Over time U.S. investments in clinical research have grown substantially, despite the fact that the rate of increase in funding slowed between 2003-2007. Industry was the largest funder of biomedical research in 2008 ($38.4 billion) and within the public sector, the National Institutes of Health (NIH) invested $27.9 billion in biomedical research (Dorsey et al., 2010). The 2009 American Recovery and Reinvestment Act provided a one-time infusion of $1.1 billion in new funds for comparative effectiveness research, which includes pragmatic randomized trials;2 and the 2010 Patient Protection and Affordable Care Act created a sizable trust fund to sustain this support.
Growth in funding and the ongoing rapid pace of scientific discovery have led to a major increase in the number of clinical studies, with
1 This section of the summary is based on presentations by Deborah Ascheim, Associate Professor, Department of Health Evidence and Policy, Mount Sinai School of Medicine; Annetine Gelijns, Co-Chair, Department of Health Evidence and Policy, Mount Sinai School of Medicine; and Juan Lertora, Director, Clinical Pharmacology, NIH Clinical Center.
2Explanatory randomized trials generally measure the effects of a treatment (its efficacy) under ideal conditions, often using carefully selected individuals treated in a research setting. Pragmatic trials measure the effects of a treatment (its effectiveness) in the everyday clinical practice of community physicians, and their purpose is to provide information about the choice of treatments (Roland and Torgerson, 1998). Key features of pragmatic clinical trials include that they: (1) select clinically relevant alternative interventions to compare, (2) include a diverse population of study participants, (3) recruit participants from heterogeneous practice settings, and (4) collect data on a broad range of health outcomes (Tunis et al., 2003).
the number of randomized trials increasing 35 percent between 2005 and 2010.
The design, conduct, and analysis of clinical trials involves a broad range of public and private institutions, including academic medical centers, hospitals, community practices, pharmaceutical and medical device industries, voluntary health organizations, and contract research organizations (CROs). The NIH Clinical Center in Bethesda, Maryland, is a unique example of an organization wholly devoted to clinical trials. The Center conducts a robust program of intramural clinical trials, with more than 1,250 physicians and the bed capacity for 240 inpatients, all of whom are participants in one of the Center’s more than 1,400 active research protocols. Most of the Center’s intramural clinical trials are phase I or phase II research—that is, early, proof-of-concept studies, rather than later, confirmatory research, which is generally phases III and IV (see Appendix B: The Clinical Trials Process for definitions of phases 0 through IV).
Multiple Challenges to the Clinical Trials Enterprise
Despite increases in both funding and effort and an understanding of the requirements for successful clinical trials (see Box 2-1), the enterprise continues to falter, said Annetine Gelijns, Co-Chair, Department of Health Evidence and Policy, Mount Sinai School of Medicine. Some of the barriers to participation in clinical trials by key groups—patients, academic institutions, community physicians, and researchers themselves—and potential ways to overcome them are described in the succeeding paragraphs.
Barriers to Patient Recruitment
Clinical trials experience significant obstacles in patient recruitment.
• Well-defined clinical trial goals and target patient population
• Realistic patient accrual strategy
• Community outreach and education, building trust
• Community/patients advisory boards
• Well-informed community physicians and other health care professionals
• Effective patient retention strategy
• Protocol implementation support team
a Material presented by Juan Lertora, Director, Clinical Pharmacology, NIH Clinical Center.
For example, according to Gelijns, clinical trials funded by the National Cancer Institute encounter these stumbling blocks (IOM, 2010b):
• It can take more than 2 years from the time planning for a trial begins to enrollment of the first patient.
• The recruitment of qualified patients in clinical trials is often a slow process; insufficient patient recruitment can delay or cause a trial to be cancelled.
• Forty percent of trials end prematurely and are not published, a substantial waste of resources.
• Less than 5 percent of adult cancer patients are enrolled in clinical trials.
• Even fewer patients who are elderly, women, or minorities are enrolled.
As another example, trials involving a common cardiovascular disease, atrial fibrillation, also experience major issues in terms of slow accrual of patients. This holds for explanatory (phase I and II) trials and confirmatory (phase III) trials, but is perhaps even more salient for pragmatic trials that make head-to-head effectiveness and safety comparisons in a broad range of patients essential for developing the evidence base for everyday clinical decision making. Pragmatic trials are much less common than efficacy trials. The latter trials are clearly important to define the benefits of a novel intervention. However, the relative paucity of pragmatic trials has led to a situation where the evidence base underlying much of clinical practice is inadequate. A recent review of American College of Cardiology (ACC)/American Heart Association (AHA) guidelines found that a large proportion of recommendations in current cardiovascular guidelines were based on expert opinion or case studies, and not evidence from randomized clinical trials comparing therapies (Tricoci et al., 2009). According to Gelijns, factors contributing to problems in recruitment of both efficacy and effectiveness trials include
• lack of awareness among physicians and patients that relevant trials are available;
• lack of awareness of the benefits of engaging in clinical trials;
• maintaining clinical equipoise,3 especially when treatment arms are very different and patients or physicians have strong preferences for one therapy over the other;
3 Equipoise is the point at which a rational, informed person has no preference between two (or more) available treatments (Lilford and Jackson, 1995). In clinical research, the ethical concept of equipoise is satisfied when genuine uncertainty exists as to the comparative therapeutic benefits of the therapies in each arm of a clinical trial.
• maintaining equipoise is even harder for pragmatic trials as the treatment evaluated is typically available, and reimbursed for by payers, outside the clinical trial setting; and
• continued patient participation when the treatment is available outside the trial environment.
Gelijns noted that positive developments that may help overcome these barriers are
• greater collaboration between voluntary health organizations and clinical trial investigators to use centralized trial registries to reach out to patients with specific conditions and their physicians;
• outreach to community organizations (for example, the National Heart, Lung, and Blood Institute [NHLBI] has engaged in intensive community-based efforts to reach people with sickle cell disease);
• creation of user-friendly tools that can use data in the electronic health record (EHR) to identify eligible patients at the point of care;
• making the public aware that trial participants could receive superior care, as a result of increased follow-up and monitoring; and
• efforts to help people understand the current uncertainties about the relative benefits of various treatment options and the value that clinical trials provide to clinical decision making.
Weak Institutional Support
A number of disincentives to conducting clinical trials are found within academic medical institutions, Gelijns said, one result of which is a declining pipeline of new clinical investigators. Some of the reasons individuals do not pursue clinical investigation in the academic medical setting include
• economic disincentives such as inadequate reimbursement for trial activities and reduced time available to provide patient care;
• logistical disincentives that derive from a lack of infrastructure and support; and
• cultural disincentives, which manifest themselves in the decisions of tenure and promotion committees that often have a bias against clinical trials activity as a prestigious or beneficial academic pursuit.
Lack of Community Physician Involvement
There is also a lack of engagement on the part of community physicians. For example, the results of trials are often not generalizable
to the community physician’s practice, as they are based on carefully selected patient populations and specialized delivery settings, Gelijns and Deborah Ascheim, Associate Professor, Department of Health Evidence and Policy, Mount Sinai School of Medicine, observed. However, Gelijns added, opportunities to create new partnerships and networks between academic and community-based physicians are being actively explored. Such networks could provide a “clinical laboratory” in which multiple trials can be conducted without having to create a new infrastructure for each individual trial. This structural approach would increase both the efficiency and the relevance of trials.
Another barrier for community physicians and patients, according to Gelijns, is a lack of third-party reimbursement for “experimental” therapies. More than a decade ago, Medicare agreed to cover the “routine care costs” for patients participating in qualified randomized trials (notably NIH-sponsored trials). However, private insurers’ reimbursement policies remain highly variable. This constitutes a major obstacle to trial participation, according to Gelijns.
Finally, some barriers to participation are inherent in the design of trials and are relevant to patients, researchers, and community physicians, said Gelijns. A review of more than 10,000 industry-sponsored trials found that the number and frequency of trial-related procedures such as laboratory tests or patient questionnaires has increased substantially over time (Getz et al., 2008). As the number and frequency of unique procedures patients must undergo have increased, and as the demands of protocol administration also have become much greater for investigators, Gelijns said, the resultant “trial execution burden” may be a major deterrent to participation for patients and potential researchers alike.
The analysis of trial results generally relies on statistical measures of “central tendency”—averages—and their strength, as measured by standard deviations and significance tests. An unavoidable weakness of this approach, Juan Lertora, Director, Clinical Pharmacology, NIH Clinical Center, commented, is that its generalizability is limited because the “real-world” patient population is heterogeneous, so results of a treatment are likely to be more scattered than in a study population of demographically and clinically similar individuals. Broadening analytic methods used in trials may make them more applicable to real-world populations and thus more useful for community physicians.
Innovations in trial design may simplify protocol design and speed the trials process, especially in the case of exploratory trials, said Gelijns. Such developments require close interaction between statisticians, inves-
tigators, and the Food and Drug Administration (FDA) to ensure they will generate evidence acceptable in the approval process. Gelijns concluded by noting that, while clinical trials are a critical link between scientific discovery and clinical practice and essential to ensuring that Americans reap the potential benefits of the nation’s enormous investment in biomedical research, much can be done to improve the effectiveness of this system. She emphasized that a broad spectrum of stakeholders is involved in this process and will be critical to its success.
NIH Resources for Extramural and Intramural Researchers
Lertora noted that NIH can share with outside researchers some resources it has developed to support its own intramural clinical trials. These resources can facilitate protocol development and trial implementation for researchers who need support beyond what their home institutions can offer. Making such resources available is in line with the recommendation to have the Clinical Center open its doors to extramural investigators (NIH SMRB, 2010). Tools available to external researchers include the following:
• NIH Clinical Center staff developed a sophisticated computer-assisted protocol development tool called ProtoType4 (Wanjek, 2008). The tool was created by NIH Clinical Center staff, including Philip Lightfoot, and with the support of Dr. Jon McKeeby and NIH Clinical Center Director, Dr. John Gallin. This web-based tool, 8 years in initial development and launched in 2008, allows for a standardized approach to protocol development, is flexible in terms of incorporating new “rules,” increases the efficiency of Institutional Review Board (IRB) and scientific committee reviews, and creates a full protocol history. The paperless system can also import image libraries, reference databases, and so on. Furthermore, it supports collaboration and is helpful in training new investigators.
• A computer-based algorithm known as the “Investigational New Drug (IND) Wizard” is part of ProtoType and helps investigators determine whether an IND filing is required for protocol implementation, which is especially helpful for researchers investigating potential new indications for existing drugs (“drug repurposing”).
4 The ProtoType tool can be found at: http://prototype.cc.nih.gov/prototype10/contents/login/pw_login_screen.aspx (accessed October 10, 2011).
Intramural clinical investigators at the NIH Clinical Center are assisted by the NIH Office of Communications, Patient Recruitment, and Public Liaison regarding prescreening of potential patients, establishing contact with volunteers and physicians, and providing ongoing recruitment assistance. Another resource available to intramural investigators is the long-standing Clinical Center healthy volunteer program with a registry of over 50,000 people that serves as a source for control groups and for the study of basic human physiology, pathophysiology, and pharmacology.5
“We need to create an environment in which the clinical study is no longer the patient’s last resort, but the first resort, and participation in a clinical study becomes part of the standard of care.”
—Angela Geiger, Alzheimer’s Association
The workshop’s first panel asked experts from diverse perspectives— regulators, payers, patients, and pharmaceutical companies—to comment on the problems and opportunities in clinical trials described in the opening presentations. Much of the discussion focused on drug development.
Clinical Trials and New Drug Development
At the outset, clinical trials’ importance in the regulatory process was underscored by panelist Leslie Ball, Acting Director of FDA’s Office of Scientific Investigations. Clinical trials produce the data that support FDA determinations whether a drug should be approved, and they determine what information will appear on the product label. While FDA regulations are designed to promote the public’s health and protect people from unsafe, ineffective, or poor-quality drugs, the agency also has the mission of helping to speed innovations. As a result, according to Ball, FDA attempts to avoid rules and oversight practices that unintentionally introduce inefficiency and unnecessary resource use into the drug development process, and the agency recognizes that clinical researchers benefit from predictability all along the path of drug development: review, approval, and monitoring.
5 Recently, a national registry for healthy volunteers for phase I clinical trials was proposed, in order to avoid overreliance on a small pool of individuals (some of whom treat trial participation, which pays from a few hundred to a few thousand dollars, as a profession). A limited volunteer pool poses excess risk to the individuals and may compromise a research study if treatment effects are confounded by recent exposure to another investigational agent (Resnik and Koski, 2011).
Evolution in clinical trials management is posing some new challenges for pharmaceutical manufacturers and regulators. Among them are trends toward globalization of clinical trials, recruitment of patients overseas, outsourcing trial components, and an increasing complexity of trial protocols. There is also a need to determine how best to integrate information from EHRs into a trial’s analytic structure.
To create more efficiencies in this dynamic environment, Ball noted that FDA is engaged in efforts to
- increase dialogue among drug developers, academia, patients, and regulators;
- prioritize regulatory oversight activities and inspections based on evaluative assessments of risk and protocol implementation anomalies that really matter;
- foster a proactive quality risk management approach to the design, conduct, monitoring, data management, and reporting of clinical trials that builds quality in from the beginning of a trial’s development;
- standardize and harmonize processes and regulations, both domestically and internationally;
- develop standards to facilitate use of EHRs in trials (Kush, 2011); and
- improve trial methodology through public-private partnerships (for example, the Clinical Trials Transformation Initiative [CTTI], http://www.trialstransformation.org/, and the Analgesic Clinical Trials Innovation, Opportunities, and Networks [ACTION] initiative).
One reason trials are increasingly difficult to conduct is that the U.S. standard of health care is generally high, commented Richard Murray, Head of Global Center for Scientific Affairs, Merck & Co. It is increasingly difficult to identify patients to participate in studies who have not yet had some type of treatment for their condition and, because new treatments are likely to represent only incremental improvements, large numbers of patients are needed to achieve statistical validity.
For the management of chronic diseases and conditions, studies may need to be relatively long in order to identify safety issues that could arise when a drug must be taken for months, years, or, possibly, a lifetime (IOM, 2011). Disadvantages of longer trials are that patients are harder to retain in the trial, the risk of investigator fatigue rises, and, from the manufacturer’s point of view, the “clock is ticking” on the period of patent protection, said Murray.
According to Murray, recruitment barriers encountered include pub-
lic mistrust of clinical trials (and, in general, some mistrust of science and medicine by many Americans) and mistrust of the motives of the pharmaceutical industry. A horizon issue that might further erode trust, he said, is the use of foreign research subjects. This trend could be perceived as posing ethical problems, especially if the treatments being tested are not available in the overseas patients’ own country.
Box 2-2 lists some of the principles and practices that workshop participants and discussants noted they have found helpful in recruiting and retaining patients in clinical trials. According to Greg Simon, Senior Vice President, Patient Engagement, Pfizer Inc., Pfizer has moved the “make it convenient” notion a large step forward with a new, online clinical trial, called REMOTE, which hopes to enroll about 600 patients in 10 states. The trial, which has FDA approval, will be testing a new drug and using mobile phone and Internet technology to facilitate participation, avoiding repeated visits to trial sites. Pfizer will compare results of this study to those of a similar traditional, center-based trial to determine whether
• Develop a strong recruitment strategy involving community leaders.
• Start recruitment early—before time to start the trial.
• Respect local culture.
• Engage caregivers.
• Involve racially and ethnically diverse physicians in the recruitment process and in the trial itself (as a longer-term strategy, Merck has worked with the National Medical Association to increase clinical research training among African American physicians).
• Make it easier and more convenient for people to be part of the trial, through such practices as reducing the required number of blood draws and reimbursing for, or even arranging, travel.
• Give people credit for checking in and filling out some of their information remotely.
• Clearly explain the importance of the trial for the public’s health, which makes it easier to overcome public skepticism and lack of understanding.
a Based on presentations by Carol Horowitz, Associate Professor, Department of Health Evidence and Policy, Mount Sinai School of Medicine; 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; Richard Murray, Head of Global Center for Scientific Affairs, Merck & Co.; and Greg Simon, Senior Vice President, Patient Engagement, Pfizer Inc., and workshop discussions.
the mobile model can improve the access, timeliness, quality, and safety of its research.
A number of broader problems with the current clinical trials enterprise were described by Simon:
• Broken feedback loops. Trials produce knowledge that never reaches physicians, and physicians learn specifics about drug performance that never find their way back to researchers or become incorporated in future trials. Stronger feedback loops would help not only in identifying unexpected problems with a drug, but also in discovering unexpected positive outcomes.
• The need for greater coordination and data sharing among academia, government, clinicians, and industry, in order to encourage new innovation.
• Uncertain expectations regarding how much information is needed to get a drug to patients and what information should be collected once a drug reaches the market. Simon noted that even a 10,000-person trial cannot reveal how a drug will operate in millions of people.
• The need for increased funding for postmarketing surveillance, perhaps by allowing Prescription Drug User Fee Act (PDUFA) monies to be used for this purpose.6
In response to the discussion of whether private or public payers should increase their support for the clinical trials enterprise, Bruce Vladeck, Senior Advisor, Nexera, Inc., described the need for strong, meaningful trial standards to overcome the problem of trials of widely varying quality. He commented that criteria are needed to identify trials both of adequate scientific quality and sufficient importance, in terms of the public’s health, to warrant public support. He also added that there needs to be strict compliance with research conflict-of-interest policies. Workshop participants discussed the respective roles of randomized clinical trials and observational studies in the generation of medical evidence and the assessment of a treatment’s safety. Randomized clinical trials are useful for conditions that require small trials in subgroups of people with similar genetic mutations. In addition, randomized trials may still be the best way
6 Enacted in 1992 and renewed in 1997 (PDUFA II), 2002 (PDUFA III), and 2007 (PDUFA IV), this law authorizes FDA to collect fees from companies that produce certain human drug and biological products and includes certain performance and other standards in connection with the regulatory process. In fiscal year 2010, FDA collected just under $552 million in PDUFA fees from manufacturers, and these fees provided more than 60 percent of financial support for the drug approval process (FDA, 2011a).
to assess the efficacy of new drugs, as opposed to providing a definitive safety assessment, Jeffrey Drazen, New England Journal of Medicine, said. Rare side effects or drug-drug interactions require time and large numbers of heterogeneous patients to discover. Because such events are relatively rare, Ball said, safety is being monitored through postmarketing surveillance that relies on patient registries and observational studies. FDA’s Sentinel Initiative, for example, actively queries diverse automated health care data holders—EHR systems, administrative and insurance claims databases, registries, and the like—to track the safety of drugs, biologics, and medical devices once they reach the market (FDA, 2011b).
Patient Recruitment by Voluntary Health Organizations
Voluntary health organizations are proving effective in involving patients and the broader public in trials that involve relatively rare diseases for which cure remains unavailable. For example, there are only about 30,000 people living with cystic fibrosis (CF) in the United States. To reach this population, the Cystic Fibrosis Foundation has worked with the medical community to establish more than 110 CF care centers nationwide, about 80 of which now can conduct clinical trials, said Joan Finnegan Brooks, President, Patient-Focused Market Research, and a person with CF. This network has successfully reached out and engaged the CF community to participate in clinical trials, resulting in several new therapies to manage the disease.7
Similarly, the Alzheimer’s Association has promoted patient involvement in clinical studies (Alzheimer’s disease affects some 5.4 million Americans). The Alzheimer’s Association’s FY 2009-FY 2011 strategic plan explicitly includes patient participation in clinical studies as a part of the strategy to accelerate research progress. The Alzheimer’s Association has taken a number of concrete steps in accord with this strategy, said Angela Geiger, Chief Strategy Officer, Alzheimer’s Association. The organization has developed a health care provider outreach campaign to increase awareness of Alzheimer’s trials, and in July 2010 it launched TrialMatch, a web- and telephone-based service that has connected more than 2,500 people with some of the 131 Alzheimer’s clinical trials under way around the country, resulting in at least 115 people enrolled. Almost half of those who complete a TrialMatch profile are caregivers who say they “want to give something back” by participating in studies; the next biggest group is healthy volunteers. The desperation many patients and families feel
makes them willing to participate in a trial, even if they understand it will not help them, said Geiger.
According to Simon, in an environment where scientists are many and trial participants are scarce, patient groups are starting to organize their own clinical trial networks and offer them to scientists, “because they realize their registries, their tissue banks, their biobanks, and their experience are the key resources.”