Race and Medicine

Race and Medicine

Issue: May 25, 2019
May 23, 2019
17 min read

Efforts to reduce racial disparities in cancer care aim to ‘make the invisible visible’

Issue: May 25, 2019
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

The trends are clear — not only do ethnic and racial minority groups face poorer cancer outcomes, they also are less frequently invited to participate in clinical trials, more often experience delays in treatment and initiation of palliative care, and are less likely to be offered genetic testing.

Study data even show that, on average, physicians spend less time in visits with minority patients.

To understand these subtle but pervasive racial inequalities, it may help to remember how recently racism was openly expressed within the American health care system.

Karen M. Winkfield, MD, PhD
Karen M. Winkfield

“I think people tend to forget that segregation was not that long ago,” Karen M. Winkfield, MD, PhD, past chair of ASCO’s Health Disparities Committee and associate professor of radiation oncology at Wake Forest Baptist Comprehensive Cancer Center, told HemOnc Today. “I trained at institutions where you can actually see the vestiges of the ‘Negro wards,’ which were oftentimes put in the basement. These patients didn’t have the same access to care. That happened well into the 1970s.”

Some structural remnants of segregation persist.

Kristen Pallok, MD, recent graduate of Rush Medical College in Chicago, explained how, for instance, the South and West of Chicago remain areas of “concentrated disadvantage.”

“There are only two Commission on Cancer-accredited hospitals in the whole South Side of Chicago,” Pallok, first author on a recent perspective published in The New England Journal of Medicine on structural racism, told HemOnc Today. “A patient who is living in South Side neighborhoods of concentrated disadvantage might have to travel for miles to get to a high-quality center.”

David A. Ansell, MD, MPH, and Kristen Pallok, MD
As long as structural racism remains intact, racial disparities in cancer care will persist, according to a NEJM perspective that included David A. Ansell, MD, MPH, and Kristen Pallok, MD as two of its authors, along with Fernando De Maio, PhD (not pictured). “When you attack the structural underpinnings of the system, you can actually make a difference,” Ansell told HemOnc Today.

Source: Adam Daniels.

As long as this structure remains in place, racial disparities likely will continue, perspective co-author David A. Ansell, MD, MPH, senior vice president for community health equity at Rush University Medical Center, told HemOnc Today.

“When you concentrate poverty and you concentrate race, you have different health care experiences. That’s just a fact,” he said. “It’s tied to funding and capitalization. It is structural, because it’s built into the policies and procedures, the way we fund health care and the way we do things.”


HemOnc Today spoke with oncologists, social psychologists and health disparities experts about what drives racial discrimination in the health care system, obstacles to quality care in minority communities, and what is being done to recognize and eliminate these differences.

‘Multivariate in the extreme’

Structural racism is only one facet of the complex and deep-seated issue of cancer care discrimination.

Rather, it involves an inextricable mix of socioeconomic, cultural, psychological and institutional factors, according to Louis A. Penner, PhD, social psychologist and emeritus professor in the department of oncology at Wayne State University.

“People ask me why black patients with cancer have worse outcomes than whites,” Penner told HemOnc Today. “They ask me, ‘Well, is it socioeconomic?’ I say ‘yes.’ They ask, ‘Is it due to racial prejudice?’ and I say ‘yes.’ Finally, I say to them, ‘We could do this all day, because there is no single cause of this.’ The problem is multivariate in the extreme.”

The largest racial disparities exist for treatable cancers, because the more treatable a cancer is, the more discretion the patient and clinician have regarding the treatment plan, allowing for bias-driven decisions, Penner said. He used an example of prostate cancer, often a slow-progressing, treatable disease.

“The physician might, for various reasons, look at a patient and say, ‘The side effects of a prostatectomy right now are probably worse than the cancer,’ and the patient may agree,” Penner said. “However, in lung cancer, for instance, the patient can’t say, ‘Well, let’s wait and see.’ There aren’t as many options. The patient almost always gets immediate treatment. Thus, it is less likely that physicians’ and patients’ pre-existing thoughts and feelings would affect treatment decisions in lung than in prostate cancer.”

Another important driver of inequalities in cancer care is what Penner terms “race-related attitudes,” ingrained and often unconscious ideas among both patients and clinicians.

“These are the attitudes that physicians and patients bring to these health care interactions that are reflective of race relations in the United States,” he said.

Many problematic race-related behaviors are the products of implicit bias — attitudes or stereotypes about a minority group that unconsciously affect behavior, Penner said.

Implicit bias — an instant, unconscious and reflexive reaction — persists in society at large and in health care.

In a 2012 study, Sabin and colleagues evaluated data on 404,277 individuals from Harvard University’s Project Implicit database, which includes results of website visitors who took the Race Attitude Implicit Association Test during a 28-month period. Of the test-takers, 2,535 were MDs.


Results showed that physicians demonstrated an implicit preference for white Americans vs. black Americans. In all cases except for black MDs, the implicit bias surpassed self-reported bias.

“MDs, especially white, male MDs, had much higher implicit bias than the general public,” Penner said.

Simply making physicians aware of their implicit biases is not enough to correct this behavior, Penner said. Instead, physicians need to change their ways of perceiving and communicating with black patients.

“To me, the major thing we can do is teach physicians two things. First, teach them how to see the patient as an individual,” he said. “The patient isn’t a black woman; the patient is Mrs. Brown. She has four children and works as a teacher. Generally, in social psychology, the way you reduce racial bias is to individuate the person; that is, to see them as a unique individual. You can teach physicians to do that.”

A second skill physicians can learn to improve is patient-centered communication with all patients.

“If you can communicate more effectively with a minority patient, they aren’t going to react as negatively,” Penner said.

Medical mistrust: ‘A perfect storm’

Another significant obstacle to quality care for minority patients is medical mistrust, which often occurs in situations of racial discordance.

Mistrust of physicians can lead to less confidence in the physician and the treatment, which may result in lack of patient adherence to the regimen.

“Chemotherapy often makes you sick, and it costs money,” Penner said. “You’ve gone through a cycle of chemotherapy, you feel like hell, you go off, you feel better. For the next chemotherapy, you may have a 20% copay. If you don’t have confidence in the doctor and you don’t have confidence in the treatment, what we believe happens is you are much more likely to drop off treatment.”

Additionally, if a black patient doesn’t show much confidence in a treatment approach, a physician may perceive this as unwillingness to cooperate with the plan. This may exacerbate the problem, by confirming physician stereotypes about poor compliance among minority patients.

A 2016 study by Penner and colleagues examined the effects of oncologists’ implicit biases on the attitudes of black patients.

“The reason we thought this was important to study is because if you have cancer and you’re going to see an oncologist, one possibility is that you don’t really care what that physician says or does,” he said. “All you want is to be cured.”

However, his findings suggested otherwise: Even when faced with a potentially life-or-death situation, black patients with cancer expressed dislike and mistrust of doctors with higher implicit bias. Additionally, the study found that physicians’ implicit bias had an indirect effect on patients’ confidence in the recommended treatments.

Louis A. Penner, PhD
Louis A. Penner

“If the physician is high in implicit bias and the patient is high in mistrust, the physician sees it as significantly less likely that the patient will tolerate the treatment. And, the patient has less confidence in the treatment, as well,” he said. “It’s a perfect storm.”

Penner said this vicious cycle can be interrupted by teaching better communication skills to the clinician, the patient or both. As clinicians begin to better engage with black patients, the patients may respond with greater trust, he said.

“Another thing we’ve done with some success is to teach the patients to be more assertive,” Penner said. “If the patients are more assertive, they’re going to cause the doctor to individuate them more.”

The scope of the problem

Even as cancer research has led to more advanced, targeted cancer treatments than ever before, not all patients are able to reap the benefits of these efforts, as demonstrated by their persistently worse survival outcomes, lack of accrual onto clinical trials and lower rates of genetic testing.

According to the American Cancer Society, black patients have the highest death rates and shortest duration of survival of any racial or ethnic group in the U.S. for most cancers.

Studies have shown worse outcomes among black patients for various types of cancers; Murphy and colleagues evaluated data from 27,828 patients with pancreatic cancer and found that although black patients are disproportionately affected by pancreatic cancer, they were less likely than white patients to undergo pancreatic resection.

Similarly, although black women have the highest rate of mortality from cervical cancer, study results presented at Society of Gynecologic Oncology Annual Meeting on Women’s Cancer showed they are less likely than white women to receive brachytherapy boost as part of their treatment.

“When we stratified our analysis by whether black patients received brachytherapy or not, we found that black patients who received brachytherapy or maximal treatment did as well as white patients,” Martin T.H. King, MD, director of brachytherapy clinical operations at Dana-Farber Cancer Institute, told HemOnc Today when the study was presented. “But when they didn’t receive brachytherapy, they seemed to do worse than white patients. Basically, what this tells us is that a lot of this disparity in survival outcomes is related to undertreatment.”

Additionally, the American Cancer Society reported that black women are significantly more likely to die of breast cancer than white women, a disparity that has increased over time. African American men have more than double the likelihood of dying of prostate cancer than white men, and African Americans have a higher incidence of colorectal cancer.


Other minority groups demonstrate higher rates of other cancers.

Hispanic and native American/Alaska native women have a higher prevalence of cervical cancer than women of other races or ethnicities. Native Americans and Alaska natives have the highest prevalence of liver and intrahepatic bile duct cancer, followed by Hispanics and Asian/Pacific Islanders. Kidney cancer is more prevalent among American Indians and Alaska natives than other racial or ethnic groups.

Despite these higher incidences, minority patients are widely excluded from clinical trials.

Studies have shown that although black patients are as willing as whites to participate in clinical trials, they are less frequently invited to participate, and those who are enrolled are more likely to be insured and have a higher socioeconomic status.

Less than 8% of clinical trial enrollees are Hispanic, even though Hispanics represent 17% of the U.S. population.

According to Winkfield, many black and immigrant patients present to hospitals that mainly treat uninsured or underinsured patients. These hospitals often lack the clinical trials that are conducted at major academic institutions.

“Clinical trials allow patients to have access to state-of-the-art care and can sometimes provide access to new therapies at no additional cost to the patient,” she said. “It may also mean obtaining a higher acuity of care from your providers, if the study requires being seen on a more frequent basis.”

Winkfield added, however, that implicit bias often is another factor in the low accrual of black and Hispanic patients in clinical trials.

“Just as with Spanish-speaking patients, the perception is that African Americans are going to be harder to enroll or that it’s going to take longer to explain things,” she said. “Minority patients are disproportionately impacted by lower socioeconomic status — but that doesn’t mean they aren’t intelligent.”

Black patients are also less likely to be offered genetic testing for some cancers.

McCarthy and colleagues administered a survey to 3,016 women in Pennsylvania and Florida diagnosed with stage I or stage II breast cancer, their medical oncologists (n = 808) and their surgeons (n = 732).

Results — published in Journal of Clinical Oncology — showed black women were less likely to undergo genetic testing (OR = 0.66; 95% CI, 0.53-0.81) and less likely to be advised to do so by their surgeon or oncologist (OR = 0.66; 95% CI, 0.54-0.82).

Also, care appeared to be “highly segregated,” with physicians who treated white women reporting they had practices with mostly white patients, and physicians who treated black women reporting practices comprised primarily of black patients.


The racial disparity in genetic testing may be partly due to insurance limitations, according to Beverly Moy, MD, MPH, associate professor of medicine at Harvard Medical School, clinical director of the breast oncology program specializing in medical oncology at Massachusetts General Hospital and HemOnc Today Editorial Board Member.

“Inadequate health insurance that doesn’t offer some of the most essential services related to cancer is certainly an important contributor to racial inequities in cancer care,” Moy told HemOnc Today. “ASCO has issued policy statements about important provisions that Medicaid should cover, including routine costs of clinical trial participation and cancer prevention testing without copays. Meaningful Medicaid reform is needed to remedy this situation.”

Gaps in palliative care

Inequalities in cancer care are not only evident in terms of curative treatment, but also in the palliation of symptoms.

A retrospective, population-based analysis published in Pediatrics detailed socioeconomic differences in the extent of palliative care provided to children with complex chronic conditions, including cancer. In the review of 8,654 children who died in California between 2000 and 2013, researchers found discrepancies in care related to income, race/ethnicity and age.

For instance, living in a low-income neighborhood increased likelihood that a child would die in the hospital and undergo a medically intense intervention at the end of life. Hispanic and “other” race or ethnicity was associated with hospital death and more than two markers of intense end-of-life care.

Studies have also revealed differences in the amount of time oncologists devoted to palliative care discussions with minority patients.

Cardinale B. Smith, MD, PhD, associate professor in the division of hematology and medical oncology at Tisch Cancer Institute and Brookdale department of geriatrics and palliative medicine at Mount Sinai Hospital, discovered this while studying a communication skills intervention for oncologists.

“We audio-recorded the visits after the intervention to see if it had an impact,” Smith told HemOnc Today. “One of the things we identified was a marked discrepancy in the amount of time the physicians spent with their minority vs. nonminority patients.”

Smith said that, on average, the clinicians spent 17 minutes with nonminority patients and 14 minutes with minorities. She said given the magnitude of that conversation, the lack of time spent was unexpected.

“These are incurable cancer patients who have started their full-blown treatment and were coming in for their first scan after starting treatment,” she said. “So, at the time of that visit, it’s either ‘things are going well, let’s talk about the future,’ or ‘things aren’t going well.’ So, it’s a surprising finding.”


Smith said she doesn’t think her findings were driven by financial factors: all patients were able to access care, and different types of hospitals were represented.

There were likely many factors associated with these findings, including implicit bias and medical mistrust, Smith said, adding that an overall workforce shortage in palliative care remains to be addressed.

“There aren’t going to be enough bodies to do the work, so we really have to focus on educating those clinicians at the frontlines who are treating these illnesses,” she said. “We need them to have at least a basic level of palliative care knowledge so they can do a baseline amount of this work.”

Employing technology

Because it is such a complicated and multifactorial issue, racial disparity in cancer care will likely need to be tackled through various approaches.

Samuel Cykert, MD
Samuel Cykert

Samuel Cykert, MD, professor of medicine in The University of North Carolina at Chapel Hill’s division of general internal medicine, has found a way to apply digital technology to this effort.

Cykert developed a three-pronged, race-aware intervention and evaluated it in a 5-year study conducted at five cancer centers. The study focused on patients with early-stage non-small cell lung cancer, a population he chose for its bleak prognosis if left untreated.

“It struck me that even with early-stage lung cancer, if it’s left untreated, your chances of dying within a year are 30% to 40%, and your chance of dying within 4 years is 100%. That’s a strong motivation to get treated.”

Prior to introducing the intervention, the rates of treatment for early-stage lung cancer were 78% for white patients at participating institutions compared with 69% for black patients.

To develop the intervention, researchers consulted with members of the black community who had undergone cancer treatment.

“One thing we heard was that this can’t be left up to individuals, because responsibility gets diffused,” Cykert said. “It has to be a system-based intervention.”

Members of the community also emphasized the need for transparency and accountability at each stage of the intervention.

Based on this feedback, researchers developed a program that consisted of three components: a real-time warning system utilizing electronic health records, race-specific input to clinical teams on their rates of treatment completion, and a nurse navigator to interact with patients over the course of their treatment.

Researchers collected data from the EHRs on routine patient physician appointments, appointments for radiology testing, biopsies and treatment.

“We could extract files that looked at when these things were scheduled, and then whether the treatment or the procedure was completed,” he said. “If a patient missed an appointment, we knew it within 24 hours, and a warning would come up. We also programmed expected milestones in care so we could inform clinicians when patients adherent to appointments experienced inertia in care.”


Another component of the intervention was the hiring of trained nurse navigators who would see the warnings and take appropriate action within the hospital system.

“We had a group of navigators who were specially trained in the history of system-based racism, the history of implicit bias, and the history of barriers to care that were worse for African Americans than whites,” Cykert said. “These navigators would see the warnings. If the warning was a patient warning, the navigator would engage the patient and draw them back into care.”

The intervention yielded a significant increase in treatment rates for both black patients (96.5%) and white patients (95%).

Cykert said he hopes more institutions will use technology and knowledge of implicit bias to implement programs designed to reduce racial disparities.

“With all these digital data available, we can build systems that can be used in near-real-time to improve care and mitigate disparities,” he said. “Also, we can measure institutional bias and differential care. We can think about an intervention and do it.”

‘Making the difference’

In their recent analysis, Pallok, Fernando De Maio, PhD, and Ansell illustrated structural racism through the breast cancer battle of one 60-year-old, African American woman in Chicago.

The woman, concerned with a painful breast lump, presented to a community ED due to her lack of insurance. Because the woman previously had normal mammograms, the emergency physician did not conduct diagnostic testing or follow-up. The clinician suspected an infection and sent the patient home with antibiotics.

When the lump remained and additional imaging tests suggested breast cancer, a general surgeon removed the lump and suggested mastectomy. The patient was not referred to an oncologist or even told the stage or type of her breast cancer.

“She had a mammogram, and then was referred to see a surgeon at the hospital,” Pallok said. “They basically told her, ‘This is what we’re going to do, and this is the procedure you’re going to have.’ In this case, it was an excisional biopsy.”

Ansell said this series of events departed from standard of care in many ways.

“First, she got a screening mammogram, which was incorrect. It should be a diagnostic mammogram that includes an ultrasound,” Ansell said. “Then the recommendation should be for a needle biopsy and staging workup, and based on that needle biopsy, treatment is specifically prescribed. Instead, she just had the lump removed, and then went back for the mastectomy, which was not necessary at that time.”

Fortunately for this patient, a navigator from the Metropolitan Chicago Breast Cancer Task Force was placed at that hospital. Part of the navigator’s responsibility was to review abnormal imaging tests and reach out to the patient who had the abnormality.


“I call this ‘navigating to quality,’” Ansell said. “The patient said this interaction happened just in time, just before she got her breast removed.”

The patient was spared the unnecessary surgery, but other patients may not be so lucky.

“The point we make about this case is, this is how structural racism works,” Ansell said. “It’s one thing to put out words like ‘structural racism’ — people may or may not have a reaction to that. In this case, we showed exactly how it works.”

Ansell said he and his co-authors consider structural racism to be a subcategory of structural violence.

“It’s structural because it is built in to the process and procedures of institutions and it is violent because people die unnecessarily as a result,” he said. “We’ve been caught up in a way of thinking about disease as only biological, without taking into account how neighborhood conditions determine who lives and who dies.”

In addition to the use of navigators, the Metropolitan Chicago Breast Cancer Task Force also evaluates hospitals in terms of meeting quality benchmarks, such as properly training mammography technicians, and works with the hospitals to improve quality.

“Over time, with people readily sharing quality data, we were able to show improvements in these institutions,” Ansell said. “As a result of this and other activities, the mortality of black women in Chicago has dropped more than in any city in the United States.”

Breast cancer mortality has dropped by 13% among black women in Chicago, and the gap in care between black and white women has dropped by 20%, a disparity reduction not seen in any other American city with a large black and white population.

These improvements were made by recognizing the unacknowledged and often invisible structural racism that was built into Chicago’s health care system, Ansell said, adding that these more ambiguous forms of racism can potentially cause more harm than overt racism.

“In some ways, it’s more deeply disturbing, because it’s almost hardwired into our day-to-day systems,” he said. “We need to undo these systems that almost invisibly perpetuate inequities. That’s why one of our recommendations in the study is ‘make the invisible visible.’”

Ansell said the success of Chicago’s task force has shown that although racism is a deeply embedded part of American culture, the idea of eradicating it is not an impossible dream.

“It’s not immutable, it’s not intractable,” he said. “What we found in Chicago is that when you attack the structural underpinnings of the system, you can actually make a difference.” – by Jennifer Byrne

Click here to read the sidebar article.


Alimena S, et al. Abstract 11. Presented at: Society of Gynecologic Oncology Annual Meeting on Women’s Cancer; March 16-19, 2019; Honolulu.

Johnston EE, et al. Pediatrics. 2019;doi:10.1542/peds.2018-2228.

McCarthy AM, et al. J Clin Oncol. 2016;doi:10.1200/JCO.2015.66.0019.

Murphy MM, et al. Cancer. 2009;doi:10.1002/cncr.24433.

Pallok K, et al. N Engl J Med. 2019;doi:10.1056/NEJMp1811499.

Penner LA et al. J Clin Oncol. 2016;doi:10.1200/JCO.2015.66.3658.

Smith CB, et al. Abstract 19. Presented at: Palliative and Supportive Care in Oncology Symposium; Nov. 16-17, 2018; San Diego.

Sabin J, et al. J Health Care Poor Underserved. 2009;doi:10.1353/hpu.0.0185.

Wujcik D and Wolff SN. J Health Care Poor Underserved. 2010;doi:10.1353/hpu.0.0251.

For more information:

David A. Ansell, MD, MPH, can be reached at 1725 W. Harrison St. #263, Chicago, IL 60612; email: david_ansell@rush.edu.

Samuel Cykert, MD, can be reached at 145 N. Medical Drive, CB #7165, Chapel Hill, NC 27599-7165; email: samuel_cykert@med.unc.edu.

Beverly Moy, MD, MPH, can be reached at 55 Fruit St., Yawkey 9A, Boston, MA 02114; email: bmoy@mgh.harvard.edu.

Kristen Pallok, MD, can be reached at 600 Paulina St., Suite 403, Chicago, IL 60612.

Louis A. Penner, PhD, can be reached at Mid-Med Lofts, Suite 3000, 87 E. Canfield St., Detroit, MI 48201; email: pennerl@karmanos.org.

Cardinale B. Smith, MD, PhD, can be reached at 1190 5th Ave, New York, NY 10029; email: cardinale.smith@mssm.edu.

Karen M. Winkfield, MD, PhD, can be reached at 1 Medical Center Blvd., Winston-Salem, NC 27157.

Disclosures: Ansell, Cykert, Moy, Pallok, Penner, Smith and Winkfield report no relevant financial disclosures.