The clinician-researcher, an individual who both conducts research and provides direct services (also referred to as the clinical scientist and the scientist-practitioner), is considered to be an important figure in health research. It is often stated that the field would stagnate without the involvement of researchers who have direct clinical experience with the health conditions and service systems being studied (1,2). Several recent articles have discussed the declining number of clinically trained individuals in health and behavioral health research and the need for new initiatives to change this pattern (2,3,4,5,6). Although most have focused on the declining number of physicians in research, the clinical psychology literature has also noted this trend (7,8). Several discussions have specifically highlighted the need for more clinician-researchers to conduct patient-oriented research, which is defined as research that involves direct contact with human participants (1,6,9,10).
At the same time, some have expressed concern over the ethical challenges posed by the involvement of clinicians in research that includes human participants (11,12,13,14). Much of this discussion has focused specifically on financial conflicts of interest that often arise in the context of research; however, equally important are conflicts of interest that arise as a result of what has been described as the clash in agendas between the clinician and researcher roles (12,13,14).
In this Open Forum, we critically examine the advantages and challenges of the role of the patient-oriented clinician-researcher at the interface between behavioral health treatment (the service arena that includes both mental illness and addiction treatment) and behavioral health research. We define the patient-oriented clinician-researcher as an individual who both provides behavioral health services and conducts research involving human participants that is related to the provision of such services. We focus on the interface between research and treatment because we believe that this is precisely the arena that poses both the greatest potential for the patient-oriented clinician-researcher to make a special contribution and the greatest risk for ethical conflict.
It is not necessarily self-evident that individuals with clinical training and active clinical involvement can make an "essential" (2) contribution to behavioral health research. Currently, a majority of researchers funded by the National Institutes of Health (NIH) and the National Institute of Mental Health have Ph.D. degrees and are predominantly nonclinicians (6,9). These individuals, who come from nonclinical disciplines such as sociology and epidemiology, typically have stronger training in research methodology, statistics, and theory than physicians, clinical psychologists, nurses, and social workers. If "good research" is defined as conforming to a classic model (14) in which existing theories are tested by using sound research methodology, it stands to reason that those with stronger backgrounds in theory and research methods would be better candidates to conduct research.
A compelling counterargument, however, is that in an applied field such as behavioral health, researchers without clinical experience or any direct exposure to the clinical phenomena or service systems they study necessarily miss many of the real-world issues that can inspire innovative and relevant research (2,7,15); a historical example is the study of substance abuse among people with severe mental illness, which was not studied until clinicians raised the issue (16,17).
In behavioral health research, direct work experience in the behavioral health system can provide insights beyond what has already been discussed in the literature that can inform several lines of research—for example, how the "black box" of managed care influences direct service delivery (18). Other questions will arise out of a direct understanding of what is most relevant to those who conduct treatment in routine settings, such as the types of psychotherapy that would be most useful to practitioners.
Another area in which patient-oriented clinician-researchers can make a major contribution is in the practical improvement of coordination between research teams and research sites. The importance of this type of coordination has been recently recognized in the substance abuse field: the National Institute on Drug Abuse and the Center for Substance Abuse Treatment have established a number of initiatives designed to encourage collaboration between researchers and practitioners to facilitate the conduct of studies that are relevant to routine clinical practice in real-world settings (19). Patient-oriented clinician-researchers may be uniquely qualified to anticipate problems in communication and coordination that can arise when research studies are conducted in these settings. They may also have greater credibility when explaining a study to and negotiating with clinical staff who will function as essential facilitators or inhibitors of the research study. Thus they can serve as a "bridgers" (20) between the research and practice worlds.
The patient-oriented clinician-researcher can also make an important contribution to the quality of clinical services, because involvement in research and clinical practice can facilitate an interactive flow of ideas between the clinical and research domains.
Specifically, the patient-oriented clinician-researcher can make an important contribution to clinical service quality by facilitating the dissemination and implementation of evidence-based treatment approaches. Currently there is much discussion of the lack of implementation of evidence-based treatment approaches in routine mental health and addictions treatment settings (21,22,23). Corrigan and colleagues (22) highlighted the role of barriers to implementation of such approaches, which include resistance from line staff and administrators because they do not understand research, as well as a lack of user-friendly manuals that allow easy translation of empirically tested approaches into real-world settings. Patient-oriented clinician-researchers with active ties to clinical settings can be a strong force in breaking this impasse; they will not be unknown entities to line staff and administrators and will not be easily dismissed as "academics" who "do not understand what we do." They can help to communicate the importance of evidence in terms that staff and administrators understand and value. Similarly, patient-oriented clinician-researchers are better aware of the challenges involved in implementing empirically validated treatments into routine settings and can help develop dissemination strategies that address these challenges.
Several challenges are associated with integrating the clinician and researcher roles in patient-oriented research. In general, the challenges created by the combined clinician-researcher role arise from a clash in agendas, tasks, and "ways of being" that create practical and ethical conflicts.
Research ethics discussions have consistently noted that there can be conflict in clinical studies between the interests of the individual participant and the scientific aims of the study (11,12,1324). The patient-oriented clinician-researcher may experience an internal clash between the clinical mandate to act in the patient's best interest (beneficence) and the scientific mandate to pursue truth with all appropriate rigor (scientific autonomy) (12). In clinical research, the "therapeutic misconception" (25) occurs when participants assume that the goal of clinical research is always therapeutic and when they do not understand, for example, that they may be randomly assigned to a study condition that is not in their best personal interest. This misconception is even more likely to occur if the study investigator is someone whom participants know and trust as a treatment provider and who is asking them if they would like to participate in a research study.
Several authors (11,12,1326) have discussed ways to address this issue. A first approach recommends that patient-oriented clinician-researchers avoid having direct research contact, such as recruiting or interviewing, with individuals with whom they have a therapeutic relationship (26). This recommendation does not address the more subtle ways in which a clinician's involvement can sway an individual to participate or remain in a study if a therapeutic relationship is defined as an individual psychotherapy or psychopharmacology relationship; for example, a patient may see the name of a trusted clinician on a list of investigators. Trust in clinical investigators has been found to be an important reason that many individuals with mental illness decide to take part in research (27). Although this trust is not necessarily misplaced, it can still be abused. Therefore, although we agree that patient-oriented clinician-researchers should not recruit individuals with whom they have individual therapy or psychopharmacologic relationships, there is clearly a need to supplement this recommendation with other approaches to adequately address conflicts of interest.
A second approach that has been recommended is for patient-oriented clinician-researchers never to conduct research with vulnerable individuals, which includes all persons with mental disorders, when there is any risk of exploitation or harm (12). Decisions about when a study may be exploitive of vulnerable individuals are delegated to external decision makers, such as patient advocates and institutional review boards. This position takes away individual autonomy from participants and can therefore be too restrictive. Although institutional review boards are necessary for general oversight, this approach is not likely to be widely useful for resolving many specific conflicts of interest.
A third approach is for patient-oriented clinician-researchers to fully disclose their conflicts to potential research participants (13,24). Although disclosure is a necessary element that should be incorporated into all approaches to address conflicts of interest, it is limited in that disclosure can be provided in a token or pro forma manner along with simultaneous suggestions that it is not relevant. If the overriding message to participants is that there is no real conflict, then it will not factor into their decision process.
A fourth approach is for clinician-researchers to seek an integration of their dual roles and develop a "coherent moral identity" that promotes good ethical judgment (11). This position dictates that clinician-researchers be aware of themselves and the possibilities for exploitation in their studies and take extra care to develop a relationship with participants that minimizes the possibility for therapeutic misconception. This position argues that clinician-researchers be aware that they have an ethical responsibility to both the individual client and to society at large. The "integrated identity" approach to dealing with conflicts of interest also provides a method for addressing situations in which a patient-oriented clinician-researcher overidentifies with a specific client, client population, or treatment system at the expense of the scientific value of a study. Miller and colleagues (11) discussed how such a scenario may lead a researcher to subvert essential procedures, such as random assignment.
Although the separation of dual relationships, full disclosure, and external review are all important steps toward addressing ethical conflict, the establishment of integrated identity is ultimately the most comprehensive means of balancing and prioritizing ethical issues and resolving conflicts of interest. We acknowledge that there are legitimate concerns that this model places a great deal of responsibility on the individual clinician-researcher to develop good judgment. Recommendations for dealing with ethical conflicts of interest that stress education and self-awareness are not always clear-cut and can leave room for clinician-researchers to continue to allow biases to lead to exploitation or poor science. However, although awareness does not automatically solve all problems, it sets the stage for using judgment to take appropriate steps.
In addition to the ethical challenges discussed above, there are practical reasons that it is often difficult to successfully function as a patient-oriented clinician-researcher. These challenges stem from the confusion or conflict that often occurs when an individual functions in multiple roles simultaneously—termed "interrole conflict" by social psychologists (28). In discussing these issues, we do not mean to suggest that they are representative of the experience of all clinician-researchers, many of whom enjoy the opportunity to shift roles (29). We characterize role confusion as both external and internal.
External role confusion. A major practical challenge for the patient-oriented clinician-researcher is clarifying for others what his or her job really is. A typical schedule may consist of spending a portion of time in a clinical setting and the remaining time conducting research. Whether the patient-oriented clinician-researcher chooses to coordinate his or her research at the clinical site or in another office is usually determined by resources and where the individual is able to work most productively. Either way, however, there will be problems explaining to others what one is doing. If patient-oriented clinician-researchers choose to conduct research in their clinical office, they will have difficulty explaining to clients and other staff with little understanding of the researcher role why they are not available for clinical work at the moment. If a problem situation arises, the pull to participate clinically will be very strong, and the research component of the patient-oriented clinician-researcher's duties will suffer. Matters are helped somewhat by the use of separate offices for research and clinical work, but this too can create challenges when clients and staff are not familiar with the research role.
Internal role confusion. Internal role confusion results primarily from the fact that research and clinical work consist of fundamentally different tasks and "ways of being" in a work environment, including elements such as work schedule, location, and accepted attire. Research typically involves less rigidity in these elements. A patient-oriented clinician-researcher who fluctuates between these two types of work identity may have difficulty adjusting to the constant shift. Although some may find the variety enjoyable, others will find themselves conflicted or confused about their work identity. In some cases the conflict or confusion will lead an individual to gravitate toward the work role that he or she is most comfortable with (either loose or rigid) and may make it tempting to abandon either clinical work or research.
Fostering integration to address role confusion. As in the case of ethical conflict, we believe that the development of an integrated clinical and research identity is the best way to deal with the practical role conflict issues discussed above. The development of an integrated identity will require a fair amount of introspection; clinician-researchers will need to examine their allegiance to an underlying moral or service principle to guide their work and develop an awareness of the importance of both research and clinical service delivery to serve this principle. An integrated identity will ensure that both science and practice remain balanced and that both beneficence and scientific integrity exist in harmony rather than conflict.
Most discussions of the process of training clinician-researchers have emphasized that clinicians should be trained in research in order to build research skills on a "core" of clinical knowledge and identity (4). This training approach is probably insufficient to develop integration, because in most cases those who teach research methods and statistics are not clinician-researchers, and the aspiring clinician-researcher will receive little encouragement to see research and clinical work as domains that should be integrated. The implicit message of this approach is that an individual needs to choose one area and stick with it. In contrast, we believe that training should focus on teaching clinicians to become patient-oriented clinician-researchers. Within such a model, all important topic areas would focus specifically on how they are experienced and dealt with by patient-oriented clinician-researchers.
To foster commitment to an underlying moral principle for clinical and research work, flexibility about how work is performed, and openness to knowledge from both clinical and research arenas, we suggest that trainees meet with patient-oriented clinician-researchers who agree to be mentors and role models.
Training programs can offer seminars to ensure that trainees will be able to use clinical experience to inform research and research knowledge to inform clinical work. Seminars could also highlight the need for patient-oriented clinician-researchers to act as agents to disseminate evidence-based practices. To ensure that trainees are able to manage internal and external role confusion and conflicting time demands, we suggest that they be given opportunities to meet with patient-oriented clinician-researcher mentors and role models to discuss practical issues in clinical research. Listening skills to maintain openness to both clinical and research-based knowledge are needed, as well as "bridging" skills to enhance communication between the clinical and research world. Such skills could be learned in courses about how to improve training and communication skills and to apply skills derived from motivational interviewing to enhance listening skills.
In addition to these skills, trainees should have knowledge of the ethical complexities and potential conflicts in clinical research. Programs could expose them to literature on ethical problems and engage them in discussions about addressing ethical dilemmas with senior clinician-researchers.
These training approaches could be flexibly applied to various types of curricula, including graduate training, postdoctoral training, career awards, and intensive training conferences. Regardless of the level of intensity, however, a key element is always the involvement of self-identified patient-oriented clinician-researchers who can serve as role models. For example, existing NIH-funded postdoctoral programs and career awards targeted at clinician-researchers could require role-modeling or apprenticeship opportunities with established patient-oriented clinician-researchers. Graduate training programs in fields such as clinical psychology could incorporate a course early in the training sequence on the integration of research and clinical practice (30) and ensure that patient-oriented clinician-researchers are present among the faculty. For those unable to take advantage of long-term training opportunities, NIH could fund intensive training conferences.
Fostering integrated identities would help individual patient-oriented clinician-researchers to develop internal models for prioritizing how to resolve role and ethical conflicts whenever they occur by staying focused on the underlying principles that direct their work. This would reduce the incidence of ethical violations and help protect research participants from harm. On a systemic level, training would also provide a degree of insurance against the drift away from the integrated clinician-researcher role, which would help maintain the number of clinician-researchers in the field.
We believe that the patient-oriented clinician-researcher can play a very important role in improving the quality of both behavioral health treatment and research. Patient-oriented clinician-researchers can serve as effective "bridgers" between the research and practice communities and can facilitate both the development of clinically relevant research and the dissemination of evidence-based treatments into routine clinical services. We recommend that role integration, through appropriate training, be encouraged to address both ethical and practical conflicts that can arise as a result of the combined roles. Such training can help protect both the vulnerable individuals who participate in research and the need for patient-oriented clinician-researchers in the field of behavioral health to continue to operate as double agents who can enrich the quality of both services and research.
This work was supported in part by grant K23-MH066973-02 to Dr. Yanos from the National Institute of Mental Health, grants R01-DA15978 and R01-DA15537 to Dr. Ziedonis from the National Institute on Drug Abuse, and grant KD1 TI12549 to Dr. Ziedonis from the Substance Abuse and Mental Health Services Administration. The authors thank Fred Huang, M.D., Kim Mueser, Ph.D., David Mechanic, Ph.D., Joel Yager, M.D., and Jonathan Krejci, Ph.D. for helpful comments.
Dr. Yanos is affiliated with the department of psychiatry at the University of Medicine and Dentistry of New Jersey-New Jersey Medical School in Newark and with the Institute for Health, Health Care Policy, and Aging Research at Rutgers University in New Brunswick, New Jersey. Dr. Ziedonis is with the department of psychiatry at the University of Medicine and Dentistry of New Jersey-Robert Wood Johnson Medical School in Piscataway. Send correspondence to Dr. Yanos at the University of Medicine and Dentistry of New Jersey, BHSB, D-level, 183 S. Orange Avenue, Newark, New Jersey 07101 (e-mail, yanosph@umdnj.edu).