HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

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

Pediatric Radiology at the Millennium¹

¹ From the Department of Diagnostic Radiology, Yale University School of Medicine, 333 Cedar St, SP2-332, New Haven, CT 06520 (H.P.F., J.T., A.M.C., D.S.K.); the Department of Economics, Yale College, New Haven, Conn (H.P.F.); the Department of Radiology, Schneider Children’s Hospital, Albert Einstein College of Medicine, Long Island Jewish Medical Center, New Hyde Park, NY (J.C.L.); and the Department of Research, American College of Radiology, Reston, Va (J.S.S.). From the 1998 RSNA scientific assembly. Received August 2, 2000; revision requested September 13; final revision received January 9, 2001; accepted February 6. Supported in part by the American College of Radiology Research Bureau. Address correspondence to H.P.F. (e-mail: howard.forman@yale.edu).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To evaluate the current practice of and job market for pediatric radiologists in the United States and Canada with a 1998 survey and compare findings with those of surveys from 1980 and 1989.

MATERIALS AND METHODS: Surveys were mailed to the 728 active members of the Society for Pediatric Radiology. Questions covered professional practice, subspecialization, and involvement in evolving technologies. A pediatric radiology help-wanted index was generated from a diagnostic radiology help-wanted index.

RESULTS: Increasing involvement in the interpretation of computed tomographic, ultrasonographic, and magnetic resonance images was found among the 57% (411 of 728) of the members who responded. The attainment of a certificate of added qualification in pediatric radiology was found in nearly three-fourths of the membership, and 85% (348 of 408) had completed a fellowship. More than half of the respondents were involved in interventional procedures amid a continued increase in volume; 24% (100 of 409) of the membership, as compared with 7% in the 1980 survey, was practicing in a community or "other" setting. Subspecialization within pediatric radiology had diminished, and there was a larger percentage of female pediatric radiologists, particularly among the newest members. The job market was robust, having recovered substantially since the middle 1990s.

CONCLUSION: The practice of pediatric radiology has evolved, with increasing involvement in advanced techniques and modalities, as well as a greater presence in community settings. The help-wanted index supports the recent discussions of a personnel shortfall.

Index terms: Economics, medical • Radiography, in infants and children • Radiology and radiologists, socioeconomic issues, pediatric radiology

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In 1980 and 1989, surveys were conducted among the pediatric radiologists in the United States and Canada to determine the evolving nature of their practice in the face of dramatic technologic change (1,2). At the time of the later survey, it was apparent that technologic innovation, with the introduction of magnetic resonance (MR) imaging, computed tomography (CT), ultrasonography (US), and nuclear medicine, had been embraced by the specialty. Further, the ultimate responsibility for imaging of the pediatric age group was being assumed by these specialists rather than by the traditional technology-oriented imagers that had served the adult population. At the time of the earlier survey, it was not apparent that this would occur, as the domain of pediatric radiologists appeared to be limited to radiographic interpretation and fluoroscopic examination. As changes in pediatric imaging have continued, further query of the membership appeared to be in order.

A help-wanted advertising index for the field of radiology has been introduced and validated as a measure of the employment market (3,4). Results of one report (5) suggested that pediatric radiologists are in great demand. The purpose of our study was to evaluate the current practice of and job market for pediatric radiologists in the United States and Canada with a 1998 survey and compare findings with those of surveys from 1980 and 1989.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Survey
A survey was mailed to each of the 728 active members of the Society for Pediatric Radiology in early 1998. The survey was modeled after the 1989 survey, with questions asked about current subspecialization, type of institution, involvement in performance and interpretation of various newer studies, availability of information technology, and demographic data about the member.

We asked members about how they found their current jobs, when they entered the society, and other demographic data regarding their training.

With regard to interventional procedures, we asked about involvement, current practice, and changes in volume. Further, we sought information about the current certificate of added qualification (CAQ) status of the member.

When possible, comparisons with the prior surveys were made. Statistical comparisons were made by using the ² test, when applicable, and a P value less than .05 was considered to represent statistical significance.

Help-wanted Index
A pediatric radiology help-wanted index was generated from a diagnostic radiology help-wanted index by coding all advertisements in the classified advertisement sections of the journals American Journal of Roentgenology and Radiology from January 1991 through December 1999. Three of the authors (H.P.F., A.M.C., and D.S.K.) performed all of the coding, but each advertisement was only coded once, by one of the three. When questions arose, the three authors would reach consensus on correct coding. Each advertisement was classified according to three features: (a) academic versus private practice, (b) geographic region, and (c) radiologic subspecialty. Multiple advertised positions could derive from a single advertisement if it met the following criteria: (a) the advertisement distinctly sought more than one position and (b) the positions could be distinctly described. For example, an advertisement stating that positions were offered with subspecialty interest in vascular and interventional radiology, chest, and mammography was listed as one position with general subspecialty listing. On the other hand, if a department was seeking new faculty for its vascular and interventional radiology section, its chest section, and its mammography section, that would be listed as three separate positions.

Rules followed in coding each advertised position included the following:

1. Advertisements for part-time positions were listed as such but were coded as jobs for purposes of the overall index.
   
2. Locum tenens positions were not included in this study. Positions that stated that the term of employment was less than 1 year were similarly excluded from the analysis.
   
3. Veterans Administration hospitals were coded as academic centers because most have academic affiliations.
   
4. If a university-affiliated group described itself as private or partnership track, it was listed as such. Otherwise, university affiliation with academic rank implied academic coding.
   
5. If a position was distinctly advertised as being split evenly between two subspecialty areas, then it was entered as such. Otherwise, it was listed as the dominant subspecialty; it was listed as general if multiple subspecialties were listed.
   
6. Chairman’s jobs were listed as "other" in terms of subspecialty.
   
7. Chief-of-service jobs at a Veterans Administration or smaller hospital affiliate of a larger program were listed according to the subspecialty requirement of the job or were categorized as general, when appropriate.
   
8. If a radiology job was specifically being recruited for as an administrative position, albeit with a small clinical component, it was listed as "other" in subspecialty.
   
9. Nonradiology jobs were not entered into the database. With the exception of chairman’s positions, jobs that did not include a clinical component were excluded from the database.
   
10. No attempt to screen for repeat jobs (in the same issue or subsequent issues) was made. To remove a potential source of bias, even if an advertisement was identified as a duplicate, it was included.
    
11. The regions of the country used for coding were as follows: a)  Northwest (six states): Alaska, Washington, Montana, Wyoming, Idaho, and Oregon. b) Southwest (seven states): Nevada, Utah, Arizona, New Mexico, Texas, Hawaii, and Oklahoma. c) Midwest (13 states): North Dakota, South Dakota, Nebraska, Kansas, Missouri, Iowa, Minnesota, Wisconsin, Illinois, Indiana, Michigan, Ohio, and Colorado. d) Northeast (10 states): Pennsylvania, New York, New Jersey, Delaware, Connecticut, Vermont, New Hampshire, Maine, Rhode Island, and Massachusetts. e) Southeast (13 states and the District of Columbia): Maryland, the District of Columbia, Virginia, West Virginia, Kentucky, Tennessee, North Carolina, South Carolina, Georgia, Alabama, Mississippi, Arkansas, Louisiana, and Florida. f) California (one state). g) Non–USA positions were excluded because they were rare.
    
12. The subspecialties that were specifically coded were mammography, thoracic imaging, abdominal/cross-sectional imaging, interventional/vascular imaging, neuroradiology, pediatric imaging, emergency radiology, musculoskeletal imaging, nuclear medicine, and general, as noted earlier. Modality-defined positions such as US or MR imaging were coded as "abdominal/cross-sectional."
    

Data Analysis
Data were accumulated for each journal separately and then combined for analysis. Entries advertising for pediatric radiology jobs were abstracted from the general set and evaluated separately. Data were further summarized on an annual basis and plotted, for display, on a monthly basis, with a 12-month rolling average (computed by averaging the current month and the prior 11 months of data) superimposed to remove or decrease seasonal variation.

Statistical analysis of the index for validation has been performed and reported, previously (3,4). We report, here, current applicable data.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Survey
Four hundred eleven responses to the 1998 survey were received, representing a 57% (411 of 728) response rate, as compared with the previous response rate of 66% in 1979 and 64% in 1989. All respondents did not answer every question; thus, the denominator for some queries was less than 411.

Of the total, 64% (242 of 379) of respondents were male, 54% (222 of 409) practiced in a freestanding children’s hospital, and 74% (303 of 410) had a CAQ in pediatric radiology. Eighty-five percent (348 of 408) had done a formal pediatric radiology fellowship, 66% (265 of 402) had use of e-mail, 6% (24 of 401) had a nonpediatric CAQ, and approximately half (199 of 407) had use of teleradiology, with a similar proportion (212 of 407) having access to a picture archiving and communication system, or PACS.

One of the new areas of inquiry regarded involvement with interventional procedures. We surveyed the involvement and recent subjectively reported changes in individual procedural volume. Slightly more than half (207 of 402) of the members reported involvement in interventional procedures. Seventy-one percent (284 of 401) of those surveyed reported that many of the interventional procedures are performed, at least in part, by nonpediatric radiologists; 72% (276 of 383) reported an increase in the total volume of interventional procedures during their career. Only 9% (33 of 383) reported a decrease, with the remainder (19% [72 of 383]) reporting no change.

Twenty-four members (6% of 401) reported that they had an additional CAQ (or added qualification in the case of nuclear medicine). Three had a vascular/interventional CAQ, 10 had added qualification in nuclear medicine, and 11 had a CAQ in neuroradiology.

Table 1 shows the hospital affiliation for each of the three survey periods. Pediatric radiologists are increasingly finding themselves in community and "other" practices: 7% in the original series, 15% in the 1989 series, and 24% (100 of 409) currently.

When we categorize the members according to year of entry into the Society for Pediatric Radiology, arbitrarily choosing those who became members in 1987 and after (n = 197) to compare with those who became members before 1987 (n = 201), we find several differences that are summarized in Table 4. CAQs are more prevalent in the older than in the younger group, albeit not by a statistically significant amount. Consistent with our observation that a large percentage of the membership is newly placed in community hospitals and other settings, we found that the newest members of the Society for Pediatric Radiology were disproportionately present in those settings. The newest members showed greater involvement in cross-sectional imaging than did the senior group. Sixty-one percent (123 of 197) of the older members were involved in the interpretation of MR images, a modality that entered clinical usage after most had finished their training. Finally, there has been a shift in the relative proportion of male and female pediatric radiologists. Whereas nearly three-fourths (138 of 187) of the older group are men, nearly half (85 of 179) of the younger group are women.

The number of job advertisements in pediatric radiology over time is represented graphically in Figure 1. The number of absolute advertised positions placed peaked in June and July of 1992 with 17 advertised positions, was zero on 12 occasions in 1995 and 1996, and then achieved a new higher peak of 20 in November 1999. As with the findings for the general help-wanted index (3,4), the pediatric radiology index exhibited seasonal variation, although it was less clear than in our overall index, in which the middle of the academic year had the most advertisements. In addition, the 12-month rolling average demonstrated peaks and troughs comparable with those of the overall help-wanted index; the rolling average had a peak of 12.3 advertised positions in November 1992 and a nadir of less than one advertised position in the spring of 1996, which was somewhat later than the overall help-wanted index nadir in July 1995, with a new rolling average peak of 12.5 in December 1999. Unlike the overall index, the pediatric help-wanted index trend exhibited a second minor trough in the fall of 1997 and winter of 1998.

View larger version (41K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Graph shows the help-wanted index of pediatric radiology from 1991 to 1999. The raw data are represented by the bars, and the 12-month rolling average is represented by the connected line of diamonds. The first 11 months are an average of the first months in the series because data were available for only 11 months of the 1st year. Note that the most recent raw numbers have exceeded the previous peak from 1992, whereas the rolling average is approaching this peak.

View larger version (35K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Graph shows the 12-month rolling average of academic () and private () advertisements from 1991 to 1999. Note the relatively outsized increases in advertisements for academic positions during the last year, which is similar to that for 1992.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The shift to more female pediatric radiologists is also a notable trend. Even in the most recent data available about residency training, only 26% of trainees in diagnostic radiology are women (7). Thus, pediatric radiology as a subspecialty has a relatively higher number of women, as compared with the pool of available radiologists. A future survey could seek to learn the reasons for this finding.

The trend away from self-identification as a subspecialist within pediatric radiology is an unexpected finding. In the 1989 survey, more than one-fourth of those surveyed identified themselves as subspecialized in gastrointestinal radiology, musculoskeletal radiology, uroradiology, chest radiology, US, and CT. In the current survey, no single category was listed by more than one-fourth of the responding members, and only US exceeded 20% of those surveyed. No single category, including MR imaging, showed a net increase in self-identified specialization. It is probable that, as the specialty has evolved, the tendency to identify oneself as a true specialist has taken on a more rigid meaning. Because most pediatric radiologists are now comfortable with most of the modalities and organ systems, true subspecialization is less common. It may also be the case that the current increase in presence in community hospitals has required greater breadth of abilities and less of an emphasis on subspecialty expertise. This is speculation, as it was not queried.

The data on interpretation show that the trend toward increasing interpretation of all advanced imaging modalities has continued. Net increases were found in all settings in the interpretation of CT, US, and MR images. Interestingly, the highest overall level of individual interpretation of each modality was not seen in the children’s hospital setting. Typically, the highest figures and the greatest increases were seen in the university hospital and the community hospital setting. Thus, in an environment such as a children’s hospital where there is true subspecialization (meaning that not all faculty interpret all modalities), the overall numbers will be lower.

The differences in the practices and the demographics of the older and younger pediatric radiologists (inferred only, as we did not ask for age and instead had to rely on a self-identified year of entry into the Society for Pediatric Radiology) were frequently statistically significant. The younger group is more female, more involved in cross-sectional imaging, and more likely to practice in a community or outpatient setting. Most striking is the fact that the older group is more likely to currently hold a CAQ in pediatric radiology. Although this is partly due to the fact that a CAQ is not awarded at completion of a fellowship (one must complete 1 year of practice after fellowship and then pass oral board examinations), it is impressive that more than 75% of the senior members have their CAQ.

Among the younger members, nearly all interpret CT and US images, and 90% are involved in the interpretation of MR images. This is an increase from the earlier survey, when less than 80% of the overall membership interpreted CT and US images and less than 50% interpreted MR images.

In terms of clinical work, it was notable that 6% had an additional CAQ in neuroradiology, nuclear medicine, or vascular and interventional radiology. Further, the increase, perceived by the membership, in their role in pediatric interventional cases makes this area one that will continue to evolve. Certainly, the unusual pediatric radiologist with added skills in interventional radiology is valuable (8,9).

Although the help-wanted index for the pediatric radiologist subpopulation is limited by a small sample size, the trends bear attention. Pediatric radiology as a specialty was later to recover from the nadir of the mid-1990s. We believe this finding to be meaningful. During the early part of the 1990s, through active recruitment and academic marketing, many trainees were encouraged to enter the pediatric radiology job market. Many of the personnel trained during this time found themselves entering the job market at a time of overall diminished demand in the mid-1990s. Whereas an earlier perceived shortage of pediatric radiologists was probably real, the overall hiring of all radiologists was diminished (3,4), and thus, fewer jobs, on the margin, were available. One could argue that the pediatric job market does not use traditional venues of advertising. Although we agree that a majority of jobs are found by other means, as discussed later, Figure 2 shows that real trends can be identified through surveys that confirm the anecdotal evidence. In our case, it is notable that the current dramatic undersupply of pediatric radiologists (5) is concordant with the new high reached in the pediatric radiology help-wanted index.

Because a substantial part of the data in this article addresses the job market, the findings in the survey about seeking and finding a job were of interest. Although there is much anecdotal evidence, to our knowledge, the means pediatric radiologists have used to find their current positions has not been quantified previously. The respondents indicated that there is no one predominant way to find a position; many different means are often used for a single position. The respondents indicated that journal advertisements are seldom (only 14 [3%] of the 411 respondents indicated that this was the primary source) of definitive value in finding a position.

One might ask why, then, do departments continue to advertise in the absence of a satisfactory return on investment? Further, one might ask how valuable is the help-wanted index, if these advertisements are not useful? To the first question, one might reasonably conclude that advertisements are not all placed to fill a position. In many locales, there is an equal opportunity employment requirement in hiring; and, thus, positions must be advertised, even when a candidate is already identified. For the latter question, it is important to understand the nature of help-wanted indices. As Forman et al (3) and Covey et al (4) have indicated, the help-wanted index is a proxy for the job market. Each practice may advertise in one journal or both, or multiple times in either or both. Such variation, although acknowledged, is not of concern. Economic theory predicts that, as long as there is a cost associated with advertising, employers will weigh such costs in their decisions about the number, size, and distribution of printed advertisements. It is also well known that such indices represent the tip of the iceberg, in most cases, and thus only represent a small fraction of the market. Still, their changes are an indication of the market, as a whole. Even when an advertisement is placed for equal opportunity employment purposes, it clearly indicates a position, filled or not.

As with much of health services research, there are many limitations to this type of empiric research. Surveys are useful, but, by their nature, they reflect only the respondents. There has been an increasing percentage of the membership that has declined to fill out this survey. It is possible that their answers are different from those of the respondents. Further, the survey is only of members of the Society of Pediatric Radiology. It is possible that there are clinically active pediatric radiologists who have different practice patterns and are not members of the society. Given the nature of this survey and the desire to preserve anonymity, we have not attempted to further survey nonrespondents. Former trends have persisted, and supporting evidence from other areas of the literature and personal communications are validating.

The change to more formal training in pediatric radiology, including certification by the Accreditation Council on Graduate Medical Education, may have led to changes in actual clinical experience during the fellowship, and this may account for some of the changes noted in the survey. Although this could have been tested for by querying the respondents about specific aspects of training, it was not included in the current survey.

In conclusion, the specialty of pediatric radiology is experiencing a new surge in demand at precisely the time that there are fewer new trainees. The involvement of pediatric radiologists with all forms of diagnostic imaging has increased dramatically during the past decade, and there is continued evidence that current supply is outstripped by demand.

	FOOTNOTES

 
² Current address: Department of Diagnostic Radiology, Sick Children’s Hospital, Toronto, Ontario, Canada.

³ Current address: Department of Diagnostic Radiology, Memorial Sloan-Kettering Hospital, New York, NY.

⁴ Current address: Department of Pediatrics, Massachusetts General Hospital, Boston.

Abbreviation: CAQ = certificate of added qualification

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

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Leonidas JC, McCauley RGK, Faerber EN. Pediatric Radiologists in the United States and Canada: involvement with newer imaging modalities. Radiology 1981; 138:235-237.[Abstract/Free Full Text]
2. Forman HP, Leonidas JC, Kirks DR. Clinical activities of pediatric radiologists in the United States and Canada: 10-year follow-up. Radiology 1990; 175:127-129.[Abstract/Free Full Text]
3. Forman HP, Kamin DS, Covey AM, Sunshine JS. Changes in the market for diagnostic radiologists as seen through a help-wanted index. AJR Am J Roentgenol 2000; 174:933-988.[Abstract/Free Full Text]
4. Covey AM, Sunshine JS, Forman HP. The job market in diagnostic radiology 1999: updated findings from a help wanted index of job advertisements. AJR Am J Roentgenol 2000; 175:957-961.[Abstract/Free Full Text]
5. Slovis TL. New horizons in pediatric radiology. Radiology 2000; 216:317-320.[Free Full Text]
6. Ecklund K, Share JC. Extension of academic pediatric radiology to the community setting: experience in two sites. Pediatr Radiol 2000; 30:3-6.[CrossRef][Medline]
7. Hricak H. Women, careers, and academic radiology. Acad Radiol 2000; 7:485-486.[CrossRef][Medline]
8. Chait P. Future directions in interventional pediatric radiology. Pediatr Clin North Am 1997; 44:763-782.[CrossRef][Medline]
9. Towbin RB. Pediatric interventional procedures in the 1980s: a period of development, growth, and acceptance. Radiology 1989; 170:1081-1090.[Free Full Text]

This article has been cited by other articles:

				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]

				E. A. Jewett, M. R. Anderson, and G. S. Gilchrist The Pediatric Subspecialty Workforce: Public Policy and Forces for Change Pediatrics, November 1, 2005; 116(5): 1192 - 1202. [Abstract] [Full Text] [PDF]

				M. L. Mayer and A. C. Skinner Too Many, Too Few, Too Concentrated?: A Review of the Pediatric Subspecialty Workforce Literature Arch Pediatr Adolesc Med, December 1, 2004; 158(12): 1158 - 1165. [Abstract] [Full Text] [PDF]

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