Hospital at Home-Plus: A Platform of Facility-Based Care

Abstract

OBJECTIVES:

To describe the evolution of a hospital at home (HaH) program to a HaH with a 30-day posthospitalization transition period (HaH-Plus) and results of a retrospective review of cases.

DESIGN:

After launching HaH-Plus, we used the same interdisciplinary clinical team to provide acute home-based care for a broader range of home-based acute-level services than originally conceived in the Hospital at Home model. These included a palliative care unit at home (PCUaH), an observation unit at home (OUaH), a post-acute care rehabilitation at home (RaH), and a program for the hospital averse – those patients needing to be in the hospital but who refuse.

SETTING:

Urban health system.

PARTICIPANTS:

Individuals 18 years or older residing in specified catchment area with Medicare fee-for-service or accepted Medicare/Medicaid Advantage plans requiring facility-based care.

INTERVENTION:

Provision of facility-based acute-level care at home to 685 participants.

MEASUREMENTS:

Length of stay, readmission, and mortality.

RESULTS:

HaH-Plus cared for 685 individuals. The PCUaH had the oldest participants (mean age 87), and all groups were predominantly female and dually eligible for Medicare and Medicaid. Diagnoses and length of stay were similar in all groups except that those in RaH had a larger group of diagnoses, than those accepted in to HaH-Plus and those in OUaH had a shorter stay. Rate of readmission was highest for RaH (19%). Mortality during the active treatment episode was highest for PCUaH and hospital averse as compared to HaH-Plus, OUaH and RaH.

CONCLUSION:

Providing a broader range of facility-based care in the home has significant advantages for patients and increases the scalability of HaH. Developing a spectrum of services was possible by leveraging a robust,24-hour HaH team. Community- and home-based care could become a greater part of the U.S. healthcare system if a platform of HaH services along with advances in technology and payment models were developed.

INTRODUCTION

Hospitalization is associated with poor outcomes for older adults.^1-3 Individuals transitioning from 1 site of care to another are at risk of errors, which can result in morbidity and hospital readmission.^4-6 Substantial efforts have been made to make the hospital environment safer for older adults;^7-9 a complementary approach has been to move acute care into the home.

Studies have shown that individuals have a better care experience and lower costs, receive better-quality care, and have fewer hospital-associated complications with Hospital at Home (HaH) than with usual hospital care, especially older adults.^10-12

In 2014, Mount Sinai received a Health Care Innovation Award from the Center for Medicare and Medicaid Innovation (CMMI) to implement and study HaH, evaluate its clinical effectiveness, and develop a payment model for HaH in fee-for-service Medicare.

The HaH leadership team developed and implemented HaH-Plus, which consists of HaH (in which people receive hospital-level services at home as a full substitute for the care they would have otherwise received in the hospital) along with care during a 30-day post-acute care follow-up period. HaH-Plus was built on the foundation of a home-based primary care (HBPC) practice. Shortly after launching HaH-Plus, we realized that there was an opportunity to use our acute home-based medical care expertise to provide a broader range or platform of services in the home than would usually be provided in a healthcare facility. Doing this has significant advantages for patients and substantially increases the scalability and effect of HaH. The aim of this article is to describe the development, evolution, and out-comes of the HaH-Plus platform.

METHODS

Initial Program Design – HaH-Plus

The program was originally conceived solely as a substitution HaH care delivery model. Program clinicians screened all patients in the emergency department (ED) or observation unit during program admission hours (8 a.m.—3 p.m., M-F) at Mount Sinai Hospital, a large, urban, quaternary teaching hospital.

Individuals enrolled in HaH-Plus were aged 18 and older, had fee-for-service Medicare or 1 HaH contracted managed care insurance plan, and lived in Manhattan. They were medically eligible for the HaH-Plus program if they had a qualifying acute medical condition that required inpatient care: community-acquired pneumonia, congestive heart failure, chronic obstructive pulmonary disease or asthma, urinary tract infection, cellulitis, deep vein thrombosis or pulmonary embolism, hyperglycemia, or hypertensive urgency. They had to meet the Milliman Care Guidelines for medical necessity for hospital admission.¹³ Individuals were ineligible if they were likely to require intensive care or additional monitoring that could not be provided in the home, which was assessed through a home safety questionnaire. Individuals who could not care for themselves were required to have family or aide support. The CMMI grant funded salaries of all program staff and services that vendors provided, diagnostic and laboratory services, supplies and medications, and durable medical equipment (DME). Other costs were billed to Medicare (e.g., emergency department charges, outpatient medications filled during the stay, other post-acute care).

Eligible individuals identified in the ED who consented to HaH-Plus care were transported home in an ambulance or car service. They received daily physician or nurse practitioner (NP) visits until they were deemed clinically eligible for discharge. They received at least 1 but often 2 daily visits from a registered nurse (RN), according to their plan of care. They also received a home visit from a social worker on admission and as needed thereafter. Additional services, such as home health aides and occupational and physical therapy, were provided in the home as needed.

Diagnostic studies such as electrocardiograms, laboratory studies, plain films, and ultrasounds and therapeutics such as intravenous fluids or medications, oxygen, and DME were provided in the home. Diagnostic modalities that could not be provided in the home were available using the hospital’s outpatient services. HaH-Plus staff were avail-able 24 hours a day by telephone, and in the event of an urgent issue, the team physician could request an urgent visit from a physician, NP, or RN. In addition, providers who were certified to direct paramedics could request community paramedicine encounters to provide urgent visits over video conference with the attending physician on call and the subject and paramedic in the home. This was used infrequently (6 times) during this period. If a person was clinically unsafe to remain at home, they were returned to the hospital.

After participants were deemed clinically eligible for discharge from acute HaH (the active phase), they were followed for an additional 30 days as part of HaH-Plus (post-active or post-acute phase). During HaH-Plus, participants were stratified into high and low risk of readmission and received varying levels of RN and social work support, including telephone and in-person visits for 30 days, participants could also be referred to a certified home health agency if required. High-risk individuals included those who were high users of the ED or hospital before HaH, had unmet activity of daily living or instrumental activity of daily living needs at time of discharge, or were dually eligible for Medicare and Medicaid. Clinicians could also designate individuals as high risk. The rest of the team remained available to participants and their primary care providers with 24-hour in-person and telephone support and the capacity to perform a home medical visit, diagnostics, and therapeutics to complete recovery from the acute illness and ensure a safe transition to primary care.

The HaH-Plus Team

The team consisted of clinical staff and a variety of community-based vendors and built upon the strengths of an existing HBPC program.¹⁴ Using the resources and relationships already established allowed the HaH-Plus team to work with clinicians skilled in providing care in the home. In addition, the team used vendors familiar with HBPC.

Evolution of the HaH-Plus Program

The HaH-Plus program saw its first patient in November2014. Over the following 3 years, it evolved to include a suite of home-based medical care services that would otherwise have been provided in facilities. This evolution grew out of needs we identified for patients and the program. The same team of clinicians and administrative staff treated all service lines (Figure 1).

Figure 1:

Diversifying the Hospital at Home - Plus Suite of Services: Current and Future

SNF: Skilled Nursing Facility

Broadened the Scope of HaH-Plus Qualifying Conditions

Outpatient and ED clinicians referred multiple individuals who wanted to participate in HaH-Plus but did not have HaH-Plus qualifying diagnoses or meet clinical exclusion criteria. We realized that the infrastructure needed to provide safe care for the initial set of qualifying diagnoses could be expanded to other diagnoses with similar treatment approaches. We broadened the list of qualifying acute diagnoses gradually over the 3 years and it now includes over 69 Medicare Severity Diagnosis-Related Groups (MS-DRGs).

Expanded Hours and Sites

The program launched at 1 hospital with limited hours of 9 a.m. to 3 p.m. As the initial pilot was expanded, HaH-Plus expanded admission hours to 5 p.m. and initiated Saturday admissions. One year after launch, the program expanded to2 other hospitals in the system, with a fourth hospital added1 year later. The larger team, with expanded hours and number of participating hospitals, allowed greater infiltration of practice culture across the health system.

Palliative Care Unit at Home

We received referrals for individuals with end-stage diseases who were near the end of life and not receiving hospice care. These individuals did not fit HaH-Plus clinical inclusion criteria because their acute illness acuity was too great, but many of them desired hospital-level care but did not want to be admitted to the hospital. In deference to their care preferences, we admitted them to HaH-Plus and focused care on symptom management, quality of life, and working with them and their families to enroll in hospice when appropriate. As expected, many of these individuals died during the HaH-Plus stay or during the subsequent 30 days. This clinical experience led us to formally develop the first service line to be borne out of HaH—the palliative care unit at home (PCUaH).

Hospital Averse at Home

As awareness of the HaH-Plus program spread, clinicians called to request direct admission to HaH-Plus from the home. Individuals with clear diagnoses and sufficient examination were admitted into HaH-Plus, but there were many individuals who were too acutely ill to qualify medically for HaH-Plus but refused evaluation in the ED or, if already in the ED, refused to be admitted. This prompted development of Hospital Averse at Home. We accepted these individuals if we felt that we could temporize a clinical situation that might ultimately require a hospital stay. These individuals’ outcomes were tracked separately because a successful HaH-Plus stay could end in hospitalization.

Observation Unit at Home

One barrier to program recruitment was implementation at Mount Sinai of an observation unit. Individuals who would have traditionally required short hospital stays and who might have been suitable for HaH-Plus were being admitted to the observation unit; many were eventually admitted fora traditional hospital stay after prolonged observation unit stays. To address this, we developed an observation unit at home (OUaH) program to target individuals in the ED who were most likely to have a prolonged observation unit stay that was likely to convert to a hospital admission. If they required a longer stay, we converted them from OUaH to a HaH-Plus admission in a manner that paralleled the approach hospitals take to these individuals.

HaH-Plus Extended Hours

The second recruitment challenge was that a substantial number of individuals were being missed for HaH-Plus admission because they were being admitted to the hospital at night and on weekends. We were able to expand our daytime admitting hours into the early evening, although later evening and nighttime admissions were still missed. To remedy this, we hired physician assistants who worked in the ED to improve case identification and to hold individuals over in the ED fora HaH-Plus admission the following day.

Post-Acute Rehabilitation at Home

Finally, we developed post-acute rehabilitation at home (RaH), which targeted individuals slated for a subacute rehabilitation stay after an acute hospitalization regardless of their hospital diagnosis. During this project, we did not have individuals in the HaH service lines transfer to the RaH program. Rather than transfer from the acute hospital to a skilled nursing facility, where subacute rehabilitation is provided, such individuals would return home and receive home-based intensive physical therapy up to 6 days per week. In addition to physical, occupational, and speech therapy, they received clinician, nursing, and social work visits as needed. They were then transitioned to a “plus” period to complete 30 days of participation and received the same support as HaH-Plus participants with RN and social work follow-up, both in-home and telephonic, with physicians and NPs available as needed to ensure recovery from their acute illness and safe transition to primary care.

Data Collection

A retrospective review was conducted of everyone enrolled in the program from November 14, 2014, through August31, 2017. The same team of inpatient coders that code all admissions for Mount Sinai Hospital into Diagnosis-Related Group categories coded all charts. Admission diagnoses and clinical outcomes, including length of stay and30-day ED visit and hospital admission rates, were recorded. The Icahn School of Medicine at Mount Sinai institutional review board approved all data collection.

RESULTS

HaH-Plus cared for 685 individuals over 34 months. These individuals are described in Table 1 according to service line; 321 were cared for in the traditional HaH group(acute hospital), 264 in the RaH group, 33 in the Hospital Averse group, 41 in OUaH, and 26 in the PCUaH. Most were female and elderly, those in the PCUaH were the oldest (mean age 87). A majority had Medicare, and a high percentage were dually eligible for Medicare and Medicaid. Individuals admitted to OUaH who subsequently required longer treatment were converted to the HaH-Plus group;23 (35.9%) underwent this conversion and are counted in the acute group.

Table 1:

Demographics of the patients in each service line

		Acute Hospital (n=321)	Observation Unit at home (n=41)	Palliative Care Unit at home (n=26)	Hospital Averse (n=33)	Rehab at home (n=264)
Mean Age (sd)		77.0 (16.4)	81.7 (11.5)	87.2 (10.3)	85.8 (12.8)	84.5 (8.7)
Age 18-49 (%)		6.2	2.4	0	0	0.4
Age 50-69 (%)		19.3	7.3	3.9	15.2	4.9
Age 70-89 (%)		51.1	61.0	46.2	33.3	67.4
Age 90+ (%)		23.4	29.3	50.0	51.5	27.3
Female (%)		72.0	58.4	65.4	69.7	65.9
Payor (%)	Fee-for-service Medicare, non-duals	39.6	43.9	57.7	45.5	77.3
	Dual Eligible	42.4	43.9	42.4	51.5	22.7
	Medicare Advantage	8.1	4.9	0	0	0
	Medicaid managed care	8.4	7.3	0	0	0
	Private*	1.6	0	0	3	N/A

^*we accepted one private insurance from 11/14-6/17 for all but the Rehab at home

Individuals were admitted with similar diagnoses in the acute, OUaH, PCUaH, and Hospital Averse groups, although at differing rates (Table 2). Those admitted to RaH had a broader range of diagnoses because of the nature of individuals receiving rehabilitation care. Those admitted to HaH-Plus were triaged and admitted from the ED, 1 of the hospital ambulatory sites, or directly from home after referral from their primary care provider. Individuals in the HaH (63.2%) and OUaH (56.1%) groups were most often admitted from the ED, whereas those in the PCUaH (84.6%) and Hospital Averse (72.7%) groups most often came from home, reflecting their avoidance of the ED. The disciplines caring for the participants were similar across the groups except that those in the RaH used more physical therapy.

Table 2:

Most common diagnoses, location of triage and resource utilization in each service line

		Acute Hospital (n=321)	Observation Unit (n=41)	Palliative Care Unit (n=26)	Hospital Averse (n=33)	Rehab at home (n=264)
Most Common Diagnoses (%)	UTI	21.8	12.2	11.5	0	8.0
	Pneumonia	18.7	17.1	38.5	15.2	6.8
	Cellulitis	14.0	2.4	3.9	15.2	1.9
	Dehydration	12.5	4.9	23.1	21.2	0.4
	CHF	11.5	17.1	15.4	21.2	2.4
	Asthma	4.7	7.3	0	3	0
	COPD	7.5	7.3	0	0	0.8
	Mechanical fall	0	0	0	0	15.6
	Musculoskeletal pain	0	0	0	0	4.9
	Other medical	9.3	31.7	7.7	24.2	42.4
	Other surgical	0	0	0	0	15.9
Triage Location (%)	ED	63.2	56.1	15.4	12.2	14.0
	Clinic	5.3	14.6	0	12.1	0
	Home	21.2	29.3	84.6	72.7	2.3
	Hospital	0	0	0	0	83.0
	Observation unit	10.3	0	0	3	0.8
Number of Visits during the Active Period, mean (SD)	Provider (MD/NP)	3.8 (1.9)	1.9 (0.4)	4.9 (3)	4.4 (2.9)	2.2 (0.8)
	Nursing	3.2 (2.7)	0.6 (0.7)	3.9 (3.6)	4.5 (5)	2.7 (2.3)
	Physical Therapy	0.2 (0.5)	0 (0.2)	0 (0)	0.1 (0.3)	10.4 (5.3)
	Occupational Therapy	0 (0.1)	0 (0)	0 (0)	0 (0)	2.4 (3.1)
	Speech Therapy	0 (0.1)	0 (0)	0 (0)	0.1 (0.3)	0.3 (1.1)
	Social Work	0.7 (0.6)	0.3 (0.5)	0.5 (0.5)	0.6 (0.7)	1.2 (0.8)
Number of Visits during the Plus Period, mean (SD)	Provider (MD/NP)	0.3 (0.3)	0.2 (0.2)	0.5 (0.5)	0.2 (0.2)	0.1 (0.1)
	Nursing	0.7 (0.7)	0.8 (0.8)	0.8 (0.8)	0.8 (0.8)	0.2 (0.2)
	Physical Therapy	0 (0)	0 (0)	0 (0)	0.1 (0.1)	0.1 (0.1)
	Occupational Therapy	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)
	Speech Therapy	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)
	Social Work	0.3 (0.3)	0.5 (0.5)	0.2 (0.2)	0.4 (0.4)	0.1 (0.1)

UTI: Urinary Tract Infection, CHF: Congestive Heart Failure, COPD: Chronic Obstructive Pulmonary Disease, ED: Emergency Department, MD Medical Doctor, NP Nurse Practitioner

Length of stay was similar for the acute, (3.4 ± 2.1) PCUaH (3.9 ± 2.7), and Hospital Averse (3.4 ± 2.8) groups, whereas average length of stay was 1.0 ± 0.3 days for the OUaH and 14.0 ± 6.4 days for RaH (Table 3). If the individuals subsequently required a higher level of care than could be provided at home during active treatment, they were transported from home to the ED. Participants and families could also request transfer to the ED if they wished, and they could be returned back home from the ED once they were stabilized or they could be admitted. Of the individuals admitted to HaH-Plus, those who were Hospital Averse were the most likely to be admitted to an inpatient service (15.2%); none of those in the PCUaH group were sent to the hospital. This did not apply to those in the RaH group because they had already been discharged from their acute inpatient stay.

Table 3:

Length of Stay, Escalations and readmissions of the service lines

	Acute Hospital (n=321)	Observation Unit (n=41)	Palliative Care Unit (n=26)	Hospital Averse (n=33)	Rehab at home (n=264)
LOS, days (sd)	3.1 (2.0)	1.0 (0.2)	3.8 (2.8)	3.4 (2.8)	14.0 (6.4)
Escalations (%)	8.7	2.4	0.0	15.2	N/A
30-day ED visits (%)	5.7	4.9	4.2	6.7	4.7
30-day Hospital Readmissions (%)	9.2	0.0	0.0	10.0	19.0
Mortality-Acute (%)	0.9	0.0	7.7	6.1	1.1
Mortality- 30 day postacute (%)	2.8	4.9	37.5	3.3	2.3

LOS: Length of stay

The rate of ED visits during the 30 days after discharge was relatively similar across the service lines, although the rate of inpatient readmission during the 30 days was highest for RaH (18.9%). There were no 30-day readmissions in the OUaH or PCUaH group. Mortality during the active treatment episode were highest for those in the PCUaH and Hospital Averse groups. Within the 30 days after an active treatment episode, mortality was highest for those in the PCUaH group (37.5%) (Table 3).

DISCUSSION

Building HaH infrastructure to serve a single purpose such as HaH-Plus limits HaH dissemination potential and was a motivating force to develop the HaH-Plus platform. Development of a spectrum of home-based services that would otherwise be provided in facilities allowed the program to diversify patients while leveraging a, 24-hour acute care team that could respond to a breadth of clinical problems across populations. The broad range of HaH-Plus services increases its relevance and appeal to health systems looking to improve clinical and financial outcomes and makes the program a valuable partner to health systems. This allows the program to be customized to local needs and stakeholders. The broader range of home-based services allows for more flexibility, a more person-centered approach, and a higher program volume, which makes staffing and budgeting more predictable.

The HaH-Plus platform was initiated by substituting HaH-Plus for an inpatient admission. Identification of gaps in the current healthcare delivery system that could be filled by caring for people at home with the assets of HaH led to the development of additional models in the platform. For example, PCUaH developed out of the needs of individuals who met hospice criteria, had not elected hospice, and needed higher levels of clinical care. The service allowed them to receive necessary treatments and facilitated goals-of-care discussions with families and caregivers while still providing care consistent with their poor prognosis. We are developing a pediatric HaH program and a postsurgical HaH for surgical procedures that require postoperative hospitalization.

As this suite of services grows, it is important to remember that each subgroup may have different expected outcomes. In our example of PCUaH, none of these individuals were sent to the hospital, consistent with the program goal of caring for people at home at the end of life. Similarly, the much higher 37.5% mortality for those in PCUaH reflects their poor prognosis at enrollment. Despite years of data and growing national interest in HaH, implementation of HaH in the United States has been limited. Lack of a payment model in fee-for-service Medicare has hampered dissemination, although building the appropriate HaH clinical and administrative infra-structure has also been a challenge.

Successful treatment at home of individuals with an acute illness requires a skillset that includes acute hospital care, home-based care, and a strong focus on coordination of care and transitions. Programs require an infrastructure that can respond to abrupt changes in clinical status and needs when certain clinical resources are not readily available. Leveraging the experience of a home-based primary or palliative care program can help create that infrastructure. Despite Mount Sinai’s long experience and presence in home-based care, we struggled to identify the appropriate individuals for this HaH model. We had to refine our entry and exclusion criteria continuously and eventually formed multiple new arms of the program, as described above. To meet the varied needs of individuals receiving hospital-level care at home, HaH programs should have clinicians with knowledge of care delivery in the home and sufficient numbers of them to provide that care around the clock. Other challenges we faced included appropriate team staffing and productivity and how to balance the unpredictable and immediate enrollment needs of our patients with a static staffing model. Furthermore, many of the ser-vices provided depended on relationships with various vendors, and these contracts needed to be updated and refined as our models evolved.

As the U.S. healthcare system shifts and the debate continues about access, value, and efficiency, we believe that HaH will play a role in the ongoing discussion about the future of U.S. hospitals. The suite of services in the HaH-Plus platform at Mount Sinai grew organically and was funded completely by our CMMI grant in its first 3 years. As this drew to a close, working toward sustainability has been a major part of our work, and we continue this, working with national and regional payors and our home institution. As HaH grows, HaH adopters around the country and around the world will develop additional variants of the HaH model of care. A diverse HaH platform will serve a future vision of the hospital in which the traditional bricks and mortar facilities will be large intensive care units and places where people go for high-level procedures; much else will be provided in the community and abetted by future development in technology, payment models, and a change in the culture of the U.S. healthcare system that values community- and home-based care. This shift of care in the community will require training our providers, aligning with our community partners, and shifting the current framework that usually results in hospitalization. Our experience with and growth in our Hospital at Home model has shown that this shift is feasible, safe and provides care that justifies the operational and cultural overhead needed to begin.