| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Hospital of the University of Pennsylvania, 3400
Spruce St., Philadelphia, PA 19104.
2 Department of Diagnostic Radiology, Yale University School of Medicine, New
Haven, CT 06510.
3 Department of Radiology, Wake Forest University School of Medicine,
Winston-Salem, NC.
4 Research Department, The American College of Radiology, Reston, VA
20191.
5 Department of Economics, Yale College, and Yale University School of
Management, Yale University, New Haven, CT 06510.
Received May 13, 2005;
accepted after revision May 25, 2005.
Address correspondence to D. D. Saket.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Three sources of data are presented and compared with previous data: vacancies in academic radiology departments as of July; the ratio of job listings to job seekers at a major placement service; and the number of positions advertised in Radiology and the American Journal of Roentgenology.
RESULTS. Vacancies in academic radiology departments averaged 3.9 in 2004, down 29%, and decreased for all subspecialties as compared with 2001, but the number of vacancies remained very similar to that for 2003. Job listings per job seeker were 1.1 in 2004, stable over the past 2 years but at the lowest level since 1997. The overall number of positions advertised declined by 14% in 2003 compared with 2002 and by an additional 17% in 2004, reaching the lowest level since 1998. In 2004, 45.3% of positions advertised were academic. Comparing 2003-2004 with 2001-2002, all geographic regions exhibited absolute declines in advertisements except the Northeast, which showed a 1.5% increase. Absolute increases occurred for musculoskeletal and emergency radiology positions. Statistically significant proportional decreases occurred for general radiology, vascular/interventional radiology, and pediatric radiology.
CONCLUSION. Three separate data sources confirm a substantial and broad-based multiyear decline in the strength of the demand for diagnostic radiologists, with some shifting in relative demand for subspecialties. It is not clear if the decrease continued in 2004 or if 2004 demand was similar to that of 2003. Data are relative and do not indicate the employment market is weak in absolute terms.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recent studies have shown that relative radiologist demand has peaked and may, in fact, now be declining [3-5]. Definitive characterization of the direction of movement of the radiologist job market, especially over the past 2 years, however, is lacking. In this article, we present the most recent findings regarding the status of the radiologist job market, focusing primarily on the updated 2003-2004 data from our 14-year help wanted index of job advertisements [3, 6, 7]. These data are supplemented by two other sources: a 2004 survey of vacancies in academic radiology departments and an analysis of the American College of Radiology's (ACR) Professional Bureau records, which are compared with previously published data [4]. Our discussion focuses on the possible reasons for the trends observed and their policy implications.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
In this article, the newly obtained data from 2003-2004 are compared directly with those from 2001-2002. Findings from the years 1991-2000 are also included in certain figures when they help to illustrate long-term trends. When comparing data across time periods, we performed statistical analysis using a two-tailed comparison of proportions. Values for p of less than 0.05 were considered to be statistically significant.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ACR Professional Bureau Data
As Figure 3 shows, the
number of job listings per job seeker has remained constant over the past 3
years (2002-2004) at 1.1, but it is at its lowest level since 1997. Although
this ratio is far below the peak value of 3.8 in the year 2000, it is still
higher than it was from 1991 to 1997 when it fell to as low as 0.2 in
1994.
Help Wanted Index for Job Advertisements
Totals—A total of 8,098 positions in diagnostic radiology
were advertised during the 24-month period from January 2003 to December 2004,
specifically 4,430 in 2003 and 3,668 in 2004. This total represents a 25.7%
decrease relative to the total of 10,904 positions advertised from 2001
through 2002 or a decrease of 1,403 advertisements per year. The 3,668
advertisement total for 2004 represents the lowest number since 1998.
The average number of positions advertised decreased to 369 per month during 2003 (14% decrease vs 2002) and then decreased further to 306 per month during 2004 (additional 17% decrease vs 2003). For 2003-2004, a monthly peak of 467 advertisements occurred during January 2003, whereas the monthly nadir of 276 occurred in March 2004, which was the lowest monthly rate since August of 1999 (Fig. 4). The 12-month rolling average, which is calculated to minimize seasonal variation, showed a consistent decline from 488 in February 2002 to 305 in November 2004, with a negligible increase to 306 in December 2004. By November 2004, near the end of our data time frame, the rolling average had fallen to levels not seen since October 1999 (Fig. 4).
|
|
Type of practice—From January 2003 through December 2004, 57.3% of the total jobs advertised were for private practice positions and 42.7% were for academic positions. This proportion of academic positions compared with 39.8% during the prior 2-year period from 2001 to 2002 is a statistically significant increase (p < 0.01). In fact, the proportion of academic positions advertised climbed to 45.3% in 2004, the highest percentage since 1993 (Fig. 6). Although the proportion of academic versus private practice advertisements increased, the absolute numbers of advertisements for both academic and private practice positions declined in 2003-2004 versus 2001-2002, reflecting the decrease in the number of overall advertisements. The number of private practice position advertisements declined from 6,562 to 4,639 (29.3% decrease), while the number of academic position advertisements decreased less, from 4,342 to 3,459 (20.3% decrease).
|
In addition to the national data, the regional ratio of academic-to-total advertisements, a measure of relative regional demand for academic versus private practice radiologists, was calculated for all geographic regions for the time periods 2003-2004 versus 2001-2002 (Fig. 7). The 2003-2004 academic regional ratios were greatest in the Northeast (47.2%) and Southeast (46.1%), which both showed statistically significant increases (p < 0.01) versus the 2001-2002 time period, when they were 38.4% and 39.8%, respectively. The academic regional ratio was 41.9% in the Midwest, down from 45.5% (p < 0.01). This ratio exhibited no statistically significant change in the Southwest (37.8% vs 33.5%), Northwest (31.5% vs 32.2%), and California, where it was the lowest (27.5% vs 29.3%).
|
Geographic location—Several geographic trends are noted when the data from 2003-2004 are compared with those from 2001-2002 (Table 2 and Fig. 8). The Northeast now accounts for the greatest percentage of positions advertised (31.1%), having surpassed the Midwest (26.0%) and Southeast (21.8%) during the most recent 2-year period. Both the Northeast (31.1% vs 22.8%) and Southwest (9.3% vs 7.6%) exhibited statistically significant increases in their relative proportions of advertisements as compared with 2001-2002. The percentage of advertisements from California increased slightly (6.1% vs 5.7%), but this finding was not statistically significant. On the other hand, the Midwest (26.0% vs 31.1%), Southeast (21.8% vs 25.6%), and Northwest (3.9% vs 7.1%) all showed statistically significant proportional declines. The absolute numbers of positions, however, decreased in all geographic regions except the Northeast, which exhibited a 1.5% absolute increase.
|
|
Similarly, examination of the academic data in isolation reveals a statistically significant redistribution of advertisements to the Northeast (34.5% vs 22.0%, p < 0.01) from the Midwest (25.5% vs 35.5%, p < 0.01). There was no statistically significant change in the composition of overall academic advertisements with respect to all other regions: the Southeast (23.6% vs 25.6%), Southwest (7.3% vs 7.2%), Northwest (3.0% vs 5.6%), or California (4.2% vs 3.9%).
Subspecialty trends—Subspecialty data for the 2003-2004 versus 2001-2002 time periods are summarized in Table 3. Although general radiology still comprises the largest proportion (24.1%) of advertisements (Fig. 9), it continues to display the largest percentage decline and absolute decline for all categories, as it stood at 31.1% of advertisements in 2001-2002 versus 43.4% in 1991-1994 [7]. Statistically significant proportion declines between 2003-2004 and 2001-2002 were also observed for vascular/interventional radiology (11.4% vs 14.0%), pediatric radiology (3.6% vs 4.5%), and "other" positions (1.0% vs 1.5%). Absolute declines were noted for all subspecialties except emergency radiology, which almost doubled (201 advertisements vs 111), and musculoskeletal radiology (585 vs 514). These findings correspond to proportional increases in emergency radiology from 1.0% to 2.5% of overall advertisements and, in musculoskeletal radiology, from 4.7% to 7.2%. Absolute declines but statistically significant proportional increases, meanwhile, were observed for abdominal/cross-sectional radiology (17.2% vs 16.0%), mammography (12.9% vs 9.7%), neuroradiology (10.8% vs 9.7%), and nuclear medicine (4.7% vs 3.6%).
|
|
Analysis of the 2003-2004 subspecialty data shows statistically significant differences in the subspecialty compositions of academic versus private practice advertisements. Academic practices exhibited greater relative demand for neuroradiologists (13.3% vs 9.0% of total advertisements, p < 0.01), vascular/interventional radiologists (13.0% vs 10.2%, p < 0.01), chest radiologists (6.7% vs 1.2%, p < 0.01), and emergency radiologists (3.6% vs 1.7%, p < 0.01). Private practices, on the other hand, maintained higher relative demand for general radiologists (30.3% vs 15.7%, p < 0.01) and part-time radiologists (1.2% vs 0.3%, p < 0.01). The proportion of advertisements for the other radiology subspecialties did not show statistically significant differences when academic were compared with private practices.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
What is not clear is whether the easing of demand pressure has stabilized or whether demand pressure is continuing to ease. Although all three employment market indicators are significantly down when 2004 is compared with 2001, little has changed between 2004 and 2003 with regard to the academic vacancy and Professional Bureau statistics. Pointing in the other direction are job advertisements in 2004, which were clearly fewer than in 2003, although, as noted, there is a distinct possibility the decrease in advertisements had stopped by the end of 2004. Complicating the analysis is the possibility that increased use of electronic media and "head-hunting" agencies is leading to a decline in job advertisements in print journals without there being an actual decrease in demand.
Second, although the absolute demand for academic radiologists has decreased, academic radiology positions comprise an ever-increasing proportion of overall advertisements, steadily increasing from 34.0% in 1999 to 45.3% in 2004. This finding probably is due primarily to significant decreases in demand for private practice positions, which historically have been more sensitive to overall market changes. However, in interpreting the changing ratio of academic-to-private practice advertisements, it is important to recall that, in response to equal employment opportunity (EEO) regulations, academic departments are likely to advertise positions whether journal advertisements are a relatively productive channel for finding employees. This is less true for private practices. Thus, the observed smaller sensitivity of academic job advertising to changing market conditions may reflect the idiosyncrasies of advertising academic positions rather than employment market reality.
Of note, academic radiologists are in greatest demand in the Northeast, likely reflecting the relative greater number of training programs, probable greater difficulties in attracting and retaining faculty, and the greater utilization of imaging in that region [1]. This academic shortage appears to be the primary reason why the overall absolute demand for radiologists in the Northeast, as measured by the number of advertisements, continues to increase despite declines in the rest of the country.
Third, general radiology positions continue to exhibit steep absolute and proportional declines, now comprising 24.1% of overall advertisements as compared with 43.4% during 1991-1994. The greatest increase in demand in 2003-2004 as compared with 2001-2002 was for musculoskeletal radiologists (proportional demand up 53%), mammographers (up 33.0%), and abdominal/cross-sectional imagers (up 8.9%), which include body MR-trained individuals. Interestingly, the help wanted index showed significantly decreased demand for pediatric radiologists, which is in conflict with the academic survey data and anecdotal reports, and should be interpreted with caution due to the small size of the pediatric radiology market data in this study. The subspecialty findings reported in this article are generally in-line with the survey data obtained by Meghea and Sunshine [5], which showed that, in 2003, musculoskeletal radiologists, mammographers, and pediatric radiologists were the only subspecialists who, on average, desired a decrease in working hours, although they did not necessarily represent the radiologists currently working the greatest number of hours [5]. Of note, our article also showed a doubling of the absolute demand for emergency radiologists, which may reflect the trend toward subspecialization of emergency radiology services, particularly in academic settings [10-13].
Study Limitations
Each of the three information sources has important limitations, which have
been discussed in detail in our previous articles
[3,
4,
6,
7]. The general validity of the
help wanted index and its application to the radiology job market have been
described in previous articles
[3,
6,
14-18].
The most important factor to keep in mind is that a help wanted index is not
an exact, but instead an indirect, measure of demand pressures in labor
markets. As such, only general trends can be described, and absolute numbers
mean little.
In addition, help wanted data are best used for short-term analysis, and caution should be taken when using help wanted index data to predict long-term trends. As noted, apart from any actual change in labor markets, in-print job advertising may have an inbuilt downward trend in this electronic age. Also, the factors driving the employment market are complex and variable.
However, all three data sources in this study point to the same conclusion—namely, a significant decrease in demand for diagnostic radiologists over the past few years. We therefore judge this basic finding to be sound.
Policy Implications
Easing of the demand side of the radiologist job market shows that drastic
measures are not needed to increase the supply of diagnostic radiologists. The
market is cyclical and seems to have corrected largely on its own. The major
causes of the changes in the labor market are probably several. In our
previous article [4], we
postulated eight factors that might be responsible including productivity
gains, turf gains by other specialties
[19], and outsourcing of
on-call responsibilities.
The easing of the radiologist shortage has occurred despite sharp rises in diagnostic imaging utilization. Bhargavan and Sunshine [1] recently reported that total imaging volumes have been increasing at an 8% annual rate and up to 15% per annum for cross-sectional techniques such as MRI and CT [1]. The number of practicing radiologists, on the other hand, has increased only 1.5% per year [1], and there is no evidence that radiologists are working longer hours (currently averaging 49 hr per week, matching the desired workload) [5]. Thus, the decreased demand for radiologists likely reflects productivity gains and growth of imaging by other specialties. In this context, continued penetration of PACS and teleradiology, provided by U.S.-trained radiologists, are possibly contributing to radiologists' ability to handle their increasing workload. These changes not only may allow individual radiologists to be more efficient, but also may potentially permit work to be distributed more effectively among radiologists, especially with regard to geographic location, subspecialty expertise, and night-time work. However, systematic quantitative information on their precise effects is lacking.
If PACS and teleradiology are important factors in increased productivity, it is important to note that they are likely to provide only a one-time boost in productivity as they become widespread. Once they are fully disseminated, and thus no longer providing further gains in productivity each year, the serious radiologist shortage seen at the beginning of this decade may reappear.
Of great concern is the effect of nonradiologist interpretation of imaging studies. Medicare data from 1993-1999 showed that relative value units (RVUs) increased by 6.9% over this period among radiologists and 32.4% among nonradiologists [20]. Because income sources for other specialists have continued to be squeezed, there is no reason to expect that these trends have not continued or worsened over the past 5 years. In fact, Levin et al. [21] have recently reported that, as of 2002, radiologists received only 48.3% of Medicare payments for imaging, as compared with 52.1% in 1997 [21]. It appears that nonradiologists, led by cardiologists [21], have filled some of the void created by the recent radiologist shortage.
As the market evolves, one would expect changes in relative demand and supply of radiologists, geographically and according to subspecialties. As demand in some areas continues to increase relative to the whole, it is prudent to use this information in guiding our trainees toward increasing future demand. Further, it is prudent to reflect on the potential causes for the supply-demand imbalance and to encourage payers to appropriately alter relative compensation of different types of procedures to reattain market equilibrium. This is most notable, in the current survey, for the specialty of breast imaging/mammography.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  T. L. Ebbert, C. Meghea, S. Iturbe, H. P. Forman, M. Bhargavan, and J. H. Sunshine The State of Teleradiology in 2003 and Changes Since 1999 Am. J. Roentgenol., February 1, 2007; 188(2): W103 - W112. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Meghea and J. H. Sunshine Retirement patterns and plans of radiologists. Am. J. Roentgenol., December 1, 2006; 187(6): 1405 - 1411. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Licurse, D. D. Saket, J. H. Sunshine, C. D. Maynard, and H. P. Forman Update on the diagnostic radiology employment market: findings through 2005. Am. J. Roentgenol., September 1, 2006; 187(3): W249 - W254. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
