Five common pitfalls in commercial ACO shared risk arrangements

 Download the PDF

Across the U.S. health system, providers are transforming their delivery models in conjunction with the rapid expansion of "value-based" payment models. The end goal is noble and consistent with the “Triple Aim” of healthcare: provide better quality of care while managing overall costs and improving the health of populations. In many instances, the means to that end is the adoption of alternative payment models. These alternative payment models provide financial incentives to support higher-value healthcare.

One common "value-based" payment model is a shared risk agreement. As this name implies, these arrangements look to offer providers a share of any savings that can result from more efficient, higher-value healthcare. Providers might also have to share in the losses that could result from poor delivery and/or coordination of care.

Unfortunately, these models are not always designed in a manner that provides adequate alignment of financial incentives between provider and payer. While some agreements are marketed as “50/50” agreements, several commonplace provisions can leave the provider with far less than 50% of the savings. Even worse, many agreements expose providers to significant risks that they are not equipped to handle.

For these agreements to be transformational, as intended, providers and payers must work together to construct agreements that adequately reward providers, are economically viable over a multiyear period, and strive to transfer care management risk, not insurance risk. Below are five common pitfalls of shared risk agreements that can make it difficult to meet these objectives and put the provider at substantial risk.

1. Rebasing away your efforts

Target rebasing, the approach often used to develop the annual cost target under shared risk agreements, is a critical element, which has the potential to substantially reduce the aggregate share of savings an accountable care organization (ACO) receives over the course of an agreement. A common approach is to use the most recent year of experience as the basis for projecting the cost target for the next performance year. However, as illustrated in Figure 1, it is easy to demonstrate a realistic example where the aggregate share of savings an ACO receives under this rebasing method is only around 16%. This is a very different outcome from what may seem like a “50/50” risk-sharing agreement on paper.

Figure 1: Illustrative example showing the impact of rebasing over a five-year period

	Year 0	Year 1	Year 2	Year 3	Year 4	Year 5	Average
(1) Actual claims PMPM	$500	$525	$551	$579	$608	$638	$580
(2) Target assuming 7% trend		$535	$562	$590	$619	$650	$591
(3) Gross savings [(2) – (1)]		$10	$11	$11	$11	$12	$11
(4) ACO's shares [50% of (3)]		$5.00	$5.38	$5.59	$5.62	$5.93	$5.50
(5) Projected costs without ACO*		$535	$572	$613	$655	$701	$615
(6) Total savings produced by ACO [(5) – (1)]		$10	$21	$34	$48	$63	$35
(7) ACO's aggregate share of savings [(4) / (6)]		50%	25%	17%	12%	9%	16%
*Projected costs without ACO assumes costs would have increased at 7% annually without the ACO’s management initiatives.

As demonstrated in Figure 1, an annual rebasing approach immediately recaptures any savings an ACO generates. This means that an ACO needs to continue to generate savings in addition to the prior year’s savings to generate savings under the agreement. The opportunity to continue reducing costs is diminished across multiple years. There are multiple alternative rebasing approaches which can be used to delay or minimize this recapturing of savings and significantly increase an ACO’s chances of receiving a fair share of the savings.

2. Failing to dig into the trend methodology

The method used to project baseline costs to the performance period (i.e., develop the target for the performance period) has a significant impact on the outcome of shared risk arrangements. It is critical for ACOs to understand the methodology behind this trend assumption and ensure that it is applicable to the specific circumstances of the ACO, such as changes in operating expenses and population mixes. Most importantly, the ACO should have the ability to review and mutually agree upon the trend assumption.

Often, an ACO’s trend is compared with a general market trend or similar indices, which do not fully reflect the ACO’s unique characteristics. For example, a particular provider may have historically been reimbursed at a lower rate than the market and, thus, more likely to require a higher fee schedule increase, even though it is still operating efficiently compared with the market. In this type of situation, a market trend will not reflect an ACO’s specific fee schedule increase and possibly penalize providers even in a scenario where care is more efficiently delivered. There are many other reasons why an ACO’s costs will not trend at the same rate as “similar” populations, and those differences should not drive the success or failure of an ACO under these agreements.

3. Overlooking the unintended consequences of minimum savings/loss corridors

Many commercial risk agreements contain a minimum savings/loss corridor provision, which prevents or limits settlements within a certain threshold around the cost target. Typically, this is expressed as a percentage of the cost target. The intention of these corridors is to prevent payments to providers resulting from random cost variation rather than care management. However, in practice, these corridors often impede providers’ ability to share in the savings they are generating. At minimum, providers should assess the underlying claims volatility based on the nature of the population (e.g., commercial, Medicare) and the expected number of attributed beneficiaries.

The most obvious impact this has on providers’ shared savings is that risk corridors can both reduce likelihood of payment and the potential amount of payment. Assuming a risk corridor of 2% for a 25,000-member commercial population, our analysis suggests the likelihood of receiving a payment may be reduced by as much as 30% to 40%. The impact the corridor will have on the amount of shared savings will greatly depend on whether or not there is a "hard" or "soft" risk corridor. For hard corridors, savings are only paid out in excess of the corridor. Assuming the provider achieved 3% in savings and there is a 2% risk corridor, this reduces disbursements by 66%. This can be detrimental to the economic viability of the alternative payment model. On the other hand, a soft corridor pays out "first dollars" in excess of the benchmark on the condition that the corridor has been exceeded. Using the same example, the provider would share in the full 3% savings. This is a significant improvement compared with the hard corridor.

4. Viewing quality adjusters solely from a clinical perspective

Most shared risk agreements have some form of a quality adjustment that links shared savings/losses to the quality of care provided (i.e., the higher the quality, the greater the share of savings an ACO can earn). Quality adjustments are primarily intended to ensure that a certain level of quality is maintained and care is not withheld to keep costs low. However, quality adjustments often make it difficult for an ACO to receive a true “50/50” split of any savings generated.

The quality adjustment mechanism can vary widely from payer to payer, but there are many commonalities among the various models, which ACOs should pay close attention to, such as:

• Relevance of measures: It is critical for quality measures to be applicable to the membership at-risk and the data available. For example, if Rx data isn’t available for a large portion of the population (as is often the case with self-funded employers using separate pharmacy benefit managers), quality measures requiring Rx data should be limited or removed.
• Application of quality scores: What happens if an ACO improves or maintains high quality when there is a deficit in a performance year? In many shared risk agreements, the answer is nothing. An ACO’s share of a savings is typically reduced if quality targets are not met, yet the ACO’s share of losses is not similarly reduced if quality targets are met. Successful quality performance should apply symmetrically to both savings and losses (i.e., higher quality scores should translate to a reduction in an ACO’s share of deficits).
• Diversity of measures: Shared risk agreements often aim to keep the number of quality measures low enough to ease administrative burden and focus quality efforts on a concise subset of services. While this approach is ideal, it is important to ensure the measures are not overly concentrated around specific conditions or specialties. For example, if the majority of quality measures are related to diabetes and cardiovascular services, an ACO’s performance under these agreements is heavily dependent on success or failure of just two subsets of services, which may only represent a small portion of overall system expenditures and/or patients.
• Number of opportunities: It is important to understand the number of opportunities (i.e., the number of patients or episodes) that the ACO currently has under each measure and ensure there is a minimum threshold in place so the ACO’s success or failure isn’t based on a measure where there are only a handful of opportunities.
• Setting realistic targets: The financial success of a shared risk agreement is highly dependent on the achievability of the quality targets. At a minimum, the targets should be based on historical information relevant to the ACO and the current marketplace, and not based on theoretical targets or nationwide benchmarks that are not reflective of the local market in which the ACO operates.

5. Assuming perfection in the process

An often overlooked yet vital component of a shared risk agreement is the ability for the ACO to audit the data and calculations the payer is using in development of the cost targets and performance measurement. There are many moving parts in shared risk agreements such as continuously changing provider rosters, changes in member attribution, and changes in underlying claims systems. There are many opportunities in the data collection and processing for a breakdown to occur. Seemingly innocuous nuances in the data can often turn out to be icebergs exposing a much larger underlying issue. An ACO can perform several key checks during implementation and on an ongoing basis to identify issues. However, this process cannot be done effectively without the proper provisions in the contract requiring regular and timely detailed data transfers and obligating the parties to follow a process to resolve data issues. Even with the proper contractual terms, a formal, vigilant internal review process is needed to evaluate the data received.

Conclusion

Shared risk arrangements were developed as a means to better align financial incentives with the end goal of more efficient, higher-value healthcare. Unfortunately, many of these arrangements fall short on aligning financial incentives. Each of the common provisions outlined above could be detrimental to a successful provider/payer partnership if not structured appropriately.

With the U.S. healthcare system undergoing one of the largest transformations in recent decades, now is the opportunity to transition to better aligned, equitable payment models. It is critical providers and payers work together to avoid these all too common pitfalls.