| 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, New Haven, CT.
Received October 21, 2005;
accepted after revision November 3, 2005.
Address correspondence to R. S. Lewis
(rlewis{at}acr.org).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We analyzed data from the American College of Radiology's 2003 Survey of Radiologists, a large, stratified random sample survey that achieved a 63% response. Responses were weighted to make them representative of all radiologists in the United States.
RESULTS. Approximately 10% of all radiologists, or 2,700-2,800
radiologists, are breast imaging specialists, but 61% of radiologists
interpret mammograms, and only approximately 30% of mammograms are interpreted
by breast imaging specialists. Of radiologists who reported that breast
imaging was their primary specialty, only 21% took a fellowship in the field
(much lower than for other subspecialties), 59% spent 
50% of their
clinical work time in the specialty, 82% interpret 2,000 mammograms
annually, and only 11% (also well below other subspecialties) report that the
main subspecialty society (the Society of Breast Imaging) is one of the two
most important professional organizations for them. On average, breast imaging
specialists, like other radiologists, report that their workload is about as
heavy as desired. Their level of enjoyment of radiology does not differ
significantly from average.
CONCLUSION. Breast imaging appears not to be as strongly organized to raise awareness of and support for its problems as are other subspecialties. Although others find evidence of likely future problems, breast imaging specialists are not currently overworked or less satisfied in their profession than other radiologists, despite relatively low revenue generation and a particularly high risk of a malpractice lawsuit.
Keywords: breast imaging • mammography • mammography interpretation • practice of radiology
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The survey had more detailed questions about the performance of breast imaging and mammography than about any other field of radiology. This article, an in-depth portrait of breast imaging specialists and of the interpretation of mammography, is one of the first full-scale papers to be produced from the 2003 survey.
Breast cancer is by far the leading cause of non-skin cancer in U.S. women and is second only to lung malignancies as a cause of cancer death. After 30 years of increasing breast cancer death rates, these rates are falling, probably due to widespread screening with quality mammography and other medical advances. Thus, the availability of adequate numbers of qualified breast imagers to interpret mammograms and perform other breast imaging is a concern not only of radiologists but throughout the health care community. For example, the Institute of Medicine (IOM) recently published a report, Improving Breast Imaging Quality Standards [13], on this topic.
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The questionnaire for the 2003 survey consisted of 36 items; many items contained multiple subitems. Questionnaire items and topics were elicited from two rounds of canvassing of ACR physician and staff leaders, winnowed according to priorities indicated by top leadership, and pretested in two large surveys conducted in autumn 2002, with refinements made after each pretest.
The survey sample, a random sample composed of four strata, was taken primarily from the American Medical Association's (AMA) Physician Masterfile, a reasonably complete listing of all allopathic physicians in the United States, whether or not AMA members, but it also included a sample of osteopathic radiologists obtained from the American Osteopathic College of Radiology. The sample included residents, fellows, and retirees, not merely posttraining, professionally active physicians.
The survey was administered by our contractor, the Survey Research Center of the University of Virginia, by mail between March and August 2003, with nonrespondents being sent up to four remailings, and other steps taken to boost the response rate. The response rate achieved was 63%, with 1,924 usable responses. Responses were weighted so that the weighted statistics would be representative of the answers that would have been received if all physicians in the United States in the four strata had been surveyed and had responded.
Our leading tool to minimize data deficiencies was the designation of the 12 items on the questionnaire that were judged most crucial as "core questions." When questionnaires were returned, our contractor checked that these 12 items were indeed answered and made three designated consistency checks involving them. If there were any problems with the core items, the Center telephoned the respondent to try to obtain the missing response(s) or to resolve the consistency problems. Data used in this report have also been cleaned and edited to further minimize deficiencies. For example, when survey respondents indicated that they performed a specific type of breast imaging, but failed to respond to a question about performing breast imaging in general, we amended the data set to indicate that they did perform breast imaging in general.
Definition of Variables
The definition of most variables has been detailed in a previous paper
[15] and also is apparent from
the Results section and the tables. We detail here only the definitions that
we used of radiologists who perform breast imaging.
To provide the fullest profile of radiologists who perform breast imaging, this article contains information on many different definitions of who is a breast imaging specialist. The definitions are as follows:
In addition, to more broadly portray breast imaging specialists and especially radiologists who perform mammography, we present information on the following categories of radiologists:
Analysis Methods
All information presented in the remainder of this article, unless
otherwise noted, is based on weighted data and refers to physicians from all
sampling strata combined. Where numbers of physicians are given (Figs.
1 and
2), an adjustment was made for
item nonresponse—that is, for the failure of a limited percentage of
respondents to answer each question—so that the numbers are, as always,
representative of what answers would have been if all physicians of interest
in the United States had responded.
Reported standard errors and tests of statistical significance are
calculated taking into account not only the weighted nature of the data but
also the complex survey design—that is, that responses came from
distinct strata. The SEs are calculated with the Statistical Analysis System
(SAS) software procedure "surveymeans." All data analysis was
conducted with SAS software, release 9.0 (SAS Institute). Because most tables
(all but Table 1) have
comparisons between breast imaging specialists and five comparison groups of
radiologists, we used a two-tailed z-test with p < 0.01 as the
criterion of statistical significance, in keeping with the Bonferroni
inequality [17]. In two
instances in which visual inspection of the data showed relatively large
differences between breast imaging specialists and other radiologists but the
differences were not significant at p < 0.01, we applied the more
common criterion of p < 0.05; these instances are explicitly noted
in the text. SEs for percentages in the tables can be approximated by the
formula:
 |
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
30% of their clinical work time doing breast
imaging. The smallest of the highly useful estimates of the number of breast
imaging specialists results from including only those who spend 3
70% of their clinical work time performing breast imaging. Some 1,300
radiologists, or 5% of the profession, meet this definition. More than two
thirds of radiologists do at least some breast imaging
(Fig. 2). Sixty-one percent
perform mammography. However, only 28% of radiologists perform 2,000 or more
mammographic examinations per year, and only 7.2% perform 5,000 or more
mammographic examinations per year. Table 1 contains data on the relationship between pairs of characteristics of radiologists who perform breast imaging. Each row of the table refers to those radiologists with the characteristics listed at the left of the row, and each cell tells what percentage of these radiologists also have the characteristic listed at the top of the column in which the cell appears. For example, the table shows, in the first row, that of those who took a breast imaging fellowship, three fourths report breast imaging as their primary specialty, four fifths spend 3 30% of their clinical work time in the specialty, and more than two thirds perform at least 2,000 mammograms per year. Other highlights of the table include the following: Of those who report that breast imaging is their primary specialty, one fifth did a fellowship in the field, approximately three fourths spend 3 30% of their clinical work time in the specialty, and more than four fifths perform at least 2,000 mammographic examinations annually. Of those who spend 3 30% of their clinical work time in the specialty, one fourth did a fellowship in the field, half spend 3 70% of their clinical work time in the specialty, and half perform at least 5,000 mammograms per year. Of radiologists who perform mammography, only one sixth report breast imaging is their primary field and a similar fraction spend 3 30% of their clinical work time performing breast imaging. Even among those who perform 3 2,000 mammograms a year, the corresponding fractions are only approximately one third.
Tables 2 and S1 present the demographic characteristics of radiologists who perform breast imaging and the corresponding demographics of five comparison groups of radiologists. (All supplementary tables may be viewed online at www.arjonline.org.) Breast imaging specialists are, on average, about 50 years old, which is about the same as other subspecialists but younger than nonsubspecialists. However, radiologists with breast imaging fellowships are, on average, younger (average age, 44 years).
By most definitions of who is a breast imaging specialist, the percentage who are women— at least 50%—is far higher than the percentage of all radiologists who are women (18%). However, the percentage of women among radiologists who perform mammograms is not particularly high except among those who perform at least 5,000 mammograms annually.
Regionally, the largest percentages of breast imaging specialists practice in the Northeast, whereas for radiologists overall, the South has the largest numbers (Table S1).
Types of Procedures Performed by Breast Imaging Specialists
Concentrating on the two most useful definitions of breast imaging
specialists (those who report this as their primary specialty and those who
spend 3 30% of clinical work time in the field), we find
approximately 75-90% perform sonographically guided breast biopsy,
stereotactic breast biopsy, and localization for surgical biopsy (Tables
3 and S2). In contrast, 18-35%
of subspecialists who are not breast imaging specialists and 28-58% of
radiologists who are not subspecialists do so. About half of breast imaging
specialists perform fine-needle aspiration, compared with 16% of
subspecialists who are not breast imaging specialists and 21% of
nonsubspecialists. Some 35-40% of breast imaging specialists were using
computer-aided detection, compared with less than one sixth of other
radiologists. Statistics for breast MRI were similar. Breast imaging
specialists perform on average approximately five of the total of seven
categories of breast imaging procedures on which we obtained data, compared
with an average of 1.7 for other subspecialists and 2.6 for
nonsubspecialists.
|
Table S3 divides the full range of procedures performed by radiologists into seven major categories such as nuclear medicine, MRI, CT, and breast imaging. Breast imaging specialists, on average, perform approximately four of the seven major categories of procedures, whereas, on average, other radiologists perform approximately five (Tables 3 and S3). Excluding breast imaging, the categories most often performed by breast imaging specialists are radiography or fluoroscopy, sonography, and CT. All are performed by more than half of breast imaging specialists. Nonetheless, other radiologists are more likely to perform radiography or fluoroscopy and CT than are breast imaging specialists and, as would be expected, are less likely to perform breast imaging.
The Work of Breast Imaging Specialists
Full-time radiologists who are breast imaging specialists report working,
on average, 49–50 hours per week, somewhat less than radiologists in
general (53 hours on average) (Tables
4 and S4). A third or more of
breast imaging specialists work part-time, more than the 20% statistic for all
radiologists. In contrast, breast imaging specialists who work full-time on
average take the same amount of vacation time per year—about 7
weeks—as radiologistsin general, and on average take the same annual
number of days for professional education andsociety meetings (9.5 days).
|
By most definitions of who is a breast imaging specialist, these
radiologists work at an average of approximately three distinct locations,
similar to other radiologists (Tables
4 and S4), including both other
subspecialists and nonsubspecialists. However, those who work 50% of
their clinical time in the field work at an average of 2.3 locations, fewer
locations than other radiologists.
Overall, approximately 13–15% of breast imaging specialists would like less work, even if that meant their income would decrease proportionately; and a broadly similar percentage would like more work with a corresponding increase in income. Both statistics are broadly similar to those for radiologists overall (Tables 4 and S5). Finally, breast imaging specialists are like other radiology subspecialists and nonsubspecialists in reporting, on average, a level of enjoyment in working as a radiologist that is approximately halfway between "enjoyvery much" and "enjoy somewhat."
Distribution of Work Time of Breast Imaging Specialists
By some definitions of who is a breast imaging specialist, well over 80% of
them report spending some time in clinical practice in hospitals, which is not
significantly different from radiologists overall (89%) (Tables
5 and S6). However, only 77% of
those who spend 30% of their clinical work time in breast imaging report
spending some time in clinical practice in hospitals, which is significantly
less than radiologists overall.
|
Approximately 55–60% of breast imaging specialists report spending some percentage of their clinical time at nonhospital sites, which is higher than some comparison categories of subspecialists. For those breast imaging specialists who are involved in clinical practice in nonhospital settings, the average percentage of work time spent at those sites is three fifths or more, which is greater than the 46% average for all radiologists.
One fourth to one fifth of breast imaging specialists spend time teaching, which is less than the 40% statistic for other subspecialists. Also, breast imaging specialists who report spending some time teaching spend less time teaching (5-6% on average) than other radiologists who teach (12-13%). Breast imaging specialists are also less likely to spend time in research (8-10% spend time in research) than other subspecialists (19%).
The Practices in Which Radiologists Who Perform Breast Imaging Work
Approximately 40% of breast imaging specialists work in private
multispecialty practices, which is higher than for other subspecialists (25%)
(Tables 6 and S7). Conversely,
13% of breast imaging specialists work in primarily academic practices, fewer
than other subspecialists (24%). Note, however, that 20% of the breast imaging
specialists who spend 50% of their clinical work time in their specialty
are academics.
|
Approximately 35% of breast imaging specialists work in private
radiology-only practices, which is not statistically significantly different
from the 41% for other subspecialists. However, significantly fewer than 35%
of radiologists who studied under a breast imaging fellowship or who spend
50% of their clinical work time in breast imaging are in private practice
(13% and 25%, respectively).
A bit more than one third of breast imaging specialists work in practices located primarily in the main city of a large metropolitan area (an area with population > 1 million), similar to other subspecialists, but more than nonsubspecialists (17%) (Tables 7 and S8). In contrast, radiologists who interpret mammograms, but not a very high volume of mammograms (< 5,000 annually), are less likely to work in the main cities of large metropolitan areas.
|
In other respects, the location of breast imaging specialists' practices is
like that of other subspecialists. In particular, both breast imaging
specialists and other subspecialists are much less likely to work in primarily
nonmetropolitan practices than nonsubspecialists, 27% of whom are in
nonmetropolitan practices. On the other hand, radiologists who interpret
mammograms are not infrequently in nonmetropolitan practices (except those who
perform 5,000 mammograms annually).
By the most useful definitions of who is a breast imaging specialist, 15-18% of breast imaging specialists work in practices that serve only hospitals, less than for other subspecialists (27%) (Tables 7 and S9). Conversely, 16-23% of breast imaging specialists work in practices that serve nonhospital sites only, much more than the 6% for other subspecialists. About two thirds of breast imaging specialists work in practices that serve both hospitals and nonhospital sites, similar to radiologists overall and to other subspecialists.
Based on the most useful definitions of who is a breast imaging specialist,
breast imaging specialists work in practices that perform, on average, about
five of the six major categories of radiology procedures, excluding breast
imaging (Tables 8 and S10). For
some definitions of breast imaging specialist, this is, by a statistically
significant (p 
0.01) amount, fewer than the average of 5.6 of
the major categories of non-breast imaging procedures performed in the
practices of all other subspecialists. Table S10 contains specifics for
radiography and fluoroscopy, a low-tech category, and MRI, a high-tech
category. A smaller percentage of breast imaging specialists are in practices
providing radiography or fluoroscopy than is true for nonsubspecialists, and a
smaller percentage of breast imaging specialists are in practices providing
MRI than is true for other subspecialists.
|
Breast imaging specialists work in practices with an average size of approximately 20 radiologists (Tables 8 and S11). This is not significantly different from other subspecialists but is larger than for nonsubspecialists (average oractice size, 12 radiologists). The 25th percentile practice size for breast imaging specialists is approximately 8 radiologists, and the 75th percentile size is almost 30 radiologists (Table S11).
When they are away, 75% or more of breast imaging specialists have coverage
from a member of their group who is in the same subspecialty. For those who
spend 30% of their clinical work time in breast imaging, the number is
85%, higher than the 73% that is charateristic of non–breast imaging
subspecialists.
About two thirds of breast imaging specialists are in practices owned entirely by members of the practice (Tables 8 and S11). This is about the same as the percentage for all radiologists.
Mammograms and the Radiologists Who Interpret Them
The 2003 survey indicates that approximately 38 million mammograms were
performed annually in the United States in 2002–2003. Radiologists who
report that breast imaging is their primary specialty interpreted 33% of these
and radiologists who spend 30% of their clinical work time in breast
imaging interpreted 28% (Fig.
3).
|
5,000 mammograms a
year—constitute 12% of all radiologists who interpret mammograms but
interpret more than one third of all mammograms
[13].
|
|
|
Overall, 62% of all radiologists interpret mammograms (Table 9) [13]. The percentage is significantly lower (43%) for radiologists in practices in the main cities of large metropolitan areas and is significantly higher for radiologists in practices in the suburbs of small metropolitan areas (77%) and in nonmetropolitan areas (88%) [13].
Forty-five percent of mammograms are interpreted by radiologists in private, nonacademic radiology practices, and one third of mammograms are interpreted by radiologists in private, nonacademic multispecialty practices (Fig. 6). Similar percentages of all the radiologists who interpret mammograms are in these types of practices. Academics are 4.5% of all radiologists interpreting mammograms, and they interpret, in total, 6.5% of mammograms (Fig. 6).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Similarly, it is important to note that radiologists who interpreted fewer than 1,000 mammograms in the preceding year, although they interpreted only 6% of all mammograms, are fully 25% of all radiologists who interpret mammograms [13]. If volume standards were raised so that these radiologists did not qualify for interpreting mammograms, a major decline in access would result, and that would probably lead to a substantial decline in the percentages of women 40 years and older undergoing mammography. In short, the United States presently has little alternative to current arrangements.
A greater percentage of mammograms could be interpreted by breast imaging specialists without impairing access or disproportionately increasing the workload of radiologists in this subspecialty if a large increase in the number of radiologists entering the field were achieved. However, such an increase is extremely unlikely because of two current realities: Mammography is by far the leading source of malpractice lawsuits against radiologists, and mammography is a relatively low-revenue activity within radiology. Tort reform and reimbursement increases, respectively, could alter these circumstances, but neither change is prominent on the horizon.
With existing arrangements, breast imaging specialists are not working longer hours than other radiologists or feeling more burdened or less satisfied (Table 4). Thus, concerns about a current shortage of breast imaging specialists seem misplaced if current arrangements are acceptable. However, the combination of unfavorable factors described in the previous paragraph suggests a serious threat of a shortage in the future. Indeed, Farria et al. [18] find evidence of a gathering storm. A large proportion of breast imaging fellowships are unfilled, and breast imaging specialists are not recommending the field as a career, with the reasons described in the preceding paragraph prominent in their rationale.
Breast imaging fellowships appear to be a relatively new training
opportunity, as shown by the relatively young age of those who have completed
these fellowships. This is probably the reason that only a relatively small
fraction of breast imaging specialists have taken a fellowship in breast
imaging—21% of those for whom breast imaging is their primary specialty
and 24% of those who spend 30% of time in the field. In contrast, for
other subspecialties we have studied (pediatric radiology and interventional
radiology), the corresponding fellowship statistics are 70-80%
[15,
19].
The percentage who report that the SBI is one of the two most important professional organizations to which they belong is also relatively low—11-14% compared with approximately 45-55% for the most prominent specialty societies for interventionalists and pediatric radiologists [15, 19]. The relative lack of formal advanced training and professional society involvement among breast imaging specialists supports the notion that breast imaging is a less integrated specialty than others. Perhaps this arises partly from the large percentage of part-timers in the field, who are likely to have relatively strong involvement outside the profession. In any case, the data suggest that breast imaging specialists may not be as strongly organized to express their needs and to raise awareness and support for their problems and issues as other specialists are. This is particularly troubling given that failure to diagnose breast cancer is the most common cause of malpractice lawsuits against radiologists and that mammography is less well reimbursed than other radiology procedures.
The low percentage of academic radiologists who interpret mammograms (Table 9) shows that academic practices are more subspecialized than others. The low percentage of radiologists interpreting mammograms in nonacademic government-owned practices probably reflects the small percentage of women 40 years or older who are among Veterans Affairs and military health system patients.
The fact that by one important definition (spending 30% of clinical
work time doing breast imaging) breast imaging specialists are less likely to
spend any of their clinical work time in hospitals than other subspecialists
probably reflects the widespread existence of freestanding breast imaging
centers and the development of mammography as an in-community service.
The high percentage of part-timers among breast imaging specialists may in part be due to the fact that a high percentage of breast imaging specialists are women and a far greater percentage of women radiologists than men work part-time. Still-prevalent sex role norms about childrearing responsibilities probably explain the much higher percentage of women who work parttime, particularly because on-call duties are typically less required or not required for part-timers.
All in all, it is encouraging, given that mammography pays relatively less well than other radiology services and that it is the most frequent cause of malpractice lawsuits, that breast imaging specialists enjoy radiology as much as other radiologists do.
Study Strength and Limitations
Like other studies, ours has both strengths and limitations. Major
strengths include the following: The data are from a large, carefully
conducted survey that achieved a high response rate through intensive
follow-up. Weighting adjusted for nonresponse bias—that is, differences
between respondents and nonrespondents—by the characteristics used in
the weighting (age and ACR membership), and logistic regression analysis
showed no significant nonresponse bias in terms of sex or region. Paying
careful attention to the completeness of our sample, we included osteopathic
radiologists, who are approximately 3% of all radiologists, and nuclear
medicine specialists with a major connection to radiology. Multiple steps
improved date quality.
The survey nonetheless has noteworthy limitations. As with almost any survey, statistics drawn from it may have inaccuracies from at least three sources: sampling variability (the likely size of these inaccuracies is measured by the SE); nonresponse bias (but only with respect to characteristics not considered in the weighting or logistic regressions); and incorrect or illogical responses (some still remain despite careful and extensive data cleaning).
Acknowledgments
We thank R. James Brenner, president of the Society for Breast Imaging, who
provided helpful comments.
Some of the material in this paper was originally developed by us for the Institute of Medicine (IOM) and some of that material appears in the IOM report Improving Breast Imaging Quality Standards [13]. This material appears here through an agreement with the IOM.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  D. L. Miglioretti, C. C. Gard, P. A. Carney, T. L. Onega, D. S. M. Buist, E. A. Sickles, K. Kerlikowske, R. D. Rosenberg, B. C. Yankaskas, B. M. Geller, et al. When Radiologists Perform Best: The Learning Curve in Screening Mammogram Interpretation Radiology, September 29, 2009; (2009) radiol.2533090070v1. [Abstract] [Full Text]
|
|
|
|
 |
|
 D. L. Miglioretti, R. Smith-Bindman, L. Abraham, R. J. Brenner, P. A. Carney, E. J. A. Bowles, D. S. M. Buist, and J. G. Elmore Radiologist Characteristics Associated With Interpretive Performance of Diagnostic Mammography J Natl Cancer Inst, December 19, 2007; 99(24): 1854 - 1863. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
