Substantial evidence suggests the number of clinician-scientists has declined over the past several decades.2,5,7,11,14 Owing to a growing concern about the about the decline of orthopaedic clinician-scientists, Dr. Douglas Jackson, under the umbrella of the Orthopaedic Research Society (ORS) and American Academy of Orthopaedic Surgeons (AAOS) organized the first Orthopaedic Clinician-Scientist Steering Committee in March 2000 to identify the major obstacles to young clinician scientists and to propose solutions (a second meeting was held in February, 2001). The task force's summary8 suggested the major deterrents included: (1) lack of successful mentors and role models; (2) lack of nurturing environments; (3) educational debt; (4) institutional emphasis on primary care rather than subspecialization; (5) lack of financial support for additional training and education; and (6) peer pressure to share clinical responsibilities. The task force suggested the AAOS leadership “must make a bold and concerned effort if we wish to maintain orthopaedic clinician-scientists …”.
The task force realized we had relatively little information on the actual numbers of orthopaedic clinician-scientists and the scope of their activity, and in May of 2000, the AAOS and ORS accordingly funded a study to gather that information.3 The data was established by two approaches: mining publicly available federal (National Institutes of Health [NIH] and National Library of Medicine) databases, and an extensive survey of all 830 ORS members who were also active members of the AAOS. Only 64 of these 830 had received as much as $100,000 of funding from the NIH as principal investigators over a 10-year period (1992-2001) and only 22 had received as much as $1 million of NIH funding. More disturbing was a substantial decreasing downward trend in NIH funding of orthopaedic surgeons. Clinical demands were identified as the primary obstacle to pursuing a career as clinician-scientists by 75% of the orthopaedic researchers, and only 17% suggested lack of research funding was a primary obstacle. In fact, 86% had sought external funding and 70% had been successful. However, the funding provided no more than 5% salary support, an inadequate amount for release from clinical demands.
In response to the recommendations of the AAOS/ORS steering committee, the AAOS established the Clinician Scientist Development Program Committee in 2002. The newly established committee wondered which elements allowed certain individuals to succeed as clinician-scientists. This observational study was intended as a followup survey to define characteristics of the successful orthopaedic clinician-scientist.
MATERIALS AND METHODS
We designed a survey instrument with questions that would provide useful information to prospective clinician-scientists and departmental chairs wishing to recruit or maintain clinician-scientists. We presumed some of the questions or answers would overlap, but it was important to consider the successful clinician-scientist from both points of view. We defined the successful clinician-scientist as one who had as principal investigator achieved at least $100,000 NIH funding over a 10-year period.3 Of those 64 individuals, six had retired, leaving 58 orthopaedic surgeons who were active AAOS and ORS members. These 58 individuals were sent the survey from the AAOS.
The 11-page survey included 27 question groups, considered in the following four categories: (1) training, clinical research background, and practice (three question groups and six questions); (2) clinician-scientist research activities (seven question groups and 12 questions); (3) clinician-scientist training perceptions (four question groups and 10 questions); and (4) funding and service opportunities (13 question groups and 21 questions). The survey instrument is available online (supplemental materials are available via the Article Plus feature at www.corronline.com. You may locate this article then click on the Article Plus link on the right).
For all data reported we had at least an 85% response rate except in the case of the data on salary.
From the 58 individuals we received 46 responses (79%). All raw data is available online (supplemental materials are available via the Article Plus feature at www.corronline.com. You may locate this article then click on the Article Plus link on the right).
Twenty six of 40 (65%) began their research career as undergraduates or medical students. Only seven of 40 (18%) began research as a resident, and only four of 40 (10%) as a faculty member. Twenty of 45 (44%) had specialized training in research and 18 of 20 (90%) of these had an experience greater than 12 months. A large majority (90%) suggested that a minimum of 6 months training was important.
Of the programs these respondents represented, only 13 of 45 (29%) required a year of research for one or more residents. One research resident was selected in 4 of these programs, two in 3, and three in 2.
Nearly 34 of 46 (74%) spent less than 20% of their time in research. Twenty eight of 46 (61%) stated they had protected time, although generally for less than 20 to 30% of their total time. The ideal amount of research time for these individuals would be at least 30% time for 43 of 45 (96%), and 25 of 42 (60%) suggested this would be feasible in their setting. Most (38 or 44, 86%) thought some minimum time commitment necessary to the career of a clinician-scientist.
The eight questions related to salary were answered only by 10 to 32 of the 46 respondents, but the data does suggest up to 20% of support comes from departmental sources in 60% (6 of 10 respondents) and from institutional 24% (5 of 21 respondents) independent of direct clinical earnings. Over 90% (20 of 22) of the respondents obtain at least 5% salary support from external funding and nearly 40% (seven of 22) obtain over 30% salary support externally. Nearly half (21 of 46) suggest their salary is the same as their colleagues while the same number stated it was lower. A majority (31 of 36, 67%) stated their research activities affected the volume of patients and the gross and net clinical billings.
A majority (26 of 43, 65%) suggested the level of their current research funding was about right. Ninety three percent are principal investigators on at least one externally funded project. Industry provided external funding for 87%, federal agencies for 52%, and foundations for 49% of these externally supported studies. Other sources of funding (30% gifts) constituted smaller percentages of the total.
Nearly 60% (25 of 43) of these clinician-scientists have unshared laboratory space, with 56% (14 of 25) of those having more than 1000 square feet of space. Of those sharing space, 90% do so with three or more other scientists.
A large majority (41 of 46, 89%) of these individuals spent more than 10 days away from work each year for scientific meetings, highlighting the importance of this activity. A large majority (41 of 44, 93%) served on grant review panels and 70% of these were federal agencies. Other grant review panels included the Orthopaedic Research and Education Foundation (48%), other foundations (56%), industry (17%), and other (unstated) funding sources (13%).
When asked in a series of questions to define a clinician scientist, 40 of the 46 respondents selected an individual seeking and obtaining external funding as the primary feature. A majority (39 of 43, 90%) thought formal training of at least 6 months an important factor, and of lesser importance was whether the individual worked on their own or collaborated with others (42 of 46, 93% stated they do collaborate).
A substantial number (20 of 46, 44%) thought their colleagues resented the time they spent in research, and 76% (13 of the 17 who answered the question) thought their colleagues perceived research activities to be less important than clinical productivity. Nearly half (8 of 17) suggested the perceptions of their colleagues had a negative impact on the amount of time they spent in research.
A decline in the number of clinician scientists over several decades has been documented,4,11 although a similar decline in the number of orthopaedic surgeon clinician-scientists has not. Whether there is a decline, we can state there are few orthopaedic clinician-scientists considering only 64 orthopaedic surgeons received $100,000 of funding from NIH as a principal investigator over 10 years and only 22 received $1 million,3 making it clear the number is small. This second survey was conducted as an ongoing effort to develop better strategies for encouraging clinician-scientists by characterizing successful individuals and documenting the obstacles to success.
We note several important limitations of this survey. First, we presumed all clinician-scientists would be members of the Orthopaedic Research Society. We believe this assumption reasonable for those individuals conducting basic or translational science but perhaps not for those conducting exclusively clinical research. Given the two sorts of research can involve different time and resource requirements, the survey results likely apply only to those involved in basic or translational research. Second, we had a 79% response rate, which while quite good for surveys, might not be fully representative. The earlier and larger study (with a response rate of 51%),3 which included all of the individuals in the current study, demonstrated good agreement between responders and nonresponders in two crucial areas, publications and funding, so we presume the 79% response rate is also reasonably representative. Third, actual working conditions do not always lend themselves to the simple categorizations of a survey instrument and might not reflect the range of obstacles encountered. Fourth, the eight questions related to salary were answered by only between 10 and 32 of the 46 respondents and individuals answering one question might not have answered a related question. Furthermore, this survey was not designed to ascertain detailed information on salaries. Therefore, data related to salary should be viewed cautiously and is presented only to introduce what might be an important issue in facilitating clinician-scientists. Finally, the perception of resentment is only that from the respondents. We did not survey colleagues for their perceptions nor are we certain a survey would accurately identify whether and how resentment was present.
There are several important reasons to enhance the number of orthopaedic clinician-scientists. First, advances in clinical practice continue to arise more often from the laboratory rather than the clinic. In orthopaedics, these advances include tissue engineering, biologics, gene therapy, and imaging. Unless clinicians are involved in the development of the techniques, they are less likely to be efficiently transferred to the clinic. Second, many current surgical techniques will become obsolete and be replaced with more biological solutions. The AAOS steering committee on the clinician-scientist estimated a need for at least three to six new clinician-scientists each year.8 Third, research for musculoskeletal diseases in general, and for orthopaedic fields in particular, is vastly underfunded in proportion to the burden of disease and the overall NIH budgets. The National Institute of Arthritis, Musculoskeletal, and Skin Diseases (NIAMS) receives only about 2.7% of NIH funding, and orthopaedic surgeons less than 0.12%.3 The National Institute of Arthritis, Musculoskeletal, and Skin Diseases is arguably the largest source of musculoskeletal funding at the NIH, although other institutes also contribute. Assuming the PhD collaborators of surgeons in orthopaedic departments receive 10 times the funding of the surgeons themselves, the amount is still small, especially considering musculoskeletal diseases constitutes a burden of disease of about $100 million per year.10 Research spending is in part related to the numbers of effective grant proposals submitted, and without properly trained orthopaedic clinician-scientists, this amount is unlikely to substantially increase.
If one wishes to encourage the development of clinician-scientists, it seems intuitive that programs should carefully consider and mentor those individuals involved in research at an early stage. Our study suggested a majority of the successful clinician-scientists had their first research experience as undergraduates (50%) or medical students (15%). Based upon surveys of 293 selected students in the medical doctorate program of the Harvard-Massachusetts Institute of Technology Division of Health Sciences and Technology, 75% held faculty appointments and 68% of MDs and 86% of MD-PhDs were engaged in research.1 These data and ours provide circumstantial evidence that those staying in research were involved at an early part in their career. This evidence does not address the question of whether those involved in research at an early point in their career are more or less likely to be involved later. Although a large minority (about 40%) of our respondents had at least 1 year of research training, the data also does not address the question of whether individuals spending a year or more of dedicated research time during residency are more likely to stay in research later in their career. This would require the longitudinal followup of all residents who have spent a year or more in dedicated research.
Leon Rosenberg, former dean of the Yale School of Medicine who has written on the clinician-scientist11-13 and speaking on the medical education system, commented, “We have failed to create a culture of inquiry.” Although our survey provides no direct supportive data, the small numbers of orthopaedic surgeons engaged in research from the previous survey3 and the small numbers of successful orthopaedic clinician-scientists in this survey tend to support that contention. Of the nearly 16,000 members of the AAOS,9 a tiny minority engage in research at the NIH level as principal investigators. Thus, one must consider whether in orthopaedic surgery we have failed to create a culture of inquiry and whether clinical endeavors are not valued more highly than research.
Realizing the selection bias in this group (ie, individuals in those programs most likely to provide training), it seems clear research options for residents are limited, although this survey provides no definitive answers on this point, and a survey of training programs might provide better information on training availability. Information on the funding of these research years was too limited to be valid, although nine individuals noted their departments provided the support.
Orthopaedic surgeons tend to be busy clinicians, regardless of participation of research, and nearly three-fourths (74%) of these successful clinician-scientists spent less than 20% of their time in research. This is despite the fact that 60% stated they had protected time, although in general this protected time is less than 20% to 30% of their total time. Obviously, clinical or administrative expectations eroded into that protected time, yet 95% suggested the ideal amount of research time for them would be at least 30% time and 60% suggested this would be feasible in their setting. These figures might seem contradictory, but suggest other factors prevent them from spending time as they wish. Many individuals may work in a permissive, if not supportive environment and still not spend as much time in research as they think ideal. This may be explainable in part by the resentment many clinician-scientists feel, with the effect of contributing more time to clinical endeavors and less to research. Most (86%) thought some minimum research time commitment necessary to the career of a clinician-scientist and 73% thought 20% to 40% was critical. Interestingly, only 27% thought greater than 40% important, suggesting most thought substantial clinical involvement also important.
As noted, 44% of our respondents believed their colleagues resented their participation in research. Given these individuals are those who succeed in obtaining external funding (particularly highly competitive NIH funding) and might bring greater prestige to departments, the rate of resentment might well be higher for those less successful individuals who struggle to develop a research career. Such resentment may play a major role in discouraging young individuals.
The investment in musculoskeletal research from the orthopaedic community is small compared with the burden of disease and the potential financial resources. If we presume the nearly 16,000 members of the AAOS net $350,000 per year, the income of that group is about $5.6 billion. Based on several estimates,6 the investment by the orthopaedic community (primarily academic departments) is about $23.5 million (Table 1). This reflects an investment in our future of about 0.4%, a figure that would be low even if these estimates are off by several fold. This investment likely reflects in part a perception by orthopaedic practitioners that research is not important to our field. We do not know whether or how this perception may relate to a failure on the part of the research community to adequately communicate the relevance and importance of their contributions. We also do not know whether this reflects Rosenberg's broader view that we have failed to create a culture of inquiry. Regardless, these sorts of perceptions undoubtedly must be addressed to make substantial progress in creating a small but crucial cadre of orthopaedic clinician-scientists.
Given data from a previous report3 and this data, it is clear the major impediment for orthopaedic clinician-scientists is the demand of clinical responsibilities. These successful individuals spend less time in research than they think ideal. Achieving external funding on the other hand, does not seem to be the major problem. We question whether the larger problem relates to the relatively low value orthopaedic surgeons in general place upon original inquiry. Creating supportive environments for clinician-scientists should be a primary goal of the orthopaedic community.
We thank Ms. Sylvia Watkins-Costillo, in the Department of Research and Scientific Affairs of the AAOS for her invaluable help finalizing the wording and format of our survey questions, and for tabulating all data; and we thank Ms. Kimberly Scroggs for sending surveys and reminders to recipients and for collecting the data. This project was conducted under the auspices of the AAOS Clinician Scientist Development Program and is published with their permission.