Abstract
Context
For end-stage liver disease (ESLD) patients, care focuses on managing the life-threatening complications of portal hypertension, causing high resource utilization.
Objectives
To describe the end-of-life trajectory of hospitalized ESLD patients in Medicare.
Methods
Using a 5% random sample of Medicare fee-for-service beneficiaries, we performed a retrospective cohort study, identifying hospitalized ESLD and heart failure (HF) patients (2007–2011). Index hospitalization end points included mortality, discharge to hospice, and length of stay. Postdischarge end points included all-cause mortality, rehospitalization, hospice enrollment, and days alive and out of hospital (DAOH). Follow-up was at one and three years after index hospitalization discharge. A reference cohort of decompensated HF patients was used for baseline comparison.
Results
At one year, the ESLD cohort (n = 22,311) had 209 DAOH; decompensated HF (n = 85,397) had 252 DAOH. Among ESLD patients, inpatient mortality was 13.5%; all-cause mortality was 64.9%. For these outcomes, rates were higher in those with ESLD than HF. In the ESLD group, rehospitalization rate was 59.1% (slightly lower than the HF group), hospice enrollment rate was 36.1%, and there were higher than expected cancer rates. For hospice-enrolled patients, the median length of time spent in hospice was nine days. The HF cohort had lower hospice enrollment, but more days enrolled.
Conclusion
The results of this study show that morbidity and mortality rates associated with end of life in ESLD are substantial. There is an acute need for alternative approaches to manage the care of ESLD patients.
Keywords: End-stage liver disease, end-of-life trajectory, morbidity, mortality
Introduction
Liver disease is the 12th leading cause of mortality in the U.S. and is often associated with a high symptom burden and protracted course.1–3 Recent trends show an increasing prevalence of cirrhosis beginning in the early 2000s, and although hepatitis C has traditionally been the leading etiology for cirrhosis development, the introduction of direct-acting antiviral medications for hepatitis C treatment will eventually cure this entity.4 The ongoing exponential growth in obesity has led to nonalcoholic fatty liver disease replacing hepatitis C as the leading etiology for cirrhosis. As a result of changes such as these, predictions suggest that an increased number of older adult patients will develop end-stage liver disease (ESLD).5,6
ESLD is associated with multiple complications of portal hypertension including ascites, hepatic encephalopathy, and gastroesophageal varices. ESLD requires frequent outpatient visits and hospitalizations and causes more than 150,000 hospital admissions annually.7 One consequence of longstanding decompensation is disability, which often creates a complex psychosocial situation.8 Care for ESLD patients has traditionally focused on treating the underlying etiology of cirrhosis, management of portal hypertension complications, and evaluation for transplant. The complications of portal hypertension are generally manageable, but as the disease progresses, patients tend to experience a clinical decline with increased symptom frequency and severity.9 Despite efforts to manage complications, many ESLD patients ultimately succumb to their liver disease with only a small proportion ever achieving successful transplantation. Based on 2015 statistics, approximately 6000 patients achieve successful transplantation each year, but more than 15,000 patients remain waitlisted.10
Palliative care offers an approach that helps patients effectively manage symptoms and the psychosocial aspects of living with a life-threatening disorder, while simultaneously allowing aggressive management for those who need advanced care.11 The transition to hospice may occur on the palliative care continuum when aggressive care is no longer indicated. Palliative medicine has been widely used in oncology, improving quality of life through reduced symptom burden, increased satisfaction with medical care, and prolonged survival.12,13 Yet when compared to cancer, ESLD has a distinctly different end-of-life (EOL) disease trajectory. As previously described by Murray et al. in 2005 and further supported in other studies, cancer is more commonly characterized by an acute decline, whereas advanced chronic illnesses like ESLD have more of a fluctuating deterioration, characterized by periods of stability that are interspersed with acute episodes of decline.14–16 These differences have made it challenging to understand the appropriate time to offer palliative care interventions in ESLD and other chronic organ failure situations.17,18
The primary objective of this study was to describe the EOL trajectory of hospitalized ESLD patients in a Medicare fee-for-service population by examining morbidity and mortality compared to a reference cohort of heart failure (HF) patients. Similar to ESLD, HF patients often have a prolonged course with acute episodes of decompensation. There is growing recognition that palliative care and hospice can effectively help patients manage symptoms and the psychosocial aspects of living with a chronic disease, as well as decrease hospitalizations. These benefits have been well described in HF literature, leading to the choice of HF as a reference cohort.19–26
Methods
Data Source
For this retrospective cohort study, we use claims data and corresponding denominator files from the Centers for Medicare & Medicaid Services for a nationally representative 5% sample of Medicare beneficiaries. These research-identifiable files allow longitudinal follow-up and include inpatient institutional claims, outpatient institutional claims, hospice claims, and noninstitutional professional service claims for periods of enrollment in traditional fee-for-service Medicare. The denominator files include information about patient demographics, program enrollment, and mortality.
ESLD Cohort
The ESLD cohort included all patients with at least one hospitalization between January 1, 2007 and December 31, 2011 with a diagnosis code for chronic liver disease, cirrhosis, or hepatic decompensation (Table 1) in any position on the inpatient claim. If the inpatient claim did not include a diagnosis code for hepatic decompensation, then this code had to be on an outpatient or carrier claim in the year before hospitalization. Patients were excluded if they had a diagnosis of secondary liver cancer (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] diagnosis code 197.7) on any claim in the year before the hospitalization. We required patients to be enrolled in fee-for-service Medicare at hospital discharge, as well as during the prior year. If a patient has multiple eligible hospitalizations, we selected the earliest for inclusion in our study. We included all adult patients aged 18 years and older; therefore, our cohort contained patients receiving Medicare due to disability.
Table 1.
Characteristics of the Adult ESLD Cohort at Index Hospitalization
| Variable | Overall (n = 30,597) |
|---|---|
| Age (yrs), mean (SD) | 70.9 (13.3) |
| Gender, male | 14,516 (47.4%) |
| Race | |
| Black | 3842 (12.6%) |
| White | 24,701 (80.7%) |
| Other/Unknown | 2054 (6.7%) |
| Complications of liver disease | |
| Spontaneous bacterial peritonitis | 688 (2.2%) |
| Variceal bleeding | 1489 (4.9%) |
| Hepatic encephalopathy | 4553 (14.9%) |
| Ascites | 21,593 (70.6%) |
| Hepatorenal syndrome | 728 (2.4%) |
| Procedures related to liver complications | |
| Chemoembolization | 815 (2.7%) |
| Endoscopic procedure | 14,090 (46.1%) |
| Radiofrequency ablation | 93 (0.3%) |
| Transjugular intrahepatic portosystemic shunt | 171 (0.6%) |
| Etiology of liver disease | |
| Viral hepatitis | 4592 (15.0%) |
| Alcohol | 4829 (15.8%) |
| Other | 10,870 (35.5%) |
| Unknown | 16,148 (52.8%) |
| Other medical history | |
| Cancer | 9101 (29.7%) |
| Hepatocellular carcinoma | 922 (3.0%) |
| Chronic kidney disease | 12,197 (39.9%) |
| Chronic obstructive pulmonary disease | 14,737 (48.2%) |
| Coronary artery disease | 15,329 (50.1%) |
| Diabetes mellitus | 14,883 (48.6%) |
| Heart failure | 15,048 (49.2%) |
| Hypertension | 26,125 (85.4%) |
ESLD = end-stage liver disease.
HF Cohort
We identified a comparison cohort of patients hospitalized with decompensated HF. In comparison with ESLD, there was a large number of HF patients, so the HF cohort included patients aged 65 years and older with at least one hospitalization between January 1, 2007 and December 31, 2011 with a primary diagnosis code for HF (ICD-9-CM diagnosis codes 428.x, 402.x1, 404.x1, 404.x3). For this comparison, we restricted the ESLD cohort to patients aged 65 years and older to create comparable cohorts most representative of the majority of Medicare beneficiaries. Similar to the ESLD cohort, these patients must have been enrolled in fee-for-service Medicare at hospital discharge, as well as during the prior year; the earliest hospitalization was selected if multiple hospitalizations occurred. Patients in the ESLD cohort were not eligible for inclusion in the HF cohort. In both cohorts, the included hospitalization is referred to as the index hospitalization.
Patient Characteristics
To describe both ESLD and HF patients in our study, we summarize demographic information (age, sex, and race) from the denominator file and comorbid conditions identified using diagnosis codes from inpatient, outpatient, and professional service claims in the year before the index hospitalization. Coding for these conditions was based on previously published algorithms.27,28
In addition, we describe several disease-related concepts among the ESLD cohort using coding algorithms based on diagnosis and procedure codes (Supplemental Table, available at jpsmjournal.com) on the index hospitalization claim and all inpatient, outpatient, and professional service claims from the prior year. Complications of liver disease included ascites, hepatic encephalopathy, hepatorenal syndrome, spontaneous bacterial peritonitis, and variceal bleeding. The etiology of liver disease was identified as viral hepatitis, alcohol, other, or unknown (the latter being determined if none of the listed codes were found). Procedures related to liver complications included chemoembolization, endoscopic procedures (esophagogastroduodenoscopy, endoscopic retrograde cholangiopancreatography, and endoscopic ultrasound, colonoscopy), radiofrequency ablation, and transjugular intrahepatic portosystemic shunt. We also identified hepatocellular carcinoma in this cohort.
Outcomes
Outcomes associated with the index hospitalization included mortality, discharge to hospice, and length of stay. The length of stay was ascertained based on the index hospitalization admission and discharge dates. Mortality before discharge and discharge to hospice were determined from the disposition code on the index hospitalization claim.
Postdischarge outcomes included all-cause mortality, hospice enrollment, rehospitalization, and days alive and out of hospital (DAOH). The follow-up periods of interest were one and three years after discharge from the index hospitalization. Only those patients discharged alive from the index hospitalization were included in the analysis of postdischarge outcomes. We determined mortality during the follow-up period using death dates in the Medicare denominator files. We determined new hospice enrollment during the follow-up period and total days in hospice from hospice claims. We identified rehospitalizations using inpatient claims subsequent to discharge from the index hospitalization. Rehabilitation stays were not counted as rehospitalizations. For the ESLD cohort, we also identified hospitalizations for hepatic transplantation (ICD-9-CM procedure code 50.51, 50.59). For the HF cohort, we identified hospitalizations for cardiac transplantation (ICD-9-CM procedure code 33.6, 37.51, 37.52). We determined the number of DAOH during the follow-up period based on the death date (if present), and the duration of all subsequent hospitalizations. Patients who enrolled in Medicare managed care during the follow-up period and those whose follow-up was censored by the end of data availability (December 31, 2011) were excluded from the measurement and reporting of DAOH and days in hospice.
We examined baseline characteristics for cancer rates in the adult ESLD cohort (Table 1) and performed a subgroup analysis of ESLD patients with and without cancer to compare differences in outcomes.
Statistical Analysis
We summarize patient characteristics for all cohorts (ESLD ≥ 18 years, ESLD ≥ 65 years, and HF ≥ 65 years) using means with SDs for continuous variables and frequencies with percentages for categorical variables. We tested for differences between the ESLD and HF cohorts using Kruskal-Wallis tests for continuous variables and chi-square tests for categorical variables.
We present in-hospital mortality and discharge to hospice using frequencies and percentages, and we use chi-square tests to test for differences between groups. We present length of stay, DAOH, and days in hospice using means with SDs or medians with interquartile ranges, using Kruskal-Wallis tests to test for differences between groups. Calculations of the incidence of mortality at one and three years were based on Kaplan-Meier estimates, and we used log-rank tests to test for differences between groups. For all rehospitalization outcomes and for hospice enrollment, we estimated incidence at one and three years using the cumulative incidence function, which accounts for the competing risk of mortality and tested for differences between the groups using Gray tests.29
Analyses were performed using SAS version 9.3 (SAS Institute, Inc., Cary, NC). The institutional review board of the Duke University Health System approved the study.
Results
Baseline characteristics for the ESLD adult study cohort are listed in Table 1, and the observed outcomes for this cohort are reported in Table 2. Inpatient and all-cause mortality were 11.8% and 42.0%, respectively. DAOH at one year was 213.4 for this cohort.
Table 2.
Observed Outcomes for the Adult ESLD Cohort
| Variable | Overall |
|---|---|
| Index hospitalization mortalitya | 3616 (11.8%) |
| Index hospitalization, discharged to hospicea | 1650 (5.4%) |
| Index hospitalization, length of stay (days)a | |
| Mean (SD) | 9.7 (11.9) |
| Median (Q1, Q3) | 6 (4, 12) |
| Postdischarge mortality, all-causeb | |
| One year | 10,588 (42.0%) |
| Three years | 13,606 (61.1%) |
| Hospice enrollmentb | |
| One year | 5921 (23.1%) |
| Three years | 7209 (31.2%) |
| Rehospitalization, all-causeb | |
| One year | 15,896 (62.3%) |
| Three years | 17,876 (74.5%) |
| Hepatic transplantationb | |
| One year | 112 (0.5%) |
| Three years | 153 (0.7%) |
| Days in hospice, among those enrolledc | |
| Mean (SD) | 45.0 (104.2) |
| Median (Q1, Q3) | 9 (3, 36) |
| Days alive and out of the hospital, one yeard | |
| Mean (SD) days | 223.2 (147.2) |
| Median (Q1, Q3) | 301 (49, 358) |
| Days alive and out of the hospital, three yearsd | |
| Mean (SD) days | 530.9 (453.7) |
| Median (Q1, Q3) | 448 (48.5, 1054) |
ESLD = end-stage liver disease; Q1 = quarter 1; Q3 = quarter 3.
Calculated among all patients.
Calculated among all patients discharged alive; hospice enrollment results include the 6.1% of patients discharged alive who were discharged to hospice.
Calculated among all patients discharged alive with complete information for the follow-up period who were observed to have enrolled in hospice during the three-year follow-up period.
Calculated among all patients discharged alive with complete information for the follow-up period.
Table 3 compares baseline characteristics of the ESLD and decompensated HF cohorts. The percentages of each comorbid condition were statistically different between the two groups, but the most striking difference was the rate of cancer, which was 34.7% in ESLD and 17.5% in HF. The most common cancers were colon and liver.
Table 3.
Characteristics of the ESLD and HF Cohorts at Index Hospitalization, Patients Aged ≥65 Yearsa
| Variable | ESLD (n = 22,311) | HF (n = 85,397) |
|---|---|---|
| Age (yrs), mean (SD) | 77.4 (7.6) | 81.7 (8.0) |
| Gender, male | 9491 (42.5%) | 32,129 (37.6%) |
| Race | ||
| Black | 1956 (8.8%) | 8192 (9.6%) |
| White | 19,021 (85.3%) | 73,787 (86.4%) |
| Other/Unknown | 1334 (6.0%) | 3418 (4.0%) |
| Other medical history | ||
| Cancer | 7733 (34.7%) | 14,983 (17.5%) |
| Chronic kidney disease | 8568 (38.4%) | 37,107 (43.5%) |
| COPD | 10,784 (48.3%) | 53,712 (62.9%) |
| Coronary artery disease | 12,100 (54.2%) | 65,786 (77.0%) |
| Diabetes mellitus | 10,707 (48.0%) | 42,743 (50.1%) |
| HF | 11,743 (52.6%) | 85,397 (100.0%) |
| Hypertension | 19,574 (87.7%) | 81,081 (94.9%) |
ESLD = end-stage liver disease; HF = heart failure; COPD = chronic obstructive pulmonary disease.
P-value for all comparisons <0.001.
Table 4 compares the end points for the ESLD and HF cohorts. There is greater in-hospital death in ESLD patients at 13.5% compared to 3.7% in HF. The all-cause mortality in ESLD at one year (46.2%) and three years (64.9%) was greater than the all-cause mortality in the HF cohort, which was 34.8% at one year and 60.8% at three years. Both cohorts had high rehospitalization rates. The length of stay for the adult ESLD cohort was 9.9 days compared with 5.1 days for the HF cohort. DAOH were 209.5 in ESLD and 252.1 in HF.
Table 4.
Observed Outcomes for the ESLD and HF Cohorts, Patients Aged ≥65 Yearsa
| Variable | ESLD | HF |
|---|---|---|
| Index hospitalization mortalityb | 3021 (13.5%) | 3127 (3.7%) |
| Index hospitalization, discharged to hospiceb | 1446 (6.5%) | 2323 (2.7%) |
| Index hospitalization, length of stay (days)b | ||
| Mean (SD) | 9.9 (10.6) | 5.1 (5.1) |
| Median (Q1, Q3) | 7 (4, 12) | 4 (3, 6) |
| Postdischarge mortality, all-causec | ||
| One year | 8396 (46.2%) | 26,761 (34.8%) |
| Three years | 10,534 (64.9%) | 40,315 (60.8%) |
| Hospice enrollmentc | ||
| One year | 5029 (27.3%) | 13,956 (17.9%) |
| Three years | 6042 (36.1%) | 19,688 (28.9%) |
| Rehospitalization, all-causec | ||
| One year | 10,795 (59.1%) | 49,862 (63.9%) |
| Three years | 12,192 (71.0%) | 58,289 (79.6%) |
| Hepatic transplantationc | ||
| One year | 33 (0.2%) | — |
| Three years | 41 (0.3%) | — |
| Cardiac transplantationc | ||
| One year | — | 18 (<0.1%) |
| Three years | — | 23 (<0.1%) |
| Days in hospice, among those enrolledd | ||
| Mean (SD) | 43.5 (100.9) | 62.9 (126.0) |
| Median (Q1, Q3) | 9 (3, 33) | 13 (3, 59) |
| Days alive and out of the hospital, one yeare | ||
| Mean (SD) days | 209.5 (150.8) | 252.1 (135.8) |
| Median (Q1, Q3) | 273 (30, 357) | 332 (129, 362) |
| Days alive and out of the hospital, three yearse | ||
| Mean (SD) days | 496.5 (455.7) | 581.6 (432.1) |
| Median (Q1, Q3) | 353 (29, 1046) | 586 (124, 1059) |
ESLD = end-stage liver disease; HF = heart failure; Q1 = quarter 1; Q3 = quarter 3.
P-value for all comparisons <0.001.
Calculated among all patients.
Calculated among all patients discharged alive; hospice enrollment results include the 7.5% of ESLD patients and 2.8% of heart failure patients discharged alive who were discharged to hospice.
Calculated among all patients discharged alive with complete information for the follow-up period who were observed to have enrolled in hospice during the three-year follow-up period.
Calculated among all patients discharged alive with complete information for the follow-up period.
Table 5 details a subgroup analysis of the full ESLD cohort, including adults older than or equal to 18 years of age with and without cancer, given the high proportion of ESLD patients with cancer from the characteristics identified at the index hospitalization (Table 1). As expected, outcomes including in-hospital death, postdischarge mortality, hospice mortality, and DAOH were poorer in those with cancer compared to those patients who were cancer free. There was decreased rehospitalization with twice as many hospice enrollments at one and three years in the cohort with cancer compared to those without cancer. Table 6 provides the subtypes of primary cancer found in the ESLD cohort.
Table 5.
Observed Outcomes for the Full ESLD Cohort
| Variable | ESLD Patients (Overall) | ESLD Patients (w/o Cancer) | ESLD Patients (w/Cancer) |
|---|---|---|---|
| Index hospitalization mortalitya | 3616 (11.8%) | 2368 (11.0%) | 1248 (13.7%) |
| Index hospitalization, discharged to hospicea | 1650 (5.4%) | 828 (3.9%) | 822 (9.0%) |
| Postdischarge mortality, all-causeb | |||
| One year | 10,588 (42.0%) | 6474 (36.4%) | 4114 (55.6%) |
| Three years | 13,606 (61.1%) | 8645 (55.9%) | 4961 (73.7%) |
| Hospice enrollmentb | |||
| One year | 5921 (23.1%) | 3223 (17.8%) | 2698 (35.9%) |
| Three years | 7209 (31.2%) | 4073 (25.4%) | 3136 (45.2%) |
| Rehospitalization, all-causeb | |||
| One year | 15,896 (62.3%) | 11,486 (63.7%) | 4410 (59.1%) |
| Three years | 17,876 (74.5%) | 13,016 (76.9%) | 4860 (68.5%) |
| Hepatic transplantationb | |||
| One year | 112 (0.5%) | 80 (0.5%) | 32 (0.5%) |
| Three years | 153 (0.7%) | 113 (0.7%) | 40 (0.6%) |
| Days in hospice, among those enrolledc | |||
| Mean (SD) | 45.0 (104.2) | 50.5 (117.5) | 37.1 (80.5) |
| Median (Q1, Q3) | 9 (3, 36) | 9 (3, 38) | 10 (3, 32) |
| Days alive and out of the hospital, one yeard | |||
| Mean (SD) | 223.2 (147.2) | 238.6 (142.4) | 185.5 (152.0) |
| Median (Q1, Q3) | 301 (49, 358) | 321 (83, 360) | 185.5 (19, 352) |
| Days alive and out of the hospital, three yearsd | |||
| Mean (SD) | 530.9 (453.7) | 578.2 (451.3) | 414.6 (438.2) |
| Median (Q1, Q3) | 448 (48.5, 1054) | 604 (77, 1063) | 195 (19, 953) |
ESLD = end-stage liver disease; Q1 = quarter 1; Q3 = quarter 3.
Calculated among all patients.
Calculated among all patients discharged alive; hospice enrollment results include the 7.5% of ESLD patients and 2.8% of heart failure patients discharged alive who were discharged to hospice.
Calculated among all patients discharged alive with complete information for the follow-up period who were observed to have enrolled in hospice during the three-year follow-up period.
Calculated among all patients discharged alive with complete information for the follow-up period.
Table 6.
Most Common Cancer Diagnoses Among ESLD Cohort
| Primary Cancer | n (%) |
|---|---|
| Breast | 1415 (12.5) |
| Lung | 1183 (10.4) |
| Colon | 2267 (20.0) |
| Stomach | 509 (4.5) |
| Pancreas | 1241 (10.9) |
| Ovary | 1163 (10.3) |
| Endometrial | 375 (3.3) |
| Peritoneal | 530 (4.7) |
| Skin | 227 (2.0) |
| Prostate | 1324 (11.7) |
| Blood/lymph | 1678 (14.8) |
| Bladder | 575 (5.1) |
| Kidney | 514 (4.5) |
| Livera | 1971 (17.4) |
ESLD = end-stage liver disease.
Liver cancer not limited to HCC but includes all primary hepatic malignancies.
Discussion
Our study demonstrates significant morbidity and mortality associated with EOL in ESLD with reduced DAOH, high in-hospital mortality, and high all-cause mortality. The baseline characteristics of the ESLD cohort are consistent with the published literature with regard to demographics and common complications of liver disease.30–32
To better understand and quantify the EOL trajectories for ESLD, the novel objective measure of DAOH was used in this analysis. The theory behind using this measure is that DAOH can embody both the morbidity and mortality of a disease entity. All end points are not equal; for example, having a self-limited gastrointestinal bleed may have different implications than suffering from refractory encephalopathy with multiple hospitalizations. Capturing the different clinical implications of these two scenarios is more accurately accomplished with DAOH compared to commonly used time-to-event measures.33,34
Surprisingly, there was a higher than expected rate of hospice enrollment for ESLD at hospital discharge and during follow-up. There was a hospice enrollment of 31.2% at the three-year follow-up for ESLD patients aged ≥18 years, which suggests that a significant number of patients are being identified and referred for these services (Table 2). One explanation for these results was the high rate of cancer seen in this population. In the ESLD cohort, 29.7% of patients had a diagnosis of cancer; only 3.0% had a diagnosis of hepatocellular carcinoma (Table 1). We performed a subgroup analysis of all adult ESLD patients with and without cancer and found that having a cancer diagnosis increased the rate of discharge to hospice from 3.9% to 9.0% during the index hospitalization (Table 5). For patients with ESLD without cancer, there was also a significant increase in the rates of hospice enrollment at one-year (17.8%) and three-year (25.4%) follow-up, compared to those ESLD patients with cancer (35.9% at one year and 45.2% at three years; Table 5). When stratifying the types of malignancy comprising the cancer diagnosis, we found many aggressive tumors such as pancreatic, lung, and ovarian, as well as more common malignancies such as breast, prostate, and colon cancer (Table 6). This finding suggests that the presence of cancer may have played a significant part in the clinical decompensation of this patient population and caused confounding associated with higher rates of hospice enrollment than would be expected in ESLD alone. Only 3.9% of ESLD patients without cancer were discharged to hospice following the index hospitalization, and more than one-third of this patient population died within one year; this suggests that we are still missing a large cohort of patients that may benefit from palliative care and hospice referral (Table 5).
Another important consideration in hospice referral is the amount of time that patients are able to receive these benefits. ESLD patients had a median time of nine days in hospice before death, which is 50% lower than the national average of 18.5 days for all-comers enrolled in hospice.35 Enrolling patients earlier in the course of decompensation would allow them to use the extensive services offered through hospice for a longer period of time; these services include 24-hour access to nursing, a multidisciplinary team to address active issues, and increased availability of medical equipment and medications paid for through Medicare with minimal costs to the patient or caregiver. Earlier enrollment may also benefit the health care system by decreasing costs via lower rates of rehospitalization, invasive procedures, and other futile measures.36,37
We chose a cohort of patients with decompensated HF as a reference group because the burden of HF is well understood and the disease trajectory bears similarity to ESLD. Recent evidence suggests that palliative care and hospice can provide valuable services to address unmet needs in HF, raising the prospect that patients with ESLD may also benefit from more aggressive use of these services.22,38 This study has shown that decompensated liver disease carries significant morbidity and mortality, highlighted by a prolonged hospital length of stay (seven days) and high index hospitalization mortality (13.5%). All-cause mortality and readmission rates in patients with ESLD are similar to rates in HF patients, which suggests that the early use of palliative medicine and hospice in ESLD should be considered.
Limitations
Our study design had several limitations. First, although validated coding algorithms were used, patients may be missed or inappropriately included because of clinician errors in coding. Inaccuracies in coding may also impact characteristics and outcomes data, thereby resulting in underrepresentation or overrepresentation. Second, patients who were not enrolled in Medicare fee-for-service were not analyzed, which may result in biases for those with additional third-party coverage. Third, the ESLD and HF cohorts were created differently, establishing patient populations with decompensated disease, and excluding patients with chronic stable disease. Given the high rates of morbidity and mortality in both cohorts, this objective was achieved, yet appropriate patients with decompensated HF may have been excluded. Nevertheless, preliminary data were assessed based on prior literature to ensure that frequencies were consistent and suggest that these limitations did not create significantly skewed data. Fourth, the full adult ESLD cohort had similar outcomes to the ESLD cohort ≥65 years, which suggests generalizability among the groups; however, the generalizability of the data to ESLD patients outside of Medicare is unknown. Given that decompensated liver disease is often an indication for disability benefits and, consequently, Medicare coverage, one would expect a significant proportion of the patient population to be represented in this study, lessening the effect of this limitation. Finally, there was a subset of patients with both cirrhosis and HF as comorbid conditions. We decided to include these patients in the ESLD cohort if they had a code for hepatic decompensation. Volume overload complications would be difficult to assign to one disease entity vs. the other, due to the physiologic overlap; nonetheless, complications from gastrointestinal bleeding, encephalopathy, spontaneous bacterial peritonitis, and hepatorenal syndrome are often more clearly associated with hepatic failure. As a result, if clinicians chose to code hepatic decompensation, then these particular patients were deemed appropriate for ESLD inclusion. Ultimately, it was decided to include these patients given the goal of trying to better understand the ESLD EOL trajectory, which is often associated with multiple comorbid conditions in need of consideration during disease management.
Future Directions
Although there are increasing initiatives to examine the role of early palliative care in disease entities such as HF, there has been little research on palliative care in relation to liver disease. Hepatologists have primarily been responsible for the management of liver disease patients; whereas some symptoms such as variceal bleeding and ascites can be straightforward to treat, other issues such as psychosocial aspects and advanced care planning may be challenging, leaving patients with a poor quality of life. Our study suggests that ESLD patients have high morbidity and mortality and may benefit from palliative care and hospice. More aggressive research is needed to identify the subsets of ESLD patients that are most likely to benefit from palliative care services. Future prospective studies should focus on examining health care utilization and patient-reported quality of life outcomes in the setting of early palliative care for liver disease patients.
Supplementary Material
Acknowledgments
This work was supported internally by the Duke Clinical Research Institute.
Footnotes
Disclosures
The authors declare no conflicts of interest.
References
- 1.Alvarez MA, Cirera I, Solà R, et al. Long-term clinical course of decompensated alcoholic cirrhosis. J Clin Gastroenterol. 2011;45:906–911. doi: 10.1097/MCG.0b013e3182284e13. [DOI] [PubMed] [Google Scholar]
- 2.Das K, Das K, Datta S, et al. Course of disease and survival after onset of decompensation in hepatitis B virus-related cirrhosis. Liver Int. 2010;30:1033–1042. doi: 10.1111/j.1478-3231.2010.02255.x. [DOI] [PubMed] [Google Scholar]
- 3.Planas R, Ballesté B, Alvarez MA, et al. Natural history of decompensated hepatitis C virus-related cirrhosis. A study of 200 patients. J Hepatol. 2004;40:823–830. doi: 10.1016/j.jhep.2004.01.005. [DOI] [PubMed] [Google Scholar]
- 4.Beste LA, Leipertz SL, Green PK, et al. Trends in burden of cirrhosis and hepatocellular carcinoma by underlying liver disease in US veterans, 2001–2013. Gastroenterology. 2015;149:1471–1482. doi: 10.1053/j.gastro.2015.07.056. [DOI] [PubMed] [Google Scholar]
- 5.Kanwal F, Kramer JR, Duan Z, et al. Trends in the burden of nonalcoholic fatty liver disease in a United States cohort of veterans. Clin Gastroenterol Hepatol. 2015;14:301–308. e1–2. doi: 10.1016/j.cgh.2015.08.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Vernon G, Baranova A, Younossi ZM. Systematic review: the epidemiology and natural history of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis in adults. Aliment Pharmacol Ther. 2011;34:274–285. doi: 10.1111/j.1365-2036.2011.04724.x. [DOI] [PubMed] [Google Scholar]
- 7.Wigg AJ, McCormick R, Wundke R, Woodman RJ. Efficacy of a chronic disease management model for patients with chronic liver failure. Clin Gastroenterol Hepatol. 2013;11:850–858. doi: 10.1016/j.cgh.2013.01.014. [DOI] [PubMed] [Google Scholar]
- 8.Bajaj JS, Thacker LR, Wade JB, et al. PROMIS computerised adaptive tests are dynamic instruments to measure health-related quality of life in patients with cirrhosis. Aliment Pharmacol Ther. 2011;34:1123–1132. doi: 10.1111/j.1365-2036.2011.04842.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.D’Amico G, Pasta L, Morabito A, et al. Competing risk and prognostic stages in cirrhosis: a 25-year inception cohort study of 494 patients. Aliment Pharmacol Ther. 2014;39:1180–1193. doi: 10.1111/apt.12721. [DOI] [PubMed] [Google Scholar]
- 10.Organ Procurement and Transplantation Network. Data. U.S. Department of Health & Human Services Organ Procurement and Transplantation Network; 2015. [Accessed February 9, 2016]. Available at: https://optn.transplant.hrsa.gov/converge/data/ [Google Scholar]
- 11.World Health Organization. [Accessed February 9, 2016];Cancer: WHO definition of palliative care. 2016 Available at: http://www.who.int/cancer/palliative/definition/en/
- 12.Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non-small-cell lung cancer. N Engl J Med. 2010;363:733–742. doi: 10.1056/NEJMoa1000678. [DOI] [PubMed] [Google Scholar]
- 13.Glare PA. Early implementation of palliative care can improve patient outcomes. J Natl Compr Canc Netw. 2013;11:S3–S9. doi: 10.6004/jnccn.2013.0212. [DOI] [PubMed] [Google Scholar]
- 14.Murray SA, Kendall M, Boyd K, Sheikh A. Illness trajectories and palliative care. BMJ. 2005;330:1007–1011. doi: 10.1136/bmj.330.7498.1007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Murray SA, Sheikh A. Palliative care beyond cancer: care for all at the end of life. BMJ. 2008;336:958–959. doi: 10.1136/bmj.39535.491238.94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kendall M, Carduff E, Lloyd A, et al. Different experiences and goals in different advanced diseases: comparing serial interviews with patients with cancer, organ failure or frailty and their family and professional carers. J Pain Symptom Manage. 2015;50:216–224. doi: 10.1016/j.jpainsymman.2015.02.017. [DOI] [PubMed] [Google Scholar]
- 17.Bernabeu-Wittel M, Murcia-Zaragoza J, Hernández-Quiles C, et al. Development of a six-month prognostic index in patients with advanced chronic medical conditions: the PALIAR score. J Pain Symptom Manage. 2014;47:551–565. doi: 10.1016/j.jpainsymman.2013.04.011. [DOI] [PubMed] [Google Scholar]
- 18.Poonja Z, Brisebois A, van Zanten SV, et al. Patients with cirrhosis and denied liver transplants rarely receive adequate palliative care or appropriate management. Clin Gastroenterol Hepatol. 2014;12:692–698. doi: 10.1016/j.cgh.2013.08.027. [DOI] [PubMed] [Google Scholar]
- 19.Hupcey JE, Penrod J, Fogg J. Heart failure and palliative care: implications in practice. J Palliat Med. 2009;12:531–536. doi: 10.1089/jpm.2009.0010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gadoud A, Jenkins SM, Hogg KJ. Palliative care for people with heart failure: summary of current evidence and future direction. Palliat Med. 2013;27:822–828. doi: 10.1177/0269216313494960. [DOI] [PubMed] [Google Scholar]
- 21.Cheung WY, Schaefer K, May CW, et al. Enrollment and events of hospice patients with heart failure vs. cancer J Pain Symptom Manage. 2013;45:552–560. doi: 10.1016/j.jpainsymman.2012.03.006. [DOI] [PubMed] [Google Scholar]
- 22.Bekelman DB, Hutt E, Masoudi FA, Kutner JS, Rumsfeld JS. Defining the role of palliative care in older adults with heart failure. Int J Cardiol. 2008;125:183–190. doi: 10.1016/j.ijcard.2007.10.005. [DOI] [PubMed] [Google Scholar]
- 23.Gavazzi A, De Maria R, Manzoli L, et al. Palliative needs for heart failure or chronic obstructive pulmonary disease: results of a multicenter observational registry. Int J Cardiol. 2015;184:552–558. doi: 10.1016/j.ijcard.2015.03.056. [DOI] [PubMed] [Google Scholar]
- 24.Low J, Pattenden J, Candy B, Beattie JM, Jones L. Palliative care in advanced heart failure: an international review of perspectives of recipients and health professionals on care provision. J Card Fail. 2011;17:231–252. doi: 10.1016/j.cardfail.2010.10.003. [DOI] [PubMed] [Google Scholar]
- 25.LeMond L, Allen LA. Palliative care and hospice in advanced heart failure. Prog Cardiovasc Dis. 2011;54:168–178. doi: 10.1016/j.pcad.2011.03.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Goodlin SJ. Palliative care in congestive heart failure. J Am Coll Cardiol. 2009;54:386–396. doi: 10.1016/j.jacc.2009.02.078. [DOI] [PubMed] [Google Scholar]
- 27.Birman-Deych E, Waterman AD, Yan Y, et al. Accuracy of ICD-9-CM codes for identifying cardiovascular and stroke risk factors. Med Care. 2005;43:480–485. doi: 10.1097/01.mlr.0000160417.39497.a9. [DOI] [PubMed] [Google Scholar]
- 28.Quan H, Sunararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43:1130–1139. doi: 10.1097/01.mlr.0000182534.19832.83. [DOI] [PubMed] [Google Scholar]
- 29.Gray RJ. A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat. 1988;16:1141–1154. [Google Scholar]
- 30.United States Census Bureau. Quick Facts: United States. [Accessed February 9, 2016];United States Census Bureau web site. 2015 Available at: http://www.census.gov/quickfacts/table/PST045215/00.
- 31.Runyon BA AASLD Practice Guidelines Committee. Management of adult patients with ascites due to cirrhosis: an update. Hepatology. 2009;49:2087–2107. doi: 10.1002/hep.22853. [DOI] [PubMed] [Google Scholar]
- 32.Garcia-Tsao G, Sanyal AJ, Grace ND, Carey W. Prevention and management of gastroesophageal varices and variceal hemorrhage in cirrhosis. Hepatology. 2007;46:922–938. doi: 10.1002/hep.21907. [DOI] [PubMed] [Google Scholar]
- 33.Ariti CA, Cleland JG, Pocock SJ, et al. Days alive and out of hospital and the patient journey in patients with heart failure: insights from the Candesartan in heart failure: assessment of reduction in mortality and morbidity (CHARM) program. Am Heart J. 2011;162:900–906. doi: 10.1016/j.ahj.2011.08.003. [DOI] [PubMed] [Google Scholar]
- 34.Wong YW, Fonarow GC, Mi X, et al. Early intravenous heart failure therapy and outcomes among older patients hospitalized for acute decompensated heart failure: findings from the Acute Decompensated Heart Failure Registry Emergency Module (ADHERE-EM) Am Heart J. 2013;166:349–356. doi: 10.1016/j.ahj.2013.05.014. [DOI] [PubMed] [Google Scholar]
- 35.National Hospice and Palliative Care Organization. [Accessed February 9, 2016];NHPCO’s facts and figures: Hospice care in America. (2014). Available at: http://www.nhpco.org/sites/default/files/public/Statistics_Research/2014_Facts_Figures.pdf.
- 36.Obermeyer Z, Makar M, Abujaber S, et al. Association between Medicare hospice benefit and health care utilization and costs for patients with poor-prognosis cancer. JAMA. 2014;312:1888–1896. doi: 10.1001/jama.2014.14950. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Kelley AS, Deb P, Du Q, Aldridge Carlson MD, Morrison RS. Hospice enrollment saves money for Medicare and improves care quality across a number of different lengths-of-stay. Health Aff (Millwood) 2013;32:552–561. doi: 10.1377/hlthaff.2012.0851. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Ivany E, While A. Understanding the palliative care needs of heart failure patients. Br J Community Nurs. 2013;18:441–446. doi: 10.12968/bjcn.2013.18.9.441. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.