Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2211001499
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Anzilotti, K.
Right arrow Articles by Forman, H. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Anzilotti, K.
Right arrow Articles by Forman, H. P.
(Radiology. 2001;221:87-91.)
© RSNA, 2001

Health Policy and Practice

Relative Attractiveness of Diagnostic Radiology: Assessment with Data from the National Residency Matching Program and Comparison with the Strength of the Job Market1

Kert Anzilotti, MD, Daniel S. Kamin, MD 2, Jonathan H. Sunshine, PhD and Howard Paul Forman, MD, MBA

1 From the Department of Diagnostic Radiology (K.A., D.S.K., H.P.F.), Yale University School of Medicine, 333 Cedar St, SP2-332, New Haven, CT 06520; the Departments of Economics, Yale College, and General Management, Yale School of Management, New Haven, Conn (H.P.F.); and the Research Bureau, American College of Radiology, Reston, Va (J.H.S.). Received September 7, 2000; revision requested October 20; revision received January 24, 2001; accepted February 26. Address correspondence to H.P.F. (e-mail: howard.forman@yale.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
PURPOSE: To quantify medical students’ preferences for various specialties and to compare shifts in specialty preference with overall employment market prospects.

MATERIALS AND METHODS: By using National Residency Matching Program data, a previously validated help-wanted index (HWI), and American Medical Association (AMA) salary data, the authors quantified medical students’ preferences for various specialties and examined the association of shifts in specialty preference with overall trends as seen in the strength of the diagnostic radiology match and its relationship with the job market. The authors established a proxy for the relative attractiveness (RA) of diagnostic radiology (relative to all specialties) in the match, which was calculated by subtracting the percentage fill rate for all specialties from that for radiology. The RA values were plotted for 1990–2000 and compared with trends in the HWI, American College of Radiology data, and AMA salary data.

RESULTS: The RA of diagnostic radiology varied greatly during the past 10 years, with a low in 1996 and a return to its high in recent years. There is a relationship between the RA and economic vitality of diagnostic radiology, with the RA lagging behind the HWI and AMA salary data by 2 years.

CONCLUSION: Medical students appear to have an in-depth understanding of the economic forces at play in the health care job market and incorporate this information into their choice of a specialty.

Index terms: Radiology and radiologists • Radiology and radiologists, departmental management


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Each year, the National Residency Matching Program conducts the "match," which seeks to optimize the placement of medical students and other physicians in institutions and training programs (in 2000, there were 26,360 students and physicians and 3,769 institutions and training programs) (1). Although the overall match rate or percentage of positions filled on the day of the match currently exceeds 90%, it varies considerably across specialties, with some filling fewer than 50% of their positions and other specialties filling almost all positions (1). Furthermore, year-to-year variations can be substantial. It is clear from each year’s data that there are preferences expressed by medical students, but, to our knowledge, this has never been fully explored. Although anecdote has suggested that the changes in health care financing that occurred during the prior decade may have influenced these preferences, to date this has not been quantified or compared. The purpose of our study was to quantify medical students’ preferences for various specialties by using the publicly available matching data and to compare shifts in specialty preference with overall employment market prospects.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Publicly available data from the National Residency Matching Program were collected for the years 1990–2000 (1). We chose to evaluate the specialty of diagnostic radiology in comparison with the overall match, as well as one reference specialty, internal medicine. We chose internal medicine because it is the single largest matching specialty.

For these specialties, we collected specific data about the total number of residency positions available and the total number of residency positions filled, including all 1st and 2nd postgraduate year positions combined. Data were obtained only from the National Residency Matching Program Web site, and there were no missing data for the years studied. We then calculated the percentage of residency positions filled (all specialties) and determined the percentage filled in diagnostic radiology and internal medicine. We use the term fill rate in this article to refer to this value. These values were calculated for each year from 1990 through 2000.

We calculated the relative attractiveness (RA) of a specialty by subtracting the overall fill rate from the specialty’s fill rate. The RA reflects the relative fill rate of a specialty compared with that of all other specialties combined. Given that more medical students will apply to a specialty if they perceive it as more attractive, an increased fill rate in the match directly corresponds to the RA of that specialty. Furthermore, changes in the RA can be compared between specialties to determine any trends. The RA values for diagnostic radiology and internal medicine were then calculated, plotted for the 11 separate matches from 1990 to 2000, and compared to determine any trends in these data.

The annual RA values for diagnostic radiology were then compared with the previously reported help-wanted index (HWI) from Forman et al (2) and Covey et al (3) (Fig 1) and with American Medical Association (AMA) income data from the same periods (4). The HWI used by Forman et al (2) is a composite tally of job advertisements in Radiology and the American Journal of Roentgenology and has been previously shown to directly track the strength of the job market in radiology. The AMA income data are published annually. We calculated a ratio of salaries for the specialty to salaries for mean physicians. Alternatively, we took the means for the specialty and divided that by the mean for all physicians to generate a ratio. The HWI data were available for 1991–1999, and AMA salary data were available for 1990–1997.



View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Graph shows HWI data plotted monthly from 1991 to 1999.

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Relative Attractiveness of Specialties
Although the number of residency positions available during the 10-year period remained relatively stable, the overall fill rate increased (Fig 2). The largest number of positions offered annually were in internal medicine, with an annual average of 4,702; the average number of positions offered annually in diagnostic radiology was 896. There was a relative peak in the number of diagnostic radiology positions offered in 1996, with a subsequent decline in 1997 and then a plateau to the present (Fig 3).



View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Chart shows the total positions offered (bars) and filled ({blacklozenge}) through the National Residency Matching Program for all specialties combined for 1990-2000.

 


View larger version (25K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Graph shows the total radiology positions offered (including postgraduate years 1 and 2) through the National Residency Matching Program from 1990 to 2000.

 
The RA of diagnostic radiology varied substantially during the past 10 years (Fig 4). In 1991, the RA of diagnostic radiology was at its high of 18. It then continued to trend downward until its nadir of -23 in 1996. Since 1996, the RA of diagnostic radiology has trended upward, with 1999 being the 1st year that the RA reentered positive territory at 3. Given the disparity in the number of residency positions offered, the RA of internal medicine has had a proportional but inverse variation when compared with that of diagnostic radiology. The RA of internal medicine reached its high in 1995, which was almost the same time that diagnostic radiology reached a low. Furthermore, these curves had opposing trends during the entire period analyzed.



View larger version (33K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Graph shows the RA of radiology ({blacksquare}) and internal medicine ({blacklozenge}) plotted by year for 1990-2000. Data under the graph correspond to that plotted on the graph. im = internal medicine, rad = radiology.

 
Diagnostic Radiology: RA versus HWI and Salary Data
Comparison of the RA of diagnostic radiology during the 10-year period with the HWI from the same time (1991–1999 only) shows a relative relationship, with the RA lagging behind the HWI by approximately 2 years (Fig 5). As demonstrated with the HWI, the job market for diagnostic radiology was strongest at the beginning of the 1990s, decreased to a low in 1994, and then recovered to exceed its original peak by the end of the decade. The RA of diagnostic radiology shadowed this trend, delayed by only 2 years. Similarly, the same relationship can be seen when the RA of diagnostic radiology is compared with median income data for radiologists from the AMA (stated as a ratio of radiologists’ income and the median income of all physicians) (Fig 6). Again, an association is suggested, with a lag time of approximately 2 years (AMA income data were available only through 1997).



View larger version (26K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Graph shows the RA (left scale, {blacksquare}) and HWI (right scale, {blacklozenge}) plotted by year. Data for RA were obtained from 1990 to 2000; data for the HWI were obtained from 1991 to 1999.

 


View larger version (25K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6. Graph shows the RA versus salary ratio (ratio of the median salary of radiologists to the median salary of all physicians) (4).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
REFERENCES
 
During the past decade, the practice of medicine has been dramatically affected by economic changes. The simultaneous explosion of managed care and/or managed competition and the changes instituted by the Health Care Finance Administration in introducing the resource-based relative value system for reimbursement of Part B (Supplemental Medical Insurance) Medicare have had a substantial effect on the lives of all clinical physicians. This has been especially true in the field of diagnostic radiology. Reduced reimbursements for procedures, national political attention directed at health care costs and quality, increased utilization of imaging in all spheres of medical practice, and rapid improvements in imaging technology leading to dramatic changes in the practice of patient care have all had a major effect on radiologists (5–8). These market forces, as well as perceived and feared future reductions in reimbursement, have had profound effects on hiring and salary trends within the profession (8).

By using the HWI, Forman et al (2) and Covey et al (3) have shown the substantial changes in hiring trends that have taken place in diagnostic radiology from 1991 to 1999 (Fig 1). A dramatic reduction in hiring was seen in the mid-1990s, followed by a rebound in hiring that surpassed previous peaks. HWIs, such as those published by Forman et al and Covey et al, have been shown to be accurate indicators of hiring trends in other industries (9,10), and the trends shown with these data have been confirmed by Sunshine et al (8), who analyzed survey data from the American College of Radiology (11–17).

In this study, we used a new metric, RA, to show the varying success of two specialties and the match as a whole in an effort to determine medical students’ reasons for choosing diagnostic radiology as a specialty. The RA of a specialty has several absolute meanings. A negative value implies that a specialty is less attractive than the mean, whereas a positive RA means that the specialty is more attractive. Perhaps more important is the overall trend of the RA of a specialty. A negative trend reflects worsening conditions for a specialty—although it may still have a positive value overall—and a positive trend corresponds to a specialty’s increasing attractiveness regardless of its current value. Thus, there are really two components to each year’s RA: (a) the absolute value that indicates whether a specialty is more or less attractive than all specialties combined and (b) the change from the prior year, which is indicative of improvement or decline in attractiveness.

From the perspective of diagnostic radiology, although 1999 was the 1st year in which the RA returned to positive territory, the positive trend during the past 4 years bodes well for the specialty, and the continuation of this trend into 2000 supports this finding. On the basis of the recent indications on the job market (2–4), this trend would be expected to continue. The current upswing in the job market is likely a trend back toward regular hiring practices after a period in which employers resisted hiring new radiologists until they could assess how new market forces such as managed care and health care reform would affect their work volume (18–20).

Contrary to what was predicted, imaging volume has increased rather substantially during the past decade (21). This, coupled with the decrease in the number of medical students matching in diagnostic radiology in the mid-1990s and the "catch-up" hiring that is now going on, has created an imbalance in supply and demand. Furthermore, it seems this may not be corrected for several years. This prediction is based on the fact that the current upswing in the number of students entering diagnostic radiology will not be reflected in the job market for at least 5 years (likely more as certificate of added qualification requirements and the need for fellowship training continue to increase the total training time).

The close relationship between the RA of diagnostic radiology and the HWI suggests that medical students likely incorporate such market data in their decision process for choosing a specialty. In other words, the economic strength of a specialty plays a role in a medical student’s decision to enter that specialty. Given the long training period (supply chain), however, the current economic viability of a specialty (demand) is likely to be irrelevant when current medical students enter the job market.

In 1996, there was a sharp increase in the number of diagnostic radiology positions offered. The exact reason for this increase is not clear. Several hypotheses, however, can be entertained. In the transition from diagnostic radiology being a 4-year residency to a 4 + 1 clinical year tract, programs may have had to increase the number of positions offered to keep a steady number of incoming residents. If this were the reason for the increase, however, there should have been a proportional decrease in the number of positions offered in a subsequent year. This proportional decrease did not occur. Furthermore, our data included all positions, including those for postgraduate years 1 and 2 combined, for all years analyzed.

Second, residency programs may have been expecting governmental or market-imposed decreases in the number of diagnostic radiology residency positions. Thus, they may have attempted to increase the number of residents they had by increasing their allocated number of positions, thereby protecting themselves from any future proportional mandated decreases. It seems unlikely, however, that enough programs would have done this to create the increase that was seen. This increase in positions could not have come at a worse time, as 1996 was the year that the RA for diagnostic radiology reached its low for our observation period.

There are several limitations to the RA index. First, as the overall fill rate approaches 100%, the RA value becomes less meaningful because there can be less substitution between specialties. Second, out-of-match and postmatch scramble arrangements in which medical students and residency programs fill positions without participating in the match cannot be accounted for with our methods. We assume, however, that these types of arrangements should not affect the meaning of RA values each year. In most instances, out-of-match arrangements do seem to fill all available positions. Future survey research could assess this further.

On the basis of our findings, and allowing for these limitations, we believe that medical students have an in-depth understanding of the economic forces at play in the health care job market and that they use this information when choosing a specialty. The 2-year lag between changes in the diagnostic radiology job market and salary averages and their effect on medical students’ decisions is likely a reflection of the period between when a medical student chooses a specialty and the time of the actual match.

Given the increasing overall demands for our services and the current limits placed on government-funded residency positions, it is likely that a genuine shortage of radiologists will exist in the not-too-distant future. Market forces, as previously described, should bode well for the economic viability of the specialty, and the RA, in the short run at least, should maintain its current peak. Policy makers and graduate medical education planners may wish to consider these data to allow for the appropriate level of trainees and future radiologists. Furthermore, this metric can be used to evaluate all specialties as to the desire of trainees to enter the field.


   FOOTNOTES
 
2 Current address: Department of Pediatrics, Massachusetts General Hospital, Boston. Back

Abbreviations: AMA = American Medical Association, HWI = help-wanted index, RA = relative attractiveness

Author contributions: Guarantor of integrity of entire study, H.P.F.; study concepts and design, H.P.F.; literature research, K.A.; data acquisition, D.S.K., H.P.F., K.A.; data analysis/interpretation, H.P.F., K.A.; statistical analysis, H.P.F., K.A.; manuscript preparation and definition of intellectual content, H.P.F., K.A.; manuscript editing and revision/review, H.P.F., J.H.S., K.A.; manuscript final version approval, H.P.F., K.A.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 

  1. NRMP match data 1990–2000.; Available at: www.nrmp.aamc.org/nrmp/. Accessed May 2, 2000..
  2. Forman HP, Kamin DS, Covey AM, Sunshine JH. Changes in the market for diagnostic radiologists as measured through a help wanted index. AJR Am J Roentgenol 2000; 174:933-938.[Abstract/Free Full Text]
  3. Covey A, Sunshine JH, Forman HP. The job market in diagnostic radiology 1999: updated findings from a help wanted index of job advertisements. AJR Am J Roentgenol 2000; 175:957-961.[Abstract/Free Full Text]
  4. American Medical Association. Physician socioeconomic statistics 1999–2000 Chicago, Ill: American Medical Association, .
  5. Weiner JP. Forecasting the effects of health care reform on US physician workforce requirement. JAMA 1994; 272:222-230.[Abstract]
  6. Simon CJ, Dranove D, White WD. The effect of managed care on the incomes of primary care and specialty physicians. Health Serv Res 1998; 33:549-569.[Medline]
  7. Friedenberg RM. Graduate medical education and physician supply. Radiology 1998; 206:585-587.[Free Full Text]
  8. Sunshine JH, Bushee GR, Vydareny KH, Shaffer KA. Graduates speak: the employment experience of 1995 graduates of diagnostic radiology and radiation oncology training programs. Radiology 1997; 203:695-704.[Abstract/Free Full Text]
  9. Boschan C. Job openings and help wanted advertising as measures of cyclical fluctuations in unfilled demand of labor In: The measurement and interpretation of job vacancies. National Bureau of Economic Research. New York, NY: Columbia University, 1966; 491-518.
  10. Cohen MS, Solow RM. The behavior of help wanted advertising. Rev Econ Stat 1967; 49:108-110.[CrossRef]
  11. Burkhardt JH, Sunshine JH, Shaffer KA. The employment market for 1996 diagnostic radiology and radiation oncology graduates: training program director’s point of view. AJR Am J Roentgenol 1997; 169:333-337.[Abstract/Free Full Text]
  12. Deitch CH, Sunshine JH, Chan WC. Hiring by radiology groups in 1994. Radiology 1996; 198:359-364.[Abstract/Free Full Text]
  13. Bushee GR, Sunshine JH, Chan WC, Shaffer KA. The demand side of the job market for diagnostic radiologists and radiation oncologists: hiring by physician group in 1995. AJR Am J Roentgenol 1996; 167:303-309.[Abstract/Free Full Text]
  14. Mallick R, Leader SG, Sunshine JH, Shaffer KA. Hiring by radiology groups in 1996. Radiology 1997; 205:479-486.[Abstract/Free Full Text]
  15. Crewson PE, Sunshine JH, Schepps B. The situation of diagnostic radiology training programs and their graduates in 1997. AJR Am J Roentgenol 1998; 171:919-922.[Abstract/Free Full Text]
  16. Lalman D, Porter S, Sunshine JH, Bushee GR, Schepps B. Initial employment experience of 1996 graduates of diagnostic radiology and radiation oncology training programs. AJR Am J Roentgenol 1998; 171:301-310.[Abstract/Free Full Text]
  17. Sunshine JH, Kassing P, Shaffer K, Janower M. The job market for radiology residents and fellows graduating in 1994. AJR Am J Roentgenol 1994; 163:1305-1308.[Abstract/Free Full Text]
  18. Jacoby I, Meyer GS. Creating an effective physician workforce marketplace. JAMA 1998; 280:822-824.[Free Full Text]
  19. Ruhnke GW. Residencies and employment under managed care: a medical student’s view. Health Aff (Millwood) 1996; 15:113-117.[CrossRef][Medline]
  20. Dunnick NR. Employment in radiology is exceptionally high. Acad Radiol 1998; 171:301-310.
  21. Sunshine JH, Bushee GR, Mallick R. U.S. radiologists’ workload in 1995–1996 and trends since 1991–1992. Radiology 1998; 208:19-24.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 RadiologyHome page
K. Horst and R. B. Gunderman
There Is More to Life than Lifestyle
Radiology, March 1, 2006; 238(3): 767 - 771.
[Full Text] [PDF]

Home page
 JAMAHome page
G. W. Rutecki, E. R. Dorsey, and D. Jarjoura
Lifestyle as a Factor in Medical Students' Career Choices--Reply
JAMA, December 10, 2003; 290(22): 2941 - 2942.
[Full Text] [PDF]

Home page
 JAMAHome page
E. R. Dorsey, D. Jarjoura, and G. W. Rutecki
Influence of Controllable Lifestyle on Recent Trends in Specialty Choice by US Medical Students
JAMA, September 3, 2003; 290(9): 1173 - 1178.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Roentgenol.Home page
D. D. Saketkhoo, A. M. Covey, J. Sunshine, and H. P. Forman
Updated Findings from a Help Wanted Index of Job Advertisements and an Analysis of the Policy Implications: Is the Job-Market Shortage for Diagnostic Radiologists Stabilizing?
Am. J. Roentgenol., October 1, 2002; 179(4): 851 - 858.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Anzilotti, K.
Right arrow Articles by Forman, H. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Anzilotti, K.
Right arrow Articles by Forman, H. P.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE