Health Care Is Failing the Most Vulnerable Patients: Three Underused Solutions

“High-need high-cost (HNHC)” patients, “frequent fliers,” and “high utilizers” are buzzwords that describe medically and functionally impaired adults who incur high health care expenses. ¹ This population contributes substantially to the skewed distribution of health care spending in the United States, where 5% of patients account for 50.4% of spending. ² Care for chronically ill patients with behavioral health comorbidities and low socioeconomic status is one of the US health care system’s greatest challenges.

HNHC patients present thorny problems to health care organizations seeking to improve performance on health care’s “Triple Aim,” which has 3 components: (1) improving the patient experience of care, including quality and satisfaction; (2) improving the health of populations; and (3) reducing the per-capita cost of health care. ³ HNHC patients incur 4 times the average annual health care expenditures, are more likely to continue to incur high costs over time, and report poorer experiences of care than the US general population. ¹ Low income is associated with a higher prevalence of chronic disease and higher health care costs among HNHC patients compared with the US general population, primarily driven by preventable use of acute care. ^4,5 The large percentage of HNHC patients with a behavioral health comorbidity in particular often have unmet medical needs, poor access to specialists, and high health care expenditures. ⁶ As clinicians struggle to meet the needs of these patients within traditional fee-for-service health care, a sense of futility often takes hold, which contributes to burnout and higher rates of clinician turnover in settings where HNHC patients congregate compared with other settings.

We aimed to spearhead a redesign of health care delivery for the HNHC population by identifying effective but underused interventions for HNHC patients based on 3 major sources of risk to their clinical stability: medical and functional complexity, behavioral health comorbidity, and low socioeconomic status (Figure). ^{1,6 -8} Our institution uses a structured design-prototyping method, adapted from human-centered design thinking, ⁹ that begins with an in-depth review of medical and industrial literature. ^10,11 This review is followed by direct field observation and systematic discussions with stakeholders and subject matter experts to further refine our understanding of unmet needs, to uncover options for improvement from other systems of care or closely related industries called the “adjacent possible,” and to catalog value-enhancing care delivery innovations at pioneering provider sites. After we defined the unmet needs and potential innovative solutions to those needs, we scored and prioritized the solutions based on their projected impact on cost of care, patient experience, quality of care, and ease of implementation. We used insights gained from this process to build macroeconomic models and formulate distinct interventions that we then coalesced into a novel health care delivery model. We iteratively refined this model through discussions with experts in health care systems and the care of HNHC patients. At the conclusion of this process, the health care delivery model consisted of 3 components: shifting acute care from hospitals, economically integrating behavioral health into primary care, and addressing 2 social determinants of health that are amenable to intervention at the individual level: housing and access to transportation for medical care.

Figure

Conceptual framework of high-need high-cost population with psychosocial vulnerability. Data sources: Hayes SL et al, ¹ Hayes SL et al, ⁶ Hedden et al, ⁷ and Henry J. Kaiser Family Foundation. ⁸

Shifting Acute Care From Hospitals

HNHC patients incur higher rates of preventable emergency department (ED) visits and subsequent hospitalizations than the general population. Such services cost Medicaid and Medicare as much as $13 billion annually. ^{12 -14} Subject to physician-determined guidelines for assuring patient safety, innovative health care organizations, typically in partnership with emergency medical services (EMS) and/or law enforcement, are safely shifting such care to 2 lower-cost settings: mobile integrated health–community paramedicine (MIH-CP) programs and sobering centers.

Mobile Integrated Health–Community Paramedicine Programs

MIH-CP programs allow for the use of round-the-clock paramedics’ expertise to expand care delivery to patients’ homes. The 2 key factors to successful MIH-CP programs that are acceptable for patients and their caregivers are (1) additional training of paramedics to foster diagnostic and triage skills and (2) closed communication between paramedics and the physician supervising the program. The latter ensures continuity of care that enables primary care providers to organize needed services before a clinical deterioration leads to a costly ED visit or hospitalization. Together, these factors allow for safe expansion of paramedics’ scope of practice. ¹⁵ MIH-CP programs may provide services ranging from responding to urgent care needs to performing hospital discharge follow-up visits for patients at high risk for readmission. ¹⁵ In a study of 980 beneficiaries eligible for both Medicare and Medicaid with considerable medical, behavioral, and social needs as well as average per-member, per-month spending of $2600, ¹⁶ MIH-CP programs yielded net annual savings of $454 720, had improved patient satisfaction, and led to fewer ED visits among 85% of study participants. ^17,18 An evaluation of several MIH-CP programs serving 2515 people in California estimated cost savings from avoiding ED visits and hospital readmissions at approximately $2.4 million from 2015 to 2017. Most savings accrued to Medicare and Medicaid, as well as hospitals serving a large number of Medicare beneficiaries. ¹⁹

Sobering Centers

Sobering centers are designed to offer a safe environment for short-term recovery for intoxicated people who do not require the high level of care provided in an ED. Sobering centers typically accept intoxicated people from law enforcement and/or EMS and may provide referrals to other social services in the community. ²⁰ From 2006 to 2014, the number of ED visits for acute alcohol intoxication increased 50%, from 1.8 to 2.7 million visits nationally, and generated annual costs of $9 billion. ²¹ Compared with the $1428 average cost of a treat-and-release ED visit for any condition (2017 Medical Expenditure Panel Survey), ²² sobering centers provide comparable services for $265 per visit for alcohol intoxication. ²⁰

Communities such as San Antonio, Texas, report that savings from sending people to sobering centers also accrue to criminal justice services and other publicly funded service organizations and, thereby, to taxpayers. ²³ Although no data on the use of EDs for acute intoxications are available for the HNHC population, the prevalence of alcohol and other substance use disorders in this population, which is reported to be as high as 47%, ²⁴ suggests that wider testing and refinement of sobering centers may be warranted.

Integrating Behavioral Health Into Primary Care

Unaddressed behavioral health conditions are associated with 2.5 to 3.5 times higher total medical expenditures among patients with a comorbid behavioral health condition compared with patients without a comorbid behavioral health condition. ²⁵ Most patients with behavioral health conditions present in primary care settings, ²⁶ indicating that these settings could serve as a practical access point to address the physical and behavioral health needs of HNHC patients. Of the various behavioral health integration models, the Collaborative Care Model (CoCM), which has been implemented nationally, has the largest evidence base demonstrating clinical efficacy and cost effectiveness. ^27,28 Patients enrolled in the CoCM were twice as likely as patients in usual care to have a significant reduction in depressive symptoms. ²⁹ Growing research demonstrates the efficacy of the CoCM with other behavioral health conditions, such as posttraumatic stress disorder and substance use disorders, and in improving physical symptoms of comorbid chronic medical conditions. ^{30 -32}

The realized cost savings from the CoCM are subject to initial investments in staffing and training, resulting in net costs of $522 per patient in the first year. ³³ The intervention adds 2 new roles to the primary care team: a behavioral health manager and a psychiatric consultant. ³⁴ Efforts to use telepsychiatry as part of the CoCM ³⁵ may help reduce on-site staffing demands and lower costs. Ultimately, the CoCM generated an estimated $3363 net savings per patient in total health care costs during a 4-year period. ³³ In 2018, Medicare adopted reimbursement codes to allow primary care providers to bill for CoCM services, allowing for increased financial viability in fee-for-service systems. Financial modeling shows that the CoCM consistently increases net revenue for primary care practice sites. Now boosted by supplemental Medicare payment, the CoCM may be primed for rapid adoption. ³⁶

Addressing 2 Key Social Determinants of Health

Although more than 95% of US health care spending is for medical care, ³⁷ social factors, environmental factors, and individual behavior have a greater influence on health and well-being than direct medical services. ³⁸  US health professionals, including the American College of Physicians, have called for action to address the social determinants of health. ³⁹ Moving from recognition to action is challenging, but evidence suggests that 2 promising programs warrant wider testing and refinement: funding nonemergency medical transportation (NEMT) and housing the highest-cost persistently homeless population.

Funding Nonemergency Medical Transportation

Poor access to transportation leads to missed medical appointments, poor clinical outcomes, and high health care spending. Unmet NEMT needs cause 3.6 million people to miss or delay medical care annually. ⁴⁰ Downstream effects of this postponed care include progression of untreated chronic disease, preventable use of costly higher-acuity care, and reduced clinician productivity. ^{41 -43} Evidence suggests that transportation assistance is cost-effective and, in some instances, cost saving. ⁴⁰ Compared with traditional NEMT, such as taxi vouchers or direct provision of ride services, transportation networks that leverage digital technologies promise improved savings and patient experience. A 2016 pilot program between CareMore and Lyft demonstrated a 30.0% reduction in wait time, 32.4% reduction in average per-ride cost, and 80.8% satisfaction rate after initiation of the pilot program. ⁴⁴ Modern NEMT brokers, such as Circulation, present additional advantages by contracting with transportation network companies such as Uber and Lyft, as well as with various other transportation options (eg, wheelchair-accessible vehicles with door-to-door service) that can be deployed based on patient needs. Wider use and assessment of such services are needed because indiscriminately offering transportation is unlikely to yield a sustainable return on investment, even if limited to HNHC patients. ⁴⁵

House the Highest-Cost Persistently Homeless

Despite high levels of health care, adults experiencing homelessness have poor health outcomes. Compared with the general population, adults experiencing homelessness have a higher prevalence of chronic disease and are 3 to 4 times more likely to have a shorter life expectancy. ⁴⁶ Up to 50% of adults experiencing homelessness have a mental illness and/or substance use disorder. ⁴⁶ The distribution of total public spending among people experiencing homelessness is also skewed; for example, 5% of people experiencing homelessness accounted for 47% of total public expenditures in an analysis involving 7 health care, criminal justice, and social services agencies in Santa Clara County, California. ⁴⁷

Targeted permanent supportive housing (PSH) with comprehensive physical health, behavioral health, and social services for people experiencing homelessness could improve health outcomes and be in the financial interest of the health care and criminal justice systems. One type of PSH, Housing First, ⁴⁸ separates the provision of housing from preconditions of sobriety or compliance with treatment. Outcomes from Housing First programs are equivocal, with anecdotal evidence and quasi-experimental studies suggesting cost savings that have not been confirmed in randomized controlled trials. ⁴⁹ However, a PSH intervention that selectively targeted the 10% of people experiencing homelessness with the highest health care costs and other social services projected reductions of 58.1% in total health care spending and 99.6% in criminal justice system spending, excluding the cost of the intervention, compared with propensity-matched controls. ⁵⁰ Using predictive models, the number of people experiencing homelessness that are targeted can be adjusted based on the supply of PSH and desired cost savings. ⁴⁷ Even including the cost of the intervention would result in net reductions in total health care and criminal justice system spending. However, the exact net reduction per system would depend on the cost distribution. Although federal regulations prohibit matching of state Medicaid dollars spent on room and board, state Medicaid dollars can be used for on-site supportive services, which are shown to be critical to the effectiveness of PSH. ^49,51

Public Health Implications

Despite substantial discussion of the challenges of better serving HNHC populations, most health care organizations have not used effective strategies for several reasons. First, several regulations act as barriers. In particular, widespread implementation of MIH-CP programs and sobering centers is hindered by regulations that mandate the transport of all patients to the ED and limit the scope of practice of nonphysician providers such as paramedics. For example, California’s Health and Safety Code (HSC §§ 1797.52, 1797.218) limits paramedics’ scope of practice to providing care in emergencies only (ie, preventive visits cannot be performed) and requires transport of patients to an ED or an acute care hospital. ¹⁹ Second, financial incentives are misaligned. The relationship between the health care system and social determinants of health has been called the “wrong pockets” problem. ⁵² This term refers to the institution that pays the upfront costs for an intervention not receiving the financial benefit of the intervention’s increased costs, resulting in sustained inaction. Finally, all strategies demand challenging but necessary collaboration among diverse stakeholders in financing, standards, and licensing. In many settings, establishing fruitful relationships among stakeholders requires cultural change (eg, appreciation of diverse perspectives, development of mutual trust in other professions’ protocols and standards, common goals) that is difficult to achieve.

Our design process identified 3 underused solutions that are likely to improve value if selectively applied to population subsets where success has been demonstrated by innovative peers. More rapid adoption, assessment, and refinement of these 3 underused solutions will require federal and state regulatory reforms, as well as private-sector movement to value-based purchasing. Policy options include the use of Medicare and Medicaid funds to pay for EMS that does not result in ED transport, such as the alternative payment model currently pilot-tested under the auspices of the Centers for Medicare & Medicaid Services ⁵³ and states’ expansion of the scope of EMS providers’ practice. All payers, particularly Medicaid, should follow Medicare’s lead in enabling primary care–based payment for integrated behavioral health services. Lastly, to address key social determinants of health, policy makers should test relaxation of current prohibition of the use of Medicaid funds to house the highest-cost adults experiencing chronic homelessness in severe states of chronic illness or add comparative cost-effectiveness to the US Department of Treasury’s criteria for community benefit spending by nonprofit hospitals. Although addressing policy shortfalls will have an impact on achieving eventual change, clinicians and their health care organizations can move forward with 3 tangible changes that will make an immediate difference in the lives of our most poorly served patients and in our stewardship for the high publicly funded spending that they incur.