I hate it when I fall behind in my journal reading. Of course, it happens all the time, as you might expect, with my time sandwiched between running my lab, writing grants, seeing patients, and operating. Sometimes, though, I get a chance to try to catch up a bit. Such was the case the other day, but unfortunately I came across an article that almost made me wish I hadn't. It was a study published in the February issue of Annals of Surgery1 and it showed that the situation is much worse than I expected. it also shows that I may be a rarer bird than I thought I was, and not just because of my admittedly strange personality. What the study showed is that surgeons are hugely under-represented in the ranks of physicians holding NIH grants and that the situation is getting worse.
The study, carried out by a group from UCSF and NHLBI, systematically examined data regarding success rates for applying for NIH grants among surgeon-scientists and compared them to non-surgeon-scientists. The results were not pretty--if you're a surgeon.
Basically, what the investigators did was to obtain research project grant (RPG) data from the NIH Consolidated Grant Applicant and Fellow File (CGAFF) or from the NIH website. The CGAFF contains data about investigators who have applied for grants and contracts from the NIH and other PHS agencies since the very beginning, way back the grant system was first established in 1937. They then obtained medical school manpower data from the American Association of Medical Colleges (AAMC) faculty roster. Total NIH RPG application and award data were compared with application and award data from all surgical investigators (including PhDs working in surgical departments) and then with surgical investigators with an M.D. degree. This allowed them to break down the data into applications and success rates by surgical and nonsurgical investigators between 1982 and 2004. (Note: 2004 was the year I managed to compete successfully for my first and thus far only R01.)
The first thing that stood out for me was just how small a component of total NIH awards. Surgical awards peaked at 3.5% of total awards in 1992 and have been steadily declining since then to an all-time low of 2.2% of total awards in 2004, which is the last year for which data were available. Also notable was that, between 1982 and 2004, the rate of increase in the number of awards was 71.2% for non-surgeons, 41.4% for all surgical investigators, and only 23.8% for surgical investigators with M.D.'s. It was first speculated that the reason for the low representation of surgeons in the NIH funding profile was because of a lower success rate between surgeons and non-surgeons (25% versus 29%, respectively), but further analysis indicated that that relatively small difference was not enough to account for the discrepancy. What might account for the discrepancy is the much lower rate of increase in surgical applications for NIH grants, which was only a little more than half the rate of increase for non-surgical applications (a 67% increase versus a 124% increase) over the time period examined. Meanwhile, consistent with a trend in all NIH-funded investigators that has gotten the NIH worried, the age at first NIH grants for surgical investigators was 41.8 years, and on a more encouraging note resubmission rates for surgical investigators didn't differ significantly from the overall rate. The bottom line was that the underrepresentation of surgeon-investigators in the ranks of NIH-funded investigators was primarily accounted for by the low and diminishing rate of submitting applications by surgeons at the very time when the number of new applications to the NIH is skyrocketing.
I agree with many of the authors' speculations about why surgeons submit so few grant applications to the NIH relative to other physician-investigators and relative to PhD-investigators. There are three big ones that I see.
Clearly, the most significant impediment to a surgeon doing research is time pressure. Surgery has always been a very time-intensive specialty. Certainly other specialties are also time-intensive, but the different aspect of surgery is that in addition to the cognitive skills necessary to practice it, it is a very technical specialty. In other words, practice makes perfect, and for the particularly demanding surgical specialties, it's very difficult to remain skilled without being active operating. This is something that hasn't changed in 23 years; in fact, the situation has arguably gotten worse with the influx of technology into surgery, including advanced laparoscopic procedures with a high learning curve and a requirement for a high degree of technical prowess and robotics. The authors don't even mention this; instead they tried to blame resident work hour reductions for the increased time pressure on surgeon-investigators. Give me a break. Resident work hour restrictions and the shift in workload to attendings that has resulted from it may well be a factor, but at best it's probably a minor one. To try to blame a trend that goes back at least to when I was a resident on a change in residencies that has only been in place around four years is a specious argument. Indeed, the authors' own data would tend to argue against resident work hour restrictions, as the data clearly show that the decline in surgeon applications for NIH funding is a trend that goes back to at least the mid-1990s, long before mandatory work hour restrictions were in place anywhere other than New York state.
To me, the second strongest impediment is economic pressure. Since I became a physician 20 years ago, the financial model of paying for medical and surgical care has changed radically due to the managed care revolution of the 1980s and 1990s. In the good old days, surgery was lucrative enough that many academic departments of surgery had fairly generous slush funds to fund research by junior faculty, allowing them to develop preliminary data to use in grant applications. However, with managed care and ever-declining reimbursements adjusted for inflation, in order to generate the same clinical income, surgeons have to work a lot harder. Worse, some departments expect their surgeons to cover their entire salary with billing, along with secretarial support and rent for their offices. That's almost impossible to do as a half-time surgeon, much less a quarter-time surgeon, which would be better for doing credible lab-based translational research. I happen to be extremely fortunate to have a cancer institute and department of surgery that are willing to subsidize my research activities to the tune of my being able spend more time in the lab than out of it, but such situations are rare. I also have no illusions that my freedom to do research would go away pretty fast if I ever lose my research funding, and I especially have no illusions that I'm a lock to hang onto my research funding, given the current budget woes at the NIH.
There is also the problem of research training. Unlike many medical specialties. surgeons receive their research experience after the second year of a five year residency, at which point they take either one or two years in a research lab, whereas most other specialties take their training after residency, which makes their transition into a research-oriented career more natural. One thing surgeons do share with their non-surgical colleagues is that they're research training is generally too brief to prepare them to compete successfully for NIH grants. The difference is that not only is a surgeon's research training too brief but by the time he or she finishes his residency it's three years old. For most areas of research it might as well be ten. Couple this with the long and arduous nature of surgical training, many new surgeons, even ones who are very interested in research, aren't too thrilled with the prospect of more years of research training. Too many of them may have little choice anyway, given the record debt levels that medical students are finishing with.
The authors go on to describe a number of potential strategies to cultivate surgeon-investigators. First, they have to be recognized as valuable. Naturally, I'm not an unbiased source, but a surgeon brings some unique attributes to the translational research endeavor. Indeed, the authors argue as much, and they are correct:
Along with other clinical specialists, surgeons have an important intimate familiarity with patients and their diseases. Surgeons have access to tissue for banking and are in a position to study and apply device-based and molecular therapies that require local delivery. Furthermore, the surgical personality, which is inherently impatient with the status quo, may be important in helping to drive translational efforts.
To me, perhaps the key values that surgeons bring to translational research is access to tissue for molecular studies and the ability actually to assist in the design and testing of new medical devices. The "surgical personality," however, is not entirely an unalloyed good. It is true that surgeons tend to be driven and unsatisfied with the status quo. That's good. It's also true that surgeons tend to be tinkerers. That's also good. The downside is that we are sometimes too impatient to develop the preliminary data necessary in the development of new molecular and targeted therapies, which is odd, given how amazingly patient surgeons can be while doing long and involved surgical procedures.
Although the proposal that some minimum number of grants be reserved for surgical investigators will never fly (after all, why surgery and not other specialties?) Some of the strategies make sense, particularly restructuring the surgical residency program for aspiring surgeon-researchers to make the research more relevant to becoming an independent investigator. Another idea is to cultivate surgical centers of excellence in research in order to train new surgeon researchers in a less haphazard manner. This idea has potential in that such centers would attract students interested in research, but I see a downside as well in that it could produce a "two-tier" surgical training system. Finally, certainly mentorship is critical, but the problem is that there are fewer and fewer mentors who can guide aspiring young surgical researchers. It's a well-documented statistic that there is a major attrition in surgical researchers with age, with many simply giving up and "going straight clinical."
Perhaps the most depressing thing about this analysis is that one likely explanation for such a low number of surgeons applying for NIH grants is that there is a perception in surgical departments that the odds are so long that it's just not worth the effort, that the hurdles in the way are impossible to jump. This is a problem that could potentially be solved with more money, but enough money to solve it is unlikely to be forthcoming and would certainly be fiercely competed for by other specialties. The real problem is systemic and cultural. We as academic surgeons who are established (and basic scientists will find it hard to believe that I, with my one R01 that I haven't renewed yet, count as "established" as a surgical investigator) must somehow find a way to attract, encourage, and support new surgeons interested in research. At the current rate of decline, I fear that by the time I retire the NIH-funded surgical investigator will be as rare as a brown hair in my head in a sea of gray will be.
1. Mann, M., Tendulkar, A., Birgir, N., Howard, C., Ratcliffe, M.B. (2008). National Institutes of Health Funding for Surgical Research
. Annals of Surgery, 247(2), 217-221.
I'm not sure I see as big a problem. It seems like the raw number of surgeon grantees has actually gone up by quite a bit during the last 23 years (23% according to your numbers). I assume the raw number of practicing surgeons has not increased by 23%?
If I am correct, then a greater percent of surgeons are doing research compared to 23 years ago.
The NIH budget more than doubled over the past 23 years making entirely new fields of study. There is no reason why one group of investigators in one subfield should match the NIH-wide growth rate. I'm not sure why it's relevant that there are now more physician non-surgeon researchers. There are entirely new fields of medicine along with the growth of MD/PhD programs (few of whom have the time/sanity to specialize in surgery). A better comparison would be between surgery an another "old" field like obstetrics or pediatrics.
Am I missing something here? Are there much larger gaps in evidence based medicine for surgery than other medical fields?
I'm a PhD in a surgery department. I think the number of PhD's in situations like mine has gone up, so that much more research is being funded and performed in surgery departments but by PhD's rather than the MD's. Another thing to consider is that non-NIH funding for research in surgery departments is probably fairly common. In my department, many MDs are funded through clinical revenue, Department of Defense funding, the VA, ACS, etc.
Personally, I find the knee jerk shifting of blame to a reduction in resident hours disturbing and dishonest. Practice makes perfect, yes. But no amount of practice will do anyone any good after they've been up all night, or two nights. the 80 hour work week is only a STEP in the right direction. This type of thinking puts patients at risk and is nothing more than the "well, it was good enough for us" mentality. The old timers should realize that their system wasn't working, won't work, and shouldn't ethically be attempted. A physician who has been up all night has about as much awareness and coordination as a drunk person. This has been studied, but apparently some doctors don't believe in peer reviewed studies. I hear this argument all the time on the news and radio. Old timers lamenting the current state of affairs with, "well, they're just not seeing as many cases as we used to." What difference does it make how many cases you see if you're so tired that you can barely keep yourself from falling into the sterile field while you're holding the retractor for hours at a time? It just makes me want to shake them when they say these things.
To me, perhaps the key values that surgeons bring to translational research is access to tissue for molecular studies and the ability actually to assist in the design and testing of new medical devices.
I agree with this but I think there is a fundamental shift occurring in the training of biomedical investigators. One of the most important researchers in the 20th century, in my mind, was Judah Folkman, an MD who went on to be a pioneer in angiogenesis research. Additionally, at my institution, most of the 'big names' who were Nobel laureates and made seminar contributions to their fields were also MDs. However, these people were all trained much earlier in the century at a time when PhDs did mostly fundamental basic science that may not have been very 'biomedical'. The training available now for PhDs is far more advanced in getting scientists (PhDs) without clinical training to tackle clinically relevant problems and to think like a clinician (with all due respect to the MDs that indeed do good science). For instance, in my PhD training I did a fellowship where I tackled a number of diseases in depth, rounded with docs, observed surgeries, and met and interviewed patients. While I don't feign to think that this can supplant an MD, I do feel very well equipped to tackle disease research and to apply for funding with a focus on a given pathology.
The issue, therefore, may be in the fundamental training of MDs, PhDs, and MD-PhDs. Other institutions- including Hahvahd- are implementing programs designed to make PhDs tackle not just basic science problems in disease/physiology but also in translational medicine/science.
I don't doubt that you do excellent surgery and research (although since I don't know your real name I can't 'Pubmed' you) but the dynamic nature of science, medicine, and healthcare means that the numbers of MDs (and even MD/PhDs) doing research may very well be dwindling in certain specialties.
Hm. people complaining about the 80 hour week remind me a lot of "when I was a kid, we had to walk to school!" etc.