| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Research Department, American College of Radiology, 1891 Preston White Dr.,
Reston, VA 20191.
2 Department of Diagnostic Radiology, Yale University School of Medicine, New
Haven, CT.
3 Present address: College of Human Medicine, Institute for Health Care Studies,
Michigan State University, East Lansing, MI.
Received April 24, 2006;
accepted after revision May 8, 2006.
Address correspondence to J. H. Sunshine
(jsunshine{at}acr.org).
Abstract
 Top Abstract IntroductionLess Work for Radiologists...More Radiologists in Practice...More Work Completed per...Materials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We measured the contribution of eight possible explanations, predominantly using simple quantitative analyses. We analyzed published data, data on the volume of imaging from Medicare and from the Medical Expenditure Panel Survey, data on residents and fellows from the American College of Radiology's (ACR) membership department, data on residents from the American Board of Radiology, data from the ACR's 1995 and 2003 Surveys of Radiologists, and data from interviews about nighthawk services.
RESULTS. From these data sources, we determined the following. Total imaging and imaging by radiologists continued to grow rapidly—by > 20% from 2000 to 2003 (measured in relative value units), which was somewhat faster than in the years preceding 2000 when the shortage was building. Foreign imagers took on a negligible portion of the workload. No reductions in retirement occurred among radiologists during 2000-2003, a 10-20% decrease in the annual number of residency graduates occurred, and no increase in residents going directly into the workforce rather than taking a fellowship was noted. Radiologists' average annual work hours were relatively constant, increasing by perhaps 2%. Work done per hour—that is, productivity—increased sharply (by approximately 15%) during this period.
CONCLUSION. Increased productivity is the predominant explanation of how the radiologist shortage eased. The contribution of other factors was, in comparison, small or even in the opposite direction.
Keywords: productivity • radiologist shortage • radiologist work hours • teleradiology • workload
|
TopAbstractIntroductionLess Work for Radiologists...More Radiologists in Practice...More Work Completed per...Materials and MethodsResultsDiscussionReferences |
|---|
|
|
However, over the ensuing years, the shortage eased considerably. By 2003, the ratio of jobs to job seekers at the ACR Professional Bureau fell to approximately 1:1 (Fig. 1), and the number of advertisements fell by approximately 20% (Fig. 2).
How could the shortage have eased, given that imaging work grows much more rapidly than the number of radiologists? Simply as a matter of arithmetic, it would seem that one or more of the eight developments discussed in the following sections must have occurred [3]. For convenience, we group these eight developments into three categories: less work for radiologists than expected, more radiologists in practice than expected, and more work completed per radiologist than expected.
Less Work for Radiologists Than Expected
|
TopAbstractIntroductionLess Work for Radiologists...More Radiologists in Practice...More Work Completed per...Materials and MethodsResultsDiscussionReferences |
|---|
Number 2 is that the share of the total imaging workload increased for professionals other than radiologists, so much so that radiologists saw little increase in their workload.
Number 3 is that the work was being outsourced abroad through international nighthawk services, decreasing the amount of work for U.S. imagers.
More Radiologists in Practice Than Expected
|
TopAbstractIntroductionLess Work for Radiologists...More Radiologists in Practice...More Work Completed per...Materials and MethodsResultsDiscussionReferences |
|---|
|
For number 6, we investigated whether fewer residents took fellowships, thereby putting a substantial number of radiologists into the posttraining workforce a year earlier than expected.
More Work Completed per Radiologist Than Expected
|
TopAbstractIntroductionLess Work for Radiologists...More Radiologists in Practice...More Work Completed per...Materials and MethodsResultsDiscussionReferences |
|---|
Finally, number 8 on our list was productivity. Were radiologists getting more work done per hour? This might result from changes such as the introduction of new technologies (be they high-tech such as PACS or low-tech such as wet-reading telephone lines), increased efficiency of evening and weekend work hours through the use of nighthawk services, or changes in practice operations.
In this article, we present findings on which of these eight possible mechanisms have, in fact, had substantial effects on the radiologist job market.
|
TopAbstractIntroductionLess Work for Radiologists...More Radiologists in Practice...More Work Completed per...Materials and MethodsResultsDiscussionReferences |
|---|
For possible developments 1 and 2, we tabulated data on Medicare imaging services from Medicare's annual Physician/Supplier Procedures Summary (PSPS) Masterfile (previously known as the BMAD, or Part B Medicare Annual Data file). This file contains a record of services received by Medicare beneficiaries who have Part B fee-for-service enrollment. We tabulated the total number of imaging procedures received by these beneficiaries and the total physician work relative value units (RVUs) involved in these procedures. We divided by the number of Part B fee-for-service beneficiaries to obtain per capita data. More details of the data set and our methods were published previously [4, 5].
Data of comparable quality or detail do not exist for the people younger than 65 years. We used the Medical Expenditure Panel Survey (MEPS), a nationally representative sample of the general U.S. population, for studying development number 1 to ascertain if trends it observes in those younger than 65 years are similar to those it reports for those 65 years and older. More information about this data set and our methods were published previously [5].
For possible development number 3, we held conversations with a convenience sample of individuals involved in nighthawk services to gain an understanding of these services.
For possible development number 4, we used results from one of our recent articles, which studied trends in radiologists' retirement patterns [6].
For possible development number 5, we obtained the number of physicians completing diagnostic radiology residency training each year from the ACR's membership department. We also obtained from the American Board of Radiology (ABR) the number of first-time takers of the diagnostic radiology oral board examination each year. This is the ABR's best measure of trends in the number of physicians completing radiology residency each year (Hattery R, personal communication, March 21, 2006) because trainees are not permitted to take this examination before the last year of their residency, and only a few do not take it then.
For possible development number 6, we obtained data from the ACR's membership department on how many of each year's residency graduates proceeded to a fellowship. However, the set of fellowships included is basically limited to those accredited by the Accreditation Council for Graduate Medical Education (ACGME), and a substantial proportion of fellowships do not fall into this category.
For possible development number 7, we compared data on radiologists' work hours from ACR's 2003 Survey of Radiologists with corresponding data from the ACR's 1995 Survey of Diagnostic Radiologists and Radiation Oncologists. The 1995 survey is the closest predecessor to the 2003 survey that obtained information on work hours, and the two surveys are very similar. Details of both and of their methods were published previously [4, 6-9].
For possible development number 8, we computed trends in work performed per hour by analyzing published data on radiologists' workload from Bhargavan and Sunshine [9] together with data on radiologists' work hours from the 1995 and 2003 ACR surveys.
|
TopAbstractIntroductionLess Work for Radiologists...More Radiologists in Practice...More Work Completed per...Materials and MethodsResultsDiscussionReferences |
|---|
|
Table 2 shows that the rate of growth of per capita imaging in the population of those younger than 65 years was similar to that for those 65 years and older. Assuming the same growth rate, both in RVUs and in procedures, in the population younger than 65 years as Medicare data show for the population 65 years and older, and recognizing that the U.S. population grows by approximately 1% annually, we estimated that between 2000 and 2003 total imaging in the United States increased by approximately 23% (3% population growth added to 20% RVU growth, Table 1) in physician work RVUs and 18% (3% + 15%) in procedures.
|
Number 2: Growth of Radiologists' Workload
Table 1 shows that imaging
performed by radiologists, measured in physician work RVUs, increased by 19%
per Medicare beneficiary between 2000 and 2003. Measured in procedures, the
increase was 11% per beneficiary. Again, both increases were larger than in
1997-2000.
Relevant information measuring total workload per radiologist (not the total for all radiologists in the United States) for patients of all ages has been published [9] for a similar time span. This previous study showed a 9% increase in procedures per radiologist and a 15% increase in RVUs per radiologist over the period 1998-1999 to 2002-2003.
Number 3: International Teleradiology
Firms in other countries provide after-hours interpretations. However, most
of the radiologists involved are Americans that these firms station abroad to
interpret images when it is daytime for them, although it is after hours in
the United States. Thus, foreign radiological interpretation activity does not
decrease the workload that U.S. radiologists need to handle, it just relocates
some of the U.S. radiologists to places outside the United States.
Exceptions are few and very small. Our conversations indicate these consist mostly of foreign-origin, U.S.-trained radiologists who have returned to their home countries and perform some long-distance interpretations. Our rough estimate is that the size of the workforce involved is perhaps 0.1% or less of the size of the U.S. radiologist workforce (which is some 25,000-30,000 radiologists) [8, 10].
Number 4: Radiologists' Retirement
A detailed analysis [6]
showed radiologists' retirement patterns in 2003 were the same as in 2000.
There was no net movement toward delayed retirement or increased return to
work after temporary retirement.
Number 5: Number of Residency Graduates
Because most graduates of diagnostic radiology residency go on to a
fellowship, most enter the work force a year after they complete their
residency. Viewed in light of this reality, the ACR membership data on the
annual number of residency graduates (Fig.
3) show that more radiologists entered the workforce when the
severe shortage was developing—that is, in the late 1990s and
2000—than in the subsequent years when the shortage was easing (most
prominently, in 2002 and 2003). Specifically, those entering the workforce
from 1998-2000, the years when the shortage was increasing, were the residents
who graduated in 1997-1999 and numbered approximately 1,000-1,100 annually. In
contrast, those entering the workforce in 2001-2003, when the shortage was
easing, were the residents who graduated in 2000-2002 and numbered
approximately 800-900 annually. The ABR data (Gerdeman A, personal
communication, March 28, 2006) show a similar pattern, with approximately 900
to 1,000 first-time takers of the diagnostic oral examination in the years
when the shortage was developing and approximately 800 to 900 in the years
when it was easing (Fig. 3).
(Recall that this statistic is the ABR's best measure of trends in the number
of residency graduates.)
Number 6: Residents Taking Fellowships
A summer 2004 survey of diagnostic radiology residents graduating in 2005
found 87% intended to continue on to a fellowship
[11]. This is a higher
percentage than was found in the late 1990s; studies of 1997 and 1999
graduates found that 80-83% continued on to fellowships
[12]. The ACR data on
fellowships filled each year (Fig.
3) show a relatively constant number in the late 1990s and in the
current decade.
Number 7: Radiologists' Work Hours per Year
ACR surveys show that between 1995 and 2003, average weekly hours of
full-time diagnostic radiologists increased by 5%, but the number of weeks
they worked per year decreased by 3% (Table
3). Thus their annual work hours increased by approximately 2%.
Over the same 8-year period, the average radiologists' workload increased by
18% in procedures and 36% in RVUs
[9].
|
Number 8: Work Done per Hour
The statistics in the previous paragraph imply that over the 8-year period
procedures per work hour increased by 16%—the net effect of an 18%
increase in procedures in a work year that has 2% more hours. Similarly, the
implication is that RVUs accomplished per work hour increased by 34%, with 36%
more RVUs in 2% more hours.
|
TopAbstractIntroductionLess Work for Radiologists...More Radiologists in Practice...More Work Completed per...Materials and MethodsResultsDiscussionReferences |
|---|
In brief, the growth of imaging did not lessen; therefore abatement in imaging growth is not an explanation of why the radiologist shortage eased.
Number 2: Did Radiologists' Workload Stagnate?
Although imaging by professionals other than radiologists has been growing
faster than imaging by radiologists (Table
1), radiologists' workload grew rapidly in the early years of the
current decade. To be specific, the Medicare data show a 19% increase in RVUs
per beneficiary performed by radiologists between 2000 and 2003. If the same
19% growth occurred for the population younger than 65 years, then, given the
1% annual growth of the U.S. population, the increase in total workload would
have been approximately 22%. Alternatively, one can form an estimate of growth
of radiologists' total workload from the Bhargavan et al.
[2] study, which has workload
data from the entire population but not for a fully identical period. This
report showed a 15% increase in RVUs per radiologist. Given the approximately
1.5% annual growth of the radiologist workforce, which would be approximately
a 6% increase in the number of radiologists over the 4-year period involved,
the total growth in workload (including both growth in workload per
radiologist and growth in number of radiologists) would again be somewhat more
than 20%.
Clearly, radiologists' workload did not stagnate. Rather, it grew rapidly, although somewhat less rapidly than total imaging. Therefore, stagnation in radiologists' workload cannot be an explanation of why the shortage eased.
Number 3: Has Work Moved to Foreigners?
Although international teleradiology exists, it is largely being performed
by American radiologists stationed offshore. Thus, its effect in reducing the
amount of work American radiologists have to perform is negligible, and this
effect cannot explain the easing of the radiologist shortage.
Another aspect of international teleradiology—and of domestic nighthawk arrangements as well—may be more relevant. One nighthawk radiologist who keeps reasonably busy during his or her work shift replaces multiple on-call radiologists, each of whom is not fully occupied. This arrangement increases the efficiency of radiologists' work. To put the situation in somewhat different words, one work "day" (actually, the "day" is a night shift) replaces multiple work "days" (which are on-call shifts that are not fully busy). Thus, fewer workdays (more accurately, "work shifts") are needed to perform the same amount of imaging.
Number 4: Did Radiologists Delay Retirement?
No change was found in retirement patterns, so delay of retirement or
return to work after temporary retirement offer no explanation of the easing
of the radiologist shortage.
Number 5: Are There More Residency Graduates?
The number of radiologists graduating from residency and moving into the
workforce was, if anything, larger when the shortage was becoming severe than
in the years when the shortage eased. Thus, an increase in the number of
residency graduates is not an explanation of why the shortage eased.
Number 6: Are Fewer Residents Taking Fellowships?
The data on fellowships are fragmentary, but do not show an increase in
residents skipping a fellowship and thereby augmenting the workforce. To be
more specific, although it would be desirable to have survey data for more
years in the current decade, not solely for 2005, the 2005 data show a record
high percentage of residents going on to a fellowship. The other set of data,
those on ACGME-accredited fellowships, omits the large number of fellowships
that are not ACGME-accredited but, again, these data show no falloff in
residents going into fellowships. Thus a falloff in residents taking a
fellowship does not seem to be an explanation of the easing of the radiologist
shortage.
Number 7: Have Radiologists' Work Hours Increased?
Our data show a small increase of approximately 2% in radiologists' average
annual work hours between 1995 and 2003. (This is the most relevant period for
which we can obtain data.) Over the same period, the work done each year per
radiologist increased a great deal more—by 36% in RVUs and by 18% in
procedures. Thus, longer hours played only a small role in each radiologist's
ability to do more work per year—a role that accounts for only
approximately 6% of the increase in annual RVUs per radiologist and 11% of the
increase in annual procedures per radiologist.
Number 8: Are Radiologists Doing More Work per Hour?
The implication of the preceding paragraph is that the dominant explanation
of how radiologists are handling the increased workload they face is that they
are accomplishing more in each hour they work. Increased productivity is the
source of approximately 90% or more of the increase in the average
radiologist's annual work output.
Conclusion
The shortage of radiologists eased considerably during 2000-2003 (Figs.
1 and
2). However, in this period,
the growth of imaging in the United States and the growth of the imaging
performed by U.S. radiologists continued. As described previously, we estimate
the total work performed by all radiologists in the United States increased by
slightly more than 20% in this period.
An unexpected growth in the number of radiologists in practice is also not an explanation of the easing of the shortage. We did not find a decrease in retirement among radiologists, an increase in the number of residents graduating, or an increased tendency for residents to skip fellowship and thereby get into the workforce earlier. However, the long standing and expected growth in the number of radiologists, which is approximately 1.5% a year, or approximately 5% over the 3 years of interest, was taking place. This explains some of the radiology profession's ability to undertake more work. But this growth in the number of radiologists was only approximately 25% as large as the total increase in work done by radiologists.
For the most part, what has happened is that each radiologist is, on average, doing more work. Data from ACR surveys show an increase of approximately 15% in annual RVUs per radiologist over a time span roughly matching that of interest in our study (2000-2003). Given that the different estimates come from different data sources, this survey-based estimate is in relatively good agreement with independently derived estimates of an increase in total workload of all radiologists in the United States of slightly more than 20% and an increase in the number of U.S. radiologists of approximately 5%.
The number of hours radiologists work each year has changed little, which means they are getting more work done predominantly by doing more work per hour rather than by working more hours. In other words, increased productivity is clearly the dominant reason the radiologist shortage eased.
However, knowing that increased productivity is the reason the shortage eased still leaves two important questions unanswered. First, how is it that productivity increased rapidly enough in 2000-2003 to not only keep up with the growing workload but also to ease the shortage, whereas in the years preceding 2000 productivity did not increase rapidly enough to keep up with workload growth (although that growth was slower), resulting in a severe shortage? Second, what specific features of radiology practice led to higher productivity? Some possibilities are high-tech, such as PACS, or low-tech, such as having support staff hang and take down films. We hope to answer these questions through future research.
Our data have several limitations. The most notable include the fact that MEPS does not have data on inpatient imaging; its sampling variability is substantial relative to the changes in utilization observed over periods of a few years; and its sonography data for 2000 are anomalous. Also, we have no measure of radiologists' work hours specifically for 2000. Last, the data on fellowships are less extensive in either date or coverage than would be desired.
However, despite these limitations, these data are sufficiently unambiguous to show that increased productivity is, undoubtedly, the dominant reason the radiologist shortage eased.
Acknowledgments
We thank Robert Hattery and Anthony Gerdeman for explaining which ABR data
are most relevant to our questions and then for furnishing these data. We
thank the ACR membership department for undertaking special tabulations of
data on residents and fellows and Mythreyi Bhargavan for assistance, most
prominently with the MEPS data.
|
TopAbstractIntroductionLess Work for Radiologists...More Radiologists in Practice...More Work Completed per...Materials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  D. L. Weiss and C. P. Langlotz Structured Reporting: Patient Care Enhancement or Productivity Nightmare? Radiology, December 1, 2008; 249(3): 739 - 747. [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. Pickhardt, C. Hassan, A. Laghi, D. H. Kim, A. Zullo, F. Iafrate, and S. Morini Is There Sufficient MDCT Capacity to Provide Colorectal Cancer Screening with CT Colonography for the U.S. Population? Am. J. Roentgenol., April 1, 2008; 190(4): 1044 - 1049. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Wright, J. B Killion, J. Johnston, V. Sanders, T. Wallace, R. Jackson, D. Snodgrass, and R. A Redd RAs Increase Productivity Radiol. Technol., March 1, 2008; 79(4): 365 - 370. [Full Text] [PDF]
|
|
|
|
 |
|
 R. Steinbrook The Age of Teleradiology N. Engl. J. Med., July 5, 2007; 357(1): 5 - 7. [Full Text] [PDF]
|
|
|
|
|
|
H. P. Forman The radiology job market: help wanted. Am. J. Roentgenol., November 1, 2006; 187(5): 1147 - 1148. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), number of
residents who continued with Accreditation Council for Graduate Medical
Education (ACGME)-accredited fellowships (
), and number of first-time
takers of diagnostic radiology oral board examination (
).