Transparent charts and opening the problem list to patients

The other day my electronic health record (EHR) introduced a new feature that startled me. The entire problem list was included on the after-visit summary, the printout that is given to the patient at the end of the visit.

I love the after-visit summary. It documents key features of the outpatient visit and provides a written record of medications changed, referrals made and immunizations given. It also allows the physician to enter free-text instructions (“Please bring all your medicines with you to your next appointment”) or encouragement (“Congratulations on your 4-lb weight loss. Keep up the good work!”). But seeing the whole problem list included has raised a set of questions I have been thinking about since I was a medical student: Whom does the chart belong to, in whose language or idiom should notations be made, and what level of control should the patient have over the content of the physician's notes?

As a resident, I was taught that the chart belongs to the patient. But as a practical matter, it doesn't. The physician, the practice or the medical center maintains physical control of the paper chart at all times. While a patient can request to see his chart, most patients don't know how to make the request, and I suspect that many who do are unsure how their physicians would feel about it. From time to time in paper-chart days, a patient would take the chart from the tray on the door. I'll admit that my heart would drop when I came into an exam room and found a patient leafing through his record.

Why are we made uncomfortable by patients reading their charts? Perhaps the most benign reason is that many of us use our notes as reminders to ourselves and document our impressions and intentions in a telegraphic shorthand. The notes are meaningful to us and intelligible to other clinicians but may require lots of explanation to make sense to a nonclinician. With time at a premium, the prospect of translating a note into simple English can seem burdensome. Our notes may record events or diagnoses, although accurate, that patients regard as unflattering or pejorative.

In addition, because chart notes memorialize the thinking of the clinician, they often record an elaborate differential diagnosis, including unlikely but dire considerations that the physician wants to keep track of. Doctors develop their own approaches to warning patients of worrisome diagnostic possibilities and to sharing bad news; most times we don't want our raw thoughts communicated to patients almost as they occur.

Of course, there are less legitimate reasons that we may prefer not to share the chart with patients. When I was a medical student, an intern was disciplined for writing a history and physical that began “Mr. R. is a 68-year-old ...” followed by an elaborate, multicolored cartoon of a turkey. Regarding the chart as our own, private space, sometimes we are sloppy, even unprofessional, in our characterizations of patients and the experiences they relate to us.

Then there is the whole issue of language. The structures that we use to describe the emergence of an illness or to document the course of a disease are not forms that the patient or her family would use to tell the same story. The chart and the 2 narrative conventions that comprise it, the history and physical and the progress note, have their own vernacular, much of which sounds odd to patients.

When I was teaching the “doctoring” course at the University of California, San Francisco, I would take a history and perform a limited physical exam in front of the first-year class. The subsequent week the patient would return and the class would compare their writeups to those of several faculty members. One year while I was discussing the “chief complaint,” the patient broke in and said, “I wasn't complaining. You asked what was bothering me and I told you.”

Worse, some of our language sounds borrowed from a courtroom or a police report: “Mr. T. denies exertional shortness of breath but admits regular 3-pillow orthopnea.” Even the chief complaint is contested: Should it be expressed in the patient's own words, as many of us were taught, or can we substitute medical terminology? Sometimes we even decide that the symptoms the patient says brought him to the clinic (“sore ankle”) are not important enough to be the chief complaint and swap in the “real problem” (severe microcytic anemia).

From time to time, patients attempt to control chart entries. In the early years of the epidemic, people with HIV regularly requested that information about their diagnosis not be included in the chart, usually because of the implications for subsequent insurability. I have had patients request, or sometimes demand, that information about extramarital or same-sex sexual contact be excluded. I even had a patient who didn't want “constipation” listed on his problem list.

Increasingly, the chart is becoming a shared workspace. As clinical teams, both actual and virtual, become more prevalent, it is likely that more authors will be making entries in the chart. As a result, there will be less segregation of content according to the background of the author (progress notes, nursing notes and so on). And who is a more central member of the team than the patient?

In a study published in 2012 in Annals of Internal Medicine, Tom L. Delbanco, MD, MACP, and colleagues described an experiment in which patients were given access to outpatient notes written by their primary care physicians. This was a popular innovation; essentially all the patients wanted to continue with open notes after the end of the experimental period. And of interest, none of the physicians who had volunteered to participate in the pilot chose to stop. There were significant points of difference, however. A third of the patients believed that they should be able to approve the contents of their notes, while the overwhelming majority of the physicians disagreed.

That first day in clinic when the problem list was included in the after-visit summary, I had a patient ask me about “dyspnea” when it appeared as one of her problems. A trickier moment arose with a patient whose first problem was listed as “atypical somatoform disorder.” My EHR has a “rename” function that I use all the time. It allows me to substitute a more natural or complete name for a problem instead of the one I have chosen from the predefined menu of “available” problem names. There is also a “share with patient” check box; unchecking the box results in suppression of the problem in views available to the patient, including the after-visit summary. I have thought long and hard about this box.

I began my internship on the medicine service at San Francisco General Hospital, and my first attending physician was the chief of the service, Hibbard E. Williams, MD, FACP. In orienting us, Dr. Williams explained that he liked to round at the bedside. Accordingly, a responsibility of the interns was to explain to newly admitted patients that the physician team would come in the morning and an intern or medical student would present the problem that had brought them to the hospital. Dr. Williams asked us to invite the patients to listen as the history was recounted and to feel free to ask for explanations of unfamiliar terms and to correct misstatements.

I had not been long at San Francisco General, but a day or two was plenty to convince me that many of my admissions would have alcohol or drug use as the root of their medical problems. So I asked Dr. Williams if we were always to lay out our diagnostic thinking in front of the patient. He allowed that there were a few circumstances in which we would not explore the full differential diagnosis at the bedside but then asked me about the situations I was envisioning.

I said, “What if I think that the patient is an alcoholic?”

He looked at me for a moment and then said, “If the patient has been on your service for 6 hours and doesn't know that you think he is an alcoholic, you're not doing your job.”

It seems that I am still learning this lesson.