- Issue: With an incentive to provide high-quality care while controlling costs, accountable care organizations (ACOs) may focus on patients who require the most resources and are most at risk for encountering serious problems with their care. Understanding how ACOs approach care for complex patients requires examination of their organizational strategies, contracting details, and leadership structures.
- Goals: Describe the specific strategies employed by ACOs that have comprehensive care management programs and processes for complex patients.
- Methods: Cross-sectional descriptive analysis of the fourth wave of the National Survey of ACOs.
- Key Findings: Most ACOs report having comprehensive chronic care management processes or programs in place to manage people with complex needs. More labor-intensive interventions, however, are rare. Few ACOs report having advanced programs for engaging patients, in-home visits after hospital discharge, or evidence-based services for patients needing mental health or addiction treatment.
- Conclusion: While ACOs have increased their efforts to target populations with complex care needs, there is a need for more varied approaches to improving care delivery.
People with complex care needs account for nearly one-fifth of all health care spending, even though they comprise only 1 percent of patients.1 These are individuals with multiple chronic conditions or functional limitations; people whose conditions carry significant nonmedical needs; and frail older adults. Their mental health, physical health, and social needs require coordination across numerous providers, family caregivers, and social service agencies. Traditional fee-for-service payment models rarely reimburse for the coordination, care management, and team-based care that this population needs.
Emerging models of health care payment and delivery, such as ACOs, present an opportunity to improve quality of care and lower costs for people with complex needs and a range of medical and social issues. Unlike fee-for-service, ACO contracts award providers with bonus payments tied to cost and quality performance for their assigned patients. So far, ACOs have achieved modest reductions in health care spending (with Pioneer and Next Generation ACOs producing more promising results than ACOs in the Medicare Shared Savings Program), and care quality has improved without raising costs — an increase in the value of health care.2
Given their incentives to reduce cost and improve quality, ACOs often employ care management programs that follow evidence-based strategies for increasing the value of care delivered to people with complex needs. These strategies include:
- identifying people who are at high risk for adverse clinical events (often referred to as risk stratification)
- separating high-risk patients into subgroups with common needs (segmentation)
- improving care transitions across settings
- engaging individuals and their families in care decisions
- using programs that help patients address chronic illness.3
Understanding the variation in ACOs’ use of these strategies is a useful first step in determining a standard of care for this population.
To better gauge ACO efforts to manage care of people with complex needs, we analyzed responses to the fourth wave of the National Survey of ACOs, fielded in 2017–18. Wave 4 included several questions regarding the use of evidence-based approaches to managing care for this population, including risk stratification, segmentation, improvement in care transitions, engagement of patients and families in care decisions, and chronic condition management. Based on responses to those questions, we assessed the extent to which organizations have adopted approaches with special relevance for complex populations. We also created an index of ACOs’ ability to simultaneously implement these approaches. For a full list of survey questions used in this analysis, see Appendix Table 1.
Care management has emerged as a leading evidence-based approach to meeting the multifaceted needs of people requiring complex care.4 The National Survey of ACOs allows us to assess the landscape of chronic care management programs and processes and identify where ACOs may need to pay more attention to the strategies incorporated in care management programs.
To characterize the current state of evidence-based approaches for the care of people with complex needs, we grouped ACOs by their overall use of care management processes and programs — that is, by whether they reported having either comprehensive programs or, alternatively, few or no programs. We then analyzed ACOs’ uptake of specific evidence-based approaches for the care of patients with complex needs.
In the survey, most ACOs (63%) report having comprehensive care management programs and processes (referred to here as simply “care management programs”) in place, based on their response of 7 or higher on a 9-point scale (Exhibit 1). In contrast, 33 percent report they have only “some” care management programs in place, while 4 percent say they have “few or no” such programs.
Below, we stratify ACOs according to their implementation of care management programs to compare approaches to three related strategies important for the care of complex populations: patient identification, patient engagement, and care transitions. This information will help ACO leaders, health care delivery and population health experts, and policymakers understand the range of activities that could be incorporated into care management programs to provide better care for people with complex needs.
Identification and Engagement
People with complex needs are heterogeneous in terms of diagnoses, nonmedical needs, and functional status.5 Health care organizations often try to identify patients at highest risk for poor outcomes so they can deploy limited resources where they are likely to have the greatest impact. By identifying people with complex needs (often through analysis of electronic medical records), providers can better deliver targeted, proactive care. Over half (52%) of ACOs reporting comprehensive care management programs state that they have an advanced system in place to identify and target patients using predictive risk stratification for their attributed patients (Exhibit 2), compared with only 29 percent of other ACOs.
A related but distinct approach to tailoring scarce resources for complex populations involves patient segmentation, or the grouping of high-risk patients along similar clinical or nonclinical needs based on administrative data.6 Such segmentation is used to design more effective interventions.7 Most survey respondents (66%), regardless of their ACO’s care management score, report that their organization segments high-risk patients into subgroups based on common needs, such as frailty, mental illness, or combination of chronic conditions.
Engaging and activating patients has been shown to decrease long-term health care costs and to improve health care experiences.8 Yet even among ACOs with comprehensive care management programs, clinician training in activation and engagement is limited, as are programs to encourage patient involvement in care decisions. Only two of every five (38%) ACOs with comprehensive care management programs report that most or all their affiliated clinicians are trained in patient activation and engagement methods. And just 38 percent of these ACOs have advanced programs in place for clinicians to encourage ACO patients to be actively involved in decisions involving their care and in self-managing their conditions. The percentages are even lower for other ACOs, only 16 percent of which respond that most or all their clinicians are trained in patient engagement and 15 percent of which report having comprehensive programs to encourage patient involvement.
The National Survey of ACOs asked organizations about seven components of evidence-based interventions designed to reduce risk of readmission for hospitalized individuals who are transitioning to home or a postacute care facility. ACO-reported use of interventions requiring one-on-one interaction between providers and patients is low. Despite being a prominent component of evidence-based care transition models, in-home follow-up is the least commonly used strategy for patients undergoing a care transition.9 Only 21 percent of surveyed ACOs with comprehensive care management programs report that most or all patients receive an in-home follow-up visit within 72 hours of discharge. And only 11 percent of the remaining ACOs said that most or all patients receive this in-home visit.
Use of similarly labor-intensive programs is notably lower in ACOs that do not report comprehensive care management programs. For example, use of a postdischarge care manager or health coach is twice as common in ACOs with comprehensive care management programs (62%) as compared with other ACOs (31%).
As shown in Exhibit 3, other strategies to reduce risk of hospital readmission — particularly processes facilitated through technology or automation — were more common in ACOs.
Organizational Characteristics of ACOs Reporting Comprehensive Chronic Care Management Approaches
The organizational characteristics included in the survey, including leadership structure and contracts with payers, vary according to ACOs’ self-reported level of care management. The majority of ACOs in both groups report having a Medicare contract: 82 percent of ACOs with comprehensive care management programs versus 85 percent for other ACOs. However, it is slightly more common for ACOs with comprehensive care management programs to have Medicaid ACO contracts compared to other ACOs (25% vs. 20%). A similar gap exists for commercial ACO contracts (75% vs. 67%).
There are only modest differences between the leadership structures of ACOs with comprehensive care management programs and ACOs with no, few, or some care management programs. Fifty-two percent of ACOs in the former group are led by physicians, 32 percent are jointly led by hospitals and physicians, 9 percent are hospital-led, and 8 percent have other leadership arrangements such as coalitions or state, regional, or county organizations. The leadership structure of ACOs in the latter group are 56 percent physician-led, 25 percent jointly led, 9 percent hospital-led, and 10 percent have other leadership arrangements.
More than half of people with complex needs are estimated to have significant behavioral health comorbidities, suggesting that behavioral health services should be integrated with their primary care.10 In fact, there is growing evidence that effective integration of these services into primary care can improve behavioral health outcomes. These strategies, broadly labeled “collaborative care,” are similar to strategies used to improve care transitions. For example, collaborative care programs often employ care managers to address medical and nonmedical needs; have a consulting mental health clinician; and use a symptom registry to track mental health symptoms.11 In the survey, ACOs with comprehensive care management programs are more likely to use these strategies than ACOs without them, particularly with regard to care managers for nonmedical needs (75% vs. 53%) (Exhibit 4). However, less than 30 percent of NSACO respondents report using registries to track mental health symptoms, even though these registries are considered a critical component of evidence-based collaborative care models.
Care Across Multiple Domains
To better assess the simultaneous uptake of three domains above — patient identification, patient engagement, and care transitions — we formed a measure based on three questions, one from each domain (Exhibit 5). We computed how many ACOs reported advanced approaches for three, two, one, or none of the domains.
ACOs reporting comprehensive care management programs more frequently have advanced approaches in one or more of the three domains than other ACOs. However, 17 percent of these ACOs report no advanced approach for patient identification, patient engagement or care transitions, and only 21 percent have evidence-based strategies in all three areas. More than half (55%) of ACOs that do not report having comprehensive care management programs indicate they have zero advanced approaches to patient identification, patient engagement, or care transitions.
Discussion and Policy Implications
More attention is being paid to improving the quality of care for people with complex needs while slowing growth in their medical costs. The adoption of evidence-based strategies to care for this population, however, varies widely across ACOs. Even among ACOs that have comprehensive care management programs, relatively few have in place multiple evidence-based strategies to enable those programs to succeed. And while the majority of ACOs identify and segment patients with complex needs, these strategies are more common in ACOs that also report comprehensive care management programs. The majority of ACOs have room to increase or enhance their programs to support patients with complex needs.
While there is wide variation in the evidence-based care received by people with complex needs, processes that incorporate one-on-one interaction are notably less common than other processes. Only 30 percent of ACOs report that most or all clinicians receive training and have processes available to them to encourage patient engagement and activation.
For people with complex physical conditions, co-occurring behavioral health needs require additional integration of behavioral health services into other general medical care.12 Despite the known behavioral health challenges of people with chronic conditions, ACOs incorporate collaborative care strategies unevenly. This implies that ACOs perceive their care management programs as comprehensive, even when their efforts to integrate behavioral health services with other care are modest. The findings above suggest there is opportunity for further collaboration between behavioral health providers, social service agencies, health systems, and payers to more fully address the care needs for complex patients.
Finally, while research suggests that physician-led ACOs are more likely to generate spending reductions to achieve shared savings compared with other ACOs,13 organizations with comprehensive care management programs are not more likely to be physician-led than ACOs with less comprehensive care management programs.
Although the majority of ACOs report in the survey that they have comprehensive chronic care management programs and processes for people with complex needs, there are several evidence-based strategies in which ACOs are not investing. There are many opportunities for ACOs to increase uptake of evidence-based strategies to address the needs of complex populations. Additional research is needed, however, to understand the value of more labor-intensive and costly programs, such as those related to patient engagement, one-on-one care transitions, and integration of physical and behavioral health services, as well as the outcomes associated with these programs. It is also important to learn how ACOs can accelerate their adoption of proven strategies.
How We Conducted This Study
The most recent National Survey of ACOs was conducted by SSRS from July 20, 2017, to February 15, 2018. The survey was completed online by most respondents (77%). A paper survey with a subset of questions also was provided to potential respondents in December 2017 and again in January 2018. After screening out ineligible and overlapping organizations, our sample included an estimated 862 ACOs; we contacted an average of 3.2 potential respondents at each organization. Our outreach methods included email, phone calls, and physical mailings. About 55 percent of the sample returned a survey and 48 percent completed at least half of the core survey questions. ACOs with a Medicare ACO contract had a 69 percent response rate while ACOs without a Medicare contract had a 36 percent response rate.
This survey is the fourth wave of the NSACO. We identified our sample through multiple sources including Centers for Medicare and Medicaid Services data, internet data collection, professional networking, and information from Leavitt Partners. Among the wave 4 survey responses, 394 ACOs provided data for all variables related to this report. Respondents typically had leadership roles in the ACO including ACO Executive Director (29%), ACO Vice President (7%), ACO Chief Executive Officer (15%), ACO Medical Director (9%) and Other (e.g., Chief Operating Officer, ACO Director, and Account Manager) (41%).
The NSACO variables used for this report:
|To what extent are chronic care management processes and programs in place to manage patients with high-need, high-cost chronic illnesses?||9-point Likert scale
|For patients attributed to the ACO, to what extent is a system in place for predictive risk stratification?||9-point Likert scale
|Do you segment high-risk patients into subgroups based on common needs (e.g., frailty, mental illness, similar combinations of chronic conditions)?||
|To what extent are clinicians trained in patient activation and engagement methods and techniques (e.g., two-way communication, motivational interviewing, etc.)?||9-point Likert scale
|To what extent are processes in place for clinicians to encourage ACO patients to be actively involved in decisions involving their care and self–management of their conditions?||9-point Likert scale
|For how many of your ACO-attributed hospitalized patients undergoing a care transition to home or a post-acute care facility receive the following services to reduce the risk of readmission?
|Medicare contract||Based on screeners|
|Medicaid contract||Based on screeners|
|Commercial contract||Based on screeners|
|Which of the following best describes the leadership structure of your ACO?||
|Do any providers in your ACO use the following strategies to integrate primary care and treatment for depression and/or anxiety?
|To what extent are systems in place to assure smooth transitions of care across all practice settings including hospitals, long–term care, home care, adult day care, and community–based health and social services as needed?||9-point Likert scale
Note: Except for variables based on 9-point Likert scales, “seen but skipped” and “don’t know” responses were collapsed into the “No” or “None” categories.
Care Management Across Domains
To summarize ACO activity in each of the domains discussed above, we summarized answers to three questions. These three questions were selected to allow comparisons between wave 4 and prior NSACO survey waves in future work: a) “For patients attributed to the ACO, to what extent is a system in place for predictive risk stratification?” b) “To what extent are systems in place to assure smooth transitions of care across all practice settings including hospitals, long–term care, home care, adult day care, and community–based health and social services as needed?” and c) “To what extent are processes in place for clinicians to encourage ACO patients to be actively involved in decisions involving their care and self–management of their conditions?” For each of these three questions, a response of 1–6 was assigned 0 points and a response of 7–9 was assigned 1 point, yielding a range of 0 to 3 possible points per ACO.
Wave 4 of the National Survey of ACOs was supported by the Commonwealth Fund (Grant no. 20160616), the National Institute of Mental Health of the National Institutes of Health (Grant no. R01MH109531), and the California Health Care Foundation (Grant no. 20249). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
1. Susan L. Hayes et al., High-Need, High-Cost Patients: Who Are They and How Do They Use Health Care? A Population-Based Comparison of Demographics, Health Care Use, and Expenditures (Commonwealth Fund, Aug. 2016); and Peter Long et al., eds., Effective Care for High-Need Patients: Opportunities for Improving Outcomes, Value, and Health (National Academy of Medicine, 2017).
2. David Muhlestein et al., “Recent Progress in the Value Journey: Growth of ACOs and Value-Based Payment Models in 2018,” Health Affairs Blog, Aug. 14, 2018.
3. Hayes et al., High-Need, 2016; Long et al., Effective Care, 2017; and Center for Health Care Strategies, Programs Focusing on High-Need, High-Cost Populations, fact sheet (CHCS, Apr. 2016).
4. Christine Vogeli et al., “Multiple Chronic Conditions: Prevalence, Health Consequences, and Implications for Quality, Care Management, and Costs,” Journal of General Internal Medicine 22, Suppl. 3 (Dec. 2007): 391–95; and Timothy W. Farrell et al., Care Management: Implications for Medical Practice, Health Policy, and Health Services Research (Econometrica, for the Agency for Healthcare Research and Quality, Feb. 2015).
5. Karen E. Joynt et al., “Segmenting High-Cost Medicare Patients into Potentially Actionable Cohorts,” Healthcare 5, no. 1–2 (Mar. 2017): 62–67.
6. Hayes et al., High-Need, 2016; and Joynt et al., “Segmenting High-Cost,” 2017.
7. 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.
8. Judith H. Hibbard and Jessica Greene, “What the Evidence Shows About Patient Activation: Better Health Outcomes and Care Experiences; Fewer Data on Costs,” Health Affairs 32, no. 2 (Feb. 2013): 207–14; Jessica Greene et al., “When Patient Activation Levels Change, Health Outcomes and Costs Change, Too,” Health Affairs 34, no. 3 (Mar. 2015): 431–37; and Susan L. Ivey et al., “Patient Engagement in ACO Practices and Patient-Reported Outcomes Among Adults with Co-Occurring Chronic Disease and Mental Health Conditions,” Medical Care 56, no. 7 (July 2018): 551–56.
9. Keane K. Lee et al., “Post-Discharge Follow-Up Characteristics Associated with 30-Day Readmission After Heart Failure Hospitalization,” Medical Care 54, no. 4 (Apr. 2016): 365–72.
10. Joynt et al., “Segmenting High-Cost,” 2017; Blumenthal et al., “Caring for High-Need,” 2016; David Mancuso, Daniel J. Nordlund, and Barbara Felver, Frequent Emergency Room Visits Signal Substance Abuse and Mental Illness (Washington State Department of Social and Health Services, June 2004); and Susan L. Hayes, Douglas McCarthy, and David C. Radley, “The Impact of a Behavioral Health Condition on High-Need Adults,” To the Point (blog), Commonwealth Fund, Nov. 22, 2016.
11. Jürgen Unützer et al., “Collaborative Care Management of Late-Life Depression in the Primary Care Setting: A Randomized Controlled Trial,” Journal of the American Medical Association 288, no. 22 (Dec. 11, 2002): 2836–45; and Emily Woltmann et al., “Comparative Effectiveness of Collaborative Chronic Care Models for Mental Health Conditions Across Primary, Specialty, and Behavioral Health Care Settings: Systematic Review and Meta-Analysis,” American Journal of Psychiatry 169, no. 8 (Aug. 2012): 790–804.
12. Hayes, McCarthy, and Radley, “Impact Behavioral,” 2016.
13. David Muhlestein, Robert Saunders, and Mark McClellan, “Medicare Accountable Care Organization Results for 2015: The Journey to Better Quality and Lower Costs Continues,” Health Affairs Blog, Sept. 9, 2016; and J. Michael McWilliams et al., “Medicare Spending After 3 Years of the Medicare Shared Savings Program,” New England Journal of Medicine 379, no. 12 (Sept. 20, 2018): 1139–49.