Can research and care: be ethically integrated?
Subject: Medical ethics (Research)
Medical research
Medicine, Experimental
Authors: Largent, Emily A.
Joffe, Steven
Miller, Franklin G.
Pub Date: 07/01/2011
Publication: Name: The Hastings Center Report Publisher: Hastings Center Audience: Academic; Professional Format: Magazine/Journal Subject: Biological sciences; Health Copyright: COPYRIGHT 2011 Hastings Center ISSN: 0093-0334
Issue: Date: July-August, 2011 Source Volume: 41 Source Issue: 4
Topic: Event Code: 310 Science & research
Product: Product Code: 8000200 Medical Research; 9105220 Health Research Programs; 8000240 Epilepsy & Muscle Disease R&D NAICS Code: 54171 Research and Development in the Physical, Engineering, and Life Sciences; 92312 Administration of Public Health Programs
Accession Number: 268403465
Full Text: Medical ethics assumes a clear boundary between clinical research and clinical medicine: one produces knowledge for the benefit of future patients, while the other provides optimal care to individuals right now. It also assumes that the two cannot be integrated without sacrificing the needs of the current patient to those of future patients. But integration could allow us to provide better care to everyone, now and in the future.

Historically, clinical medical research and clinical medical care were viewed as similar and as intertwined. Physicians considered research on patients to be an integral part of the care they delivered. (1) After all, the good physician constantly made observations and formulated and tested hypotheses; he turned to the literature and experimented with different treatments to find the one most efficacious for the patient in front of him. Moreover, he learned from these experiences to improve his treatment of future patients. On this model, both medical care and research were scientifically guided, therapeutically oriented activities conducted within the physician-patient relationship. (2)

In 1979, however, the Belmont Report disentangled research and care, defining clear boundaries between the two. (3) Central to the distinction was the idea that the purpose of clinical research is fundamentally different from that of clinical medicine: whereas medical care focuses on providing optimal care to individual patients, clinical research is primarily concerned with producing generalizable knowledge for the benefit of future patients. Other defining characteristics of clinical research include distinctive research methodologies--such as placebo controls, randomization, and blinding--that are anathema to medical care because they sacrifice personalization of care in order to minimize bias; the use of procedures that hold no prospect of medical benefit for the research participant but whose associated burdens and risks are justified in light of their scientific value; and a unique investigator-subject relationship that is best understood in contrast with the physician-patient relationship. (4)

Thirty years after Belmont, the sharp distinction between research and care is becoming increasingly blurred. Medical care and clinical research are extensively, deliberately, and routinely integrated in numerous settings. Significantly, this integration is occurring not just within the physician-patient relationship, but also at the level of health care systems. For example, use of electronic medical records within the Veterans Administration has furthered evidence-based practice and research by allowing clinical data to be aggregated for health services research. (5) Practice-based research networks, such as the "primary care practice-based research networks," hold the potential to enable the performance of randomized, controlled trials and other types of research in primary care offices nationwide. (6) This is a situation for which there is no historical precedent.

To a great extent, recent technological advances motivate and enable research-care integration. Health information technology, particularly electronic medical records, provides researchers access to patient data in a more comprehensive and less labor-intensive manner than ever before. It facilitates hypothesis formulation and testing, as well as the evaluation of millions of patients' experiences nearly in real time. (7) Cultural shifts, such as the growing appeal of evidence-based medicine and comparative effectiveness research, have also motivated research-care integration. Finally, there is perceived social value in integration. Integration offers an opportunity to improve medical care, to contain health care costs without sacrificing quality, to remove many barriers to effective research, and to increase patient satisfaction by enhancing access to promising treatments during periods of evidence development.

In 2007, the Institute of Medicine Roundtable on Evidence-Based Medicine wrote: "We must quicken our efforts to position evidence development and application as natural outgrowths of clinical care." (8) With its emphatic "we must," this statement implies that there may be an obligation to integrate research and care. Yet a systematic argument identifying the basis and scope of an ethical obligation to integrate research and care has not been advanced. Although research can provide care for an individual that is good, even optimal, some aspects of research participation are often not in an individual's best interests because of the loss of personalization and the inclusion of procedures that are not beneficial to the individual. The tension between patients' individual interests and the aim of research to promote public good must therefore be satisfactorily resolved. Furthermore, if the integration of research and care is desirable to the extent that it is feasible, many ethical questions arise regarding the design of a maximally integrated system. Depending in part on how it is accomplished, integration can challenge conventional understandings of both research and care. It is increasingly apparent that existing conceptual and ethical tools to account for this are underdeveloped.

In what follows, we imagine a plausible setting in which research and care are integrated in ways that build on and transcend existing practices. This thought experiment is not intended to outline a model of research-care integration that would be readily accepted in the marketplace; rather, the thought experiment allows us to explore why research-care integration is ethically plausible and to explore the ethical questions that arise in such a maximally integrated environment. Our objective is twofold: to show that the ongoing integration of research and care is both justified and desirable and to illustrate that systematic integration requires the rethinking of concepts, tools, and regulations familiar to clinical medical and research ethicists. The thought experiment can be set up so that it characterizes either one health plan within a decentralized system of health plans, as in the United States, or as a universal, single-payer, national health service. Although we focus primarily on developing and justifying the former model, we will briefly discuss the way in which the latter also might be justified. The thought experiment is unlikely to be feasible in either form, but the ethical considerations that we address are applicable to less systematic efforts at integrating research and care.


The Research-Integrated Health System is a hypothetical full-service, not-for-profit integrated delivery system consisting of a network of community clinics and hospitals modeled after the Kaiser system. The clinics and hospitals are staffed by salaried employees. The RIHS mission is to integrate care and research thoroughly in order to provide high-quality, evidence-based medicine to current and future patients and to contribute to the good of both the enrolled population and the broader public through evidence development. The RIHS model is predicated on the philosophy that outcomes for patients collectively are optimized when the practice of medicine is based on current knowledge and that research participation can provide good, even optimal, treatment for a given patient, depending on the design of the trial and the patient's other treatment options. The emphasis on research-care integration is clearly spelled out in the RIHS charter and also in the informational literature provided to potential and current enrollees.

In order to fulfill its mission, the RIHS seeks to achieve integration of care and research in two primary ways. First, it is a learning health system. (9) As is already occurring in some integrated health systems, such as the VA, clinically relevant data from all patient encounters within the RIHS, including diagnostic assessments and treatment interventions, are entered into electronic medical records and systematically analyzed. The experiences of thousands of patients within the RIHS network are continually examined to learn from their experiences and to identify advantageous changes in patient care. Some of these data-driven activities, such as safety monitoring, quality assurance, and policy analysis, are generally not considered human subjects research, but many such activities fall under the current human subjects research framework. Medical records are also used to conduct patient-oriented observational research, to develop research questions, and to prioritize research projects within the RIHS. No informed consent is needed for this medical records research, as this is transparently made a condition of health plan participation. Second, an important innovation in RIHS is that patient care is routinely delivered in the context of clinical trials, including randomized, controlled trials, as described in detail below. Currently, systematic integration of clinical trials with routine care occurs only in narrow medical niches--most notably, pediatric oncology. The results of these trials are used as the basis for developing the evidence-based care guidelines followed by RIHS clinicians and are regularly published in peer-reviewed journals to contribute to the general improvement of medical care.

Prescribed and invited trials. There are two kinds of clinical trials within the RIHS: "prescribed" and "invited." Trials qualify for prescribed status if they are judged to be compatible with optimal medical care in light of current knowledge and if associated research procedures pose no more than minor net risks, compared with usual care. (10) Minor net risks are analogous to the U.S. regulatory category of "minimal risk": the probability and magnitude of harm or discomfort are judged to be no greater than what average, healthy people normally encounter in daily life or in routine physical or psychological examinations. (11) A minor net risks ceiling limits the incremental risks of care delivered within the context of a prescribed trial, over and above those of typical care. Prescribed trials include randomized, controlled trials designed to answer clinically important questions regarding the relative merits of two or more medically indicated treatments, such as antibiotic drugs to treat meningitis or packages of treatment interventions to treat low back pain. Prescribed trials might also include combination versus monotherapy with standard treatments, as well as selected randomized or single-arm trials evaluating novel therapeutics for conditions without validated treatments (see the sidebar, "Examples of Prescribed Trials"). When such a trial is enrolling participants in the RIHS, the default is that physicians prescribe trial enrollment to eligible patients rather than recommending a preferred treatment.

Given that treatment within prescribed trials occurs routinely and that this policy is transparent to RIHS enrollees, there is no need for a full, research-specific process of informed consent for these trials. Rather, physicians engage patients in the less formal consent process that (with notable exceptions) typifies medical care. Patients can refuse trial participation, just as they can refuse conventional medical treatment; however, if a prescribed trial is enrolling patients and a given patient meets eligibility criteria but refuses enrollment, treatment outside the trial is reimbursed but with a substantially higher copayment. If a patient initially agrees to participate in a prescribed trial but later chooses to withdraw, she will be required for the duration of the trial to pay a substantially higher copayment for standard treatment for the condition under investigation, or her copayment may be prorated, depending on how long she participated in the trial. For those who participate in research, some or all of the copayment associated with a prescribed trial would be waived as an incentive and a token of appreciation.

A second kind of trial is the "invited" trial. Invited trials include those for which there is greater uncertainty about risks and benefits of treatments under investigation, as well as those that involve more than minor net risks to participants from research interventions that are uncompensated by a prospect of direct benefit (see the sidebar, "Examples of Invited Trials"). Trials comparing two medically indicated treatments that differ substantially in invasiveness, burden, or type of side effects and many randomized, controlled trials of investigational or off-label agents or procedures are also included in the invited category. New treatments may be offered provisionally through invited trials while the RIHS collects additional information on safety and efficacy to determine if the treatment will eventually be incorporated into the member benefits package. Conventional, research-specific informed consent is obtained for invited trials, and patients are free to decline participation without penalty and to receive standard, validated treatment (if available) as a reimbursed alternative. RIHS physicians, however, must discuss the availability of invited clinical trials with eligible patients and encourage them to participate.

Organizational oversight. To foster clinically relevant research, optimally integrate research findings into practice, and protect the interests and rights of patients, the RIHS boasts a multifaceted system of research review and oversight. Some elements of this system will be familiar, and others are innovative. The RIHS has chartered a research advisory committee that is responsible for overseeing the RIHS's clinical research portfolio. This committee periodically reviews the portfolio to ensure that it addresses questions of importance to the membership and is responsive to members' medical needs, that the research resources of the RIHS--financial, human, and other--are being used wisely, and that the benefits and burdens of the research activities are distributed equitably across the enrolled population. Furthermore, the research advisory committee conducts research to evaluate the quality of research studies completed and patient satisfaction throughout and after participation in trials. The committee chair, as well as a majority of members, are RIHS enrollees selected democratically by other members. (12) The research advisory committee ensures community accountability for research performed within the RIHS and provides the public perspective on challenging issues facing the RIHS. In this respect, it is similar to the Citizens Council of the U.K.'s National Institute for Health and Clinical Excellence.

An RIHS science review committee, comprised of physicians, scientists, statisticians, academics, and other researchers, is charged with initially evaluating the clinical and scientific merit of all intervention studies conducted with RIHS patients. Although some members of the committee are selected from within the RIHS, others are independent consultants. Trials can either be RIHS-sponsored studies initiated by RIHS investigators, or suitable clinical trials funded by public or private sources, including pharmaceutical, biotechnology, and device manufacturers. Studies that meet established criteria of scientific merit and clinical relevance are forwarded to the RIHS institutional review board for review.

The RIHS has an established IRB that includes robust representation from among the group of enrollees. The IRB reviews and approves protocols for all trials conducted with RIHS patients and decides whether a given trial will be classified as prescribed or invited on the basis of net risks, burdens, and scientific uncertainty. The IRB also assumes responsibility for determining which of the clinical trials warrant formal ongoing monitoring by the RIHS data safety and monitoring board, which is charged with protecting the well-being of trial participants and the scientific integrity of trials. Observational research using the electronic database is part of an omnibus protocol, amended at various times as new studies are proposed by RIHS investigators. Careful protections of privacy and confidentiality are instituted for these studies, and the RIHS charter stipulates that information unearthed in the course of research cannot be used to deny coverage. For example, genetic tests conducted during research cannot lead to genetic discrimination.

The RIHS clinical oversight committee is responsible for reviewing data from observational studies and clinical trials conducted within RIHS and elsewhere. The clinical oversight committee then determines which results are clinically significant and how the findings should be incorporated into clinical care. The committee drafts clinical care guidelines so that all care delivered reflects the best available evidence. If a new treatment is found to be better (safer or more effective) than existing treatments in research conducted within the RIHS, the clinical oversight committee works to ensure that the treatment is incorporated into the RIHS benefits package. If the clinical oversight committee identifies research results as promising but not sufficient for changing practice, it can propose a study to the research advisory committee or to RIHS investigators and collaborators. The clinical oversight committee is primarily comprised of RIHS physicians and researchers, but the enrollees are also represented.

Finally, though it is not an oversight body per se, the office of patient and participant advocacy offers a variety of research-related services to RIHS enrollees, including explanation of research participant rights, including the right to withdrawal; verification of which trials are prescribed and which are invited; assistance in preparing research advance directives and assigning durable power of attorney for research; monitoring of informed consent for research; participation in oversight committees, including the IRB and the research advisory committee; and publication of a quarterly newsletter updating enrollees on ongoing research and findings. Members of the office of patient and participant advocacy--including lawyers, ethicists, social workers, and clinicians--are also available to respond to patient concerns and complaints that result from research participation. In particular, the office of patient and participant advocacy is the recipient of appeals from enrollees or from RIHS clinicians who wish to argue that for clinical or other reasons it is not appropriate for a particular patient to incur a financial penalty for choosing standard treatment rather than enrolling in a prescribed trial. The RIHS recognizes that there will be some valid reasons for otherwise eligible individuals to opt out of prescribed research participation or to withdraw from research without penalty, and the appeals process offers a procedurally fair process for identifying and acting upon these. Because participation in prescribed trials often occurs on a short timeline--for example, an antibiotic is prescribed for a respiratory infection that requires immediate treatment for the patient's comfort and well-being--appeals can be fast-tracked so that there is no interference with an individual's receipt of clinically indicated care.

Financial arrangements. Enrollees have a set insurance premium and modest copayments for physician visits and prescriptions. Copayments apply to all services that are standard of care. Copayments may be waived to incentivize or show appreciation for participation in a prescribed trial. As indicated above, substantially higher copayments are generally required for patients who refuse to enroll in prescribed clinical trails. Copayments typically do not apply to experimental or investigational services that are rendered in the course of an invited trial, though patients may be required to copay for services that are standard of care within that trial. The RIHS budget has allocated funds for research purposes; these are used, in part, to finance invited trials. As discussed above, the research advisory committee works to ensure that the resources of the RIHS are allocated appropriately. If any trial--prescribed or invited--is cosponsored by the RIHS and, for example, a device or drug manufacturer, the cosponsor is expected to offset the cost of that trial to the RIHS. Net revenue from external research sponsors is used to lower the plan's costs to enrollees.

All physicians affiliated with the RIHS are salaried employees. According to the terms of their employment and in order to limit conflicts of interest, they cannot receive recruitment bonuses when their patients enroll in clinical trials. Furthermore, they cannot have financial ties with industry sponsors of research conducted within the RIHS. Patients, however, may receive payment in exchange for participation in invited research, either to reimburse them for expenses incurred or occasionally to incentivize participation. The RIHS IRB reviews all payments to patients to ensure that they do not constitute an undue influence.

Is RIHS Desirable?

Research and care can be integrated in a number of ways, each of which can be imagined along a spectrum of increasing manipulation of patient care. As outlined above, the RIHS arguably constitutes the most systematic and thorough integration of research and care ethically conceivable. Would RIHS be a desirable innovation? A persuasive argument could be advanced to support such a mode of delivering medical care for two primary reasons: it systematically implements the concept of evidence-based medicine, and it advances socially valuable research.

Benefits from the patient perspective. An essential question in considering the ethical and practical merits of the RIHS is how outcomes for patients in such a system would compare with those of patients in traditional medical practice, in which physicians individualize treatment in light of the best available evidence. This is an empirical question that, in principle, could be answered if RIHS were a real model of medical care.

In thinking about the question, however, we should guard against biased presumptions. One might think that patients necessarily "sacrifice" the quality of their care by virtue of being treated within the context of research. (13) To be sure, such patients may have to accept some level of burden and perhaps discomfort from nonoptional care, such as an extra blood draw to gather data purely for research; they may be asked to accept random selection of treatment in some circumstances; or they may be expected to accept some reasonable level of risk for the benefit of developing knowledge. Whatever "sacrifice" of personalized care is involved in research-care integration does not, however, imply that patients will fare worse on clinically significant outcomes, as compared with standard medical care. Sonia Shah holds that "There is an inherent perversity in controlled clinical trials, which is that by relying on the contrast between two groups treated differently, one group must suffer worse outcomes than the other," (14) but strictly speaking, this is not true. Outcomes may be similar for both groups. But even assuming that clinically significant differences will often occur, this situation would be perverse only if we knew at the outset which of the compared treatments was better. If we did, though, there would be no point in conducting the trial. Moreover, standard medical care includes treatment alternatives. When physicians select a given treatment among a set of medically indicated options, their patients may fare better or worse than other patients who receive another standard option. Hence, differential outcomes also are inherent in routine medical care.

Indeed, we hypothesize that patients enrolled in RIHS would do better, at least in the long run, by participating in research aimed at improving medical care. Lamentably, many treatments that are proven effective are not incorporated in everyday patient care. It has been shown that only 14 percent of research findings filter down to everyday practice; for those that do, the process takes an average of seventeen years. (15) Implementation of evidence-based guidelines has been linked to improved patient outcomes across numerous specialties. More patients receive scientifically validated treatments when evidence-based medicine is utilized, and fewer are exposed to ineffective, unnecessary, or harmful treatments. RIHS is designed both to advance systematically evidence-based medicine through acquisition of evidence and to hasten implementation of evidence-based medicine at the patient-provider level.

Presuming that individuals enrolled in RIHS have outcomes that are at least as good as those of individuals receiving standard care, enrollees have access to other benefits. In the aggregate, utilizing evidence-based medicine within the RIHS may result in coverage of more beneficial services or in reduction of premiums by eliminating procedures that are harmful, ineffective, or inferior to others. Note that these benefits can accrue independent of cost-effectiveness analysis, a topic we have bracketed. Additionally, there is an increasing demand among the public for rapid access to medical innovations, but this must be balanced with the need for better information on the innovations' risks and benefits. (16) Invited trials offer RIHS practitioners and enrollees a means of achieving such a balance. Integrating research into care allows enrollees access to new, promising therapies during a period of evidence collection. Additional evidence can be used to determine if interventions meet the RIHS standards for coverage, providing a transparent process for enrollees to see and participate in coverage decisions.

Benefits from the social perspective. The most socially valuable advantage that can be anticipated from maximal research-care integration is the attainment of more generalizable knowledge to benefit future patients. In a system where research and care are coextensive, research becomes more effective. It has been reported that 72 percent of clinical studies are delayed more than one month due to unfilled enrollment. (17) More than one in five trials sponsored by the National Cancer Institute fail to enroll a single subject, and only half reach the minimum sample size needed for a meaningful result. (18) Prescription of trial participation accompanied by financial incentives will doubtless enable many important trials to take place. Because RIHS clinicians encourage eligible patients to participate in invited trials, these trials may also accrue subjects and reach completion more quickly. At present, researchers may aim for a smaller sample size than is scientifically desirable given the difficulties of recruiting large numbers of patients; RIHS offers a mechanism to achieve greater statistical power and enhanced generalizability. It is unethical to enroll patients in trials and expose them to research risks if scientifically valid, socially valuable data will not result. Therefore, RIHS offers a model to make the very conduct of clinical trials more ethical. (19)

Several additional benefits bear mentioning. First, the RIHS model will help counteract the double standard for evidence in our current system, under which new drugs can be licensed only on the basis of favorable randomized, controlled trials, whereas innovative invasive procedures can be introduced into practice on the basis of "clinical experience." The RIHS will not approve any treatments for standard medical care without adequate evidence. In addition, the population-based nature of study results from RIHS research may more accurately reflect what will happen in practice because the data will be less subject to selection and other biases due to differences in who has access to or decides to participate in the trials. Groups that have been historically underrepresented in clinical research, including children, pregnant women, and minorities, will be automatically included in observational research and will be more likely to participate in RIHS trials. Finally, patients routinely benefit from the research participation of others, and free riding occurs when an individual receives a benefit while refusing to share the burden of obtaining it. (20) In RIHS, the benefits and burdens of research will be spread more evenly throughout the population, reducing (though not eliminating) concerns about free riders who benefit from clinical research without contributing to its development. (21)

Is RIHS Justified?

Having outlined the benefits of research-care integration, it is clear that many significant benefits will accrue to individuals and to society when research and care are integrated maximally. Now, the essential task is to ask: Do these benefits outweigh the risks?

The risks of RIHS. If the integration of research and care is conducted according to the standards outlined above, RIHS poses few novel risks to patients. Medical risks, such as those encountered in a conventional care-centered environment, will remain the same or--more likely--will be reduced, owing both to quality improvement initiatives that seek to improve outcomes and reduce medical errors and to continuous observational and clinical research that will identify safer interventions. However, RIHS enrollees will confront some incremental research-related risks, burdens, and inconveniences. The physical and emotional risks of research procedures are worth highlighting. In keeping with the principles of ethical research, RIHS trials should seek to minimize risks to the extent possible while maintaining scientific integrity. Furthermore, the use of medical records raises many salient questions about privacy and confidentiality. These must be carefully managed in accordance with existing ethical guidance.

The acceptability of risks and benefits. We believe that the systematic integration of research and care, particularly as it relates to extensive use of evidence-based medicine, will result in improved patient outcomes. The incremental risks that RIHS enrollees obligatorily assume--that is, participation in prescribed trials and observational research--are minimal and will be readily outweighed by the benefits. For invited trials, patients will continue to be able to determine if the risk-benefit ratio is acceptable to them. RIHS offers clear, significant social benefits that outweigh the minimal net risks that enrollees will be required to assume. Not only will research be conducted more efficiently and have greater generalizability, but the problem of free riders who benefit from clinical research without participating will largely be resolved. Research-care integration according to the model of the RIHS is ethically justified.

Potential Objections

Maximal integration of research and care as envisioned here challenges our existing conceptual models of both clinical research and clinical medical care and raises many important ethical questions. Therefore, a number of objections may be anticipated.

Exploitation. Prescribed trials challenge the conventional ethical prescription that participation in research should be voluntary--enrollment is the participant's choice alone. An obvious concern is that linking care to research participation, as occurs within RIHS, is exploitative. According to the going understanding, exploitation occurs when A uses B as a mere means or when A takes unfair advantage of B. (22) Neither of these definitions seems applicable in the context of the RIHS. Enrollment information for prospective RIHS enrollees transparently communicates the integration of medical care and research. Patients who do not want to participate in such a plan can choose other conventional health plans. But even if there were no alternatives to an RIHS-type health plan (a possibility considered below), we contend that patients would not be exploited by this arrangement. Patients are not used as a mere means in prescribed trials; rather, they have a reasonable expectation of clinical benefit stemming directly from participation in the trial. Furthermore, because prescribed trials must not exceed a modest net risk threshold, as determined by the IRB, and because a robust appeals process exists to address legitimate exceptions, the RIHS does not take unfair advantage of patients. Patients are not being asked to make excessive sacrifices for others. Because of the constraints placed on prescribed trials to make them hew more closely to standard clinical care, the research-care distinction fades and obviates this concern.

Coercion. Prescribed trials also challenge the conventional notion that research participants should be able to decline enrollment in a study or to withdraw from a study at any time without penalty. On the RIHS model, prescribed trials cannot be declined without repercussions because treatment outside prescribed trials requires a significant copayment. A charge of coercion--threatening to deprive someone of something to which they have a right--may seem relevant in these circumstances.

Whether participation in prescribed trials is incentivized by reducing copayments for participants or by substantially increasing copayments for those who refuse to participate, individuals are "losing" money--though, admittedly, there is a strong framing effect that distinguishes the two. Yet it is not clear that RIHS enrollees have a right to receive standard care outside a prescribed trial at the customary expense. At the time of enrollment, it is made clear that in some cases, patients will have a right to treatment at low or no cost only within a prescribed trial. If individuals consent to this model of care, they are not being coerced within any particular trial. If RIHS is the only choice, as in a single-payer health care system in a democratic society, this implies a legitimately enacted societal determination that the right to care at customary (or no) expense entails a reciprocal responsibility to participate in certain classes of research. Such responsibility cannot reasonably be described as coercive. Furthermore, prescribed trials are consistent with the standard of medical care; no patients are required to participate in trials that involve withholding standard care (such as placebo-controlled trials of new treatments for conditions that have proven effective treatment options) or that involve more than minor net risks. Thus, patients are not coerced within trials. Finally, patients may choose to receive standard care outside of the trial by paying a higher copayment, while those who believe that for individual reasons they or their patient should not be required to join a prescribed trial will have access to a procedurally fair appeals process for adjudicating such claims.

Distortion of the physician-patient relationship. The relationship between physicians and patients in the RIHS differs profoundly from the relationships that are characteristic of medical care. The traditional view, widely endorsed, is that physicians should always act in the best interests of the present patient. Although exceptions to this view are extensively acknowledged, little systematic guidance has been offered to assist physicians in identifying which exceptions are appropriate and how they should take into account the interests of others when providing patient care. (23) This lack of ethical guidance will be brought into stark relief within the RIHS. RIHS physicians will routinely collect data for observational research and bring the interests of third parties to bear on the physician-patient encounter. Moreover, treatment will be provided--indeed, prescribed--within clinical trials and routinely depart from the ideal of personalized care. The overlap of research and medical duties has yet to be justified. (24) One should not assume, however, that the conventional Hippocratic model is best. RIHS offers, on balance, a system in which the benefits for patients and the wider society outweigh the disadvantages. Although they do not completely evade the potential for moral conflict between dual loyalties to research and care, RIHS physicians are able to proceed in good conscience, recognizing that their patients have chosen to receive medical care in the context of research and have given fully informed consent.

From Thought Experiment to Application

The purpose of this thought experiment is to show that integration of research and care is ethically plausible--not to show that a health plan like RIHS would succeed in the marketplace. Still, it is important to acknowledge that few individuals--if given a choice between RIHS or a plan with a comparable commitment to implementation of evidence-based medicine but without the prescribed research component--would choose to join RIHS. It would be unreasonable to appeal to altruism alone; rather, individuals must have a personal interest in choosing research-integrated care. It is possible that the RIHS might have better patient outcomes than any alternative, given the potential clinical benefits that flow from trial participation and the strong institutional commitment to guiding care by means of research; however, it would be difficult to demonstrate this rigorously. In the absence of clear and convincing evidence of superior outcomes, it seems unlikely that many people would opt for RIHS.

This has two implications worthy of further consideration: 1) maximal research-care integration may only be achievable within a universal health system, and 2) although maximal integration is unlikely to be feasible in a fragmented system driven by consumer choice, such as exists in the United States, research-care integration is nevertheless justified and desirable.

Fragmented system. Although there are barriers to achieving maximal integration of research and care in a fragmented health system, integration is still being pursued, as many innovative models attest. For example, with its network of clinics and hospitals in communities across Iowa, Wisconsin, and Minnesota, the Mayo Health System is in many ways similar to the RIHS. The esteemed Mayo Clinic had a research budget of $523 million in 2008, and roughly 80 percent of Mayo physicians are involved in research. (25) Importantly, the Mayo Health System does not provide insurance to individuals and therefore does not condition coverage on participation in research. Without this tool, what incentives could be used to draw research and care closer together and to foster a culture of shared responsibility, in which developing evidence is a cause common to health care professionals and the public? Presumably, many individuals choose Mayo because it offers access to promising new therapies and favorable health outcomes driven by application of evidence-based medicine. Other ways to integrate research and care include large, simple trials with more flexible inclusion/exclusion criteria; the Center for Medicare and Medicaid's system of coverage with evidence development; and primary care practice-based research networks. (26) We should continue to pursue methods like these.

Universal health system. The RIHS model of research-care integration is justified most easily when enrollees have a choice among health plans. Yet we believe a national health service (single-payer system) or a government insurance plan in which care was modeled on the RIHS could also be ethically justified. Medical records research can be ethically conducted without informed consent. (27) Similarly, participation in registries and trials as a condition of coverage already occurs in some contexts--for example, Medicare's program of "coverage with evidence development"--and has been found ethically acceptable. (28) As for invited trials, enrollees will participate only if they give fully voluntary informed consent. Thus far, a single-payer system modeled on the RIHS appears unproblematic; however, the inclusion of prescribed trials would be without precedent. Without a choice among health plans, consent to enroll cannot do the same moral work of advance authorization for participation in prescribed trials. How, then, might "mandatory" participation in interventional research be justified?

There are reasons to believe that a national health service modeled on the RIHS thought experiment could be justified along the lines of a hypothetical social contract. Individuals behind a Rawlsian "veil of ignorance," lacking knowledge about their health status but interested in receiving optimal medical care, arguably would choose the RIHS model as the basis for organizing their health system. (29) Posing a question to a "prospective citizen," Daniel Wikler asks which of two societies you might prefer to join:

Implemented as a single-payer system, the RIHS is comparable to the second society. While tradeoffs exist, the benefits of the second society seem clearly to outweigh the associated burdens. The state would provide universal care with the requirement that care be delivered within a context that maximally develops knowledge for the good of all and does not impose any more than minimal net risks on individuals. A single-payer system modeled on the RIHS would apportion the burdens and benefits of clinical research participation fairly within the population. Moreover, a single-payer system would do more to reduce free riders by limiting the possibilities for individuals to benefit from evidence-based medicine without contributing to it.

The RIHS thought experiment shows that a large, integrated health care system can combine research and care to a much greater extent than is already occurring and in a way that advantages present and future patients individually and collectively, without violating fundamental rights. Moreover, these benefits to the individual and to the collective outweigh any novel risks introduced when clinical research is conducted concurrently with the delivery of care. For these reasons, further integration of research and care is a worthy goal of health system design.


The opinions expressed are those of the authors and do not reflect the position or policy of the National Institutes of Health, the Public Health Service, or the Department of Health and Human Services.

Examples of Prescribed Trials

* Mrs. Jonsen visits her primary care provider, Dr. Chen, because she has a chronic cough. A chest x-ray confirms that she has pneumonia and requires antibiotics. Dr. Chen tells Mrs. Jonsen that the RIHS is currently conducting a clinical trial comparing two Food and Drug Administration-approved antibiotics to see which is more effective in treating pneumonia. The trial is not blinded, but it is randomized. Therefore, Dr. Chen calls the pharmacy, and the research pharmacist tells Dr. Chen which of the two antibiotics to prescribe for Mrs. Jonsen. Mrs. Jonsen is informed about the benefits and risks of the selected treatment in the informal manner of routine drug prescription.

* Billy and his mother go to Dr. Balch, Billy's pediatrician, to have Billy vaccinated for seasonal flu. Dr. Balch informs Billy and his mother that RIHS is conducting a trial to compare the effectiveness of an inhaled form of the flu vaccine to a traditional injection in children. Dr. Balch tells Billy's mother that randomization has occurred at the physician level and that all of the patients in his practice are receiving the flu shot. Billy receives the shot.

* Mr. Ross has recently been diagnosed with seasonal affective disorder. Dr. Stevens, Mr. Ross's psychiatrist, provides information on photo-therapy and other behavioral modifications to improve mood. She also tells Mr. Ross that if he is interested in medication, he must participate in a double-blinded, placebo-controlled trial evaluating the effectiveness of antidepressants in correcting seasonal affective disorder. Dr. Stevens reiterates that neither she nor Mr. Ross will know which treatment Mr. Ross is receiving, and Mr. Ross may receive a placebo. Dr. Stevens reminds Mr. Ross that, under the terms of his health plan, an antidepressant is not available outside the trial unless he chooses to incur a higher copay. Mr. Ross declines to participate in the trial, stating his preference to try nonpharmacologic interventions first.

Examples of Invited Trials

* Ms. Richardson recently injured her elbow playing tennis. Dr. Smith, her orthopedist, suggests conservative treatment--rest, ice, an elbow brace, and a prescription anti-inflammatory. Ms. Richardson agrees to this plan of treatment. Before Dr. Smith writes the prescription, he tells Ms. Richardson that she can receive a Food and Drug Administration-approved anti-inflammatory or she can participate in a double-blind, randomized, controlled trial comparing the approved anti-inflammatory to a promising new anti-inflammatory. She expresses interest in participating, gives her informed consent, and obtains her prescription from the research pharmacist.

* Dr. Tate, a retired rheumatologist, has been suffering from osteoarthritis for several years. He has now made an appointment with Dr. Smith to determine if he would benefit from knee surgery. Dr. Smith invites Dr. Tate to participate in a trial comparing outcomes from a widely used but never validated knee surgery to a sham surgery intervention. Dr. Tate, appreciating the need for rigorous evaluation of the surgical procedure and the modest risk of the sham surgical intervention, agrees after careful consideration to enroll in the trial.

* Mrs. Williams is pregnant with her first child. The baby has been diagnosed via sonogram with a heart defect requiring surgical correction. Dr. Reynolds, the obstetrician, invites Mrs. Williams to participate in an RIHS study examining health outcomes after in utero heart surgery. Dr. Williams assures Mrs. Williams that her child can have reimbursed open-heart surgery after birth regardless of whether she participates in the trial. Mrs. Williams asks for more time to discuss the trial with her husband.


Steven Joffe's work was supported by a faculty scholar award from the Greenwall Foundation.


(1.) F. Lemaire, "Patient Care versus Research: Does Clinical Research Provide Individual Benefit to Patients Enrolled in Trials?" Current Opinion in Critical Care 10, no. 6 (2004): 565-69.

(2.) F.G. Miller and D.L. Rosenstein, "The Therapeutic Orientation to Clinical Trials," New England Journal of Medicine 348 (2003): 1383-86.

(3.) The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research (Washington, D.C.: Department of Health, Education, and Welfare, 1979).

(4.) F.G. Miller, "Revisiting the Belmont Report: The Ethical Significance of the Distinction between Clinical Research and Medical Care," APA Newsletter 5, no. 2 (2006): 10-14; S. Joffe and F.G. Miller, "Bench to Bedside: Mapping the Moral Terrain of Clinical Research," Hastings Center Report 38, no. 2 (2008): 30-42.

(5.) Institute of Medicine Roundtable on Evidence-Based Medicine, "The Learning Health Care System Workshop Summary," paper presented at The Learning Health Care System, Washington, D.C., 2006.

(6.) W.M. Tierney et al., "A National Survey of Primary Care Practice-Based Research Networks," Annals of Family Medicine 5, no. 3 (2007): 242-50.

(7.) J. Slutsky, "Moving Closer to a Rapid-Learning Health Care System," Health Affairs 26, no. 2 (2007): w122-w124.

(8.) Institute of Medicine Roundtable on Evidence-Based Medicine, "The Learning Health Care System Workshop Summary."

(9.) Ibid.

(10.) D. Wendler and F.G. Miller, "Assessing Research Risks Systematically: The Net Risks Test," Journal of Medical Ethics 33 (2007): 481-86.

(11.) U.S. Department of Health and Human Services, Protection of Human Subjects, 45 CFR 46, Subparts A-D.

(12.) E.J. Emanuel, The Ends of Human Life (Cambridge, Mass.: Harvard University Press, 1991).

(13.) S. Hellman and D. Hellman, "Of Mice but Not Men: Problems of the Randomized Controlled Trial," New England Journal of Medicine 324 (1991): 1585-89.

(14.) S. Shah, The Body Hunters (New York: The New Press, 2006), 83.

(15.) Tierney et al., "A National Survey of Primary Care Practice-Based Research Networks."

(16.) S.D. Pearson, F.G. Miller, and E.J. Emanuel, "Medicare's Requirement for Research Participation as a Condition of Coverage: Is It Ethical?" Journal of the American Medical Association 296 (2006): 988-91.

(17.) Center for Information and Study on Research Participation, "101 Facts about Clinical Research," information/documents/101 factsaboutcli nicalresearch.pdf (accessed September 21, 2009; document no longer available on site); N.S. Sung et al., "Central Challenges Facing the National Clinical Research Enterprise," Journal of the American Medical Association 289 (2003): 1278-87.

(18.) G. Kolata, "Lack of Study Volunteers Hobbles Cancer Fight," New York Times, August 2, 2009.

(19.) E.J. Emanuel, D. Wendler, and C. Grady, "What Makes Clinical Research Ethical?" Journal of the American Medical Association 283 (2000): 2701-2711.

(20.) G.O. Schaefer, E.J. Emanuel, and A. Wertheimer, "The Obligation to Participate in Biomedical Research," Journal of the American Medical Association 302 (2009): 67-72.

(21.) T. Almy, "Meditation on a Forest Path," New England Journal of Medicine 297 (1977): 165-67; F. Mosteller, "Innovation and Evaluation," Science 211 (1981): 881-86.

(22.) A. Wertheimer, Exploitation (Princeton, N.J.: Princeton University Press, 1996).

(23.) D. Wendler, "Are Physicians Obligated Always to Act in the Patient's Best Interests?" Journal of Medical Ethics 36 (2010): 66-70.

(24.) Ibid.; C. Weijer and P. Miller, "Therapeutic Obligation in Clinical Research," Hastings Center Report 33, no. 3 (2003): 19-28.

(25.) Research at Mayo Clinic,, accessed June 2, 2010.

(26.) Tierney et al., "A National Survey of Primary Care Practice-Based Research Networks."

(27.) F.G. Miller, "Research on Medical Records without Informed Consent," Journal of Law, Medicine, and Ethics 36, no. 3 (2008): 560-66, at 567.

(28.) Pearson, Miller, and Emanuel, "Medicare's Requirement for Research Participation as a Condition of Coverage: Is It Ethical?"

(29.) J. Rawls, A Theory of Justice (Cambridge, Mass.: Harvard University Press, 1971).

(30.) D. Wikler, "Ethical Considerations in Randomized Clinical Trials," Seminars in Oncology 8, no. 4 (1981): 437-41.
In the first, doctors always give
   their patients the best care they
   can, while in the second patients
   are sometimes slighted in the interest
   of medical progress. The state
   of the art of medicine, however, is
   more advanced in the second than
   in the first.... Patients in the second
   society lack some of the feeling
   of security that comes with the
   belief that one's doctor will always
   do what he or she believes best for
   his or her patient; those in the first
   society have less of the security
   that one's doctor's beliefs about
   treatment are sound. (30)
Gale Copyright: Copyright 2011 Gale, Cengage Learning. All rights reserved.