Phase I: patient interviews

In 2005-06, we conducted interviews with 50 FFS Medicaid patients who were identified with predictive methodology used by our current Hospital to Home program. This process allowed us to define remediable medical and social risk factors that might be addressed by an intervention to improve health and reduce admissions and costs among the frequent hospital users identified. Through this process we were able to identify hospitalization risk factors among frequently admitted patients identified by the algorithm. This approach was paramount in informing our intervention model on ways to improve care for this patient population.

During the interviews we discovered that the majority of patients had no usual source of primary care; the ED was the single most common source cited for care. Unstable housing, substance abuse, and mental illness also emerged as strong drivers of hospitalization. The 50 interviewed patients consumed a per-person average of nearly $40,000 in Medicaid funds during the prior year, most of which were spent on hospital admissions. As a point of reference, mean New York State Medicaid payments for all Medicaid enrollees totaled $7,583 during the same year (KFF 2007b). Prior research has demonstrated that such patients will generate even higher costs in the 12 months following discharge if no intervention is undertaken to reduce health services use (Billings et al. 2006). This was indeed the case for our study population whose costs more than doubled in the year following study enrollment (Raven et al. 2009).

Phase II: pilot, small-scale, local intervention

Based on our interview data, in addition to site visits to and discussions with leadership from similar programs around the country, we developed a program model we felt would best serve the needs of the patient population, and decided to pilot it for a small number of patients to determine feasibility and effectiveness. The pilot program was structured to begin at the point of enrollment, whether at the patient’s bedside in the hospital, or in a homeless shelter utilizing a flexible, patient-centered, intensive care management model with a multidisciplinary team approach. Although the program is tailored to the needs of each patient, it is designed to be responsive to local institutional patient flow, and thus readily adaptable to other settings.

Care management teams composed of social workers, community based care managers (CBCMs), and dedicated clinicians work to meet patients’ needs in both health care and community settings. The model incorporates motivational interviewing, harm reduction, access to housing through our community partners utilizing a Housing First approach (Larimer et al. 2009), and provides cellular phones for patients when needed to better enable regular communication with program staff.

For the pilot, we enrolled 19 Medicaid FFS patients between 18 and 64 years of age who were admitted to our urban public hospital and identified by Billings’ predictive case-finding algorithm as being high-cost and high-risk for future hospital readmission. All were male, a majority were substance users (17 of 19), and almost half (8 of 19) were homeless. They were evaluated to identify needs for things such as transportation to and advocacy during medical appointments, mental health and substance use treatment, and home visits. A community housing partner, Common Ground Community, initiated housing applications in-hospital for homeless patients. CBCMs facilitated appropriate discharge plans, then worked closely with the patients in the community using a harm reduction approach.

The 19 patients had a total of 64 inpatient admissions in the 12 months before the intervention versus 40 inpatient admissions in the following 12 months, which is a 37.5% reduction. A majority (11 of 15, or 73.3%) had fewer inpatient admissions in the year after the intervention compared to the prior year. Overall, ED visits also decreased, while outpatient clinic visits increased. Yearly Medicaid hospital reimbursements declined an average of $16,383 per patient. Because it was a pilot, the program was not to scale. However, subtracting our overall annualized pilot intervention costs from annual per-person Medicaid savings, the pilot resulted in over $5,000 savings to Medicaid per patient. Given these promising results, we have now expanded the model to serve more patients at additional hospitals to see if the pilot’s success can be replicated.

New York State Department of Health chronic illness demonstration project: hospital to home (H2H)

Chronic Illness Demonstration Project

Building on the momentum from the pilot work, in 2007, the New York State Medicaid Office of Health Insurance Programs decided to fund seven Chronic Illness Demonstration Projects (CIDPs) across New York State. The goal of each three-year CIDP is to find and enroll FFS Medicaid patients identified as high-cost and high-risk for hospitalization in the 12 months following discharge, and to work with patients to improve their health and well-being while reducing costs to Medicaid, primarily through reductions in costly hospital admissions. The New York City HHC was awarded one of six CIDP contracts, in essence providing an opportunity to create and test the effectiveness of our program model on a much larger scale for some of the most vulnerable patients in New York City.

In April 2009, Hospital to Home was initiated by the HHC and implemented by three of its member hospitals: Bellevue Hospital Center in Manhattan, Woodhull Hospital in Brooklyn, and Elmhurst Hospital in Queens. While all a part of the same corporation, each of these facilities differs from one another in terms of layout, population served, surrounding community resources, and embedded hospital resources. However, the program model, messaging, philosophy, and staffing patterns are uniform across facilities. The experience of implementing a single program across three individual facilities has been, and will continue to be, an essential part of determining whether such a program can be implemented at multiple distinct sites (and later disseminated to other facilities) while maintaining its singular programmatic goals and policies.

Why is this important for the future of health care and Medicaid?

Currently, there is much discussion about the importance of the medical home and the formation of Accountable Care Organizations. If such models for care are to succeed for health care providers serving complex, high cost patients, strategies such as those employed by H2H will be needed. Providers will increasingly be responsible for their patients across a continuum of care that extends beyond the walls of the clinic or hospital. To be competitive, they will have to grapple with the issue of following patients outside the traditional confines of the health care system into the community, into their homes or shelters, necessitating interaction and collaboration with community-based providers, and if they exist, family members and friends. This is a notable difference from the current standard of care. Providers and policy makers in areas with high concentrations of Medicaid-insured and other potentially vulnerable populations must begin to think now about how to contain costs and optimize health care utilization and outcomes for such patients if they are to have a chance to succeed.

Hospital to Home addresses exactly these issues. We prioritize two key concepts: care coordination and accountability. Our program motto is “we’ll meet you where you are” and we mean this both literally and figuratively. Our staff conducts much of its work not in the hospital, but in the community, aiming to surround patients with needed services and support as they move in and out of the health care system.

Health system deficiencies H2H addresses to coordinate care

As previously mentioned, our H2H enrollees are algorithmically identified FFS Medicaid patients with many health and social needs, and serious health and social stressors that interfere with their ability to interface effectively with the complex system on which their health and social well-being depend. H2H aims to help patients overcome specific inadequacies in the health system that fragment care and contribute to health care system inefficiency.

Currently, much of health care is siloed and commonly not co-located with behavioral health services (Druss & Newcomer 2007; Druss & von Esenwein 2006). Patients with mental illness have unaddressed chronic disease and preventive health needs (Raven et al. 2009), and organizations governing substance use services and mental illness, which often share patients with hospitals and health care systems, are distinct from one another and subject to restrictive funding streams that often make management of co-occurring disorders difficult (KFF 2007b). Perverse incentives that financially reward hospitalizations and underpay outpatient and primary care provide inadequate incentives to prioritize basic care coordination and preventive care (Larimer et al. 2009). Finally, a general lack of financial incentive for provision of post-discharge follow-up in the community, combined with an often precariously housed patient base, make continuity of care difficult (Druss & Newcomer 2007). As these realities are unlikely to change in the near future, we factored these challenges into our program design with the intent of eventual dissemination.

H2H rationale and program model

The essential mechanism by which we function is to specifically address these system misalignments at the individual level by engaging patients with a consistent, multidisciplinary care management team (along the lines of the Medical Home model) including a key “go-to” person (the CBCM), so that care can be managed and coordinated across multiple locations and providers. The CBCM is similar to a “patient navigator” where there is extensive literature that has demonstrated effectiveness. The main role of patient navigators is to provide basic care management, eliminate barriers to care, provide emotional support, and link the patient to appropriate community supportive agencies and resources (Dohan & Schrag 2005; Wells et al. 2008).

Our program philosophy is to “meet patients where they are” both literally and figuratively. As such, care team staff aims to show patients, many of whom have had negative experiences within the health care system, that improved care management and coordination via H2H can allow for a different, more positive experience, both within and outside of the health care system. This helps patients to become more engaged in and hopeful about their own care. Intervention staff focuses not only on care team-identified goals, but also patient identified goals over both the short and long term. By improving care provision via needed coordination and oversight, we anticipate that we will be able to improve health status and patient satisfaction as we demonstrated in our pilot work, while averting hospital admissions that are costly to Medicaid.

H2H staffing

Each of the three participating HHC hospitals is staffed with a social work supervisor (LCSW) who oversees a staff of CBCMs. CBCMs are required to have a minimum of a high school diploma and relevant experience working in direct patient services. CBCMs are required to build their own caseloads, which are capped at 25 patients. They are the most important single point of contact for the patients. In addition, we have a full-time housing coordinator to work with our homeless patients to obtain housing and some degree of dedicated primary care at the participating hospitals.

Once a patient is enrolled, CBCMs work with them to complete an in-depth health assessment to identify immediate, short, and long term goals to help improve health and well-being and optimize health care utilization. From this, a care plan is created. Goals range from obtaining glasses or dentures, to connecting to stable primary care, cutting down on the amount of alcohol consumed, and obtaining housing. We issue cell phones to patients who don’t have other ways to keep in contact with our staff, and we use motivational interviewing techniques and take a harm-reduction approach.

The care plan and interventions needed are tailored for each patient depending on his or her unique needs and desires. Some patients obtain care within HHC, and others outside of it. Many have people who contribute to their care, including case managers at methadone programs or in shelters, psychiatrists, or family members.

Each week, the care team at each hospital, consisting of the CBCMs, social work supervisor, and participating physician, holds care management rounds. During rounds, each CBCM presents their cases which are reviewed and discussed by the team and next steps for the coming week are determined. In addition, each month we hold program-wide care management rounds that include the above mentioned staff in addition to the Housing Coordinator, site coordinators from each hospital, and program management and administration. These meetings provide the team with an opportunity to see one another and exchange ideas. Each site presents a challenging case for group discussion, and we review program updates and any administrative issues.

H2H finances

At present, H2H is not financially feasible for HHC. Ironically, because of the way the health care payment system is currently structured, we are the only CIDP participant who actually stands to lose revenue by fulfilling the CIDP objectives. For hospital systems, admissions (unless they occur within 30 days of discharge, a marker for poor discharge planning and follow-up which is penalized by insurers) are the largest source of revenue for hospitals. Conversely, outpatient appointments, which would be an ideal setting for care coordination and information gathering regarding social risk factors, are reimbursed at comparatively low rates. As a result, to maximize revenue, patients are booked for 15-minute appointment slots, hardly enough time to address a laundry list of health and social issues, let alone coordinate with outside agencies to improve follow-up in the community.

If HHC stands to lose money by cutting Medicaid costs, why would we participate? First, it’s the right thing to do. Currently, many of the patients eligible to enroll don’t receive the coordinated health and social care that could exist under health care reform, and they are suffering for it; the financial incentives to undertake the cultural and logistical changes needed to expand the locus of care simply do not exist. Secondly, when improved payments to health care providers for care coordination and accountability do come to fruition in the next few years, incentives will be more appropriately aligned. Health care provider groups, which could/should include social workers, nurses, mid-level providers, and possibly, partnerships with providers at community-based organizations, will determine what services are needed to optimize patients’ health and social outcomes, and will also bear financial risk for the patients they serve. H2H allows HHC to develop and test, for more vulnerable patients not properly engaged with the system, viable strategies for the future.

How do we get patients to join H2H?

In an ideal world, patients would be eager to find and take part in an organization focused on improving care. While a small number of our eligible patients come to us via response to a mailing, most high-cost, high-risk patients eligible for our program must be found, and it takes some detective work, due largely to the fact that we must rely on whatever basic demographic data is available in the Medicaid database. This leaves us with many inaccurate addresses and phone numbers, or missing data. Once those obstacles are surmounted, upon hearing about what our program does, most patients are interested, and we have a very low refusal rate (under 10%) for participation. As few similar programs currently exist, our program is appealing to most eligible enrollees. In the coming years, however, it is likely that competition for patient participation in similar programs will increase, particularly for the newly insured either through Medicaid or state exchanges.

Many patients are distrustful of the health care system due to past experiences. How do we engage them? First, we hire staff members who want to do this work with this particular patient population, and most are hired based on experience in working with our target population. Many staff are multi-lingual and all have access through our facilities to interpreter services, both on- and off-site 24 hours a day, seven days a week. In addition, our staff undergoes rigorous training in motivational interviewing techniques, recognizing various aspects of medical and social issues they will encounter with our patient population, and our singular program philosophy. Because our motto is “we’ll meet you where you are” staff are trained to listen to and elicit the patients’ needs and goals, rather than impose our own standards onto them. When a CBCM first meets a patient, they are trained to attempt to fill some immediate need for that person, even if it’s just a cup of coffee or an ear to help to establish rapport. CBCMs are also trained to help patients identify small, short term, achievable goals (such as establishing primary care, obtaining eyeglasses, coming to our staff offices) in addition to more challenging long term goals (including weight loss and tobacco cessation) so that small victories can be focused upon and recognized.

In order to help this process, we have a “patient necessities fund” comprised of petty cash that can be used for small purchases to fill immediate needs that would otherwise go unaddressed (such as socks, a calendar, or a coat). Because we anticipated from prior research that a significant portion of patients would be homeless, we contracted with the New York City Department of Homeless Services (DHS) to assist with outreach and enrollment of homeless eligible enrollees, and we employ a full-time Housing Coordinator skilled in outreaching to and working with the unique needs of the homeless.

Our patient alert system allows us to receive real-time e-mail notifications when eligible individuals register at any HHC outpatient clinic, ED, or are hospitalized at any HHC hospital. Because our staff can meet with patients to engage them within the hospital setting, they often have the opportunity to explore the reason for the current visit or prior visits, and to bring an immediate focus to some aspect of their health or care that our program can address. In addition, our staff has a willingness to focus on work that is not health care related. Much of their day-to-day work might consist of tasks not directly connected to the health care system, but that affect our patients’ health.

How do we use technology to coordinate care and increase accountability?

In order to help coordinate care, we rely on technology a great deal. All of our staff have Blackberries that enable immediate e-mail and text services to help coordinate care on password-protected devices. We also provide our patients who do not have phones with mobile phones, and pay for their minutes. Patients must sign a contract upon receipt of the phone for tracking purposes, and staff emphasizes that H2H phones are to be used primarily for staying in touch with the program or related matters. Minute usage and potential overages are monitored by our administrative staff, and overages are viewed as an opportunity to discuss any potential issues with the patient. CBCMs train patients on how to use the phones and keep them charged, and they add relevant program numbers into the phone.

As mentioned previously, our patient alert system allows patient-services staff to receive real-time notification of a patient’s presence at any acute-care facility in the HHC system. As a part of this program, we worked with HHCs Medicaid HMO to adapt their current database to work for H2H. The result is a program-specific database (distinct from the patient’s electronic medical record) that our staff uses to enter all information about our enrolled patients. This database allows us to generate reports that assist with staff management and oversight regarding outreach and enrollment, proper and up-to-date documentation, our ability to follow through with initial and ongoing goals and referrals for our patients, and other issues such as billing Medicaid. Our staff also has access to the electronic medical record at their base hospital which aids in care coordination and management.

Patient vignette

Ms. W is a 43-year-old woman with morbid obesity, insulin-dependent diabetes, neuropathy and nephropathy, hepatitis C, hypertension, and asthma. She is a smoker and has a past history of IV drug use; she suffers from depression, and has had suicide attempts in the past. Prior to enrollment she was homeless in a shelter and had frequent hospital admissions for high blood-sugar and related infections, in addition to difficulty remembering to take all of her medications.

The H2H program and her care manager provided her with a cellular phone, outpatient appointment reminders, and frequent moral support. Our Housing Coordinator assisted with housing placement, and she moved into a permanent supportive housing environment where H2H arranged for a home health aide for medication assistance. Through H2H she was assigned to a primary care physician who has linked to the various subspecialists she requires. She has also been attending the monthly H2H support groups. Since enrollment, she has had only one hospitalization, although she continues to have difficulty with adequate blood-sugar control.

Dissemination and replication

By the conclusion of the project, we will have developed a user-friendly description of the intervention that includes staffing patterns and all elements of service delivery that can be disseminated to stakeholders. Lessons learned from this project have already begun to be disseminated to interested parties within and outside of HHC. Importantly, the New York State Department of Health is conducting an independent analysis of the six Chronic Illness Demonstration Projects currently being implemented across the state, and the results will help to inform the future direction of Medicaid regarding this patient population. Finally, because we are partnering with several community-based organizations in New York City, we plan to work with them to disseminate our project results as well. Replication of this project by other institutions will be facilitated by the fact that our intervention is designed to be flexible, enabling providers to tailor the intervention to individual patient needs at different hospitals or health care settings.

References

Billings, J., Dixon, J. M., Mijanovich, T., et al. (2006) Case finding for patients at risk of readmission to hospital: development of algorithm to identify high risk patient. British Medical Journal, 333, 327–332.

Billings, J., & Mijanovich, T. (2007) Improving the management of care for high-cost Medicaid patients. Health Affairs, 26, 1643–1654.

Dohan, D., & Schrag, D. (2005) Using navigators to improve care of underserved patients: current practices and approaches. Cancer, 104, 848–855.

Druss, B.G., & von Esenwein, S.A. (2006) Improving general medical care for persons with mental and addictive disorders: systematic review. General Hospital Psychiatry, 28, 145–153.

Druss, B., & Newcomer, J. (2007) Challenges and solutions to integrating mental and physical health care. The Journal of Clinical Psychiatry, 68(4), e09.

Kaiser Family Foundation. (2007a) Medicaid, A Primer. The Henry J. Kaiser Family Foundation, Melano Park, CA. Retrieved from http://www.kff.org/medicaid/7334-02.cfm.

Kaiser Family Foundation. (2007b) State Health Facts. The Henry J. Kaiser Family Foundation, Melano Park, CA. Retrieved from http://www.statehealthfacts.org/cgi-bin/healthfacts.cgi? action=profile&area=New+York&category=Health+Coverage+%26+Uninsured&subcategory=Health+Insurance+Status&topic=Total+Population.

Larimer, M.E., Malone, D.K., Garner, M.D., et al. (2009) Health care and public service use and costs before and after provision of hoursing for chronically homless persons with severe alcohol problems. Journal of the American Medical Association, 301, 1349–1357.

Raven, M., Billings, J., Goldfrank, L., et al. (2009) Medicaid patients at high risk for frequent hospital admission: real-time identification and remediable risks. Journal of Urban Health, 86, 230–241.

Sommers, A., & Cohen, M. (2006) Medicaid’s High Cost Enrollees: How Much Do They Drive Program Spending. The Henry J. Kaiser Family Foundation, Melano Park, CA. Retrieved from http://www.kff.org/medicaid/7490.cfm.

Wells, K., Battaglia, T., Dudley, D., et al. (2008) Patient navigation: state of the art or is it science? Cancer, 113, 1999–2010.

Stay updated, free articles. Join our Telegram channel

Join