I

Atpgttttrpboitcm

a

Vl

1

Research Residents’ Choices ofAcademic Radiology Careers

Bruce J. Hillman, MDa, Spencer B. Gay, MDa

Purpose: In 1993, the University of Virginia instituted a separate research residency program in theResidency Match (one resident per year). As part of the program, research residents took their second year awayfrom clinical service for research under mentors and enrolled in a postgraduate educational program. Thisarticle evaluates what fraction of these residents eventually chose research careers.

Methods: The article reviews the career choices of 51 residents (8 research residents, 43 conventional clinicalresidents) who entered residency training between 1993 (the year the research residency program began) and1999 (the last year for which postfellowship data were available) and compares the rates of research residents andconventional clinical residents choosing academic careers.

Results: Of the 8 research residents, 5 are currently in academic practice, but 2 of the 8 were dismissed fromthe program for poor clinical performance. Of the 43 clinical residents, all but 2 are in community practice; 6of these residents either were dismissed or left the program voluntarily. Research residents had greater difficultywith board examinations than clinical residents.

Conclusions: A dedicated research residency can increase the rate of trainees ultimately choosing academiccareers, but there are significant risks for these trainees that they will not be able to sufficiently accommodateboth clinical and research training roles.

Key Words: Resident education, research, academic careers

J Am Coll Radiol 2005;2:1030-1032. Copyright © 2005 American College of Radiology

inatwatytitf“srstdhrsa

NTRODUCTION

cademic radiology is the engine for the continuation ofhe specialty. Not only do academic residency programsroduce the vast majority of future practicing radiolo-ists, but they also train the individuals who will conducthe research that fuels technological advancement andhat underpins future practice. Numerous past publica-ions have mourned that so few radiology trainees (orhose of other specialties) choose academic—specifically,esearch—careers, speculated on why this is the case, androposed strategies to reverse this trend [1–4]. There isroad agreement among the authors of these articles thatne major obstacle is a lack of dedicated research train-ng, so that graduates of radiology residencies are unableo compete for grant funds and associated time free fromlinical responsibilities once they assume faculty appoint-ents.One approach to addressing improved research train-

Department of Radiology, University of Virginia, Charlottesville, Va.

Corresponding author and reprints: Bruce J. Hillman, MD, University ofirginia Health System, Department of Radiology, PO Box 800170, Char-

eottesville, VA 22908; e-mail: bjh8a@virginia.edu.

030

ng is a dedicated “research residency” that provides sig-ificant research training in addition to clinical trainingnd meets qualifications for clinical competence and cer-ification by the American Board of Radiology. In 1993,e initiated such a program for one resident each year,

ccepted into our program through a separate listing inhe residency match. These residents underwent first-ear training similar to that undergone by their peers inhe conventional clinical program, which included anntroduction to all major technologies and organ sys-ems. During the second year, the research residents wereree from daytime clinical responsibilities (they tooknight-float” call and attended clinical conferences); theypent the second year of their residencies in either labo-atory or clinical research under the supervision of as-igned mentors. The research residents were also requiredo enroll in a postgraduate program, which provided theidactic background for the specific research paths theyad chosen. At the end of their research year, researchesidents rejoined the conventional program, taking theame number and types of clinical rotations as their peersnd graduating the program in the same year as their

ntering class, with the exception that they did not par-

© 2005 American College of Radiology0091-2182/05/$30.00 ● DOI 10.1016/j.jacr.2005.07.013

tw

rtt

M

Wigteigwpeoady

itipdcw

R

Ftc7egwi

imOaisaCK

tD

rfp

pBdpclncwc

D

Trtigiirmrmwo

cfdwwmcssafbdydrro

sps

Hillman, Gay/Research Residents’ Choices of Academic Radiology Careers 1031

icipate in our usual 1-month research requirement, norere they allowed the usual 3 months of electives.This article tracks the career choices of our research

esidents and compares them with the choices made byheir peers in the conventional clinical program duringhe same time frame.

ETHODS

e generated a spreadsheet of all individuals acceptednto our research and conventional clinical residency pro-ram beginning in 1993 and ending with those initiatingheir residencies in 1999 (all later residents are still inither residency or fellowship training). Items includedn the spreadsheet were residents’ years of entry into andraduation or dismissal from the program, whether theyere admitted into the research or traditional clinicalrogram, their results on American Board of Radiologyxaminations, and where they currently practiced at leastne year after graduation from our residency program (tollow time for at least 1 year of fellowship). Only resi-ents past fellowship training were included in the anal-sis.

Data were collected from our residency files and fromnformation we keep in support of our alumni associa-ion mailing list. In a few cases, we tracked down missingnformation from individuals by telephone. The data areresented observationally as ratios of the number in aca-emic practice over the total number matriculated. Be-ause of the small number of individuals studied, thereas no attempt to perform a statistical analysis.

ESULTS

ifty-one residents entered our training program duringhe period from 1993 to 1999. Eight of these were ac-epted into the research residency (1 per year during theyears of this evaluation, plus an additional resident who

ntered the regular clinical program in 1992 but wasranted permission to transfer into the research residencyhen it was initiated in 1993), whereas 43 matriculated

n the clinical program.Five of the 8 research residents are currently practicing

n academic settings. However, 2 residents were dis-issed from the program for poor clinical performance.ne of the 2 dismissed research residents found a slot in

nother radiology residency program, completed train-ng, and is among the 5 in academic practice. Our re-earch residents who chose academics currently practicet Northwestern University, Wake Forest University, thehildren’s Hospital of Philadelphia, the Memorial Sloanettering Cancer Center, and the Mayo Clinic.Of the 43 residents who graduated from our conven-

ional clinical program, 2 are in academic practice, at

uke University and Emory University. Two clinical t

esidents were dismissed from the program for poor per-ormance during the study period, and 4 others left forersonal reasons.Another measure frequently used to evaluate residents’

erformance is how they performed on the Americanoard of Radiology examinations. Two research resi-ents initially failed sections of the written boards butassed on repeat examination, and 2 research residentsonditioned sections of their oral board examinations butater became board certified. Another research residentow in practice has yet to take the oral examination. Inontrast, only 1 traditional clinical resident failed theritten examination during the same period, and none

onditioned or failed the oral boards.

ISCUSSION

his experience demonstrates that a dedicated radiologyesearch residency program in which matriculants signalheir interest in undergoing significant research trainingn advance of their acceptance into the program canuide trainees into choosing academic careers. However,t also is evident that there are significant risks for thesendividuals who must devote energies to both clinical andesearch training. Although research residents muchore frequently chose research careers, they also had a

elatively high rate of dismissal for poor clinical perfor-ance; among those who remained in the program, thereas a much higher risk for having to repeat written andral board examinations before becoming certified.

It is unclear whether these results are related to ourhoices of whom to admit into the research program or aunction of particulars of the program itself. One of the 2ismissed residents was an unmatched medical studenthom we recruited the year the research residency slotent unmatched. The other had a low ranking on ouratch list. Most likely, both application evaluation pro-

edures and the intensity and nature of combining re-earch and clinical training played a role. The criteria forelection into the research residency differed from thosepplied to our conventional clinical residency in that theormer required a demonstration of interest in researchy past performance, an interest in research displayeduring the interview, and a willingness to commit to aear of research signified by ranking the research resi-ency program on a resident’s match list. Because theesearch residency was a separate listing in the match, ouresearch residency rank list was distinct and separate fromur rank list for the clinical program.

Our research residency confers benefits but also placesignificant added stress on trainees matriculating in thatrogram. Even during the year a resident is away fromcheduled clinical rotations, he or she is expected to con-

inue to advance clinically through night call experiences

atrir

tasaIpo

lddpaitmtywiisps

dcottlqrt

ue

tstitrodacrdswfn

tmrt

R

1

2

3

4

5

6

1032 Journal of the American College of Radiology/Vol. 2 No. 12 December 2005

nd attendance at clinical conferences. Given the reduc-ion from 4 to 3 clinical years, faculty members felt thatequiring a research resident to participate in these activ-ties was essential to his or her eventual success as aadiologist.

In addition, research residents simultaneously carryhe full course load of master’s-level graduate studentsnd work on multiple research projects. When the re-earch residents return to clinical rotations in their thirdnd fourth years, they continue to work on their research.ndeed, a number of their projects have become majorrotocol design and statistical resources for the researchf other residents and even faculty members.

Despite some negative aspects of our results, we be-ieve that overall, the goals we had for our research resi-ency have been met. There are several reasons why ra-iology trainees fail to choose academic careers that ourrogram largely supercedes. First, even though radiologyttracts many of the very brightest medical students intots training programs, many are from the start directedoward a lifestyle that they imagine is embodied in com-unity practice. Although there are exceptions, many of

hese residents fail to become intellectually engaged be-ond rote memory; they largely address their efforts to-ard learning clinical radiology and passing board exam-

nations. Our research training program selects residentsndependently of our conventional clinical program, en-uring that there is sufficient interest on the part of ap-licants by their foreknowledge that research will be aignificant aspect of their training.

Going beyond aspects of selection, our program ad-resses one of the other factors that help determine careerhoice: experience in research during training [5,6]. Inur opinion, one reason so many talented trainees findhe prospect of an academic career so daunting is thathey feel unprepared to develop their own research andead the interdisciplinary teams that modern science re-uires. Our research residency provides a yearlong expe-ience in research and supports that experience with in-

ensive didactic education that provides greater

nderstanding of research tools, design, and analysis thanxperience alone.

The major concern for readers of this study should behat the sample was so small and that we observed only ahort period of time. Although this is true, this is a topichat does not lend itself well to studying large groups ofndividuals, because the number of similar programs inhe United States is scant, few have the duration of expe-ience that we describe, and there is no general consensusn how these programs should be organized. Still, theifferences in the career choices of our research residentsre strikingly different from those of our conventionallinical trainees. We intend to follow the careers of ouresearch residents into the future to see if the trendsepicted in this article continue, to what extent our re-earch residents persist in their academic careers, andhether they are successful in obtaining extramural

unding and publication (3 of the 6 currently have sig-ificant federal or foundation funding).We offer this experience as evidence that if we believe

hat the development of academic careers is an urgentandate for the specialty, the promotion of dedicated

adiology research residency programs is one approachhat bears consideration.

EFERENCES

. Alderson PO, Breslin LB, Becker GJ, et al. Enhancing research in academicradiology departments: recommendations of the 2003 Consensus Confer-ence. Radiology 2004;232:405-8.

. Hillman BJ, Putman CE. Fostering research by radiologists. Radiology1992;182:315-8.

. Goldman E, Marshall E. Research funding: NIH grantees—where have allthe young ones gone? Science 2002;298:40-1.

. Hoffman JM, Staab EV, Begg L, Croft BY, Menkens AE, Sullivan DC.Training the next generation of imaging scientists and clinicians. AcadRadiol 2000;7:678-80.

. Hillman BJ, Fajardo LL, Witzke DB, Cardenas D, Irion M, Fulginiti JV.Influences affecting radiologists choices to enter academic or private prac-tice careers. Radiology 1990;174:561-4.

. Hillman BJ, Witzke DB, Fajardo LL, Fulginiti JV. Research and researchtraining in academic radiology departments. Invest Radiol 1990;25:587-

90.