| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Both authors: Research Department, American College of Radiology, 1891 Preston White Dr., Reston, VA 20191.
Received September 22, 1999;
accepted after revision October 25, 1999.
Address correspondence to P. E. Crewson.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Data for 1244 practicing diagnostic radiologists were gathered from the American College of Radiology's 1995 Survey of Radiologists. Radiation oncologists, the retired, and residents and fellows were excluded. Responses were weighted to represent all practicing diagnostic radiologists in the United States and were compared with a similar previous survey.
RESULTS. More than one quarter (28%) of diagnostic radiologists are subspecialists. Diagnostic radiology subspecialists were more likely than generalists to have fellowship training, work in center cities, work in large groups, and be employed by an academic institution. They were also generally younger and provided a narrower range of imaging services than generalists.
CONCLUSION. Although most diagnostic radiologists are generalists, subspecialization will likely continue to grow.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In an ongoing effort by the American College of Radiology (ACR) to develop a better understanding of the radiology profession, this study explores the heterogeneity of diagnostic radiology from the perspective of subspecialization and fields of practice. This is the fifth in a series of studies from the ACR's 1995 Survey of Radiologists. The first study was an overview [2], the second investigated women in radiology [3], the third studied characteristics of group practices [4], and the fourth investigated satisfaction in the radiology profession [5].
After a brief introduction to the methods used in the present inquiry, this study will expand on previous, less detailed, overviews of subspecialization [2,3,4,6,7,8] by investigating the distribution and characteristics of diagnostic radiology subspecialists, changes in subspecialization since 1990, and associations between fellowship training and clinical experience after training.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Only practicing (i.e., posttraining and professionally active) diagnostic radiologists were included in this study. This definition of practicing diagnostic radiologists excluded radiologists who were either in residency or fellowship programs, who were retired, or who were specializing in radiation oncology. After these exclusions, the total nonweighted number of respondents was 1244.
Weighting was used to account for differential sampling and response rates of various groups in the sample. The weighted data represent an estimate of what the responses would be if all diagnostic radiologists had been surveyed and had responded. The total weighted number for this subgroup of diagnostic radiologists was 21,440. For tests of statistical significance, the weighted number was adjusted downward to represent the actual sample size so as not to inflate the probability that a finding is statistically significant.
Data from ACR's 1990 Survey of Radiologists were also used in this study. The 1990 survey, a stratified random sample of 2804 diagnostic radiologists, radiation oncologists, and nuclear medicine specialists, achieved a 69% response rate [6, 7]. Many of the questions used in the 1990 survey are identical in working to those asked in the 1995 survey, which allowed us to directly compare results based on the 1990 data with those based on the 1995 data. Again, only practicing diagnostic radiologists were included in the analysis based on the 1990 data, resulting in 1037 survey respondents (20,582 weighted). In subsequent analyses, the sample and weighted numbers for both the 1990 and 1995 surveys vary depending on the response rates to individual survey questions (i.e., respondents often fail to answer all survey questions).
Variables
The key variable in this study was diagnostic radiology subspecialty.
Subspecialty was determined by responses to a question that asked survey
participants to select one "major activity" that best described
what they do in clinical practice. Those who selected general diagnostic
radiology or general radiology (both diagnostic and therapeutic) were
classified as diagnostic radiology "generalists"; and those who
selected one of eight diagnostic radiology subspecialty fields were
categorized as diagnostic radiology "subspecialists." These
subspecialties included CT, interventional or vascular radiology, mammography,
MR imaging, neuroradiology, nuclear medicine, pediatric radiology, and
sonography.
Another variable recorded all imaging techniques presently practiced. This variable collected multiple responses covering diagnostic radiography and fluoroscopy, CT, interventional or vascular radiology, mammography, musculoskeletal radiology, MR imaging, neuroradiology, nuclear medicine, pediatric radiology, and sonography.
The variable measuring years of experience was based on the number of years since each respondent started their first permanent posttraining position. Practice size was based on the number of posttraining radiologists in the respondent's main practice and excluded residents and fellows. The definitions of other variables and response categories are evident in the tables and text.
Some comparisons between generalists and subspecialists are similar to what has been reported in previous studies using the same data [2,3,4]. This overlap is necessary, however, as a reference point for the more detailed analyses presented here; and the exclusion of "other" categories and radiation oncologists will cause some baseline estimates in this paper to vary from previous reports.
Statistical Techniques
Although this study was intended primarily to serve as a descriptive
analysis of subspecialization, statistical tests were conducted, when
appropriate, to help identify meaningful (statistically significant)
associations between substantively important variables. Statistical
differences between the characteristics of subspecialists and generalists were
evaluated using chi-square analysis for categoric comparisons. Analysis of
variance was used for comparisons involving arithmetic means, such as age,
experience, and number of imaging techniques. In conjunction with analysis of
variance, Bonferroni t tests were used to reveal the most substantive
differences between subspecialties.
In reporting results, we used p equal to or less than 0.01 (two-tailed test) as the measure of statistical significance. This conservative p value was used to adjust for the problem of multiple comparisons. Although it is common to conduct significance tests and report unadjusted p values for multiple comparisons [9], as the number of significance tests increases so does the probability that a statistically significant difference will be found by chance alone in at least one of the comparisons. One can adjust for this by dividing the p value threshold (such as 0.05) by the number of set-wise comparisons to give the value necessary to claim statistical significance.
Because estimates of statistical significance derived from the chi-square
technique can be inflated by a large sample size, only differences that were
both statistically significant (p 
0.01) and exhibited at least a
minimal level of association (0.10 as measured by Cramer's V) were reported.
The Cramer's V coefficient can range from 0 (no association) to 1 (perfect
association). Although many other measures of association are available for
the chi-square technique, Cramer's V has the additional quality of not being
affected by sample size or the number of comparison cells in a two-dimensional
table.
In many cases, the subspecialty sample size was extremely small. To assist in evaluating the reliability of the percentages reported for subspecialty subgroups, Table 1 includes margins of error. Based on a 95% confidence interval, these margins of error represent how far the actual percentage in the population (e.g., if we had data for all diagnostic radiologists in the United States) may range from the percentage reported from our sample. Although margin of error is reported only in Table 1 for percentages, the implication that margin of error increases as sample size decreases remains true for all descriptive statistics.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
The characteristics of generalists and subspecialists are reported in Table 1. Individual characteristics include sex, fellowship training (yes or no), age, and years of experience after training (measured by the number of years since starting their first permanent position after training). Practice characteristics include work status (full- or part-time), group type, number of practice locations, group size, and number of imaging techniques performed. Environmental characteristics are limited to primary workplace (center city, suburb, small metropolitan area, nonmetropolitan, or rural) and region of the country (East, Midwest, South, or West).
Statistically significant differences between generalists and all other diagnostic radiologists who were subspecialists were found for the individual characteristics involving fellowship training, age, and years of experience. Subspecialists as a group were more likely than generalists to have fellowship training (76% versus 40%) and subspecialists, on average, were younger (45 versus 49 years old) and consequently had fewer years of experience (11 versus 15 years) than generalists.
Statistically significant variations were also found for practice and environmental characteristics. Subspecialists as a group were more likely than generalists to work full-time (95% versus 88%), practice in an academic institution (43% versus 9%), and practice in a larger group (an average of 19 versus 10 radiologists per group). As would be expected, subspecialists provided a narrower range of services than generalists. The average number of imaging techniques was 8.8 for generalists compared with 6.8 for all subspecialists grouped together. This number ranged from a high of 7.8 for interventional or vascular subspecialists to 3.1 for nuclear medicine subspecialists.
Statistically significant variations were found between generalists and subspecialists for both environmental characteristics. Subspecialists as a group were more likely than generalists to work in a center city (44% versus 17%) and less likely to work in a nonmetropolitan or rural area (4% versus 21%). Subspecialists were also more likely to be located in the East (29% versus 25%) or West (25% versus 18%) than were generalists. Differences between subspecialists and generalists for sex and number of practice locations did not reach statistical significance.
Other statistically significant differences were evident when investigating differences separately between each of the eight subspecialties identified in Table 1 and generalists. Women radiologists accounted for only 11% of generalists, but they represented more than 30% of those subspecializing in either CT (34%), mammography (36%), pediatric radiology (38%), or sonography (32%). Although 40% of generalists had fellowship training, such training was much more common for subspecialists. With the exception of mammography and sonography, more than 68% of other subspecialists had fellowship training. In addition, compared with the mean age of 49 years for generalists, radiologists specializing in interventional or vascular procedures, MR imaging, and sonography were, on average, about 7 years younger. Overall, radiologists with the most experience were found primarily practicing in the areas of general diagnostic, nuclear medicine, and pediatric radiology. The most dramatic difference in experience was found between interventional or vascular subspecialists (8 years) and general diagnostic radiologists (15 years).
Practice characteristics also differed according to major clinical activity. Although 88% of generalists worked full-time in 1995, more than 96% of those specializing in CT, interventional or vascular procedures, MR imaging, and neuroradiology worked full-time. Although most generalists (58%) practiced in one or two locations, most MR imaging (51%) and neuroradiology (52%) subspecialists practiced in three or more locations. As would be expected, the average practice size for generalists (10.3) was almost half that of many of the subspecialists. With the exception of mammography (14.9) and nuclear medicine (17.7) subspecialists, all other subspecialists worked in groups that were statistically larger than the average group size for generalists.
Variations were also found in the environmental characteristics involving workplace location and region of the country. Most diagnostic radiologists subspecializing in CT (56%), pediatric radiology (65%), and sonography (68%) were located in central cities of large metropolitan areas. In comparison, generalists were much more likely to be found in either small metropolitan (40%) or rural (21%) areas. Regional differences also appeared among subspecialists. Diagnostic radiologists subspecializing in CT (44%) or mammography (55%) were more likely to be found in the East, whereas a large share of sonography subspecialists (50%) practiced in the West.
Table 1 is limited to
reporting the distribution (percentages) of characteristics in generalist and
subspecialty subgroups; however, additional information can be gained by
investigating subspecialization from the perspective of the diagnostic
radiologists' characteristics. As an example, in comparing subspecialization
by age of the diagnostic radiologist, we found that diagnostic radiologists
older than 54 years were significantly less likely to be subspecialists
(18%) than diagnostic radiologists less than 45 years old (
33%).
Diagnostic radiologists working full-time were more likely to be
subspecialists (30%) than diagnostic radiologists working part-time (15%); and
only 7% of diagnostic radiologists working in nonmetropolitan or rural areas
were subspecialists compared with 50% of diagnostic radiologists working in
central urban areas.
Change in Specialization Between 1990 and 1995
Data from the 1990 and 1995 surveys of radiologists were used to compare
the distributions and characteristics of generalists and subspecialists
(Table 2). A statistically
significant difference existed between 1990 and 1995 in the overall
distributions of subspecialization in diagnostic radiology. The percentage of
diagnostic radiologists who identified their major clinical activity as
general diagnostic radiology decreased from 79% in 1990 to 72% in 1995,
whereas those who identified their major clinical activity as either
interventional or vascular radiology or neuroradiology increased by three
percentage points each. Other subspecialties showed little, if any, change
between 1990 and 1995.
|
With one exception, no statistically significant differences existed between 1990 and 1995 in the individual and practice characteristics of generalists and subspecialists. Diagnostic radiologists who specialized in mammography in 1995 had significantly fewer years of experience as practicing radiologists (11.8 years) than those who specialized in mammography in 1990 (24 years). Although other observable variations were seen between 1990 and 1995 in the characteristics of subspecialists, none reached the threshold necessary for statistical significance: several of the subspecialties had small sample sizes, resulting in large margins of error.
Fellowship Training and Clinical Practice After Training
As previously reported in job market studies
[10,11,12],
most (>70%) of new graduating residents take fellowship training before
entering the job market. Little is known, however, about the association
between fellowship training and clinical practice after training. To address
this issue, reports of specialized training in a fellowship program were
compared with major clinical activity.
As of 1995, 50% of all diagnostic radiologists reported fellowship training, a thirteen percentage-point increase from what was reported in the 1990 Survey of Radiologists by the ACR [6]. Overall, 15% of all diagnostic radiologists reported fellowship training in body imaging, 12% reported fellowship training in interventional or vascular procedures, and 11% reported fellowship training in neuroradiology. The percentage of all diagnostic radiologists with fellowship training in any one of the other subspecialty fields did not exceed 6%.
A statistically significant association was seen between major clinical activity and fellowship training. As an example, more than half of diagnostic radiologists subspecializing in either interventional or vascular radiology, MR imaging, neuroradiology, nuclear medicine, pediatric radiology, or sonography had fellowship training specifically related to their major clinical activity. This figure ranged from a high of 91% of all neuroradiology subspecialists with fellowship training in their subspecialty to a low of 4% of mammography subspecialists with related fellowship training.
As previously discussed, past studies of graduates of diagnostic radiology training programs have shown that fellowship training has become the norm for many. To differentiate the experiences of more recent (1990s) graduates of diagnostic radiology training programs from those who entered the profession earlier, the fellowship rates for diagnostic radiologists were dichotomized into those with 5 years or less and those with more than 5 years of posttraining experience.
Although 54% of all diagnostic radiologists with more than 5 years of experience had no fellowship training, only 35% of all diagnostic radiologists with 5 or fewer years of experience had no fellowship training. In addition, diagnostic radiologists with 5 or fewer years of experience were twice as likely to have training in body imaging (CT, MR imaging, sonography) (25%) as those with more than 5 years of experience (12%). Overall, the distribution for radiologists with 5 or fewer years of experience was similar to what has been reported by recent graduates [1]. Generally, too few cases were studied to develop meaningful statistical comparisons between the two experience levels for each of the subspecialty categories. However, for generalists, a threefold difference was seen between the two experience levels in the percentage having body imaging fellowships.
A comparison was also conducted to evaluate the percentage of those with fellowship training who were able to subspecialize in the same imaging technique their training was in. Diagnostic radiologists with fellowship training in neuroradiology were the most likely (75%) to later subspecialize in the same imaging technique, followed by those with fellowship training in pediatric radiology (48%), interventional or vascular radiology (39%), body imaging (27%), nuclear medicine (17%), and mammography (8%).
Because of the practical demands facing diagnostic radiologists to provide a wide range of radiology services, subspecialization did not completely exclude performing other imaging techniques (Table 3). A comparison between imaging techniques performed and primary clinical practice showed substantial crossover between subspecialties. Excluding nuclear medicine, more than 80% within each subspecialty also performed conventional radiography and fluoroscopy; and more than 50% performed sonography. As evidenced in Table 3, nuclear medicine subspecialists appear to be the most specialized group of diagnostic radiologists. Only 59% of nuclear medicine subspecialists also performed conventional radiography and fluoroscopy, almost 26 percentage points fewer than the other subspecialties. Involvement by nuclear medicine subspecialists in any other imaging technique was considerably less.
|
Data Limitations
Although data collection was completed in May 1995
[2], this study provides a
valuable reference point for measuring trends in the subspecialization of the
radiology profession. However, one should not interpret these findings as
representative of the current status of subspecialization. In fact, the
changes evident between 1990 and 1995 suggest subspecialization is likely to
be greater today than was reported from the 1995 study. In interpreting our
findings, remember that the data are based on radiologists' self-reports and
perceptions.
Most subspecialties have low sample size, resulting in a large margin of error. As a result, this study was conservative both in the statistical analyses and the conclusions presented. In addition, the measure of subspecialization was demanding in that it required radiologists to spend at least 50% of their time performing one type of clinical practice or imaging technique to qualify for a subspecialty. Most subspecialty societies have less demanding requirements for membership.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although fellowship training is not universal, most subspecialists have fellowship training in their subspecialty, and this trend is likely to increase if the rates of fellowship training and subspecialization reported for younger, less experienced radiologists continue. Overall, however, the evidence reported from the 1995 Survey of Radiologists suggests generalists will continue to represent a majority in the diagnostic radiology profession. Even though subspecialization is likely to grow—fueled by younger fellowship-trained radiologists—the number of diagnostic imaging techniques in widespread use today is unlikely to decrease in the near future, nor will the demand for a broad range of general diagnostic radiology services.
Acknowledgments
We thank Barbara Schepps, committee chair, and the Radiologist Resources
Committee of the ACR Commission on Human Resources for oversight and helpful
suggestions.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. S. Baxi, L. Liberman, C. Lee, and E. B. Elkin Breast Imaging Fellowships in the United States: Who, What, and Where? Am. J. Roentgenol., February 1, 2009; 192(2): 403 - 407. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. S. Lewis, J. H. Sunshine,, and M. Bhargavan A portrait of breast imaging specialists and of the interpretation of mammography in the United States. Am. J. Roentgenol., November 1, 2006; 187(5): W456 - W468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Merewitz and J. H. Sunshine A Portrait of Pediatric Radiologists in the United States Am. J. Roentgenol., January 1, 2006; 186(1): 12 - 22. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Sunshine, R. S. Lewis, and M. Bhargavan A Portrait of Interventional Radiologists in the United States Am. J. Roentgenol., November 1, 2005; 185(5): 1103 - 1112. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. T. Bramson and G. A. Taylor SOS: Can We Save Pediatric Radiology? Radiology, June 1, 2005; 235(3): 719 - 722. [Full Text] [PDF]
|
|
|
|
|
|
Y. S. Cypel and J. H. Sunshine Basic Characteristics of Radiology Practices: Results from the American College of Radiology's 1999 Survey Am. J. Roentgenol., August 1, 2003; 181(2): 341 - 349. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Sunshine, Y. S. Cypel, and B. Schepps Diagnostic Radiologists in 2000: Basic Characteristics, Practices, and Issues Related to the Radiologist Shortage Am. J. Roentgenol., February 1, 2002; 178(2): 291 - 301. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. J. Becker 2000 RSNA Annual Oration in Diagnostic Radiology: The Future of Interventional Radiology Radiology, August 1, 2001; 220(2): 281 - 292. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Enzmann, P. M. Anglada, C. Haviley, and L. A. Venta Providing Professional Mammography Services: Financial Analysis Radiology, May 1, 2001; 219(2): 467 - 473. [Abstract] [Full Text]
|
|
|
|
|
|
L. F. Rogers Missing the Meeting: Technologic Innovation and Work Demand Versus Payment Initiatives Am. J. Roentgenol., May 1, 2000; 174(5): 1189 - 1189. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
