Accounting for the High Enrollment of African Americans on Winship Cancer Institute’s Myeloma Clinical Trials

Abstract

Background

Thirty-four percent of Multiple Myeloma (MM) clinical trial participants at Winship Cancer Institute (Winship) are African American (AA); however, AAs make up only 4.5 percent of myeloma clinical trial participants in the United States. Given our high enrollment, we aimed to measure AAs’ trust in providers and identify if clinical trial enrollment barriers exist.

Methodology

A member of the ethics research team surveyed AA patients who had consented to a MM clinical trial at Winship. Three validated surveys were used: Trust in Medical Research (TMR); Human Connection (THC) which measures how much patients feel they are heard and valued by their physicians; and the Duke Intrinsic Religiosity Scale (DUREL) which measures strength of religious engagement and belief. The survey also included questions about the impact of side effects, distance to the trial center and trial related costs on the decision to participate in clinical trial.

Results

Ninety-two percent (61/67) of patients approached consented. The mean TMR score and the mean THC score were significantly higher (P-value < 0.001) than the results obtained in key national surveys (TMR 14.9 compared to 11.65; THC 57.7 compared to 54.6). These two surveys were significantly correlated, meaning trust and human connection increase or decrease in tandem. The 3 religiosity subscale results showed high religiosity (3.84, 4.36, and 4.35 with 5 being the highest score). The mean scores of the importance of the investigational agent’s side effects, trial costs, and distance to trial center on the decision to enroll in a clinical trial were also high (8.5, 7.8, and 6.5, respectively, with 10 being the most important).

Conclusion

In our study population, high trust and human connection overcame other trial participation barriers: strong religious beliefs and concerns about side effects, costs, and travel distance. We present a roadmap to guide investigators to increase human connection, and hopefully trust.

Introduction

Multiple myeloma (MM) is the most common hematologic malignancy among African Americans (AA) in the United States [1]. It is estimated that 7810 new cases of multiple myeloma will occur among black people in 2022, and 2530 will succumb to the disease [2]. The exact cause of MM remains unknown. Despite uncertainty about the etiology of MM, standard of care treatments and clinical trial research allow patients to manage the disease for years [3].

Clinical trials are crucial for providing evidence for the efficacy of novel multiple myeloma therapies as well as for improving medical knowledge about the condition itself. In 2015, a promising clinical trial was conducted for the ground-breaking myeloma drug NINLARO. Considering the prevalence of MM in AAs, it was unfortunate that only 1.8% (13/722) of the trial’s participants were AA [4]. Nationally, AA enrollment in myeloma clinical trials remains low - only 4.5% of total enrollment [5,6]. The clinical trial community must work to overcome barriers to trial enrollment so that a representative number of AAs participate in myeloma trials.

The most common barrier to AA participation in clinical trials is a lack of trust. Numerous studies have found that AAs do not trust the doctors who are treating them [5,7–11]. AAs also have been noted to have suspicion or fear of experimentation rooted in historical events, such as the Tuskegee Syphilis Study [12]. Typical barriers compounding this mistrust are concerns about extra financial or logistical burdens, a lack of access or opportunity to participate, fear of toxicity or potential side-effects, strict eligibility criteria which can differentially affect African Americans and other minorities [13], and provider barriers like implicit bias or lack of human connection in the doctor-patient relationship [10,14–16].

Thirty-four percent (265/781) of participants on Winship myeloma trials between 2018 and 2021, the timeframe of eligibility for this study, were African American. This percentage is typical of myeloma trial enrollment at Winship with AAs comprising 34% (94/280) of trial participants between 2014 and 2017. To begin to account for Winship’s success in enrollment of AAs to MM trials, we conducted this study to evaluate patients’ level of trust with medical researchers. We also sought to identify and examine enrollment barriers existing at Winship.

Methods

STUDY DESIGN

This survey study was conducted at Emory Winship Cancer Institute and was performed with the Protocol Review and Monitoring Committee (PRMC) and Emory Institutional Review Board approval (STUDY00002697). Our methodology was qualitative description [17].

SELECTION OF PATIENTS

African American, adult, English speaking, Multiple Myeloma patients who had consented to a clinical trial at Winship between 2018 and 2021 were eligible. Two methods were used to identify eligible patients. Nurses and clinical trial coordinators in the Phase 1 unit and infusion center of Winship identified eligible patients and we queried Winship’s OnCore Clinical Trials Management System. A study team member approached each potential participant at their next appointment, explained the study in detail, provided the potential participant with the verbal consent form and offered participation. If the potential participant consented to participate in the study, the survey was conducted in person or virtually. Consent was documented by completion of the interview.

STRUCTURED SURVEY

The survey’s design was based on key literature, which identified validated scales or questions that had already been successfully applied by other researchers to the domains of this study. The domains are described in Table I. The three validated scales used in the survey were the Trust in Medical Research scale (TMR) produced by Hall et al. [18], the Human Connection (THC) scale developed by Mack et al. [19], and the Duke Intrinsic Religiosity Scale (DUREL) [20]. Adapted questions from Lannin were used to further assess religious influence on participants [21]. The survey also included questions adapted from Advani et al, to assess the impact of side effects, distance to the trial center and trial related costs on the decision to participate in clinical trial [21]. In total, the survey questions consisted of 28 Likert-scale questions, and four true or false questions. The questionnaire is available in Supplemental Materials. Sociodemographic variables were assessed, including age, gender, education level, employment status, and income level (Table II).

Table I:

Survey Domains

Domain	Description	Assessment Tool
Trust in Provider	Baseline measure of general trial participants’ trust in medical research	Validated 4-question scale designed by Weinfurt and Sugarman [18]
Comfort with Provider	To assess trial participants’ comfort with their providers	The Human Connection Scale (THC) designed by Mack et al. [19]
Harms and Logistics of Clinical Trials	Assess the importance of harms, and logistics of clinical trials in deciding to participate in clinical trials	Clinical trial participation assessment questions used by Advani et al. [9]
Religiosity	Religious belief impact on participation in clinical trials.	Closed-ended Likert-scale questions; Duke University Religion Index (DUREL) 5-item measure and questions adapted from Lannin et al. [21]
Demographic Information	Age, race, sex, employment, income, marital status, insurance status, religiosity	Standard Demographic Sheet

Table II:

Demographics

African American	61
Female (n, %)	31 (50.8%)
Education (n, %)
Some high school	2 (3.3%)
High school graduate/GED	18 (29.5%)
Some college	14 (23%)
College graduate	15 (25%)
Some post-graduate	2 (3.3%)
Post-grad/Prof degree	10 (16.4%)
Employment status
Employed Full-Time	16 (26.2%)
Employed Part-Time	4 (6.6%)
Retired	29 (47.5%)
Not Employed	2 (3.3%)
Disabled	10 (16.4%)
Annual Household Income
Less than $19,999	6 (9.8%)
$20,000-$59,999	22 (36%)
$60,000 - $79,999	12 (19.7%)
More than $80,000	14 (23%)
Don’t know	3 (4.9%)
Refused	4 (6.6%)

PROVIDER MEETING

We attended a myeloma working group meeting by videoconferencing to ask myeloma providers to describe the factors that contributed to their success in enrolling AA patients. Only providers who consented verbally to participate in our study were included in the discussion. These participants included myeloma physicians, clinical research coordinators, registered nurses, and mid-level providers.

Since a discordance between provider and patient ethnicity and race has been cited as a reason for low enrollments, we asked the providers if they would be willing to identify their ethnicity and race by email.

OUTCOMES AND MEASURES

The primary objective was to examine the level of trust in AA myeloma trial participants at Winship using TMR. TMR is a four-item questionnaire that measures patient confidence in the safety, fidelity, and honesty of research. The items are graded on a scale of 1 to 5, with some items reverse coded so that higher scores indicate greater trust in medical researchers and lower scores indicate greater mistrust. We compared the mean score for our participants to the TMR’s mean score (11.65) for the African Americans in the Hall national study. The Hall study participants were mostly white (n=3623, 89% white). The mean score for the white participants in the Hall study was 12.46 and the overall mean was 12.4.

Secondary objectives aimed to identify any existing barriers present at Winship. We assessed participants’ connection with their providers, levels of importance of clinical trial related factors (i.e., costs, distance, side effects), and religiosity. The sixteen item Human Connection (THC) Scale developed by Mack et al was utilized to assess the extent to which patients felt a sense of mutual understanding, caring, and trust with their physicians The population of the Mack study included 217 participants, 85% white and 8% AA, with a mean THC score of 54.7 (Range: 16 to 64). The five item Duke University Religion Index (DUREL) and adapted questions from Lannin et al. were incorporated to evaluate religiosity. The three components of the DUREL include organizational religious activity, non-organizational religious activity, and intrinsic religiosity (or subjective religiosity). Organizational religious activity (ORA) involves public religious activities such as attending religious services or participating in related religious activity (prayer groups, Scripture study groups, etc.). Non-organizational religious activity (NORA) consists of religious activities performed in private, such as prayer, scripture study, watching religious TV or listening to religious radio. Intrinsic religiosity (IR) assesses degree of personal religious commitment or motivation. IR involves pursuing religion as an end-in-itself. The DUREL measures each of these dimensions separately as reported in Table III. Lannin’s four true or false questions assessed participants’ religious beliefs in relation to medical research [21]:

Table III:

Mean Scores of Survey Reponses

Survey	Mean	Maximum
TMR (n=61)	14.9	20
THC (n=61)	57.7	64
DUREL-ORA (n=59)	3.84	5
DUREL-NORA (n=59)	4.36	5
DUREL-IR (n=59)	4.35	5
Distance (n=61)	6.54	10
Cost (n=61)	7.54	10
Side Effects (n=61)	8.48	10

1. If a person prays about cancer, God will heal it without medical treatments,

2. Since God will determine whether or not I will die from my cancer, I must just believe in Him and trust Him rather than look for research to cure my cancer.

3. God can work through doctors to heal people

4. God can work through clinical trials to heal people

Adapted questions from Advani were used to measure overall importance of trial related factors on the decision to join a clinical trial. These factors included concerns about costs, distance to the site, and side effects.

STATISTICAL ANAYLSIS

Mean scores on the Trust Scale among participants in this study were compared to 11.65, the average of the AA population (n=3623, 89% white) in the Hall study, using a one-sample t-test. Association between continuous measurements trust and human connection was assessed using Pearson’s correlation coefficient. Group comparisons were assessed using chi-squared tests, two-sample t-tests, or non-parametric equivalents where appropriate. For the Human Connection Scale and DUREL Index, and other measures, we conducted descriptive analyses. Statistical significance was assessed at the 0.05 level, and all tests were two-sided unless otherwise noted.

Results

PARTICIPANT DEMOGRAPHICS

We presented the survey to 67 patients, and 92% (61/67) participated. Lack of interest, being unwell, or lack of compensation were the main justifications for declining this study. Survey participants’ demographics are listed in Table II. Thirty-seven myeloma providers consented to discuss their thoughts on why they were so successful in recruiting AAs. These participants included 5 physicians (14%), 5 nurses (14%), 3 regulatory specialists (8%), 11 clinical research coordinators (30%), 5 data scientists (14%), and 8 others (22%). Two of the myeloma physicians were Non-Hispanic white, 1 Hispanic White and 2 South Asian. No physician self-identified as African American. Seven research coordinators (6 African American and 1 Non-Hispanic white) and 4 nurses (2 African American and 2 Non-Hispanic white) shared their demographics.

SURVEY RESULTS

The mean Trust in Medical Research (TMR) score among the African American population at Winship was 14.9, significantly higher than the 11.65 observed for Hall’s national sample of AAs used to validate the scale (P < 0.001) and higher than the score for the white participants in the Hall study (12.46) [18]. The average human connection (THC) score was 57.7, which was significantly higher than the 54.6 found in a sample of 217 patients in the Mack trial, 85% of whom were white (P -value < 0.001). The Pearson correlation coefficient between the TMR score and the THC score is 0.39, demonstrating a substantial positive association (P-value = 0.002) between human connection scores and trust scores. The DUREL scales had high mean scores indicating that participants view their religion as the framework for their lives and are active participants in religious activities (Table III).

Several participants chose not to respond to the Lannin poll questions, with 51/61 responding to question 1, 52/61 responding to question 2, and 59 responding to questions 3 and 4. Seventy-three (41/52) percent of those polled disagreed with the assertion that if someone prayed about cancer, God would naturally heal it without the need for medical intervention. Seventy-nine percent (41/52) of participants disagreed with the statement, “Since God will decide whether or not I will die from my cancer, I must believe in Him and trust Him rather than look for science to cure my cancer.” Ninety-seven percent (57/59) of participants agreed that God can heal people through clinical trials and doctors, suggesting that strong religious beliefs, did not deter our population from trial enrollment. The average importance of side effects, cost and distance were 8.48, 7.8, and 6.5 respectively, on a scale of 0–10, with 10 being of highest importance or influence on decision to enroll in a clinical trial.

PROVIDER DISCUSSION

At our meeting with myeloma providers at Winship, we shared our preliminary data and asked the providers for their thoughts on the factors contributing to the high AA enrollment at Winship. The providers’ responses are outlined below.

First, the providers pointed out that Emory has both a nursing and medical school and the curricula includes a focus on awareness of social issues of healthcare, which has led to intentionality in creating trust with AAs.

I’ve been at Emory for four years, and I also went here for school. So, I might be a little bit biased, but I think its [high enrollment] possibly due to the fact that we’re teaching hospital and we encourage learning and education, and I just think that that sometimes makes for an environment that is very aware of issues that are maybe not obvious to some. We talked a lot about the African American population and a general distrust of health care. [Female, registered nurse]

Providers also stressed that they treat every person with respect.

“ So I went to every provider that treats Myeloma, and asked the same question, what do we do differently than the others to see this higher accrual? So the one unanimous answer that I got was we respect our patients so we don’t treat anybody any differently. And that has led us to build the trust and faith in a system” [Male, Physician]

Providers also mentioned that trust is built from the very first visit.

A clinical trial participation does not start on the day that a patient is progressing. A clinical trial starts on the day that you see a patient. So that trust builds on over time and time and time. [Male, physician]

I don’t think it’s [trust] a one-time deal. I think it’s something you build up over a period of weeks and months, multiple visits. [Male, physician]

The providers stated that involving the care team early on helps to increase trust.

I think probably having all the team meet the patient as early as possible during their management, the nurse and the nurse practitioner and anybody else involved in the team, sometimes the pharmacist as well. I think that means a lot to the patient to know that there’s a team of people trying to work for his or her best benefit. [Male, physician]

Involving family members may help to increase patient understanding and lead to greater trust and higher enrollment of AA in clinical trials at Winship.

Getting the family involved is a good idea. [Male, physician]

Every single time when there’s a relapse that happens, we want the best option that is available out there for that patient… [if] involving a family member would yield more of an understanding [about clinical trials] that potentially is a time where I involve family for discussions. [Male, physician]

LIMITATIONS

This study has several limitations. The study was conducted at a single institution in the United States. Therefore, it may not be generalizable, but we hope that establishing a human connection can help underrepresented groups worldwide feel more comfortable participating in clinical trials. Participants were allowed to refuse to answer any question or end the surveys at any time which led to the collection of partial data for a few participants (1 participant skipped four questions from the DUREL scale, 1 participant ended the survey early before completing the DUREL, several participants skipped questions from the Lannin section as reported above). To account for this, results have been labeled with the number (n) of participants completing that section. Another limitation is participants who refused trials could not be interviewed, so we do not have the perspective of patients who decided not to enroll in clinical trials. It is also worth noting that the data from the validated scales used in the survey did not specify whether the participants had participated in a clinical trial.

DISCUSSION

When we used the TMR scale to assess trust, we discovered high trust in providers with 72% (44/61) of participants agreeing or strongly agreeing with the statement, “I completely trust doctors who do medical research.” Further, when asked, “How much do you trust your doctor?” 79% (48/61) of participants responded, “A great deal.” When we looked at the importance of clinical trial-related factors, we discovered that barriers did exist including high level of concern about side effects, trial-related costs, and the distance to the cancer center. We assessed religion using the DUREL scale and discovered that our participants were also very religious. Even though these barriers existed in this population, they did not deter participants from enrolling in a clinical trial. We suggest this may be due to the high level of trust and comfort with provider.

The high correlation between trust and human connection among our study population may suggest that a way to increase trust levels of AA patients is to increase human connection, or the feeling of being heard and valued by the provider. It was clear during our discussion with the providers that establishing a human connection with their patients is very important: respecting every patient, not treating anyone differently due to race or ethnicity but treating everyone as human beings worthy of respect and worthy of being listened to and engaged with. This may partially explain why Winship is more successful in enrolling African Americans.

Based on this discussion with Myeloma providers , we propose a roadmap to establishing a human connection (Figure 1). The categories in the roadmap (Rapport, Compassion, Communication, Trust, and Caregiver) are based on questions on the THC scale as shown in Figure 1. In our study, questions constituting the rapport component, listening to the patients’ concerns, being honest, and openminded, scored high in our study sample. The next step on our roadmap is being compassionate; seeing the patient as a whole person, caring about them, being concerned about their quality of life, and offering hope at every visit. Good communication is another key step in our roadmap. The provider should strive to be an excellent communicator with African American patients, clearly communicating explanations and suggestions and making the patient comfortable with asking questions. The lowest score on THC was to the question “How often does your doctor ask how your family members are coping with your illness?” 26% of patients responded, “Never” and 28% responded, “Sometimes.” A step to establishing a strong human connection is concerning yourself with the caregiver or family members of the patient. Training medical providers to incorporate this roadmap into their bedside manner and routine conversation with patients could be one way to establish human connection, promote therapeutic alliance, and increase trust, though future testing is needed.

Figure 1: Roadmap to Building a Human Connection.

*This figure is a depiction of our roadmap, based on the human connection scale and the discussion with Myeloma providers at Winship.

Conclusion

The results from this study suggest that the African American Myeloma clinical trial participants at Winship Cancer Institute have high levels of trust in their providers, successfully overcoming this major clinical trial barrier. Though our study population had high religiosity, and was concerned with side effects, costs, and distance, these barriers were apparently overcome by trust and human connection. Since having a human connection was correlated with trust, we propose that a roadmap to improve human connection may increase trust and trial enrollment in many settings. Our future steps include getting feedback on our roadmap. If the roadmap is considered useful, we hope to disseminate it, and thus help to increase clinical trial enrollment of underrepresented groups.