The Fund supports networks of state health policy decision makers to help identify, inspire, and inform policy leaders.
The Milbank Memorial Fund supports two state leadership programs for legislative and executive branch state government officials committed to improving population health.
The Fund identifies and shares policy ideas and analysis to advance state health leadership, strong primary care, and sustainable health care costs.
Keep up with news and updates from the Milbank Memorial Fund. And read the latest blogs from our thought leaders, including Fund President Christopher F. Koller.
The Fund publishes The Milbank Quarterly, as well as reports, issues briefs, and case studies on topics important to health policy leaders.
The Milbank Memorial Fund is is a foundation that works to improve population health and health equity.
December 10, 2024
Issue Brief
Grace Flaherty
Publication
Dec 2, 2024
Oct 9, 2024
Jul 10, 2024
Back to Publications
Health care costs have been rising faster than inflation for decades, meaning health care has been taking up larger and larger proportions of government, employer, and household budgets and crowding out other critical priorities. To draw attention to the problem of health care affordability and increase systemwide health care cost transparency and accountability, some states are taking action through cost growth target programs.
State health care cost growth target programs aim to control health care spending by setting an annual goal for the maximum rate at which health care costs should increase that is tied to income and/or economic growth. In addition to reporting spending growth at the state, insurer market, and individual insurer levels, states with cost growth target programs report spending growth for large provider organizations, such as health systems and hospitals. Reporting on provider organization cost growth is an important strategy for keeping providers accountable for constraining spending growth. However, accurately attributing spending to these large provider entities can be challenging.
Attribution challenges can have significant downstream impacts, especially in states with cost growth programs that include performance improvement plans and/or financial penalties for providers. If spending is incorrectly attributed, providers may face unfair penalties or be required to implement improvement plans based on inaccurate assessments of their performance. This can undermine trust in the program and create unintended financial strain on health systems that are incorrectly identified as high spenders. Conversely, some providers may avoid deserved accountability. Ensuring accurate attribution is essential to maintaining the integrity and effectiveness of these state programs.
This issue brief describes these attribution challenges and highlights Connecticut’s attribution approach involving the collection and utilization of providers’ tax identification numbers (TINs). We also highlight Massachusetts’s and Oregon’s approaches. All three approaches (Connecticut’s, Massachusetts’s, and Oregon’s) are valid and worthy of consideration by states, but we focus on Connecticut’s approach in this brief because it gives the most concerned parties (provider organizations) the central role in defining their network.
Attribution refers to the process of assigning an insurer enrollee or member to the clinician principally responsible for their care. To obtain provider organizations’ cost growth target data, most state cost growth target programs ask insurers to take four steps: (1) assign spending to members, (2) attribute members to individual clinicians, (3) link those clinicians to the provider organizations for which the state is assessing cost growth target performance, and (4) report providers’ aggregated spending to the state (See Figure 1.). These steps, and step 3 in particular, can present challenges for insurer data submitters and can concern provider organizations whose cost growth is being reported.
Insurers’ challenges include incomplete information on clinician affiliations, complexities arising from changes in TINs, provider acquisitions and other changes in organizational affiliations, and discrepancies between state reporting requirements and total cost of care (TCOC) contracts between insurers and providers.
Most states ask insurer data submitters to attribute spending to a list of provider organizations developed by the state. States typically include large provider entities that can be reasonably expected to influence total health care costs, such as medical groups, health systems, federally qualified health centers, and independent practice associations. Some states identify provider entities by whether they have a TCOC contract. Other states include provider entities deemed large enough to have a TCOC contract, whether or not they have one. These states usually define the provider organizations by name and associated state-assigned identification number.
Unless an insurer holds a TCOC contract with the provider organization, it may not know which clinicians are associated with the provider organization or clinicians’ TINs. (See the sidebar TINs vs NPIs for more details on TINs.) Some large provider organizations, such as federally qualified health centers, may have only one associated TIN through which all services are billed.
Provider organizations such as independent physician associations or health systems with affiliated community networks include many clinicians and practice groups, each potentially with their own TINs. Insurers have their own internal provider rosters for individual providers in their network, but without a provider directory that matches individual clinicians and practice groups with large provider organizations, insurer data submitters in states that define their list of provider organizations must rely on their individual contracting structures to assign providers to large provider entities.
Figure 1. Process for Attributing Spending to Large Provider Entities
Source: Angeles J. Making health care more affordable: a playbook for a state cost growth target. The Peterson-Milbank Program for Sustainable Health Care Costs. https://www.milbank.org/publications/making-health-care-more-affordable-a-playbook-for-implementing-a-state-cost-growth-target/measuring-performance-against-the-target/. Published January 10, 2023. Accessed November 18, 2024.
Attribution challenges are further complicated when a provider organization changes its TINs. If the old TIN still appears in the state’s cost growth target reporting, it can cause confusion for both the state and the provider. Similarly, a provider organization may claim that a particular TIN is not affiliated with the organization or that it is not responsible for the spending attributed to it. A provider organization, for example, might deny responsibility for spending attributed to a particular TIN because of complex relationships between different entities within the organization, such as independent practices or newly acquired groups.
Attribution accuracy can also be affected when a provider organization acquires another organization or clinician group. Following an acquisition, both states and insurers must ensure that post acquisition spending for clinicians affiliated with the acquired organization are reported under the new parent provider organization. To do so, states must keep their list of large provider organizations updated (e.g., remove the provider organization that was acquired and indicate in state data specifications under which parent organization it was subsumed) and insurers must update their records in a timely way.
TINs vs NPIs
Stakeholders may reference both tax identification numbers (TINs) and national provider identifiers (NPIs) when discussing attribution. TINs and NPIs are both identification numbers associated with health care providers, but they serve different purposes. A TIN is a nine-digit number used by the Internal Revenue Service to track tax obligations and payments. TINs are issued to provider organizations for tax reporting and financial identification purposes. TINs usually represent business entities rather than individual providers. For example, a large health care system or a multispecialty group practice would have a TIN that encompasses all the practices and locations operating under that entity.
An NPI is a unique 10-digit number issued by the Centers for Medicare and Medicaid Services to health care providers. NPIs are used to identify individual clinicians and organizations in health care transactions, such as billing, claims processing, and electronic health records. There are two types of NPIs. Type 1 NPIs are assigned to individual clinicians, such as physicians, dentists, nurses, and other solo practitioners. Each individual provider has a unique NPI that identifies them personally. Type 2 NPIs are assigned to organizations, such as hospitals, group practices, nursing homes, and other health care entities.
States have focused on identifying the TINs associated with large provider identities, rather than NPIs, because insurers typically rely on TINs to identify providers for contracting and pull data for their cost growth target data submissions.
Insurers are accustomed to sending reports to the provider organizations with which they contract in alignment with their contractual TCOC arrangements. However, cost growth target programs request that all spending for an insurer’s members be attributed to large provider entities — and not just spending for members attributed to providers in accordance with a TCOC arrangement. (The only exception is for members whom insurers are unable to attribute to a specific clinician, usually due to a lack of utilization.) This poses three main challenges for the purposes of cost growth target attribution:
Provider roster management refers to the process of maintaining an up-to-date list of health care clinicians in an insurer’s network and their associated provider organization. Provider roster management is typically the responsibility of both the insurer and the provider organization. The insurer is responsible for maintaining an accurate, up-to-date list of clinicians in their network, while the provider organization must regularly update the insurer with any changes (e.g., new providers joining, providers leaving, changes in specialties or locations). Insurers are legally required to maintain provider rosters and meet network adequacy standards (e.g., time and distance criteria to ensure that beneficiaries can access care within reasonable geographic boundaries). Beyond these legal requirements, effective provider roster management is crucial for accurate billing and patient referrals. Provider roster management also impacts the extent to which insurers can accurately attribute spend ing to a prescribed list of provider organizations in cost growth target data submissions.
Provider roster management is challenging for both provider organizations and insurers. From a provider organization’s perspective, informing insurers of frequent changes can be time-consuming and administratively burdensome. From the insurer’s perspective, there is no standardized process for maintaining a provider roster, and it can be hard for insurers to handle the variation in how provider organizations submit roster updates. Insurers and provider organizations commonly ascribe blame to one another for inaccuracies in insurer-maintained rosters.
Delays in updates to provider rosters and inaccuracies in provider organization rosters may lead to discrepancies in insurers’ cost growth target data submissions. No state has attempted to improve provider organization roster management as a part of its cost growth target program, but it could be an area of future work. State-developed provider registries, such as the Massachusetts Health Quality Partners’ Massachusetts Provider Database (MPD), may help with identifying relationships between individual providers and provider organizations; however, registries like the MPD are typically only updated on an annual basis and therefore are less useful for day-to-day roster management between insurers and provider organizations.
Table 1. Potential Explanations for Differences between a Provider Organization’s Cost Growth Target Member Months and TCOC Member Months
The Connecticut Office of Health Strategy (OHS) has faced provider attribution challenges since it began collecting cost growth target data from insurers in 2021. The insurer data submitters in Connecticut reported difficulty attributing both fee-for-service and TCOC spending to Connecticut’s list of large provider entities, which OHS calls Advanced Networks. Similarly, the Advanced Networks questioned why the member months and spending in their cost growth target performance did not align with those reported to them by insurers in TCOC reports.
In 2024, OHS convened an Attribution Work Group comprising insurer data submitters and Advanced Network representatives to discuss the attribution challenges and potential solutions. The Attribution Work Group process led OHS to implement two changes to its cost growth target data submission methodology, starting with the collection of 2022–2023 spending data. The changes seek to address two of the challenges defined earlier in this brief: (1) identifying individual clinicians and practice groups associated with a large provider organization, and (2) bringing greater transparency to differences with existing TCOC reporting. These changes were informed by Oregon’s and Massachusetts’s attribution methodologies (see sidebar). The changes included:
In Connecticut’s updated process, Advanced Networks submit their TINs to OHS, who compiles the TINs into one list and distributes them to payers to use to attribute spending to the Advanced Networks in their cost growth benchmark data submissions (see Figure 2).
OHS has identified two obstacles so far in its new TIN collection effort:
Figure 2. Connecticut’s Revised Provider Attribution Process
Oregon’s and Massachusetts’s Attribution Approach
Massachusetts and Oregon each take a unique approach to attributing spending to provider organizations in their cost growth target programs. Both states ask insurers to attribute and report spending for provider organizations using a hierarchal method with three tiers: members choosing a primary care provider (tier 1), members assigned to a primary care provider through insurer-provider contracts (tier 2), and members assigned to a primary care provider based on the insurer’s own utilization-based attribution methodology (tier 3). Unattributed members are reported separately.
The two states differ in their approach to identifying the provider organizations. Oregon requires insurers to self-identify provider organizations (i.e., Oregon does not provide a list of defined provider organizations), requires insurers to submit known TINs for each provider to which a member is attributed, and separately asks insurers to report all known TINs for each large provider organization. Insurers first attribute members to a primary care provider, then roll up member spending to the highest provider organization level (e.g., hospital, health system, medical group, federally qualified health center). Insurers report the provider organization’s spending (for a given year, line of business, and attribution tier) in one row of the data submission template. Additionally, insurers must list all associated TINs for that provider organization in a separate tab, which the state later uses to combine all insurer files. In this way, Oregon’s approach puts the burden on the state (rather than on insurers) to clean and match provider organizations’ spending for analysis, public reporting, and accountability.
In contrast, in Massachusetts, insurers report spending by physician group and local practice group using state-defined identifiers. Massachusetts instructs insurers to report physician group data based on their individual contracting structures with providers.
Attribution of cost growth performance to provider organizations has proven challenging in state cost growth target programs due to the complexity of accurately linking their members’ individual primary care providers to larger provider organizations. The primary problem is that insurers often lack comprehensive information regarding which clinicians are associated with large provider entities, especially when they do not hold TCOC contracts with those organizations. Without access to essential data such as TINs, insurers struggle to correctly attribute spending to the appropriate provider organizations as defined by the state.
Changes in TINs and acquisitions or affiliations introduce more complexity to attribution efforts. Additionally, discrepancies between cost growth target reporting and TCOC arrangements further complicate attribution. Insurers may report different member months and spending figures for the same provider organizations under cost growth targets versus TCOC arrangements, due to variations in contract terms, employer opt-outs, and the exclusion of self-insured spending. Provider roster management adds another layer of complexity. Insurers depend on up-to-date provider rosters to accurately attribute spending, but inconsistencies and delays in roster updates can result in inaccurate data submissions. The lack of a standardized process for maintaining and communicating provider rosters exacerbates this issue.
Connecticut’s collection and utilization of TINs and hierarchical attribution is a step toward addressing some, but not all, of these challenges. By requiring TINs from Advanced Networks and requesting detailed attribution data, Connecticut is working to improve the accuracy and transparency of attribution in its cost growth target data. The state’s efforts highlight the critical need for collaboration between insurers and provider organizations to overcome the inherent complexities of attribution in cost growth target programs. One key challenge that remains unaddressed in Connecticut’s approach is timely and accurate roster management, by both providers and insurers.
Understanding and addressing these attribution challenges is essential for states to successfully implement their cost growth target programs and ensure that health care spending growth is accurately monitored at the provider organization level. By addressing attribution challenges, states can more confidently hold provider organizations accountable for exceeding spending targets, ensuring that any required performance improvement plans or penalties are based on correct data. Accurate attribution also helps reduce disputes over which organizations are responsible for spending and increases trust and transparency in cost growth target programs.
