The Decisional Dilemma: Identifying and Addressing High Utilization and Cost Appropriately

A key challenge for practitioners (which we define as healthcare providers, payers, and health systems), as well as researchers, is that there is no consensus on a definition of HNHC patients, those high-need patients for whom high costs or use are potentially preventable or modifiable. The most common approach is to define the HNHC population as the patients with the highest healthcare use or costs during a historical time period, but no consensus has been reached on the appropriate duration of the time period and the number and types of services.^{12, 13} Also, focusing on use and cost alone can obscure detecting future high-need patients. Many high-cost patients use this level of care for a relatively short period of time, and it is the appropriate choice for treating their condition (e.g., certain types of cancer or orthopedic surgery patients).¹⁷ A second approach is to identify HNHC patients based on diagnoses, such as individuals with multiple co-occurring chronic conditions.^{18, 19} Relying chiefly on a diagnostic criterion, however, ignores that not all high-need patients have high costs that are preventable or modifiable. A large percentage of high-need patients use care appropriately, even if it may be costly (i.e., needed healthcare is received through ongoing relationships with primary care and specialist providers). Often, factors co-occurring with physical diagnoses, including functional limitations, mental health conditions, substance use, and social risk factors, are important drivers of high healthcare use and cost.²⁰ Overall, HNHC patients are those patients with chronic conditions for whom reducing use of high-cost healthcare services in favor of other types of care is theorized to result in better care and better outcomes. For the purposes of this review, we focus on studies of patients with high healthcare costs or use. Within this larger group of studies, we will highlight analyses that more directly identify patients whose use of healthcare services would be considered inappropriate and is potentially preventable or modifiable.

Description of the HNHC Patient Population

A common approach to describing the magnitude of the HNHC patient population nationally is in relation to the concentration of healthcare costs or use disproportionately attributed to a specific percentage of the population. National data from the Agency for Healthcare Research and Quality (AHRQ)–sponsored Medical Expenditure Panel Survey reports that, in 2016, patients ranked in the top 5 percent by their healthcare expenditures accounted for 50 percent of total healthcare expenditures.²¹ Among these top 5 percent of spenders, 42 percent were 65 years of age or older. Persons in the top 5 percent of spending were more likely to have at least one inpatient stay during the year with inpatient stays accounting for 40 percent of their healthcare expenditures.²¹ The Pew Charitable Trust found that the top 5 percent of Medicaid patients nationally use 60 percent of Medicaid funding.²² In relation to ED use, AHRQ’s Healthcare Cost and Utilization Project data reported that among Medicare beneficiaries 65 years of age or older, approximately 5 percent of the population accounted for 16 percent of ED visits and 14 percent of ED charges.²³ Among Medicare beneficiaries younger than 65 years, 6 percent accounted for 26 percent of ED visits and 24 percent of ED charges. Further, a systematic review on frequent ED use across payers found that, across studies, approximately 5 to 8 percent of patients visiting the ED were frequent users, with this small group accounting for 21 to 28 percent of all ED visits.²⁴

The HNHC patient population can be further characterized in relation to the prevalence of multiple chronic conditions. According to the Center for Health Care Strategies (CHCS), 80 percent of high-cost Medicaid beneficiaries have three or more chronic conditions.²⁵ CHCS further found that among Medicaid-only beneficiaries with a disability, each additional chronic condition was associated with an average increase in costs of $8,400/year. Among Medicaid-only beneficiaries, the most common diagnostic pairs of diseases among the highest cost patients are cardiovascular-pulmonary (30%), cardiovascular-gastrointestinal (25%), cardiovascular-central nervous system (25%), central nervous system-pulmonary (24%), and pulmonary-gastrointestinal (24%).²⁵ Based on Centers for Medicare & Medicaid Services administrative data from 2011, the burden of multiple (≥2) chronic conditions among Medicare beneficiaries is high, accounting for over two-thirds of the fee-for-service population, while those with six or more conditions account for 14 percent of the population.²⁶

What Is Known From Other Systematic Reviews

The present work builds on and extends earlier systematic reviews. We identified three reviews focusing on characteristics associated with being HNHC patient and eight reviews of interventions for these patients.

The three reviews of studies of HNHC patient characteristics identified their populations differently. One focused on high-cost patients,²⁷ the second on frequent ED users,²⁸ and a third on ED users whose visits were identified as primary care sensitive, preventable, inappropriate, or unnecessary.²⁹All three reviews elucidated factors associated with being HNHC, organized according to Andersen’s model of predisposing, enabling, and need characteristics.³⁰ The study of high-cost patients found that having multiple chronic conditions, mental illness, and advancing age were all associated with higher healthcare costs.²⁷ Patient income had a variable effect. Frequent ED users were more likely to have a fair or poor self-perception of their health coupled with mental health and substance abuse concerns.³¹ Characteristics associated with high primary care–sensitive ED use included being female, being black, and having Medicaid coverage.²⁹

The present review modified the approach used in these reviews in several ways. The Wammes et al. review of high-cost patients included studies from multiple countries, while we limit our work to the United States, so that our findings are generalizable to the United States’ healthcare coverage and delivery systems.²⁷ We add to the literature on characteristics associated with high healthcare use by separately examining high inpatient or outpatient use, as well as high ED use. We also include studies of chronic disease subgroups. Finally, while the earlier reviews organized their findings using the Anderson chronic disease framework, we group characteristics associated with being HNHC patients using the typology developed by the National Academy of Medicine.²⁰

Multiple reviews have synthesized the literature on interventions that are hoping to change the utilization patterns, decrease costs, and improve clinical outcomes among HNHC patients. One review by Raven and colleagues (2016) examined studies of interventions, implemented in the United States, intended to reduce ED visits.³² They found that case management reduced ED use and increasing ED copayments had mixed results. We identified six other reviews focusing on ED use.^{12, 13, 33–37} Of these, five were not limited to studies conducted in the United States so the findings would not necessarily be applicable to the U.S. healthcare system.^{12, 13, 33, 34, 36} A sixth review of ED interventions was limited to studies conducted in the United Sates but allowed the weaker pre-post designs that did not control for confounding.³⁷ Similarly, a review of interventions to address frequent hospitalization³⁸ and one of intensive outpatient care programs³⁹ also included pre-post designs.

Some of the limitations of earlier reviews of intervention studies are addressed in the present review. In prior work, the institutional settings in which studies were conducted were often not adequately described.¹² Also, information about social determinants of health, including socioeconomic status, behavioral health status, housing stability, and racial/ethnic demographics, was often missing. When systematic reviews included important social variables, analysis was limited to one or two factors. For example, several key systematic reviews (e.g., Soril et al.,¹³ Edwards et al.,¹⁵ and Baker et al.¹⁴) contain scant detail regarding social risk factors of health and the settings within which interventions occurred.

Attempts to synthesize the literature on interventions for HNHC patients are challenged by the absence of a consensus in the field of a unifying classification schema or taxonomy organizing key dimensions of these complex interventions into meaningful distinctions for grouping and separately considering their components.⁴⁰ Intervention classification in recent systematic reviews regarding high utilizers differed significantly. One systematic review stratified interventions by home based, clinic based, and primary care augmentation.¹⁵ Another categorized interventions into case management, individual care plans, and information sharing.¹³ Baker et al.¹⁴ stratified interventions by population type: adults (1) with two or more chronic diseases, (2) with one chronic disease and depression, and (3) identified at risk for high healthcare utilization.

Payer-Level Interventions

Payers are invested in reducing healthcare costs and inappropriate use by HNHC patients. Various alternative models of health services delivery can support HNHC patient interventions. For example, accountable care organizations (ACOs) align financial incentives and health outcome accountability to promote care coordination among a large group of healthcare providers across the care continuum (ambulatory outpatient, acute care, post-acute care, and home health settings). ACOs use a variety of strategies to accomplish these goals. They support care coordination by increasing physicians’ access to a uniform source of complete information about patients’ health service use and health status. Techniques such as population segmentation and risk stratification can be used to focus resources on HNHC patient populations. Together, these enhancements can increase coordination among healthcare settings, particularly during transitions between settings in episodes of care.

One example of a payer-driven model is home-based primary care. This model moves the delivery of the majority of primary care from an ambulatory office-based setting to a patient’s residence.⁴¹ Intended to be a comprehensive care delivery model, this model typically involves a team-based approach and combines home-based care for medical needs with intensive management and care coordination. Home-based care programs serve a patient population that has a high probability of being HNHC because participants commonly have complex chronic conditions, functional limitations, and physical restrictions in accessing traditional office-based care. Home-based primary care may benefit HNHC patients if such approaches can better address patients’ specific needs, values, and preferences.

Health System– and Provider-Level Interventions

Many complex healthcare interventions are organized at the health system and provider levels. Health system–level interventions are often designed to support changes at the patient population level by promoting a culture that furthers program goals and organizes intervention activities across participants.⁴² Rather than focusing on individual patients, system-level participants are generally clinics or practices, which in turn are often expected to implement provider- and patient-level interventions. Provider-level interventions target care providers to change how they treat patients. Examples of provider-level interventions include the distribution of evidence-based guidelines and protocols, provider education and training, and provider feedback reports.

At the health system and provider levels, data are often used to report healthcare service use that may indicate that a patient is HNHC and needs additional services. One example is ED alerts. ED alerts inform clinicians that a patient has received care in the ED, which can help them identify patients whose patterns of care might be considered HNHC and in need of additional support services. Another strategic use of data to identify HNHC patients is “hotspotting.” This strategy uses data to identify individuals whose high costs are outliers, understand the problem, dedicate resources, and design effective interventions.⁴³ Hotspotting interventions typically include multidisciplinary, coordinated care that treats the whole patient, including the nonmedical and social determinants that affect health, such as housing, mental health, substance abuse, and emotional support. The underlying idea is that intensive and highly individualized care addressing the unique needs of an HNHC patient, including social determinants of health, will improve clinical outcomes and reduce health expenditures beyond the cost of the high-intensity, highly individualized hotspotting intervention.⁴⁴

Cross-Sector Interventions

Although not traditionally considered in the domain of healthcare, supportive services addressing social risk factors can affect the success or failure of complex interventions for HNHC patients. Interventions that focus on cross-sector issues intend to bridge the gap between engaging healthcare professionals who deliver direct care and addressing individual patients’ nonmedical needs. A variety of roles, such as care managers, social workers, community health workers, patient navigators, and peer-to-peer networks, can directly provide or connect patients with needed support services.

Support services are intended to address patient barriers related to resource limitations, health literacy, and treatment adherence, thereby improving patient outcomes. Relatedly, staff providing supportive services can assist with social risk factors that have been identified as predisposing factors limiting access to care. Predisposing factors may include low income, poor education, poor nutrition, homelessness, and lack of transportation. For example, stable housing could improve patients’ ability to interact consistently with their healthcare providers and social support systems, which, in turn, could increase adherence to their treatment plan.⁴⁰

Complex Interventions for HNHC Patients

Intervention strategies that target HNHC patients encompass the approaches discussed above. We grouped interventions for HNHC patients, using the scheme proposed by Bodenheimer to group care management programs, into seven categories based on the primary setting in which the intervention is delivered (with several modifications).⁴⁵ We included system-level transformation models that affect a larger population of patients, but we only report findings for those patients who are HNHC. We did not include hospital discharge models because they are outside the scope of this review. Most of the interventions designed to target the complex needs of HNHC patients are themselves complex and are often a hybrid of various models, so some may not neatly fit into one of these seven categories.

• System-level transformation model: Clinics, practices, and other organizations modify how care is delivered to all patients.
• Telephonic/mail model: Care is delivered remotely by telephone or mail.
• Community-based model: Care provider meets with the patient wherever the patient is located.
• ED-based model: Patients are recruited and receive care in the ED.
• aICU (ambulatory intensive caring unit) model: Care is received in a separate high-risk clinic or through a high-risk team within a primary care clinic.
• Primary care model: Care is embedded in one or more primary care practices.
• Home-based care model: Care is delivered in the patient’s home.