How Hospitals Can Control Medicaid ‘Super-Users’

Insights Results

Overview of article

• This article introduces a released article by The Center for Medicaid and CHIP Services (CMCS), which evaluated how hospitals can cut costs and improve patient care by setting up care delivery models that target Medicaid patients who have complex, unaddressed health issues and a history of frequent encounters with healthcare providers. Specifically, CMCS said the success of 10 programs that have targeted Medicaid super-utilizers indicate that early intervention and primary care can save money while improving the health outcomes of these complex patients. The article also suggests 6 questions hospitals and states should consider prior to launching similar programs

Key takeaways/implications

• The following 6 questions are presented for review by states and providers interested in launching super-utilizer programs:
  1) Should we pursue a super-utilizer program in our state? – A state should consider if a super-utilizer program in their state is likely to be successful in improving beneficiary outcomes and reducing unnecessary spending. This involves considering 4 items: 1) Identifying the major super-utilizer subpopulations within the state; 2) Identifying a provisional set of factors driving high-utilization among these populations; 3) Assessing the feasibility of eliminating unnecessary utilization through a set of targeted interventions to address those factors; and 4) Estimating both the potential cost and savings associated with a program that is able to address those drivers and reduce unnecessary utilization. Answering these questions will require to first analyze claims data to identify the potential super-utilizer population. In particular, states will want to ensure that this population represents “impactable” costs which may include patients who have multiple emergency department visits for mental illness or substance use disorders. The state can then begin to identify types of interventions that would be necessary to address the needs of these populations. Finally, the state should estimate the initial and recurring costs of establishing and maintaining the program’s infrastructure including web-based provider portals with patient data (allowing providers and programs to sort patients by number of recent hospitalizations and ED visits to consider patients’ utilization patterns and develop tailored interventions to meet their needs), real-time utilization data like a state health information exchange, and decision support tools to help care managers use these data to identify and prioritize high risk individuals
  2) What payers are involved? Although Medicaid is the most common insurer for super-utilizers, other payers should be considered. Developing a super-utilizer program focused on Medicaid population offers the advantages of allowing states and managed care organizations to use existing Medicaid data systems and analytic tools and to design targeted services. However, partnering with other payers may provide greater incentives to providers to refer and partner with super-utilizer programs. When partnering with other payers, data base and logistical challenges in adapting program services to the different needs of different populations must be considered
  3) Who provides the services and what is their relationship to primary care providers? Partnership with primary care providers should be considered. Approaches to partnership include: A) Centralized where care managers or outreach workers employed or contracted by the state or Medicaid managed care organization are embedded in primary care practices; B) Supportive networks where not-for-profit, community based organizations provide care managers to support a network of primary care practices in their region; and C) Community-based care teams where interdisciplinary teams including nurse care managers, social workers and behavioral health workers based on communities visit patients in their homes and community settings. Ambulatory intensive care units are another approach and offer more comprehensive services in specialized settings. These programs can provide short-term interventions (providing comprehensive medical, mental health addiction treatment and social services for a limited duration before transitioning to a primary care site) or take over care of the patients permanently (including an interdisciplinary staff comprised of physicians, nurse care managers, social workers, pharmacists and behavioral health specialists with extensive experience care for medically and social complex patients)
  4) What is the targeting strategy? First, the target population must be identified. There are 7 proposed targeting approaches: A) Targeting based on high observed-to-expected costs; B) Targeting specific patterns of care; C) Targeting very high levels of utilization (it should be noted that high levels of spending alone is often a flawed approach as this is likely a sign of legitimate and necessary medical treatment for a high-cost condition rather than high rates of inpatient or outpatient care); D) Targeting based on referrals and follow-up investigation; E) Excluding candidate clients with medical conditions associated with high but non-preventable costs (e.g., cancer or acute trauma); F) Targeting by presence of risk factors associated with high, preventable costs (e.g., substance use disorders, homelessness); or G) Targeting by community
  5) What services are provided? – The selection of services should be guided by the needs of the individuals selected for program and access to primary care and behavioral health in the community. Services may include: in-person medical care, in-person behavioral healthcare, assistance with social needs and health coaching. Segmenting individuals into subpopulations (e.g., primarily ED v. primarily inpatient) can allow programs to tailor their services to individual needs. Effective super-utilizer programs use real-time data to identify potential patients, engage them while they are still at the hospital or in the ED, and follow-up with existing clients enrolled in the program in a timely way when they are hospitalized or visit an ED. Programs obtain real-time data either through Health Information Exchanges (HIEs) or admission/discharge/transfer feeds from partnering hospitals. Program staff review real-time data on a daily basis, sending care managers to meet potential and existing clients while they are hospitalized and responding quickly to existing emergent needs. Additionally, capacity, infrastructure, and physical location of primary care and behavioral health services in a community are key determinants of the scope of services provided and must be considered. After the program identifies the core services, it will need to carefully consider the type of staff who will be able to provide these services to patients most effectively. Finally, feedback loops are a key function for a super-utilizer program in determining program improvement
  6) How is the program funded? States should consider how to pay super-utilizer programs for services provided directly to Medicaid patients and how to fund other essential program components such as program planning, management and evaluation. Existing programs that serve Medicaid beneficiaries use several different payment mechanisms that warrant consideration, such as Medicaid Case Management Payment, Multi-Payer Case Management Payment, Per-episode of care payment for program services, per-member per-month payment to managed care organization, and shared savings for total cost of care
• There 7 ways in which CMS has worked to enhance data resources for states that are focused on Medicare-Medicaid enrollees: 1) Enhanced federal match for design, development, and implementation of MMIS; 2) Enhanced federal match for health information exchanges; 3) Administrative contract; 4) Medicaid Health Homes; 5) Integrated care models; 6) Targeted case management; and 7) Medicare data access and assistance