It seems to me that this intellectual spectrum is exhibited by the relationship between an internist, a surgeon, and an academic surgeon. Our medical pals delight in poring over long lists of medications which, after correcting for creatine clearance, can uncover a negative drug–drug interaction. For a surgeon, however, memorizing the PDR is painfully dull. We fancy ourselves doing something. We may logically doubt the presence of an occult abdominal abscess, but we prefer to explore the patient, dwindling in the ICU from multiple organ failure, rather than to diligently document deterioration and demise. As surgeons, with gratifying frequency, we win. When we don’t, however, it is hard to pin the blame on others. Like FDR we must “… admit it frankly …” and go on. But just “going on” should not be sufficient. Repetition of error cannot be counted as experience. The responsibility of the academic surgeon is to codify wins, catalogue failures, and share these experiences in a rigorously disciplined scientific fashion.
In this issue of Annals of Surgery, Drs. Robertson, Klingensmith, and Coopersmith conclude that research time during a surgical residency is good. This study gloriously confirms my own bias—so, I like it. I believe that the exercise of disciplined observation, garnered during 2 “research years” buried in the middle of a surgical residency, will mature a surgeon into a superior clinician even if he/she never does another lick of “science.”
Physics is easy. Each time you drop a brick out of the window, it goes down time after time. Clinical investigation is less predictable and therefore harder to “get right.” And with psychosocial investigations, we are almost surprised when someone can repeat our work. The authors of “Long-term Outcomes of Performing a Postdoctoral Research Fellowship During General Surgery Residency” have accepted a formidable challenge. Like my bias, their subliminal goal is to validate a surgical research experience. As FDR said, we inhabit the “… place of moral leadership.” As academic surgeons, we leap at the opportunity to examine our best practices, and make them even better. And identify the less effective practices, and change them. But practice-changing conclusions obligate persuasive analysis of a disciplined collection of credible data. That is what you learn during surgical research training. So how did Drs. Robertson, Klingensmith, and Coopersmith do it?
Like me, they clearly began with the conclusion that surgical research training is good. Their chosen study population was the Barnes Surgical Residency Program during a 15-year period of peak academic performance. Selling the Barnes residency as representative of all surgical programs nationally is analogous to depicting the constitutional views of the Founding Fathers as representative of all Americans. Or, personally, I like to think of Olympic medalists as reflective of the way I run or swim. So, I won’t challenge this epidemiologic three-pointer. The authors then chose a “postdoctoral research fellowship” as their independent variable. “Success” or “contributions” or “happiness” as an academic surgeon are difficult to quantify, so the authors settled upon the elusive downstream “extramural or NIH research funding” as their dependent variable.
So, in this study, the authors have smoked out 23 former surgical residents who subsequently received independent faculty funding and 13 who were awarded NIH grants. Not bad! Robertson, Klingensmith, and Coopersmith then exhibit the fruits of their own research training in the selection of their study population, or dictatorial denominator. This pivotal denominator decision completely controls the statistical impact of their conclusions. About 1000 surgeons are certified by the American Board of Surgery annually; so during the 15-year span of this study, 15,000 surgeons graduated from surgery programs nationally. Thirteen NIH grant recipients out of 15,000 surgeons (0.13%), not too impressive, so let’s find a better denominator. How about just the surgical residents graduating from the Barnes program? Barnes finishes about 7 categorical residents annually for a total of 105 surgeons in 15 years. Now 13 of 105 (or 13.5% academic success), that’s more impressive! But can the authors do even better? They canvassed notably responsive surgical graduates with a questionnaire examining their academic success and 75 (71%) of 105 answered. Now, 71% is a gentleman’s C minus or a passing grade, so what’s the problem? The wisdom, and indeed beauty, of this sample size reduction strategy is clearly that the folks who returned for your high school reunion were the ones with jobs (something to be proud of), not the homeless guys. Now, we’re making progress. Thirteen of 75 respondents is a dazzlingly commendable 20% NIH funding success.
With gratifying and enviable frequency, Barnes surgical residents who weathered a tour in the laboratory did achieve subsequent extramural research funding. Now the ponderously pedantic, pocket-protected pragmatist might propose that the authors restrict their conclusions to: “A postdoctoral research fellowship accomplished by a surgical resident at Barnes conferred a 35% likelihood of subsequent independent funding and a 20% chance of NIH support.” How dull. This doesn’t apply to me, so why should I care? So, in an imaginative spasm of analytical discipline, the authors have promoted the dependent variable from “downstream independent or NIH funding” to “long-term outcomes …” (please note the authors’ title).
Personally, I would never accuse the authors of promoting their research conclusions beyond the limits of their data. I do, however, sense some intellectual momentum here. Although this is an inspirationally slippery slope, permit me therefore to further amplify the authors’ very solid conclusions just one small, but highly credible, additional step from “long-term outcomes” to “world peace” (please see my title).2