Adolescent Medicine: Title and subTitle BreakWorkforce Trends and RecommendationsAdolescent Medicine

Improving adolescent health is an urgent national priority in the United States. Of the 467 national health goals included in Healthy People 2010, one hundred seven pertain to adolescents; of these, 21 are considered critical and center on injury, violence, substance use, mental health, reproductive health, and chronic disease.¹ Although some of these objectives can be achieved using traditional models of causation and prevention, others represent conundrums in which adolescents with the greatest needs have poor access to care or receive care provided by physicians and other health care professionals who are inadequately trained to provide the needed services.²

In 2009, the National Research Council and Institute of Medicine published a report entitled Adolescent Health Services: Missing Opportunities.² The report addressed the increasingly diverse, complex, and unmet health care needs of adolescents in the United States and outlined research directions to inform the development of systems of care to better meet these needs. Subsequent studies and commentaries have documented the low rates of preventive care among adolescents and have called for improvements in insurance, reimbursement, access, professional training, screening, and outcome measures.^{3 - 5} All solutions, however, rely on a core workforce of physician experts in adolescent medicine (AM) to educate residents about the care of adolescents, provide consultation and specialty care for adolescents, pursue research, and prepare AM fellows to assume these responsibilities on completion of training.

Academic leaders, professional organizations, and public advocates for adolescent health have expressed increasing concern about the current and future capacity of the AM physician workforce.² The numbers of AM physicians and fellows in training (designated by the Accreditation Council for Graduate Medical Education [ACGME] as subspecialty residents) are low in comparison with most other subspecialties; however, the number of adolescents in the United States and the complexity of their problems has increased, and the scope, the reason for the workforce problem, and the solutions remain elusive.

In an effort to develop evidence-based strategies to strengthen the AM workforce, we reviewed data pertaining to the numbers of adolescents, AM physicians, and AM fellows in the United States; trends in the number of fellows since board certification and program accreditation were instituted; salaries of AM physicians and comparable subspecialties; and job availability for AM physicians. Our primary sources were the US Census Bureau, US Department of Health and Human Services, American Board of Pediatrics, American Board of Internal Medicine, American Board of Family Medicine, ACGME, and PubMed, using the search terms adolescentmedicine, fellowship training, training programs, manpower, and workforce from January 1968. Salary data were obtained from the 2006-2007 Association of Administrators of Academic Pediatrics and the Medical Group Management Association Report on 2007 salaries, which does not report data by academic level. Our objectives were to (1) elucidate the scope of the AM workforce problem, (2) present evidence supporting or refuting factors thought to influence AM career decisions, and (3) recommend strategies designed to strengthen the AM physician workforce.

There will be an estimated 53.2 million adolescents aged 10 to 19 years in the United States by 2050 compared with 40.7 million in 2000 and 29.4 million in 1980.⁶ In 2009, there were 43.0 million individuals aged 15 to 24 years in the United States, representing 14% of the total population.

Seventeen percent of adolescents aged 12 to 17 years have special health care needs and 10% have functional limitations due to chronic health conditions.⁷ Another 18% are overweight, and 65% do not meet the recommended levels of physical activity. Binge drinking in the past month is reported by 20% of adolescents aged 16 to 17 years and 36% of those aged 18 to 20 years. One half of sexually active 12th-grade students do not use condoms. Despite these health problems and health risks, 16% of individuals aged 15 to 17 years have not seen a health care professional and 37% have not received preventive health care in the past 12 months. Among adolescents who do access care, most do not receive the anticipatory guidance or brief office interventions from adolescent health care professionals who have learned effective behavioral and public health skills shown to improve health behavior.^{2 - 5} In its 2009 report, the Institute of Medicine concluded that better health care services for adolescents requires improved training in adolescent health for primary care professionals who deliver those services and an increase in the number and availability of AM physicians for training, consultation, research, and program development.²

Certification in AM was first issued in 1994 by the American Board of Pediatrics and the American Board of Internal Medicine and, in 2001, by the American Board of Family Medicine. In 2009, of the 744 certified AM physicians, 547 (73.5%) were pediatricians,⁸ 66 (8.9%) were internists,⁹ and 131 (17.6%) were family medicine physicians.¹⁰ The number of pediatricians certified in AM ranked 13th of the 14 subspecialties whose examinations were administered by the American Board of Pediatrics in 2008. Of 967 first-time applicants to the 2008 pediatric certifying examination who planned pediatric fellowship training in the coming year, only 20 (2.1%) identified their subspecialty as AM.⁸ The Table compares physician workforce and population data for the 3 age-based subspecialties: AM, neonatal-perinatal medicine, and geriatric medicine. The number of physicians certified ever column refers to the total number of first-time certifications issued in each field by any board, regardless of subsequent recertification.^{8 - 11} The population data for ages 10 to 24 years (ie, adolescent medicine), annual births (ie, neonatal-perinatal medicine), and ages 65 years and older (ie, geriatric medicine) are obtained from the 2008 US Census Bureau.⁶ Population per physician is calculated as the ratio of the patient population served by each field to the number of physicians certified ever within each field. The ratio for AM is 94 times higher than the ratio for neonatal-perinatal medicine and 26 times higher than the ratio for geriatric medicine.

Accreditation of AM fellowship programs by the ACGME began in 1998. Many of the estimated 40 to 50 programs that existed before 1998 offered 1 year of clinical training with little or no research training.^{12 - 13} Although the American Board of Internal Medicine and American Board of Family Medicine require a 2-year fellowship without a research component for board eligibility, most AM fellowship programs are based within departments of pediatrics and structure their programs to adhere to the American Board of Pediatrics' requirements with 3 years of training that incorporates research training and creation of a scholarly product. Some observers have suggested shorter fellowships; however, the skills needed to provide teaching and academic leadership in the field cannot be acquired in a 1-year fellowship. After the establishment of the ACGME certification in 1998, the number of fellowship programs declined, reaching a plateau of 24 to 26 programs during the subsequent 10 years.¹⁴ The annual number of AM fellows at levels 1 to 3 was 64 in 1998 and 67 in 2010.⁸ Thus, there has been modest variability but no sustained change in the number of AM fellows in the past 12 years. In comparison, the annual number of fellows at levels 1 to 3 in all pediatric subspecialties increased steadily from 1863 in 1997-1998 to 3607 in 2009, a 94% increase. The number of pediatric nephrology fellows increased from 65 to 151, and the number of pediatric rheumatology fellows increased from 24 to 87 during the same period. Both subspecialties had similar numbers of fellows as AM in 1995.

Between 1998 and 2010, the annual number of AM fellows at levels 1 to 3 with core specialty training in internal medicine or family medicine ranged from a peak of 10 (16% of all AM fellows in 2001) to a trough of zero. The annual percentage of AM fellows at levels 1 to 3 who were women increased from 64% in 1998 to 82% in 2010. In comparison, the annual percentage of fellows at levels 1 to 3 in all pediatric subspecialties who were women increased from 47% in 1998 to 61% in 2010. The number of men choosing higher-salaried pediatric subspecialties, for example, neonatology, cardiology, critical care, and emergency medicine, continued to increase during this period yet not as much as the number of women. Some pediatric subspecialties with salaries comparable to those of AM (see “Job Availability, Career Choice, and Salaries” in this section) also saw an increase in the number of male fellows (nephrology and rheumatology), but others saw no change (endocrine) or a decrease (infectious diseases and AM). Men in general pediatrics or in a pediatric subspecialty were more likely to report earning potential as 1 of the 2 most important reasons for choosing a postresidency career, compared with their women counterparts.¹⁵

In 1998, pediatric department chairmen identified AM as 1 of 8 specialties in which the number of faculty to be hired in the next 5 years would exceed the number of graduating fellows.¹⁶ Twelve years later, the projected deficit in AM physicians appears to have been accurate, as there were 23 graduating AM fellows and 22 AM physician positions posted on one Web site alone, maintained by the Society for Adolescent Medicine.

Rarely does one factor determine career choice for pediatric residents.¹⁵ Factors identified as contributing to career choice include structured lifestyle, interest in specific diseases/patient populations, interest in a research/academic environment, role models, and financial considerations. Exposure to role models was important in some AM fellows' career choice.¹⁷ Furthermore, a program undertaken by both national professional organizations of pediatric nephrologists and pediatric rheumatologists to increase exposure to potential role models and their specialty through resident scholarships to their annual scientific meetings during the past decade has been temporally related to the increase in fellows in those specialties.

It has been unclear whether salary levels for AM specialists play a role in low numbers of fellows in training. From the 2006-2007 Association of Administrators of Academic Pediatrics data, the national average AM physician's 2006 salary at the assistant professor level, including base salary and bonus, was $122 000. This was ranked 16th of 20 pediatric subspecialties, with the same salary as pediatric endocrinology and a higher salary than child development, infectious diseases, and genetics specialists. At the same time the average AM salary was within $5000 of pediatric nephrology, rehabilitation, and rheumatology. Another database, the 2007 Medical Group Management Association Report, reported that the mean AM salary was $139 807; the median was $128 225, ranking it 15th of 19 pediatric subspecialties, above pediatric endocrinology, pediatric hospitalist, clinical laboratory and immunology, and child development and within $10 000 of pediatric hematology-oncology, infectious diseases, rheumatology, and genetics. There was an excess number of fellows in training related to projected faculty positions available in endocrinology and infectious diseases. The increase in numbers of endocrine and infectious diseases fellows in training compared with AM fellows in training cannot be explained by simple supply and demand. Job availability is a broader and more complex issue than the data here represent. Many posted positions in AM are in academic centers, yet fellows enter a broader array of jobs in clinical teaching, reproductive health, college health, school-based health, and other public sector jobs than those represented in sites and journals reviewed by the authors. Several academic center–based positions appear to be filled slowly, at both entry and leadership levels, suggesting that the supply is not easily meeting the demand. If this were the case, then supply and demand would dictate that salaries to attract academically based AM specialists should be increasing. They are not. Salaries for pediatric endocrinologists in 2007 were lower than those for AM physicians. The amount of debt increased 37% from 1997 to 2002 for all physicians graduating from pediatric residency programs.¹⁸ Salary and debt, albeit factors in selection of careers for some pediatricians entering higher-compensation subspecialties (cardiology, critical care, and neonatology), do not appear to explain the differences in the number of fellows entering AM compared with other pediatric cognitive subspecialties. This is consistent with reports that earning potential was the fifth most common reason cited for choosing a postresidency career for physicians choosing lower-compensation pediatric subspecialties.¹⁵ The relative importance of earning potential in making a career choice may be different for pediatricians and internists. McDonald and colleagues¹⁹ reported that the number of residents choosing a career in an internal medicine subspecialty was slightly lower for those with $100 000 to $150 000 in debt compared with those with no debt (57.5% vs 63.5%). These data do not address reasons for career choices for family medicine physicians.

AM physicians require training and skills to provide complex clinical care, teaching, research, and program development in contrast to primary care physicians who require skills to provide clinical preventive services for adolescents.

The goal of subspecialty fellowship training should be to produce the next generation of leaders, teachers, and physician scientists who can be the driving force for innovative education to (1) teach AM fellows and primary care residents and contribute to the training of other primary care professionals, including nurses, nurse practitioners, and physician's assistants, about adolescent health care; (2) perform cutting-edge research; (3) provide clinical care for complex adolescent problems; and (4) help create comprehensive adolescent health care services. Drawing the distinction between adolescents' health problems that can be diagnosed and treated in primary care versus those that require referral to an AM physician is essential. For example, a primary care physician would be expected to perform the initial evaluation and identify the diagnosis of a patient with anorexia nervosa. That physician would then be expected to treat the patient in consultation with an AM physician.

From the findings presented herein, we have developed the following recommendations with the goal of increasing the number of physicians entering AM fellowships.

The academic successes of AM during the past 2 decades are marked by board certification, fellowship program accreditation, residency curricula creation, and the evolution of a remarkably respected scientific journal, the Journal of Adolescent Health. These same accomplishments have increased professional and public recognition of unmet population needs and the specialists who can help address them. The adolescent population is large, diverse, underserved, and characterized by increasingly complex medical and behavioral issues. Meeting their health care needs is a national priority. Primary care professionals who treat adolescents want and need adolescent-specific training in anticipatory guidance, screening, counseling, and management of common adolescent problems. A larger workforce of AM physicians is needed to provide this training, consult on complex medical and psychosocial issues when requested, and lead research efforts that will advance knowledge in the field. Developing this workforce will require improved recruitment into fellowship training; mentorship, policies, and resources that support trainee and faculty diversity; and articulation of the skills that define an AM physician.

Correspondence: Dr Hergenroeder, Texas Children's Hospital, 6701 Fannin St, Ste 1710.00, Houston, TX 77030.

Financial Disclosure: None reported.

Funding/Support: Partial support for this project was provided by grant T71 MC00009 from the Maternal and Child Health Bureau (Drs Emans and Hergenroeder).

Additional Contributions: We acknowledge Gail McGuinness, MD, and James Stockman, MD, from the American Board of Pediatrics for their consultation on this project and Edie Moore, Society for Adolescent Medicine, for accessing archives of the Journal of Adolescent Health.