top of page

Don't take your eye off the bundle

Updated: Apr 20, 2023

Both Medicare and commercial versions are evolving in some surprising ways.

By Amanda Berra, Chief Education Officer, Union Healthcare Insight

Remember circa 2015, when Medicare’s Comprehensive Care for Joint Replacement (CJR) program appeared, and bundled payments for episodic risk were the talk of the industry?

Today, bundles are no longer novel, don’t generate as many trade press articles or research studies as they once did, and the data on uptake and effectiveness can best be described as “patchy”[i]. In total, this makes it all too easy to lose track of where we are on them now.

Yet bundles are an important part of the overall healthcare strategy landscape, and their situation is always changing. So let’s take a quick spin around Planet Bundled Payments to ensure we’re all up to date.

1. What’s happening in Medicare bundles?

BPCI-A is being reworked

If you’re a citizen of the healthcare world, you’re already aware of Medicare’s bundled payment program, Bundled Payments for Care Improvement Advanced Model (BPCI-A).

What you may not have been tracking is the fact that the program was set to sunset in December of 2023, but has been extended to 2025. Acute care hospitals, physician groups, and Medicare ACOs may apply through May 31 of this year to participate in the next round; current participants can just sign amended participation agreement. [ii]

The purpose of the BPCI-A rework is to improve program economics—from CMS’s point of view

For those saying, ‘Wait a minute, why would BPCI-A need a redesign to save money? I thought bundles were the alternative payment model (APM) that have been most reliably shown to reduce cost?' Here’s the story.

Bundle effectiveness study results vary, but most show savings, especially if the bundles are surgical rather than medical; effect on quality tends to be neutral to positive. By the way, from my point of view, a muted pattern of cost/quality improvements is actually strikingly positive, given how close-to-impossible it is to study the effect of just one APM type in a sea of different APM models—at least here in the U.S.[iii]

Since bundles seem reasonably effective at reducing episode cost, you could be pardoned for thinking that must translate into savings for the payer—in this case, the federal government—because, that would make common sense. And also because many in the industry think bundled payments in general boil down to either a straightforward discount/price cut, or else a 'gainshare', in which, if money is saved, that savings is split between the payer and the provider.

But never forget healthcare’s "funhouse economics." In this part of the funhouse, BPCI-A had a bonus/penalty structure set around a benchmark. If you came in under the benchmark, you received a bonus; if you didn’t, you received a penalty. Benchmarks varied from provider to provider.

Good news for BPCI-A: Total episode spending has (so far) come in lower than the comparison group—with surgical bundles, in general, saving way more than medical bundles—in model years 1-3. Year 3 is the most recent year for which detailed net savings data is available.[iv]

Bad news for BPCI-A: The program's benchmarks have so far been set too high, i.e., made too attainable, such that many more providers received bonuses than penalties. Thus CMS both saved money (on episodes) and lost money (on the penalty/reward mix)—with the net effect of losing money.

To make BPCI-A yield actual savings, in the two-year extension, CMS is going to reduce target benchmarks, i.e., lower the limbo stick. In fact, it already did so retroactively for year 4 and forward in BPCI-A – causing consternation among providers[v], and high-profile analytics provider Signify Health (recently acquired by CVS Health) to declare the program financially unsustainable from a provider/enabler point of view. Signify has closed its episodes of care analytics business and exited its activities supporting providers in BPCI-A.[vi]

Might CMS make a future round of BPCI-A mandatory? Well...maybe.

Back in September, 2020, CMMI announced that the next round of BPCI-A would likely be made mandatory, starting as early as 2024[vii]. Presumably that timeline would now look more like 2026, the end of the new two-year (voluntary) extension.

No mandatory version of BPCI-A has yet been announced. If it did come along now, it would a relative surprise because there has been such a long period of quiet on the bundle front. If you were betting against it, you might say: Well, the program hasn't net-net achieved savings. Also, with only about a third of eligible hospitals participating in even one clinical episode (and I'm sure a much lower proportion of physician groups), it would mean drafting most of the industry into a program in which they have declined to voluntarily participate. That would mean a major lobbying battle.

On the other hand, the goal is to save CMS money. And, one gets the sense from CMS and its contractors that they feel, with a little more tinkering, they can figure out how to consistently come out ahead on the voluntary version of BPCI-A [iv]. If CMS can do that, presumably during the two-year extension, then mandatory bundles start to seem like a reasonable possibility. Especially because the mandatory version of the same program might end up being a much more robust driver of federal savings.

As Rachel Hollander, senior manager at McDermott+Consulting put it, “There is an acknowledgement that voluntary models are subject to selection bias, which can limit potential savings. This may mitigate in favor of more mandatory models going forward. CMMI signaled its intention to do just that in its 2021 Strategy Refresh, which cited voluntary participation as a key challenge limiting participation and potential savings in past models and as a design flaw that the agency plans to address in future models.”

2. What’s happening in commercial/employer bundles?

Commercial insurer plan bundle uptake, in the context of plan products for multiple employers, is still overall low, for all the reasons it always has been. For example: commercial bundles often involve a negotiated-rate discount in exchange for volume; negotiated rates tend to be the way they are for a reason; and volume steerage is quite difficult to pull off. And the only way most commercial bundles can work for plans, employers, and providers all at once is by directing enrollees to a “best of the best” single provider in a region (not even just a market).[ix] That is a hard fit with most commercial plans, which need to contract with many providers. Picking and choosing to that extent makes contracting conversations with the rest of the provider partners contentious.

Employers, however, continue to be interested in bundles, with a slow-but-steady uptake pattern. But I'll qualify that statement by noting that available data tends not to track bundles per se, but rather use of ‘centers of excellence’ (COEs). ‘Using COEs’ can mean anything—including, but definitely not limited to, the commercial version of bundles[x]. In this way, COE use, commonly reported by benefits consultants and other industry groups, is camouflaging the size of the subset of these arrangements that are paired with a bundled payment.[xi]

Three factors are keeping the heat up under commercial bundles:

#1 Employers are concerned about healthcare cost trend, and disillusioned with high deductibles, wellness programs, PBMs, and other failed levers to control cost growth. Meanwhile commercial bundles keep racking up impressive case studies in which spending per episode is curbed.[xii] There is even some research to suggest that bundles can act as a de-facto reference price – one employer that contracted with a single, local provider for a discounted bundle soon found all the other providers in the market lowering their negotiated rates to match that of the COE.[xiii]

#2 The emergence of third-party bundle constructors, who can make bundled payment/COEs more feasible from the employer viewpoint. Companies such as Carrum Health and Trascarent can identify the COE, help with contracting, and even integrate with the health plan/TPA.

#3 A new wave of concern among large employers, especially since the passage of the Consolidated Appropriations Act (CAA), about their state and/or federal government-mandated fiduciary role in actively managing health benefits with an eye toward transparency, efficiency, and quality.

Sarah Raali, a partner at McDermott Will & Emery who practices healthcare benefits law told us, “With the CAA and additional health plan laws and regulations, Congress and the DOL appear to be underscoring that employers have a responsibility to ensure that their health plan is cost-effective, offers their participants quality care, and complies with applicable laws and regulations. Including transparency in coverage, mental health parity, reasonable compensation to service providers, and the prohibition on gag clauses in service provider contracts.”

Such a general mandate does not push employers to adopt bundled-payment-COEs per se, however, it adds to the general momentum in their direction — because part of a major lure of commercial bundles is the greater transparency and predictability in both cost and quality that they offer.

Employers are beginning to look at a type of bundle they have long avoided: oncology care

One interesting change on the employer-sponsored bundles front is a spike of interest and activity around oncology bundles.

The first bundles to get commercial momentum—and the ones that are still most common today—were and are specialty procedures such as joint replacement. These are natural candidates because they are relatively simple and clear-cut to bundle. They also offer clear potential ROI wins for employers, based on their volume, price variability, the potential to shift to lower-cost sites, along with issues with appropriateness of care (e.g., it is common to be recommended for surgery when in fact physical therapy may be lower-cost and more effective).

In contrast, in oncology, the keys to success in bundling weren’t, and aren’t, as clear. As Carrum Health Market VP Deirdre Saulet told us, “Oncology care is really hard to bundle – private payers have been trying to do payment reform in oncology for a long time, through many mechanisms.”

And equally, for a long time, the market wasn’t ready. Saulet said: “For the last decade, I have been talking about value-based employer-provider partnerships in oncology with employer executives, and for most of that time, they said they were nowhere near ready. They said things like ‘Cancer is too complex, too personal, too scary’. They let it go for then.”

But now, cancer is the top driver of employer healthcare cost. As Saulet puts it, “Basically, it’s a behemoth they can’t avoid.”

Prices are the main driver of increased oncology care spend; bundles offer a robust solution

Back to Saulet: “It’s about the price of everything in oncology. We have had all this innovation and the cost of cancer care will keep rising as the population ages and as people live longer. Cancer care has become a chronic disease for many. But while we have made so many advances, we have to confront that they are expensive—and there is little oversight or regulation on price.” Not to mention the drug value chain, which flows through PBMs, is notoriously murky and inefficient.

There are other employer-friendly tactics in the market to try to rein in the cost of oncology, or any other clinical episode—tactics that may make employers feel less worried about overstepping in a high-stakes, complex, personal clinical area for employees and families. For example, cancer navigation services and second-opinion platforms are both non-bundle options that help, by way of assisting with accurate diagnosis and evidence-based treatment plans.

But ultimately, bundles, if they can be pulled off, are a more direct and robust way of tackling the problem, because they both limit price and steer enrollees to the most rigorously determined, top-quality, evidence-based providers. As price continues to be a pain point, and third-party enablers continue to lower the feasibility barrier for employers (or even, in the future, more health plan partnerships), I think it’s reasonable to expect the commercial bundle space to grow.

In sum

Outside the specialized world of bundles, much of the industry may have taken its eye off the evolution of this model – not least because of patchy data and the complexities of Medicare’s main program, BPCI-A. But given the long term cost/price challenges that healthcare payers face, and the potential bundles have shown to decrease cost/price (with the aforementioned caveats about BPCI-A), it seems likely that this model will become increasingly relevant in the future.


Union is a researcher-founded insight company. We understand the history, incentives, and forces that drive the healthcare industry. We also have a passion for learning and teaching others about how it all works. Want to chat? Reach out at


[i] So many data challenges. Including: 'Bundles as a share of spending' is obscured by the fact that the best source on this, HCP-LAN’s alternative risk payment model survey, lumps bundles in with other forms of risk. Commercial bundle uptake among employers is obscured by benefits and business groups collecting data on ‘COE use’ (which can mean anything). To my knowledge, there is no current, prevalence-oriented survey of bundle uptake among providers that spans commercial, Medicare, and Medicaid bundles. There is no unified source that I know of looking at Medicaid bundles. In general, far more information is available about Medicare bundles than commercial or Medicaid – but, even in the Medicare world, not only is the data extremely complex to deal with, but also only selected data points are available. I notice that to get data for their recent JAMA article on reconciliation payments in BPCI model year 4 , Ryan et al had to submit a Freedom of Information Act request (!). See Ryan, A. et al, “Reconciliation Payments in the Bundled Payments for Care Improvement Advanced Program and Reductions in Clinical Spending Needed for CMS to Avoid Financial Losses,” JAMA Network (research letter), April 11, 2023, available at: [ii] American Hospital Association (AHA). [iii] See page 59 for MedPAC’s discussion of ‘cohort contamination’ – i.e., all providers of any size are participating in different VBC activities, making it really hard to pick the effects of any given single one out of the noise: [iv] Lewin Group for CMS. "In Model Year 3, the BPCI Advanced Model reduced average standardized episode payments by $1,028 per episode, or 3.8% of the baseline mean, relative to the comparison group". Lewin quote that, to me, suggests they think that with some more tinkering, the program could be made sustainable: ““Target prices may have been too high for medical clinical episodes but were generally more accurate for surgical clinical episodes in Model Years 1 and 2 and Model Year 3. The target price methodology remained unchanged in Model Years 1 through 3 but changed in Model Year 4 [2021]. Therefore, estimates of savings may differ for Model Year 4 and will be reported in the Fifth Evaluation Report.” Lewin Group for CMS. “CMS Bundled Payments for Care Improvement Advanced Model: Fourth Evaluation Report,” March, 2023. Available at: [v] Wickline, A. "Surgeons Fold Against Medicare's Stacked Deck" (Opinion). STAT, August 25, 2022. Available at: [vi] Bagasse, J. "Signify Health Exiting CMS's Bundled Payment Program," Healthcare Finance News, July 11, 2022. Available at: [vii] Daily, R. "Big Medicare Bundled Payment Changes Could Cut Participation, Advisors Say". Healthcare Finance News, Sept, 2020 [viii] [ix] This is the case because, the provider in question has to put in a ton of work, may price the bundle at a discount compared to rest of commercial book of business, and most importantly, it becomes a COE on the basis of its ability to filter out avoidable surgeries before they happen--right answer for the patient and the payer, not for FFS economics [x] In commercial bundles, employers contract with a provider COE; the provider takes on the work of ‘convening’ all the players across the continuum, and the employer steers its enrollees to that COE for the relevant procedure or diagnosis [xi] Per COE data (such as it is), 2021 use of COEs among employers includes maternity (31%), CV (46%), Oncology (46%), MSK (47%). Source: Business Group on health, “LARGE EMPLOYERS’ HEALTH CARE STRATEGY AND PLAN DESIGN SURVEY,” 2023. Available at: [xii] Whaley, C. et al. “An Employer-Provider Direct Payment Program Is Associated With Lower Episode Costs45% per procedure savings when procedures were performed through Carrum Health,” Health Affairs, March, 2021. Available at: [xiii] Commercial bundles forcing prices down, a de facto reference pricing model “One recent study by Christopher Whaley and colleagues examined a BPCI-like program implemented among self-insured employers for four surgical procedures. After implementation, episode prices for these procedures saw a 10.7 percent relative reduction. This reduction was seen across all providers performing those services for the firms’ employees, not just those participating in the bundles.” “Value-Based Payment As A Tool To Address Excess US Health Spending, " Health AffairsResearch Brief, December 1, 2022.DOI: 10.1377/hpb20221014.526546 Available at: ://


Join our mailing list to see future posts

Thanks for submitting!

bottom of page