[Skip to Navigation]
Sign In
Invited Commentary
September 22, 2021

Navigating the Complex Ecosystem of Race, Ethnicity, Structural Racism, Socioeconomic Factors, Medical Care Delivery, and End-of-Life Care—Casting Away the Compass to Make a Map

Author Affiliations
  • 1Department of Anesthesiology, Perioperative and Pain Medicine, Stanford University, Stanford, California
  • 2Division of Primary Care and Population Health, Palliative Care Section, Department of Medicine, Stanford University, Stanford, California
JAMA Netw Open. 2021;4(9):e2126348. doi:10.1001/jamanetworkopen.2021.26348

Patients dying of metastatic cancer may prioritize ameliorating pain, minimizing suffering, extending life, or exhausting all possible life-prolonging interventions. As patient prognosis, preferences, or experiences change, these goals may further evolve. Existing evidence suggests that patients from ethnic and racial minority groups are more likely to undergo aggressive interventions at the end of life and that disparities exist in both quality of care and use of palliative care services.1 More research is needed to better understand the rationale and ecologic factors associated with treatment decisions among patients from racial and ethnic minority groups at the end of life.

Deeb et al2 present the results of an epidemiologic study assessing data from the Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project national inpatient sample in the US from 2010 to 2017. The team created a database of patients with metastatic cancer who were hospitalized and died during the hospitalization (defined as a terminal hospitalization) and explored potential associations between aggressive end-of-life treatments and patient ethnicity, race, socioeconomic status, payer type, age, hospital type (teaching vs nonteaching), and geographic region. Aggressive inpatient end-of-life services were described as the receipt of systemic therapy, invasive mechanical ventilation, admission through the emergency department as a proxy for unplanned care, prolonged hospital length of stay (median, 6 days; interquartile range [IQR], 3-12 days), and/or increased total charges billed to insurance (median, $43 681; IQR, $17 973-$97 110). Among more than 58 million assessed records, the authors identified 321 898 hospitalizations of patients with metastatic cancer, with 21 335 patients considered to have terminal hospitalizations. In brief, the increased use of all aggressive inpatient end-of-life services was associated with public rather than private or other insurance coverage as well as admission to urban nonteaching and teaching hospitals rather than rural hospitals. Admission through the emergency department, receipt of mechanical ventilation, longer hospital stays, and higher charges, but not receipt of systemic therapy, were associated with non-White race and ethnicity.

The authors’ findings2 support those of previous studies1,3 in which minority race or ethnicity and lower socioeconomic status have been associated with higher rates of aggressive end-of-life interventions. Although these data highlight potential structural racism within the medical care system and beneficially illuminate the importance of studying associations between outcomes, race, ethnicity, and socioeconomic factors, they offer a limited view of a much more complex and nuanced ecosystem. First, by conflating these identifiers, the opportunity for a more delineated analysis is missed. Second, although the authors discuss that after controlling for socioeconomic and insurance status, high-cost, low-value care persisted among patients from minority groups, readers can appreciate neither potential directionality in, nor concurrence with, other unmeasured confounders. Identifying the etiologic factors associated with racial and ethnic disparities at the end of life is challenging, and these disparities are often associated with myriad inputs, particularly those within socioeconomic factors. Moreover, the aggressive end-of-life care that occurs among patients from racial and ethnic minority groups may not reflect a preference or value system but rather the ecosystem in which they receive their care. A higher prevalence of comorbid conditions and/or presentation with a higher severity of illness may be the predominant factors associated with the need for critical care, mechanical ventilation, or a longer hospital stay that incurs higher charges. Although the authors performed a subanalysis of payer and hospital type, future studies may better distinguish the factors within this complex ecosystem through stratification of patients by multiple potential confounders, such as illness severity scores or comorbidity indices.

These aggressive inpatient end-of-life services are actually a surrogate measure for proactive communication about end-of-life care. Indeed, only preemptive conversations discussing these treatments can lead to the choice to proactively avoid them. Existing data suggest that racial and ethnic minority patients are less likely to have do-not-resuscitate orders and advance directives suggestive of preemptive and proactive conversations.4 In addition, racial and ethnic discordance between patients and practitioners may further complicate such cancer-related communication.5 Thus, by the very nature of being in a minority group, individuals may be hampered in their ability to have meaningful preemptive conversations that avoid aggressive end-of-life treatments.

These outcomes of aggressive inpatient end-of-life services and low-value care for patients dying of cancer are not only inherently limited in scope but also imply a potentially inaccurate agenda. Quality measures used to retrospectively analyze end-of-life cancer care through secondary data sets were originally developed through a combination of literature reviews and expert and layperson focus groups and have since been gradually adopted into use as general quality measures. Yet these outcomes were chosen precisely because their coordination with International Classification of Diseases codes yields practical results in retrospective assessments of large data sets; at best, they offer limited and problematic insight into the quality and appropriateness of proactive care delivery. Emerging data suggest that aggressive end-of-life treatments, even those provided in an intensive care unit, may result in a death that is valued and prioritized by some patients and families.6 Personal, ethnic, cultural, and/or religious beliefs about appropriate end-of-life care also vary among not only patients and their family members but also clinicians.7 Furthermore, individuals as well as cultural, ethnic, and/or religious communities may strongly advocate for the intrinsic value of life—that there is no distinction between killing and allowing death to happen—and thus may find the treatment limitations required to meet the quality measures discussed to be anathema. For these individuals, high-quality and goal-concordant care would actually consist of these aggressive inpatient end-of-life services that are posited as having low value. The reality of providing compassionate, goal-concordant, and culturally sensitive end-of-life care is more nuanced and multifaceted than what can be reflected in broad-stroke quality measures primarily selected for secondary data set analyses.

Systemic racism is both pervasive and understudied in medical care. We strongly advocate for more and better research to assess how and when the medical care system provides culturally inappropriate or low-quality care to any minority community. Nowhere is the intersection between culture, ethnicity, race, and medical care delivery more complex and nuanced than at the end of life. To provide culturally appropriate and high-quality end-of-life care and reduce potential structural racism in the health care system, we need a more detailed map that illuminates actual and culturally meaningful associations between beliefs and end-of-life care choices, one that starts to untangle the myriad existing confounders. We need research that not only uncovers existing disparities but also illuminates potentially actionable means to reduce them. At best, studies using secondary data sets are but a cloudy compass; they may suggest a potential direction for future research but offer little insight into the actual ecosystem to be explored. Moreover, quality measures used in such secondary data analyses are rarely chosen by the minority community and may be implicitly biased and culturally insensitive. Proactive research that actively engages with minority communities is needed to better explore the complex intersection of race, ethnicity, socioeconomic factors, structural racism, health care delivery, and end-of-life care priorities. Only when outcomes are culturally sensitive and meaningful, when minority communities are able to collaborate with research and researchers, and when studies point out not just the existence of but also the amelioration strategies for care inequities will we start to navigate our way through this most complex of ecosystems to achieve more optimal, beneficial, compassionate, culturally sensitive, and equitable end-of-life care for all.

Back to top
Article Information

Published: September 22, 2021. doi:10.1001/jamanetworkopen.2021.26348

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Boateng A et al. JAMA Network Open.

Corresponding Author: Rebecca A. Aslakson, MD, PhD, Division of Primary Care and Population Health, Palliative Care Section, Department of Medicine, Stanford University, 1215 Welch Rd, Modular A, MC 5411, Stanford, CA 94305 (aslakson@stanford.edu).

Conflict of Interest Disclosures: None reported.

Additional Contributions: The authors thank David Magnus, PhD, Director of the Stanford Center for Biomedical Ethics at Stanford University, for his insights that aided in writing this manuscript. Dr Magnus did not receive compensation for his assistance.

Johnson  KS.  Racial and ethnic disparities in palliative care.   J Palliat Med. 2013;16(11):1329-1334. doi:10.1089/jpm.2013.9468PubMedGoogle ScholarCrossref
Deeb  S, Chino FL, Diamond LC, et al. Disparities in care management during terminal hospitalization among adults with metastatic cancer from 2010 to 2017. JAMA Netw Open. 2021;4(9):2125328. doi:10.1001/jamanetworkopen.2021.25328
Miesfeldt  S, Murray  K, Lucas  L, Chang  CH, Goodman  D, Morden  NE.  Association of age, gender, and race with intensity of end-of-life care for Medicare beneficiaries with cancer.   J Palliat Med. 2012;15(5):548-554. doi:10.1089/jpm.2011.0310PubMedGoogle ScholarCrossref
Erickson  SE, Vasilevskis  EE, Kuzniewicz  MW,  et al.  The effect of race and ethnicity on outcomes among patients in the intensive care unit: a comprehensive study involving socioeconomic status and resuscitation preferences.   Crit Care Med. 2011;39(3):429-435. doi:10.1097/CCM.0b013e318206b3afPubMedGoogle ScholarCrossref
Penner  LA, Dovidio  JF, Gonzalez  R,  et al.  The effects of oncologist implicit racial bias in racially discordant oncology interactions.   J Clin Oncol. 2016;34(24):2874-2880. doi:10.1200/JCO.2015.66.3658PubMedGoogle ScholarCrossref
Rolnick  JA, Ersek  M, Wachterman  MW, Halpern  SD.  The quality of end-of-life care among ICU versus ward decedents.   Am J Respir Crit Care Med. 2020;201(7):832-839. doi:10.1164/rccm.201907-1423OCPubMedGoogle ScholarCrossref
Curlin  FA, Nwodim  C, Vance  JL, Chin  MH, Lantos  JD.  To die, to sleep: US physicians’ religious and other objections to physician-assisted suicide, terminal sedation, and withdrawal of life support.   Am J Hosp Palliat Care. 2008;25(2):112-120. doi:10.1177/1049909107310141PubMedGoogle ScholarCrossref
1 Comment for this article
Guidance vs Grievance
Dilipsinh Solanki, MD |
I am a retired Oncologist with almost 50 years of practice: 20 years in academia, the rest in a 600-physician single specialty group in a suburban and small town practice.

I was proud of my ability to communicate softly yet candidly with my patients about their cancers at the start and at the end. I developed a skill to talk to them in the “medical language” that they could grasp and yet have a clear idea about what a treatment can and cannot do at every phase of their cancer. Most of the time, I could not communicate
with my Hispanic patients except thru an interpreter which always left me dissatisfied at the start and the end.

There are many issues involved here. The main ones are the lack of communication skills among the younger generation trained in the “high tech, keep doing something” mindset, time constraints (faster to order a test or treatment than to talk to the patient and family at length about an uncomfortable topic), pressure from the patient or family to do “something," self doubt, etc).

The authors here have reduced all this to racism instead of providing a more thoughtful analysis of possible reasons which could not be readily discerned in a retrospective study. The authors want us to accept that they know exactly what is behind these findings.

Perhaps JAMA Network Open should look for “real world” experts on matters such as this.