Abstract
PURPOSE:
To determine whether the type of delivery system is associated with intensity of care at the end of life for Medicare beneficiaries with cancer.
PATIENTS AND METHODS:
We used SEER registry data linked with Medicare claims to evaluate intensity of end-of-life care for patients who died of one of ten common cancers diagnosed from 2009 through 2014. Patients were categorized as receiving the majority of their care in an integrated delivery system, designated cancer center, health system that was both integrated and a certified cancer center, or health system that was neither. We evaluated adherence to seven nationally endorsed end-of-life quality measures using generalized linear models across four delivery system types.
RESULTS:
Among 100,549 beneficiaries who died of cancer during the study interval, we identified only modest differences in intensity of end-of-life care across delivery system structures. Health systems with no cancer center or integrated affiliation demonstrated higher proportions of patients with multiple hospitalizations in the last 30 days of life (11.3%), death in an acute care setting (25.9%), and lack of hospice use in the last year of life (31.6%; all P < .001). Patients enrolled in hospice had lower intensity care across multiple end-of-life quality measures.
CONCLUSION:
Intensity of care at the end of life for patients with cancer was higher at delivery systems with no integration or cancer focus. Maximal supportive care delivered through hospice may be one avenue to reduce high-intensity care at the end of life and may impact quality of care for patients dying from cancer.
INTRODUCTION
In 2019, more than 600,000 cancer-related deaths are expected to have occurred across the United States. The end-of-life period represents an important opportunity for in-depth communication between patients and physicians to identify individual preferences and goals of care. Likely reflecting multiple factors including patient and family preferences, the quality, cost, and intensity of end-of-life care for patients with cancer varies widely, with many patients receiving expensive, high-intensity care.1-4 Importantly, high-intensity end-of-life experiences may offer little improvement in quality of life and are often associated with lower patient and family satisfaction.5-9 Enrollment in hospice services at the end of life may be one method of transitioning to lower intensity care.
Current health care policies focusing on quality at the end of life correlate lower intensity care with higher quality. Continued efforts to link the delivery of high-quality care to reimbursement in emerging health care policies, such as the Oncology Care Model, further emphasize the important role the health care delivery system plays in improving care at the end of life for patients with cancer. Many believe the type of health system a patient is treated in may impact end-of-life care through an emphasis on care coordination or through the provision of oncology-specific services that enhance end-of-life education, communication, early planning, and decision making for patients with cancer.10-13 Integrated delivery networks (IDNs) that have a specific emphasis on communication and coordination of care across the disease continuum may elicit patient preferences earlier and have improved methods and motivation for decreasing high-intensity care at the end of life, such as more robust hospice enrollment. Examples of IDNs include Kaiser Permanente, Intermountain Healthcare, and Mayo Clinic. Alternatively, health systems focused on caring for patients with cancer, such as National Cancer Institute– and Commission on Cancer–designated programs, may have optimized processes to anticipate, navigate, and plan for end-of-life care for patients with cancer. Whether these characteristics translate to differential outcomes at the end of life for Medicare beneficiaries with cancer remains unknown. Moreover, the extent to which hospice enrollment, which may mitigate the use of high-intensity care at the end of life, is associated with the structure of a delivery system is undetermined.
In this study, we use SEER registry data linked with Medicare claims data to evaluate the association between health care delivery system structure and performance on seven nationally recognized end-of-life quality measures for patients dying from cancer. We examine delivery system types including integrated systems, systems with a cancer focus, systems that are both integrated and certified cancer centers, and systems that are neither. We hypothesize that patients with cancer treated at integrated delivery systems, cancer-focused delivery systems, or systems with both an integration and cancer focus will receive lower intensity care at the end of life.
PATIENTS AND METHODS
Data Sources
We used 3 data sets to perform these analyses. First, we used SEER-Medicare–linked claims data from 2008 through 2014 to identify eligible patients, demographic and clinical information, and our outcomes of interest. Within SEER-Medicare data, we specifically included and analyzed claims from the Medicare Provider Analysis and Review, Carrier, Outpatient, and Hospice files. Second, we used the American Hospital Association Annual Survey to determine hospital characteristics, including teaching status, number of hospital beds, and designation as an accredited American College of Surgeons Commission on Cancer program. Finally, we used IQVIA’s (Durham, NC; formerly IMS Health) Health Care Organization Services data set to determine level of integration for specific hospitals and health systems, which allowed us to group individual hospitals and clinics into larger delivery systems.
Identification of the Study Sample
We identified patients age 66 to 99 years old who died of breast, colon, lung, liver, esophageal, ovarian, prostate, bladder, kidney, or pancreatic cancer between 2009 and 2014 using the Patient Entitlement and Diagnosis Summary and File of the SEER-Medicare data. From this file, we also obtained the date of death to establish the last 12 months of life. To ensure complete claims data for the study cohort, we excluded patients who did not have continuous enrollment in Medicare Parts A and B or who were enrolled in a health maintenance organization at any point during our study interval.
Definition of Delivery System Types
For this study, we defined 4 distinct types of delivery systems based on level of integration and cancer focus. Using previously described methodology, we identified a delivery system as integrated when it was listed in Becker’s Hospital Review’s list of the top 100 integrated delivery systems.14,15 A delivery system was considered cancer focused when it was accredited by the American College of Surgeons Commission on Cancer or part of a National Cancer Institute–designated cancer center.16 Each delivery system was then assigned to 1 of the following 4 distinct types: integrated, cancer focused, both cancer focused and integrated, or neither cancer focused nor integrated.
We used a plurality attribution model to assign each patient to a delivery system based on the number of Medicare Provider Analysis and Review (MedPAR) claims. Patients with no MedPAR claims were excluded from the study. For patients with equal numbers of MedPAR claims, we assigned delivery system based on the longest relationship calculated by the longest length of stay at a hospital.
Identification of End-of-Life Quality Measures and Outcomes
Measures to assess quality of care at the end of life for patients with cancer have been developed and endorsed by the National Quality Forum and ASCO (Table 1).17 We evaluated each measure at the patient level using claims-based algorithms to identify higher intensity of care defined by each quality measure (eg, multiple hospitalizations in the last 30 days of life represents higher intensity care; Appendix and Appendix Table A1, online only). For each measure, a higher percentage represents a greater proportion of patients receiving higher intensity care. Finally, we created an all-or-none measure where patients were classified as receiving higher intensity of care at the end of life if they met any measure of high-intensity care.18
TABLE 1.
Nationally Endorsed End-of-Life Quality Measures
Statistical Analysis
First, we compared patient characteristics from the SEER-Medicare files (ie, marital status, age, race, ethnicity, rurality, sex, type of cancer, and county poverty rate) and hospital characteristics from the American Hospital Association Annual Survey (ie, bed size and teaching hospital status) according to delivery system type using χ2 tests.
Second, we fit multivariable logistic regression models to estimate the percentage of patients meeting each individual quality measure of high-intensity end-of-life care for each of the delivery system types. We adjusted for marital status, age, race, ethnicity, rurality, sex, cancer type, county poverty rate, comorbidities using Hierarchical Condition Categories, hospital bed size, teaching hospital status, and dual eligibility. In these models, we used Huber-White sandwich estimators to obtain robust SEs to account for the correlated nature of our data. In addition, we separately modeled our all-or-none measure for each delivery system type.
Next, we evaluated the relationship between hospice utilization and receipt of high-intensity care across all delivery systems using multivariable logistic regression models and adjusting for the same covariates as in the primary model. We then used χ2 tests to compare receipt of high-intensity end-of-life care with utilization of hospice for all patients stratified by delivery system type. Finally, we performed a sensitivity analysis to determine whether the differences in utilization of high-intensity end-of-life care were primarily driven by hospice utilization. For this analysis, we used a multivariable logistic regression and adjusted for the same variables as in our primary model, but we also included the percentage of hospice utilization at the delivery system in the model.
RESULTS
We identified 100,549 patients meeting inclusion criteria who died of one of ten cancers between 2009 and 2014 in our SEER-Medicare cohort. Overall, 47% of patients received care in a cancer center–only delivery system, 9% were treated in a system that was integrated only, 23% were treated in a delivery system that was affiliated with both a cancer center and an IDN, and 21% were treated in a delivery system that was neither a cancer center nor an IDN.
Patient and Delivery System Characteristics
Table 2 lists descriptive statistics for patients treated in the various delivery system subtypes. The distribution of cancer type did not differ significantly across delivery systems, with lung and colorectal cancers composing > 60% (48% lung and 14% colorectal) of deadly cancers at all delivery systems. Similarly, the proportions of male and female patients were similar across delivery system subtypes. Health systems that were not affiliated with a cancer center or an IDN had fewer hospital beds (66% with < 200 hospital beds v 15% of cancer centers with < 200 beds; P < .001), were less likely to be teaching hospitals (21% v 57% of cancer centers; P < .001), treated a higher proportion of racial and ethnic minorities (83% white v 90% white at IDNs and 8% Hispanic v 4% Hispanic at IDNs; P < .001 for both), treated patients with a greater poverty burden (27% of patients in the highest quartile of census poverty level v 18% at cancer centers; P < .001), and were more likely to treat patients from a rural location (32% v 14% at cancer centers; P < .001).
TABLE 2.
Patient Characteristics Stratified by Delivery System Type
Multivariable Models
The results of multivariable models examining the relationship between delivery system type and receipt of high-intensity care at the end of life are presented in Figure 1 as adjusted percentages and in Appendix Table A2 (online only) as adjusted risk ratios. Delivery systems that were not affiliated with a cancer center or IDN had higher rates of hospitalization in the last 30 days of life (11.3%), of death in an acute care setting (25.9%), and of not utilizing hospice in the last year of life (31.6%; all P < .001). Patients treated at a delivery system affiliated with a cancer center had the highest likelihood of receiving chemotherapy in the last 14 days of life (3.9%; P < .001). Rates of intensive care unit admissions in the last 30 days, of multiple emergency department visits in the last 30 days, and of short duration of hospice utilization (< 3 days) were not significantly associated with delivery system type.
Fig 1.
Multivariable model examining the relationship between the type of delivery system and receipt of high-intensity care at the end of life. (*) P < .05. CC, cancer center; ICU, intensive care unit; IDN, integrated delivery network.
For our all-or-none measure reflecting overall receipt of any high-intensity care at the end of life, delivery systems that were affiliated with a cancer center and an IDN performed best (47.7% received higher intensity care), whereas delivery systems with no affiliation performed slightly worse (51.4% received high-intensity care; P < .001).
Hospice Utilization
For each measure examined, patients who used hospice services in the last year of life were less likely to receive high-intensity end-of-life care (Fig 2). This is reflected in fewer patients with multiple hospitalizations and emergency department visits in the last 30 days of life, fewer intensive care unit admissions, and lower use of chemotherapy in the last 14 days of life. In particular, 75.5% of patients not enrolled in hospice died in an acute care setting compared with only 2.4% of patients who used hospice (P < .001). Unadjusted comparisons of the association between higher intensity care at the end of life and hospice utilization stratified by delivery system type showed a similar pattern. Our sensitivity analysis showed that hospice utilization was not the main driver of other measures of high-intensity care because the size and significance of the differences across delivery systems persisted despite adjustment for hospice utilization in the model.
Fig 2.
Relationship between use of hospice services and receipt of high-intensity end-of-life care. ICU, intensive care unit.
DISCUSSION
Our study has two principal findings. First, the type of delivery system had only a modest impact on the intensity of care that patients with cancer received at the end of life. At delivery systems without a cancer center or IDN affiliation, patients experienced marginally higher intensity of care at the end of life compared with the other delivery system types. This may reflect fewer disease-specific programs, less infrastructure to support care coordination and well-aligned communication, and limited resources for specific end-of-life services. Second, the use of hospice is associated with intensity of care, with patients enrolled in hospice receiving less intense care at the end of life across all delivery system types.
Our finding showing only modest impact of delivery system on care delivery at the end of life is consistent with several studies evaluating the cost and quality of cancer care. In prior work, we demonstrated minor differences in spending according to delivery system integration across the initial, continuing, and end-of-life phases of care for patients with cancer.15 Intensity of end-of-life care remains heavily influenced by patient treatment preferences and physician attitudes.19 Delivery systems may be uniquely positioned to influence care through systems-based programs to avoid readmissions, improve quality of care and coping strategies, and foster end-of-life care planning and symptom management.20-23
Hospice utilization as a means to facilitate lower intensity care at the end of life and cost savings has been well studied; however, results of prior analyses show that hospice mediates a small portion of the overall variation in end-of-life spending.24,25 Despite uncertain savings, benefits of hospice utilization importantly include higher quality of end-of-life care and greater satisfaction on the part of the family.9,26 Previous work evaluating the use of hospice at the end of life highlighted a potential challenge for patients living in rural areas where there may be limited availability of hospice resources.27 In this study, we found that hospitals with no cancer center or IDN affiliation are located in more rural areas and that these delivery systems are less likely to use hospice. However, these differences in hospice utilization according to delivery system type are relatively small, and perhaps access to and utilization of hospice are improving nationally.
Our study has several limitations. First, cancer centers and IDNs, in particular, are not monolithic. The degree to which each delivery system focuses on and implements aspects of care coordination and communication, specifically for patients with cancer at the end of life, likely varies widely across health systems. Second, quality of care at the end of life is highly personal and depends on human interactions between patients, families, physicians, and other providers within the context of a complex health system. We acknowledge this sensitivity by focusing on the intensity of care provided at the end of life, not by describing whether that care was high or low quality. Although nationally endorsed claims-based quality measures can reflect broad assessments of intensity of care and are often used to assess quality in emerging health policies such as the Oncology Care Model, more granular data, including patient and family preferences and experiences, will be necessary to improve quality of care at the end of life at the individual patient level.28 Third, although cancer-focused delivery systems may have improved ability to anticipate, navigate, and plan for end-of-life cancer care, there are also forces that may increase the intensity of care with greater availability of clinical trials, which may promote chemotherapy and other high-intensity care at the end of life. In addition, patients who seek care at cancer-focused systems may have goals that are more aligned with high-intensity end-of-life care. Finally, prognostication and the ability to accurately predict the timing of death in patients with cancer remain an ongoing challenge for patients and providers. This uncertainty complicates decisions of when to pursue lower or higher intensity of care, such as choices to pursue another round of chemotherapy or consider hospice enrollment.
These limitations notwithstanding, our findings have important implications for patients, physicians, health system leadership, and policymakers. First, the type of care that patients receive at the end of life may be influenced by the health system where they are treated. Seeking care at a delivery system with a cancer center or IDN affiliation that emphasizes early communication and uses a patient-centered approach may prevent high-intensity care and potentially optimize quality of care at the end of life compared with unaffiliated systems. Specifically, health systems could focus on early communication with patients and families to develop end-of-life plans congruent with their values. This may include routine communication among care navigation teams, oncologists, and primary care providers to potentially avoid emergency department visits, hospitalizations, and intensive care unit admissions, which can spiral into high-intensity care that may not be in accordance with patient and family goals. Making social services, pain and symptom management, palliative care, and hospice resources readily available to patients may allow for earlier access to lower intensity care, when appropriate, thereby avoiding unwanted high-intensity care for patients near the end of life. Similarly, physicians and health system leaders can work together to build infrastructure that supports early end-of-life education, communication, consideration of hospice services, and decision making for patients with cancer. Despite unclear financial benefits, hospice utilization may be one mechanism for improving quality and lowering intensity of care at the end of life. Finally, policymakers should continue to encourage the delivery of high-quality care at the end of life through policies that require measurement of and that financially reward adherence with nationally recognized and endorsed quality measures, such as programs like the Oncology Care Model. Perhaps stronger, well-constructed incentives will drive more significant improvements in quality of care than we see from delivery system type alone. Because most health systems have a more diffuse approach to improving care coordination and communication, creating real change in utilization at the end of life may require specific programs focused on this outcome.
Overall quality of care at the end of life for patients with cancer is impacted only modestly by delivery system type, but clear opportunities for improvement exist, with many patients still receiving high-intensity and high-expense care at the end of life. Although care at the end of life remains and will always be deeply personal for patients, local and health system factors, including a better understanding of care delivery, and patient and physician preferences at the end of life are needed to inform how health systems can play a role in supporting the delivery of patient-centered, high-quality care for patients with cancer. Maximal supportive care delivered through hospice may be one avenue to reduce high-intensity care at the end of life and impact quality of care for patients dying from cancer. As payment systems transform to value-based payment systems, policies that promote and reward providers who make hospice available for their patients may meet both aspects of the value equation.
ACKNOWLEDGMENT
Supported by National Cancer Institute Grant No. 1-R01-CA-174768-01-A1 (D.C.M.). Presented in part at the 2019 American Urological Association Annual Meeting, Chicago, IL, May 3-6, 2019.
APPENDIX
Outcome 5. Chemotherapy
Claim Level Flag Criteria
MedPAR.
Any one of the dgn_cd starts with “V581” or any one of the SRGCDE1-SRGCDE25 is in the following ICD9 list:
'0010' '1770' '9925' '9928'
“3E00X05” “3E00X0M” “3E01305” “3E0130M” “3E02305” “3E0230M”
“3E0300P” “3E03303” “3E03305” “3E0330M” “3E0330P” “3E0400P”
“3E04303” “3E04305” “3E0430M” “3E0430P” “3E0500P” “3E05303”
“3E0530M” “3E0530P” “3E0600P” “3E06303” “3E0630M” “3E0630P”
“3E0A305” “3E0F305” “3E0F705” “3E0F805” “3E0G305” “3E0G705”
“3E0G805” “3E0H305” “3E0H705” “3E0H805” “3E0J305” “3E0J705”
“3E0J805” “3E0K305” “3E0K705” “3E0K805” “3E0L305” “3E0L705”
“3E0M305” “3E0M705” “3E0N305” “3E0N705” “3E0N805” “3E0P305”
“3E0P705” “3E0P805” “3E0Q005” “3E0Q305” “3E0Q705” “3E0R305”
“3E0S305” “3E0V305” “3E0W305” “3E0Y305” “3E0Y705” “XW03351”
“XW04351”
'96400','96408', '96409', '96410', '96411', '96412','96412',
'96413', '96414', '96414', '96415', '96416', '96417', '96418',
'96419', '96419', '96420', '96421', '96422', '96423', '96424'
'96425', '96520'
NCH (Carrier).
Any one of pdgns_cd, dgn_cd1 to dgn_cd12 starts with 'V581' or any hcpcs within the following HCPCS list:
'0519F' '36823' '51720' '61517' '95990' '95991' '96400' '96401'
'96402' '96403' '96404' '96405' '96406' '96407' '96408' '96409'
'96410' '96411' '96412' '96413' '96414' '96415' '96416' '96417'
'96418' '96419' '96420' '96421' '96422' '96423' '96424' '96425'
'96426' '96427' '96428' '96429' '96430' '96431' '96432' '96433'
'96434' '96435' '96436' '96437' '96438' '96439' '96440' '96441'
'96442' '96443' '96444' '96445' '96446' '96447' '96448' '96449'
'96450' '96451' '96452' '96453' '96454' '96455' '96456' '96457'
'96458' '96459' '96460' '96461' '96462' '96463' '96464' '96465'
'96466' '96467' '96468' '96469' '96470' '96471' '96472' '96473'
'96474' '96475' '96476' '96477' '96478' '96479' '96480' '96481'
'96482' '96483' '96484' '96485' '96486' '96487' '96488' '96489'
'96490' '96491' '96492' '96493' '96494' '96495' '96496' '96497'
'96498' '96499' '96500' '96501' '96502' '96503' '96504' '96505'
'96506' '96507' '96508' '96509' '96510' '96511' '96512' '96513'
'96514' '96515' '96516' '96517' '96518' '96519' '96520' '96521'
'96522' '96523' '96524' '96525' '96526' '96527' '96528' '96529'
'96530' '96531' '96532' '96533' '96534' '96535' '96536' '96537'
'96538' '96539' '96540' '96541' '96542' '96543' '96544' '96545'
'96546' '96547' '96548' '96549' 'C8953' 'C8954' 'C8955' 'G0355'
'G0357' 'G0358' 'G0359' 'G0360' 'G0361' 'G0362' 'G0370' 'J7150'
'Q0083' 'Q0084' 'Q0085' 'S5019' 'S5020' 'S9329' 'S9330' 'S9331'
'S9425' 'C1086' 'C1166' 'C1167' 'C1178' 'C9012' 'C9110' 'C9127'
'C9205' 'C9207' 'C9213' 'C9214' 'C9215' 'C9217' 'C9218' 'C9235'
'C9257' 'C9262' 'C9414' 'C9415' 'C9417' 'C9418' 'C9419' 'C9420'
'C9421' 'C9422' 'C9423' 'C9424' 'C9425' 'C9426' 'C9427' 'C9429'
'C9431' 'C9432' 'C9433' 'C9437' 'C9440' 'J0594' 'J0894' 'J8510'
'J8520' 'J8521' 'J8530' 'J8560' 'J8565' 'J8600' 'J8610' 'J8700'
'J8705' 'J8999' 'J9000' 'J9001' 'J9010' 'J9017' 'J9020' 'J9025'
'J9027' 'J9033' 'J9035' 'J9040' 'J9041' 'J9045' 'J9050' 'J9055'
'J9060' 'J9062' 'J9065' 'J9070' 'J9080' 'J9090' 'J9091' 'J9092'
'J9093' 'J9094' 'J9095' 'J9096' 'J9097' 'J9098' 'J9100' 'J9110'
'J9120' 'J9130' 'J9140' 'J9150' 'J9151' 'J9170' 'J9171' 'J9178'
'J9180' 'J9181' 'J9182' 'J9185' 'J9190' 'J9200' 'J9201' 'J9206'
'J9207' 'J9208' 'J9211' 'J9230' 'J9245' 'J9250' 'J9260' 'J9261'
'J9263' 'J9264' 'J9265' 'J9266' 'J9268' 'J9270' 'J9280' 'J9290'
'J9291' 'J9293' 'J9300' 'J9303' 'J9305' 'J9307' 'J9310' 'J9315'
'J9320' 'J9328' 'J9330' 'J9340' 'J9350' 'J9351' 'J9355' 'J9357'
'J9360' 'J9370' 'J9375' 'J9380' 'J9390' 'J9999' 'Q2017' 'Q2024'
'S0087' 'S0088' 'S0115' 'S0116' 'S0172' 'S0176' 'S0178' 'S0182'
'96400','96408', '96409', '96410', '96411', '96412','96412',
'96413', '96414', '96414', '96415', '96416', '96417', '96418',
'96419', '96419', '96420', '96421', '96422', '96423', '96424'
'96425', '96520';
Outpatient.
Any one of the dgn_cd starts with 'V581' or any hcpcs within the same list as of NCH.
TABLE A1.
Coding Definitions for Each Quality Measure
TABLE A2.
Adjusted Risk Ratios (with 95% CI) for Each Measure for Each Delivery System Type
AUTHOR CONTRIBUTIONS
Conception and design: Lindsey A. Herrel, Deborah R. Kaye, Chandy S. Ellimoottil, David C. Miller
Financial support: Lindsey A. Herrel, David C. Miller
Provision of study materials or patients: David C. Miller
Collection and assembly of data: Lindsey A. Herrel, David C. Miller
Data analysis and interpretation: Lindsey A. Herrel, Ziwei Zhu, Jennifer J. Griggs, Deborah R. Kaye, James M. Dupree, David C. Miller
Manuscript writing: All authors
Final approval of manuscript: All authors
Accountable for all aspects of the work: All authors
AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
Association Between Delivery System Structure and Intensity of End-of-Life Cancer Care
The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/op/authors/author-center.
Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).
Jennifer J. Griggs
Employment: Anglona Corporation
Stock and Other Ownership Interests: Anglona Corporation
Deborah R. Kaye
Employment: Blue Cross Blue Shield of North Carolina (I)
Research Funding: Blue Cross Blue Shield of Michigan
Travel, Accommodations, Expenses: MedReviews
James M. Dupree
Stock and Other Ownership Interests: Lipocine
Other Relationship: Blue Cross Blue Shield of Michigan
David C. Miller
Research Funding: Blue Cross Blue Shield of Michigan
No other potential conflicts of interest were reported.
REFERENCES
- 1. doi: 10.1001/archinternmed.2010.394. Bergman J, Saigal CS, Lorenz KA, et al: Hospice use and high-intensity care in men dying of prostate cancer. Arch Intern Med 171:204-210, 2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. doi: 10.1002/cncr.25217. Wright AA, Mack JW, Kritek PA, et al: Influence of patients’ preferences and treatment site on cancer patients’ end-of-life care. Cancer 116:4656-4663, 2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. doi: 10.1377/hlthaff.2011.0650. Morden NE, Chang C-H, Jacobson JO, et al: End-of-life care for Medicare beneficiaries with cancer is highly intensive overall and varies widely. Health Aff 31:786-796, 2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. doi: 10.1377/hlthaff.2018.0015. Keating NL, Huskamp HA, Kouri E, et al: Factors contributing to geographic variation in end-of-life expenditures for cancer patients. Health Aff 37:1136-1143, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. doi: 10.1200/JCO.2009.26.3863. Wright AA, Keating NL, Balboni TA, et al: Place of death: Correlations with quality of life of patients with cancer and predictors of bereaved caregivers’ mental health. J Clin Oncol 28:4457-4464, 2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. doi: 10.1001/jama.2015.18604. Wright AA, Keating NL, Ayanian JZ, et al: Family perspectives on aggressive cancer care near the end of life. JAMA 315:284-292, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Ellershaw J, Ward C: Care of the dying patient: The last hours or days of life. BMJ 326:30-34, 2003. [PMC free article] [PubMed] [Google Scholar]
- 8. doi: 10.1001/jama.284.19.2476. Steinhauser KE, Christakis NA, Clipp EC, et al: Factors considered important at the end of life by patients, family, physicians, and other care providers. JAMA 284:2476-2482, 2000. [DOI] [PubMed] [Google Scholar]
- 9. doi: 10.1200/JCO.2016.68.9257. Kumar P, Wright AA, Hatfield LA, et al: Family perspectives on hospice care experiences of patients with cancer. J Clin Oncol 35:432-439, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. doi: 10.1001/jamaoncol.2019.0291. Kiely BE, Stockler MR: Discussing prognosis, preferences, and end-of-life care in advanced cancer. JAMA Oncol 5:788-789, 2019. [DOI] [PubMed] [Google Scholar]
- 11. doi: 10.1001/jamaoncol.2019.0292. Paladino J, Bernacki R, Neville BA, et al: Evaluating an intervention to improve communication between oncology clinicians and patients with life-limiting cancer. JAMA Oncol 5:801-809, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. doi: 10.7326/0003-4819-154-4-201102150-00004. Kelley AS, Ettner SL, Morrison RS, et al: Determinants of medical expenditures in the last 6 months of life. Ann Intern Med 154:235-242, 2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. doi: 10.1016/j.jpainsymman.2016.11.012. Rocque GB, Dionne-Odom JN, Sylvia Huang C-H, et al: Implementation and impact of patient lay navigator-led advance care planning conversations. J Pain Symptom Manage 53:682-692, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. doi: 10.1016/j.juro.2016.07.040. Herrel LA, Kaufman SR, Yan P, et al: Health care integration and quality among men with prostate cancer. J Urol 197:55-60, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. doi: 10.1200/JOP.2017.027730. Kaye DR, Min HS, Norton EC, et al: System-level health-care integration and the costs of cancer care across the disease continuum. J Oncol Pract 14:e149-e157, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. National Cancer Institute: NCI-designated cancer centers. https://www.cancer.gov/research/nci-role/cancer-centers.
- 17. doi: 10.1093/intqhc/mzi061. Earle CC, Neville BA, Landrum MB, et al: Evaluating claims-based indicators of the intensity of end-of-life cancer care. Int J Qual Health Care 17:505-509, 2005. [DOI] [PubMed] [Google Scholar]
- 18. doi: 10.1001/jama.295.10.1168. Nolan T, Berwick DM: All-or-none measurement raises the bar on performance. JAMA 295:1168-1170, 2006. [DOI] [PubMed] [Google Scholar]
- 19. doi: 10.1016/j.jpainsymman.2019.04.014. Duberstein PR, Kravitz RL, Fenton JJ, et al: Physician and patient characteristics associated with more intensive end-of-life care. J Pain Symptom Manage 58:208-215.e1, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. doi: 10.1200/JOP.18.00595. Johnson PC, Xiao Y, Wong RL, et al: Potentially avoidable hospital readmissions in patients with advanced cancer. J Oncol Pract 15:e420-e427, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. doi: 10.1093/annonc/mdy488. Nipp RD, El-Jawahri A, Ruddy M, et al: Pilot randomized trial of an electronic symptom monitoring intervention for hospitalized patients with cancer. Ann Oncol 30:274-280, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. doi: 10.1200/JCO.2017.75.6676. Hoerger M, Greer JA, Jackson VA, et al: Defining the elements of early palliative care that are associated with patient-reported outcomes and the delivery of end-of-life care. J Clin Oncol 36:1096-1102, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. doi: 10.1200/JCO.2017.73.7221. Greer JA, Jacobs JM, El-Jawahri A, et al: Role of patient coping strategies in understanding the effects of early palliative care on quality of life and mood. J Clin Oncol 36:53-60, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. doi: 10.1377/hlthaff.2016.0683. Wang S, Hsu SH, Huang S, et al: Longer periods of hospice service associated with lower end-of-life spending in regions with high expenditures. Health Aff 36:328-336, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. doi: 10.6004/jnccn.2016.0107. Wang S-Y, Hall J, Pollack CE, et al: Associations between end-of-life cancer care patterns and medicare expenditures. J Natl Compr Canc Netw 14:1001-1008, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. doi: 10.1002/cncr.32188. Trevino KM, Prigerson HG, Shen MJ, et al: Association between advanced cancer patient‐caregiver agreement regarding prognosis and hospice enrollment. Cancer 125:3259-3265, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. doi: 10.1089/jpm.2006.9.1292. Virnig BA, Ma H, Hartman LK, et al: Access to home-based hospice care for rural populations: Identification of areas lacking service. J Palliat Med 9:1292-1299, 2006. [DOI] [PubMed] [Google Scholar]
- 28. doi: 10.1200/JCO.2012.45.7473. Ryoo JJ, Ordin DL, Antonio ALM, et al: Patient preference and contraindications in measuring quality of care: What do administrative data miss? J Clin Oncol 31:2716-2723, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.ASCO ASCO QOPI Reporting Registry. https://practice.asco.org/sites/default/files/drupalfiles/QCDR-2019-Measure-Summary.pdf
- 30.National Quality Forum Palliative care and end-of-life care. https://www.qualityforum.org/Topics/Palliative_Care_and_End-of-Life_Care.aspx