Thank you for the opportunity to testify today. I have been studying the Medicare post-acute care populations for over 20 years, most of which has been to provide either Congress or CMS and ASPE research to address their Medicare payment and quality monitoring policies for beneficiaries using long term care hospitals, inpatient rehabilitation hospitals, skilled nursing facilities, home health agencies, and hospices. My dissertation, which I completed over 20 years ago, addressed the issue of fragmented financing and delivery system for these populations and I feared it was a trite topic then, because people had been actively trying to “integrate” services and funding streams since the 1970s. Yet, here we are in the new century, and these issues still aren’t resolved.
One difference today is that the conversation has changed. Instead of discussing cost constraints, ensuring access to services, and thinking about the costs of individual parts of each insurance program we are now, as a nation, focusing on people- on developing person-centered systems of care. This is not just semantics. This refocus underpins an entire revolution occurring in health care delivery today. The Medicare program, and the private insurers, are moving towards value-based payment and delivery models. Payments are being tied to outcomes – in the traditional FFS programs, providers are losing payments if their patient is being hospitalized for avoidable adverse events. People are starting to recognize that adverse events that lead to hospitalizations, such as dehydration, nutritional deficiency falls with injury, or those resulting from medication omission or complications could be avoided with a little extra follow up support. By holding hospitals and others responsible for avoidable hospitalizations that occur within 30 days after hospital discharge, the providers are being forced to look beyond their front door in thinking about care coordination and patient support.
Many initiatives are underway in both the public and private sectors to redesign healthcare and achieve the Triple Aim of better population health, improved outcomes, and lower costs. Both public and private insurers have established Accountable Care Organizations and Primary Care Medical Homes which give physicians and hospitals responsibility for monitoring outcomes and reducing rehospitalizations as well as managing chronic disease and providing preventive services. Certain programs have targeted the dual-eligible populations to integrate financing and delivery for those who meet poverty guidelines. The Administration for Community Living has supported initiatives to improve transitions of care for older Americans and disabled populations. Many resources are being brought to bear…..if you or your physician meets one of the qualifying program parameters.
But the problem is broader than a select payer group. The LTC population are the frailer, less helthy portion of the Medicare population, not necessarily impoverished but most likely among the providers’ “high-cost” beneficiaries. These Medicare populations are using the physicians and hospitals participating in the ACOs and other redesign initiatives. And all of them are using hospitals that are now being held accountable for reducing rehospitalization and infection rates as part of their traditional fee for service quality monitoring programs.
These value-based approaches or outcomes-oriented incentives are changing how providers practice medicine and giving them an incentive to be accountable for the patient outside of the office visit or hospital admission. This is changing the conversation.
Who are the LTSS populations?
Long Term Care has been synonymous with dual-eligibility for years because Medicaid has traditionally been the primary insurer for LTSS services – nursing facilities, personal care services, group homes and other non-medical supports that Medicare and other insurers do not cover. Almost 70% of Medicaid spending for LTSS populations is for LTSS. Impressively, the majority of these costs today are for community-based services such as HCBW, Personal Care, and to a smaller extent home health. But the population is aging and not all people who need LTC services are impoverished. Some are just older or disabled and need some support to remain safely in the community. The Administration for Community Living sponsors support services for all older adults and persons with disabilities, regardless of income or other “qualifying” medical conditions. But the funding is quite limited.
So who are these populations? They tend to have multiple chronic conditions (e.g., high blood pressure, heart disease, diabetes, cancer, and stroke) and their Medicare costs per person increase with the number of conditions. About 37 percent of all Medicare beneficiaries have at least 4 chronic conditions. The average Medicare cost/person with 4-5 chronic conditions is $12,174, compared to only $9,738 per average beneficiary. Among the 14% of beneficiaries who have 6 or more conditions, Medicare spends over $32,000/year.
Much of the spending is accounted for by hospitalizations. On average, only 20 percent of the beneficiaries have a hospitalization during the year. However, among those with 4 -5 chronic conditions, almost 30 percent are hospitalized at least once and among them, 3 percent have 3 or more hospitalizations / year. Those with 6 or more chronic conditions experience even higher hospital use. And the rehospitalization rate within 30 days is even higher – ninety percent of the readmissions were among those with 4 or more chronic conditions; the majority (70 percent) were just among the 14 percent of beneficiaries with 6 or more conditions. These are the high cost Medicare populations.
Reframing the Issue.
Instead of asking how we can fund LTC services, it is time to ask how providers can use LTSS providers to better manage their Medicare high-cost populations. In this new accountable world, providers have the same incentive as insurers to manage their patients’ outcomes for 30 days following hospital discharge, the period in which most rehospitalizations occur (Gage et al, 2009). Instead of ending their fiscal responsibility at the hospital front door, providers now are more at risk for the high cost readmission ($11,594 per stay – almost fifty percent more than the index admission cost). Hospitals, physicians, ACOs now have an incentive to put a case manager in place to avoid being penalized for a readmission. Alternatively, they could contract out with a community-based organization to send a nurse into the home following their discharge to the community ($114.35/visit). If the patient qualifies for home health, they could refer all high risk beneficiaries into home health for at least one visit following discharge home. The cost of one nursing visit ($115) relative to the cost of an avoidable hospitalization ($8,531 for an index; or higher for a readmission) seems like a wise investment in improving the quality of care provided to these populations.
By reframing today’s discussion to focus on helping the Medicare providers achieve better outcomes, and ultimately improve their payments for their high cost populations, the medical providers will have an incentive to work with the LTC community to reduce the likelihood of hospitalization. If paying them a relatively minimal amount protects the hospital or ACO from a much larger loss when payment adjustments are made to reflect quality of care, they should have an incentive to create the system that is needed to improve outcomes. This is truly patient-centered care and not silo-based approaches to care redesign. Bundled payment initiatives and other risk-transferring models, such as ACOs and Primary Care Medical Homes are providing these incentives to better build a patient-centered care system.
A small number of these accountable entities are beginning to work with the broader health and LTSS communities to establish programs that target high risk populations who may be more likely to be rehospitalized within 30 days. Some are starting to work with LTSS providers in the CMS Innovation Models, ACL Transition initiatives, and Aligning Forces for Quality, among other efforts. But more can be done in the existing FFS environment to manage these high cost populations, and in doing so, the support system for the LTC population will improve without increasing public program costs. For a small cost (1-2 nursing visits), hospitals and other accountable entities can (and are) contracting with community-based organizations to:
- make sure the patient is safely in their home after discharge,
- reconcile the patient’s discharge medications with pre-admission medications
- check the home for structural barriers that may cause falls (loose rugs, etc.)
- set up transportation for follow-up physician visit and grocery shopping
- review the patient’s understanding of their discharge instructions using tools such as the CTM-3 or the PEM
- send the primary care physician information about recent services the patient has received and any changes in medication
- call in new prescriptions and notify the pharmacist of changes in medications
Changes to the Medicare Benefit?
Your question to me today was focused on whether the Medicare SNF or HH benefits needed modification to better meet the needs of these populations. Traditionally, these benefits have always been strictly defined as acute care benefits. Medicare covers up to 100 days in the SNF following a related hospitalization of at least 3 days in length. Some argue that the SNF services could be more effectively used as a lower cost inpatient resource for nursing or therapy if the 3 day prior stay were removed. This argument suggests that eliminating the 3 day prior hospitalization rule would allow short term inpatient nursing to be provided in a SNF without requiring a higher cost hospitalization. A change like this would only benefit patients whose medical needs require 24 hour nursing for a period but not at the intensity of medical management given in a hospital. Many nursing homes have shown through managed care contracts that they can provide this level of nursing and avoid hospitalizations. However, removing the 3 day prior hospitalization requirement could also result in a “woodwork” effect if LTC beneficiaries living in NFs could flip over to Medicare coverage at-will. Any changes in this requirement would need to be restricted to patients meeting a certain level of medical complexity to qualify for SNF services without a hospitalization. This type of change would be beneficial to these LTC populations by eliminating the need for the ambulance ride, the long wait in the emergency room upon arrival, and the potential stay under observation that currently does not count as a trip to the hospital.
Modifying the home health benefit is even more complicated. The benefit has always provided skilled nursing, physical therapy, occupational therapy, speech pathology, social work, and aide services to any beneficiary who is homebound and has at least intermittent skilled nursing or physical therapy needs and a physician’s certified referral. These patients do not have to have been hospitalized; in fact, over half (64 percent) of the home health population are direct admits from the community. Also notable is the relatively low proportion of the community entrants who are Medicaid-eligible. While the duals account for a much larger share of the community entrants than the PAC users, they still represent less than half of the community entrants (42 percent). This underscores that the population needing LTC services are not all impoverished or able to gain services through Medicaid programs.
The community entrants appear to have higher LTSS needs than the PAC admissions. For example, aide visits account for a majority of the services provided in 11 percent of the community admit episodes while only 4 percent of the PAC episodes have a majority of unskilled services. The community entrants also have a higher proportion of dementia conditions (29 percent compared to 21 percent of PAC). These data suggest the Medicare home health benefit is already a strong part of the LTSS system. Any expansions of the benefit to further support the LTC populations would increase the Medicare HH budget but may be offset by reducing hospitalizations among high risk populations.
In closing, I would encourage the continued movement toward patient-centered accountability which is moving away from insurance-based constraints and introducing incentives for providers to better coordinate care across the continuum. Incorporating the community-based providers in local discussions about health care outcomes at the community level will benefit all parties – the accountable providers, the beneficiaries, the caregivers, and the payers.