharing resources

For detailed case studies of bundling programs using effective gainsharing models, read our study Succeeding Under Bundled Payments (especially pages 39 to 46). Additionally, you can test your own gainsharing models using the Inpatient Bundled Payment Impact Calculator.

As always, if you have any questions, please feel free to call or email me directly.