Abstract
Expanding health care costs lead those at financial risk for the burgeoning expenses – payers, patients, policy makers, and health systems – to find fast solutions. Medicaid expansion has led states to become increasingly responsible for these costs, at the expense of other fundamental social services. Massachusetts, for example, was one of the first states to provide near-universal health insurance coverage for residents through its Medicaid plan, MassHealth. In 2007, the first year of the program, MassHealth accounted for 27% of the state budget. In fiscal year 2019, MassHealth is estimated to account for 36% of the state budget, leading to cuts in programs such as mental health, education, local aid, and transportation. 1 MassHealth aims to decrease the total cost of care (TCOC) for its population and, as a result, hospitals and clinicians are asked to prioritize and leverage strategies thought to impact these goals, focusing on high-cost utilization through reducing inpatient admissions and readmissions, reducing emergency department visits, improving access to behavioral health services, and reducing overutilization of radiology and laboratory diagnostics.
If achieved, these goals likely will benefit both cost and quality of care. The groups asked to implement them – including MassHealth Medicaid Accountable Care Organizations (ACOs) – often are comprised of entities with financial and other incentives that are misaligned with the goals. For example, many of the MassHealth ACOs are affiliated with hospitals. Hospitals are historically analogous to auto repair shops: they wait for a car to break down and present to the shop with a (costly) problem before getting involved in solutions or early intervention. Hospitals have considered preventive care, population health, and the social needs involved beyond their scope. Aligning incentives among varied and historically misaligned groups – including hospitals and health systems, frontline providers, payers, community-based organizations, and the populations they serve – is critical to achieve health care cost and quality improvements.
In our individual roles as a primary care family physician, hospitalist, and emergency department physician in an urban academic tertiary hospital, we have witnessed the effect of misaligned incentives on several patients we have served. GG is one these patients. GG is a 28-year-old male who first presented to the family medicine clinic for primary care as a teenager. He has moderate persistent asthma, and because of the hierarchy of needs in his life – with school, work, and community engagement superseding his health – he often did not have medications readily available and missed multiple appointments. He also is the victim of trauma and its consequences: he witnessed the deaths of several friends killed by gun violence, self-medicated with substances, and stopped coming to primary care and pulmonology visits, despite outreach efforts by clinic personnel. When GG ran out of asthma medications, he presented to the emergency department; by the time of arrival, he was short of breath, often with low oxygen levels, and required treatment for acute asthma exacerbations with nebulizers and oral steroids. Given that GG was never admitted to the hospital for these exacerbations, one can assume that they could have been both prevented with appropriate medications and/or addressed in an outpatient setting. And while prednisone, a steroid, is effective in treating asthma, the medication also can have significant side effects, including a mild sense of euphoria and steroid-induced psychosis. In fact, the medical literature describes a nearly 6% incidence of steroid-induced psychosis in the adult population. 2
Over time, GG increasingly presented to the emergency department with deteriorations in his mental health and social circumstance. His behavior became more erratic; as a result, he became estranged from his family, homeless, and inappropriate with the emergency department staff, at times becoming verbally and physically assaultive. In one recent month, GG presented to the emergency department 12 times. He continues to miss scheduled, clinic-based primary care, psychiatry, and pulmonology appointments. His outpatient care team collaborates in an interdisciplinary effort to reengage GG, deploying an array of wraparound services. These services include patient navigation and care management to help with housing insecurity, integrated behavioral health to help with his mental health and anger issues, pharmacy teams to deliver his medications, and appointments to visit his primary care and pulmonology providers for his worsening asthma. Despite the service offerings of primary care, we are not able to engage GG in a meaningful way. We understand GG's hierarchy of needs is led by psychosocial problems that, if adequately addressed, might allow us to engage in medical care. We understand that GG has a history of trauma – possibly even from the health care system – and that a clinic-based, health-centered model may not be appropriate for his care. Even more, we know – from GG and from population literature – that investment in social needs can have greater impact on health outcomes than medical expenditure. 3 –5 And yet, GG's episodic care in the emergency department continues to reflect the absence of health and social services that could truly lower his TCOC.
When evaluating the multiple medical stops in GG's health care journey, we note that each health care provider has done his or her individual job thoughtfully and diligently. The emergency department focuses on GG's individual encounters and carries out its mandate to evaluate, manage, and treat his medical and nonmedical issues each time he presents. The primary care team reaches out by phone to GG regularly and develops medical plans for his care. Each team recognizes the social drivers and barriers to health and is inspired to help GG with his care. And yet, within this traditional delivery model, we do not have the structures or impetus to develop a health and social system that will address the needs of high-cost, high-need patients like GG. For example, the hospital is not yet at financial risk for GG's emergency department visits or the consequences of his homelessness and behavioral health conditions; in other words, currently, the hospital is reimbursed by Medicaid for each emergency department visit GG makes. Without an incentive to change system behavior, hospitals will continue to operate as repair shops.
As Medicaid and other ACOs begin the journey to downside risk, we have an opportunity to optimize and enhance the system to provide patient-centered, end-to-end, integrated health and social care. Social services are losing funding at the state level while the health care budget is increasing. And although health systems have traditionally shied away from delivering social services, they could pay for them and/or partner with community-based organizations. Perhaps, in GG's case – if financially responsible for his care – the health system to which he is attributed would have reason to do so.
Although Medicaid ACOs are beginning to build the support and incentives for MassHealth patients who are “super-utilizers,” a term reserved for the small proportion of patients who account for the largest proportion of health care costs (in the US, the top 1% of patients incur more than 20% of health care costs, and 5% of patients incur approximately 50% of total costs 6,7 ) there is much to be done about the thousands of patients like GG who do not yet meet that super-utilizer threshold. One promising step in this direction is medical education and clinical training. The next generations of frontline providers are learning to diagnose and treat the psychosocial and psychiatric issues that drive health outcomes. Every primary care and emergency medicine clinician in our hospital is now trained to ask patients about social determinants of health, such as stable housing, access to medications, food insecurity, substance use disorders, and educational aspirations. Starting last year, our hospital also became the sixth in the country to open a residency to train family medicine–psychiatry residents. These physicians will have the biopsychosocial training of primary care physicians and be versed in psychiatric sequelae of regular medications, helping to guard us from side effects of commonly used medications such as prednisone.
Despite these advances in clinical training, the health system continues to fall short of meeting the needs of patients like GG. When looking at TCOC reductions and improved health care outcomes for our patients, we must align our incentives across the health and social systems. We need to build processes that facilitate real-time shared medical decision making and warm transitions of care among hospital and outpatient care teams during each patient encounter. We need to partner with community organizations – such as shelters, substance use treatment centers, and outpatient behavioral health services – and use the health system to finance organizations that improve health outcomes and cost for their covered lives. As educators, we must continue to train our students to work in a cooperative model, and to think creatively about how to meet health needs outside of the walls of the repair shop. As policy makers and insurers, we must incentivize this end-to-end approach rather than episodic, fee-for-service care. Collectively and collaboratively, we can begin to develop a health and social system that will improve the lives of our high-cost, high-need patients.