Introduction

Slow accrual to cancer clinical trials has been identified as the greatest impediment to winning the “war on cancer” [1]. It is estimated that less than 5% of all adult cancer patients enter clinical trials [2–4]. Even lower rates of enrollment have been reported in underserved populations [5]. Poor physician–patient communication has been identified as a key contributor to low accrual [6–8]. Although identified as a significant challenge [9], oncologists need to communicate effectively the purpose, technical aspects, and implications of trials to patients [6, 10, 11], and to assist eligible patients in decision making. Moreover, patients commonly fail to understand and correctly recall the information they receive about trials they are asked to consider [12, 13].

Informed consent requires that patients understand the information that has been provided and be competent to weigh possible courses of action. Informed consent is intended to protect patient autonomy and ensure that patients have an active role in making treatment decisions [14–17]. However, the complex language and excessive detail of some clinical trial information statements and consent forms may confuse rather than enhance patient understanding [18]. Unfortunately, patients frequently do not understand the rationale for trials and may not recall that they are receiving treatment in a clinical trial [12, 13]. Although cancer patients typically express high informational needs regarding available treatment options, their preferences for decisional involvement are variable [19–23]. Unfortunately, the literature suggests that physicians are not effectively ascertaining patients' informational or decision-making needs [24, 25]. To date, efforts to improve the quality of communication about informed consent have typically targeted improvement of physician communication, yet studies exploring the content of clinical trial discussions reveal that physicians remain dominant in these consultations [26–28]. Without appreciation of patients' informational and decision-making needs, it is not surprising that informed consent and patient satisfaction with treatment decision making remain suboptimal [24, 25]. Thus, there is a need to understand more fully the patient-level factors that contribute to high quality informed consent regarding clinical trial participation.

Various methods have been described in the literature to assist doctors and patients in discussing clinical trials and aiding decision making including clinical trial-specific decision aids, educational videos, and communication skills training for oncologists. However, rigorous research in this area is in its infancy, and the efficacy of these methodologies is yet to be proven. Question prompt lists (QPLs) [29–32] have demonstrated a significant promise in aiding doctor–patient communication by promoting an active patient participation in their general, surgical, and palliative oncology consultations. QPLs consist of a written sample of questions separated into content categories (e.g., diagnosis and prognosis). Previous studies have shown that QPLs are useful in the general oncology setting as an aid to patient question asking, particularly regarding prognosis. In addition, patient outcomes are improved when patients ask questions and their oncologist endorses questions [29–31]. Salient questions will undoubtedly vary across oncology contexts, and therefore, researchers have developed QPLs with different question content for palliative and surgical oncology contexts [33, 34]. Investigators [33, 34] have used focus groups with patients in their target populations to develop new questions specific to these areas and then pilot tested these new QPLs in the clinical setting. The results of pilot testing have shown QPLs to be acceptable, understandable, and valued by patients [33, 34]. The first author (RB) previously collaborated to develop and conduct initial testing of a QPL that targeted patient information needs specific to informed consent to cancer clinical trials in an Australian oncology setting. Due to cultural and health system differences, the results of this preliminary work needed validation in the US setting.

Thus, to develop further a clinical trial-specific QPL by gathering a broad range of patient views, we conducted focus groups with two distinct groups of cancer patients, those who had participated (trial experienced) and those who had not participated (trial naive) in a clinical trial. Our goal was to explore patient views about their clinical trial information needs and decision-making processes, and obtain feedback about the utility and completeness of the previously developed QPL. In this paper, we present data about the role of the QPL in patients' treatment decision making.

Methods

Participants

Oncologists

Six medical oncologists from three services—breast, lung, and genitourinary (GU) oncology (two from each service)—at a comprehensive cancer center in New York City agreed to use their outpatient clinics to recruit study participants.

Patients

Participants were adult cancer patients who were treated at the participating cancer center. They were recruited consecutively during follow-up visits. All participants were classified as either “trial experienced” or “trial naive.” Exclusion criteria consisted of: (a) age less than 18 years, (b) not English speaking, and (c) cognitive or physical impairment rendering the patient incapable of providing informed consent to participate in this study.

Procedure

Each of the six participating oncologists and the principal investigator (RB) randomly selected ten patients who had participated in a clinical trial and ten who had not participated in a clinical trial in 12 months prior to the study period, from each oncologist's patient lists. Our goal was to generate a list of 120 patients. As one oncologist had a small number of patients who participated in a trial, only five of his patients were selected. Thus, 115 eligible patients were identified. Of this total of 115, 55 had participated in a clinical trial, and 60 patients had not. From this available pool of 115 patients, we aimed to recruit between 24 and 32 patients (see sample size calculation below). The research assistant (RA) mailed a recruitment letter to patients that explained the study purpose, an invitation to attend a focus group, and a response form that patients could return via a prepaid envelope if they did not wish to be contacted about the study. The RA called those patients who had not returned a response form within 2 weeks of the initial mailing and discussed study participation. Interested patients gave verbal assent to participate and were scheduled for a focus group. Patients who attended the focus group were compensated $25 for their time and effort.

Focus group procedure

The focus groups were designed and conducted according to well-established methodologies [35, 36]. The investigators developed guides for each focus group containing a set of relevant topics and subsidiary questions. In the focus groups with patients who had participated in a clinical trial, we asked participants to answer questions retrospectively, i.e., as experienced participants. Patients who had not participated in a clinical trial were asked to answer questions prospectively, i.e., as trial naive. The focus group guide questions covered: needs for information when making trial decisions, the process of trial decision making, and the utility of the QPL (see Table 1 for sample questions).

Table 1 Focus group probes
Full size table

Immediately prior to the focus group, participants signed informed consent and completed a brief demographic questionnaire that included questions about age, gender, marital status, education level, occupation, nationality, first language, and whether the participant had previously participated in a clinical trial. Two authors, ES and RB, jointly moderated each focus group. The focus group guides were used as tools to moderate the discussions and allowed time for participants to discuss issues that were not included. Each focus group was audio-recorded and subsequently transcribed. The QPL was distributed during the focus group after a discussion of information needs and decision-making processes had occurred. The QPL is presented as Appendix 1.

Qualitative analysis plan

The research team reviewed and interpreted the data using thematic text analysis with an inductive, data-driven approach [37–39]. ATLAS.ti was used to manage the data coding [40]. Consistent with this method, each member independently developed codes to represent the underlying meaning of the text. The research team then met regularly to compare codes and achieve consensus. During these meetings, the team achieved consensus about code names and meanings and, through this process, developed a codebook. We ultimately created a codebook consisting of 90 descriptive codes. Our codes were descriptive in nature in that they each represented a description of distinct phenomena we identified through our coding process [39]. We identified a set of 27 codes that were most relevant to issues relating to clinical trial decision making and three codes within the category of assessment of the QPL description. We worked together in our consensus meetings to describe these recurrent themes and provide illustrative examples of each theme from our focus group data. Rigor in our qualitative analysis was derived from successive rounds of iterative consensus work among multiple team members who analyzed the focus group transcripts [41].

Sample size consideration

Based on established sample size recommendations for focus groups [35, 36] with the goal to achieve theoretical saturation (no new themes emerging), we aimed to recruit 11 patients each to four focus groups resulting in 44 potential participants. We estimated an attrition (no show) rate of approximately 20% leaving six to eight participants per group. Thus, our expected sample size was 24–32 participating patients.

The study received approval by the institutional review board at the participating institution.

Results

Participant characteristics

Twenty patients were recruited into four focus groups (two trial experienced and two trial naive). We recruited 9/55 trial-experienced patients and 11/60 trial-naive patients. An overview of our recruitment process of the 20 participants, including reasons for nonparticipation, is presented in Fig. 1. After four focus groups, no new themes emerged; therefore, as we had achieved our aim of theoretical saturation, recruitment ceased.

Fig. 1
figure 1

Sampling schema and flowchart of recruitment for question prompt list focus groups

Full size image

Patients were mostly female (70%), with an average age of 60 years. Eight patients had been diagnosed with lung cancer, eight with breast cancer, and four with genitourinary cancer. Most patients (12/20) had advanced disease. Of nine trial-experienced patients, six had participated in a phase II trial and three in a phase I trial. Five of the 55 trial-experienced patients had participated in a phase III study; however, none agreed to participate in the focus groups (see Table 2).

Table 2 Demographic and disease characteristics of patient sample
Full size table

Decision-making themes

Our qualitative analysis revealed that two factors predominantly influenced patient decision making regarding clinical trials—the outcome of patients' comparative assessment of perceived risks versus benefits of a trial and the level of trust patients had in their doctors' recommendation regarding whether a trial would be an appropriate medical option for them. The severity of a patient's disease influenced trial decision making only for trial-naive patients. We then compared the similarities and differences between trial-experienced and trial-naive patients for each of these decision-making factors.

Assessment of perceived risks versus benefits of a trial

Overall, trial-experienced patients were more likely to regard clinical trials more positively than trial-naive patients, who tended to focus more on the potential risks from trials. Many trial-experienced patients expressed faith in the survival benefits of the trial treatment. These patients expressed a desire to “do whatever it is to survive” and were minimally concerned with potential negative side effects, financial costs, or the possibility that there might be a potential conflict of interest being industry-sponsored.

In contrast, trial-naive patients were more likely to have negative perceptions about clinical trials and would not be receptive to a clinical trial without concrete evidence about personal survival benefit. One trial-naive patient stated, “I don’t know if I would want a clinical study unless I really knew that maybe something good would come out of it. Otherwise I would feel as if it was just something they just wanted some data on.” Another trial-naive patient declined to participate in two trials that she was offered, as she was not convinced that the trial would be the best treatment option for her.

“So when I was asked to do two different clinical trials. And honestly my big question was ‘what’s in it for me…I didn’t want to be Saint Teresa or anybody. I really didn’t. I just said no if this isn’t like the best thing that’s gonna have a good outcome for me.’”

The primacy of trust in physician recommendations

Trust in their doctor's recommendation regarding the benefit of a clinical trial was also central to trial decision making for both trial-experienced and trial-naive patients. Trial-experienced patients sought their doctors' recommendation through two primary ways. First, many patients directly asked for their doctor's expert recommendation and made their trial decision accordingly. Expert opinion was especially vital for patients who expressed feeling overwhelmed with the decision. One trial-experienced patient stated, “So I was trying to balance the risks and she’s the expert and I’m not…plus you’re so overwhelmed that you just want somebody to tell you what to do.” Second, some trial-experienced patients reported asking for a recommendation indirectly by using proxy questions. An example was asking whether the physician would recommend the trial to a close family member. Some trial-experienced patients reported that receiving a positive or negative answer to this proxy question could be a deal breaker in their decision making, as noted below.

“Because imagine what that would say if your doctor sat there and said no. If my wife were in your shoes I would not allow her in this trial…I mean I don’t know that I’d give it another thought…. It’s almost like that question ‘is this worth me going forward with?’”

Trial-naive patients indicated that, if they developed trust in their doctors' medical judgment, they would be likely to follow the doctors' guidance. One trial-naive patient reported, “I came here for only the best…If he came to me, my doctor, and said let’s do a clinical trial, I would do it. I trust him.”

Other trial-naive patients regarded the manner in which their physician presents the trial opportunity to them as central in engendering trust with the doctor's recommendation. One patient stated that she would feel comfortable with her doctor's judgment if the physician openly shared trial information and answered questions completely. Trial-naive patients differed in their views about the physician introducing a trial as a treatment option. Some patients would view their doctor as “looking out for” them if they suggested they consider a trial. On the other hand, one patient disliked that her own doctor presented her with a trial, as during standard treatment discussions, she had followed the doctor's treatment recommendation and been a “loyal patient.” The trial decision-making process challenged her to make up her own mind about participation.

Severity of disease

As trial-naive patients were asked to discuss the factors that would influence their decision making if they were presented with the option to join a clinical trial, many stated that, if their disease were at a later stage or if they had exhausted other treatment options, they would be receptive to join a trial. Several of these trial-naive patients stated that they would be open to a trial if a trial was the last resort to improve their disease, or if they had already “gone through the arsenal” of other treatments. Severity of disease was not an important trial decision-making factor for trial-experienced patients.

Utility of the QPL for decision making

Patients viewed the QPL as a valuable tool to support their decision making regarding clinical trials. Patients divided the potential utility of the QPL into three categories: (a) preclinical trial consultation, (b) during a consultation, and (c) postconsultation.

Preconsultation benefits included the ability of the QPL to save patient's time, in that they would not need to prepare questions on their own. The QPL presents patients with a menu of potential issues to discuss with a physician and thus prepares patients for a medical consultation. One trial-naive patient said of the questions in the QPL, “I found the questions themselves informed my thinking…the way they were phrased made me realize something or gave me information. And I would be very pleased to get that.” Some patients found the questions educational as they include trial-related terminology (e.g., randomized, blinded, response rate) that patients may not have known. One trial-naive patients noted that “the fact of even using the word, randomized, in here. It tells me something that I wouldn’t…may not have known.”

During consultation, benefits included enhanced the ability of patients to have meaningful discussions with physicians and helped patients use the limited time they had with their doctors to their best advantage. One trial-naive patient commented on the time aspect saying “I like the idea that if I had this (the QPL) in advance, I could just narrow it down to what’s really important to me and focus her and my energy because you do have a limited time period. So I think I would like this for everything.”

Patients also reported postconsultation benefits including that the QPL is a concrete takeaway that patients can review and be better prepared for a follow-up discussion. One trial-naive patient reported: “So if you had a piece of paper that outlines what it (the trial) is, and then you had these questions, then you could call back at a pre-determined time to talk to someone about it. And then it would help you make your decision, but I think these questions are very thoughtful and very good.” In addition, patient mentioned that seeing the question list might help them to remember the details of their clinical trial discussions with doctors.

Discussion

The aim of this study was to further develop a QPL for clinical trial discussions and assess its perceived utility to aid decision making using focus groups with trial-experienced and trial-naive patients. Cancer patients find it difficult to make decisions about clinical trials [42, 43]. Consistent with prior work, our results show that clinical trial decision making is primarily influenced by a trade-off between hope for personal treatment benefit and the risks of side effects [44], and by trust in the physician [45]. Most studies show that patient preferences for decision making vary with most well patients preferring a collaborative role [46–48], yet many sick patients prefer the doctor to make decisions on their behalf [49]. Most of our sample had stage 3 or 4 cancers and expressed a desire to make a trade-off guided by a trusted clinician who would steer them to an optimal treatment decision. Although shared decision making about treatment options is a well-accepted approach [46, 50, 51], paternalistic decision making was desired by our sample. Trial-naive patients indicated that they would choose a clinical trial only when all other options were exhausted. This is likely explained in part by these patients' negative perceptions of trials in comparison with the trial-experienced patients. It may also be that trial-naive patients perceived that they had more available options than the experienced patients who opted for the trial.

For patients with advanced disease, decision making about whether to enroll in a clinical trial is complex. Patients' decision-making processes may change as their illness progresses. Patients in our sample were enthusiastic about the QPL. Trial-experienced patients endorsed that it would have been helpful in their trial decision making. Trial-naive patients expressed a desire to use such a tool if they were faced with a choice about whether to join a trial. Much of the previous QPL, research has focused on the usefulness of a QPL within the cancer consultation as an aid to activate patient question asking [29, 30, 33, 34]. Focus group patients also noted the usefulness of the QPL outside of the consultation as a means of preparing for the consultation and as an education aid. Moreover, they expressed that the QPL could provoke patients to consider novel issues, gain salient information, be better informed for future trial discussions, and help make a trial decision. Future research could usefully measure the efficacy of the QPL to achieve these potential benefits.

These promising results suggest that a clinical trial QPL would be a valuable aid for patients facing difficult treatment decisions. Future research is underway to test the utility of the QPL in actual consultations, and to determine its impact on question asking and consultation communication. We will also assess the potential impact on decisions to accrue to a clinical trial.