Table 2.

Summary of Synthesized Studies and Key Findings.

Author (Year)	Study design (evidence quality)	Population delineation (sample size)	Socio-demographic profile of participants	End-of-life related measures	Elements of access addressed						Access disparities identified by demographic
					Approachability	Acceptability	Availability	Affordability	Appropriateness	Key findings	Age	Sex/ Gender	Race/ Ethnicity	Education	SES/ Insurance	Disease/ Severity	Other
Quantitative Studies
Andrews, et al (2014)19	Cross-sectional (III, C)	– Young muscular dystrophy patients (2) – Family caregivers (32)	– 28% Hispanic caregivers – 6% of caregivers spoke better Spanish than English – 88% female caregivers	– Disease severity – Service utilization	–	–	–	✓	✓	– Disease severity, per capita income, family history associated with service utilization – Race/ethnicity and level of acculturation not significantly associated with service utilization	–	–	–	–	✓	✓	✓
Brown, et al (2018)8	Retrospective cohort analysis (III, B)	– Patients who died from chronic illness (22,068)	– 82% White, 6% Asian, 6% Black, 2% Hispanic, 2% Native American – 57% Male, 43% Female – 27% had a 4-year college education or higher	*Within the last 30 days of life: – Intensive care unit admission – Use of mechanical ventilation – Receipt of CPR	–	–	–	✓	✓	– Race/ethnicity, SES, and lower income significantly associated with all 3 types of high-intensity care in last 30 days of life – Lower levels of education significantly associated with mechanical ventilation and ICU care in last 30 days of life – Racial/ethnic minority groups more likely to be admitted to ICU and receive mechanical ventilation than white non-Hispanics	–	–	✓	✓	✓	–	–
Gidwani-Marszowski, et al (2018)20	Retrospective cohort analysis (III, C)	– Veteran patients who died of solid tumors (87,251)	– 87% White – 99% Male – 52% Medicare-reliant, 46% VA-reliant	*Within the last 30 days of life: – Quality of care – Chemotherapy use – Intensive care unit/hospital admission or death in the hospital – Number of days in the hospital	–	–	✓	✓	✓	– Medicare-reliant patients had more intensive (lower quality) EOL care than VA-reliant patients in terms of chemotherapy, hospital and ICU admissions, number of days in the hospital and death in the hospital – Race, urbanicity associated with more ED visits and hospitalizations	–	–	✓	–	✓	–	✓
Givens, et al (2010)21	Retrospective cohort analysis (III, B)	Heart failure patients (98,258)	– 88% White, 9% Black, 1% Hispanic – 39% Male – 76% had median income <$45,000	– Hospice entry	–	✓	✓	–	–	– Blacks/Hispanics used hospice less for any diagnosis than Whites after adjusting for income, illness severity, and urbanicity – For heart failure, Hispanics had longest median duration (19 days), followed by Blacks (14 days), and Whites (13 days) – Comorbidity burden, advanced age, ED visits/hospitalizations, and local hospice density significantly associated with hospice utilization	✓	–	✓	–	–	✓	✓
Haines, et al (2018)22	Retrospective analysis (III, B)	Adult trauma patients (2,966,444)	*Among those discharged to hospice: – 89% White – 52% Female – 71% Medicare reliant	– Disposition to hospice – Hospital LOS	–	✓	–	✓	✓	– Blacks, Hispanics, and Asians significantly less likely to receive hospice than Whites – Age, gender, and type of disease significantly associated with disposition to hospice – Medicare patients transferred to hospice sooner than uninsured patients who had longer hospital LOS	✓	✓	✓	–	✓	✓	–
Hoerger, et al (2019)23	Descriptive (III, B)	States (50)	*At the state-level, on average: – 69% White – 50.6% Female – $56,406 Median income	– PC access – Personality (Openness)	–	✓	✓	✓	–	– Greater PC access in states where residents were older, White, had higher SES, and were politically liberal – States with higher levels of openness had greater PC access	✓	–	✓	–	✓	–	✓
Hui, et al (2012)24	Retrospective case-control (III, B)	Patients who died of advanced cancer (816)	*Of patients who had a PC referral: – 58% White, 22% Black, 14% Hispanic, 5% Asian, 1% Other – 53% Female – 64% had a college education or higher	– PC consults and referrals	–	✓	✓	–	✓	– Patients with PC consults were younger, female, and more likely to be married – Patients with GYN, breast, and GI cancers were more likely to have PC access – Majority of patients had an average of 20 medical team encounters prior to PC referral and/or did not see PC at all before death	✓	✓	–	–	–	✓	✓
Kumar, et al (2012)25	Cross-sectional (III, B)	Cancer patients (313)	– 77% White, 18% Black, 2% Asian, 2% Hispanic, 1% Native American – 72% had a college education or higher	– Use of SPCS (counseling, cancer support group, cancer rehabilitation, and PC consult)	–	✓	–	✓	–	– Most commonly used SPCS was counseling (30%); least common were PC consultation (8%) and cancer rehabilitation (4%) – SPCS users were more likely to be female and have a higher level of education – Type of cancer was significantly associated with SPCS use	–	✓	–	✓	–	✓	–
Lee, et al (2013)26	Descriptive (III, C)	Pancreatic cancer patients (1,008)	– 15% Black – 65.3% Male	– Surgery and adjuvant therapy use for locoregional disease – Chemotherapy use for distant disease	–	–	–	–	✓	– Race was not significantly associated with surgical resection, adjuvant therapy/chemotherapy use – Severity of disease was significantly associated with surgical outcomes	–	–	–	–	–	✓	–
Okafor, et al (2016)27	Retrospective analysis (III, B)	Patients with metastatic malignancies of the colon (217,055)	*Among White patients: – 48% Female – 58% Medicare reliant; 6% Medicaid reliant *Among Black patients: – 40% Female – 54% Medicare reliant; 21% Medicaid reliant	– Colorectal stent utilization – Hospital LOS	–	–	–	✓	–	– Medicare was the most common payer for the procedures across all races; Blacks incurred the highest average total charges ($156,876) – Higher SES was significantly associated with greater utilization – Lower SES was significantly associated with greater hospital LOS and charges	–	–	✓	–	✓	–	✓
Sharma, et al (2015)28	Retrospective analysis (III, B)	Metastatic cancer patients (6,288)	– 69% White, 19% Black, 6% Hispanic, 7% Other *Of patients reliant on Medicaid: – 20% Hispanic, 17% Black, 3% White	– Receipt of inpatient PC consultation	–	–	✓	–	✓	– Black/lower SES patients were significantly more likely than Whites to have an inpatient PC consult and to be referred to hospice – Advanced illness and increased frequency of hospitalizations were significantly associated with greater odds of having an inpatient PC consult	–	–	✓	–	✓	✓	–
Silveira, et al (2011)29	Observational (III, B)	U.S. counties (3,140)	*At the county level, on average: – 9% Black, 6% Hispanic – 51% had a high-school education	– Hospice program availability	–	–	✓	✓	–	– Hospice availability was significantly associated with urbanicity – Population size, median household income, race, and certificate of need – positively predicted hospice count per county – Area, ethnicity, education, and age negatively predicted hospice count per county	✓	–	✓	✓	✓	–	✓
Singh, et al (2017)30	Retrospective observational (III, B)	Stroke patients (395,411)	– 69% White – 52% Female *Of patients who had a PC encounter – 7% White, 5% Hispanic, 4% Black	– PC encounter – Death during hospitalization	–	✓	✓	–	✓	– Age, sex, and race were significantly associated with PC use – PC encounters were significantly associated with large, urban teaching, and non-profit hospitals	✓	✓	✓	–	–	–	✓
Stewart, et al (2016)31	Descriptive (III, B)	Hepatocellular cancer patients (33,270)	– 38% White, 26% Hispanic, 8% Chinese, 8% Black, 6% Vietnamese, 4% Filipino, 3% Korean, 1% American Indian/Alaska Native	– Receipt of surgical treatment – Cause-specific and all-cause mortality	–	✓	–	✓	✓	– Age, race, gender, and SES were significantly associated with odds of receiving surgical treatment – Lower mortality was significantly associated with younger age, female gender, earlier stage disease, higher SES, and later time period of diagnosis	✓	✓	✓	–	✓	✓	–
Watanabe-Galloway, et al (2014)32	Descriptive (III, B)	Patients who died from colorectal cancer (34,975)	*Non-Hispanic White: – 91% of rural beneficiaries, 91% of micropolitan beneficiaries, and 83% of metropolitan beneficiaries *Female sex: – 52% of rural beneficiaries, 51% of micropolitan beneficiaries, and 53% of metropolitan beneficiaries	– ER visits, inpatient hospital admissions, and number of ICU days in the last 3 months of life – Proportion of people who used hospice – Proportion of people who enrolled in hospice less than 3 days before death, among those who ever used hospice	–	✓	✓	✓	✓	– Older age and female sex were significantly associated with fewer ER visits and hospitalizations, greater hospice use – Race/ethnicity, lower SES, and rurality were significantly associated with less hospice use, greater use of ED, inpatient care, and ICU in last 3 months of life	✓	✓	✓	–	✓	–	✓
Worster, et al (2018)33	Retrospective analysis (III, C)	Patients referred to PC services (3,207)	– 65% White, 27% Black, 3% Hispanic, 3% Asian/Pacific Islander, 1% Other – 52% Female	– Time to inpatient PC consult – Disposition to hospice – Hospital LOS	–	✓	–	–	✓	– Race was not a significant – predictor of time to PC consult, hospice enrollment, or increased hospital LOS	–	–	–	–	–	✓	–
Qualitative Studies
Isaacson, et al (2015)34	Qualitative (III, B)	Health Care Providers (7)	– 57% Native American/Alaska Indian, 43% non-Native	– Perspectives on PC specific to Native Americans – Availability of PC on reservations	✓	✓	✓	✓	–	– Cited lack of funding, and infrastructure as main challenges for implementing PC widely – Both Native and non-Native providers noted misconceptions regarding PC/hospice as a barrier to patient engagement	–	–	–	–	–	–	✓
Johnston, et al (2019)35	Qualitative (III, B)	Multiple stakeholders (Patients, caregivers, providers, CHWs) (24)	– 13% Black patients – 4% caregivers – 46% providers – 13% CHWs	– Views and attitudes toward patient navigation and PC use for Blacks with advanced malignancies	✓	✓	–	–	✓	– Stated CHWs help bridge barriers in PC to enhance patient-centered care in a culturally sensitive manner – Inconsistent messaging and branding is a significant barrier to PC use	–	–	✓	–	–	–	✓
Kavalieratos, et al (2014)36	Qualitative (III, B)	Primary care, cardiology, and palliative care providers (18)	– 67% Physicians – 33% Non-physician providers	– Perceived factors that impede PC referral and access for heart failure patients	–	–	✓	–	–	– Unpredictable trajectory of heart failure can prevent patients from seeking out PC services – Providers had limited knowledge regarding PC and its use in heart failure	–	–	–	✓	–	✓	–
Mixed Methods Study
Periyakoil, et al (2016)37	Mixed- methods (III, A)	Patients and families (387)	– 37% White, 51% Asian American, 12% Black – 67% Female – 64% had a college education or higher	– Barriers to quality EOL care	✓	✓	–	✓	–	– 61% reported barriers to EOL care for persons in their culture/ethnicity – Common barriers noted were finances, communication, and beliefs – Patients with no formal education found financial challenges most difficult, followed by communication barriers between doctors and patients	✓	✓	–	✓	✓	–	✓

CPR = Cardiopulmonary resuscitation; PC = Palliative Care; LOS = Length of stay; CHW = Community health worker; EOL = End of life; SPCS = Supportive and palliative care services.