August 13, 2012

New Thoughts on EHR Design
By Greg Goth
For The Record
Vol. 24 No. 15 P. 10

The results of a late 2011 workshop reveal interesting insights into how EHR components can become more patient friendly.

There may be no cliché more overused in business circles than “think outside the box.” But if ever it could be applied without embarrassment, it may be in relation to EHR design. A recent spate of workshops, conferences, and guidelines are publicizing the idea that check-mark boxes, procedural code fields, and bare-bones mandated interoperability regulations will not meet the needs of the many disparate audiences that require access to medical data.

“The key thing is to not give up,” says Lauren Zack, director of usability at athenahealth. “It was a little frustrating going to industry conferences and hearing, ‘This is just so complex. Healthcare is different; it’s just too complex.’ In my mind, it’s not more complex than trying to figure out how to let all of America file their taxes. Break it down into pieces and the goal is that the patient will be able to consume the information—and we have to help them do that.”

Zack is not speaking as a starry-eyed optimist. Having been on the frontlines of efforts to address the problem, she recognizes the widely disparate needs of different user groups that will require access to any given EHR as well as the entrenched cultural and economic forces that have created a siloed nightmare for patients trying to gain access to their data let alone well-designed presentations of them.

Standards Just the Starting Point
Foremost among the realizations the HIT community is learning as EHR design progresses from its elementary stylistic beginnings is the role government agencies, industry associations, and other stakeholders will play in facilitating consensus. Prime examples of such activities can be seen from the recent design workshops held by the National Institute of Standards and Technology (NIST) and the California HealthCare Foundation (CHCF).

Glen Moy, a senior program officer at the CHCF, says the efforts for improving user-centered design and meaningful use interoperability mandates offer the entire spectrum of those involved in EHRs a golden opportunity to achieve symmetry in improving pervasive information exchange. To that end, he helped organize and lead a November 2011 design workshop called Project Synapse in which representatives from leading EHR vendors, nontraditional information portals, and industry associations discussed how best to address crafting a patient-friendly continuity-of-care document (CCD).

“There are all these federal incentives to get providers to purchase and implement EHRs and incentives to get physicians to communicate and interact with patients using EHR components,” Moy says. “We could wait until everything is settled and interoperable and at that point you can focus on usability or design, but it seems like you lose the impetus of a lot of people and a huge opportunity. There have been complaints for years about the usability of EHRs, or the lack thereof.”

“In my experience with EHR design, I’ve found that EHRs generally don’t map well to the needs and mental models of the people who try to use them,” says Abbe Don, health team lead for design firm IDEO, which cosponsored the Synapse workshop with the CHCF. “Using an EHR system typically requires a very steep learning curve and hours of training and ongoing customer support. Medical teams’ workflows are disrupted, and unnecessary strain is put on doctor-patient encounters.”

Ultimately, according to Zack and Sally Okun, health data integrity manager of the Web community Patients Like Me, the original concept behind the CHCF workshop—designing a patient-friendly CCD—took a backseat to a more flexible conceptualization of how to present essential data elements to a wide variety of users.

“It was extremely helpful to have the session be facilitated by an outside design firm. That really opened up peoples’ eyes to the fact that even if you have a set of standards that are very restrictive like the current CCD, breaking the bounds of the regulated disciplined document … was really helpful to a lot of folks,” Zack says. “It was great watching people’s heads explode just a little bit: ‘That’s not a CCD.’ And once we got to ‘Can you imagine, let’s call it something else, and what could it be?’ was just a fantastic turning point.”

Yet exactly how flexible EHR designers can be may be constrained by the very standards that have evolved to ensure a common format. “Patients Like Me is not constrained by the EHR standards,” Okun says. “We can play around with new ideas and test the envelope a little bit. When you’re an EHR vendor, there are conventions you have to stay with before you can push out the boundaries of what you can offer to your clients.

“The charge for the IDEO workshop was trying to come up with a consumer/patient-oriented CCD, and the funny thing was we sort of came away feeling like it’s almost impossible to do that because the conventions of that document are geared so much toward the clinician and not the patient.”

Ironic Lessons?
Okun says the CHCF/IDEO workshop resulted in another revelation for Patients Like Me’s designers: Patient-friendly presentations are often as unusable for physicians as structured formats are for patients. For example, Patient Like Me’s graphics-oriented Doctor Visit Sheet, which allows patients to create a PDF chart that includes subjective observations on their general feelings of well-being as well as more objective measures such as weight and lab results, is not as useful for clinicians as it was envisioned to be, according to Okun.

“The more I used it in the field on a couple examples with clinicians, I could see how not useful that patient view was to them,” she says. “What they depended on was a quick look at data; quantifiable data is really important or else they kind of get caught in the weeds. There’s a balance we’re trying to strike there so the patient can take their data to their doctor but in a form the clinician can fully appreciate in that eight to 10 minutes they have with the patient. Our doctor visit sheet will likely get evaluated this year to be a little less patient focused and a little more clinician focused.”

Don says IDEO has noticed similar patient-physician dichotomy in its research into user-centered EHR design. “We learned that patients and vendors are well served by graphical elements that allow them to view health information over time and that physicians are more accustomed to seeing labs results in data tables,” she says. “They are looking for one element, such as bold type or red type, to call out only those values outside of the normal range.”

It must strike anyone with any familiarity with the EHR landscape that designing a one-size-fits-all presentation is an impossible task. However, according to Matt Quinn, a usability scientist at NIST, healthcare is far behind the curve in observing the principles of usability design and testing.

Quinn recently gave a presentation about improving the usability of HIT security applications to an audience of about 100 people. “Not a single person in that audience had a human factors background, a cognitive science background, industrial engineering background, design background—none of it,” he says. “The important point here is integration of the user-centered design process into the development process. It’s like baking a cake and forgetting the flour—it’s hard to put it in later. What we have are systems that were designed by engineers and now that they’re widely used—or trying to be widely used—we’re finding that going back and making the changes is very hard. It’s something you have to do in a systematic way.”

Quantifying User-Centered Design
Quinn, who previously worked for the Agency for Healthcare Research and Quality, says the agency’s experience with EHRs revealed two consistent shortcomings: a lack of formal application of user-centered design principles and a paucity of diversity in user testing.

“UCD [user-centered design] is not just having perspective users design stuff but a wheel of understanding the users’ needs, workflows, and environments; engaging the users in the design process; setting performance objectives; and finally testing and evaluating,” Quinn says. “We found this was going on in pieces and parts but not formally as it is in lots of other industries that have usable stuff. And there were also not enough of the right kind of experts in this field. The other big shortcoming was there was a lack of diversity in user testing. That means a representative set of users was not being engaged in the concept of designing the systems or in testing and evaluation.”

What’s more, he says, EHR designers need not venture far to find that diverse user cohort. “In many ways, the healthcare workforce, both clinical workers and administrative workers, including those on the back end of these systems, is very much representative of America if not even more elderly, more diverse, more English as a second language,” Quinn says. “So that’s a challenge in itself.”

NIST has released numerous guidelines on EHR usability issues and best practices, including the following:

• NISTIR 7741, NIST Guide to the Processes Approach for Improving the Usability of Electronic Health Records (www.nist.gov/manuscript-publication-search.cfm?pub_id=907313);

• NISTIR 7742, Customized Common Industry Format Template for Electronic Health Record Usability Testing (www.nist.gov/manuscript-publication-search.cfm?pub_id=907312); and

• NISTIR 7804, Technical Evaluation, Testing and Validation of the Usability of Electronic Health Records (www.nist.gov/manuscript-publication-search.cfm?pub_id=909701).

“We’re not prescribing how to design systems,” Quinn says. “What we’re saying is whatever you do, test it and record results in a standard way so you can communicate key human performance characteristics.”

He says there has been noticeable progress in vendor and professional association collaboration on the conceptual principles of user-centered design as EHR deployment has received more emphasis. However, it’s still too early to tell whether these partnerships are yielding the desired results. “[At the 2011 workshop], one of the things we really tried to pull out of them was what are areas where you can collaborate? We understand competitive pressures and we know every EHR interface doesn’t work the same and probably shouldn’t, but are there any examples of putting the accelerator and the brake pedal in consistent places? And the resounding response was no.”

Wild Card in Play
Yet Quinn, Zack, Okun, and others say the amount of cross-pollination is accelerating. Design methodologies may not be identical, but there are numerous similarities, including on-site interviews of various user communities, observation of technologies in use, asking selected users to keep journals of their experiences with certain technologies, and focusing on cutting-edge “extreme users.”

“[Extreme users] tend to highlight issues more quickly than people in the center of a typical bell curve,” Don says. “In the case of the Synapse project, that meant we interviewed patients who are very proactive in the ways they engage with the healthcare system, requiring them to connect with three or more healthcare professionals. We also interviewed ‘extreme’ physicians who are self-described mavericks. Many of them had designed their own workflows and their own practice-specific electronic medical records because they felt very strongly that commercially available products do not address their fundamental approach to patient care.”

The great wild card adding uncertainty to how quickly user-centered design principles will become more central to EHR development may be how quickly the consumer-centric model that dictates which technologies emerge victorious in other industries also envelops healthcare.

“In other industries, you have consumers who have become very good indicators of when products are usable and not,” Zack says. “For example, in the way the leading-edge consumers chose the iPhone over BlackBerry. The same thing will happen in healthcare IT, but there’s been a lag. The project with the CHCF was really interesting because it helped folks who don’t see the patient as the end user to think more about putting them in the center.”

Okun believes the traditionally defined EHR will not be the vanguard platform for patient-centered technology. Platforms such as Ginger.io’s smartphone-based passive data reporting application will engender a new era of physician-patient collaboration about which technologies best enhance their relationship, she says, adding that the best applications will be a smorgasbord of EHRs, websites such as Patients Like Me, and reporting technologies.

“The way I talk about it when I do presentations is presume your patient’s online, get some understanding of the kinds of sites you think are good for them to use, and that will help them have better outcomes,” Okun says.

Moy is pleased with the way the Synapse project engendered a spirit of collaboration and innovation around design, even if it didn’t result in a user-friendly CCD. “The goal was never to produce a ‘product,’” he says. “It was to try and introduce the notion of user-centered design into people’s consciousness and then find different avenues to continue to show what better design looks like and how products or systems can better engage patients.”

— Greg Goth is a freelance journalist from Oakville, Connecticut, specializing in technology and healthcare policy issues.