Electronic health records (EHRs) have been around for a few years. You've probably seen them, or at least seen a doctor access them: at my doctor's office every exam room has a PC, which the doctors and nurses can use to access my file, send prescription orders to the pharmacy, and probably let the NSA know how healthy I am. (It would be amazing if the NSA was not collecting EHR data, come to think of it.)

Northwestern professor Enid Montague and University of Wisconsin graduate student Onur Asan recently published a study of how interaction with EHRs can affect the way doctors and patients communicate. They videotaped 100 patients visiting 10 doctors, and monitored how much doctors and patients interacted with each other, versus how much either of them looked at the computer. The displays were "located on the wall between patient and the physician and facing toward them." I e-mailed Dr. Montague about the setup, and he kindly shared a picture of the setup. The display was mounted above a desk on an articulated arm, the sort you can pull out and swivel. The mouse was on the desk, and the keyboard on a pull-out tray underneath.

In an article for Northwestern Medical School, Montague says:

We found that physician–patient eye-gaze patterns are different during a visit in which electronic health records versus a paper-chart visit are used. Not only does the doctor spend less time looking at the patient, the patient also almost always looks at the computer screen, whether or not the patient can see or understand what is on the screen.

Doctors spent about a third of their visits looking at EHRs rather than patients (they spent about 8% of their time looking at paper charts, in contrast), which meant less attention to nonverbal cues and body language. This is important because, as Montague and Asan explain in the (firewalled) article (free abstract here):

posture and body language can provide indications of comfort or satisfaction, facial expressions can provide more objective assessments of satisfaction or emotional state, and eye gaze can illustrate attention to people or other artifacts.

In other words, EHRs tend to distract doctors from paying attention to patients, and picking up on things that they might have noticed when working with paper. Further, patients tend to take their attentional cues from doctors: as they put it, "Physician gaze largely influences patient gaze in the primary encounter."

But these effects seem to me more an artifact of the technology, and the ergonomics of interacting with PCs, than a necessary consequence of records becoming "electronic." As anyone who's worked with students in a computer lab and with tablets can testify, having to talk to students around monitors that are set up to block their view of the front of the room literally offers more obstacles than engaging students whose screens are on their desks. So it would be interesting to see how tablets versus desktop machines affect doctor-patient interactions.

Another obvious thing to investigate would be views of information that are meant to be shared by doctor and patient– in effect, visualizations that are meant to be explained by the doctor. Rather than being things the patient can't understand, and which subtly reinforce the authority of doctors by highlighting patients' ignorance, such interfaces could be useful prompts for more interaction. But whatever design improvements come, they should be focused on the doctor:

any intervention to increase eye contact, or EHR information sharing, will likely need to be targeted to the physician. These findings illustrate the importance of designing work systems that allow and encourage physicians to be patient centered. [emphasis mine]

The findings also echo the work that Richard Harper did in the late 1980s on the ways automating police note-taking and reporting could negatively affect interactions between police and victims or witnesses. Lots of early police automated systems were essentially digital forms, which you can imagine would be pretty useful in streamlining paperwork, reducing bureaucracy, etc.;' but it turns out that for some victims and witnesses (who might have just gone through something pretty traumatic), having the officer typing or trying to figure out what field they should be entering information into was very off-putting. For police, the systems meant more attention on screens and fields, and less time listening to stories, watching to see if someone seemed especially nervous of suspicious, etc.. Reducing police investigations to automated note-taking didn't do anyone any good.

(h/t to Annie Murphy Paul)