According to the American Board of Medical Specialties, the purpose of certification is "to provide assurance to the public that a physician specialist…has successfully completed an approved educational program and an evaluation process which includes an examination designed to assess the knowledge, skills, and experience required to provide quality patient care in that specialty" (1). In his history of American medical education, Ludmerer (2) cited the establishment of standards for specialty and subspecialty certification as one of the actions taken by academic medicine "to assure that medical practice was conducted at the highest possible level."
A recent article addressed the relationship between physician performance and certification by a member board of the American Board of Medical Specialties. Sharp and colleagues (3) conducted an extensive literature review that yielded 11 studies meeting the screening criteria. Of the 29 findings reported in these papers, 16 indicated a significant, positive association between certification status and clinical outcomes, and 13 demonstrated no association.
The authors noted that they were surprised that so few studies had been done but pointed out that 87% of licensed physicians in the United States have attained board certification, making it difficult to identify comparison groups of noncertified physicians. In addition, it is also challenging to obtain estimates of the patient care outcomes of individual physicians.
While the certification process is voluntary and is not linked to licensure, in recent decades the significance of certification appears to have increased as health care organizations have adopted certification as an employment requirement and insurers have made it a condition for reimbursement. Hence, it is reasonable to assume that most graduates of residency training programs will seek board certification. Writing in 1997 about the history of specialization, Stevens (4) commented, "Even in 1970, with the then-recent formation of a specialty board for family practice, it was clear that all future physicians would be specialists, typically board-certified, as indeed they are today."
The purpose of this study was to track the progress of a cohort of graduates of psychiatry residency training programs in achieving certification by the American Board of Psychiatry and Neurology (ABPN). These data provide a detailed picture of how recent graduates perform on the ABPN’s examinations.
The ABPN has a three-step process for certification in psychiatry. Candidates must first meet the credentialing criteria, including completion of residency training in programs accredited by the Accreditation Council for Graduate Medical Education and possession of an unrestricted license to practice medicine in the United States or in Canada.
The part I examination, which was first administered in 1967, is a multiple-choice examination that broadly samples the basic and clinical sciences relevant to psychiatry (5). Part II is an oral examination that assesses the candidates’ clinical skills (6). In order to ensure that candidates have a current fund of knowledge at the time of certification, a passing grade on the part I examination is valid for a period of 6 years or three opportunities to successfully complete the part II examination.
In the early 1990s, the directors of the ABPN decided to administer the part I examination closer to the end of residency training. As a result, candidates can now sit for the examination in November (about 4 months after graduation) rather than waiting until the following spring (9 months after graduation). The first fall administration was held in 1994. To qualify for that examination, candidates had to have completed residency training after Oct. 1, 1993, but no later than June 30, 1994.
The subjects for this study were the 739 new candidates who sat for the fall 1994 part I examination. In 1994, 1,433 residents completed their fourth year of residency training in psychiatry (7). Of these, approximately 370 were in child and adolescent psychiatry residency programs. Hence, the candidates for the 1994 fall examination represented 52% of the "graduating class."
As part of the application process, the applicants signed a release statement agreeing to let the ABPN release information about examination results and examination scores, provided that such data were reported in the aggregate.
Part I and Part II Results
Of these 739 candidates, 603 (82%) passed part I on the first attempt. An additional 61 (8%) passed on the second attempt, 21 (3%) on the third attempt, eight (1%) on the fourth attempt, three (<1%) on the fifth attempt, four (<1%) on the sixth attempt, and four (<1%) on the seventh attempt. Thus, a total of 704 members of the cohort (95%) passed part I during our follow-up period, approximately 8 years. Twenty-one (3%) reattempted part I and failed, and 14 (2%) made no additional attempts during our follow-up period. (There is no limit on the number of times an applicant may take the part I examination.)
A total of 697 candidates (94% of the cohort) attempted part II at least once. On the first attempt, 466 (67%) passed. An additional 121 (17%) passed on their second attempt, and 36 (5%) passed on the third attempt. Eighteen candidates in this cohort failed part II on their third attempt and took the part I examination again; 15 passed, and three failed. Of the 15, 12 took the part II examination again, and two passed on what was in effect a fourth attempt, and two passed on the fifth attempt. Hence, 627 (90%) of those who attempted part II passed it, and 70 (10%) attempted and failed.
Relationship Between Part I and Part II Performance
Of the 597 candidates who passed part I on the first attempt and took part II during the follow-up period, 410 (69%) also passed part II on the first attempt, and 187 (31%) failed. Of the 100 candidates who failed part I on the first attempt and took part II, 56 (56%) passed part II on the first attempt, and 44 (44%) failed.
The results of a chi-square analysis were significant (χ2=6.21, df=1, p=0.02), indicating that those who passed part I on the first attempt were also more likely than those who failed to pass part II on the first attempt.
Approximately 8 years after their first attempt at the part I examination, 85% of the cohort (627 of 739) were certified, and 15% (112 of 739) were not.
ABPN currently offers certificates in six subspecialties to board-certified psychiatrists who meet the requirements. The subspecialties are child and adolescent psychiatry, geriatric psychiatry, addiction psychiatry, forensic psychiatry, clinical neurophysiology, and pain medicine. Although the last two are open to psychiatrists, most of the candidates in those two subspecialties are neurologists or child neurologists.
Of the 627 members of the cohort who were certified in the follow-up period, 199 (32%) were also certified in one subspecialty, and 29 (5%) were certified in two subspecialties, for a total of 257 subspecialty certificates. Of these, 123 certificates were in child and adolescent psychiatry, 47 in addiction psychiatry, 44 in forensic psychiatry, and 43 in geriatric psychiatry.
The results of this study suggest that most recent graduates of residency training programs who attempt the ABPN process are likely to become board certified, and the majority of subjects in our cohort did so by passing both the part I and part II examinations on the first attempt. Those who passed part I on the first attempt were more likely than those who failed to pass part II on the first attempt. In addition, a substantial percentage of the cohort went on to achieve subspecialty certification.
From 1995 through 1999, the pass rates for all first-time takers were 69% for part I and 62% for part II, compared to 82% and 67%, respectively, for this cohort. These findings suggest that it may be advantageous to begin the certification process as soon as possible after the completion of residency training.
Future research should explore in more detail the relationship between performance on the ABPN examinations and other variables, such as years out of training, prior performance (e.g., United States Medical Licensing Examination scores, in-training examination scores, ratings of residency performance), and training program characteristics.
Because this cohort did not include all of the physicians who completed residency training and were eligible to sit for the fall 1994 part I examination, these data do not necessarily reflect how a full cohort would perform. It could be hypothesized, for example, that more confident candidates would opt to begin the certification process as soon as possible.
The percentage of diplomates obtaining subspecialty certification in geriatric, addiction, and forensic psychiatry may have been greater in this cohort than will be the case with subsequent groups. Because these were new subspecialties (the first examinations were administered in 1991, 1993, and 1994, respectively), subspecialty training in residency programs accredited by the Accreditation Council for Graduate Medical Education was not required for the first 5 years in which the certificates were issued. Applicants could qualify by meeting practice requirements in the subspecialty, as did the great majority of candidates during that time.
Received Jan. 3, 2002; revision received Aug. 29, 2002; accepted Sept. 19, 2002. From the American Board of Psychiatry and Neurology, Inc. Address reprint requests to Dr. Juul, American Board of Psychiatry and Neurology, Inc., Suite 335, 500 Lake Cook Rd., Deerfield, IL 60015; firstname.lastname@example.org (e-mail).