Psychiatric News
Letters to the Editor

April 2, 1999

Urgent Reform Needed in ABPN Exam Process

By Diana Dell, M.D.
Member-in-Training Trustee

With the release of the Part 1 written exam scores in mid-January and the administration of the Part II oral exam in late January, I have had a lot of communication with disgruntled APA members-in-training (MITs) and early career psychiatrists (ECPs) about "the boards."

Most writers and callers are outraged with the process. One clever fellow had gone to the Internet and found a posting showing that psychiatry had the lowest pass rate on the American Board of Medical Specialties list in 1997. I checked that Web site myself (; search for "board certification"), and he was right. In 1997 the American Board of Psychiatry and Neurology (ABPN) reported that for the "last five years" psychiatry had a pass rate of 60.7 percent. I researched the rest of that 51-page document very carefully-a pass rate of 60.7 percent was clearly the lowest rate on the list. That is pretty outrageous.

If you are a frequent reader of this column, you know that psychiatry is a second career for me. I started out in ob-gyn, in which I am both certified and recertified. The certification process is similar to psychiatry's-a written examination (taken during the last week of residency) and an oral exam (during second year in practice). I remember taking a very difficult written examination, but it seemed quite clinically relevant. The material I needed to know to practice ob-gyn safely was tested by that instrument.

That is not how I felt after writing my psychiatric exam this year. At the end of the day, I remember thinking that no matter what my score ended up being, it would not say much about my capacity to practice psychiatry safely or in any way differentiate me from other psychiatrists who chose not to take, or retake, the exam.

Clearly, the role for board certification has changed so that now it is more than just professional pride and identity: being board certified is an economic requirement for new graduates. Psychiatry training now has three stages: internship, residency, and board certification.

MIT and ECP leadership at APA has been working on several issues with the ABPN. At the top of our list has been the publication of program-specific pass rates for both Part I and Part II of the examination. We contend that trainees invest four to five years of their lives in programs and they have the right to know whether they are getting what they need to pass the certification exams. To date, the ABPN has refused, saying that training directors are opposed to it. From the training directors, I have heard four primary arguments:

  1. The 60.7 percent pass rate doesn't tell the whole story. That number doesn't differentiate between first-time takers and people who fail repeatedly; there is no indication about whether it refers to Part I or Part II, and so on.

    Answer: My point exactly. Other specialties publish those data-on the Internet. If the data are not available for psychiatry, then why not? One thing is certain: without clarifying the data, a blanket 60.7 percent pass rate does not enhance the image of psychiatry or psychiatrists. We need an open discussion about not only the pass rates, but also how the data will be presented in public formats.

  2. The pass rate in "my" program is much higher.

    Answer: But no program director I've talked to can tell me what the pass rate is for his or her program. The information comes to the program director in a form that is difficult to decipher-a seven-year aggregate designed to "protect privacy." The ABPN needs to provide pass-rate information to program directors in a clear way that can tell them where improvement is needed. Also, the information needs to be available to current as well as prospective trainees. If a medical school graduate interviews at two programs and likes both of them equally well, would he or she be more likely to choose the program with a higher pass rate? Probably so. Would that knowledge drive programs to improve? Probably so.

  3. We don't want to end up just training people to pass the boards.

    Answer: Why not? If the boards are the gold standard and essential to modern practice, shouldn't all programs be training people to pass the boards-literally and figuratively? If the boards really measure crucial clinical skills, we should undertake an effort to bring all practicing psychiatrists to the necessary level of skill to pass the boards. If the boards don't measure crucial clinical skills, which is supposed to be their raison d'Ítre, then they should be improved until they do measure them.

  4. Some programs have such poor pass rates that they would end up closing down.

    Answer: Yes, that is true.

Trainees have no leverage in this process. We need certification, and the ABPN "owns it." That ownership entitles the ABPN to construct the exam as its chooses, to schedule it when it chooses, to pick a pass rate, to set the price for its services, and to do what it wants with its data. And I don't think that the ABPN "answers" to anyone. Although it is a member of the American Board of Medical Specialties, that group has no capacity to "police" the activities of its members.

Training directors may have a little more leverage. If they are teaching the right stuff and they learn that their trainees do not have adequate pass rates, perhaps they could then exert enough pressure on the ABPN to make appropriate changes-whether those changes involve scheduling Part I to coincide with the end of residency, working toward a pass rate compatible with that of other specialties, or helping programs understand what information is critical for residents to attain.

Moreover, there is real potential for a conflict of interest. The American Board of Family Practice had the highest pass rate for the same reporting period at 91.8 percent. For every 1,000 candidates who pass, there are 89 candidates who fail and may choose to take the examination again. When the pass rate in psychiatry is 60.7 percent, 1,000 passing candidates would have come from a pool of 1,648. If all of the 648 failing candidates take the test again, even at the reduced fee offered by APBN, it generates more than a quarter-million dollars.

Is reform desirable? Yes, I think it is, but I don't know how to get ABPN's attention. Maybe this will help.