Are accountable care organizations (ACOs) potentially creating financial savings by neglecting or limiting care for serious mental illness (SMI)? A team of researchers has examined the issue for an article in Health Affairs. Based on their analysis, it appears that such is not the case; but there is complexity, which the researchers explore in their article.
In the article published in the August issue of Health Affairs and entitled “ACO Participation Associated With Decreased Spending For Medicare Beneficiaries With Serious Mental Illness,” José F. Figueroa, Jessica Phelan, Helen Newton, E. John Orav, and Ellen R. Meara look at the complexities around care for serious mental illness for Medicare patients enrolled in Medicare Shared Savings Program (MSSP) ACOs. What they find is complex and somewhat nuanced.
The authors state in their abstract at the outset of the article that “Serious mental illness (SMI) is a major source of suffering among Medicare beneficiaries. To date, limited evidence exists evaluating whether Medicare accountable care organizations (ACOs) are associated with decreased spending among people with SMI. Using national Medicare data from the period 2009–17, we performed difference-in-differences analyses evaluating changes in spending and use associated with enrollment in the Medicare Shared Savings Program (MSSP) among beneficiaries with SMI. After five years, participation in MSSP ACOs was associated with small savings for beneficiaries with SMI (−$233 per person per year) in total health care spending, primarily related to savings from chronic medical conditions (excluding mental health; −$227 per person per year) and not from savings related to mental health services (−$6 per person per year). Savings were driven by reductions in acute and post-acute care for medical conditions. Further work is needed to ensure that Medicare ACOs invest in strategies to reduce potentially unnecessary care related to mental health disorders and to improve health outcomes.”
José F. Figueroa, M.D., M.P.H., is an assistant professor of health policy and management at the Harvard T.H. Chan School of Public Health (HSPH) and an assistant professor of medicine at Harvard Medical School (HMS). He is also a practicing Internist and associate physician at the Brigham and Women's Hospital, where he serves as the Faculty Director of the BWH Medicine Residency Management & Leadership Pathway. Jessica Phelan is a statistical analyst programmer at the Harvard Global Health Institute. Helen Newton, Ph.D., M.P.H., is a postdoctoral associate in the Department of Health Policy & Management at the Yale University School of Public Health. E. John Orav, Ph.D., is an associate professor in the Department of Biostatistics at the Harvard T.H.Chan School of Public Health. Ellen R. Meara, Ph.D., is a professor of health economics and policy at the T.H. Chan School of Public Health.
But what is behind that small savings on mental health services? It turns out, things are a bit complicated. For one thing, while more than one in five U.S. adults suffers from a mental health disorder, and 14.2 million U.S. adults suffer from a serious form of mental illness such as bipolar disorder, schizophrenia, or major depressive disorder, “In the Medicare population, the burden of serious mental illness (SMI) is a concerning amount higher than in the general population, with a recent study suggesting that the prevalence of SMI in this population was nearly 23 percent,” the article’s authors note. “This may be because Medicare beneficiaries have a higher number of chronic physical conditions than the general population, which may then lead to a higher prevalence of major depressive disorder, and given the bidirectional relationship, depression may also lead to worsening of underlying chronic conditions. High rates of schizophrenia and bipolar disorder may also result in disability, which then qualifies these patients for the Medicare program. In addition, Medicare beneficiaries with SMI were also found to spend substantially more on the treatment of other chronic medical conditions, such as heart failure and diabetes, than those without, even after clinical risk adjustment. This is likely because the presence of SMI impairs the ability of patients and clinicians to effectively treat other chronic conditions,” they note.
In theory, the ACO model of care should be helpful in this regard, as, under the MSSP program, a group of clinicians and patient care organizations accepts responsibility for attributed patients across time. Still, the researchers note, “To date, there are few long-term empirical data about how patients with SMI and comorbid chronic medical conditions are faring in ACOs. Early evidence suggests that ACOs have achieved modest savings and improved quality for the general Medicare population, likely because of the financial incentives to care for patients across the entire care continuum,” they write. Still, they note, “Although some studies have examined rates of outpatient visits to mental health providers and use of psychotropic medications among people with depression, it is unclear whether ACOs yield meaningful savings among people with SMI over a longer period of time as experience caring for these patients in ACOs increases.”
The researchers write that they wanted to answer the question, “Was the implementation of Medicare ACOs, specifically the Medicare Shared Savings Program (MSSP), associated with savings among beneficiaries with SMI? If so, were these savings achieved from reductions in spending related to mental health services or related to treatment of chronic medical conditions? Finally, did enrollment in ACOs lead to reductions in health care use among those with SMI, including rates of hospitalizations, emergency department visits, and post-acute rehabilitative care use, relative to beneficiaries not in ACOs?”
So, the researchers “used a 20 percent sample of Medicare administrative claims from the period 2009–17 that included Parts A and B spending and use. Our sample was limited to Medicare fee-for-service beneficiaries continuously enrolled during the study period or until death. Demographic data were obtained from the Master Beneficiary Summary File. Claims from the Inpatient, Outpatient, Carrier, Skilled Nursing Facility, Home Health Agency, and Hospice files were used.” And the patients whose records looked at were “cohorts of patients who were attributed to MSSP ACOs that started contracts in 2012, 2013, 2014, or 2015.”
In that regard, the researchers write, “In a national study of Medicare beneficiaries, we found that participation in the Medicare Shared Savings Program between 2013 and 2017 was associated with small savings among those with SMI, including schizophrenia and related psychotic disorders, bipolar disorder, and major depressive disorder. These savings were primarily related to reductions in spending related to medical conditions and not reductions in the treatment of mental health disorders.”
And, they state, “Our findings suggest that ACO savings are primarily related to the treatment and management of medical conditions and not due to changes in spending related to mental health disorders. These findings raise important questions. On the one hand, it is possible that the observed savings may signal more efficient care under ACOs for the treatment of medical conditions. Prior work has suggested that poorly controlled medical conditions among people with mental illness are an important driver of morbidity and mortality.47 It is possible that ACOs are mitigating some of the effects of mental illness on chronic medical conditions and preventing potentially unnecessary care, as evidenced by greater reductions in ED visits, hospitalizations, and subsequent post-acute care use. Our findings are consistent with other work that has shown that ACO incentives likely motivate physician practices to lower use by investing in specific strategies, including care transitions and care coordination programs, risk-stratification interventions, and chronic disease management programs. The magnitude of the savings among people with SMI, however, is about half the savings previously reported among the general ACO population in the MSSP.”
Significantly, they write, “[W]e found no evidence to suggest meaningful reductions in spending related to mental health disorders. This may be because other work has suggested that there has been little integration of behavioral health treatment in traditional primary care health systems. More recent data suggest that only 17 percent of ACOs reported implementing all components of the collaborative care model, which is a cost-effective model to treat mental illness that combines primary care and consulting behavioral health specialists with the support of mental health registries.”
This is obviously complex, since, as the authors write, “There has… been little focus on specific quality measures that concentrate on the treatment of mental health disorders. This is a missed opportunity, given that Helen Newton and colleagues found that mental health–specific quality measures are associated with ACO reports of behavioral health integration activity and with better follow-up after mental health hospitalizations.”
Ultimately, they conclude, “We found that after five years of participation in the MSSP, beneficiaries with SMI who are treated by ACO practices were achieving some small savings, primarily related to reductions in acute and post-acute care use and spending related to chronic medical conditions and not from reductions in mental health services. Although these findings may reflect potential reductions in unnecessary care related to chronic medical conditions, further work is needed to understand the impact of ACOs on health outcomes. In addition, our work suggests that ACOs may still need to implement more strategies to reduce potentially unnecessary care related to mental health disorders.”
