Abdesalam Soudi grew up in a small village in Morocco, where his family didn’t have access to a hospital or a pediatrician. As needs arose, his mother took care of him with herbal medicines. So he never imagined himself in the position he’s in today: an expert in the sociolinguistic challenges that arise during a medical interview and how the computer in that setting influences doctor-patient relations. For three years, Soudi was a language coordinator for international patients at UPMC; he’s now a PhD faculty member in linguistics at Pitt. He also codirects a master’s-level course in cultural competence in medical education with the chair of family medicine, Jeannette South-Paul (MD ’79), who is the UPMC Andrew W. Mathieson Professor of Family Medicine.

How did you get interested in electronic health records (EHRs)?

Initially I was interested in medical discourse in general—the language exchange between doctor and patient. But then, when I started transcribing the conversations, I couldn’t ignore the sound of typing in the exam room. So that became the topic of my dissertation.

What are the issues?

The needs of the patient and the computer overlap and sometimes clash. So there is a tradeoff in attending to one or the other. ... It’s like texting and driving. If your texting is going well, your driving is suffering and vice versa. Another problem: There is a disconnect that exists between the patient and the computer. The patient doesn’t have access to the computer, and the computer, of course, doesn’t have access to the patient. They are blind to each other, which leads to problems in turn-taking and coherence. The responsibility to coordinate this three-way interaction mostly falls on the physician.

What might doctors want to keep in mind?

There are times when facing the computer is very good. But that first minute really shapes the rest of the encounter. Shake the patient’s hand, then “greet the computer.” Maybe first spend face-to-face quality time [with the patient] before logging in.

What would improve the situation?

Doctors are trained how to use the software and not how to use it in the patient care context. [Further, when designing systems], we should think about where and with whom the EHRs are going to be used: in the E.D., orthopaedics, etc. And the systems should be sensitive to fostering the doctor-patient relationship. I don’t see my work as a criticism of doctors. What I am critical of is the context in which they have to work.