The National Board of Medical Examiners (NBME), on behalf of the Federation of State Medical Boards (FSMB), is poised to introduce an additional hurdle to the sequence of USMLE steps required of all our graduates who desire to be licensed to practice medicine. That hurdle is a very ambitious, daylong Clinical Skills Examination (CSE), during which students will encounter 10 to 12 standardized patients chosen to illustrate a wide range of common problems seen in ambulatory settings.

I must confess, when I first heard that the NBME was seriously considering this move I expressed considerable reservations about the need for such an exam, as well as its practicality. As for the need, it seemed to me that we already had a perfectly satisfactory way to assess our students’ clinical skills — namely, the aggregate evaluations prepared by multiple faculty members who observe students repeatedly over the years as they exercise their developing skills on a variety of clinical rotations. Why, with all the available first-hand data, did we need yet another method to convince ourselves that a given student had acquired the skills necessary to function as a novice clinician?

Ah, but there’s the rub. On closer reflection, the need is not to convince ourselves; the need is to convince the public or, more precisely, the public’s designated agents — the licensing authorities who are charged with ensuring that doctors are qualified to deliver safe and effective care. The sad fact is that far too many practitioners get into trouble with licensing authorities because of gross deficiencies in the performance of basic clinical skills. It is the perceived need to intercept nascent physicians who lack these skills before they are licensed that generated the demand for the CSE. The intent is to offer those students who perform poorly on the CSE an opportunity for remediation while they are still in the formal phase of their education — or to deny licensure to those, presumably very few, who cannot meet expectations despite efforts to remediate.

As a profession dedicated to self-discipline and committed to acting in the public interest, we simply cannot ignore the evidence that current methods for identifying students who are not yet prepared to deal effectively with patients are falling short. Which is not to say, ipso facto, that a CSE can do the job any better. Indeed, there is no way of knowing that, certainly not prospectively.

What we do know is that a well-constructed CSE can, within the limits of psychometric certainty, evaluate some clinical skills that are not easily assessed with multiple-choice exams: namely the “basic” skills of establishing rapport, taking a history, conducting a targeted physical examination, and, not least, communicating effectively with patients. A host of other critically important clinical skills are admittedly beyond the scope of current CSE methodology and can only be evaluated by faculty; examples include the consistent display of professionalism, the skillful performance of requisite procedures, the judicious use of consultants, the appropriate deployment of community resources, and the ability to deal effectively with patients from diverse cultures. These limitations notwithstanding, the well-validated CSE planned by the NBME would add a needed measure of assurance to the public that our graduates are, indeed, prepared to meet their obligations as physicians.

So, the need may be there, but what about the practicality of the planned CSE? Administering the exam to some 25,000 students annually (including all applicants for Educational Commission for Foreign Medical Graduates certification, for whom a type of CSE is already required) will be a Herculean task. But the logistics have been thoughtfully considered and the task appears feasible.

The stickiest question remaining is the cost. Even with the large number of examinees to share the annual operating expenses, the cost per examinee is projected to be just shy of $1,000. On first hearing this number, I thought there must be a simpler, less expensive way. How about relying on each medical school to identify students with doubtful clinical skills, and then targeting the CSE on this presumably small subset? Here again, closer examination convinced me that a high-stakes licensing examination would not be credible, either to examinees or the public, unless all aspiring practitioners had to clear the same hurdles. Well then, how about a less expensive test of clinical skills? Unfortunately, there is none that meets the level of psychometric reliability required for deciding who can practice medicine.

Given their critical importance, a convincing means for assuring the public that licensed physicians possess basic clinical and communication skills is unarguably justified. Now that licensing authorities have deemed the MD degree alone no longer sufficient for this purpose, we must redouble our efforts to prepare all students to meet the new CSE standard. The AAMC, for its part, will continue to work with the NBME and the FSMB to minimize the financial burden of licensure on our students, who are already reeling under a mountain of debt.

Reprinted with permission from the December 2002 issue of AAMC Reporter, Copyright ©2002, Association of American Medical Colleges;