Introduction

Racial and ethnic minority populations in the USA continue to experience disproportionately higher cancer mortality rates [1], lower survival rates, and poorer quality of cancer treatment [2, 3]. African-Americans have the lowest 5-year survival rate for nearly all cancer sites [1]. Clinical trials (CTs) are a key step in the development of new cancer therapies, and access to CTs may reduce cancer health disparities [4–7]. Trial participants are usually monitored closely and may have better treatment outcomes than non-participants [8–10]. Yet, only 3–5 % of adult cancer patients participate in CTs, and among them, fewer than 10 % are racial and ethnic minorities [11]. African-American men are especially unlikely to enroll [12].

This underrepresentation limits the ability to generate hypotheses about differences in outcomes or to conduct meaningful subgroup analyses in order to detect between-group differences in response to treatment. This results in unfair distribution of the potential benefits and risks of trial participation and decreases generalizability of research findings [13]. Numerous factors contribute to decreased clinical trial participation among racial and ethnic minorities. Among these factors, medical distrust is the most frequently reported barrier [14, 15, 16]. Despite knowledge of the barriers to participation, there is scant information on the process of trial participation decision-making among African-American cancer patients. Among cancer patients who have been offered trial participation, key factors in the decision-making process and determinants of trial decision outcomes remain incompletely understood, when comparing those who have accepted participation (acceptors) and those who have declined (decliners).

The purpose of this study was to examine the processes and motivations of African-American cancer patients at the Johns Hopkins Sidney Kimmel Comprehensive Cancer Center (JH-SKCCC) for accepting or declining participation in cancer trials, and to elucidate the outcomes of these decisions. The study was conducted under the auspices of Enhancing Minority Participation in Clinical Trials (EMPaCT), a consortium of five comprehensive cancer centers, whose goal is to create and develop evidence-based strategies to enhance trial participation among underrepresented minorities. We used qualitative methods to discover underlying factors, processes, and outcomes related to the complex phenomenon of trial decision-making [17–19]. Based on the data and extant literature, we developed the Model of Cancer Clinical Trial Decision-making among African-American cancer patients.

Patients and Methods

In 2011, we conducted seven focus groups (32 patients = 14 acceptors; 18 decliners) with African-American patients who met the following eligibility criteria: (1) age ≥18 years; (2) diagnosed with lung, breast, prostate, or colorectal cancer ≤5 years, (3) received cancer treatment at JH-SKCCC, and (4) offered CT participation. Based on literature recommendations, we held separate groups for those who had agreed to participate in a trial (n =3) and those who declined (n =4), with 4–5 participants per group [20–22]. Patients were recruited through JH-SKCCC database reviews and clinic staff referrals available via a HIPAA waiver. In a letter signed by JH-SKCCC Director and our EMPaCT Principal Investigator, potential participants were invited to take part in the study or to opt out. Study staff then telephone-screened patients. Of the 187 patients screened for the study, 64 patients were eligible and invited to participate.

Because information on which patients are offered trial participation was not routinely reported or accessible across JH-SKCCC databases during the recruitment period, we used patient self-report to identify participants [23–25]. Information on patients who declined to participate when offered our study was collected over the phone during the screening eligibility process. Thus, we were able to compare them to study participants on demographics (such as name, date of birth), cancer diagnosis and treatment status, as well as clinical trial participation.

Participants provided written informed consent and completed a demographic questionnaire (age, gender, socioeconomic status, comorbid conditions, cancer diagnosis-related information) prior to the focus group interviews. We used a semi-structured discussion guide developed through team discussion, with minor refinements based on feedback from initial focus groups. The guide addressed the following topics: experiences with cancer diagnosis, attitudes towards trials, decision-making processes, reasons for participation/non-participation, trial barriers, and recommendations to improve trial participation. Sample questions included: “How did you make the decision to [participate/not participate]?” “What do you think African American patients and the people who care about them need to know about participating?”

An experienced facilitator/observer team with no clinical JH-SKCCC roles moderated all groups. Each group had at least one racially concordant facilitator. Groups lasted ∼90 min, were digitally recorded, and professionally transcribed verbatim. Participants received US$50 and transportation compensation. The study was approved by the Johns Hopkins School of Public Health Institutional Review Board.

Data Analysis

We conducted qualitative content and thematic analysis of the transcripts [26], using NVivo 9.0. At least two coders reviewed each transcript, using a previously described inductive approach to coding [27]. The team developed all concepts and themes, based on the final codes. The coding team (n = 4) met regularly to refine code definitions, discuss alternative interpretations, and select representative quotations for final review. Inter-rater reliability (kappa) among coders for all transcripts was >0.8. Study meetings were regularly convened to review findings and incorporate them into the model. Team members’ recommendations were incorporated into the final model.

Results

Nearly 50 % of patients who were screened and eligible participated in focus groups held at a local community-based primary care clinic. Participants were similar to non-participants with respect to age and gender. Overall, there were no clear differences in sociodemographic factors between acceptors and decliners or by gender except that acceptors were more likely to report incomes of ≤US$35,000 (Table 1). On average, male participants were older and were less likely than female participants to report incomes of <US$35,000 (Table 2). Nearly all had completed active treatment prior to study enrollment: 7/32 patients were <1 year from diagnosis, 14 within 1–2 years, 5 within 2–3 years, and 7 within 3–5 years of diagnosis. Among similar trials conducted at JH-SKCCC within our study timeframe (2008–2012), only data on trial acceptors were available. Our study population resembled African-American cancer center patient acceptors in median age (58 in our sample; 57 in the JH-SKCCC); however, compared to cancer center patients enrolled in trials, our participants were more likely to be female (57 % in our sample; 49 % in the JH-SKCCC).

Table 1 Demographic characteristics of study population by clinical trial status (n =32)
Full size table
Table 2 Demographic characteristics of study population by gender (n = 32)
Full size table

Data revealed multiple themes/subthemes (Table 3) related to trial decision-making. These were consistent among both acceptors and decliners. Participants described a process of decision balance: weighing information, influences, and options to arrive at the decision to participate or not. Participants’ information gathering, as well as intrapersonal and interpersonal influences, is a key component of decision processing. Trial decision-making was presented as either balanced or imbalanced and influenced decision outcomes, namely, satisfaction with decision-making and the presence or absence of decision-related regret. Some participants verbalized regret over what they perceived as a lost opportunity to convert the difficulties of their cancer experience into a communal good. Some decliners also expressed regret when they allowed decision partners to sway them against trial participation. Racial/ethnic disparities served as an implicit background for decision-making. Some participants described experiencing a health disparities dilemma: mistrust in research juxtaposed against a perceived obligation to participate, as a way to contribute to knowledge gains that may reduce treatment disparities. Health disparities dilemma arose as participants’ compared and contrasted participation with feelings of altruism in the face of research mistrust. Many participants expressed a personal responsibility or internal burden to participate in trials to ameliorate health and treatment disparities. However, the same participants who felt a personal responsibility to participate mentioned their mistrust in research within the same focus group discussion. Overall, achieving decision balance was a complex process encompassing perceived roles, responsibilities, research mistrust, and obligations to science or to family/friends.

Table 3 Themes from Model: Descriptions and exemplars
Full size table

We have modeled the themes and presented fuller explanations and exemplars for each theme/subtheme to accompany the model (Table 3). Participant-described barriers to trial participation were integrated into our model. Most have also been previously identified in the literature [15].

Discussion

In this study, we sought to understand the decision-making process and motivations of African-American cancer patients when making decisions about clinical trial participation. Using data from both acceptors and decliners, we were able to develop a focus group-derived Model of Cancer Clinical Trial Decision-making (Fig. 1), which depicted decision-making influences and outcomes. Broadly, information gathering, intrapersonal influences, and interpersonal influences were a part of decision processing. Further, participants’ ability to achieve decision balance after decision processing appeared to impact whether they experienced satisfaction or regret related to their CT participation decision. Decision-making themes identified in this study (Fig. 1) were both similar and different to those previously identified in the literature. Similar to previous studies, common reasons for non-participation included medical distrust, lack of knowledge about clinical trials, negative interactions with healthcare providers, cost, transportation, time, and trial side effects [16]. In contrast, individuals who chose to participate often did so because they felt support from family and friends and/or spiritual motivation to participate [16]. Family and friends were trusted sources of clinical trial information and acted as decision partners [28]. Personal relevance of the trial was also important and affected participants’ motivation to participate.

Fig. 1
figure 1

Application of the themes/subthemes used to develop the Model of Cancer Clinical Trial Decision-making

Full size image

While clinicians and researchers often acknowledge healthcare disparities, we were intrigued to discover that the existence of such disparities was factored into participants’ decision-making process. This theme, the health disparities dilemma, influenced and impacted the decision-making process for many participants (Fig. 1). As a part of the trial decision-making process, we believe the health disparities dilemma is a new contribution to the literature. As in other studies [29], participants who viewed trial participation as an opportunity to help others or ameliorate health and treatment disparities were more motivated to join clinical trials. Nevertheless, the health disparities dilemma was a source of tension and conflict surrounding trial decision-making and warrants additional exploration.

In our investigation of the decision-making process, we identified acceptors and decliners who experienced decision regret (Fig. 1). Specific reasons for decision regret, including those expressed by our participants, have not been well defined previously [19]. We found acceptors expressed regret when they felt they received inaccurate or inadequate information regarding what to expect from trial treatment. Decliners who initially feared the effect of cancer trials regretted not participating after they experienced the difficulties of standard cancer treatment (Fig. 1). Decision partners may facilitate decision satisfaction or contribute to regret, but there is little available literature describing the process by which they contribute to trial decision-making. Similarly, while altruism is well-established as a motivator for trial participation [29], perceived “loss” or regret related to altruism has not been emphasized. Because decision regret is closely associated with decision-making dissatisfaction [19], potential sources of regret warrant further investigation.

Our model (Fig. 1) integrates these concepts and may serve as a guide to develop strategies for recruitment of African-Americans into clinical trials. Strategies to improve communication about the risks and benefits of trial participation and financial feasibility have been shown to facilitate the enrollment of African-Americans in clinical trials; [14] they may also help reduce decision regret among trial participants. Furthermore, because decision partners played an important role in participants’ decisions to accept or decline trial participation, novel interventions that engage and support decision partners in assistive decision-making should be developed. Finally, the research team and approaches to clinical trial communication should be sensitive to potential participants’ ethical concerns and internal conflicts specific to existing health disparities [14].

Limitations and Strengths

A major strength of this study is that it focused exclusively on individuals who had been offered participation in a cancer clinical trial, thereby permitting qualitative comparisons of acceptors vs. decliners of participation. However, some limitations should also be considered. Because of our sampling limitations, our findings cannot be generalized to African-Americans, or to other underrepresented groups in clinical trials [30]. Although we achieved data saturation [31], model specificity may be limited. While we relied on patient self-report of trial offerings, prior studies among middle-aged adults have shown high validity of self-reported data compared to registries of other measures such as cancer screening [23] and history [24, 25]. We attempted to offset potential social desirability bias by including both acceptors and decliners in the study. Finally, participants’ narratives provided a helpful opportunity to corroborate whether patients were offered trial participation.

Conclusion

Our findings highlight the multiple and complex considerations surrounding trial participation from the perspectives of African-American cancer patients. Our model presents a framework for understanding the relationships between information gathering, intrapersonal perspectives, and interpersonal influences in trial decision-making. Our model may be used to further explore key factors that influence the process of decision-making and the achievement of positive decision outcomes rather than merely focusing on recruitment success. The model should be confirmed in additional African-American populations and other underrepresented populations, including members of other minority groups. Our model may be used to guide the development of interventions to improve the process of trial decision-making and thereby improve decision outcomes.