Chapter 2.  The Pediatric Rheumatology Workforce:  Current Supply

Current Supply of Pediatric Rheumatologists:  Numbers

In the United States, pediatric rheumatology is among the smallest of the clinical pediatric medical subspecialties (Table 1).  The American Board of Pediatrics (ABP) first offered a certifying exam in pediatric rheumatology in 1992; as of December 2002, the Board has certified 192 pediatric rheumatologists. 

Table 1:  Number of Board Certified Physicians by Pediatric Subspecialty

1 Source:  American Board of Pediatrics available at http://www.abp.org/STATS/WRKFRC/Menu1.htm
² Source: 2003 American Board of Pediatrics Diplomate File

Not all pediatric rheumatologists certify, however.  Data from the 2004 American College of Rheumatology (ACR) Membership Directory reveal that 185 non-trainee physicians in the U.S. identified themselves as pediatric rheumatologists (i.e., Board-certified and non Board-certified) or reported being Board-certified in pediatric rheumatology.  In addition, there were 25 pediatric rheumatology trainee members of the ACR.  ACR and ABP data sources are described in Appendix C.

[D]

Distribution of Pediatric Rheumatologists

 While access to pediatric rheumatology care may be constrained for a variety of reasons, the most conspicuous reason for decreased access is the small number of these providers and their uneven distribution. ^{4, 11}  Past studies have shown that the overwhelming majority of pediatric rheumatologists practice in academic rheumatology settings; ^{6, 12} and nearly all pediatric rheumatologists practice in metropolitan areas. ¹¹  Based on 2004 ACR data, approximately 3 percent of counties in the United States currently have one or more pediatric rheumatologists involved in patient care (Figure 1) and 13 States have none at all: Alabama, Alaska, Arizona, Idaho, Maine, Montana, Nevada, New Hampshire, North Dakota, South Carolina, South Dakota, West Virginia, and Wyoming. ⁶

Table 2: Percent of Pediatric Population Living Within Selected Distances of Board Certified Pediatric Subspecialists, American Board of Pediatrics

The current distribution of pediatric rheumatologists creates a situation in which a substantial portion of the under-18 population in the United States lives more than 50 miles from a provider (Table 2).  Thirty-five percent of the pediatric population in the United States lives more than 50 miles from the nearest pediatric rheumatologist; ¹¹ approximately 18 percent live 100 or more miles from such a provider.  In contrast, less than 10 percent of the pediatric population lives 100 or more miles from a provider for 11 of 16 pediatric subspecialties studied.  The average population-weighted distance between a county in the United States and a pediatric rheumatology provider is 57.9 miles, making it one of the least geographically accessible of the pediatric subspecialties (Table 3).

Table 3:  Average Population-Weighted Distance to the Nearest Provider by Pediatric Specialty

In Figure 2, a Lorenz curve is used to depict the equality of the distribution of pediatric rheumatologists versus the distribution of the under-18 population in the United States.  Over 70 percent of the pediatric population lives within a county that lacks a pediatric rheumatologist; approximately 60 percent of rheumatologists are located in counties where only 10 percent of the pediatric population lives.  Based on the area between the 45-degree line and the Lorenz curve, the Gini coefficient is used to quantify inequality and ranges from 0 (in cases of perfectly even distribution) to 1 (in cases of perfect inequality).  For pediatric rheumatologists in the United States, the Gini coefficient equals .84, suggesting a very inequitable distribution.

Figure 2:  Cumulative Distribution of Pediatric Rheumatologists in United States Counties Weighted by the Population Under-18 Years of Age

[D]

Population

It is important to consider the ratio of pediatric rheumatologists to the under-18 population at a market level.  Because of the low incidence rates of pediatric rheumatic diseases and the geographic concentration of providers, the relevant market for a pediatric subspecialist is likely quite large.  Using the Metropolitan Statistical Areas (MSAs) as a proxy for a market, the relative supply of pediatric rheumatologists was compared to the relative supply of other pediatric subspecialists.  These ratios use “head counts” of providers rather than counts that adjust for percent time involved in patient care because individual level data on percent time in patient care do not exist for all providers.  Furthermore, the percent of time an individual spends in other professional activities likely varies with supply.  That is, providers at institutions with more pediatric rheumatologists may spend more time in research than those in institutions with a single pediatric rheumatologist.  Thus, a single adjustment (i.e., considering every rheumatologist to be involved in patient care at 0.6 FTE) is inappropriate and will not change the relative differences across MSAs.

Only 23 percent of MSAs in the United States have a pediatric rheumatologist available (Table 4).  For all rural (i.e., non-metropolitan) areas, the number of pediatric rheumatologists per 100,000 children under 18 years of age is 0.01.  For all MSAs there is one pediatric rheumatologist per 100,000 children on average.  In the 40 most-populated MSAs, there is one pediatric rheumatologist per 233,000 children on average.  The ratio varies widely in these most populated MSAs from 0.09 per 100,000 children under 18 in Riverside-San Bernardino, California to 1.62 per 100,000 children under 18 in Cincinnati, Ohio.

Table 4:  Ratio of Board Certified Physicians to Under-18 Population (in 100,000) by Pediatric Subspecialty

Source: 2003 ABP Diplomate File

MSA:  Metropolitan Statistical Area

Across all pediatric subspecialties, pediatric rheumatology has the lowest ratio in non-MSAs, the third lowest ratio for all MSAs, and the second lowest ratio in the 40 most populous MSAs.  Some of the differences in supply across specialists reflect, in part, differences in the incidence of diseases treated by these various providers.  The relevant market area may differ across large and small pediatric specialties, too.

Pediatric Rheumatologists’ Perceptions of the Pediatric Rheumatologist Supply

The Arthritis Foundation (AF), in conjunction with the American College of Rheumatology (ACR), created and fielded a survey of pediatric rheumatologists and internist rheumatologists in the United States in March 2004, hereafter referred to as the AF/ACR Survey.  Detailed information on this survey, including descriptive statistics, is provided in Appendix E.  In this survey, pediatric rheumatologists were asked to assess the current supply of pediatric rheumatology care locally and nationally.  Nearly two-thirds of responding pediatric rheumatologists reported a local shortage of pediatric rheumatology care and all respondents reported a national shortage of pediatric rheumatology providers (Table 5).  Responding pediatric rheumatologists were also asked to specify the average wait time for an appointment in their practice:  less than 1 week, 1 to 2 weeks, or 2 or more weeks.  Sixty-five percent of responding pediatric rheumatologists reported that the wait time for an initial patient appointment exceeded 2 weeks in their practice. 

Respondents were asked to select from among a list of potential causes of the shortage; the majority of providers indicated that poor reimbursement contributed to the current shortage.  This finding is not surprising given that Medicaid patients comprise one-third of pediatric rheumatologists’ patients and a recent study showed that the Medicaid-to-Medicare fee ratio was 0.69 (i.e., Medicaid pays 69 cents for every dollar paid by Medicare) in 2003. ¹³ Other common factors cited as contributing to the shortage included poor working conditions and salary concerns.  Among those providers who wrote a specific concern in the open-ended section (n=30), many stated that lack of exposure to pediatric rheumatology during training and lack of mentors contributed to the current shortage.

There was near universal agreement among responding pediatric rheumatologists that the current shortage had important consequences for patients, including increased wait times, delays in diagnosis and treatment, misdiagnosis, and inappropriate treatment.  Among those who responded in the open-ended section (n=11), most reported that the involvement of other physician providers (i.e., general pediatrician and internist rheumatologists) in the care of children and poor outcomes were adverse consequences of the current supply and distribution of pediatric rheumatologists.

Table 5: Pediatric Rheumatologists’ Perceptions of Workforce Shortage, AF/ACR Survey (n=104)

The results of the AF/ACR survey also provided useful insight into how retirement and other activities influence pediatric rheumatologists’ involvement in patient care.  Just over one-fourth of pediatric rheumatologists have decreased their time in clinical care in the 5 years preceding retirement, with an average reduction of 32.2 percent in their patient care hours (Table 6).  The primary reason reported is having obtained salary support from a research source (39.3 percent).  Many providers also reported decreasing their patient care time because another pediatric rheumatologist joined their practice (32.1 percent); they changed employers or career (17.9 percent), or other reasons (25.0 percent).  Of note, seven pediatric rheumatologists reported decreasing time due to retirement or semi-retirement and one responding pediatric rheumatologist was excluded from analyses because s/he had completely retired.

Table 6: Changes in Patient Care Time, Pediatric Rheumatologists AF/ACR Survey

About one-third of physicians plan to decrease their time in clinical care in the next 5 years with an average planned decrease in clinical hours of 33.1 percent.  The primary reason for planning a decrease in time is obtaining salary support from a research source (45.5 percent); however, many also report retirement (21.2 percent), salary support from a business source (18.2 percent), and other reasons (27.3 percent).  About 15 percent (n=6) of responding pediatric rheumatologists indicated that they expected their clinical rheumatology involvement to decrease due to funding from another non-pediatric rheumatologist clinical source.  Those shifting to another clinical area reported expected percent reductions in rheumatology patient care time ranging from 5 percent to 50 percent; thus, they would still be involved in pediatric rheumatology care part-time. 

Pediatric Residency Directors’ Perception of the Adequacy of Pediatric Rheumatologist Supply

Pediatric residency directors oversee the residency training of all pediatric residents in the United States; as such, they have a unique perspective on the current and future supply of general pediatricians and pediatric subspecialists.  In a 2004 survey of pediatric residency directors, described in detail in Appendix E, pediatric residency directors were asked to describe the adequacy of pediatric supply in their catchment area.  The majority of those responding to this question felt that either the supply was inadequate (41.7 percent) or the supply was adequate to allow patient care but inadequate to allow time for research and teaching responsibilities (26.0 percent).  Significantly more directors in institutions that lacked a pediatric rheumatologist felt supply was inadequate (65.0 percent vs. 23.6 percent, p=0.001).  The majority of directors similarly felt that the statewide supply of these providers was inadequate (48.8 percent) or adequate for patient care only (14.5 percent).  Programs without pediatric rheumatologists were significantly more likely to describe the statewide supply as inadequate (61.8 percent vs. 38.9 percent, p<0.01); however, only 12.5 percent of programs with a staff pediatric rheumatologist described the statewide supply as adequate for patient care as well as other responsibilities.

Many residency directors expressed concerns about the shortage of pediatric rheumatologists in their facilities.  Some relevant quotes from the open-ended “comments” section of the survey are provided here.

• “Pediatric rheumatologists are in short supply. We often have to rely on adult rheumatologists to consult on our inpatients, and always have to send away [patients] for outpatient referrals.”
• “Pediatric rheumatologists are like gold.”
• “We desperately need a ped[iatric] rheumatologist and have now for the 2nd yr [year] in a row secured a visiting prof[essor] in same through the Amer[ican] Coll[ege] of Rheumatology ... a finger in the dike both for our pts [patients] care and our housestaff education.”
• “There appears to be a tremendous need for trained pediatric rheumatologists.”
• “There is clearly a shortage of Pediatric Rheumatologists.”
• “We need a ped[iatric] rheum[atologist] in the DC metro area.”

Clearly, pediatric residency directors share pediatric rheumatologists’ sentiments that the current supply of pediatric rheumatologists is inadequate for patient care and medical education.

Both from the perspectives of current pediatric rheumatologists as well as pediatric residency directors who oversee the training of pediatricians, the current supply of pediatric rheumatologists in the United States is inadequate.  Several factors may contribute to inadequate supply, including training capacity, salary concerns, and competing professional demands.  These are discussed, in turn, in the rest of this chapter.

Pediatric Rheumatologist Training

To become a pediatric rheumatologist, one must pursue 10 years of training after completion of their undergraduate degree:  4 years of medical school, a 3-year pediatric residency at an accredited institution and a 3-year pediatric rheumatology fellowship at an accredited program.  After successful completion of this training, a physician is eligible to take the Pediatric Rheumatology Board certification exam.

Pediatric Training Program Requirements

As of July 2002, a pediatric rheumatology program must meet multiple requirements to gain accreditation.  According to the Accreditation Council for Graduate Medical Education  (ACGME) ¹⁴ , they include:

• The program must provide 3 years of continuous training;
• There must be at least two Board-certified pediatric rheumatologists on faculty;
• Physicians in related disciplines, particularly pediatric orthopedics, must be available at the institution for consultation and collaboration;
• Registered physical and occupational therapists must be available;
• The patient population must be sufficiently large and varied to provide residents exposure to both common and uncommon rheumatic diseases; and
• Full support services must be present at the facility, including nuclear medicine, pediatric rehabilitation services, and clinical immunology and electromyography laboratory services.

These requirements are designed to ensure that physicians completing this training are proficient in the diagnosis and treatment of children and adolescents with rheumatic diseases.  While some rheumatic diseases exist in both pediatric and adult populations others are unique to children.  A pediatric rheumatology fellowship provides trainees with the knowledge and skills needed to treat these diseases in the physical, emotional, and developmental contexts of childhood and adolescence.  Exposure to related physician disciplines and allied health professions provides trainees with opportunities to learn the roles of these providers.  Fellowship training also prepares physicians to function as educators and researchers.  

Pediatric Rheumatology Training Programs

According to the American College of Rheumatology (ACR), there are currently 23 pediatric ACGME-accredited rheumatology fellowship programs in 14 States. ¹⁶  According to the American Board of Pediatrics, 19 physicians entered their first year of pediatric rheumatology fellowship training in 2003 and 10 entered their third year of fellowship training.  Over the past 6 years the total number of pediatric rheumatology trainees in the United States has increased from 24 trainees in 1998 to 47 in 2003 (Table 7).  This increasing trend in the number of trainees has been noted for all pediatric subspecialties in recent years. ¹⁷

Table 7:  Pediatric Rheumatology Fellowship Trainees by Medical School and Gender, American Board of Pediatrics

Source:  American Board of Pediatrics ¹⁵

In January 2004, program directors or their administrative assistants were contacted and asked to provide information about the number of first year and total pediatric rheumatology fellowship positions available at their institution as well as the number of positions currently filled (Table 8).  Three-quarters of available pediatric rheumatology fellowship positions were filled in 2003.  The reasons for the failure to completely fill available fellowship slots is not clear; however, some programs did report that insufficient funding constrained the number of slots they could fill.  Because cognitive pediatric subspecialties often do not generate sufficient revenue to support fellows for their entire 3 years of fellowship, programs often depend on grant funds to support their training programs.  Support for the clinical year is variable from institution to institution and may impact the ability to attract and retain fellows.  Because the availability of funding in this field has been limited for several years, the ability of programs to offer pediatric rheumatology in the future is often uncertain. ¹⁸  Given these factors it is not clear that the goal of substantially increasing the number of pediatric rheumatologists can be achieved without dedicated funding sources.

Table 8:  Total Number of Pediatric Rheumatology Fellowship Slots by Program

The Role of International Medical Graduates

The role of international medical graduates (IMGs) in pediatric rheumatology training is unclear; in 2003, over one-quarter of pediatric rheumatology fellows were IMGs.  A past report suggested that many pediatric rheumatology fellowship programs relied on IMGs to fill about half the positions. ¹⁹  Over the past 6 years, IMGs represented 32-46 percent of all pediatric rheumatology trainees (Table 7).  Pediatric rheumatology training programs are, consequently, relatively reliant on IMGs.  

Professor Salary Concerns

As previously mentioned, salary concerns were frequently cited as a cause of the current shortage of pediatric rheumatologists.  In a recent report from the Medical Group Management Association, the median salary for a pediatric rheumatologist at the assistant professor level was $115,022, which was comparable to other cognitive pediatric subspecialties (i.e., those specialties that do not perform procedures); however, it is far below salaries for intensivist and procedural pediatric specialties, such as neonatology ($155,202), pediatric critical care ($144,933), pediatric cardiology ($149,159). ²⁰  Furthermore, the average pediatric rheumatologist’s salary is comparable to the salary of a general pediatrician ($113,343) that has not spent an additional 3 years in training.  Thus, the financial return on the educational and time investment to become a pediatric rheumatologist is low.  No estimates were available for salaries at higher academic ranks due to the small sample sizes; salaries for more junior positions were also not available.

The Impact of Competing Professional Demands

The supply of pediatric rheumatologists is particularly sensitive to the effects of competing professional demands.  Pediatric rheumatologists are significantly more involved in research and teaching and spend a smaller percentage of their time in patient care than their internist peers.  This is due largely to differences in the practice locations of pediatric and internist rheumatologists:  the majority of internist rheumatologists are in private practice, but the majority of pediatric rheumatologists practice at academic medical centers where they are generally expected to see patients, perform research, and educate trainees.     

The 2004 ACR Membership File lists up to three professional activities for each provider.  While 93 percent of practicing (i.e., excluding trainees) pediatric rheumatologists were involved in patient care, only 67 percent listed patient care as their primary professional activity. Nearly three-fourths of all pediatric rheumatologists listed teaching as one of their professional activities while one-quarter and one-half listed basic and clinical research as an activity, respectively.  Past studies have shown that pediatric rheumatologists are significantly less likely than their internist peers to list patient care as a primary professional activity and significantly more likely to be involved in teaching and research. ¹¹  Thus, the same pool of pediatric rheumatology providers that diagnose and treat children and adolescents with rheumatic diseases is also responsible for medical education and research. 

The 2004 AF/ACR Survey, like previous studies, found that responding pediatric rheumatologists spent a substantial percentage of their time in research and teaching (Figure 3).  Likewise, previous studies have found that pediatric rheumatologists spend significantly less time in patient care ^{21, 22} and see significantly fewer patients per week than internist rheumatologists. ²²  Again, these discrepancies reflect differences in the practice locations of these providers as well as differences in the average complexity of adult versus pediatric patients with rheumatic diseases.  Consequently, even when a pediatric rheumatologist is geographically accessible, their availability for patient care may be constrained.

Figure 3:  Average Distribution of Professional Effort Among Pediatric Rheumatologists (n=107) 2004 AF/ACR Survey

[D]

Summary

Less than 200 pediatric rheumatologists practice in a limited number of areas in the United States.  On average, children need to travel 57 miles to reach a pediatric rheumatologist and 20 percent of the pediatric population in the United States lives more than 100 miles from a practicing pediatric rheumatologist.  Thirteen States, including heavily populated states such as South Carolina and Arizona, do not have any pediatric rheumatology providers.  Pediatric rheumatologists unanimously feel there is a National shortage and that this shortage leads to delays in diagnosis and treatment and suboptimal care.  Pediatric residency directors, who oversee the education of pediatric residents and are well-acquainted with the current and future trends in the supply of general and subspecialty pediatricians, echo pediatric rheumatologists concerns about the supply of pediatric rheumatologists.  While the number of pediatric rheumatology trainees has increased over the last several years, one-quarter of rheumatology fellowship positions go unfilled.  Finally, pediatric rheumatologists practice primarily in academic medical centers where the competing professional demands of research and teaching limit their availability for patient care.  Efforts to address the shortage of pediatric rheumatologists must consider the effects of the multiple professional roles that they occupy.