Physician Risk-Sharing in Medicare Bundled Payment Programs

Submitted by jonpearce on Mon, 2017-04-10 15:40

I was quoted in an article entitled "Bundled Payments: What Doctors Need to Now" in the April 2017 issue of Medical Economics, which offers some guidelines to physicians about participation in these programs. This article expands on some points made there, and provide some additional considerations for physicians who may wish to participate in these programs.

The Medicare program currently operates two major bundled payment initiatives – the Bundled Payment for Care Improvement (BPCI) program, which is a voluntary program including 47 different types of episodes, and the Comprehensive Care for Joint Replacement (CJR) program, which is a mandatory program that primarily covers hip and knee replacements. In addition, the Enhanced Payment Methods (EPM) program for selected cardiac services was announced in 2016; however it has been delayed until September 2017 and possibly later. CMS also operates the Oncology Care Model (OCM), which operates differently than the above programs.


There are two opportunities for physicians to participate in the Medicare model payment initiatives. The first is under a "gainsharing" arrangement with the healthcare provider who is the ultimate participant (meaning it has the contract with CMS) in this program. Under a gainsharing arrangement, physicians must generally meet certain quality criteria and then have the opportunity to be paid a portion of the savings that can accrue in the bundled payment program. These savings can be both internal (reductions and length of stay, savings and medical supplies, etc.) or external (reductions in utilization of post-acute providers, reductions in readmissions, etc.).

One important limitation in the amount of payment that a physician can receive is that it cannot exceed 50% of that physician’s Medicare RBRVS payments during that episode. In the example in the Medical Economics article, the physician had been paid $320.40 in Medicare payments in that episode, presumably for hospital or office visits, and was eligible to receive half of that amount, or $162.20, as a gainsharing payments for that episode. Obviously, surgical specialists have a greater opportunity to benefit from gainsharing arrangements, since their Medicare payments are higher; an orthopedist performing a hip replacement and receiving approximately $1200 for Medicare for performing the surgery could receive a $600 gainsharing payments for that episode. This requirement by CMS significantly limits the opportunities of some physicians to receive large gainsharing payments because their underlying payments from Medicare for treating those patients are relatively low.

An important advantage of gainsharing programs, though, is that they rarely expose the participating physicians to downside risk. In our experience, gainsharing payments are almost always positive. The participating provider assumes risk for high episode cost, and does not pass that risk along to the physician.

Assuming Episode Financial Risk

The other opportunity for physicians to financially participate in Medicare bundled payment programs occurs when the physician (or group of physicians) becomes the actual participant in the bundled payment program. CMS allowed physicians to participate in the initial BPCI program, in which enrollment is currently closed, but not in the subsequent CJR program. The final rules for the EPM program also limit participation to hospitals, not physicians. However, it's widely assumed that CMS will unveil a new broad-based bundled payment program in 2018, and the new administration may be more amenable to allowing physicians to be the primary contractors who would assume all episode risk.

In this situation physicians contract with CMS directly or through an intermediate party to be financially responsible for all Medicare costs incurred by patients in those episodes. Because there are no other participants in the episodes, there is no limitation on how much they can earn. Therefore, their upside earning potential is far higher in these arrangements that under gainsharing arrangements.

Unfortunately, these organizations also assume all of the downside risk of high-cost patients, which can be significant. The "episode composition" graph below shows the cost of a group of acute myocardial infarction (AMI) episodes along with the cost components of those episodes. The total of each episode is compared to a "target" established by CMS, and the participant receives a payment from CMS for the difference between the episode cost and the target if the target exceeds the cost. Conversely, the participant pays CMS the difference between the target and the episode cost when episode costs exceed the target (although extremely high cost episodes are limited by an "outlier" threshold). In this example, targets would range from between $25,000 to $35,000 depending on the DRG, meaning the participant could be liable for significant costs associated with the higher cost episodes. The potential for these losses may make participation in these types of episodes infeasible for some physician groups.

“Conveners” May Assume Risk

The BPCI program allowed non-provider organizations to act as "awardee conveners" to assist physicians and other providers in participating in this program. One major function conveners performed was to assume some level of the downside risk, which shielded physician groups from unmanageable levels of potential loss. Such organizations may reappear in future Medicare bundled payment arrangements if physicians are allowed to participate.

Hearts are Different from Hips and Knees

Many orthopedists participated in the BPCI program in the major joint replacement (MJR) episodes, and have reportedly achieved significant levels of financial success. However, other types of episodes have significantly different structures, and the ability to create savings may be quite different. The primary strategy in MJR episodes is to refocus post-discharge patient care away from institutional settings and into the home, with or without home healthcare. In addition, developing networks of post-acute providers whose care can be coordinated with that of the orthopedists has led to lower lengths of stay in post-acute facilities and consequently lower costs. In many cases these initiatives could be initiated at a policy level and require relatively little management of individual patient care. Readmissions in these episodes are relatively low (less than 10% of episodes), so management of readmissions (which is quite resource-intensive) is not usually a primary strategy in these episodes.

By contrast, the episodes involved in the EPM program (acute myocardial infarction (AMI), percutaneous coronary interventions (PCI), and coronary artery bypass grafts (CABG)) have a significantly different cost profile. Readmission rates in AMI episodes are significantly higher (around 50%) than in MJR episodes, and require a much higher level of costly personalized care management. PCI and CABG episodes generally have low use of institutional post-acute care, which reduces the opportunity to create savings. However, these episodes can have infrequent but extremely costly readmissions that can significantly add to their financial risk. Therefore, physicians considering participating and taking full financial risk for these types of episodes should carefully evaluate the opportunities and risks associated with these types of episodes, rather than assuming that they will create the same level of success as MJR episodes.