My previous blog post, “The Person Behind the Numbers,” seems to have resonated with a number of readers, several of whom have responded with supportive comments. It seems I am not the only one in this field who occasionally struggles with the challenge of maintaining one’s focus on the patient while writing about data-driven medical topics. Moreover, now that I have dipped my pen into the “remember the patient” inkwell, this issue seems to arise wherever I turn.
As I medical writer I am often called upon to develop questions and answers on a variety of topics. Sometimes these Q&A documents take the form of messaging guides for physicians, company executives, or other spokespersons who may need to prepare for media interviews. On other occasions I may be asked to insert a few self-assessment questions and answers into a continuing education course or a sales training module. In preparation for such an assignment, I recently spent some time reviewing the National Board of Medical Examiners’ guide to Constructing Written Test Questions for the Medical and Clinical Sciences (available at http://www.nbme.org/publications/item-writing-manual.html), and was struck by the following excerpt:
Questions asking for recall of isolated facts often begin by citing a disease and then asking what patient findings are expected. These questions are structured similarly to most textbooks; the examinee could look up the disease and find the answer in a single paragraph. The flaw with these items is that they seem clinically backward. Patients rarely tell their physician what disease they have and then ask the physician what their signs and symptoms are.
How true, and how obvious! And yet, how many of us fall into the trap of structuring a “clinically backward” question, as described above? Is it considered “good medicine” to include such a question in a document, even if the intended reader is a pharmaceutical sales representative, as opposed to an aspiring physician? While the writer may work diligently to ensure that the answer to such a question is medically and scientifically correct, the question itself may not reflect clinical reality.
On the other hand, I found other excerpts in NBME writer’s guide that, while presumably valid for development of cogent and pertinent questions for licensing exams, seem (to my sensibility, at least) to dehumanize the patient. For example:
Use of Real Patients. We believe it is generally better not to base multiple-choice questions on “real patients,” particularly for tests aimed at students. As a general rule, real patients are too complicated, and the elements that are complicated are not necessarily those that are important for assessment. As noted earlier, we do include window dressing (ie, incidental findings), but do not include “red herrings” (ie, information that is intended to lead examinees away from the correct answer). Unfortunately, real patients often have “red herrings” among their findings.
Use of Patient’s or Physician’s Own Words. We generally do not believe it is useful to include the patient’s own words, particularly if the examinee task is to interpret nuances of language that might be affected by tone. On the other hand, it may be useful to ask the examinee to select the most appropriate physician response to a patient by asking the examinee to choose among options phrased as open-ended, closed, or leading questions.
Yes, patients are complicated, even when they’re telling the truth about their health status and behaviors (and the NBME has more to say about patients who don’t tell the truth!). Is this why we so often create content without giving sufficient thought to how the information might affect actual patients? Isn’t it sometimes easier to write about a complex health-related topic without letting the patient get in the way? I wouldn’t presume to question the NBME’s wisdom with regard to constructing test questions; the Board has far more experience in this area than I. But by counseling against using “real patients” or patients’ “real words,” might they be telling us it’s OK to make the patient a little less “human”?
I’ll certainly be thinking about this issue a lot over the coming weeks, and I encourage you to do the same, and to send me your comments.