1577

Barriers and Facilitators to Enrollment in Cancer Clinical Trials

Qualitative Study of the Perspectives of Clinical Research Associates

Eva Grunfeld, M.D., D.Phil. Louise Zitzelsberger, Ph.D. Marjorie Coristine, M.A. Faye Aspelund, C.C.R.A.
Ottawa Regional Cancer Center, Ottawa Health Research Institute, Ottawa, Ontario, Canada.

BACKGROUND. The literature continues to report low rates of accrual to cancer

clinical trials. Previous studies have examined principally physician-related or patient-related barriers. Clinical research associates (CRAs) have a unique perspective on enrollment that has been explored very little. This study sought the views of CRAs on barriers and facilitators to accrual. METHODS. Focus groups were held at six of eight tertiary cancer centers in Ontario, Canada. Audiotapes of sessions were transcribed and subjected to content analysis by two of the authors. Emergent themes were identied. These themes are illustrated by representative quotes taken from the transcripts. RESULTS. Factors that acted as barriers or facilitators were classied into physician-related, patient-related, or system-related factors. CRAs identied physician attitudes regarding patient participation as the principal physician-related barrier. Barriers, facilitators, and modifying factors that were related to patient involvement were discussed by CRAs. Patients seemed more knowledgeable about trials than in the past and were willing to participate. System factors were considered to have the greatest impact on the ability to accrue. CRAs identied increasing trial and pharmaceutical demands coupled with tight trial time lines. Time was seen as a diminishing resource. Greater demands not only affect specic clinical trial accrual but also affect general support for trials in the cancer center and hospital. CONCLUSIONS. The impact of greater demands in a climate of decreasing health care resources is perceived by CRAs as having a negative affect on accrual. Consequently, the important process of translating potentially benecial basic research ndings into clinical practice is slowed. Cancer 2002;95:1577 83. 2002 American Cancer Society. DOI 10.1002/cncr.10862

See related editorial on pages 1397 400, this issue. Dr. Grunfeld is a career scientist of the Ontario Ministry of Health and Long-Term Care. The conclusions are those of the authors, and no endorsement by the Ministry is intended or should be inferred. Address for reprints: Eva Grunfeld, M.D., D.Phil., Ottawa Regional Cancer Center, 503 Smyth Road, Ottawa, Ontario K1H 1C4, Canada; Fax: 613-247-3511; E-mail: eva.grunfeld@ orcc.on.ca Received January 16, 2002; revision received May 7, 2002; accepted May 13, 2002. 2002 American Cancer Society

KEYWORDS: patient participation, clinician participation, clinical trials, barriers to participation, neoplasms.

linical trials are an important step in helping translate potentially benecial basic research ndings into clinical practice. However, the literature continues to demonstrate low rates of patient accrual to trials1,2 and, as a result, a slow rate of research transfer. There are a number of factors that affect patient accrual to clinical trials. Previous studies principally have examined physician and patient factors.15 A smaller number of articles looked at the impact of trial requirements6,7 or organizational factors.8 The physicians role traditionally has been seen as that of a gatekeeper, because the decision to offer participation in a trial to a patient is made initially by the patients treating physician. Physicians have been identied as one of

1578

CANCER October 1, 2002 / Volume 95 / Number 7

the principal barriers to accrual.1,2,8 In addition, patients also list many reasons for their refusal to participate in trials.9 11 Although the physician, as gatekeeper, is responsible for the initial presentation of a trial, it is the clinical research associate (CRA) who spends considerable time with patients, explaining the details of the trial (e.g., objectives, potential harms, and procedures) and obtaining informed consent. In addition to their involvement in accrual, CRAs are responsible for the day-to-day administration of trial procedures. In this capacity, they are involved directly in all facets of trial implementation. Therefore, they have a unique perspective on the multiple factors that may have an impact on accrual. The perspective of CRAs rarely has been explored in the literature. The objective of this study was to identify barriers and facilitators to the accrual of patients to cancer clinical trials by learning the views of CRAs on this subject.

MATERIALS AND METHODS

Focus groups with CRAs were held at six of eight tertiary cancer treatment centers in the province of Ontario, Canada. These centers serve a population of over 12 million and are the principal providers of tertiary cancer treatment and the principal accruers to cancer clinical trials in the province. Centers were selected for participation in the study to represent a range of perspectives (e.g., urban vs. rural, small vs. large, north vs. south). The size of the focus groups depended on the number and availability of CRAs in each center: One group had 1 participant, and the remaining groups had between 4 and 7 participants, for a total of 29 participants. Most participants were CRAs; ve participants were data managers. Although CRAs and data managers perform similar functions, CRAs have completed the certication requirements of the Society of Clinical Research Associates. Both undergo extensive initial and on-going training and implement procedures for each trial for which they are responsible. Before conducting the focus groups, a list of topics was identied by consulting the literature. A semistructured interview was developed that addressed perceived system, physician, and patient factors that may have an affect on study accrual. The same facilitator ran all of the groups to ensure consistency. Cofacilitators were responsible for operating the recording equipment, ensuring that consent forms were signed, and ensuring that all participants had an opportunity to speak during the session. Participation in the study was voluntary. Focus groups lasted between 1.00 hour and 1.25 hours. Focus groups were audiotaped, and the tapes were tran-

scribed by individuals who were employed outside the cancer center. Personal names and center references were replaced on the transcripts by asterisks to ensure condentiality. Audiotapes, transcripts, and consent forms were kept in locked cabinets; consent forms were stored separate from tapes and transcripts. Ethical approval for the study was obtained from the Ottawa Hospital Research Ethics Board. Each focus group member signed a consent form for participation and a separate consent for the use of quotes in publications emerging from the study. After each focus group, the research team reviewed their notes taken during the session. Emergent topics were discussed with subsequent focus groups. Focus groups continued to be held until the research team noted a repetition of themes. Data saturation is considered to be a stopping point for sampling in qualitative research.12,13 After the transcription of audiotapes, two researchers independently read each transcript. Using content analysis, sections of the transcript were identied as related to the topic of interest and were coded with a descriptive title. Codes were then categorized into a set of preliminary main themes and subthemes. The two researchers then compared their list of themes. Through a process of revisiting the transcripts and discussion, researchers created one comprehensive list of main themes and subthemes. Quotes that best represented themes were identied. Below, the text in brackets at the end of each quote refers to the focus group number (FG) and the page number (p) of the quote in the transcript.

RESULTS
Factors that acted as barriers or facilitators to accrual, as described by CRAs, were classied as physicianrelated, patient-related, and system-related factors.

Physician Factors
Physicians were considered important barriers to accrual, because patient involvement depends initially on their decision whether to recruit. However, compared with other factors, CRAs did not identify physician-related factors for or against accrual as prominent. The factors that were identied all related to physician attitudes toward the suitability of a patient for a specic trial, despite the patient meeting the eligibility criteria. Reasons for unsuitability included anticipated logistic problems for patients (e.g., amount of travel), concerns regarding long-term survival for elderly patients, concerns regarding patient comprehension of trial information, or physician beliefs regarding patient preferences.

Barriers to Clinical Trial Accrual/Grunfeld et al.

1579

They [physicians] have to make the decision [putting a patient on study] in regard to mental capacity, and thats their decision to makeif the patient can handle the stress and process the information. [FG5, p 5] Although physicians may agree to take part in a study, either as investigators or as coinvestigators, CRAs reported that at times they actually did not believe in a specic trial. Implicit physician attitudes toward a trial can inuence patient decision making. Physician biaswhether the physician buys into the study or not. How the physician approaches the patient about the study. Even body language or just even one word or a description between standard treatments and protocol treatments that would indicate that one would be less time consuming or easier for the patient or whatever the bias may be. [FG2, p 5]

physicians expertise, and their disease. CRAs stated that some patients felt they were considered guinea pigs and did not want to be experimented on in a trial. It was believed that a proportion of possible candidates excluded themselves from involvement by accepting without question their physicians or surgeons suggestion that they should not participate. Finally, CRAs discussed how patient acceptance of their disease inuenced decisions regarding involvement. if you have well patients, like prostate guys who are on a Phase II study; they just dont want to be bothered with all the extra attention. They want to live their life and dont even want to come to terms with the fact that they have prostate cancer. They just want to keep on going and everythings all right. Sometimes the trial is just a reminder, and its too much of a waste of time. [FG2, p 6]

Patient Factors
CRAs spend considerable time with patients in the early stages of decision making regarding a trial and, thus, are aware of the reasons why patients do or do not participate. CRAs identied a number of barriers, facilitators, and modifying factors that were related to participation. In general, patients were viewed as knowledgeable about trials and interested in participation.

Facilitators
According to CRAs, patients choose to participate in trials for a number of reasons. Some patients participate because they believe that a trial will be benecial for them. For example, it was believed that patients who had tried and failed with standard treatment participated because, for them, something was better than nothing. Some patients hoped that involvement would result in a potential cure. In addition, patients perceived that, with participation, they would receive extra monitoring and care from physicians and nurses. They have a perception that, when they are on a trial, they are cared for better, and its literally because we [CRAs] have the time as opposed to the other primary care nurses. [FG2, p 3] Some patients also became involved for the benet of future generations, whereas other patients wanted to please their physicians if there was a perception that he or she supported the trial. Two focus groups mentioned the support of pharmaceutical companies in helping facilitate involvement for patients through compensation for extra expenses incurred, such as parking or child care, or through providing laboratory kits for patients to take on vacation. The other thing is, some of the patientswant to go south and they want to be able to take their treatment with them. Some of the studies have kits they can take with them [south]. [FG4, p 10]

Barriers
CRAs identied a number of patient-related barriers. These were either logistic or attitudinal. Logistic factors were those that added stress to already burdened patients. Certain patients rejected involvement due to the burden of extra tests and questionnaires or due to the potentially greater toxicity of an investigational treatment compared with standard treatment. Residing a distance from the cancer center excluded involvement in some trials that required patients be within close proximity. Care giving responsibilities precluded participation for some patients. Often times, an ill family member will be a deterrent to a patient going on a study because theyll say, I just dont have time to deal with this. My husbands not well and I have his full care at homeTheyre giving of themselves to help someone else. Ive actually had that happen a number of times where patients have said, I would really like to participate in this but because of my home situation, Im unable to. [FG3, p 8] Attitudinal factors that were barriers to participation were related to patient views toward trials, their

Modiers
Modifying factors that were identied by CRAs were factors that may act in a dual sense either encouraging or discouraging patient involvement. For exam-

1580

CANCER October 1, 2002 / Volume 95 / Number 7

ple, CRAs mentioned that family members may push a patient either into or away from a trial. There was a concern raised in one focus group that choices were being made on behalf of the familys wishes rather than the patients wishes. I had an incident the other day where the patient is here surrounded by a row of family around him, and all the family is saying, Yes, he is going to go on the trial. Yes, he is going to ght. And the patient is saying, Yeah, whatever. Was it really his decision, was he truly informed? Debatable. [FG2, p 7] If a patients rst language was other than English or French, this was cited as another factor that may either discourage participation or have no effect. Staff at three of the centers stated that language was not a barrier. Differences were accommodated through family members, hospital-based interpreters, physicians who had the same rst language as the patients, and through multilingual materials produced by pharmaceutical companies. In one center, however, language had been a barrier to participation. Another factor that CRAs believed inuenced accrual was media coverage. It was perceived that negative effects dropped the accrual of some studies to almost zero. In other cases, positive coverage caused patients to call into centers asking about specic trials. If, for example, the IND [investigational new drug] got to the newspaper that these trials were being done, we get ooded with phone calls from all over the world. Well come, you give us the drugs. So the media has a huge impact on accrual. Drug companies involved in trials that are going slowly, thats a huge way of upping their accrual. Leak it to the media, and your sites will be ooded. [FG2, p 7]

to know about the study, and they give you a history, wanting to know if they would be eligible for the study. Theres a lot more knowledge out there in the community, and theyre coming in and saying, This is what I want. [FG3, p 5] Linked to increased knowledge is interest in trial participation. CRAs at one center mentioned that their problem was not in nding patients who were interested in participation but, rather, having too many patients who wanted to go on a study. Another center mentioned that patients wanted the option of being offered a study, and that few have refused. Although patient refusal has been considered second to physician nonrecruitment as responsible for low accrual rates, the CRAs who participated in this study did not think that patients limited accrual; rather, they felt that system factors were responsible. The patients are interested in studies. They want information. They are demanding new treatments, new ways of doing things, but we just dont have the staff, time, or money to give it to them. [FG5, p 1]

System Factors
CRAs described a climate of increasing trial and pharmaceutical requirements coupled with tight time lines. Diminishing resources, in particular time to discuss participation, were identied as potential factors in preselection bias. More demands during a period of decreasing health care resources affect not only clinical trials but also affect support for trials in the cancer centers and hospitals. Thus, system-related factors, more than physician-related and patient-related factors, were emphasized the most by CRAs as having an adverse effect on accrual.

Trial and pharmaceutical company requirements

CRAs at all centers mentioned the growing complexity of trials and accompanying requirements. Although there is a sense that regulations are put in place ultimately to safeguard patients, there also is a perception that requirements are becoming excessive. I think our trials have become more intense in the last few years. There are more requirements from people. Theyre looking for more data, more information, more scans, more X-rays, more review before they go on, more quality of life, which is a very important aspect. So, in that way, I think the trials themselves are getting larger and requiring more time and work and are becoming more intense. [FG4, p 4] Guidelines for good practice are set up for a very good reason, but those expectations fall on us

Patient knowledge of and interest in trials

CRAs commented that patients are much more knowledgeable than they were in the past about clinical trials. Media coverage and exposure to the Internet has increased patient awareness. Certainly patients are more knowledgeable. They come in with the copy off the Internet saying, This is the drug I want, why cant I have it?The information is out there for them, and they are accessing that information. I would say, I would average maybe one or two calls a month from outside the areaIve had calls from California, all over, saying, Ive heard of this drug, are you doing this study? Somebody told me you had this drug available. Ill travel up there. They also want

Barriers to Clinical Trial Accrual/Grunfeld et al.

1581

ultimatelyTheyre government regulations, actually they are international now, so [that] all research is conducted in the best way possible to help the patient. So the expectations of each study are much more demanding; absolutely everything has to be documented. [We are] even involved in monitoring patients blood counts to make sure the doses are modied correctly. These things are falling more and more on us. [FG1, p 3] Not only did trial requirements seem greater, all focus groups commented that the demands of pharmaceutical companies have increased over the past years. In some cases, demands for documentation were seen as unnecessary and excessive. Certain drug companies seem to do volume. Volume is impressive as opposed to quality. They want to know anything even if it has nothing to do with medicationlike they vomited one day because they ate bad cabbage rolls. They want when it started, when it stopped, the drugs they took for that event, and whether it was drug related. No, its not drug related, so why should we capture it? That being documented lls volumes of paper. It creates more work for us and doesnt help the treatment. [FG1, p 11] CRAs perceived that these requirements are based in part on a lack of expertise by pharmaceutical companies themselves and their staff who liaise with CRAs. Some companies will have clinical oncology experience, but more and more drug companies are getting into the oncology eld, so the people writing the protocols dont even have the background [FG6, p 15] One of the many problems with drug companies that you may know or you may not, they often hire people who have no experience. That person may have a lot of power, and they come in telling you what to do, and a lot of the time theyre wrong; they dont know what they are doing. [FG6, p 14]

you have 2 weeks to do all your planning, get your MRIs [magnetic resonance images]youre pulling strings to get it organized, and people are stressed to the max. [FG2, p 14] Overall, physicians seem busier than they were in the past. Decisions by physicians and CRAs to accrue patients or not to accrue patients were weighed against other work demands and available time. So, if youre screening somebody, you may be a little bit more selective and perhaps something like [their] psychosocial historyIm not even going to approach that patient. I dont have time to spend over 3 hours on the consent given the history of this patient when it normally can take me an hour. That can be a bias, because you dont have the time. [FG2, p 12] I dont think we really bias, or I like to think that we dont, because of the age or race or if they can speak English or not. I think its more an issue of time. [FG2, p 13] Lack of physician/CRA time is a symptom of cancer center-wide constraints. Accrual is affected because clinics are busier, and patients wait longer in the clinic for their appointments. We keep saying were so patient focused, and I believe that we really are, but it doesnt appear to be that way a lot of the times, I agreeBy the time were invited down to see the patient, and theyve been waiting hourstheyre exhausted. Sometimes, you have to just do the minimum of your jobnot the kind of job youd like to dolike giving informed consent, which is our rst contact with the patient, because theyre just too tired and absorbing nothing and are totally overwhelmed at that point. [FG1, p 2] The consequences of greater requirements due to trial complexity, pharmaceutical expectations, and lack of oncologic expertise by pharmaceutical companies have an impact not only on clinical trials staff but also on the whole system. Centers already are limited in the number of studies they can host, and extra requirements limit that number even more. There are a lot of good studies out therea lot are really interestingthat peak the interest of physicians, but it comes down to resources, like anything else. We have, I dont know how many studies on hold, until the staff can handle [them], that the center can accommodate. [FG4, p 5]

Tight time lines versus time as a diminishing resource

The time lines related to study requirements appeared to be getting shorter according to the majority of CRAs. Time was seen as a resource already in short supply; staff felt they had more work to do in less time. A lot of the requirements are that the patient has to be on the study within 5 weeks of surgery. They get referred 2 weeks later, they get treatment, and then

1582

CANCER October 1, 2002 / Volume 95 / Number 7

DISCUSSION
CRAs are involved in all aspects of running clinical trials, from completing regulatory documents and recruitment through to study completion. They interact with a number of different people, including physicians, patients, nurses, and pharmaceutical representatives, all of whom may have an impact on their ability to perform their duties. Accrual is one of the most important aspects of their job. CRAs reported a number of physician-related, patient-related, and system-related factors that had an impact on patient accrual. The roles played by physicians in encouraging or discouraging patient participation were not emphasized by CRAs in this study; however, a number of physician-related factors for or against patient accrual have been cited in the literature. They include concerns over a disruption in the patient-physician relationship, a dislike of discussions over consent, logistic constraints, clinician-researcher role conict, and issues related to patient suitability.4,8,14 In this study, CRAs identied physician attitudes as the principal physician-related barrier to accrual. CRAs identied numerous patient-related barriers to participation in trials. Because CRAs are very involved in the recruitment process, they have many opportunities to determine why patients are for or against participation. Reasons similar to those listed by CRAs also have been reported in the literature.3,4 One nding that has emerged in recent literature9,11,15 concerning patient accrual is that patients are more knowledgeable than they were in the past about trials, and they report a willingness to participate. CRAs reported that many patients ask to be involved in specic trials. Media coverage of trials and information on the Internet were linked to increased knowledge. Patients were not seen as contributing to factors that limited accrual. However, CRAs believed that system factors had the greatest impact on their ability to accrue. CRAs described a climate of increasing demands from regulatory bodies and pharmaceutical companies coupled with tighter time lines. Increased demands, however, have not been matched by increased resources. Rather, CRAs commented on how strains on the system as a whole have had an impact on clinical trials accrualthrough the necessity of balancing time to recruit with conicting demands and because of busier clinics, resulting in patients who are too tired to absorb trial information. Therefore, solutions need to be generated, not at the level of physicians and patients (who CRAs did not believe contributed limiting accrual) but, rather, in the involvement of CRAs in research design to ensure that it

is as streamlined as possible and in the commitment to research, as expressed through the allocation of resources to support trials. Physicians may choose to select among trials and may decide not to participate in trials that seem overly burdensome, as noted in previous studies.7 CRAs need to be involved in trial design from the earliest stages of development. Because they understand the day-today work of trials, they can provide input on both form design and trial procedures, helping to ensure that requirements are relevant and reasonable. A reallocation of resources that would see other personnel take over some of the routine administrative aspects of trial management would free up CRAs to concentrate on recruitment. Perhaps not only accrual itself but also other aspects of the recruitment process may benet. In a survey of adult patients with cancer (n 287 patients) on their understanding of informed consent related to involvement in a trial, higher knowledge scores were associated with a number of factors, including having a nurse present during the consent discussion.16 Due to their involvement in all aspects of trial conduct, CRAs present a unique perspective on barriers related to accrual. Due to the qualitative methods used in this study, the specic ndings presented may not be generalizable. However, they demonstrate the impact of increasing demand and diminished resources experienced by the health care system in general on accrual to cancer clinical trials. To improve accrual, future research should focus on determining characteristics of efcient trial operation and trial designwithin the restrictions of regulatory requirementsto ensure translation of potentially benecial basic research ndings into clinical practice.

REFERENCES
1. Falloweld L, Ratcliffe D, Souhami RL. Clinicians attitudes to clinical trials of cancer therapy. Eur J Cancer. 1997;33: 22212229. Siminoff LA, Zhang A, Colabianchi N, Saunders Sturm CM, Shen Q. Factors that predict the referral of breast cancer patients onto clinical trials by their surgeons and medical oncologists. J Clin Oncol. 2000;18:12031211. Swanson GM, Ward A. Recruiting minorities into clinical trials: toward a participant-friendly system. J Natl Cancer Inst. 1995;87:17471759. Ross S, Grant A, Counsell C, Gillespie W, Russell I, Prescott RJ. Barriers to participation in randomised controlled trials: a systematic review. J Clin Epidemiol. 1999;52:11431156. Primo L, Higdon R, Lim N, et al. Prospective evaluation of cancer clinical trial accrual patterns: identifying potential barriers to enrollment. J Clin Oncol. 2001;19:1728 1733. Cook Gotay C. Accrual to cancer clinical trials: directions from the research literature. Soc Sci Med. 1991;33:569 577. Foley J, Moertel CG. Improving accrual into cancer clinical trials. J Cancer Educ. 1991;6:165173.

2.

3.

4.

5.

6. 7.

Barriers to Clinical Trial Accrual/Grunfeld et al.

Winn R. Obstacles to the accrual of patients to clinical trials in the community setting. Semin Oncol. 1994;21: 112117. 9. Slevin M, Mossman J, Bowling A, et al. Volunteers or victims: patients views of randomised cancer clinical trials. Br J Cancer. 1995;71:1270 1274. 10. Tangrea J. Patient participation and compliance in cancer chemoprevention trials: issues and concerns. Proc Soc Exp Biol Med. 1997;216:260 265. 11. Ellis P, Butow P, Tattersall MHN, Stewart M, Houssami N. Randomized clinical trials in oncology: understanding and attitudes predict willingness to participate. J Clin Oncol. 2001;19:3554 3561. 8.

1583

12. Rubin H, Rubin I. Qualitative interviewing: the art of hearing data. Thousand Oaks, CA: Sage Publications, 1995. 13. Glaser B, Strauss A. Discovery of grounded theory. Chicago: Aldine, 1967. 14. Johansen M, Mayer KH, Hoover H. Obstacles to implementing cancer clinical trials. Semin Oncol Nurs. 1991;7:260 267. 15. Hancock BW, Aitken M, Radstone C, Hudson GV. Why dont cancer patients get entered into clinical trials? Experience of the Shefeld Lymphoma Groups collaboration in British National Lymphoma Investigation Studies. BMJ. 1997;314:3637. 16. Joffe S, Francis Cook E, Cleary P, Clark J, Weeks J. Quality of informed consent in cancer clinical trials: a cross-sectional survey. Lancet. 2001;358:17721777.