In this new era of value-based payment, how do health care organizations set benchmarks for cost of care? There are a number of effective strategies that organizations use singly or in combination. This APM Issue Brief explains and explores several options.
Healthcare organizations are increasingly adopting payment and contracting models that reward providers for value over volume. These alternative payment models, or APMs, can take many forms, but all include incentives for higher quality and cost-efficiency and encourage providers to more effectively coordinate patient care and manage population health.
Population-based payment models1 are one form of an APM. A population-based payment model covers all or most healthcare services for a defined group of individuals involved in a contract. Under such a model, a payer generally sets benchmarks for quality and the cost of providing care to the designated population for a set period of time. Population-based payment promotes the concept of identifying a single amount or “target” cost to cover care for all individuals in a population.
Establishing the correct baseline benchmark for cost of care is perhaps the most critically important aspect of developing any APM.
Establishing the correct baseline benchmark for cost of care is perhaps the most critically important aspect of developing a population-based APM, because it serves as the starting point for establishing the financial targets throughout the contract term. Under population-based payment, this benchmark is set at a level describing all members in the population – such as using a per-member-per-month (PMPM) cost. Some multi-year APMs recalculate this baseline benchmark on an annual basis; others establish the baseline benchmark in the first year of a contract and adjust it annually.
Setting an appropriate cost of care benchmark is a challenge for any payer and provider. An overly aggressive (lower cost) benchmark may be perceived by providers as unachievable and may dampen their willingness to accept the program and collaborate with the payer. Conversely, a level that is too high (i.e., a higher cost benchmark) may undermine a payer’s efforts to achieve the desired performance improvement goals. A successful benchmark is a proper balance between accountability and adoption – creating a “win-win” scenario for both payer and provider.
What factors should you consider when establishing a cost of care benchmark? This Issue Brief explores the most common strategies:
- A provider cost of care approach;
- A peer group cost of care approach; and
- A blended strategy that combines both of the above approaches.
1. Provider Cost of Care Approach: Using Historical Costs
The provider cost of care approach uses a provider’s prior cost experience to set a benchmark. This strategy calculates the historical cost of care for applicable medical and pharmacy services for a designated population and uses this amount as the basis for creating the baseline for a contract going forward. This calculation involves taking the following key steps:
- Define which providers are to be included in the payment model;
- Determine the designated population associated with these providers, using an appropriate attribution methodology to attribute members to providers;
- Calculate the total applicable medical and pharmacy costs incurred by these members during the defined historical period of time; and
- Divide this total by the number of member-months for the designated population over that same time period. The result is a prior, or historical, PMPM cost of care.
This provider cost of care is then used as the starting point for any specific budget adjustments or trending factors required to establish a baseline for the contract going forward. This benchmark can then be used to set an initial provider budget for the contracted population; it can also be further adjusted as necessary.2
Figure 1 describes setting a population-based benchmark using provider cost of care. Health System A is the contracting entity and includes three groups of primary care providers – Groups 1, 2 and 3. Other provider groups in the network affiliated with other health systems could also be involved in the payment initiative. Through use of an appropriate methodology, 58,625 members are attributed to the three provider groups affiliated with Health System A, based in part on their existing primary care relationship with these groups. Enrolled member months and historical costs for this population are calculated using a defined period of time — in this example, 12 months. A shorter or longer period of time may also be used.
The applicable total medical and pharmacy costs are computed for the same time period. These amounts include the cost of all covered services delivered to the designated population. Costs are typically computed using an agreed upon financial measure, such as payer-allowed claim costs. In many cases, the relative health risk of the individuals included in the population is also assessed – to ensure that comparisons account for differences in the underlying disease burden across populations. For this example and the example included in Figure 2, significant differences in risk are assumed not to exist – allowing direct comparisons of historical costs. (For help with patient attribution, risk adjusting benchmarks and budgets, see additional Qcentive APM Issue Briefs.)
Historical cost PMPM is computed for each provider group and also Health System A. The historical cost of care PMPM of $416 can then be used as a basis to determine a final benchmark for that organization, including any further adjustments for future trends and other contractual considerations.
The provider cost of care approach has a number of advantages. It can be attractive to providers because it is based on their own past experience. It is highly transparent and therefore is often seen by providers as being more acceptable. Because the model is built on a provider’s own experience managing a population, providers are more likely able to understand and drive solutions that improve cost-efficiency and quality. This can promote both short-term success and continued improvements over time.
Figure 1. Population-Based Payment: Global Budget Model Setting a Benchmark Using Historical Provider Cost of Care
Important Considerations in Using the Provider Cost of Care Method
Cost History Can Reward or Penalize. Despite its advantages, detractors of the provider cost of care approach argue that it may unfairly reward or penalize providers depending on their cost history. For historically higher-cost provider organizations, past inefficiencies are built into the baseline target and become part of the benchmark going forward. For example, if a provider’s historic costs are 20% above the average for a comparable network, a benchmark built around this experience will potentially reward the group because there is more opportunity to improve. These groups can potentially achieve more immediate savings and rewards, yet still remain above the network average.
Conversely, provider groups with historically lower costs are assigned a lower benchmark and therefore face greater challenges in creating savings and rewards – in effect being put at a contractual disadvantage relative to others due to their past success at lowering costs.
Balancing these issues and understanding the resulting incentives (i.e., how improvement is rewarded) is important in crafting contractual terms.3
2. Peer Group Cost of Care Approach: Averaging Network Costs
A financial benchmark can also be established using the experience of a provider’s peers. Rather than using historical cost of care for a particular provider, this approach calculates a benchmark using the historical cost of care for all providers and members included in that provider’s peer group or network. Peer groups can be identified based on factors such as geography or the characteristics of the patient populations involved. In all cases, the key feature of this strategy is that providers are held to a benchmark influenced not only by their own experience, but also by the experience of similar provider organizations.
Calculating the cost of care for a provider’s peers is similar to computing historical cost for an individual provider organization – with a wider group of providers and members used for the calculation. The example in Figure 2 shows that the experience of both the contracted provider organization and its peers are included in the calculation. As with the provider cost of care approach, the resulting PMPM cost can then be used to set the baseline financial benchmark for the contract and projected budgets.
As shown in Figure 2, Health System A is the contracting entity and includes three groups of primary care providers – Groups 1, 2 and 3. Health Systems B and C include additional provider groups and are identified as peers to Health System A. A period of 12 months is used to compute historical costs. A total of 153,700 members are attributed to all provider groups affiliated with the three health systems.
For each of the provider groups, total medical and pharmacy costs, including all relevant services delivered to the designated population, are computed for the time period. Historical cost PMPM is then computed. The historical PMPM costs for Health Systems A, B and C are $416, $426, and $457, respectively, and the peer-wide average historical cost PMPM is $431. Health System A appears efficient relative to the other organizations based on historical costs.
The peer historical cost of care PMPM of $431 can then be used as a basis to determine a benchmark cost of care for Health System A. Any further adjustments for future trends and other contractual needs can be applied to this amount.
Using peer costs to establish a benchmark for a provider can have advantages, including recognizing performance levels achieved by other organizations. However, some important considerations remain in using such an approach.
Comparative Risk: Care should be taken to ensure that the underlying health risk of a provider’s designated population is comparable to that of their peer providers. Providers can attract patients of different risk for various reasons, such as practice location, mix of specialization, and affiliations with specific hospitals and facilities. These risk differences will translate into differences in the expected cost of managing patient populations. In many cases, adjustments are required to recognize risk differences when using peer cost of care as the basis for setting a cost benchmark for a provider organization.4
Varied Cost Histories: Different levels of historical costs between provider organizations are another consideration. Provider organizations with costs below the historical peer average will potentially enjoy an immediate financial benefit in terms of their benchmark – in effect, receiving a benchmark above their own past experience and a reward for past performance rather than future improvement. These organizations with historically better cost performance will have a stronger incentive to participate. On the other hand, providers with a prior cost experience above that of their peers will require significant improvements to drive savings beyond their benchmark – only achieving rewards by eliminating the gap between their past performance and the peer average. Such a scenario may negatively impact their willingness to participate in an APM contract.
Figure 2. Population-Based Payment: Global Budget Model Setting a Benchmark Using Historical Provider Cost of Care
If the organizations with historically higher costs decline to participate in the APM, attempt to negotiate a higher benchmark, or participate without improving their performance, the program may end up sharing rewards only with historically lower cost providers. This imbalance will challenge efforts to achieve improvements in the cost of care and may lead to an increase in overall costs.
Transparency Challenges: Using peer costs to establish a benchmark requires a higher level of transparency from individual provider groups, which are in the best position to understand and explain their own experience and how it applies to their populations. Provider groups with concerns about such transparency may enter into extended negotiations and potentially contentious contract discussions in lieu of more collaborative approaches. An effective partnership between payers and providers is an essential element to the success of an APM model.
3. Combining Provider and Peer Cost of Care Benchmarks – a Blended Approach
Some organizations implementing a population-based APM pursue a middle ground when setting a benchmark and blend the results from both the provider and peer cost of care approaches. Such an approach combines the provider and peer cost of care findings and can be achieved in various ways. First, the historical cost of care PMPM for a provider organization and for their peers are calculated, as described above and in Figures 1 and 2.
These calculations are then blended using percentage weights to form the basis for the cost of care benchmark. Such percentages could be applied in a uniform way across all participating provider organizations, applied differentially to historically low-cost versus higher cost organizations, or be the subject of negotiations between payers and providers. Using the examples shown in Figures 1 and 2, the provider historical cost of care was $416 PMPM, while the result for peers was $431 PMPM. A blend using a 50/50 percentage weighting would yield a result of $423.50 as the basis for the cost of care benchmark for that provider organization. The health plan can use the $423.50 PMPM result as the basis for setting the benchmark, or negotiate with the provider organization to achieve a benchmark level at or as close as possible to the provider’s historical cost PMPM ($416). For multi-year contracts, the contracting organization may choose to use different weightings each year and gradually transition from historical provider costs to a more peer-based benchmark over time.
The main benefit of this approach is that it may mitigate some of the downside effects of both the provider cost of care and peer cost of care approaches. Take, for example, a provider with historically higher cost of care. Under the provider cost of care approach, this provider’s prior inefficiencies get built into the baseline benchmark, thus rewarding them for those inefficiencies and allowing them to more easily achieve targets relative to other organizations. Under the peer cost of care approach, this organization would receive a benchmark below its past experience and be challenged to close the gap and gain rewards. Using a blended model, the contract is more likely to generate performance improvement for this organization and to be more acceptable.
Important considerations exist in choosing a blended approach to setting a cost benchmark.
Transparency: Similar to the peer approach, transparency, or lack thereof, can be a challenge with a blended approach. The blended PMPM will be only partially based on the provider’s own experience, and the final calculation will be less transparent (and possibly less acceptable) to providers.
Burden on higher-cost providers: Additionally, for provider organizations that require the greatest level of improvement, the blend may not fully alleviate the burden of the significant short-term improvements required to generate savings – potentially leading to a lower adoption rate for these organizations. Adopting a multi-year strategy and gradually changing the weights used in blending may address these concerns.
Extra negotiations: The blended approach also requires additional negotiations regarding the percentage weights to be used in setting the final benchmark. The results of these negotiations could shift markedly the balance of the intended incentives and the expected savings.
Choosing the Right Approach
Selecting the right approach to calculating a cost of care benchmark for a population-based APM depends on the context for such a program, including the current level of health care costs in the market, a health plan’s market share, the distribution of providers and their negotiating power, and the presence of other past or current payment improvement initiatives. Health plans will also vary in their philosophical preferences for how to reward and reimburse providers. The willingness of employer accounts to support these programs is another consideration.
It is also important to consider the capabilities of each provider organization when setting a financial benchmark. In particular, do organizations have the appropriate infrastructure to drive practice changes to reduce costs while maintaining or improving quality of care? Can they adopt or enhance the patient coordination and population management programs necessary for success? What is the spread in historical cost of care experience across providers? Do physician leaders believe in value-based contracting as a collaborative venture to drive overall performance improvement?
Success is achieved when there is a proper balance between accountability and adoption.
All of these considerations will influence the selection of an appropriate strategy to establish a cost of care benchmark for a population-based APM. As described above, success is achieved when there is a proper balance between accountability and adoption. This balance sets the foundation for a long-term relationship among stakeholders and establishes a contract that creates value for payers, providers, and — most important — for patients.
1 Population-based payment models are often called “global budget” or “total cost of care” models. In all cases, these models involve the management of a defined population with some cost of care benchmark or target mechanism to incent the more efficient use of healthcare resources.
2 The application of these budget factors and trends may differ across models and regions, but at each step the resulting value should be evaluated to assess whether the model’s intended incentives have changed. These factors may include an adjustment for any differences in the underlying risk of the historical population used to compute the benchmark and the population to be covered by the contract. This is especially important for provider organizations with highly mobile members.
3 A related concept is the use of a provider’s own cost experience under a contract to re-base the cost of care benchmark in subsequent years of that contract – for example using Year One contract experience to set a new baseline for Year Two, rather than adjusting the Year One benchmark. Such an approach could diminish incentives for continued improvements. The issues associated with setting and trending annual budgets in multi-year APM contracts will be discussed in greater detail in a separate Qcentive Issue Brief.
4 Assessing member health risk and using measures of risk to adjust financial benchmarks and performance results is a topic of a separate Issues Brief from Qcentive.
5 Setting a benchmark based on average peer performance may be preferred by those payers seeking to drive change across less efficient provider organizations – and are in a position to mandate a transition to such a model.