In her book American Medicine and the Public Interest: A History of Specialization, Rosemary Stevens (1) writes, "Arguably, specialization is the fundamental theme for the organization of medicine in the 20th century." In Time to Heal, Kenneth Ludmerer (2) also identifies specialty and subspecialty certification as one of the positive actions taken over the last century "to assure that medical practice was conducted at the highest possible level," and he suggests that the trend toward subspecialization "rapidly accelerated" after World War II. While the observations of Stevens and Ludmerer may well be accurate, it is also the case that the movement toward specialization and subspecialization has been more intermittent than constant. There has been an ongoing debate over the past several decades about whether the U.S. health care system would be better served by more primary care physicians than more specialists and subspecialists (3), and the Accreditation Council for Graduate Medical Education (ACGME) even initiated a moratorium on the recognition of new subspecialties from 1992 to 1993 (personal communication, GL Case, June 1992). In recent years, however, as scientific knowledge and clinical advances increased, the pendulum seems to have swung in the direction of promoting new subspecialties. At least to this point in time, the American public seems very protective of its "right" to receive medical care from chosen specialists or subspecialists (4).
The American Board of Psychiatry and Neurology (ABPN) has three primary specialties: psychiatry, neurology, and child neurology. As Langsley suggested, the ABPN has been relatively conservative in its approach to subspecialties (5, 6). The ABPN offered its first subspecialty certificate, in child psychiatry, in 1959 and did not establish its second psychiatric subspecialty, in geriatric psychiatry, until 1991. Since then, however, the ABPN has recognized psychiatric subspecialties in seven other areas. The subspecialties of addiction psychiatry (1993), forensic psychiatry (1994), and psychosomatic medicine (2005) were established only for psychiatrists. The interdisciplinary subspecialty of clinical neurophysiology (1992) was established for both psychiatrists and neurologists, while pain medicine (2000), sleep medicine (2007), and hospice and palliative medicine (2008) are available for psychiatrists, neurologists, and specialists in several other medical disciplines.
The decision by the ABPN to establish a new subspecialty is based on a complicated analysis of several factors (7, 8). These include the current and recent trends in the number of psychiatrists practicing in the subspecialty, the number of current nonaccredited fellowship programs and trainees, the estimated potential for new accredited fellowship programs and trainees, the scientific and academic basis for the subspecialty, and the professional organizational support for the subspecialty. The assessment of several of these factors may be very difficult, especially estimating the potential for new accredited fellowship programs and trainees. Mistakes in this analysis can have significant political and financial consequences not only for general psychiatry and the new subspecialty but for the ABPN as well.
This article reviews the recent trends in ABPN specialties and psychiatric subspecialties. It concludes with a discussion of several of the important implications of these data as well as some of the interrelated social, economic, and political questions—the answers to which may well ultimately determine the fate of any existing or proposed psychiatric subspecialty.
We reviewed ABPN records for data on the numbers of specialty and subspecialty candidates for certification and maintenance of certification as well as for the numbers of specialty and subspecialty certification and maintenance of certification certificates awarded. The website of the ACGME provided the number of accredited training programs and number of trainees (9). The website of the American Medical Association provided data on the percent of residency graduates pursuing fellowship training (10). The annual medical education issues of JAMA provided the data on the numbers of graduates (11—18). To obtain an estimate of the percent of graduates seeking certification, the mean number of graduates in recent years was compared to the mean number of new ABPN examination candidates.
From academic years 2000—2001 to 2007—2008, the number of programs decreased from 186 to 181, and the number of graduates decreased from 1,142 to 985. From 2001 through 2008, the mean number of new candidates for ABPN certification in psychiatry was about equal to the mean number of graduates from general psychiatry and combined training programs (e.g., psychiatry-internal medicine) indicating that almost all new graduates sought certification. Of the 4,568 psychiatrists who were certified between 1994 and 1998, 80% (3,639) have maintained their certification.
The development of the four newest noninterdisciplinary psychiatric subspecialties (geriatric, addiction, and forensic psychiatry and psychosomatic medicine) is described, and background information is provided on child and adolescent psychiatry, the oldest subspecialty. It should be noted that during the early years of subspecialty certification, candidates can be credentialed based on their clinical experience in the subspecialty. After this "grandfathering" period (usually 5 years) ends, all certification candidates must complete an ACGME-accredited fellowship in the subspecialty.
In terms of interest in pursuing fellowship training, the annual survey of training program directors about their graduates' postresidency plans conducted by the American Medical Association suggests that psychiatry graduates are only moderately interested in doing so. In 2008, it was reported that 41% of psychiatry graduates intended to pursue additional training compared to, for example, almost 80% of neurology graduates (10).
Child and Adolescent Psychiatry
Child and adolescent psychiatry differs from the other subspecialties in that it is 2 years in length, and the others are 1 year. From academic years 2000—2001 to 2007—2008, the number of fellowship programs increased from 115 to 121, the number of graduates increased from 347 to 365, and about 67% of graduates sought subspecialty certification. A total of 6,932 certificates have been awarded. Of the 724 diplomates certified between 1995 and 1998, 76% (547) have maintained their certification.
On application for recognition in 1989, the number of non-ACGME-accredited programs was 29. By academic year 2000—2001, there were 62 (114% increase), and by academic year 2007—2008 the number had decreased slightly to 60. The number of fellows in 1989 is not known. In academic year 2000—2001, there were 104 fellows, and in 2007—2008, there were 80 (a 23% decrease). About 68% of graduates sought subspecialty certification (examination now administered every other year). During the "grandfathering" period (1991—1997), 68% of those who qualified via the practice track passed the certification examination compared with 91% of those who completed ACGME-accredited fellowship training. A total of 2,953 certificates have been awarded. Of the 2,425 diplomates certified between 1991 and 1998, 51% (1,245) have maintained their certification.
On application for recognition in 1990, there were 37 non-ACGME-accredited programs. In academic year 2000—2001, there were 43 (16% increase), and in academic year 2007—2008 there were 41 (5% decrease). The number of fellows increased slightly (from 61 to 63) by academic year 2000—2001 and decreased to 50 (21% decrease) in academic year 2007—2008. About 57% of graduates sought subspecialty certification (examination now administered every other year). During the "grandfathering" period (1993—1999), 82% of those who qualified via the practice track passed the certification examination compared with 98% of those who completed ACGME-accredited fellowship training. A total of 2,088 certificates have been awarded. Of the 1,776 diplomates certified between 1993 and 1998, 42% (751) have maintained their certification.
On application for recognition in 1990, there were 28 non-ACGME-accredited programs. By academic year 2000—2001, there were 38 (36% increase), and in 2007—2008 there were 41 (an 8% increase). The number of fellows increased from 32 to 72 (125% increase) in academic year 2000—2001 and stayed the same (72) in 2007—2008. About 87% of graduates sought subspecialty certification (examination now administered every other year). During the "grandfathering" period (1994—2000), 77% of those who qualified via the practice track passed the certification examination compared with 98% of those who completed ACGME-accredited fellowship training. A total of 1,814 certificates have been awarded. Of the 824 diplomates certified from 1994—1998, 54% (441) have maintained their certification.
Upon application in 2002, there were 32 non-ACGME-accredited consultation-liaison psychiatry fellowships. By academic year 2007—2008, there were 36 (a 13% increase). The number of trainees increased from 16 to 48 (200% increase). Because the "grandfathering" period ended after 2009, it is too early to estimate the percentage of graduates seeking certification. During the "grandfathering" period (2005—2009), 83% of those who qualified via the practice track passed the certification examination compared with 98% of those who completed ACGME-accredited fellowship training. A total of 1,080 certificates have been awarded.
The data presented here indicate that most recent psychiatry graduates recognize the value of ABPN certification and attempt to achieve and maintain that status. While the main reasons for wanting certification are undoubtedly different among these specialists, they almost certainly include a desire to demonstrate the competence to provide state-of-the-art quality care, obtain personal and professional recognition, become eligible for academic advancement, and fulfill credentialing requirements of service delivery systems and insurance panels.
The majority of psychiatry graduates evidently do not see the value of subspecialty training. At least to this point in time, most psychiatrists are able to practice and receive reimbursement for their clinical activities without additional training. Many psychiatrists who do pursue subspecialty training do not perceive the added value of obtaining certification in that subspecialty. The ABPN requires all candidates for certification in a psychiatric subspecialty to be certified first in psychiatry. Many fellowship-trained subspecialists evidently consider their certification in psychiatry to be adequate to meet their practice and reimbursement requirements. This seems less true for forensic psychiatrists, where certification likely increases professional credibility during court testimony and perhaps in other forensic activities as well.
The higher pass rates of fellowship graduates compared to "grandfathers" on the subspecialty certification examinations support the added value of fellowship training. In addition, it may be difficult for practicing subspecialists to find the time to adequately study for the breadth of content covered on these examinations.
The original enthusiasm for the expansion of psychiatric subspecialty programs during the 1990s seems to have tempered over the last decade. These programs were established during the era when managed care was on the rise, and many were convinced that subspecialty training and certification would be necessary for insurance credentialing and acceptance on managed care panels. Over time, however, it seems that those predictions have not proved accurate in most areas of the country.
Large numbers of practicing subspecialists desiring certification do not necessarily translate into an adequate number of fellowship programs or trainees recruited into those programs to ensure the long-term survival of the subspecialty. The numbers of "grandfathers" seeking certification have been much greater than the numbers of fellowship graduates doing so for all of the newer psychiatric subspecialties. Over the last several years the number of fellowship programs in geriatric psychiatry and addiction psychiatry have decreased slightly (by two each), and the numbers of trainees have decreased by 23% and 21%, respectively. The numbers of fellowship programs and trainees in child and adolescent and forensic psychiatry have increased slightly. Under the current ABPN examination development process, fees from about 150 candidates are required to cover examination costs. That number of candidates has not been achieved even by administering the newer subspecialty examinations every other year.
Just as subspecialty graduates other than forensic psychiatrists seek certification in smaller percentages than specialty graduates, to date subspecialty diplomates have also maintained their certification in smaller percentages than specialty diplomates. Like their colleagues who never sought certification in the first place, these subspecialty diplomates have evidently concluded that the value of subspecialty certification is not currently worth the effort and expense to maintain it.
Most specialty and subspecialty diplomates who do seek to maintain their ABPN certification have little trouble passing the maintenance of certification examinations (pass rates have exceeded 95%). These results should not be surprising, since all of these diplomates had already passed the relatively difficult Part 1 (multiple choice) and Part 2 (oral) specialty certification examinations, most are only about 10 years into their careers, and the maintenance of certification examinations focus on practical, clinically oriented content. Whether or not these same diplomates will be able to maintain these levels of performance over time remains to be seen.
While the current status of some ABPN noninterdisciplinary psychiatric subspecialties (child and adolescent psychiatry, geriatric psychiatry, and forensic psychiatry) seems fairly secure, the status of others (addiction psychiatry and psychosomatic medicine) remains in doubt. The long-term viability of any of them will most likely be determined by the following interrelated social, economic, political, and scientific forces.
The role of subspecialists in the new health care era remains to be seen. Will the public have ready access to them or will payers for clinical services encourage gate keeping by primary care physicians and specialists? There does appear to be increased public awareness of issues related to physician competency with, for example, doctor "report cards" proliferating on the Internet, health care organizations advertising their percentages of board-certified physicians, and maintenance of licensure being seriously considered by the Federation of State Medical Boards. This may put additional pressure on physicians to objectively demonstrate that they are competent to provide quality care, and board certification (and maintenance of certification) is one well-established mechanism for doing so. At the program level, a crucial factor will be the availability of stipends to support subspecialty training. It is also possible that advances in the underlying sciences may change our conceptions of the subspecialties, and in the process some may be rendered obsolete, and some may become so crucial that they are incorporated into the specialty.
In summary, while the ABPN was relatively slow to embrace the subspecialty movement, in recent years it has recognized increasing numbers of subspecialties. The interest of psychiatrists in training and certification in some of them, however, has not lived up to expectations, and whether or not any of them will thrive in the future will be determined by a complicated array of social, economic, political, and scientific factors.
Dr. Faulkner is President and CEO, and Dr. Juul is an employee, of the American Board of Psychiatry and Neurology. At the time of submission, Drs. Andrade, Brooks, Guynn, Colenda, Reus, Schneidman, and Shaw reported no competing interests. A team of researchers, including Dr. Mrazek, has developed intellectual property which has been exclusively licensed by AssureRx.