Tracking Patients in Community-Based Palliative Care through the Centers for Medicare & Medicaid Services Healthcare Innovation Project

Background

Community-based palliative care (CBPC) is a longitudinal delivery model that seamlessly integrates interdisciplinary palliative care across inpatient and outpatient care settings to people with life-limiting illnesses. ¹ The impact of CBPC has been shown to improve quality of life,^2–6 lessen symptom burden,^5,6 improve patient satisfaction,^5,7,8 decrease hospitalizations and length of stay,^3,8–12 increase hospice utilization,^8,13,14 and decrease cost of care. ¹⁵ The CBPC model plans for delivering palliative care across multiple care settings, to patients who may transition when care needs change. However, little is known about how CBPC patients first encounter palliative care service and their transition outcomes, including death or discharge.

To date, most data that track patients seen by consultative palliative care services are limited in two ways. First, much of the available data come from patients discharged from inpatient hospital based palliative care programs only. For example, the National Palliative Care Registry from the Center to Advance Palliative Care (CAPC) regularly reports patient tracking data after hospital discharge. CAPC reported of patients enrolled in palliative care programs in small hospitals (<300 beds), 14% transitioned to hospice, 18.3% died in palliative care, 22.3% were discharged to nursing facilities, and 42% were discharged home. Conversely, large hospital based palliative care programs (>500 beds) have an 18% hospice transition rate, 21.4% of patients died in palliative care, 19.1% were discharged to a nursing facility, and 41.3% were discharged home. ¹⁶ Little is known about patient outcomes of those who enter palliative care services from other settings such as home, clinic, or facility (skilled, nursing, assisted living, and residential facilities). Second, data from most specialty palliative care programs have been primarily focused on patients with a cancer diagnosis. More recently, a few studies have examined palliative care in patients with diagnoses of dementia or heart failure.^17,18 However, little work has been done examining different diagnoses in the various palliative care settings. We are also not aware of any literature that examines episodic care or reenrollment back into palliative care programs after a patient is discharged. Patients may be discharged from palliative care as they may not require ongoing services if their disease has stabilized. However, as their illness progresses these patients may require reentry back into palliative care because of increasing symptom burden, declining functional status, occurrence of new medical events, or exacerbation of illness.

In 2014, Four Seasons was the recipient of a Centers for Medicare & Medicaid Services (CMS) Healthcare Innovation (CMMI) award, “Increasing Patient and System Value with Community-Based Palliative Care (Grant# 1C1CMS331331),” to test and develop a new care delivery and payment model for palliative care through hospice organization delivery. With the Healthcare Innovation Award, Four Seasons expanded upon their existing model to implement, evaluate, and demonstrate the value of CBPC and to propose a new payment approach. One of the goals of the project is to improve healthcare outcomes through delivery of high-quality CBPC using an interdisciplinary workforce, specifically to have 30% of patients transitioning to hospice from CBPC. This article will examine the transition outcomes of participants in the program over the first two years as part of this three-year award.

Methodology

Results

As of July 31, 2016, 2842 patients were enrolled into the CMMI project. Entry points into the program occurred from the home setting (20%, majority referred by primary care physician), palliative care clinic (2%), facilities (29%), small hospitals (32%), and large hospitals (17%) (Table 1). Overall, the largest age cohort of patients was between ages 76 and 95, making up 70% of enrollees, with a larger percentage of patients being female (59%). Patients aged 65–85 made up the majority of those enrolled at the large hospital sites (∼75%) with an even distribution of male to female patients. The majority of patients enrolled in the clinic setting were between the ages of 76 and 85 (44%) with a larger percentage being male (54%) (Table 1). A primary neurologic diagnosis (including dementia and neurodegenerative disorders) accounted for the largest number of patients enrolled in the home (28%), facilities (46%), and small hospitals (28%), while cancer was the top diagnosis of the patients enrolled in the clinics (41%) and large hospitals (21%) (Table 1).

Of the 2842 patients enrolled in the CMMI project, 1004 (35%) transitioned to hospice, 367 (13%) died, 994 (35%) were discharged, and 477 (17%) were currently active patients as of October 2016 (Table 2). When separating by points of entry, of those patients admitted in collaborating facilities (nursing homes/assisted living facilities), 289 (35%) transitioned to hospice, 109 (13%) died, 226 (28%) were discharged, while 199 (24%) remain on the active caseload. The average length of stay of facility patients was 81 days. When the home setting was the point of entry, 240 (41%) of these patients transitioned to hospice, 26 (5%) died, 129 (22%) were discharged, and 184 (32%) remain on the active caseload. The average length of stay for home patients was 83 days. Of the 61 patients enrolled in the clinic, 30 transitioned to hospice (49%), 3 died (5%), 19 were discharged (31%), and 9 remain active patients (15%). Average length of stay for patients enrolled in the clinic was 88 days (Table 2).

Of those patients admitted through the small community hospitals, 312 (35%) transitioned to hospice, 134 (15%) died, 375 (41%) were discharged, and 84 (9%) are still active patients all whom were transferred to outpatient CBPC. In larger hospital settings, patients admitted into the program had the lowest rate of hospice transition (28%) and a large percentage died (20%) or were discharged (52%). Only 0.2% of these patients are currently active, and patients in larger hospitals had a very short length of stay in palliative care with an average of 4.7 days compared to 22.3 days at small hospitals (Table 2). Overall, hospice transition rates are higher and deaths in palliative care are lower with outpatient point of entry (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Hospice transitions and palliative care deaths by admission location. (A) Percentage of patients enrolled in the community-based palliative care program that transitioned to hospice (admitted to hospice). (B) Percentage of patients enrolled in community-based palliative care that died while in palliative care. *p < 0.05 with large hospital as reference.

Patients were discharged for a number of reasons, including goals of care were met, patient/family request, physician and/or facility request, unable to locate, moved out of service area, no order at new location, or discharged to another palliative care program. In facilities, small hospitals, and large hospitals, patients were primarily discharged because goals of care were met, and patient/family request was the primary reason for discharge in the home and clinic setting (Table 3). Of the 994 patients discharged, 177 (17.8%) were re-enrolled into the program (Table 4). Separating by point of initial admission, of the 177 patients 19 (10.7%) patients entered from the home setting, 1 (0.6%) from palliative care clinic, 51 (28.8%) from facilities, 94 (53.1%) from small hospitals, and 12 (6.8%) from large hospitals. The majority of patients re-enrolled back into the program in the same setting as their initial admit location. For example, of the 51 patients with an initial admission from a facility, 34 (66.7%) re-enrolled into the program from a facility and of the 94 patients with an initial admission from a small hospital, 55 (58.5%) re-enrolled from a small hospital (Table 4). Of the 177 re-enrolled patients, after re-enrollment, 62 (35%) transitioned to hospice, 17 (9.6%) died, 53 (29.9%) were discharged, and 45 (17%) were currently active patients.

Discussion

The CBPC program at Four Seasons was started in 2003 to meet the needs of those with serious life-limiting illness. With the CMS Healthcare Innovations Award, Four Seasons expanded upon this existing model to implement, evaluate, and demonstrate the value of CBPC in the region. Over the first two years, 2842 patients have been successfully enrolled and followed in the project with a projected enrollment of >5000 at completion date of September 2017. Patients who enter the program through a community setting tend to be older with a significantly higher ratio of females to males. This is a reflection of a retirement community with aging demographics in the region being offered services as an extension of hospice delivery. This also accounts for the higher rates of dementia diagnosis that are referred from nursing home facilities in the participating sites. Those that enter through a large hospital setting are younger with an equal gender ratio. On examination of hospice transition rates from point of entries, home and clinic setting are highest with rates of 41% and 49%, respectively, emphasizing the need for CBPC in the outpatient environment, where resources are lacking. Many of our home CBPC patients are home bound, have extensive caregiving needs, and lack the ability to return to their primary care physician when crisis occurs; without palliative care there is no safety net, and transfer to the emergency room/hospital is a common occurrence. This highlights the need for a community palliative care presence. Workforce shortage, lack of primary physician's palliative care skills, and challenges for financial sustainability of CBPC programs will need to be overcome to respond to increasing demands.

The transition rate to hospice in the smaller community hospitals was 35%. In contrast to the larger hospitals where the transition rate is much lower at 28%, these smaller community hospital patients tend to be much older, have multiple chronic illnesses, and are often transferred to the hospital from nursing facilities when an acute event occurs. This highlights the importance of offering palliative care in small community hospitals where hospice organizations may be best aligned to provide additional palliative care services. Larger hospitals are equipped with multiple intensive care units and act as regional sites for transfer from community hospitals so it is no surprise that the death rate with palliative care is higher (20% vs. 15%) at these sites compared to the smaller hospitals. Most hospitals have palliative care services; however, very few have outpatient follow-up due to minimal reimbursement opportunities in line with the current hospital business model. A little over a third (35%) of facility patients are transferred to hospice with 13% dying under CBPC services. In most instances, these patients were under “skilled days,” and in our service areas, it is necessary to get approval from facility administrators before people electing hospice, which adds to the higher rate of death in the nursing home settings. The need for CBPC in nursing homes will continue to rise as nursing home utilization will drastically increase with the aging population.^19,20 The importance of maintaining and fostering nursing home relationships will be key to successful partnerships in various facilities.

The average length of stay is longer in the outpatient settings versus inpatient from point of entry (Table 2). There are two reasons we observe this: first, the acuity is much greater in the hospital population leading to higher deaths, and second, inpatient palliative care programs tend to have little outpatient follow-up because of the scarcity of outpatient programs. In our smaller community hospitals, there is good continuity between the inpatient and outpatient teams, which explains the difference in length of stay between small and large hospitals (22 days vs. 5 days). In addition, in small community hospitals the hospital palliative care program and outpatient CBPC program are often run by the same organization, which leads to improved communication and collaboration, whereas this may not occur in large hospitals where providers are often employees of the hospital.

Over the first two years of the CMMI project, 17.8% of discharged patients re-enroll into the program (Table 4). This points to the fluctuating state of life-limiting illnesses, with episodic crisis and improvements; hence, resource needs are constantly shifting. As can be seen, roughly 20% require palliative care services within the year because of increasing symptom burden, declining functional status, occurrence of new medical events, or exacerbation of illness. This necessitates the availability of CBPC services across settings as disease burden can rapidly change and will be dependent on staffing models or referral patterns to match patient need to resources.

There are several limitations to this study. First, a high percentage of the patients enrolled in the CMMI project are a white older population from retirement communities in Western North Carolina; therefore, generalizability is reduced. A large percentage of patient's (26.5%) race category was unknown; this is likely due to a lack of documentation of race. In addition, this demonstration project focuses exclusively on traditional Medicare patients ≥65 years old with a limited life expectancy; since this only represents 42% of patients in Four Seasons CBPC program, further comparison is available for younger patients or those on Medicare Advantage. The highest diagnostic category in the project is neurologic with the majority of those patients having a dementia diagnosis. Neurological impairments may impact length of stay and not represent patients in the typical outpatient CBPC program. Another limitation is the low number of patients enrolled in the large hospital setting, which is a reflection of the two large hospital programs entering the project in the latter part of the second year. Referrals to outpatient palliative care from these large hospitals are limited as outpatient palliative care programs are currently forming in these communities. To increase referrals to outpatient palliative care, discharge planners will need to be educated on the availability and benefit of CBPC services.