EHRs: Getting to Meaningful Usability
 |
"With many EHRs, the marketing message is a much different story from what the reality [of the project] is," says Robert Schumacher, PhD, managing director of User Centric, a Chicago-based company that conducts usability testing on EHRs.
Usability refers to the way people interact with different products and devices, guided by the psychological discipline of human factors. The widely cited Geneva-based International Standards Organization's definition labels usability as, "the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use." In the case of EHRs, Schumacher explains, key usability criteria include intuitive interfaces, clearly represented information and functionality that minimizes errors and enhances efficiency.
Although EHRs commonly miss the mark on these standards now, Schumacher expects that to change as more users adopt these systems and demand a better experience, and because human factors research—well-established and widely applied in fields such as aviation—will bring EHRs up to speed. It's a question of when these systems will be made more user-friendly, not how to do so, he says.
What's at stake?
The meaningful use requirements in the HITECH Act operate on the premise that EHRs will make physicians more efficient, ensure more accurate documentation and ultimately improve the quality of patient care—and there are many (eventual) EHR success stories. In a number of settings, however, inadequately designed EHRs have done the opposite.Because patients have to be fully registered in the system before orders can be placed, in the ICU, medications meant for other patients might on occasion be "borrowed" for urgent use, Scanlon says, or physicians will verbally request that pharmacists release the medications without an actual electronic order in hand. The alternative, which has been reported in medical literature, according to Scanlon, is to delay a patient's urgent treatment. "In the ICU, I have to reprioritize tests based on the linear design of the EHR. It's really pretty embarrassing, if not absurd or dangerous," Scanlon adds.
User Centric's Schumacher has found EHR screens that cut off the names of certain medications, leaving physicians unsure while during orders. Another source of error can be conflated buttons, clickable bubbles representing various options that do not clearly correspond to treatment options, so that what the physician intends to order may not be what actually goes through.
Default selections jeopardize the accuracy of orders in some EHRs, while others induce incomplete information or repeat orders due to inaccessible patient information, Scanlon reports. The result is physicians devising workarounds to orders to get the EHR to truly correspond with what the physician wants to order.
This kind of EHR gerrymandering has its own set of risks, according Carl G.M. Weigle, MD, CMIO at Wisconsin Children's. Weigle and Scanlon, along with Ted Faust, PharmD, and Mari Akre, PhD, RN, form the hospital's multidisciplinary Knowledge and Solutions Architects group that is aimed at "improving the systems we put in front of our colleagues" by implementing local fixes while conducting broader usability research.
The majority of EHR usability issues are more innocuous, constituting small inefficiencies in design that, when compounded over time, require physicians to spend more time at the computer and shorter periods with patients.
Small font size, similar-looking icons that serve entirely different purposes, a lack of shortcuts, inconsistent functionality and data that are difficult to retrieve are just a few of these small agonies. "Anyone who has been in the field for any period of time has faced these issues over and over again," says Justin V. Graham, MD, MS, CMIO at NorthBay HealthCare, in Fairfield, Calif., who has helped lead the implementation of four separate EHRs during his career.
One of the primary underlying issues is that EHRs confine physicians to a linear workflow, whereas healthcare delivery—and human interaction more generally—involves significant give and take and constant adjustment by both participants. This is something computers are currently not well equipped to handle, says Dean F. Sittig, PhD, professor of biomedical informatics at the University of Texas Health Science Center at Houston.
The systems follow strict if-then logic, producing specified outputs based on physician inputs. However, care depends on interaction with patients and family members, many of whom naturally prefer that their doctor focus on them rather than the computer, Sittig says. "In general, there is a mismatch between what physicians are doing and what the EHRs are requiring. The computer often drives the patient visit, which not all physicians want."
Shared blame
Outdated design is the crux of EHRs' non-intuitiveness. "The systems were designed to emphasize billing, record keeping and inventory management," Sittig says. "To expect a system designed to do these things to facilitate physicians' thought processes at the point of care is unrealistic."Despite having dramatically shifted the dynamic of care between physicians and patients, EHRs still rely on systems designed as far back as 20 years ago.
What has allowed the systems to function in the setting of care has tended to be add-ons and extra layers of information processing, rather than wholesale redesign, according to Graham. "Vendors took the paper system [and] computerized it without thinking about the new workflow the systems enabled. They're stuck in a 1990s paradigm of software when the world has moved on," he explains.
The lack of understanding and attention paid to the systems early on has snowballed into some "horrific" practices, says Schumacher. "Most EHRs look like database elements were just thrown up on the screen for physicians to fill out," he says.
Physicians who use the systems every day inevitably know at least something about what they need an EHR to accomplish for them, which is why Scanlon insists that they be included in the design phases.
Still, not all problems originate from suppliers. Providers have paid as little attention to usability as many vendors. "It's fun and easy to pick on vendors, but traditionally, providers don't ask for highly usable devices. We ask for more bells and whistles, which adds complexity and the opportunity for additional errors," Scanlon says. The outcome, Schumacher notes, is that you get what you ask for, not what you need.
Altogether now
Providers and vendors need to step up collaboration, say Weigle and Schumacher. Although EHR developers should take greater efforts to include physicians in design, an important way for vendors to improve usability would be to spend more time observing physician workflow and relying on experts in the field of human factors, according to Weigle and Schumacher.As far as safety is concerned, Graham questions why EHR manufacturers don't follow the lead of hospitals in employing dedicated safety officers to vet their products for usability flaws that might lead to patient harm. Although this would help to minimize error, Sittig points out that added checks will mean additional clinician time spent at the computer.
Some providers likewise question whether there isn't room for a third player in the equation: either government- or organization-based standards.
"The industry has been this sort of wild, wild west of EHR vendors just creating systems," Graham says. "The marketplace hasn't held them to any standards, and purchasers often aren't sophisticated enough and don't look at usability when making purchases." Still, he, like others, remains leery of endorsing intervention. And as Scanlon points out, many of the issues associated with usability have only become prevalent in the all-out rush to meet meaningful use.
What is not debatable is the high percentage of usability errors that can be prevented, and the ways in which this can be done. Physicians are beginning to call on their peers to hold vendors accountable, so that, as markets are intended to function, the less-usable EHRs get weeded out. In this area, government could play a role in establishing minimum usability standards and thus alleviating the substantive proprietary barriers physicians face in transitioning between systems.
Ultimately, the aim of EHRs must be enhanced workflow and improved patient safety, but with minimal attention to usability and poor collaboration on both the vendor and provider sides, these systems all too easily can accomplish just the opposite.