Affordable, quality health care. For everyone.

Expanding the PACE Model of Care to High-Need, High-Cost Populations



  • Issue: High-need populations benefit from integrated care such as that offered by the Program of All-Inclusive Care for the Elderly (PACE) model. Understanding the diversity of high-need populations and where they are located can guide optimal expansion of this model.
  • Goals: Identify high-need, high-cost (HNHC) populations that may benefit from an expansion of the PACE model and determine the size and geographic distribution of these populations.
  • Methods: Conduct a literature review and form an expert advisory panel to identify key HNHC populations. Analyze Medicare and Medicaid claims data to capture the size and distribution of these populations, and the extent to which they are particularly high cost.
  • Key Findings and Conclusions: The greatest opportunity for expansion is to the population currently eligible for PACE programs. Significant opportunities exist to serve other high-need populations, some of which are particularly high-cost, and some of which may require changing how PACE is structured. Other high-need populations also could benefit from PACE if its scope were expanded and if reimbursement rates were appropriately structured to recognize variation in costs.


In recent years, the term high need (HN) has been used to describe people who have conditions that require significant levels of health care; high need, high cost (HNHC) has described those who both have the most need and make the most use of that care. There are several populations that may be identified as HNHC, including people with multiple chronic health conditions, functional limitations, and behavioral health needs.

Meeting the needs of these HNHC groups requires three steps: 1) understanding the diversity of the populations; 2) identifying integrated care programs that can best meet their needs at lower cost; and 3) spreading adoption of those integrated care programs.1 Integrated care programs provide efficient coordination of medical and mental health care as well as long-term services and supports (LTSS) that these HNHC populations may need, in ways that may be more person-centered than traditional, siloed care approaches.2

The Program of All-Inclusive Care for the Elderly (PACE) is one of the oldest and most successful models of integrating services for high-need people with acute and LTSS needs. Several studies and evaluations have demonstrated the positive effects of enrolling in PACE. Such benefits include reductions in hospitalization, rehospitalization, and emergency department use; reductions in long-term nursing facility placements; reductions in mortality; and lower rates of functional decline and better reported health status and quality of life.3

To date, PACE has been restricted to people age 55 and older who require a nursing home level of care. The PACE Innovation Act of 2015 enables the Centers for Medicare and Medicaid Services (CMS) to authorize demonstrations of PACE programs to serve other HN populations.

Using a literature review and input from a technical advisory group, we identified five high-need populations that may benefit from the types of services offered by PACE. We drew from Medicaid claims data to identify the size of these potential target populations by state. Subsets of the five populations were determined to be both high need and high cost. Understanding the size and geographic distribution of these HN and HNHC populations can support policymakers and providers in focusing on those localities with the greatest potential to benefit from a PACE expansion or other targeted services.


Which Populations Could Benefit from PACE Expansion?

An advisory group identified five HN populations and subsets of those populations that are HNHC as likely to benefit from integrated care programs such as PACE.

In addition to those currently eligible for PACE, the populations include:

  • younger adults (ages 21–54) with developmental disabilities (DD) and comorbidities
  • younger adults (ages 21–54) with physical disabilities (PD) and comorbidities
  • adults with behavioral health conditions and comorbidities
  • adults with end-stage renal disease (ESRD), comorbidity, and functional impairment.

What Are the Opportunities for PACE Expansion?

PACE can grow through three means:

  1. Scale: Increasing the number of people served by current PACE organizations in their current communities.
  2. Spread: Increasing the number of PACE organizations and number of communities served by the current PACE model.
  3. Scope: Expanding the range of populations that PACE can serve.

Our analyses found that the greatest opportunity for PACE growth is through scale. In 2012, PACE served approximately 25,000 people. Since then, PACE has more than doubled in size.4

While adding new PACE programs and spreading them to new communities is another strategy to grow PACE, such growth may be limited by state policy. Some states restrict the number of people who may be enrolled in PACE, and some limit the number of PACE programs that may be established. Other states do not offer PACE. As of March 2020, PACE programs could be found in 31 states.5 Expansion of PACE to states that do not currently offer such programs would require changes to state policy.

Expanding the scope of populations served by PACE is another growth strategy (Exhibit 1). Younger adults (those ages 21 to 54) would be an entirely new population for PACE. Those with physical disabilities may share similar physical needs with the current PACE population. However, younger adults with developmental disabilities are less likely to have the same needs as older adults currently served by PACE.

Other HN populations identified, those with behavioral health needs and those with ESRD, include adults of all ages. Many of those age 55 and older may be eligible for PACE as it currently exists.

What Are the Cost Implications of Expanding to Different Populations?

All five populations are high need and may have high costs as well.

In our analysis, we defined high-cost populations as those whose total annual costs, either Medicare or Medicaid, were in the top 10 percent for two consecutive years. Modest shares of the current PACE population had high costs (Exhibit 2). The same was true for groups with behavioral health needs and younger adults with physical disabilities. However, nearly 40 percent of the younger population with developmental disabilities and almost 70 percent of those with ESRD had high costs as well as high needs.

Because the costs associated with each group differ, expansion of PACE will require careful development of capitated payment rates, with appropriate risk adjustment that recognizes the variation in need within those groups. This will ensure adequate reimbursement for each population.

Cost implications of expansion vary significantly by state. The share of high-need populations that are also high cost ranged from 6 percent in New Mexico to 24 percent in New York (Exhibit 3 and Appendix Table 1). Expansion through the creation of new PACE programs, whether they are to serve the currently eligible or new populations, must consider both the size of the target population and its cost distribution.

Discussion and Policy Implications

The Centers for Medicare and Medicaid Services (CMS) and many states have prioritized improving care and reducing costs of care for high-need, high-cost populations by supporting integrated care models that focus on coordinating the delivery of primary, acute, and behavioral health care as well as long-term services and supports. The PACE Innovation Act is intended to support the growth of PACE by extending its scope to new populations. Significant opportunities for PACE expansion also exist by scaling its current programs and spreading programs to new locations. However, the ability to grow through scale and spread is limited in some states, which currently impose limits on PACE enrollment or restrict the number of PACE organizations in the state. Such policies will need to change before PACE can grow.

While expanding the scope of PACE to new populations offers opportunities, it may bring new challenges as well. For example, people with behavioral health needs, which include mental health and substance use disorders, may require specific types of support that are not a current focus of PACE. Additionally, the service needs of the younger and older adult populations with physical disabilities may appear similar, but younger adults may be more interested in supports for educational and employment goals, and for community integration. Providing such supports could require a significant shift from PACE’s current adult day center–based model of support.

Adequate risk adjustment of reimbursement rates will be needed to properly capture the costs of high-need, high-cost populations. This is especially important when considering expanding PACE to some adults with ESRD and to younger adults with developmental disabilities. These two groups have a significant share of the population who are high cost. They also are the two smallest populations. Together, those factors suggest that these groups may be particularly challenging as a focus for PACE expansion.


PACE has proven to be a strong model for the population it currently serves, and there are significant opportunities to expand its scale and scope to others who are currently eligible. PACE also can grow by extending, or spreading, coverage to new populations, but doing so may require significant shifts in the PACE model of care. Such shifts may be needed to address the particular care needs of targeted beneficiaries in a manner that aligns with their preferences and goals.

Regardless of which beneficiaries they choose to target, policymakers and providers will need to consider that the size of their high-need populations and their cost distributions vary by state.

How We Conducted This Study

This study identified five high-need populations that could be well served by PACE. A variety of qualitative and quantitative approaches were used to better understand the different types of high-need (HN) populations and how they are geographically distributed.

Identifying Populations of Interest

Several HN populations were identified through an environmental scan. An advisory group reviewed those populations and recommended five that were most likely to benefit from integrated care programs such as PACE. These groups were defined based on combinations of age, disability, and chronic conditions. The advisory group agreed that high-cost subgroups should be identified based on two consecutive years of being high cost (i.e., with costs in the top 10% of Medicare or Medicaid).

In addition to the HN population that is currently eligible for PACE, the advisory group recommended two groups of younger adults, and two groups of adults regardless of their age. The younger groups represent new populations for PACE, as they are not eligible currently on the basis of age. The younger group with physical disabilities may be similar in needs for support to the older population, but the group with developmental disabilities may represent a new target with different needs. Expansion to the groups with behavioral health needs and with end-stage renal disease (ESRD) would mean a targeted outreach to specific populations. Many people with behavioral health needs or ESRD may be currently eligible for PACE, and some may already be served.

Qualitative Methods

To identify HN and high-need, high-cost (HNHC) populations, we conducted an environmental scan that included information collected from publicly available, peer-reviewed literature and grey literature. We also reviewed all comments submitted in response to the PACE Innovation Act Request for Information (RFI) and conducted interviews with key advisors from current PACE programs. An advisory group recommended five populations that they believed to be particularly suitable for services under the PACE model.

Quantitative Methods

Analyses of the 2012 Medicare–Medicaid Linked Enrollee Analytic Data Source (MMLEADS) were conducted to determine the size and distribution of HN and HNHC populations by state.


MMLEADS data include comprehensive information about the eligibility, enrollment, service utilization, expenditures, chronic health conditions, and disabilities (including 27 chronic conditions and 25 mental health, tobacco, physical, and mental health disability conditions) of people who are enrolled only in Medicare, enrolled only in Medicaid, or who are enrolled in both Medicare and Medicaid. For those who are enrolled only in Medicaid, MMLEADS is limited to those who were eligible because of disabilities. The data include all qualifying individuals during 2012.

Data exclude four states (Colorado, Idaho, Kansas, and Rhode Island) because of incomplete Medicaid information for this time period.

Identifying the Populations

MMLEADS data were used to identify individuals in each of the HN populations selected for study. The authors identified each of these populations using a two-step approach similar to that used by Joynt et al.6 This approach began by identifying HN groups based on chronic health conditions and disabilities, either alone or in combination with others.

Additionally, where appropriate, the authors used data on utilization of long-term services and supports (LTSS) to identify people with needs for functional supports. These data are available only for those who are Medicaid-eligible. Functional impairment data also are available only for those who are Medicaid-eligible. Predictive models, created using data MMLEADS data linked with information from the Medicare Current Beneficiary Survey, were applied to the full MMLEADS data set to identify those people eligible for Medicare only who were likely to meet the functional eligibility criteria. High-cost populations were those whose Medicare or Medicaid expenditures fell within the top 10 percent for both 2011 and 2012.

Five high-need populations that may benefit from access to PACE

  • Currently PACE-eligible
  • Younger adults with developmental disabilities (DD) and comorbidities
  • Younger adults with physical disabilities (PD) and comorbidities
  • Adults with behavioral health conditions and comorbidities
  • Adults with end-stage renal disease (ESRD), comorbidity, and functional impairment



The authors acknowledge the participation and input of the National PACE Association (NPA). The views expressed here are those of the authors only and do not necessarily reflect those of the Commonwealth Fund or NPA.


1. David Blumenthal et al., “Caring for High-Need, High-Cost Patients — An Urgent Priority,” New England Journal of Medicine 375, no. 10 (Sept. 8, 2016): 909–11.

2. Peter Long et al., Effective Care for High-Need Patients: Opportunities for Improving Outcomes, Value, and Health (National Academy of Medicine, 2017).

3. Pinka Chatterji et al., Evaluation of the Program of All-Inclusive Care for the Elderly (PACE) Demonstration: The Impact of PACE on Participant Outcomes (Health Care Financing Administration, July 1998); Pamela Nadash, “Two Models of Managed Long-Term Care: Comparing PACE with a Medicaid-Only Plan,” Gerontologist 44, no. 5 (Oct. 2004): 644–54; Robert L. Kane et al., “Variations on a Theme Called PACE,” Journal of Gerontology: Series A 61, no. 7 (July 2006): 689–93; Louise A. Meret-Hanke, “Effects of the Program of All-Inclusive Care for the Elderly on Hospital Use,” Gerontologist 51, no. 66 (Dec. 2011): 774–85; Darryl Wieland et al., “Hospitalization in the Program of All-Inclusive Care for the Elderly (PACE): Rates, Concomitants, and Predictors,” Journal of the American Geriatrics Society 48, no. 11 (Nov. 2000): 1373–80; Micah Segelman et al., “Hospitalizations in the Program of All-Inclusive Care for the Elderly,” Journal of the American Geriatrics Society 62, no. 2 (Feb. 2014): 320–24; Jody Beauchamp et al., The Effect of the Program of All-Inclusive Care for the Elderly (PACE) on Quality (Centers for Medicare and Medicaid Services, Feb. 2008); Arkadipta Ghosh, Cara Orfield, and Robert Schmitz, Evaluating PACE: A Review of the Literature (U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, Jan. 2014); and Micah Segelman et al., “Transitioning from Community-Based to Institutional Long-Term Care: Comparing 1915(c) Waiver and PACE Enrollees,” Gerontologist 57, no. 2 (Apr. 2017): 300–8.

4. Ghosh, Orfield, and Schmitz, Evaluating PACE, 2014; and National PACE Association, PACE in the States (NPA, Apr. 2020).

5. NPA, PACE in the States, 2020.

6. Karen E. Joynt et al., “Segmenting High-Cost Medicare Patients into Potentially Actionable Cohorts,” Healthcare 5, no. 1–2 (Mar. 2017): 62–67.

Publication Details

Publication Date: October 22, 2020
Contact: Sara Karon, Senior Health Policy Researcher, Aging, Disability, and Long-Term Care Program, RTI International

Source: Sara Karon et al., Expanding the PACE Model of Care to High-Need, High-Cost Populations (Commonwealth Fund, Oct. 2020).


Senior Health Policy Researcher, Aging, Disability, and Long-Term Care Program, RTI International
Research Public Health Analyst, Aging, Disability, and Long-Term Care Program, RTI International
Public Health Analyst, Aging, Disability, and Long-Term Care Program, RTI International
Health Policy Researcher, Aging, Disability, and Long-Term Care Program, RTI International