Results of the 2001 Survey of the American Association of Academic Chief Residents in Radiology

Results of the 2001 Survey of the AmericanAssociation of Academic Chief Residents

in Radiology1

Faraz A. Khan, MD, Sanjeev Bhalla, MD, R. Gilbert Jost, MD

Rationale and Objectives.The American Association of Academic Chief Residents in Radiology (A3CR2) conducts anannual survey of residency training programs. The survey data allow comparative analyses to be performed among train-ing programs regarding resident education, benefits, clinical demands, and other resident-related issues.

Materials and Methods. Questionnaires were mailed to all accredited programs registered in the A3CR2 database (150total programs). The yearly questions address demographic information concerning the individual programs including aca-demic affiliation, number of radiologic examinations performed per resident, changes in the number of residents per pro-gram, and financial compensation. The rotating questions, which are revisited every 4 years, focused on preparation forthe American Board of Radiology examinations, educational issues, and employment outlook.

Results.Completed surveys were received from 55 programs (37.3%). Among the responses, three issues prevailed:(a) increased number of examinations performed per resident, with resultant decrease in educational time, (b) decreasedquality of education in all sections of radiology and decreased number of educational conferences, and (c) improved em-ployment outlook, but continued trend of fewer residents choosing a career in academics.

Conclusion.The current boom in the radiology job market and the increased number of radiologic examinations per-formed annually appear to adversely affect radiologic education through a decreased number of conferences and an in-creased number of radiologic examinations performed per resident.

Key Words. Radiology and radiologists, residency training, education.

© AUR, 2002

The American Association of Academic Chief Residentsin Radiology (A3CR2) includes chief residents from allaccredited programs, including university, community,and military hospitals. The organization meets annually todiscuss issues related to the training of diagnostic radiol-ogy residents. In addition, the A3CR2 conducts an annualsurvey of residency training programs.

The survey has been conducted since 1971, and thedatabase is maintained at the Mallinckrodt Institute of

Radiology. The survey is divided into two parts, yearlytopics and a group of rotating topics revisited every 4years. The yearly topics consist of general issues (eg,number of hospitals, total number of beds, salary informa-tion, residency training size and location, number ofwomen). The rotating topics of the 2001 survey includedpreparation for the American Board of Radiology (ABR)examination, educational training issues, and employmentoutlook.

MATERIALS AND METHODS

Questionnaires were mailed to chief residents at 150accredited diagnostic radiology residency programs in

Acad Radiol 2002; 9:89–97

1 From the Mallinckrodt Institute of Radiology, 510 S Kingshighway Blvd, StLouis, MO 63110. Received and accepted August 16, 2001. Address cor-respondence to R.G.J.

© AUR, 2002

89

Special Report

the A3CR2 database, including university, community,and military hospitals. Data were tabulated and ana-lyzed with commercial database software (Excel 97;Microsoft, Redmond, Wash). Information obtainedfrom the 1991–2000 A3CR2 surveys was incorporatedinto this database.

For analysis, programs were divided into four sizecategories based on the total number of residents ineach program: small (�13 residents), medium (14 –25residents), large (26 – 40 residents) and jumbo (�41residents). Programs were also divided into five regioncategories based on the location of the program: North-east (Conn, Del, Mass, Md, Me, NH, NJ, NY, Pa, RI,Vt), Southeast (Ala, Ark, Fla, Ga, Ky, La, Miss, NC,SC, Tenn, Va, WVa), Central (Iowa, Ill, Ind, Mich,Minn, Ohio, Mo, Wis), West (Ariz, Colo, Idaho, Kan,Mont, ND, Neb, NM, Nev, Okla, SD, Tex, Utah,Wyo), and Pacific (Alaska, Calif, Hawaii, Ore, Wash).The number of radiologic examinations per residentwas calculated by dividing the total number of exami-nations performed at the institution as reported on thesurvey by the total number of residents in that institu-tion and subtracting the percentage of examinationsreported as read by staff only.

RESULTS

Survey Response Rate

Chief residents from 56 (37.3%) of the 150 accrediteddiagnostic radiology residency programs in the A3CR2

database completed and returned the questionnaire. Thisresponse rate was lower than the response rate last year

and continues the trend of a less than 50% response rateas seen in the past few years. There was a 43% responserate in 1999–2000, a 24% response rate in 1998–1999, a32% response rate in 1997–1998, and a 41% responserate in 1996–1997 (Fig 1).

Number of Responding Programs and ResidentsRepresented

Chief residents from five small, 30 medium, 17 large,and four jumbo programs completed the survey (Fig 2a).Of these 56 programs, 20 were Northeast region pro-grams, 13 were Southeast region programs, 11 were Cen-tral region programs, eight were West region programs,and four were Pacific region programs (Fig 2b). Therewere 1,367 residents training in these programs at thetime of the survey.

Average Number of Radiologic ExaminationsPerformed per Resident

The average number of radiologic examinations per-formed per resident was 9,026 in small programs, 10,617

Figure 1. Response rate to prior surveys. Note the continuedtrend of low response rates.

Figure 2. Total number of residents who responded to thesurvey, according to (a) size of program and (b) region inUnited States where program is located. Numbers above barsindicate numbers of residents.

KHAN ET AL Academic Radiology, Vol 9, No 1, January 2002

90

in medium programs, 10,159 in large programs, and10,839 in jumbo programs. This represents a general in-crease in the number of examinations performed per resi-dent compared with 1996–1997 (Fig 3). The average per-centage of cases read by staff only was 20.2% and waslowest in the jumbo programs. This represents a decreasefrom 23% in 1996–1997. The average percentage ofcases read by residents only, without staff supervision,was 1.1%; this represents a decrease from 2.2% in 1996–1997. The average percentage of procedures performed byresidents only, without staff supervision, was 5.6%, whichwas not significantly changed from 1996–1997 (6%).

Changes in Residency Class SizeTo gauge the change in residency class size, chief

residents were asked to state the change in the numberof residency match positions offered in 2001 comparedwith the number in 2000. Survey responses indicatethere was a slight increase in the number of match po-sitions offered (a mean increase of 0.46 positions perprogram). This is in comparison with a slight decreasein the number of match positions offered in 1997 com-pared with 1996 (a mean decrease of 0.35 positions perprogram) (Fig 4).

Women in Diagnostic Radiology ResidencyTraining

Among the programs responding to the survey, 24.4%of all radiology residents were women (Fig 5). This rep-resents no noticeable change from 25% obtained in the

Figure 3. Average number of examinations performed perresident, according to (a) size of program and (b) region inUnited States where program is located. Gray bars indicate1996–1997 results, and black bars indicate 2000–2001 results.

Figure 4. Average change in match positions. There was asmall increase between 2000 and 2001 (black bars) in match po-sitions compared with a small decrease between 1996 and 1997(gray bars).

Figure 5. Percentage of all radiology residents who werewomen. The representation of women in radiology residency pro-grams has remained unchanged since 1991.

Figure 6. Average salary of 1st-year (bar) and 4th-year (dot)residents. Graph shows expected increases in resident remuner-ation, correlating with yearly inflationary changes.

Academic Radiology, Vol 9, No 1, January 2002 2001 A3CR2 SURVEY

91

1997 survey. In comparison, 34%–39% of U.S. medicalschool graduates in the 1990s were women. Small radiol-ogy residency programs were noted to have the highestpercentage of women residents.

Salary and BenefitsThe average salary of a 1st-year radiology resident in

2000–2001 was $36,559 (range, $30,000–$50,000), com-pared with $32,366 (range, $27,000–$42,000) in 1996–1997 (Fig 6). The average salary of a 4th-year radiologyresident in 2000–2001 was $41,553 (range, $36,000–

$51,000), compared with $37,441 (range, $31,000–$55,000)in 1996–1997 (Fig 6). These values are consistent withthe projected income based on the consumer price indexand yearly inflationary changes. Most programs (71.8%)offer a “book and travel” fund; 9% of the programs offerless than $250 per year, 61% offer $250–$1,000 per year,and 11% offer more than $1,000 per year.

Preparation for the ABR ExaminationChief residents were asked about how their programs

handle preparation for the ABR examination. Ninety-eight

Figure 7. The most important resources used by residents for the physics section of the ABR examination, according to size ofprogram. (a) Results from 2000–2001 survey. (b) Results from 1996–1997 survey. Overall, RAPHEX examinations and physicscourses rate as the most important resources in preparation for the ABR physics examination.

Figure 8. The most important resources used by residents for the written portion of the ABR examination, according to programsize. (a) Results from 2000–2001 survey. (b) Results from 1996–1997 survey. Overall, textbooks and “remembered” old test ques-tions rate as the most important resources in preparation for the ABR written examination.


92

percent of the programs surveyed offered a physicscourse to help in preparation for the ABR physics ex-amination; 76% of the residents rated this course asgood or excellent. The most important reported re-sources in preparing for the physics examination wereRAPHEX practice examination books (36%) and phys-ics courses (33%) (Fig 7). Overall, there were fewerpeople who found “remembered questions” to be usefulcompared with those who took the examination in thepast. The most important reported resources in prepar-ing for the written examination were textbooks (61%)and “remembered questions” (29%). Ninety-four per-cent had access to “remembered questions” from previ-ous examinations, and 82% found them to be at least

useful in studying for the ABR examinations (Fig 8). Morerespondents in 2000–2001 compared with 1996–1997thought the “remembered questions” were more useful.

Chief residents were also asked about preparation forthe oral portion of the ABR examination. Eighty-six per-cent of respondents reported their programs offered in-house oral board reviews, and 66% reported their pro-grams had mock oral examinations, compared with 71%and 51%, respectively, in 1996–1997. Overall, 83% ofresidents reported to have attended commercial reviewcourses, and 84% of residents were allowed time off toattend these courses. Only 41% of those who attendedthese courses received tuition compensation. These per-centages have not substantially changed over the years,

Figure 9. Average number of hours residents spent per week in (a) didactic lectures and (b) case conferences, according to pro-gram size. The reported amount of time spent in case conferences in 2000-2001 decreased compared with that in 1996–1997.

Figure 10. Adequacy of training. There was decreased satisfaction in 10 of 16 areas of training in 2000–2001 (black bars) com-pared with satisfaction in 1996–1997 (gray bars). (a) Results for cardiology, neuroradiology, neuroangiography, MR imaging, pe-ripheral angiography, interventional angiography, nonangiographic interventional, and obstetric-gynecologic radiology. (b) Resultsfor vascular US, US, pediatric radiology, nuclear medicine, thoracic imaging, musculoskeletal radiology, genitourinary radiology,and gastrointestinal radiology.


93

except that fewer programs provide tuition compensationnow compared with 1996–1997.

CurriculumChief residents were asked to rate their background in

making wise choices about equipment purchasing, super-vising quality control programs, and accurately explainingthe risks and benefits of a radiologic procedure to pa-tients. Only 25% of respondents thought they had suffi-cient training to purchase equipment or supervise qualitycontrol programs; however, 95% thought they had suffi-cient training to accurately explain the risks and benefitsof a radiologic procedure to patients. More than 75% ofrespondents desired more training to improve their abilityto purchase radiology equipment and supervise qualitycontrol programs.

Overall, the residents attended didactic conferencesan average of 4 hours per week and case conferencesan average of 4.7 hours per week (Fig 9). These num-bers are approximately 5% lower compared with thosefrom 1996 –1997. Jumbo programs provide the mosthours per week of didactic and case conference time(Fig 9).

Chief residents were also asked to rate their training invarious areas of radiology as “more than adequate,” “ade-

quate,” or “inadequate.” On average, training in all areasof radiology was rated adequate or better than adequate,with the exception of obstetric/gynecologic ultrasound(US), vascular US, cardiac radiology, and magnetic reso-nance (MR) imaging. Training was rated the highest inneuroradiology, nuclear medicine, nonobstetric US, andvascular and interventional radiology. The ratings on thequality of instruction were decreased in 10 of 16 catego-ries in comparison with 1996–1997 (Fig 10).

The 6-week lecture series at the Armed Forces Insti-tute of Pathology (AFIP) remains a central componentof training in pathology. The AFIP course was avail-able to all residents in 96% of programs, and 95% ofthe residents in the programs surveyed attended theAFIP course. This has not changed in comparison with1996 –1997.

Socioeconomic IssuesChief residents were asked a series of questions re-

garding their perceptions about socioeconomic issues.Seventy-seven percent of the survey respondents werevery well or satisfactorily informed about business andpolitical matters that affect medicine and radiology.Sixty-eight percent reported some formal instruction on

Figures 11–13. (11) Residents’ concerns about a career inacademia. Salary concerns continue to dominate resident per-ceptions of academic radiology. Gray bars indicate results from1996–1997; black bars indicate results from 2000–2001.(12) Residents’ concerns about a career in private practice. Thevolume of studies and business concerns stand out as major prob-lems in private practice radiology. Gray bars indicate results from1996–1997; black bars indicate results from 2000–2001.(13) Residents’ perceptions about the future of a career in radi-ology. There was decreased pessimism related to employmentand health care issues compared with perceptions in 1996–1997.Gray bars indicate results from 1996–1997; black bars indicateresults from 2000–2001. Vol/Rad � volume per radiologist, Appro-priate � appropriate use of imaging studies, Quality � quality ofhealth care in the United States, Ins. Coverage � problems withinsurance coverage for underinsured or uninsured people.


94

these matters by their training programs. When askedtheir thoughts about a career in academic radiology,chief residents rated low salaries, inadequate vacationtime, pressure to publish, and not enough academictime as their main concerns. In 1996 –1997, too muchpressure to publish and job security were bigger con-cerns (Fig 11).

When asked their thoughts about a career in privatepractice, chief residents rated an overly high clinical vol-ume, time pressure (throughput issues), and anxiety re-garding business matters of practice as their biggest con-cerns. Job security had decreased as a major concern;only 23% cited this in 2000–2001, whereas 67% cited itin 1996–1997 (Fig 12). Fifty-nine percent of respondentsindicated that they were likely to join a private practice,13% indicated they were likely to stay in academics, and28% were undecided.

Chief residents were also asked to rate various so-cial and economic issues with regard to future outlookon a scale of 1–5, with 3 representing “no change.”Chief residents believed there may be slight improve-ments in salaries, job security, total number of imagingprocedures, and number of imaging procedures per-formed by radiologists. They did not believe therewould be substantial changes in the quality of the workenvironment, vacation time, or quality of health care inthe United States. A decrease in the appropriate use ofimaging studies and continued problems with insurancecoverage for the underinsured or uninsured were ex-pected. In 1997, there was more overall pessimism re-garding all issues concerning the future of radiologicpractice and medicine (Fig 13).

DISCUSSION

The 2000 –2001 annual survey of chief residentsshows a continuing trend of low survey response rates.The reasons for this poor response rate are uncertain,given the recipients of the survey are a highly moti-vated group of people. Pressures from increasing clini-cal duties, board preparation, and chief resident dutiesmay explain the poor response rate. As in recent years,this survey echoes the trend of a booming private andclinical practice in radiology with increasing clinicaldemands, often to the detriment of education becauseof decreasing conference and study time. The demo-graphics of radiology residencies have not changedsubstantially: The number of positions available has

changed minimally, the number of women in the fieldhas remained stable, and salaries have remained rela-tively stable. Because of the job market and the per-ceived increased pressures in academic radiology, mostchief residents are choosing a career in private prac-tice.

Compared with findings in 1996–1997, the averagenumber of examinations performed per resident has in-creased dramatically. This is consistent with the increasedvolume noted throughout the field of radiology in bothprivate practice and academics. This may also be due toincreased after-hours demand for radiology services,which are covered primarily by radiology residents. Therehas been a slight decrease in the percentage of studiesbeing read by staff, especially at the jumbo institutions.This may be related to the pressures of academic medi-cine on the staff. There has been a slight decrease in thelimited percentage of studies being read by residentswithout staff supervision and no overall change in thesmall percentage of procedures being performed by resi-dents without staff supervision.

The residency class sizes are, on average, increasing inprograms of all size and in all regions of the country.This is a reversal of the trend noted in the mid-1990s,during the national debate over the roles of governmentand market forces in reducing the number of training po-sitions for specialists (13). In fact, there are not enoughgraduating radiology residents to fill available job posi-tions.

The percentage of women in radiology training is sta-ble, remaining below the percentage of women amongU.S. medical graduates. The reasons behind this continu-ing pattern are not clear.

From 1991 through 2001, increases in resident salarieshave generally paralleled the inflation rate as judged bythe consumer price index.

The survey once again shows that it is common prac-tice for residents to remember and transcribe ABR exami-nation questions in an effort to help future examinees pre-pare for the examination and that such “rememberedquestions” are widely available. These questions werebelieved to be an important resource in studying for thewritten diagnostic portion of the examination but not avery helpful resource for the physics examination. It isinteresting to note the variable value of this resource forthe two examinations, with remembered questions stillbeing considered a critical resource by a substantial num-ber of residents.


95

As preparation for the oral board examination, thevast majority of the surveyed radiology residency pro-grams offered an internal review course and a mockoral board examination. However, a large percentage ofresidents still attend a commercial review course out-side of their own institution. Support for commercialreview courses in terms of time off is very commonamong all programs, but funding is provided by lessthan half the programs. This high number of residentsattending the commercial review courses may be a re-flection of the perceived quality of the internal reviewcourses or may be due to anxiety caused by the oralboard examination.

The increased pressures of academic medicine haveled to decreased time for formal resident conferences.The larger residency programs have maintained theirconference time, while the smaller programs have seentheir conference time decrease compared with 1996 –1997. This may be due, in part, to the booming radiol-ogy job market and the current loss of academic radiol-ogists to private practice. Assuming that the rate ofdeparture is similar in small and large programs, theloss of any one individual will have a larger impact onthe smaller programs. The increase in volume of stud-ies read by another remaining single staff memberwould be more noticeable in a smaller program. Theeffect on conferences would also be expected to bemore noticeable.

Chief residents on average rated their training ade-quate or better than adequate in most areas of radiol-ogy, with the exception of obstetric/gynecologic US,vascular US, cardiac radiology, and MR imaging.These results are not unexpected given that obstetric/gynecologic US, vascular US, and most cardiac radiol-ogy is performed by persons in nonradiologic special-ties at most institutions. Furthermore, training in MRimaging is often perceived as a weakness, becausemost residents receive limited training in this specialtylate in the course of their residency training. Often,ample rotation time is not allotted due to pressuresfrom busier services, which need coverage. Trainingwas rated the highest in neuroradiology, nuclear medi-cine, nonobstetric US, and vascular and interventionalradiology. This is surprising considering the trainingwas rated the highest in subspecialties with certificatesof added qualification examinations, in which fellowsare frequently trained along with residents. This mayindicate that fellows provide an additional educationalbenefit for residents. The decreased overall ratings in

the quality of instruction in the majority of the catego-ries, in comparison with 1996 –1997, likely correlateswith the trend in decreased conference time. This islikely due to decreased time for teaching by faculty,which is a result of the increasing pressures of aca-demic medicine.

Chief residents appear to be better informed aboutbusiness and political matters affecting medicine and radi-ology than were residents in 1996–1997. This may be aresult of the increased stress on training related to nonin-terpretive skills in the current ABR residency trainingguidelines. Despite the increased training in noninterpreta-tive skills, only a quarter of those responding thoughtthey were sufficiently trained in issues regarding equip-ment purchase and quality control. In many academiccenters, residents are excluded from these aspects of thedepartment because of a perceived lack of interest. Per-haps there is a role for residents and their input in theseissues.

Low salaries, pressure to publish, decreasing aca-demic time, and too little vacation are perceived prob-lems with academic radiology, which drive residents tochoose a career in private practice radiology. Theycontinue to rate heavy clinical volume and anxiety re-garding business matters as their top concerns with pri-vate practice. There is much less concern about jobsecurity now in comparison with 1996 –1997. The vastmajority of committed chief residents indicated thatthey are more likely to join private practice radiology,and only a small minority of respondents are dedicatedto academic radiology.

An obvious limitation of this survey was the lowresponse rate, which may reflect the increased pres-sures and demands on the chief residents. We mustexplore alternatives to this conventional survey in orderto preserve the valuable information it may provide.Perhaps a Web-based survey would increase the re-sponse rate by eliminating the time required to com-plete it.

In conclusion, the A3CR2 survey provides valuabledata on chief radiology residents’ perceptions concerningresidency training, radiology curriculum issues, and socio-economic trends. The responses provide a metric tool thatallows for comparison among various training programsand among different time periods.

REFERENCES

1. Evens RG. Report on a survey of chief residents in academic depart-ments of radiology. Invest Radiol 1972; 27:61–62.


96

2. Heck DV, Vaughan TE, Duncan JR, Evens RG. Results of the 1997 sur-vey of the American Association of Academic Chief Residents in Radi-ology. Acad Radiol 1998; 5:224–231.

3. Oser AB, Baker SM, Wilson AJ, Evens RG. Results of the 1993 surveyof the American Association of Academic Chief Residents in Radiology.Acad Radiol 1994; 1:154–158.

4. Wise SW, Mauger DT, Matthews AE, Hartman DS. Impact of theArmed Forces Institute of Pathology radiologic pathology course onradiology resident performance on the ACR in-training and ABR writ-ten examinations. Acad Radiol 2000; 7:693– 699.

5. Collins J. Curriculum in radiology for residents: what, why, how, when,and where. Acad Radiol 2000; 7:108–113.

6. Goodman CJ, Lindsey JI, Whigham CJ, Robinson A. Diagnostic radiol-ogy residents in the classes of 1999 and 2000: fellowship and employ-ment. AJR Am J Roentgenol 2000; 174:1211–1213.

7. Wood PS, Altmaier EM, Franken EA Jr, Evens RG, Schlechte JA. Fac-tors influencing choice of academic or practice careers in radiology.Invest Radiol 1990; 25:675–677.

8. Vydareny KH, Waldrop SM, Jackson VP, et al. Career advancement ofmen and women in academic radiology: is the playing field level? AcadRadiol 2000; 7:493–501.

9. Levitt SH. Impact of managed care on scholarly activity and patientcare: case study of 12 academic radiology and radiation oncology de-partments. Radiology 2000; 216:618–623.


97

Documents

Results of the 2001 Survey of the American Association of Academic Chief Residents in Radiology