Although we are accustomed to drawing the ire of some readers, the Message from the Editor on the neurological exam generated a particularly large number of responses, both formal and informal.1 The more formal responses are now included in the correspondence section of this issue of Annals of Neurology, and we recommend them to you.2–4 We found them insightful and thought provoking, and we agree with some of the points raised.
First, we want to reiterate that we love the neurological examination nearly as much as our own children, and like them, we recognize our healthy bias to exaggerate strengths and gloss over weaknesses. Our suggestion for more evidence is really a request to give the examination more opportunities to shine and to encourage its growth.
Second, we disagree that studies of the neurological examination are necessarily poor. Yes, we have seen many poorly designed studies of specific elements of the exam—and more since the publication of the Message—but there are also many examples of elegant studies that teach us a lot about the examination.
Third, we disagree that studies attempting to find new therapies are necessarily more important. We agree that studies of treatments may seem more exciting, and they are certainly necessary, but neglecting a subject that consumes so much time and energy is also inappropriate. Just as we talk about moving toward a future in which every patient is offered participation in a research study, could every student be offered participation in a study about how we teach the examination? Costs and risks of studies of the examination are also much lower, and often the research question can be framed to produce an important advance, regardless of the result.
Fourth, we disagree that the neurological examination cannot be studied appropriately using the tools of clinical research. No question, the examination varies by setting, provider, and patient; components of the examination must be considered part of a much larger whole (including history, personality, and other parts of the exam); and gold standards are never perfect. Nonetheless, examination findings are data that we write down and communicate to others. As such, they need to be reproducible and meaningful. If a test is irreproducible in the hands of five master clinical neurologists evaluating the same patient, it's not a helpful test at least in that patient and probably many others. If five medical students can't agree on a finding, perhaps we need a better way to teach that part of the exam. We know this is a big problem for clinical trials; so why do we tolerate it in everyday care? Yes, other approaches to understanding the exam are more than welcomed, particularly given the tacit nature with which it is applied and interpreted, but there is plenty we can do with the tools of clinical research.
Finally, we agree with the notion that not everything we do requires evidence from a controlled study. The examination has been honed through years of use and repeated practice. We can all relate compelling stories about when it defied expensive testing and saved a patient. Also, it is a very inexpensive and valuable alternative to laboratory and imaging tests. The stance should be to respect it unless there is evidence otherwise, distinct from our approach to therapies.
We want to thank our critics for the great dialogue. This gave us an opportunity to question our own stance on the neurological examination and to reaffirm both our love for it and our love for data.
S. Claiborne Johnston, MD, PhD, Stephen L. Hauser, MD Editors