AMA Journal of Ethics^®

Medical malpractice policies have become so prohibitively expensive, and threats of malpractice suits so numerous, that physicians are being driven out of practice.¹ Attempts at widespread tort reform have been unsuccessful. The National Board of Medical Examiners (NBME) and Federation of State Medical Boards (FSMB) are approaching the problem from a different perspective. They want to limit malpractice litigation and, therefore, liability premiums by eliminating potentially problematic physicians before they've had a chance to practice. For several years, the NBME and FSMB have considered adding a clinical skills exam to the United States Medical Licensing Examination (USMLE) as a means of rooting out unqualified physicians-to-be. This year they announced their decision to move forward with the proposal.

The clinical skills exam will be administered to US medical students concurrent with Step 2 of the USMLE, usually taken between the third and fourth years or during the fourth year of medical school. The clinical skills exam is designed to test medical students' performance in tasks that make up the typical physician's day. In 10 to 12 thirty-minute encounters with standardized patients, students will be required to take a history, complete a physical exam, make diagnoses, counsel patients, and share their impressions on a case. Trained physician reviewers will evaluate the students' performance. To pass, students must demonstrate satisfactory clinical and communication skills. Exam administrators expect that 5-7 percent of students will be unable to pass the exam on first-attempt. One to 2 percent of students will fail even after repeated attempts. The NBME and FSMB believe that eliminating this cohort of students before they become practicing physicians will at least reduce the number of physicians that will subsequently come before state medical boards for consideration of malpractice.

A number of groups have voiced concerns about the addition of the clinical skills exam to the USMLE. The majority of these concerns have concentrated on the financial burden that the additional exam would place on the already strained budgets of medical students. The potential costs associated with the exam are enormous and merit some attention here. While an examination fee has not yet been determined, administrative costs alone are estimated at $950 per exam. Furthermore, testing centers are being established in only 6 or 7 cities nationwide, necessitating indirect expenses for travel and accommodations. Perhaps most disturbing is that the 250-500 students annually who are not expected to pass the exam after repeated attempts will still face an average educational loan debt of $100,000² and not have a license to practice medicine. These financial concerns are indeed valid, and certainly require attention. A larger question to consider, however, is whether the exam can actually accomplish what it hopes to accomplish.

The NBME and FSMB are motivated by a desire to improve overall patient-physician communication by recognizing and removing candidates who lack these skills. In the midst of what many are calling a malpractice crisis, this is an admirable goal. Insufficient evidence exists, however, to support the claim that the exam can accomplish this goal. One consideration is whether the test itself can detect the kind of skills that patients feel are lacking when they decide to sue. Errors happen in medicine, independent of a physician's excellent clinical skills. While preventing mistakes is beneficial, studies suggest that what physicians lack is the knowledge of how to handle the situation after the error has occurred. Physicians need to know how to deal with patients when something unexpected happens such that neither physician nor patient can predict how the situation will play out. A scenario in which students roleplay doctors with standardized patients who are acting cannot possibly capture the very real, spontaneous emotions and responses that accompany learning about an adverse outcome. When something goes wrong, patients, in general, want 3 things from their physician: (1) an explanation, (2) an acknowledgement or apology, and (3) reassurance that corrective measures are being taken.³ These communication skills are not assessed in the patient-physician encounters that are being simulated in this exam.

Another potentially troubling aspect of this exam is the message that it might be sending to those who pass. A few years ago, I was good enough at multivariable calculus to pass an exam. Today, I wouldn't even know where to start. The point is that the practice of medicine is more complex than riding a bike: without repeatedly using and refining the skills required to successfully interact with patients, these skills will be lost. To send the message that passing a clinical skills exam means that a student has sufficient communication skills to deal with patients before the student has completed undergraduate medical training denies the importance of emphasizing these skills throughout graduate medical training and beyond. One could argue that a physician is more strongly influenced by the way he or she sees situations handled during residency and fellowship than in undergraduate training. In fact, the recognition that certain specialties have unique communication concerns supports the idea that what students learn post-medical school is critical in determining how they deal with patients.⁴ Setting a standard where students who ultimately fail this exam are denied the opportunity to improve their skills throughout graduate training and students who pass may be satisfied that they don't need further improvement seems counterproductive to the maximization of patient-physician communication.

The clinical skills exam assigns equal weight to assessing the clinical and communications skills of students. If medical students, who pay constantly increasing tuition in medical education and leave medical schools with an average $100,000 in debt, are graduating without the skills to conduct a history and physical and generate a differential diagnosis, then stricter standards should be developed for accreditation of medical schools. Arguably, these skills should also be assessed in some way before a student graduates. The art of doctoring, however, means more than going through these motions. It is something that must be continuously learned—the promise of producing better physicians by standardized elimination of the weakest is simply a promise that cannot be kept.