We Must Not Let Clinician–Scientists Become an Endangered Species

A colleague and I were discussing the increasing number of new faculty who are clinician–teachers at our institution. Many of them are wonderful teachers, appreciated by the students and residents. We discussed how the promotion criteria for clinician–teachers require little scholarly activity compared with the substantial expectations for a tenure-track faculty member. While clinical work and teaching are not easy, the expectations for clinician–teachers are clearly based on billings, quality-of-care measures, and student evaluations. On the other hand, promotion of clinicians who are focused on research is more difficult because research productivity is far less certain. This encourages many faculty who are interested in research to opt for what they consider the less risky path of the clinician–teacher even though more clinical hours are usually required. We noted that departmental priorities and incentives increasingly seem focused on clinical efficiency and productivity. Research skills, experience, and interest appear to be getting eclipsed by the clinical mission, which is discouraging for those with interest in and curiosity about research. After he expressed his concerns, I asked him how he felt about his future as a clinician–scientist. “Clinician–scientist?” he asked. “I don’t even know what that is.” “It’s someone with a terminal clinical degree, like MD or DDS, who provides clinical care and also devotes a substantial effort to research. I think that’s what you are,” I said.* “I don’t really feel like a clinician–scientist,” he responded. “I feel more like a doctor who enjoys research, but I identify as a clinician. Most of my friends are clinicians.” “But with all of your publications and grants, you’re really a clinician–scientist. Many people think clinician–scientists are an endangered species.” He paused for a moment and then said: “I don’t know. I do my clinical work and my teaching just like everyone else. But when people have questions, I like to help find the answers, and sometimes that means doing research. I like to analyze data and write articles about what we find.” “Well, that proves my point,” I said. “Just look at your publication record. You’re probably in the top 5% of emergency physicians. And you’ve had a lot of grants too.” We compared his research productivity with that of other academic emergency physicians based on data published by Deluca et al1 and found that he actually was in the top 1% of academic emergency physicians, based on his publications and citations of his work. He sighed. “I guess you’re right. Sometimes it’s hard. I just don’t have time to finish things. I have piles of half-done studies. There aren’t enough people who can help me or who even care about what I do. With clinical work I know what time I start and when I’m done. And I like the clinical work. It’s never boring. You can actually make a difference in someone’s life. But you never know with research whether it will ever get noticed or make any difference at all. And if I did any less clinical work I’d be dangerous. I’d lose my skills.” Our conversation reminded me of the following metaphor describing the plight of the clinician–scientist trainee, which I had created for an international meeting. Imagine the clinician–scientist trainee traveling along a river toward a destination downstream that we can label “better health for all.” There are two boats that will carry this trainee downstream. One represents clinical education and practice. The other represents scientific inquiry and research. Most of the boats are filled with trainees who will remain in their original boat for the entire time. But the clinician–scientist trainee will be moving between the two boats. In the clinical education and practice boat, everyone will learn a set of skills—for example, how to fish, how to row the boat, how to avoid rocks. There will be an ongoing assessment of progress on the skills—how many fish are caught, the power and efficiency of strokes, and the occurrence of crashes against rocks in the river. The overall progress of trainees in the clinical education and practice boat as it moves along the river will also be carefully monitored. The other boat, full of research trainees, may have the same ultimate destination of better health for all, but its course is less clear. Depending upon the boat’s guide, it may stop along the shore to investigate an estuary or the sighting of birds or animals along the shore. The occupants of this boat will be responsible for locating their own provisions along the journey. The clinician–scientist trainee will have one foot in each boat as the two boats begin the journey together, but soon the boats will diverge and the trainee will have to move into one boat. At different points along the journey the trainee will be handed off to the other boat, then returned to the original boat. At some point as formal training ends, the two boats will meet up and travel together, and the new clinician–scientist will attempt to put one foot into each boat again. However, there will be turbulence, with currents pulling the boats in different directions, and the people in each boat will wonder whether the clinician–scientist is destabilizing them as they try to navigate the river together to support the clinician–scientist. The legs of the clinician–scientist will begin to stretch and lose traction as the boats increasingly drift apart. While this metaphor does not entirely encompass the challenges for the clinician–scientist, it does illustrate the difficulties of training and working in two different worlds, each of which has its own skill sets, cultures, and values. It also recognizes that there are challenges after training is complete, when the rewards and expectations for clinical work and research differ and there is attrition, since many clinician–scientists choose to devote themselves to clinical practice and abandon scientific investigations. Zerhouni et al2 recently noted: Today, the most often-cited obstacle to the development of novel and more successful therapies is the general lack of deep understanding of human pathogenesis…. The tools and methods arising from the extraordinary progress of the basic sciences—such as genomics, proteomics, and many other advances of the last decades—need to be applied directly to large cohorts who are followed for years. The tools are available, but where are the trained physicians and scientists who will dedicate their lives to such long and difficult explorations and be free of the need to generate large revenues from an increasingly cost-conscious academic health system? A 2014 workforce report from the National Institutes of Health (NIH)3 stated that during 2008–2012, there were approximately 9,000 clinician–scientists in the NIH-funded workforce—including 4,192 MDs, 4,086 MD/PhDs, 341 nurse–scientists, 253 veterinarian–scientists, and 161 dentist–scientists—and noted a declining and aging clinician–scientist workforce, providing some basis for Zerhouni and colleagues’ concerns. The report recommended addressing how to better support the transition from training to independence, increasing workforce diversity, having financial incentives such as loan repayment, and incorporating the resources of Clinical and Translational Science Awards (CTSA) programs to support training and development of early-career-development pilot grants. One approach to developing the clinician–scientist workforce has been the MD–PhD programs, many of which have been funded by the NIH. These programs have traditionally provided two years of basic science medical school education followed by three, four, or five years of graduate research training and then two years of clinical training. In this issue Harding et al4 describe the history of MD–PhD training and the Medical Scientist Training Programs supported by the NIH. The MD–PhD programs in the United States graduate about 550 clinician–scientists annually. There is an attrition rate of 10%, and time to degree averages eight years. Harding et al note that women and underrepresented groups apply to MD–PhD programs at lower rates than they do to MD programs. While 95% of graduates of MD–PhD programs go on to complete residency training, Harding et al raise concerns about the increasing time for clinical training for those programs, which typically includes several years for fellowship after residency. The increasing complexity of research also requires additional training time. Experts at national and international meetings have attempted to address the plight of clinician–scientists. Hall et al5 in this issue describe several NIH-sponsored workshops to consider new approaches to training clinician–scientists, including shifting the thinking about the training of clinician–scientists from a pipeline model to a model with the multiple on-ramps of a highway. Those on-ramps could be integrated into residency or fellowship training with master’s-level research training. This approach could augment the MD–PhD programs. Lingard et al6 in this issue studied the support for practicing clinician–scientists by conducting a narrative review of the published literature about clinician–scientists followed by interviews with expert informants to assess the completeness of the information in the published literature. Their study identified organizational and individual issues and solutions. The organizational initiatives focus on having mentors for clinician–scientist trainees, fostering a community of clinician–scientists, providing adequate protected time for research, and rewarding scientific productivity at the same level that clinical productivity is rewarded. At an individual level, the authors felt that skills related to time management and interdisciplinary collaboration were helpful. Lingard et al also identified gaps in the published literature that their expert informants had commented upon. One gap was related to how much time should be devoted to research. While there was support for 75% protected time for research, there was some concern about whether such a time commitment for research was compatible with maintenance of clinical competence, particularly in a procedure-based specialty. There was also concern that there was often structural hypocrisy, such as the prioritization of clinical needs for patients over the protected time for research due to pressures to support the clinical mission in spite of institutional commitments to support protected research time. There were also gaps in the literature related to how to measure the success of a clinician–scientist, how a clinician–scientist should be defined and recognized, and issues of clinician–scientists’ professional identity. Based on these studies, I have four suggestions about how we in academic medicine might provide better support for the clinician–scientist workforce. We must recognize and recommit to the importance of clinician–scientists in the biomedical workforce. In a fragmented health care delivery system without clear and agreed-upon goals, the contributions of clinician–scientists as a unique resource may not be clear and the case must be made for their unique value. It is likely that the synergies of the clinical and scientific perspectives provided by the clinician–scientist vary depending on the fields involved, and it would be important to identify those that are most promising. Hobin et al7 describe some of the tensions in developing synergies between basic scientists and clinician–scientists in translational research and make recommendations to improve funding, training, collaboration, and incentives. Training for clinician–scientists should better integrate clinical and research training, and must improve access and support for women and underrepresented groups. Trainees should have the opportunity to model integration of clinical and research activities that will be needed in their posttraining careers. The idea of starting and stopping clinical work for years to focus on research should be reassessed, as it allows clinical skills to deteriorate and does not provide a good model for future work balance. By integrating the research and clinical training, the trainee might be able to experience the synergies that are one of the goals of clinician–scientist training. Goldberg and Insel8 have described a reimmersion program to assist MD–PhD students with the transition from the research phase to the clinical phase that provides needed integration during the transition. This program is a good start, and we need more innovations like it. The training should also provide support for underrepresented groups through mentorship and institutional commitment that is continuous. In this issue Jagsi et al9 and Andriole et al10 describe some of the challenges for women and underrepresented groups that must be addressed to create a more diverse clinician–scientist workforce. The incentives, rewards, and payment for the work of clinician–scientists need to be better aligned with institutional and national goals for the health care system. Current fee-for-service clinical payment systems may encourage clinician–scientists to generate clinical income rather than carry out research that could improve population health and prevent disease. Research findings may result in innovations that reduce the need for certain clinical services, creating tensions between incentives for the clinician–scientist. Alternative payment models such as bundled payments may provide better alignment between research innovations and clinical care, since they reward the efficiency that research innovations can create. The work environment for clinician–scientists needs to integrate the clinical and research activities. This could occur with the development of new models supported by CTSAs and through development of teams organized around problems with both a clinical and a science component, such as those described by Ravid et al11 in this issue. Clinician–scientists are well situated to be members of large teams organized around clinical problems, where they can help coordinate the study and translation of research findings into the clinical care environment. We must address the issues expressed in the metaphor of clinician–scientist trainees with their feet in two boats. We need to find a way to reduce the tensions that are currently pulling the clinician–scientists in different directions and encouraging them to leave research for the more certain and lucrative rewards of clinical care. When we view the role of the clinician–scientist within the overall goals of the health care system, their unique capacities to contribute toward innovation, coordination, and synergy are compelling. There are now opportunities and models to better support the recruitment, training, and practice of clinician–scientists. I hope the articles in this issue of Academic Medicine will provide the impetus for renewed attention to and support of these important members of our health professions workforce.