Adaptive Trial Design: From a Recruiter's Perspective
The trial flexibility intended by the recently proposed adaptive clinical trial design is at first blush a refreshing alternative to present clinical research designs. Adaptive clinical trial design, one that allows for a predefined modification to some aspect of the clinical trial after its initiation without undermining the validity and integrity of the trial, has its merits. It has the potential to offer some increases in flexibility and adaptability needed by today’s clinical trial environment. It offers further promise of some improvements in efficiency. Yet these promises, the author contends, are impractical and somewhat deceptive. Adaptive trial design, she believes, fails to address some of the most pressing problems in patient recruitment today, ones which are worthy of address before its widespread implementation.
In their recent Monitor article “Adaptive Clinical Trial Design,” Kram & Robbins highlight the benefits of what appears to be a panacea for trial centers—a system of enhanced sample size estimation, response-adaptive treatment allocation, and early aborting of inferior trials or groups. The adaptive design, they maintain, permits greater control, at multiple points throughout the trial process, making for opportune adjustments in product development and increasing overall trial efficiency. While on its face these benefits appear to be helpful to trial process, it permits continued avoidance of several significant recruitment-related problems.
From the recruiter’s perspective, the adaptive design appears reminiscent of the problematic total quality management (TQM) and continuous quality improvement (CQI) designs of the 1980s (DiIulio et al., 1993). These similarly optimistic models promised unparalleled improvements in client feedback, efficiency and productivity levels by repeated assessments of quality throughout the business cycle. They spawned not only the development of academic and training programs but whole movements that rippled throughout the private and public sectors. Despite their initial reported success, these models ultimately petered out and resulted in little progress in the anticipated areas of efficiency and productivity. Most problematic for these designs were their inability to take into account the interests and needs of a diverse group of customers. Indeed customers in any business have different ideas about how they ought to be served and have different levels of access to resources. Moreover, TQM and CQI were shown to be more effective for businesses that have clearly identifiable products, not customer service industries (DiIulio et al., 1993). On account of similar perceived deficits in its practical application, the author contends that adaptive clinical trial design may follow suit.
Also from the recruiter’s perspective, adaptive trial design’s insufficiency for handling customer relations activities is worrisome. For instance, the design proposes the use of a rational approach to and a slow meted out, continuing enrollment rather than a “concentrated, quick signup process” for improving upon recruitment. But recruiters seek to persuade patients about the benefits of trial participation using, as stated in a recent Monitor article (Higgins, December 2007), skill and art in addition to science. When dealing with customers, there is an undeniable and oftentimes near exclusive focus on emotions. Patients have many reasons for participating in trials, some of which remain too elusive or complex to quantify, and are wholly irrational.
Additional problems related to customer service stem from adaptive trial design’s apparent failure to take into account the diverse ideas of patients about how they ought to be served. For instance, patients are not permitted to participate in the decisions about inclusion of certain elements of the trials themselves. They are not permitted to select different options for remuneration of their time and travel such as products or services in lieu of payment. In many cases patients are not permitted to have access to information about what study medications to which they were exposed. Adaptive trial design serves to perpetuate this failure of the highly bureaucratic and complex pharmaceutical industry to truly regard trial patients as customers. It also reinforces existing boundaries within and between industry departments enabling operators to become further insulated and at risk for a lack of accountability to its customers. All industry employees, regardless of their level, must strive to make customer service the focus of their activity. One possible way of moving toward obtaining this objective is by enabling patients to purchase stock in the industry as a whole, in individual community-based trial centers, or in particular medications.
Adaptive trial design poses even greater risks to the industry on account of its likely failure to identify the customer served. Much like the difficulties experienced by the federal government as it attempted to implement TQI, clinical research will likely experience difficulty identifying its customer. For instance, recruiters experience considerable difficulty on a practical level identifying the increasing numbers of customers of clinical trials involving the cognitively impaired. The first and perhaps most obvious reason for risk of self-destruction involved in adaptive trial design relates to the ethicality of addressing the unique consent requirements of this population. Most important is the need for addressing cognitively impaired people’s comprehension of trial involvement, obtaining their true informed consent to participate, and recompensing them fairly and justly for time and risk exposure. While the legal system has begun addressing the issue of the rights for the cognitively impaired who participate in trials and encourages the involvement of legal guardians, who many trial centers now view as the customer, the industry appears to lag behind in its treatment of this population in clinical trials. Present informed consent documents and practices especially for those involving the cognitively impaired fail to adequately address the issue of customer identification and generally do not respect the cognitively impaired as the true customer but defer to the wishes of their family members or caregivers.
Directly related to this risk are confidentiality protection concerns. Although all protections flow from the seemingly simple provision of confidentiality in clinical trials (Brody, 2007), there have been egregious and alarming threats and compromises to this protection under recent past presidential administrations. It is important for the industry at large, and supporters of the adaptive trial design more specifically, to reclaim this protection.
Confidentiality, in a time of increased and nebulous knowledge transmission, provides the only means of ensuring patients’ protection of medical information and even consent to participate in clinical trials (The Republican, November 2, 2007). There are several recent examples of the perils of lack of protection that underscore the importance. For instance, recent mishandling of both Social Security and Administration of Veterans Affairs electronic data underscore the importance of protections additional to a generally accepted view that materials is “sensitive.” Further, practitioners have recently reported considerable concern about and difficulty implementing the 1996 Health Insurance Accountability and Portability Act (HIPAA) (Modern Healthcare, August 14, 2007). Because no present governmental or industry language explains how the tenets of this act ought to be practically applied and their relationship to the confidentiality agreement, these personal medical history records are essentially left vulnerable to mishandling. In short, there is no control of these data, which given today’s high-tech and global environment, the author contends, are vulnerable to piracy and adulteration by unscrupulous, enterprising individuals.
Another more contemporaneous example can be found in the use of different workspaces themselves. The renovated locations in which practitioners, researchers and clinical trial staff work offer increased opportunities for violation of confidentiality protections. The increasingly technological environment in which we work permits the free exchange of materials via high-tech computers that make our lives very convenient. What practitioners and others may not fully understand is that there is at present no protection of the writings and work generated on these free range computers, either desktop or laptop varieties, which are susceptible to electronic piracy, an unregulated activity that nets approximately $200 billion annually (GAO, 2007; Montoya & Jano, 2007). If unscrupulous pirates wished to snatch materials electronically, there would be no protections afforded to and little legal expertise available for recouping the snatched. In today’s liquid intellectual environment, the only legal protection afforded public or private library users hinges on the confidentiality in medical records account activity. This confidentiality provides protection—arguably minimal, it is true—against the theft of their personal data. Until legal system advances begin to keep pace with the increasingly fluid and extralegal aspects of knowledge production, patient confidentiality provide the only means of protecting our work.
There are alternatives to adaptive design— ones that portend to provide increased efficiencies for the clinical trial workforce—which go beyond the scope of this article. Briefly, one possibility is to modify the adaptive design model to be less business-like in its efficiency. As argued above, this would allow for inclusion of behaviors and subjects that defy rational research methods but retain the rigor of these methods. Another possibility is to move clinical research in the direction of phenomenological research, a model which is sorely lacking in all areas of science (Yow, 2005), not merely clinical research. The benefit of this model is that it permits exploration of the experience of participation in clinical research, produces intriguing and expected outcomes that spur on further innovations in clinical research, and better meet the needs of clinical research subjects, particularly the cognitively impaired. Finally, a modified adaptive trial design model could provide enhanced patient consent and confidentiality protections. In this way the pharmaceutical industry would be leading the global health care environment as well as the other sectors of the U.S. economy, for it is the author’s belief that widespread implementation of adaptive trial design without proper address of problems outlined here will be destructive not only to the industry but to all engaged in the management of health-related knowledge.
References
Brody, H. (2007). Hooked. New York: Rohman & Littlefield Publishers, Inc.
DiIulio, J.J; Garvey, G.& Kettl, D.F. (1993). Improving Government Performance: An Owner’s Manual. Washington, D.C.: The Brookings Institution.
Higgins, J. (2001). Comprehensive Policy Analysis of and Recommendations for Senior Center Gambling Trips. Journal of Aging and Social Policy, 12:2.
Higgins, J. (2003). Analysis of Gambling Activities as a Programmatic Option in Senior Centers. (Unpublished Doctoral Dissertation).
Krams, M. & Robbins, H. (2007). Adaptive Clinical Trial Design. The Monitor, 21: 6.
Modern Healthcare. August 14, 2007. “HIPAA has become 'bureaucratic nightmare'.”
Montoya, I.D. & Jano, E. (2007). Online Pharmacies: Safety and Regulatory Considerations. International Journal of Health Services, 37: 2.
National Institutes of Health. Science Meets Reality: Recruitment and retention of women in clinical studies, and the critical role of relevance. January 6-9, 2003.
U.S. General Accounting Office. Intellectual Property: Risk and Enforcement Challenges. October 18, 2007.
The Republican. November 2, 2007.
Yow, V.R. (2005). Recording Oral History. California: AltaMira Press.
