Electronic medical records and interation design

A recent study on electronic medical records (EMRs) found that they may not fulfill the promise of lowered health-care costs. This  study, and the reaction to it, illustrates much of what is wrong with technology studies, and the unintended social effects of technology itself.

Many technology studies have false ideas of how web and interaction designers actually work. We collectively tend to think of technology as a “fix” that “automagically” eliminates  “waste,” even if this is not the intent of the designers themselves (which it frequently isn’t).  But as this study points out, there are far more subtle and nuanced issues relating to technology. Specifically, technology makes it easier to do some things. Is it any surprise we end up expecting more things to be done?

Let me illustrate with EMRs.

Image licensed under Creative Commons to MC4 Army on Flickr

Researchers from Harvard Medical School found that the use of electronic medical records (EMRs) is actually correlated with a higher number of diagnostic tests, such as MRIs, which in turn implies higher — not lower — health-care costs.

The authors suggest:

These findings raise the possibility that, as currently implemented, electronic access does not decrease test ordering in the office setting and may even increase it, possibly because of system features that are enticements to ordering.

This study was a quantitative analysis of medical records so did not offer any insight into why there are more tests ordered with EMRs. The authors can only speculate that the easy availability of imaging results translates into more tests being ordered. The “enticements” to order more tests could be built into the EMR systems themselves.

Anyone who has worked in interaction design will tell you that “enticement” is precisely the kind of emotion they want their users to feel. Take, for example, Stephen Anderson’s research on on “emotional design.” Anderson argues that web and application design should be “seductive” to really be successful. Trevor van Gorp also argues that designers should be aspiring to connect “affectively” with their users, and to tap into deeply held emotional experiences.

Both Anderson and van Gorp have written and spoken extensively to the user experience designer. Their ideas are au courant in the web and application design community. It is likely that some members of that community have read van Gorp’s book or use Anderson’s psychology inspired “mental note” cards in their design practice. It is just as likely that some of these people have designed the very EMR systems that strive to, surprise surprise,  “entice” physicians to order and view diagnostic tests.

Physicians are responding to a design philosophy, which is to extract from users a deep engagement. “Good” interaction design is usable, but also engaging. Instead of boring users, contemporary web and application designers are “seducing” them. Indeed, good interaction design, according to industry leader the Nielsen Norman Group, includes the principle of “explorable interfaces.” How is it any surprise at all that physicians are “exploring the interface” by ordering more tests? Good systems are designed to entice them to do exactly that.

In their story covering the Harvard study’s findings, the New York Times reports that other researchers disagree with the conclusions. The Times reports:

Dr. David J. Brailer, who was the national coordinator for health information technology in the administration of George W. Bush, said he was unconvinced by the study’s conclusions because they were based on a correlation in the data and were not the result of a controlled test.

Dr. Brailer doubts the conclusions because he does not understand how design is currently practiced, nor does he have direct input into the design principles of EMRs. If EMRs are being designed according to current ideas, they are designed not to save money, as Dr. Brailer hopes, but to entice users to explore and be engaged. Dr. Brailer clings to his scientific method here, and rightly points out that correlation does not equal causation. However, the Harvard researchers are more in tune with current design practices.

I don’t mean to knock interaction designers. Heck, some of my best friends are interaction designers! No, really. What I’m saying here is that designers design to principles. A laudable principle is to “seduce” or emotionally affect the user. This principle creates great systems. But it results in more use of systems, not less. It should come as no surprise that imaging tests represent a “seduction” for physicians, who, like all scientists, are voracious consumers of “more data.”

This case study reminds me of how we so frequently miss the mark in understanding technology. We assume it will be “efficient,” without asking how it might actually work. Worse, we routinely ignore the normative shifts that come along with cheaper and easier labour. Take, for example, house-keeping technology. We believed that the vacuum cleaner, the washing machine and the dishwasher would lead to more leisure time. What it actually lead to was higher standards of cleanliness.

Will EMRs lead to “higher standards” of imaging desire among physicians? Perhaps they already have.