The results confirmed Shaffer's hunch: On average, clinicians ordered three more tests when using the opt-out version than the opt-in or recommended versions. Docs ordered more tests expert-recommended test using the recommended design than when using the opt-in design.


"IT experts and medical professionals should work together to design these systems to reach optimal performance, which results in the best care for patients," Shaffer said, in a press release describing her findings. "A wide variety of methods exist that could improve medical lab test ordering software and would ensure that only the most appropriate, relevant lab tests for patients are ordered while saving money in the long run."

As she told Healthcare IT News, "There's  huge assumption that we've got a really good handle on how we make decisions. People don't really seem to understand that there's a lot of these subconscious processes – how things are framed, the order things are presented, whether something's checked as a default or not – that has a huge impact on our behavior and goes outside our awareness."

She and her team were motivated by a supposition that "the use of defaults would have a powerful effect on behavior," she says. "This is a phenomenon that would occur with all human beings. In the supermarket, things are arranged a certain way. We take things on the right more than on the left. It's just a larger human processing issue."

Electronic health record and clinical decision support developers should consider employing psychologists when designing their systems, she says.

"Absolutely. There's such a rich literature about how we can use these little cues. That can make a big difference in behavior. so i think there's a huge market for this.

"These are pretty influential pieces of psychological architecture, if you will," she adds.

"It really requires some complex coordination with medical experts: 'OK, what of these things is most important, should we have a pre-selected checkmark here?' You need to have the right people. but an ideal software team would include an engineer, and end-user – a physician of some type – and a psychologist or human factors engineer to think about how things would be laid out."

Asked whether she's gotten any inquiries from EHR vendors interested in her research, Shaffer laughs.

"No, not at all! I don't know if they just don't know about it, or if they think, 'Hey, business is booming as is, why add this extra layer of hassle to my life?'"

[See also: Benchmarks: Changes are afoot for clinical and business intelligence]