Abstract
Background:
Palliative care consultation to discuss goals-of-care (“PCC”) may mitigate end-of-life care disparities.
Objective:
To compare hospitalization and cost outcomes by race and ethnicity among PCC patients; identify predictors of hospice discharge and post-discharge hospitalization utilization and costs.
Methods:
This secondary analysis of a retrospective cohort study assessed hospice discharge, do-not-resuscitate status, 30-day readmissions, days hospitalized, ICU care, any hospitalization cost, and total costs for hospitalization with PCC and hospitalization(s) post-discharge among 1,306 Black/African American, Latinx, White, and Other race PCC patients at a United States academic hospital.
Results:
In adjusted analyses, hospice enrollment was less likely with Medicaid (AOR=0.59, P=0.02). Thirty-day readmission was less likely among age 75+ (AOR=0.43, P=0.02); more likely with Medicaid (AOR=2.02, P=0.004), 30-day prior admission (AOR=2.42, P<0.0001), and Black/African American race (AOR=1.57, P=0.02). Future days hospitalized was greater with Medicaid (Coefficient=4.49, P=0.001), 30-day prior admission (Coefficient=2.08, P=0.02), and Black/African American race (Coefficient=2.16, P=0.01). Any future hospitalization cost was less likely among patients ages 65–74 and 75+ (AOR=0.54, P=0.02; AOR=0.53, P=0.02); more likely with Medicaid (AOR=1.67, P=0.01), 30-day prior admission (AOR=1.81, P=0.0001), and Black/African American race (AOR=1.40, P=0.02). Total future hospitalization costs were lower for females (Coefficient=−3616.64, P=0.03); greater with Medicaid (Coefficient=7388.43, P=0.01), 30-day prior admission (Coefficient=3868.07, P=0.04), and Black/African American race (Coefficient=3856.90, P=0.04). Do-not-resuscitate documentation (48%) differed by race.
Conclusions:
Among PCC patients, Black/African American race and social determinants of health were risk factors for future hospitalization utilization and costs. Medicaid use predicted hospice discharge. Social support interventions are needed to reduce future hospitalization disparities.
Keywords: palliative care, end of life, goals of care, communication, racial disparities, hospice, intensive care
INTRODUCTION
Improving patient-clinician communication about patient values, preferences, and end-of-life care planning across racially diverse patient groups is a national priority according to the National Academy of Medicine.1 Although studies show patients across racial and ethnic groups think physicians should discuss end-of-life care issues with them (82% Black/African American, 83% Latinx, 87% White, and 82% “Other” race patients),2 racial and ethnic minority patients are less likely than White patients to receive clinician discussions that clarify personal goals and preferences for future medical care.3–5 Compared to non-Latinx White adults, for example, Latinx adults are more likely to have not engaged in such discussions because they waited for a doctor to start the conversation.2 This goal-sharing disparity is problematic because such discussions support patient decision-making and are associated with better quality of life, enhanced goal-consistent care, improved healthcare communication, and positive family outcomes.6,7 Racism and implicit and explicit clinician biases contribute to racial disparities in communication quality, satisfaction and care outcomes,8–13 which is why it is important to identify interventions that mitigate racial disparities by enabling quality communication across patient populations.
Given disparities in patient-clinician communication about end-of-life wishes3–5 and racial and ethnic minority patients’ lower odds of naming a healthcare agent to make care decisions if needed,2,14 racial and ethnic minorities with serious illness may be disadvantaged in the quality of care they receive if they are unable to make decisions themselves.2 In the context of poorer communication and factors such as socioeconomic status, access to quality healthcare, religiosity, culture, and clinician biases or racism,13,15–18 racial and ethnic minority patients undergo more care transitions near the end-of-life,19 experience more hospital readmissions,20 are more likely to have aggressive end-of-life care,2,21–23 and report lower quality end-of-life care.24–27 Evidence of racial and ethnic minority patient enrollment in hospice is mixed with some studies indicating minority patients are less likely to enroll or stay enrolled in hospice than White patients28–30 and other studies suggesting Black/African American palliative care patients31,32 and Latinx palliative care patients are as likely as White patients30 or, in the case of Latinx palliative care patients, more likely than White patients31 to be discharged to hospice. However, a complex set of variables interact to influence a person’s end-of-life decisions—not race or ethnicity alone.15,33–35 Race’s role in people’s lived experiences and its impact on health and wellbeing36,37 make race and ethnicity relevant non-causative social variables38 when seeking to understand modifiable factors that create disparities.12,15
Although research on goals-of-care consultation and advance care planning in racial and ethnic minority groups is growing, most data focuses on Black/African American patients compared to White patients.39 Data about Latinx, Asian, and self-reported Other race patients is limited, challenging clinicians’ ability to understand needs across underrepresented patient groups and generalize findings to a diverse population. Palliative care’s approach to patient-centric communication is recognized for helping patients determine care goals and potentially reduce disparities in end-of-life care,11,32 but it is unknown if racial and ethnic disparities persist among patients who receive inpatient palliative care consultation to discuss goals-of-care (“PCC”).
The purpose of this study was: (1) to compare among a racially diverse sample of seriously-ill patients who received PCC experiences during index hospitalization, the admission a patient first received PCC during the study period; and future hospitalization outcomes and costs following discharge from index hospitalization (“future costs”), which represent the volume or intensity of hospitalized care a patient received after discharge; and (2) to understand which factors independently contribute to hospice enrollment and differences in post-discharge hospitalization outcomes. Due to long-standing conceptual differences in definitions and self-identification of race and ethnicity,40–43 and evidence that societal factors—not genetic or biological factors—account for racial differences in health outcomes, our study conceptualized race and ethnicity as social-cultural factors.38,44,45 As social constructs,46 race and ethnicity serve as proxy measures for sociopolitical, cultural, structural, economic, and environmental factors and are not to be interpreted as biologic or causative factors.46,47
METHODS
Setting
This study is a secondary analysis of pre-existing clinical and administrative data from a large, retrospective observational study conducted at an urban, academic medical center in the Northeast United States that serves a socioeconomically, racially, and ethnically diverse population. The parent study found PCC was associated with reduced future hospitalization utilization and savings of over $6,000 per patient but did not assess hospitalization outcomes, costs, or hospice enrollment by race or ethnicity among PCC patients.48 Race and ethnicity were not factors in whether a patient received PCC in the parent study48 but are relevant to understanding hospitalization, cost, and hospice outcomes in the context of PCC. To supplement the parent study dataset, we used Medicaid data as a proxy for socioeconomic status49,50 and inpatient source of referral to palliative care. To measure hospitalization-related costs, direct costs were tracked forward after discharge from index hospitalization through the end of the study period using the medical center’s cost accounting system (McKesson Health Solutions, King of Prussia, PA).32,48,51 The dataset included hospitalization utilization and costs incurred at three urban hospitals in the health system following discharge from index hospitalization at the medical center.
This study was approved by the University of Pennsylvania institutional review board (August 21, 2018; authorized by 45 CFR 46.104; Category 4) and followed strict procedures to ensure ethics, security, and patient data privacy.
Participants
Patients were included if they were 18 years or older; admitted to the medical center July 1, 2014 to October 31, 2016 (the study period) for conditions other than childbirth or rehabilitation; received inpatient palliative care consultation to discuss goals-of-care; were not hospitalized at the end of the study period; and did not die during index hospitalization. The health system’s palliative care registry was used to identify which patients received a consult specifically to discuss goals-of-care. Patients were excluded if they received a consult exclusively for reasons other than goals-of-care (e.g., pain management only) or if their self-reported race or ethnicity was not documented.
Measures
Measures to describe the sample included demographic and clinical variables documented in the electronic medical record system during index hospitalization. Demographic variables included age, sex, and Medicaid status. Clinical variables included primary diagnosis, All Patient Refined Diagnosis Group (APR-DRG) risk of mortality (the likelihood of dying) and severity of illness (the extent of physiological decompensation), intensive care unit (ICU) admission, ICU admission greater than six days, receipt of oncology care, admission 30 days prior to index hospitalization, and specialty of the ordering provider who referred the patient to palliative care. Ordering providers included medical doctors, nurse practitioners, and physician assistants.
The APR-DRG variables have four subcategories (minor, moderate, major, extreme) that are systematically assigned at discharge based on procedures and diagnoses coded for billing during hospitalization.32,52 Defined by a patient’s underlying conditions and interaction of comorbidities,32,52 these variables have been found to be better predictors of in-hospital mortality than the Charlson Comorbidity Index.53 The presence and interaction of multiple serious illnesses or conditions characterize high severity of illness and high risk of mortality.54
The independent variable was race and ethnicity, self-reported by patients and documented in the electronic medical records database. This variable included Asian, Black/African American, Latinx, White, and Other (including American Indian/Alaska Native, Native Hawaiian/Pacific Islander, and patients who self-identified as Other race).
Outcomes included experiences during index hospitalization: hospitalization duration, ICU admission, ICU duration, do-not-resuscitate (DNR) documentation, documentation of care goal changes, and enrollment in hospice at discharge from index hospitalization; future hospitalization-related outcomes in the health system following discharge from index hospitalization, defined as 30-day readmissions, number of future days hospitalized, future ICU admission, number of future ICU days; and direct hospitalization costs accrued in the health system after discharge from index hospitalization (“any future costs” and “total future costs”). Direct costs represent the best estimate of the cost of providing hospital services such as nursing and allied health professional labor, supplies, testing, procedures, pharmaceuticals, and emergency department visits that resulted in hospitalization.51 Data showing hospitalization utilization and costs incurred at hospitals unaffiliated with the health system were not available and therefore not included, and have been noted as a limitation.
Data Analysis
Chi-squared tests were used to describe the sample based on patient-reported race and ethnicity (Stata v. 15, Table 1). Data regarding the specialist who referred to palliative care were missing for one patient. This patient was included in the study. Because the parent study had excluded 0.4% of patients in its sample due to missing data (representing <5% of patients), no other data were missing in our analysis.
Table 1.
Description of patients who received palliative care consultation to discuss goals-of-care (PCC) by race and ethnicity (N=1,306)
| PCC Comparison by Race | All PCC Patients N= 1,306 | Asian Patients with PCC N = 48 | Black/ African American Patients with PCC N = 383 | Latinx Patients with PCC N = 22 | White Patients with PCC N= 814 | “Other” Race or Ethnicity Patients with PCC N = 39 | P value | |
|---|---|---|---|---|---|---|---|---|
| Age | 18–34 | 78 (5.97) | 2 (4.17) | 19 (4.96) | 1 (4.55) | 55 (6.76) | 1 (2.56) | 0.77 |
| 35–49 | 179 (13.71) | 6 (12.50) | 60 (15.67) | 4 (18.18) | 104 (12.78) | 5 (12.82) | ||
| 50–64 | 331 (25.34) | 13 (27.08) | 109 (28.46) | 5 (22.73) | 196 (24.08) | 8 (20.51) | ||
| 65–74 | 368 (28.18) | 11 (22.92) | 95 (24.80) | 7 (31.82) | 241 (29.61) | 14 (35.90) | ||
| ≥75 | 350 (26.80) | 16 (33.33) | 100 (26.11) | 5 (22.73) | 218 (26.78) | 11 (28.21) | ||
| Gender | Male | 661 (50.61) | 34 (70.83) | 172 (44.91) | 12 (54.55) | 423 (51.97) | 20 (51.28) | 0.01 |
| Female | 645 (49.39) | 14 (29.17) | 211 (55.09) | 10 (45.45) | 391 (48.03) | 19 (48.72) | ||
| Medicaid Status | Yes | 142 (10.87) | 9 (18.75) | 83 (21.67) | 4 (18.18) | 39 (4.79) | 7 (17.95) | <0.001 |
| No | 1164 (89.13) | 39 (81.25) | 300 (78.33) | 18 (81.82) | 775 (95.21) | 32 (82.05) | ||
| Primary Diagnosis | Cancer | 479 (36.68) | 16 (33.33) | 110 (28.72) | 9 (40.91) | 330 (40.54) | 14 (35.90) | 0.07 |
| Cardiovascular disorder/heart failure | 232 (17.76) | 5 (10.42) | 76 (19.84) | 4 (18.18) | 136 (16.71) | 11 (28.21) | ||
| Endocrine disorder | 28 (2.14) | 1 (2.08) | 8 (2.09) | 0 (0) | 18 (2.21) | 1 (2.56) | ||
| Gastrointestinal, gynecologic, or urologic disorder | 152 (11.64) | 4 (8.33) | 46 (12.01) | 3 (13.64) | 95 (11.67) | 4 (10.26) | ||
| Infectious disease or Sepsis | 175 (13.40) | 9 (18.75) | 63 (16.45) | 2 (9.09) | 95 (11.67) | 6 (15.38) | ||
| Neurologic disorder | 81 (6.20) | 7 (14.58) | 30 (7.83) | 0 (0) | 43 (5.28) | 1 (2.56) | ||
| Respiratory disorder | 89 (6.81) | 3 (6.25) | 23 (6.01) | 3 (13.64) | 59 (7.25) | 1 (2.56) | ||
| Other | 70 (5.36) | 3 (6.25) | 27 (7.05) | 1 (4.55) | 38 (4.67) | 1 (2.56) | ||
| APR-DRG Severity of Illness 1 | Minor | 23 (1.76) | 1 (2.08) | 6 (1.57) | 1 (4.55) | 15 (1.84) | 0 (0) | 0.68 |
| Moderate | 122 (9.34) | 2 (4.17) | 36 (9.40) | 4 (18.18) | 78 (9.58) | 2 (5.13) | ||
| Major | 553 (42.34) | 25 (52.08) | 167 (43.60) | 10 (45.45) | 335 (41.15) | 16 (41.03) | ||
| Extreme | 608 (46.55) | 20 (41.67) | 174 (45.43) | 7 (31.82) | 386 (47.42) | 21 (53.85) | ||
| APR-DRG Risk of Mortality 1 | Minor | 36 (2.76) | 1 (2.08) | 11 (2.87) | 2 (9.09) | 20 (2.46) | 2 (5.13) | 0.08 |
| Moderate | 203 (15.54) | 9 (18.75) | 50 (13.05) | 7 (31.82) | 136 (16.71) | 1 (2.56) | ||
| Major | 570 (43.64) | 21 (43.75) | 179 (46.74) | 9 (40.91) | 341 (41.89) | 20 (51.28) | ||
| Extreme | 497 (38.06) | 17 (35.42) | 143 (37.34) | 4 (18.18) | 317 (38.94) | 16 (41.03) | ||
| Intensive Care Unit (ICU) Admission | Yes | 691 (52.91) | 27 (56.25) | 197 (51.44) | 9 (40.91) | 434 (53.32) | 24 (61.54) | 0.55 |
| No | 615 (47.09) | 21 (43.75) | 186 (48.56) | 13 (59.09) | 380 (46.68) | 15 (38.46) | ||
| Intensive Care Unit (ICU) Admission Greater Than 6 Days | Yes | 366 (28.02) | 16 (33.33) | 104 (27.15) | 6 (27.27) | 226 (27.76) | 14 (35.90) | 0.73 |
| No | 940 (71.98) | 32 (66.67) | 279 (72.85) | 16 (72.73) | 588 (72.24) | 25 (64.10) | ||
| Oncology care during index admission | Yes | 374 (28.64) | 14 (29.17) | 77 (20.10) | 10 (45.45) | 263 (32.31) | 10 (25.64) | <0.001 |
| No | 932 (71.36) | 34 (70.83) | 306 (79.90) | 12 (54.55) | 551 (67.69) | 29 (74.36) | ||
| Admitted 30 days prior to index hospitalization | Yes | 314 (24.04) | 13 (27.08) | 80 (20.89) | 8 (36.36) | 206 (25.31) | 7 (17.95) | 0.23 |
| No | 992 (75.96) | 35 (72.92) | 303 (79.11) | 14 (63.64) | 608 (74.69) | 32 (82.05) | ||
| Specialty of ordering provider who referred patient to palliative care for consult 2 | Cardiology | 139 (10.65) | 5 (10.42) | 47 (12.30) | 5 (22.73) | 78 (9.58) | 4 (10.26) | 0.41 |
| Gastrointestinal (GI) | 20 (1.53) | 0 (0) | 5 (1.31) | 0 (0) | 14 (1.72) | 1 (2.56) | ||
| Hospitalist/General Medicine | 177 (13.56) | 11 (22.92) | 52 (13.61) | 3 (13.64) | 103 (12.65) | 8 (20.51) | ||
| Neurology | 83 (6.36) | 3 (6.25) | 22 (5.76) | 1 (4.55) | 55 (6.76) | 2 (5.13) | ||
| Oncology (including Gyn/Onc) | 514 (39.39) | 23 (47.92) | 159 (41.62) | 6 (27.27) | 312 (38.33) | 14 (35.90) | ||
| Pulmonary | 125 (9.58) | 1 (2.08) | 30 (7.85) | 3 (13.64) | 84 (10.32) | 7 (17.95) | ||
| Surgery | 233 (17.85) | 5 (10.42) | 63 (16.49) | 4 (18.18) | 158 (19.41) | 3 (7.69) | ||
| Other | 14 (1.07) | 0 (0) | 4 (1.05) | 0 (0) | 10 (1.23) | 0 (0) | ||
Categorical variables analyzed using a Chi-squared test. Statistical significance set at 0.05, highlighted in bold font. “Other” race or ethnicity includes patients who self-reported as being American Indian/Alaska Native, Native Hawaiian/Pacific Islander, or “other” race.
APR DRG: All patient refined diagnosis related group.
Ordering providers included medical doctors, nurse practitioners, and physician assistants. One patient was missing data for source of specialist referral.
Outcomes were assessed using descriptive statistics and measures of central tendency. Categorical variables were analyzed using a Chi-squared test. The Kruskal-Wallis test was used for continuous data, which was all nonparametric. To understand what factors contributed to hospice enrollment at discharge and differences in hospitalization outcomes following discharge from a hospitalization with PCC, as identified in bivariate analysis, we conducted multiple linear regression and binomial logistic regression analyses controlling for race and ethnicity, age, sex, Medicaid status, primary diagnosis, risk of mortality, severity of illness, and admission 30-days prior to index hospitalization; and, as appropriate, Hosmer and Lemeshow’s goodness-of-fit tests. Statistical significance was set at the 0.05 level.
RESULTS
Sample Description.
The sample included 1,306 patients who received PCC during the study period (Table 1): 814 White, 383 Black/African American, 48 Asian, 22 Latinx, and 39 Other race patients. Over 80% of the sample was age 50 years or older. Patients who received PCC did not differ by race and ethnicity across demographic or clinical variables, except for sex (P=0.01), Medicaid status (P<0.001), and whether the patient had received oncology care during hospitalization (P<0.001). Males were typically more likely than females to comprise the PCC group, with Asian patients showing the greatest male prevalence (70.8% male). The lowest proportion of PCC patients to have Medicaid was White (4.8%); the highest was Black/African Americans (21.7%), with remaining groups ranging from 18.0–18.8%. Although primary diagnosis did not differ across racial and ethnic groups, Latinx (45.5%) and White (32.3%) patients demonstrated higher proportions of receipt of oncology care during index hospitalization, with Black/African Americans showing the lowest prevalence (20.1%). Black/African American ($11,144 Interquartile Range, IQR) and Latinx ($14,257 IQR) patients accrued the greatest hospitalization costs post-discharge while White patients ($856 IQR) accrued the lowest future costs ($0 median for all races) (Table 2). Do-not-resuscitate documentation was 48.3% for all patients, with higher rates among White (51.2%) and Black/African American (44.9%) patients and lower rates among Latinx (31.8%) and Asian (35.4%) patients. A majority (65.0%) of PCC patients changed their goals-of-care, with similar rates across racial groups.
Table 2.
Comparison of patients who received palliative care consultation to discuss goals-of-care (PCC) by race and ethnicity (N=1,306)
| PCC Comparison by Race | All PCC Patients N= 1,306 | Asian Patients with PCC N = 48 | Black/African American Patients with PCC N = 383 | Latinx Patients with PCC N = 22 | White Patients with PCC N= 814 | “Other” Race or Ethnicity Patients with PCC N = 39 | P value |
|---|---|---|---|---|---|---|---|
| Index hospitalization: Median number of days hospitalized (IQR)** | 10.0 (14.0) | 7.0 (13.0) | 10.0 (14.0) | 9.0 (17.0) | 10.0 (14.0) | 10.0 (16.0) | 0.58 |
| Index hospitalization: Median number of ICU days (IQR)** | 1.0 (7.0) | 1.5 (8.5) | 1.0 (6.0) | 0.0 (7.0) | 1.0 (7.0) | 2.0 (10.0) | 0.67 |
| Index hospitalization: Median direct acute care costs (IQR)** | $19,319 ($29,836) | $14,739 ($24,510) | $ 18,578 ($27,309) | $18,907 ($33,937) | $20,017 ($30,991) | $20,540 ($38,849) | 0.71 |
| Future: Any future acute care costs post-discharge from index hospitalization (% yes)* | 29.0% | 27.1% | 34.7% | 36.4% | 26.2% | 30.8% | 0.004 |
| Future: Median total direct acute care costs post-discharge (IQR)** | $0.00 ($3,432) | $0.00 ($1,560) | $0.00 ($11,144) | $0.00 ($14,257) | $0.00 ($856) | $0.00 ($13,980) | 0.07 |
| Future: Readmitted within 30 days (% yes)* | 13.9% | 14.6% | 18.3% | 18.2% | 11.6% | 15.4% | 0.04 |
| Future: Median total number of hospital days following index hospitalization (IQR)** | 0.0 (2.0) | 0.0 (2.0) | 0.0 (6.0) | 0.0 (9.0) | 0.0 (1.0) | 0.0 (7.0) | 0.004 |
| Future: Any ICU days post-discharge (% yes)* | 14.4% | 14.6% | 18.0% | 13.6% | 12.7% | 15.4% | 0.19 |
| Future: Median number of total ICU days following index hospitalization (IQR)** | 0.0 (0) | 0.0 (0) | 0.0 (0) | 0.0 (0) | 0.0 (0) | 0.0 (0) | 0.17 |
| Discharged to hospice from index hospitalization (% yes) * | 38.67% | 33.3% | 35.5% | 31.8% | 41.2% | 28.2% | 0.16 |
| DNR documented during index hospitalization (% yes) * | 48.3% | 35.4% | 44.9% | 31.8% | 51.2% | 46.2% | 0.04 |
| Goals-of-care changed during palliative care consult (% yes) * | 65.0% | 58.3% | 60.1% | 59.1% | 67.7% | 69.2% | 0.08 |
IQR: Inter quartile range. Statistical significance set at 0.05, highlighted in bold font. “Other” race or ethnicity includes patients who self-reported as being American Indian/Alaska Native, Native Hawaiian/Pacific Islander, or “other” race.
Categorical variables analyzed using a Chi-square test.
The Kruskal-Wallis test was used for continuous data, which was all nonparametric in distribution.
In bivariate analysis of hospitalization and cost outcomes during index hospitalization (Table 2), racial and ethnic groups differed in the following variables: whether a patient had any future acute care costs (representing any future hospitalization, P=0.004); the median number of hospital days following discharge from index hospitalization (P=0.004); 30-day readmission rates (P=0.04), and DNR status (P=0.04). Although median total future costs were equivalent ($0) across racial groups in bivariate analysis, there was meaningful racial variation of IQR, warranting further investigation.
Factors contributing to hospice enrollment.
In adjusted analyses, Medicaid patients were 42% less likely than non-Medicaid patients to be discharged to hospice (adjusted odds ratio, AOR=0.59, 95% CI 0.37–0.90, P=0.02) (Table 3).
Table 3.
Results of binomial logistic regression to identify variables independently associated with a patient who received palliative care consultation to discuss goals-of-care (PCC) being discharged to hospice1 (N=1,306)
| Variable | Odds Ratio (OR) | Standard Error | 95% Confidence Interval (CI) | P value |
|---|---|---|---|---|
| Age group (P=0.49)2 | ||||
| Base, 18–34 years | ||||
| 35–49 years | 0.94 | 0.57 | 0.53–1.67 | 0.83 |
| 50–64 years | 1.11 | 0.30 | 0.53–1.67 | 0.70 |
| 65–74 years | 1.13 | 0.31 | 0.66–1.92 | 0.67 |
| 75 years+ | 1.14 | 0.31 | 0.66–2.00 | 0.64 |
| Gender (P=0.11)2 | ||||
| Female | 1.22 | 0.31 | 0.97–1.53 | 0.10 |
| Race and ethnicity (P=0.46)2 | ||||
| Base, White | ||||
| Asian | 0.82 | 0.26 | 0.44–1.54 | 0.54 |
| Black/African American | 0.87 | 0.12 | 0.67–1.13 | 0.29 |
| Latinx | 0.79 | 0.37 | 0.31–2.00 | 0.62 |
| Other | 0.57 | 0.21 | 0.28–1.17 | 0.13 |
| Medicaid status (P= 0.010)2 | ||||
| Yes | 0.59 | 0.13 | 0.38–0.90 | 0.02 |
| Primary Diagnosis | 0.95 | 0.02 | 0.90–1.00 | 0.052 |
| APR-DRG Severity of Illness (SOI) (P= 0.61)2 | ||||
| Base, Minor | ||||
| Moderate | 0.60 | 0.30 | 0.23–1.58 | 0.31 |
| Major | 0.69 | 0.34 | 0.26–1.81 | 0.45 |
| Extreme | 0.76 | 0.39 | 0.28–2.07 | 0.60 |
| APR-DRG Risk of Mortality (ROM) (P= 0.08)2 | ||||
| Base, Minor | ||||
| Moderate | 1.28 | 0.55 | 0.55–2.98 | 0.60 |
| Major | 1.40 | 0.62 | 0.59–3.35 | 0.44 |
| Extreme | 1.77 | 0.83 | 0.71–4.42 | 0.23 |
| Admitted to hospital 30-days prior to index hospitalization3 (P= 0.82)2 | ||||
| Yes | 1.03 | 0.14 | 0.71–4.42 | 0.84 |
Test statistics for this binomial logistic regression model included: P = 0.01, R2 = 0.02, and the following Hosmer and Lemeshow’s goodness-of-fit test result: χ2 =7.20, (P=0.52).
When overall variables (not broken out by response categories) were analyzed in a binomial logistic regression model, test statistics included: P=0.0004, R2 = 0.02 and a Hosmer and Lemeshow’s goodness-of-fit test result: χ2 = 15.87, (P=0.04). In this binomial logistic regression model without variable subcategories broken out, primary diagnosis had a P value of 0.06.
Index hospitalization is the hospitalization in which a patient first had palliative care consultation to discuss goals-of-care (PCC) in the study period.
- Statistical significance set at 0.05, highlighted in bold font.
Factors contributing to 30-day readmission.
In adjusted analyses, patients age 75+ were 57% less likely (AOR=0.43, 95% CI 0.21–0.85, P=0.02), Medicaid patients were twice as likely (AOR=2.02, 95% CI 1.25–3.26, P=0.004), patients admitted 30 days prior were 2.4 times as likely (AOR=2.42, 95% CI 1.72–3.41, P<0.0001), and Black/African Americans were 1.6 times as likely (AOR=1.57, 95% CI 1.09–2.26, P=0.02) to be readmitted within 30 days (Table 4).
Table 4.
Results of binomial logistic regression to identify variables independently associated with a patient who received palliative care consultation to discuss goals-of-care (PCC) being readmitted within 30 days of discharge1 (N=1,306)
| Variable | Odds Ratio (OR) | Standard Error | 95% Confidence Interval (CI) | P value |
|---|---|---|---|---|
| Age group (P= 0.01)2 | ||||
| Base, 18–34 years | ||||
| 35–49 years | 0.82 | 0.28 | 0.42–1.50 | 0.55 |
| 50–64 years | 0.65 | 0.21 | 0.35–1.22 | 0.18 |
| 65–74 years | 0.59 | 0.20 | 0.31–1.14 | 0.12 |
| 75 years+ | 0.43 | 0.15 | 0.21–0.85 | 0.02 |
| Gender (P=0.59)2 | ||||
| Female | 0.89 | 0.15 | 0.64–1.24 | 0.49 |
| Race and ethnicity (P=0.06)2 | ||||
| Base, White | ||||
| Asian | 1.16 | 0.51 | 0.49–2.76 | 0.73 |
| Black/African American | 1.57 | 0.29 | 1.09–2.26 | 0.02 |
| Latinx | 1.35 | 0.79 | 0.43–4.27 | 0.61 |
| Other | 1.40 | 0.67 | 0.55–3.56 | 0.47 |
| Medicaid status (P= 0.002)2 | ||||
| Yes | 2.02 | 0.49 | 1.25–3.26 | 0.004 |
| Primary Diagnosis | 1.02 | 0.04 | 0.95–1.09 | 0.622 |
| APR-DRG Severity of Illness (SOI) (P= 0.83)2 | ||||
| Base, Minor | ||||
| Moderate | 2.21 | 1.78 | 0.45–10.82 | 0.33 |
| Major | 2.05 | 1.67 | 0.42–10.14 | 0.38 |
| Extreme | 1.78 | 1.49 | 0.34–9.22 | 0.49 |
| APR-DRG Risk of Mortality (ROM) (P= 0.96)2 | ||||
| Base, Minor | ||||
| Moderate | 0.78 | 0.39 | 0.29–2.10 | 0.62 |
| Major | 0.75 | 0.39 | 0.27–2.07 | 0.57 |
| Extreme | 0.85 | 0.48 | 0.28–2.57 | 0.78 |
| Admitted to hospital 30-days prior to index hospitalization3 (P<0.0001)2 | ||||
| Yes | 2.42 | 0.42 | 1.72–3.41 | <0.0001 |
Test statistics for this binomial logistic regression model included: P <0.0001, R2 = 0.06, and the following Hosmer and Lemeshow’s goodness-of-fit test result: χ2 =9.18, (P=0.33).
When overall variables (not broken out by response categories) were analyzed in a binomial logistic regression model, test statistics included: P<0.0001, R2 = 0.06 and a Hosmer and Lemeshow’s goodness-of-fit test result: χ2 =10.23, (P=0.25). In this binomial logistic regression model without variable subcategories broken out, primary diagnosis had a P value of 0.61.
Index hospitalization is the hospitalization in which a patient first had palliative care consultation to discuss goals-of-care (PCC) in the study period.
- Statistical significance set at 0.05, highlighted in bold font.
Factors contributing to total future days hospitalized.
In adjusted analyses, Medicaid patients (Coefficient=4.49, 95% CI 1.92–7.06, P=0.001), followed by Black/African Americans (Coefficient=2.16, 95% CI 0.47–3.85, P=0.01), and patients admitted 30 days prior (Coefficient=2.08, 95% CI 0.37–3.80, P=0.02) were independent predictors of the median number of days a patient was hospitalized post-discharge (Table 5).
Table 5.
Results of multiple linear regression to identify variables independently associated with number of future days hospitalized following discharge from index hospitalization with palliative care consultation to discuss goals-of-care (PCC)1 (N=1,306)
| Variable | Coefficient | Standard Error | 95% Confidence Interval (CI) | P value |
|---|---|---|---|---|
| Age group (P= 0.002)2 | ||||
| Base, 18–34 years | ||||
| 35–49 years | 3.23 | 1.82 | −0.35–6.81 | 0.08 |
| 50–64 years | −0.09 | 1.69 | −3.41–3.23 | 0.96 |
| 65–74 years | −0.55 | 1.72 | −3.93–2.83 | 0.75 |
| 75 years+ | −1.68 | 1.74 | −5.09–1.73 | 0.33 |
| Gender (P=0.24)2 | ||||
| Female | −1.02 | 0.75 | −2.49 to −0.44 | 0.17 |
| Race and ethnicity (P=0.15)2 | ||||
| Base, White | ||||
| Asian | −1.42 | 2.00 | −5.33–2.50 | 0.48 |
| Black/African American | 2.16 | 0.86 | 0.47–3.85 | 0.01 |
| Latinx | −0.57 | 2.90 | −6.24–5.10 | 0.84 |
| Other | 2.71 | 2.20 | −1.60–7.03 | 0.22 |
| Medicaid status (P<0.001)2 | ||||
| Yes | 4.49 | 1.31 | 1.92–7.06 | 0.001 |
| Primary Diagnosis | −0.04 | 0.16 | −0.36–0.28 | 0.792 |
| APR-DRG Severity of Illness (SOI) (P=0.57)2 | ||||
| Base, Minor | ||||
| Moderate | −0.29 | 3.11 | −6.40–5.82 | 0.93 |
| Major | 0.25 | 3.11 | −5.86–6.36 | 0.94 |
| Extreme | 0.90 | 3.23 | −5.44–7.24 | 0.78 |
| APR-DRG Risk of Mortality (ROM) (P=0.61)2 | ||||
| Base, Minor | ||||
| Moderate | 0.51 | 2.55 | −4.49–5.51 | 0.84 |
| Major | 0.51 | 2.63 | −4.95–5.37 | 0.94 |
| Extreme | 0.21 | 2.79 | −4.61–6.32 | 0.76 |
| Admitted to hospital 30-days prior to index hospitalization3 (P= 0.03)2 | ||||
| Yes | 2.08 | 2.08 | 0.37–3.80 | 0.02 |
Test statistics for this multiple linear regression model included: P<0.0001, R2 = 0.05.
When overall variables (not broken out by response categories) were analyzed in a multiple linear regression model, test statistics included: P<0.0001, R2 = 0.04. In this multiple linear regression model in which variable subgroups were not broken out, primary diagnosis had a P value of 0.81.
Index hospitalization is the hospitalization in which a patient first had palliative care consultation to discuss goals-of-care (PCC) in the study period.
- Statistical significance set at 0.05, highlighted in bold font.
Factors contributing to having any future hospitalization cost.
In adjusted analyses, patients ages 65–74 (AOR=0.54, 95% CI 0.31–0.91, P=0.02) and 75+ (AOR=0.53, 95% CI 0.31–0.92, P=0.02) were about 47% less likely whereas Black/African Americans were 1.4 times as likely (AOR=1.39, 95% CI 1.05–1.84, P=0.02), Medicaid patients were 1.7 times as likely (AOR=1.66, 95% CI 1.11–2.48, P=0.01), and patients admitted 30 days prior were 1.8 times as likely (AOR=1.81, 95% CI 1.37–2.39, P<0.0001) to have any future hospitalization cost (i.e., any future hospitalization) (Table 6).
Table 6.
Results of binomial logistic regression to identify variables independently associated with a patient who received palliative care consultation to discuss goals-of-care (PCC) having any future hospitalization cost (i.e., any hospitalization) following discharge1 (N=1,306)
| Variable | Odds Ratio (OR) | Standard Error | 95% Confidence Interval (CI) | P value |
|---|---|---|---|---|
| Age group (P= 0.01)2 | ||||
| Base, 18–34 years | ||||
| 35–49 years | 0.74 | 0.21 | 0.42–1.29 | 0.29 |
| 50–64 years | 0.67 | 0.18 | 0.40–1.13 | 0.13 |
| 65–74 years | 0.54 | 0.15 | 0.31–0.91 | 0.02 |
| 75 years+ | 0.53 | 0.15 | 0.31–0.92 | 0.02 |
| Gender (P=0.62)2 | ||||
| Female | 1.04 | 0.13 | 0.81–1.33 | 0.76 |
| Race and ethnicity (P= 0.10)2 | ||||
| Base, White | ||||
| Asian | 0.97 | 0.33 | 0.49–1.90 | 0.93 |
| Black/African American | 1.39 | 0.20 | 1.05–1.84 | 0.02 |
| Latinx | 1.56 | 0.72 | 0.63–3.85 | 0.34 |
| Other | 1.28 | 0.47 | 0.63–2.62 | 0.49 |
| Medicaid status (P= 0.004)2 | ||||
| Yes | 1.66 | 0.34 | 1.11–2.48 | 0.01 |
| Primary Diagnosis | 1.04 | 0.03 | 0.98–1.09 | 0.182 |
| APR-DRG Severity of Illness (SOI) (P= 0.71)2 | ||||
| Base, Minor | ||||
| Moderate | 0.89 | 0.48 | 0.31–2.58 | 0.83 |
| Major | 0.90 | 0.49 | 0.31–2.60 | 0.85 |
| Extreme | 0.84 | 0.47 | 0.28–2.53 | 0.76 |
| APR-DRG Risk of Mortality (ROM) (P= 0.06)2 | ||||
| Base, Minor | ||||
| Moderate | 1.23 | 0.55 | 0.51–2.95 | 0.65 |
| Major | 1.41 | 0.65 | 0.57–3.46 | 0.46 |
| Extreme | 1.86 | 0.90 | 0.72–4.81 | 0.20 |
| Admitted to hospital 30-days prior to index hospitalization3 (P<0.0001)2 | ||||
| Yes | 1.81 | 0.90 | 1.37–2.39 | 0.0001 |
Test statistics for this binomial logistic regression model included: P < 0.0001, R2 = 0.03, and the following Hosmer and Lemeshow’s goodness-of-fit test result: χ2 =13.47, (P=0.10).
When overall variables (not broken out by response categories) were analyzed in a binomial logistic regression model, test statistics included: P < 0.0001, R2 = 0.03 and a Hosmer and Lemeshow’s goodness-of-fit test result: χ2 =7.19, (P=0.52). In this binomial logistic regression model without variable subcategories broken out, primary diagnosis had a P value of 0.18.
Index hospitalization is the hospitalization in which a patient first had palliative care consultation to discuss goals-of-care (PCC) in the study period.
- Statistical significance set at 0.05, highlighted in bold font.
Factors contributing to total future hospitalization costs.
In adjusted analyses, hospitalization costs post-discharge were lower for females (Coefficient=−3616.64, 95% CI −6773.15 to −460.12, P=0.03) but greater for Medicaid patients (Coefficient=7388.43, 95% CI 1850.80–12926.07, P=0.01), patients admitted 30 days prior (Coefficient=3868.06, 95% CI 178.09–7558.03, P=0.04), and Black/African Americans (Coefficient=3856.89, 95% CI 225.51–7488.26, P=0.04) (Table 7).
Table 7.
Results of multiple linear regression to identify variables independently associated with total future hospitalization costs following discharge from a hospitalization with palliative care consultation to discuss goals-of-care (PCC)1 (N=1,306)
| Variable | Coefficient | Standard Error | 95% Confidence Interval (CI) | P value |
|---|---|---|---|---|
| Age group (P<0.0001)2 | ||||
| Base, 18–34 years | ||||
| 35–49 years | 6264.64 | 3926.86 | −1439.12–13968.39 | 0.11 |
| 50–64 years | −1451.46 | 3642.86 | −8598.06–5695.14 | 0.69 |
| 65–74 years | −2997.35 | 3705.48 | −8598.06–5695.14 | 0.42 |
| 75 years+ | −5614.69 | 3742.80 | −12957.35–1727.97 | 0.13 |
| Gender (P=0.034)2 | ||||
| Female | −3616.64 | 1608.98 | −6773.15 to −460.12 | 0.03 |
| Race and ethnicity (P= 0.255)2 | ||||
| Base, White | ||||
| Asian | −2678.86 | 4298.29 | −11111.28–5753.57 | 0.53 |
| Black/African American | 3856.89 | 1851.03 | 225.51–7488.26 | 0.04 |
| Latinx | −1848.06 | 1851.03 | −14052.96–10356.85 | 0.77 |
| Other | 5479.45 | 4733.20 | −3806.18–14765.07 | 0.25 |
| Medicaid status (P= 0.01)2 | ||||
| Yes | 7388.43 | 2822.72 | 1850.80–12926.07 | 0.01 |
| Primary Diagnosis | −297.35 | 352.52 | −988.93–394.23 | 0.402 |
| APR-DRG Severity of Illness (SOI) (P= 0.31)2 | ||||
| Base, Minor | ||||
| Moderate | −1944.85 | 6706.23 | −15101.19–11211.50 | 0.77 |
| Major | −1224.92 | 6705.63 | −14380.09–11930.24 | 0.86 |
| Extreme | 1965.31 | 6958.19 | −11685.33–15615.95 | 0.78 |
| APR-DRG Risk of Mortality (ROM) (P= 0.58)2 | ||||
| Base, Minor | ||||
| Moderate | −311.28 | 5489.41 | −11080.46–10457.90 | 0.96 |
| Major | 46.57 | 5665.88 | −11068.79–11161.94 | 0.99 |
| Extreme | 872.44 | 6000.64 | −10899.66–12644.54 | 0.88 |
| Admitted to hospital 30-days prior to index hospitalization3 (P= 0.06)2 | ||||
| Yes | 3868.06 | 1880.90 | 178.09–7558.03 | 0.04 |
Test statistics for this binomial logistic regression model included: P <0.0001, R2 = 0.05.
When overall variables (not broken out by response categories) were analyzed in a binomial logistic regression model, test statistics included: P < 0.0001, R2 = 0.04. In this multiple linear regression model in which variables are not broken out by subcategory, primary diagnosis had a P value of 0.42.
Index hospitalization is the hospitalization in which a patient first had palliative care consultation to discuss goals-of-care (PCC) in the study period.
- Statistical significance set at 0.05, highlighted in bold font.
DISCUSSION
In this study of 1,306 PCC patients, Black/African American race and two variables related to socioeconomic status, Medicaid use and prior hospital admission,55–57 were independent risk factors for greater future hospitalization utilization and cost outcomes following an admission with PCC. Compared to White patients, Black/African American patients were more likely to accrue any future cost (i.e., any future hospitalization) and be readmitted within 30 days, experienced more future days hospitalized, and accrued greater future hospitalization costs. Qualitative research is needed to understand why, when other factors are held constant, Black/African American PCC patients are more likely than White PCC patients to return to the hospital for additional care. Research is also needed to understand why racial and ethnic minorities were more likely than White patients to be hospitalized following PCC.
In our sample, Medicaid use was dramatically higher among racial and ethnic minorities (P<0.001), ranging from 18.0% to 21.7% among non-White patients compared to 4.8% White patients, and independently predicted multiple outcomes. Compared to non-Medicaid PCC patients, Medicaid patients were less likely to be discharged to hospice, more likely to accrue any future cost (i.e., future hospitalization), more likely to be readmitted within 30 days, experienced more future days hospitalized, and accrued greater future hospitalization costs. Patients admitted 30 days before PCC hospitalization were also more likely to accrue any future hospitalization cost and be readmitted within 30 days, experienced more future days hospitalized, and accrued greater total future costs. When considering PCC’s effectiveness in helping patients align future care with goals, clinicians must consider how these and other socioeconomic factors introduce barriers to achieving goals post-discharge.58–61 Low R2 values in our regression models also indicate variables not available in our study (e.g., health literacy, English language proficiency, access to high quality community-based care)58,62,63 may explain differences in post-discharge outcomes. Research is needed to understand how socioeconomic factors contribute to readmissions among PCC patients and how gaps in care following discharge or lower quality care in the community contribute.58,61
Despite well-documented racial disparities in hospice and end-of-life acute care use in the United States,2,21–23,28,29 race and ethnicity did not predict hospice enrollment among PCC patients in our study. Medicaid status was the only variable that explained discharge to hospice. Our results add to growing evidence that seriously-ill patients across racial and ethnic groups demonstrate similar preferences for comfort-oriented care when clinicians and patients engage in clear communication about prognosis, goals, and values.32,64,65 This finding is consistent with results from a recent propensity score-matched study of seriously-ill Black/African American and White patients from the same dataset, which found PCC was associated with 14- to 15-fold increases in hospice enrollment for both Black/African American and White patients compared to clinically similar Black/African American and White patients without PCC.32 Previous studies suggest patient-centered PCC communication benefits patients across racial and ethnic groups,66–68 but qualitative research is needed to understand why PCC is associated with racially similar hospice enrollment and if PCC mitigates disparities in hospice care.
Consistent with recommended standards for publishing on health inequities,44 we defined race and ethnicity as social-cultural factors, not biological factors, and acknowledged how institutional and interpersonal racism experienced by ethnic minority patients may explain differences in outcomes.44 There are policy implications for both clinical practice and institutional structures. Given the role of social determinants of health in predicting hospice and future hospitalization outcomes, interventions that address social factors and incorporate social supports are needed.60 One study found, for example, higher social support reduced the odds of minority patient readmission or death by 65.0%.61 To improve gaps in care and reduce readmissions post-discharge, primary palliative care services should be expanded58,59,69–71 and culturally-sensitive PCC communication should be adopted across settings,72–75 particularly in less affluent communities and racial minority communities less likely to have access to primary palliative care.59,76 Systems should educate community-based clinicians in PCC communication and form community resident or patient/family boards to ensure PCC interventions reflect the values and priorities of ethnic minority patients.44 Studies have shown, for example, the importance of incorporating the cultural value of familismo into palliative care for Latinx patients, which may include communicating with and engaging intergenerational family members in the overall care of patients or in decision-making.77,78
Limitations
As a secondary analysis, this study is limited by variables available in the dataset, particularly cultural and social determinants of health such as primary language, poverty level, health literacy, religiosity, and experiences with racial bias in healthcare settings.8–10,34 A second limitation is the small sample size of Latinx patients. Other palliative care studies have featured larger samples of Latinx patients using national databases and multi-site designs.31,79–81 Although these studies did not analyze future utilization or costs following PCC, their inclusion of more Latinx patients better represents the United States population and enables more robust analysis of Latinx outcomes. Future research in palliative care should include larger samples of Latinx, Asian, and other racial and ethnic minority patients underrepresented in the literature.
A third limitation was lack of access to outpatient follow-up information, which may have explained higher readmissions among non-White patients. Our study only examined hospitalization utilization and costs incurred within the health system and did not include emergency department visits that did not result in hospitalization. Because the health system is the largest acute care provider in the Northeast, it is unlikely many patients were hospitalized outside the system. Next, our study did not measure hospice use among patients who enrolled in hospice after discharge. More patients may have enrolled in hospice following PCC. Finally, our study was conducted in a racially diverse setting at a single academic medical center with significant resources and an experienced palliative care team.
CONCLUSIONS
In this study of seriously-ill patients who received palliative care consultation at a large, urban medical center, Medicaid use and hospitalization in the previous 30 days were strong and consistent risk factors for future hospitalization utilization and costs. Medicaid status was the only predictor of hospice enrollment. While race and ethnicity were insignificant for hospice discharge, Black/African American race remained an independent predictor of having any future hospitalization cost (any future hospitalization), 30-day readmission, number of future days hospitalized, and total future hospitalization costs. Despite PCC, racial groups differed in having any future hospitalization cost, 30-day readmission, number of future days hospitalized and DNR documentation, with higher readmission rates and greater costs post-discharge among Black/African American and Latinx patients and lower DNR documentation among racial and ethnic minorities. Females accrued lower future hospitalization costs than males.
Our study underscores the role of social determinants of health in predicting future hospitalization care and costs among seriously-ill patients and disproportionate social risks for Black/African American patients in explaining end-of-life outcomes. Our findings also add to a growing body of evidence that shows socioeconomic factors help explain end-of-life decisions and acute care use.32,64,65 Although palliative care consultations and care goal discussions may enable patients across racial and ethnic groups to make end-of-life decisions, interventions that address social factors are needed to reduce end-of-life inequalities. Clinicians must improve access to community-based palliative care and implement social support interventions that mediate social determinants of end-of-life care.
Funding:
LT Starr is supported by the Ruth L. Kirschstein National Research Service Award training program in Individualized Care for At Risk Older Adults at the University of Pennsylvania, National Institute of Nursing Research of the National Institutes of Health (T32NR009356). CM Ulrich is supported in part by a National Cancer Institute/National Institutes of Health Award (R01CA196131). A Perez is supported in part by the National Institute on Aging/National Institutes of Health (R01AG070351) and P30 (5P30AG059302). SH Meghani is supported in part by a National Institute of Nursing Research/National Institutes of Health Award (R01NR017853).
Footnotes
Conflicts of interest: No conflicts to report.
References
- 1.Institute of Medicine. Dying in America: Improving quality and honoring individual preferences near the end of life. Washington, DC: 2015. [DOI] [PubMed] [Google Scholar]
- 2.Clark MA, Person SD, Gosline A, Gawande AA, Block SD. Racial and Ethnic Differences in Advance Care Planning: Results of a Statewide Population-Based Survey. J Palliat Med. 2018;21(8):1078–1085. [DOI] [PubMed] [Google Scholar]
- 3.McAfee CA, Jordan TR, Sheu JJ, Dake JA, Kopp Miller BA. Predicting Racial and Ethnic Disparities in Advance Care Planning Using the Integrated Behavioral Model. Omega (Westport). 2017:30222817691286. [DOI] [PubMed] [Google Scholar]
- 4.Sanders JJ, Robinson MT, Block SD. Factors Impacting Advance Care Planning among African Americans: Results of a Systematic Integrated Review. J Palliat Med. 2016;19(2):202–227. [DOI] [PubMed] [Google Scholar]
- 5.Smith AK, McCarthy EP, Paulk E, et al. Racial and ethnic differences in advance care planning among patients with cancer: impact of terminal illness acknowledgment, religiousness, and treatment preferences. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2008;26(25):4131–4137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Houben CHM, Spruit MA, Groenen MTJ, Wouters EFM, Janssen DJA. Efficacy of advance care planning: a systematic review and meta-analysis. J Am Med Dir Assoc. 2014;15(7):477–489. [DOI] [PubMed] [Google Scholar]
- 7.Bernacki RE, Block SD, American College of Physicians High Value Care Task F. Communication about serious illness care goals: a review and synthesis of best practices. JAMA Intern Med. 2014;174(12):1994–2003. [DOI] [PubMed] [Google Scholar]
- 8.Hausmann LR, Hannon MJ, Kresevic DM, Hanusa BH, Kwoh CK, Ibrahim SA. Impact of perceived discrimination in healthcare on patient-provider communication. Med Care. 2011;49(7):626–633. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cuevas AG, O’Brien K, Saha S. African American experiences in healthcare: “I always feel like I’m getting skipped over”. Health Psychol. 2016;35(9):987–995. [DOI] [PubMed] [Google Scholar]
- 10.Benkert R, Peters RM, Clark R, Keves-Foster K. Effects of perceived racism, cultural mistrust and trust in providers on satisfaction with care. J Natl Med Assoc. 2006;98(9):1532–1540. [PMC free article] [PubMed] [Google Scholar]
- 11.Starr L, NR OC, Meghani S. Improved Serious Illness Communication May Help Mitigate Racial Disparities in Care Among Black Americans with COVID-19. Journal of General Internal Medicine. 2021;accepted, in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Boyd RW, Lindo EG, Weeks LD, McLemore MR. On Racism: A New Standard For Publishing On Racial Health Inequities. In. Health Affairs Blog. Vol 2021: Health Affairs; 2020. [Google Scholar]
- 13.Bailey ZD, Krieger N, Agenor M, Graves J, Linos N, Bassett MT. Structural racism and health inequities in the USA: evidence and interventions. Lancet. 2017;389(10077):1453–1463. [DOI] [PubMed] [Google Scholar]
- 14.LoPresti MA, Dement F, Gold HT. End-of-Life Care for People With Cancer From Ethnic Minority Groups: A Systematic Review. Am J Hosp Palliat Care. 2016;33(3):291–305. [DOI] [PubMed] [Google Scholar]
- 15.Egede LE. Race, ethnicity, culture, and disparities in health care. J Gen Intern Med. 2006;21(6):667–669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Williams DR. Race/ethnicity and socioeconomic status: measurement and methodological issues. Int J Health Serv. 1996;26(3):483–505. [DOI] [PubMed] [Google Scholar]
- 17.Cooper LA, Hill MN, Powe NR. Designing and evaluating interventions to eliminate racial and ethnic disparities in health care. J Gen Intern Med. 2002;17(6):477–486. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Melhado L, Bushy A. Exploring uncertainty in advance care planning in African Americans: does low health literacy influence decision making preference at end of life. Am J Hosp Palliat Care. 2011;28(7):495–500. [DOI] [PubMed] [Google Scholar]
- 19.Wang SY, Hsu SH, Aldridge MD, Cherlin E, Bradley E. Racial Differences in Health Care Transitions and Hospice Use at the End of Life. J Palliat Med. 2019;22(6):619–627. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Chen Y, Criss SD, Watson TR, et al. Cost and Utilization of Lung Cancer End-of-Life Care Among Racial-Ethnic Minority Groups in the United States. Oncologist. 2020;25(1):e120–e129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Yang A, Goldin D, Nova J, Malhotra J, Cantor JC, Tsui J. Racial Disparities in Health Care Utilization at the End of Life Among New Jersey Medicaid Beneficiaries With Advanced Cancer. JCO Oncol Pract. 2020;16(6):e538–e548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Brown CE, Engelberg RA, Sharma R, et al. Race/Ethnicity, Socioeconomic Status, and Healthcare Intensity at the End of Life. J Palliat Med. 2018;21(9):1308–1316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Perry LM, Walsh LE, Horswell R, et al. Racial Disparities in End-of-Life Care Between Black and White Adults With Metastatic Cancer. J Pain Symptom Manage. 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kirkendall A, Holland JM, Keene JR, Luna N. Evaluation of hospice care by family members of Hispanic and non-Hispanic patients. Am J Hosp Palliat Care. 2015;32(3):313–321. [DOI] [PubMed] [Google Scholar]
- 25.Rhodes RL, Xuan L, Halm EA. African American bereaved family members’ perceptions of hospice quality: do hospices with high proportions of African Americans do better? J Palliat Med. 2012;15(10):1137–1141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Welch LC, Teno JM, Mor V. End-of-life care in black and white: race matters for medical care of dying patients and their families. J Am Geriatr Soc. 2005;53(7):1145–1153. [DOI] [PubMed] [Google Scholar]
- 27.Rhodes RL, Teno JM, Connor SR. African American bereaved family members’ perceptions of the quality of hospice care: lessened disparities, but opportunities to improve remain. J Pain Symptom Manage. 2007;34(5):472–479. [DOI] [PubMed] [Google Scholar]
- 28.Ornstein KA, Roth DL, Huang J, et al. Evaluation of Racial Disparities in Hospice Use and End-of-Life Treatment Intensity in the REGARDS Cohort. JAMA Netw Open. 2020;3(8):e2014639. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.NHPCO. NHPCO Facts and Figures 2020 Edition. National Hospice and Palliative Care Organization; August 20, 2020. 2020. [Google Scholar]
- 30.Rising ML, Hassouneh DS, Lutz KF, Berry P, Lee CS. Hispanic Hospice Utilization: Integrative Review and Meta-analysis. J Health Care Poor Underserved. 2019;30(2):468–494. [DOI] [PubMed] [Google Scholar]
- 31.Johnson T, Walton S Jr., Levine S, Fister E, Baron A, O’Mahony S. Racial and ethnic disparity in palliative care and hospice use. Am J Manag Care. 2020;26(2):e36–e40. [DOI] [PubMed] [Google Scholar]
- 32.Starr LT, Ulrich CM, Junker P, Appel SM, O’Connor NR, Meghani SH. Goals-of-Care Consultation Associated With Increased Hospice Enrollment Among Propensity-Matched Cohorts of Seriously Ill African American and White Patients. J Pain Symptom Manage. 2020;60(4):801–810. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Levoy K, Tarbi EC, De Santis JP. End-of-life decision making in the context of chronic life-limiting disease: a concept analysis and conceptual model. Nurs Outlook. 2020;68(6):784–807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Frost DW, Cook DJ, Heyland DK, Fowler RA. Patient and healthcare professional factors influencing end-of-life decision-making during critical illness: a systematic review. Crit Care Med. 2011;39(5):1174–1189. [DOI] [PubMed] [Google Scholar]
- 35.Volandes AE, Paasche-Orlow M, Gillick MR, et al. Health literacy not race predicts end-of-life care preferences. J Palliat Med. 2008;11(5):754–762. [DOI] [PubMed] [Google Scholar]
- 36.Adler NE, Rehkopf DH. U.S. disparities in health: descriptions, causes, and mechanisms. Annu Rev Public Health. 2008;29:235–252. [DOI] [PubMed] [Google Scholar]
- 37.Idossa D, Duma N, Chekhovskiy K, Go R, Ailawadhi S. Commentary: Race and Ethnicity in Biomedical Research - Classifications, Challenges, and Future Directions. Ethn Dis. 2018;28(4):561–564. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Ford ME, Kelly PA. Conceptualizing and categorizing race and ethnicity in health services research. Health Serv Res. 2005;40(5 Pt 2):1658–1675. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Orlovic M, Smith K, Mossialos E. Racial and ethnic differences in end-of-life care in the United States: Evidence from the Health and Retirement Study (HRS). SSM Popul Health. 2019;7:100331. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Schwartz RS. Racial profiling in medical research. The New England journal of medicine. 2001;344(18):1392–1393. [DOI] [PubMed] [Google Scholar]
- 41.Hahn RA. The state of federal health statistics on racial and ethnic groups. JAMA. 1992;267(2):268–271. [PubMed] [Google Scholar]
- 42.Shanawani H, Dame L, Schwartz DA, Cook-Deegan R. Non-reporting and inconsistent reporting of race and ethnicity in articles that claim associations among genotype, outcome, and race or ethnicity. J Med Ethics. 2006;32(12):724–728. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Rivara FP, Finberg L. Use of the Terms Race and Ethnicity. Archives of Pediatrics & Adolescent Medicine. 2001;155(2):119–119. [DOI] [PubMed] [Google Scholar]
- 44.Boyd R, Lindo E, Weeks L, McLemore M. On racism: a new standard for publishing on racial health inequities. In. Vol 2021. 10 ed: Health Affairs; 2020. [Google Scholar]
- 45.Borrell LN, Elhawary JR, Fuentes-Afflick E, et al. Race and Genetic Ancestry in Medicine - A Time for Reckoning with Racism. The New England journal of medicine. 2021;384(5):474–480. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Flanagin A, Frey T, Christiansen SL, Bauchner H. The Reporting of Race and Ethnicity in Medical and Science Journals: Comments Invited. JAMA. 2021;325(11):1049–1052. [DOI] [PubMed] [Google Scholar]
- 47.Paradies Y, Ben J, Denson N, et al. Racism as a Determinant of Health: A Systematic Review and Meta-Analysis. PLoS One. 2015;10(9):e0138511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.O’Connor NR, Junker P, Appel SM, Stetson RL, Rohrbach J, Meghani SH. Palliative Care Consultation for Goals of Care and Future Acute Care Costs: A Propensity-Matched Study. Am J Hosp Palliat Care. 2017:1049909117743475. [DOI] [PubMed] [Google Scholar]
- 49.Arpey NC, Gaglioti AH, Rosenbaum ME. How Socioeconomic Status Affects Patient Perceptions of Health Care: A Qualitative Study. J Prim Care Community Health. 2017;8(3):169–175. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Casey JA, Pollak J, Glymour MM, Mayeda ER, Hirsch AG, Schwartz BS. Measures of SES for Electronic Health Record-based Research. Am J Prev Med. 2018;54(3):430–439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Starr LT, Ulrich CM, Appel SM, Junker P, O’Connor NR, Meghani SH. Goals-of-Care Consultations Are Associated with Lower Costs and Less Acute Care Use among Propensity-Matched Cohorts of African Americans and Whites with Serious Illness. J Palliat Med. 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Baram D, Daroowalla F, Garcia R, et al. Use of the All Patient Refined-Diagnosis Related Group (APR-DRG) Risk of Mortality Score as a Severity Adjustor in the Medical ICU. Clin Med Circ Respirat Pulm Med. 2008;2:19–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.McCormick PJ, Lin HM, Deiner SG, Levin MA. Validation of the All Patient Refined Diagnosis Related Group (APR-DRG) Risk of Mortality and Severity of Illness Modifiers as a Measure of Perioperative Risk. J Med Syst. 2018;42(5):81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Averill RF, Goldfield N, Hughes JS, Bonazelli J, McCullough EC, Steinbeck BA, Mullin R, Tang AM, et al. All patient refined diagnosis related groups (APR-DRGs): Methodology Overview. Wallingford, CT: 3M Health Information Systems;2003. GRP–041. [Google Scholar]
- 55.Gershon AS, Thiruchelvam D, Aaron S, et al. Socioeconomic status (SES) and 30-day hospital readmissions for chronic obstructive pulmonary (COPD) disease: A population-based cohort study. PLoS One. 2019;14(5):e0216741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Roos LL, Walld R, Uhanova J, Bond R. Physician visits, hospitalizations, and socioeconomic status: ambulatory care sensitive conditions in a canadian setting. Health Serv Res. 2005;40(4):1167–1185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Gupta N, Sheng Z. A population-based study of the association between food insecurity and potentially avoidable hospitalization among persons with diabetes using linked survey and administrative data. Int J Popul Data Sci. 2019;4(1):1102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Nelson KE, Wright R, Peeler A, Brockie T, Davidson PM. Sociodemographic Disparities in Access to Hospice and Palliative Care: An Integrative Review. Am J Hosp Palliat Care. 2021:1049909120985419. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Bazargan M, Bazargan-Hejazi S. Disparities in Palliative and Hospice Care and Completion of Advance Care Planning and Directives Among Non-Hispanic Blacks: A Scoping Review of Recent Literature. Am J Hosp Palliat Care. 2021;38(6):688–718. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60.Bradley N, Lloyd-Williams M, Dowrick C. Effectiveness of palliative care interventions offering social support to people with life-limiting illness-A systematic review. Eur J Cancer Care (Engl). 2018;27(3):e12837. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Chan B, Goldman LE, Sarkar U, et al. High perceived social support and hospital readmissions in an older multi-ethnic, limited English proficiency, safety-net population. BMC health services research. 2019;19(1):334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Barwise A, Jaramillo C, Novotny P, et al. Differences in Code Status and End-of-Life Decision Making in Patients With Limited English Proficiency in the Intensive Care Unit. Mayo Clin Proc. 2018;93(9):1271–1281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63.Barwise AK, Nyquist CA, Espinoza Suarez NR, et al. End-of-Life Decision-Making for ICU Patients With Limited English Proficiency: A Qualitative Study of Healthcare Team Insights. Crit Care Med. 2019;47(10):1380–1387. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Smith-Howell ER, Hickman SE, Meghani SH, Perkins SM, Rawl SM. End-of-Life Decision Making and Communication of Bereaved Family Members of African Americans with Serious Illness. Journal of Palliative Medicine. 2016;19(2):174–182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65.B P. Perceptions of hospice care among African Americans. J Hosp Palliat Nurs. 2011;13:281e287. [Google Scholar]
- 66.Gazaway S, Stewart M, Schumacher A. Integrating Palliative Care into the Chronic Illness Continuum: a Conceptual Model for Minority Populations. J Racial Ethn Health Disparities. 2019;6(6):1078–1086. [DOI] [PubMed] [Google Scholar]
- 67.Mayeda DP, Ward KT. Methods for overcoming barriers in palliative care for ethnic/racial minorities: a systematic review. Palliative & supportive care. 2019;17(6):697–706. [DOI] [PubMed] [Google Scholar]
- 68.Sharma RK, Cameron KA, Chmiel JS, et al. Racial/Ethnic Differences in Inpatient Palliative Care Consultation for Patients With Advanced Cancer. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2015;33(32):3802–3808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 69.Elnadry J. Home-Based Palliative Care Reduces Hospital Readmissions. Journal of Pain & Symptom Management. 2017;53(2):428–429. [Google Scholar]
- 70.Cassel JB, Kerr KM, McClish DK, et al. Effect of a Home-Based Palliative Care Program on Healthcare Use and Costs. J Am Geriatr Soc. 2016;64(11):2288–2295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 71.Ranganathan A, Dougherty M, Waite D, Casarett D. Can palliative home care reduce 30-day readmissions? Results of a propensity score matched cohort study. J Palliat Med. 2013;16(10):1290–1293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 72.Elliott AM, Alexander SC, Mescher CA, Mohan D, Barnato AE. Differences in Physicians’ Verbal and Nonverbal Communication With Black and White Patients at the End of Life. J Pain Symptom Manage. 2016;51(1):1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 73.FitzGerald C, Martin A, Berner D, Hurst S. Interventions designed to reduce implicit prejudices and implicit stereotypes in real world contexts: a systematic review. BMC Psychol. 2019;7(1):29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 74.Maina IW, Belton TD, Ginzberg S, Singh A, Johnson TJ. A decade of studying implicit racial/ethnic bias in healthcare providers using the implicit association test. Soc Sci Med. 2018;199:219–229. [DOI] [PubMed] [Google Scholar]
- 75.Starr LT, O’Connor NR, Meghani SH. Improved Serious Illness Communication May Help Mitigate Racial Disparities in Care Among Black Americans with COVID-19. J Gen Intern Med. 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 76.Chuang E, Hope AA, Allyn K, Szalkiewicz E, Gary B, Gong MN. Gaps in Provision of Primary and Specialty Palliative Care in the Acute Care Setting by Race and Ethnicity. J Pain Symptom Manage. 2017;54(5):645–653 e641. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 77.Adames HY, Chavez-Duenas NY, Fuentes MA, Salas SP, Perez-Chavez JG. Integration of Latino/a cultural values into palliative health care: a culture centered model. Palliative & supportive care. 2014;12(2):149–157. [DOI] [PubMed] [Google Scholar]
- 78.Cervantes L, Jones J, Linas S, Fischer S. Qualitative Interviews Exploring Palliative Care Perspectives of Latinos on Dialysis. Clin J Am Soc Nephrol. 2017;12(5):788–798. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 79.Islam JY, Deveaux A, Previs RA, Akinyemiju T. Racial and ethnic disparities in palliative care utilization among gynecological cancer patients. Gynecol Oncol. 2021;160(2):469–476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 80.Khan MZ, Zahid S, Kichloo A, et al. Gender, racial, ethnic and socioeconomic disparities in palliative care encounters in ischemic strokes admissions. Cardiovasc Revasc Med. 2021. [DOI] [PubMed] [Google Scholar]
- 81.Wen Y, Jiang C, Koncicki HM, et al. Trends and Racial Disparities of Palliative Care Use among Hospitalized Patients with ESKD on Dialysis. J Am Soc Nephrol. 2019;30(9):1687–1696. [DOI] [PMC free article] [PubMed] [Google Scholar]