“The one constant in trial recruitment is it will always change, and you must adapt.” —Nina Bickell, Mount Sinai School of Medicine
The case studies in this chapter, which describe patient recruitment challenges for a range of medical conditions, illustrate the role of stakeholders’ perspectives in shaping their engagement with the clinical trials enterprise. Several speakers noted that the stakeholders involved in clinical trials have unique points of view; for example, researchers are focused on answering the clinical question at hand as well as on their own career development, institutions guard their reputations and their resources, and referring physicians have multiple concerns that could include losing control of their patients’ care, as well as, in some cases, professional liability. Many workshop presenters emphasized that patients worry about a great number of issues, their health being only one of them, and every aspect of a trial protocol that makes it harder to understand, less relevant to them, and less convenient diminishes the likelihood of participation.
From an institutional and investigator perspective, not meeting enrollment numbers in a timely way can cause a clinical trial to lose momentum and can lead to other negative conditions such as investigator burnout. In the worst cases, low enrollment can cause a trial to be
1 This section is based on the presentation by Robert Michler, Surgeon-in-Chief, Professor and Chairman, and Director, Center for Heart and Vascular Care at the Montefiore Medical Center/Albert Einstein College of Medicine.
abandoned—a costly outcome that can harm the credibility of individual investigators and their institutions. The public and private organizations that fund trials look to a researcher’s and an institution’s prior history when making grant awards, and they take notice when investigators fail to meet their anticipated enrollment goals. From the investigator’s perspective, then, patient recruitment is a significant responsibility and not doing it effectively may lead to frustration, institutional concern, and even embarrassment.
Three major clinical trials in which Robert Michler, Surgeon-in-Chief, Professor and Chairman, and Director, Center for Heart and Vascular Care at the Montefiore Medical Center/Albert Einstein College of Medicine, actively participated all involved patients with serious heart disease for which surgery was a treatment alternative: Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure (REMATCH) trial, Surgical Treatment for Ischemic Heart Failure (STICH) trial, and NHLBI–funded Cardiothoracic Surgical Trials Network (CTSN).
REMATCH evaluated whether a left ventricular assist device (LVAD)2 would reduce mortality compared to optimal medical management.3 Trial participants, who were primarily Medicare patients, needed but were ineligible for heart transplants. REMATCH investigators hoped to show that an LVAD could reduce 2-year mortality by one-third to one-half, compared to optimal medical management. The investigators planned to enroll 140 patients over eight calendar quarters. After eight quarters, no more than 80 patients had been enrolled. Michler noted that enrolling centers could not meet the expense of this complex trial and the federal government would not provide additional funding to complete the study. Additional funding from the trial’s industry sponsor helped the trial investigators complete the enrollment of 128 patients over 13 calendar quarters—suggesting the high cost and length of time it can take to achieve enrollment goals.
STICH, the largest surgical trial ever conducted, involving 2,135 patients enrolled in 127 sites across 26 countries, compared coronary bypass surgery with and without ventricular reconstruction of the left ventricle to optimal medical therapy in patients with significant heart
2 An LVAD is a surgically implanted pump that helps pump blood from the heart’s lower left chamber (the left ventricle) to the rest of the body.
3 Optimal medical management included adherence to guidelines developed by a medical committee with the goals of optimizing organ performance and minimizing symptoms of congestive heart failure. The medical committee provided specific guidance on the appropriate drug therapies for this severely ill patient population. Patients also received monthly follow-up when they were out of the hospital (Rose et al., 2001).
failure.4 Enrollment began with 32 study sites in 2002 and expanded to 171 sites by 2005 because of low patient enrollment. Eventually, 44 of the sites deactivated—at great expenditure of time, energy, and some $10,000 apiece—without ever enrolling a single patient. Originally, the trial was planned for Canada and the United States, but insufficient subject enrollment in these two countries prompted expansion worldwide. The enrollment of 576 patients in Poland ultimately provided the level of patient participation to meet enrollment goals. One of the lessons learned here, said Michler, is that additional sites should be added early in the trial timeline if a trial is not meeting expected enrollment targets.
The NHLBI-funded CTSN is conducting several clinical trials concerning the surgical management of cardiovascular disease in adults. Two trials focusing on the effectiveness and safety of currently available surgical approaches to ischemic mitral valve disease plan to randomize 550 patients.5 Again, enrollment initially was sluggish, with less than one-third of the patients eligible per month actually enrolled. That rate has improved significantly, to almost half of those eligible now randomized. However, there is a range of site productivity with some of the 14 study sites enrolling three-fourths of their eligible patients, while other sites are enrolling only a fifth or a sixth of their eligible patients. Identifying, in advance, which proposed trial sites will be more assiduous in their enrollment efforts, prior to final site selection, would be of great benefit, Michler said.
According to Michler, recruitment lessons from these three trials fall into several categories:
1. Protocol issues: Enrollment may be affected if a treatment is available outside the trial. For the REMATCH trial, the investigational LVAD technology was accessible only within the trial, which served as a powerful incentive for participation for those who wished to have access to that technology.
Unanticipated costs arising from difficulties in enrollment can negatively affect other parts of the trial protocol. For example, the STICH trial protocol originally included imaging studies to measure the size of the heart, but the trial sponsors made the decision to remove them from the protocol in order to increase enrollment
4 The STICH trial involved heart failure patients in tests of (a) whether coronary artery bypass graft (CABG) surgery plus medical treatment is better than medical treatment alone and (b) whether CABG plus left ventricular volume reduction (making the heart smaller) in patients with left ventricle dysfunction is better than CABG alone.
5 The alternative treatments for mitral valve disease in the CTSN moderate ischemic mitral regurgitation trial are CABG surgery with mitral valve repair versus CABG alone (300 patients). In the severe ischemic mitral regurgitation trial, alternative treatments are mitral valve replacement versus repair with or without CABG surgery (250 patients).
efforts by using the money saved to pay clinical sites for their trial expenses.
If the protocol design is too complicated or suggests that different arms of the trial are unequal (and thus that the trial lacks equipoise), these conditions might discourage physicians from referring patients to the trial.
2. Site issues: The clinical culture at a site may affect enrollment. Examples are lack of a site champion, weak institutional interest in clinical trials, and bureaucratic hurdles (legal issues, IRB inflexibility, and so on).
3. Surgeon and referring physician issues: Education and communication are critical during recruitment and throughout a trial and communicating openly and effectively with physicians about the state of knowledge aids enrollment. Physician toolkits are useful in building the knowledge base for referring doctors (and their patients).
The stronger the referral relationships are at the outset, the better off the investigator is when initiating a new clinical trial, Michler noted. Relationships can be strengthened by keeping referring physicians informed about the progress of the trial, finding ways for them to participate without it being burdensome to them, and, when feasible, including them in publications, Michler said.
4. Communication: During discussion it was mentioned that continued communication not only aids physician involvement but also may aid in retaining trial participants. For instance, in a later discussion, workshop speaker Bernadette Boden-Albala, Co-Director, Irving Center for Clinical and Translational Research, Community Engagement Core Resource, Columbia University, said that dissemination of trial results to patients is critical. She suggested that all studies should end with dissemination of trial results back to the community in which the cohort or the trial participants were from as well as more broadly to the public. When patients are in the communication loop during and after their participation in a clinical trial, this can increase positive feelings associated with trial participation and convey that their contribution was of value. A workshop participant also noted that some people who drop out of clinical trials are convinced to do so by a family member who is a health professional—a physician, pharmacist, or nurse. The participant questioned whether health professionals are being trained to think positively about advising patients (and friends and family) to participate in trials. In many cardiovascular trials the treatment effect may appear underwhelming, Michler said, because the new treatment may not improve greatly upon what is already surgically available, instead the study results define when a specific surgical
therapy is most appropriately employed. Although in cardiovascular trials family members often become the strongest advocates for participation, many of the same issues that make it difficult for clinicians to become involved in trials also make it hard for them to recommend them to their families, he noted.
5. Participant issues: There is a need to take into account people’s motivations to participate in trials, which can include earlier access to experimental treatments, having closer clinical monitoring, and a sense of altruism. Patients from different racial and ethnic groups and socioeconomic strata have different levels of trust in the medical community. Spending time with the patient and family, and providing educational materials that are culturally and linguistically relevant is “the only way to deal with the trust problem,” said Michler (see next section).
6. Funding and reimbursement: In this era of increasingly constrained and uncertain hospital budgets, the costs of a trial may prevent hospitals from serving as trial sites. For example, a decade ago, the average cost per patient borne by hospitals in the REMATCH trial was $63,000 (excluding the costs of the device and the surgery, which were funded by the sponsors). Although since 2000 Medicare has covered the costs of treatment for participants in NIH-funded trials, other insurers may not, and not all costs borne by the institution are deemed treatment costs and thus reimbursable. Michler offered his suggestions to trial administrators to improve management of patient recruitment, listed in Box 3-1.
Principles of Community-Based Participatory Research (CBPR)6
The U.S. population groups referred to in shorthand as “racial and ethnic minorities” by 2060 will constitute the majority of the nation’s population. Bringing adequate representation of diverse societal groups, especially vulnerable populations, will be essential if all Americans are to benefit from improvements in the prevention and treatment of serious medical conditions.
Accomplishing this requires an understanding of why people want to participate in trials. “For that, we need help from insiders,” Carol Horowitz, Associate Professor, Department of Health Evidence and Pol-
6 Based on the presentation by Carol Horowitz, Associate Professor, Department of Health Evidence and Policy, Mount Sinai School of Medicine.
Steps for trial leadership:
• Refine the patient eligibility criteria.
• Understand the preparation time necessary before a trial can start.
• Meet early with the IRB and legal department to try to expedite approvals.
• To assess site enrollment capabilities prior to selection, review previous trial screening logs from each site.
• Train investigators in strategies to maximize enrollment.
• Rapidly increase the number of clinical sites if enrollment lags.
• Establish and maintain strong communication, through frequent visits and regular meetings with all sites and principal investigators.
And, at the site level:
• Identify a strong leader willing to champion the trial in the institution.
• Create partnerships with referring physicians and surgeons.
• Consider identifying multiple principal investigators, who can reach out as a colleague to a variety of physician specialists (e.g., a cardiologist as well as a cardiac surgeon) or to different population groups (e.g., a physician who is also a community leader).
a Based on the presentation by Robert Michler, Surgeon-in-Chief, Professor and Chairman, and Director, Center for Heart and Vascular Care at the Montefiore Medical Center/Albert Einstein College of Medicine.
icy, said, because the perspective potential participants will have on a proposed trial is likely very different from the researcher’s (see Table 3-1).
Researchers believe their projects involving diverse and underrepresented populations are valuable and urgent. Potential participants, however, have an earned skepticism, based on the many failed medical and social experiments of all kinds that have been conducted in their communities. According to Horowitz, when people believe they or their communities will not benefit from a study, the risk of participation rises to unacceptable levels. She remarked that the legacy of the Tuskegee syphilis study—the culmination of a long list of unethical and immoral research and treatment practices to which blacks were subjected—is neither easily dismissed nor forgotten (Washington, 2006).
Rather than sharing the perception that “this is an evidence-free world,” giving rise to an urgent need for more clinical trials, minority community members often believe that “we know why we are sick, and
|Researcher Perspective||Participant Perspective|
|Purpose||Clear, important||Dubious (“Earned Skepticism”)|
|Timeline||now, Now, NOW!||What’s the hurry?|
|Benefit||Obvious (career, grants, knowledge, health, it’s for their own good)||Unclear (drive-by, or helicopter research)|
|Risks||Minimal (no biggie, and Tuskegee was ages ago!)||Unacceptable if benefit iffy, historic abuses|
|Attitude to Research||Needed to gain knowledge||Problems apparent, resources lacking. Where did it come from?|
|Participants Should||Agree, comply||Question, contribute|
SOURCE: Horowitz and Bickell, 2011.
we don’t have the resources to do anything about it,” said Horowitz. To these communities, the researcher’s description of the planned study, which often might focus on a very narrow question that will benefit practitioners in a particular discipline, simply does not resonate.
Horowitz noted that there is a social obligation to approach the recruitment of minority patients into trials with ample forethought and a commitment to community engagement, in addition to the typical practical requirements of implementation of a protocol. Engagement should go beyond and may even precede the recruitment of participants in a particular study and, ideally, would consist of a long-term commitment to create a research-friendly community. This requires involvement with all those directly and indirectly affected by and involved in the trial: communities of patients, their clinicians, leaders of community organizations, and local opinion leaders. The latter are the hubs for social networks and the cultural insiders and thus are the people who can reach potential participants. This approach is commonly called “community-based participatory research” (CBPR).
Horowitz described three strategies that have proved useful in engaging a community in clinical research:
• Emphasize effective communication, taking into account literacy levels, vision (especially with elderly patients), and language, and ensure that information given is no more complicated, jargon-laden, or
legalistic than it absolutely needs to be.7 Effective communication requires recognition of the gap in subject-matter knowledge between researchers and prospective participants and appreciation of attitudes toward research, clinical trials, and health care providers.
• Build relationships in the community and ensure there is a mutual benefit, so that communities and participants see that the trial represents a “win-win.” Fundamentally, it means conducting trials that people want to be part of and believe is important for them.
• Anticipate practical matters. As examples, the choice between having a device implanted and taking a pill may not be an equivalent one from the point of view of patients; the study may involve costs in personal time or money that dissuade would-be participants; there may be excessive requirements for completing forms; participants may need transportation or dependent care; or the study site may be hard to get to or have limited hours. There is a need to simplify requirements and reduce burdens on participants to the extent possible.
In the discussion panel it was mentioned that while some investigators are becoming more aware of CBPR principles and how these principles could be effective in aiding participant recruitment, research sponsors do not necessarily understand them yet, said workshop participant Karriem Watson (HHS, 2011).
The National Institute on Minority Health and Health Disparities (NIMHD) has encouraged the development of CBPR, presenter Nina A. Bickell, Director, Center to Achieve and Sustain Health in Harlem, Mount Sinai School of Medicine, responded, and NIH funded the prediabetes grant Horowitz described for the workshop (see the case study on prediabetes care in the next section). The understanding was that the investigators would use CBPR methods to select the topic and create and implement a pilot study.8
A significant need in making CBPR work effectively is to build the capacity of community residents (and patients) to be grant reviewers. This requires training and a foundation of relevant information in order for CBPR to maintain robust research. For example, workshop participant Ronnie Todaro of the Parkinson’s Disease Foundation, said her organization supports patient participation throughout the research process, from study
7 The average American reads at an 8th-grade level; 40 percent of elders and half of African Americans and Hispanics read at or below the 5th-grade level (Partnership for Clear Health Communication, 2006).
8 See, for example, NIMHD’s CBPR initiative: http://www.nimhd.nih.gov/our_programs/communityParticipationResearch.asp (accessed October 10, 2011).
design, to recruitment, to dissemination of results. It also offers a 3-day learning institute for people with Parkinson’s disease, to prepare them to serve as FDA advisors, serve on IRBs, review study protocols, and educate other community members about the importance of clinical trials.
The Clinical and Translational Science Awards (CTSA) institutions also are committed to community engagement, Horowitz said.9 And, in places where there is a lot of institutional support for CBPR, there reportedly is parallel interest on the part of funders.
Case Study: Prediabetes Care
Nearly 26 million Americans have diabetes, with substantially higher rates among Hispanics and non-Hispanic blacks, compared to non-Hispanic whites (CDC, 2011). An East Harlem diabetes prevention trial is testing whether a successful, but expensive, hospital-based strategy of identifying people with prediabetes and helping them lose weight can be adapted to be delivered at the community level, using peer-led interventions. A substantial “pre-recruitment effort” began with an educational initiative, which Horowitz said was intended to (a) help area residents recognize that diabetes could lay in store for them even if they did not yet have it and (b) sensitize them to the threat of diabetes to a point at which they demanded action.
In contrast to most clinical trials, it was not the researchers, but their community partners, who chose diabetes prevention as the focus of the study. The partners selected many of the research methods and strategies; they made sure the study would resonate with the community, building trust and motivation and the relationships needed to support the research down the road; and they determined the incentive participants would receive. For its part, the research provided the community with a tangible benefit, including jobs for local people.
The community partners developed the following recruitment strategies:
9 Resolving infrastructure issues in academic medicine has been one of the concerns of the NIH-funded CTSA consortium and its 60 participating academic medical research institutes. In late 2010, NIH announced plans for a new $722 million National Center for Advancing Translational Sciences (NCATS), a proposal before Congress in summer 2011. The purpose of the new NCATS is “to catalyze the generation of innovative methods and technologies that will enhance the development, testing, and implementation of diagnostics and therapeutics across a wide range of diseases and conditions” (Collins, 2011). The CTSA consortium will be incorporated into this new Center.
• Recruitment “where the people are,” such as at major community events, in churches, food pantries, senior centers, and schools
• Holding “Stop Diabetes Day” recruitment parties
• Asking community-based organizations to champion the cause
• Leading the recruitment effort themselves (which proved the most effective)
• Asking clinicians to refer people (the least effective strategy)
Potential participants were required to fast, undergo various tests, and commit to being in a workshop and coming for follow-up. That they were willing to do this (despite not having diabetes) suggests to Horowitz that they were reasonably confident of the study’s value before actual recruitment began.
The study was a success, both in terms of recruitment and its outcomes. In three months it recruited 99 “hard-to-reach” people, most of whom were nonwhite, Spanish-speaking, unemployed, undereducated, low-income, and uninsured. The group achieved significant weight loss, which they maintained for a year.
Throughout, the researchers worked to create an upbeat environment at the study site that would make participants feel welcome and part of the team. As the researchers routinely told study participants, “This is brought to you by you.”
Case Study: Continuity of Breast Cancer Care10
An example of a trial exploring continuity of breast cancer care was presented to illustrate strategies for engagement of physicians in clinical trials. The goal of this trial was to make sure that breast cancer patients who had surgery made it back to their oncologists to obtain the adjuvant treatment that can improve survival rates. The trial subjects were breast cancer surgeons practicing in six unaffiliated New York City hospitals, for whom the researchers developed a tracking and feedback registry.
The researchers obtained the name of a key breast cancer surgeon in each hospital from its chair of surgery. They wanted to identify surgeons whom other surgeons looked up to and who could work across disciplines. These individuals were asked to be the principal investigators for the study at their specific hospital.
To convince these already-busy physicians to take on this job, Bickell said, the researchers worked to help them understand the extent of
10 The breast cancer care case study was presented by Nina A. Bickell, Director, Center to Achieve and Sustain Health in Harlem, Mount Sinai School of Medicine.
underuse of adjuvant therapy, especially among minority women.11 The researchers documented that these patterns of underuse persisted across hospitals and across surgical practices, which, according to Bickell, made the problem very salient to the surgeons and helped motivate participation.
In addition, the trial offered the surgeons a service—patient tracking— they could not otherwise have. Finally, the study team followed the “make it convenient” rule by streamlining demands on the surgeons and their practices, such as by communicating with a point person for each surgeon.
Surgeon recruitment and study outcomes were strong. Of surgeons who operate on women with breast cancer in the target hospitals, 97 percent agreed to participate, and the overall rate of adjuvant underuse dropped from 23 to 14 percent between 1999-2000 and 2004-2006 (Bickell et al., 2008).
Working with health professionals does not obviate the issue of trust, Bickell said. In this study, surgeons were particularly concerned about confidentiality and malpractice. Just as when working with a patient population, these issues had to be dealt with up front, through the trial design and recruitment processes.
Sixty years ago, clinical trials on the first drugs to treat schizophrenia and depression found such robust positive effects that the studies could be conducted with small numbers of patients. In recent decades, the situation has changed dramatically, said Kenneth Davis, President and CEO, Mount Sinai Medical Center and Professor of Psychiatry, Mount Sinai School of Medicine. The failure rate in clinical trials of new treatments for mental illnesses is disappointingly high:
• New treatments produce only small positive effects.
• Large numbers of participants respond as well (or better) to the placebo.
• Enrollment of participants is excruciatingly slow.
11 The percentages of women who needed postsurgical adjuvant therapy and did not receive it were 34 percent among black women, 23 percent among Hispanic women, and 17 percent among white women.
12 This section is based on the presentation by Kenneth Davis, President and CEO, Mount Sinai Medical Center and Professor of Psychiatry, Mount Sinai School of Medicine.
Similarly, while early trials of drugs for Alzheimer’s disease proceeded smoothly—with robust effects, small placebo responses, and short enrollment periods—the situation has again reversed.
In recent years, trials of treatments for mental disorders have faced a number of difficulties in patient recruitment that are exacerbated with respect to psychiatric conditions:
• The stigma of the condition and, in some instances, the treatment (in part a result of lurid media depictions of both mental illnesses and their treatment)
• Ethical issues, when dealing with patients at high risk of suicide
Researchers have puzzled extensively over the difficulties with clinical trials related to mental disorders and have produced several hypotheses to explain them, said Davis. Perhaps mental illnesses differ to a much greater degree than previously thought, in which case much more needs to be learned about their underlying biology so that individual patients can be linked to the correct therapy. Unfortunately, this approach has not worked for patients with Alzheimer’s disease.
It is also possible that based on new knowledge about the diagnosis, spectrum, and progression of these diseases, inclusion criteria for clinical trials have evolved to reflect these nuanced understandings of various mental illnesses. Another possible explanation for the difficulties facing clinical trials in mental health is that perhaps it was easier to recruit patients into clinical trials when there were no legitimate competing treatments. Today’s trials in mental health might attract patients who have not had relief from the now-established treatments and whose disease is more difficult to treat. Or, as workshop participant Michael Parides, Professor of Biostatistics, Department of Health Evidence and Policy, Mount Sinai School of Medicine, suggested, the design of confirmatory trials might be flawed due to weaknesses in the design and objectives of early-phase trials.
Several recent trends have changed the research environment. One such trend is the growth of CROs, which in 2010 performed an estimated $20 billion in outsourced research tasks and support for the pharmaceutical and biotechnology industries. These firms vary markedly in their ability to execute studies and their commitment to quality, rigorousness, and even speed, Davis said. Another trend is the increased use of study sites outside the United States, some of which have less experience with conducting clinical trials than do U.S. institutions.
It is sometimes difficult to know how to test new treatments for mental disorders, particularly treatments having novel mechanisms that may require novel trial designs, said Davis. Sponsoring companies and FDA
often end up at an impasse, he remarked, with the companies wanting to know what kind of trial designs FDA will accept, and FDA waiting for companies’ proposals.
Ultimately, what will be needed for successful trials for mental conditions will be to invest substantially more time in drug development, said Davis. U.S. patent law, however, fails to incentivize lengthy drug development efforts. Davis said these laws will have to be reconsidered if they are to best meet public health needs and enable real clinical breakthroughs for treatment—and especially prevention—of psychiatric conditions.
This page intentionally left blank.