Physician-scientists in anesthesiology: the all too empty pipeline

In the accompanying report, Charles Emala and his colleagues at the Anesthesia Research Council take a deep dive into the anesthesiology physician-scientist pipeline, documenting the current shortage of anesthesiologists who do research, and exploring the reasons for this shortage ¹. For anyone who believes that physician-scientists have an important role in all clinical disciplines, their observations are not encouraging. Although they found that NIH research funding and training support in anesthesiology departments has increased, only 1.5% of anesthesiology residency entrants from 2009 to 2019 (approximately 300 of 20,000) had a PhD (Figures 1E and 4B in their article). A PhD is not required for biomedical research by physicians but graduate school, unlike medical school, focuses on research training. Here we will use a PhD as a surrogate marker for estimating the number of research-trained physician-scientists entering a clinical discipline and consider MD/PhD programs as a case study on what might be done to develop more physician-scientists in departments of anesthesiology. We will also consider the survey data on hiring practices that Dr. Emala and his colleagues obtained from anesthesiology department chairs. Those data suggest that 10% or less of new faculty hires in anesthesiology in the past five years had been given protected time for a ≥50% research effort, and that startup packages for physician-scientists in anesthesiology are less robust than in other fields. We’ll return to both of these critical points later.

Why focus on MD/PhD programs? MD/PhD programs are not the sole source of research-trained physicians in the United States, but the 2014 NIH Physician-Scientists Workforce (PSW) Report noted that half of the physicians funded by the NIH had a PhD as well as an MD, a fraction that greatly exceeds their 3% representation among medical school graduates ². The data that were included in the NIH PSW report have since been complemented by the national MD/PhD program outcomes study ^3,4. That study gathered information on thousands of MD/PhD program graduates dating back to the inception of MD/PhD programs in in the 1950’s. The results showed that although nearly all MD/PhD program graduates do additional postgraduate clinical training, only 3% had chosen to become anesthesiologists. This compares with 26% who trained in internal medicine, 14% in pathology, 13% in pediatrics, and 7% in neurology. Only obstetrics and gynecology, emergency medicine, and family medicine ranked lower than anesthesiology among clinical fields that attracted MD/PhD program alumni. The percentage of MD/PhD alumni choosing anesthesiology is even lower than medical school graduates in general (4.6% in 2023) ⁵.

The authors of this essay include two (LFB and MHA) who are longstanding MD/PhD program directors (and not anesthesiologists) and two recent program graduates. DQH is an Albert Einstein College of Medicine graduate who is in his final year as an anesthesiology resident. NT is a University of Pennsylvania graduate who is a research fellow and about to transition to tenure track faculty in anesthesiology. Some MD/PhD students enter medical school with a well-defined clinical interest and stick with it, but most do not. This makes them relatively amendable to differentiation based on experiences while in training. In our experience, the reasons for the choices they eventually make vary considerably, but common themes include: 1) exposure during medical and graduate school to successful physician-scientists who work in that field, 2) assurances from mentors and advisors that a research-focused career in that area of medicine is feasible and well-supported by departmental and institutional leaders, 3) awareness that there are interesting problems yet to be solved, and 4) meaningful clinical exposure to that discipline during medical school. With few role models and limited exposure during medical school, anesthesiology has been vulnerable in each of these areas.

In 2014 when the national MD/PhD program outcome study was conducted, 75% of alumni who had become anesthesiologists were working in academic settings, but of those who were, only 21% reported being able to devote half or more of their time to research ³. That contrasts with 50% or more in pediatrics, internal medicine, and neurology. Clinical disciplines that had as low or lower fractions of academic faculty who spent more than half of their time in research included orthopedic surgery (10%), radiology (12%), urology (14%), neurosurgery (16%) and emergency medicine (19%) ³. These numbers for anesthesiology are in agreement with the survey data collected by Emala et al.

The paucity of physician-scientists in anesthesiology is not because the problems faced by anesthesiologist inside and outside the operating theater are uninteresting or unimportant. One of us who is currently a fellow in anesthesiology (NT) notes that the clinical work he does as an anesthesiologist leads him to hypotheses that can be tested in the laboratory and was one of the reasons that he chose to become an anesthesiologist. However, as MD/PhD students perform their clinical training they, like most medical students, have limited exposure to anesthesiology and are likely to encounter few physician-scientists during that exposure.

To increase exposure of MD/PhD trainees to their specialties, a number of clinical societies like the American Society for Nephrology⁶, have developed outreach programs to MD/PhD students in the mid- to late-PhD phase. Trainees are invited to attend the annual national meetings with travel awards to cover transportation, hotel, and registration expenses. The American Society of Clinical Investigation has a similar program intended for future physician-scientists in many disciplines, with an emphasis on internal medicine and pediatrics⁷, as do the neuroscience institutes at the NIH.⁸ The American Physician-Scientist Association, a student led organization, sponsors opportunities for MD/PhD, DO/PhD and MD candidates to meet with the directors of physician-scientist friendly residency programs at their annual meeting⁹.

Collectively, these societies organize specific programs to highlight research opportunities in the specialty and maintain ongoing mentoring relationships with the trainees. Anesthesia-related societies could replicate these outreach efforts to increase trainee exposure before they transition to the clinical training phase of an MD/PhD program. FAER (the Foundation for Anesthesia Education and Research) has taken a step in the right direction by offering fellowships to medical students that support an 8 week summer experience¹⁰. The FAER approach would have to be re-configured to work for MD/PhD students who in the early phase of their training are usually limited to research opportunities with potential thesis advisors.

Finally, there is a matter of salary. Whatever the current disadvantages of anesthesiology for physician-scientists, inadequate pay is not one of them. On average, anesthesiologists are paid well compared to other fields ¹¹.Ironically, this may work against those who desire a physician-scientist career. Depending on institutional priorities, budget conscious department chairs may find physician-scientists in anesthesiology to be more valuable for their clinical skills than their research skills. From a purely economic perspective, time spent doing research becomes even harder to justify when the average salary of anesthesiologists exceeds the NIH salary cap for principal investigators. Deciding how the difference will be covered presents a conundrum whose resolution reflects the values held by departmental and institutional leadership. Will the difference come from other resources? Should physician-scientists in anesthesiology be obliged to accept salaries well below those of their more clinically inclined colleagues? In an ongoing study, two of us (LFB and MHA) have tracked the relationship between median income within a clinical field and the fraction of MD/PhD program alumni in that field who are able to devote more than 50% effort to research. The data show an inverse relationship between salary and research effort. Clinical fields with higher salaries tend to have fewer people spending most of their time on research. On a research effort versus income plot, anesthesiology clusters with surgical disciplines and radiology in the higher salary, lower research effort part of the distribution.

We will close this essay with two final observations. First, thinking of the U.S. physician-scientist pipeline as being leaky throughout its length is helpful for conceptualizing the issues, shining light on an important problem, and formulating some of the remedies ¹². However, leaks can only be addressed when meaningful numbers of medical school graduates have entered the pipeline, choosing in this case to be anesthesiologists. An empty pipeline doesn’t leak, but it also doesn’t produce physician-scientists. More needs to be done during medical school to attract future physician-scientists.

Second, we think it is appropriate to ask whether anesthesiology programs should focus on becoming better stewards of the resources they already have. Over the past decade roughly 150-200 MD/PhD program graduates have become anesthesiologists. However, only a handful have achieved grant-supported, research-intensive faculty careers. If the outcomes were better for those who entered the field, more might become interested. Perhaps more attention should be paid to how to increase the success of those who train in anesthesiology, starting with the crafting of hiring packages that offer sufficient protected time and research funding¹³.^a Physician-scientists bring a unique perspective to biomedical research because of their clinical training and activities, but building a research team, obtaining research funding from the NIH and establishing independence all take time. Expecting that 2 years of 50% protected effort and a comparatively small startup package will be sufficient to achieve success flies in the face of reality. We agree with Dr. Emala and his colleagues that the problems of attracting and sustaining physician-scientists in anesthesiology are substantial. They need not be insurmountable.