On the Need for Human-Centered Design in EHRs

Bennett LauberHealth information technology (HIT) has become the hottest political issue in Washington. The healthcare industry in the United States is facing a crisis as medical facilities have spent hundreds of billions of dollars implementing electronic health record (EHR) systems, yet patients and the physicians and nurses that care for them are seeing few benefits. Congress has been holding hearings focused on detailing the problems and trying to write legislation that will provide a solution to the crisis.

The HIT interoperability bill drafted by Rep. Michael C. Burgess (R-TX) is one example. These are welcome first steps. However, none of the bills currently before Congress, and none of the hearings, are addressing the two most important issues facing medical providers today. These are lack of EHR usability, and the inability to have a patients’ entire medical record at the point of care.

The later issue has been ably addressed by Bettina Experton from Humetrix here. In this article, and future articles. we are going to address the issue of EHR usability. This topic is going to require several articles as first we need to lay out some of the current problems regarding lack of EHR usability, while explaining the importance of usability, a concept that is clearly little understood by Congress, the Office of the National Coordinator for Health IT (ONC), and even major EHR vendors.

The State of EHR Usability

More and more evidence is emerging regarding the lack of EHR usability. Speaking at the 2013 Healthcare Information and Management Systems Society (HIMSS) Conference & Exhibition, Michael S. Barr, MD, MBA, FACP, of the National Committee for Quality Assurance (NCQA) warned that: “Satisfaction and usability ratings for certified electronic health records (EHRs) have decreased since 2010 among clinicians across a range of indicators.” Barr’s presentation at HIMSS focused on “ the need for the Meaningful Use program and EHR manufacturers to focus on improving EHR features and usability.”

More recently the RAND Corporation, the JASON Task Force, and 37 professional medical societies have raised serious concerns about the lack of EHR usability. Leading medical informaticist Dr. Scot Silverstein has been raising these issues for years in his Health Care Renewal blog. Silverstein provides substantial detail in two recent articles that can be read here and here.

And it is not like EHR usability is not a known issue. The Electronic Health Record Association (EHR Association), a non-profit association of more than 40 EHR companies, created an EHR Developer Code of Conduct, which aims to encourage transparency and collaboration among EHR developers, as well as developers, providers, and industry stakeholders.

On the first page of the EHRA code of conduct, the very first item (after a general statement) is Patient Safety. The code says:

"Recognizing that patient safety is a shared responsibility among all stakeholders in an increasingly health IT-enabled, learning healthcare system: We are committed to product design, development, and deployment in support of patient safety. We will utilize such approaches as quality management systems (QMS) and user-centered design methodologies, and use recognized standards and guidelines."

What is “Usability?”

The terms User-centered design (UCD), Usability, and User eXperience (UX) have been used over the years to describe the work of the software professionals that specialize in the human-computer interaction. "Software Human Factors" is the field of study that applies the methodologies of human performance and ergonomics to software.  Instead of trying to design objects that work with the physical attributes of the human body, experts in Usability and User-centered design virtual interactions that work with the mental capabilities of human minds.

Do you remember the HP calculators of the 1970s? They were great for mathematicians, but the general public was really confused about how they worked. They were confused because in order to perform even the most basic mathematical functions people had to think differently. They had to think like the mathematicians.

Adding up a series of numbers was simple. All one had to do is key in a number, press <enter>, key in the next number, press <enter>, and then press the <+> plus key to calculate the sum of all the numbers entered.  Easy as π!

The problem with these calculators was that the design of the user interface focused exclusively on expert users and these experts were a very limited sample size. The answer to fixing the calculators was User-Centered design (UCD).  UCD is a design philosophy that creates a culture of understanding and enabling end users to perform their tasks using an information architecture and taxonomy that matches their mental model.  

After changing the user experience to match a more common understanding of arithmetic, e.g. key in a number, press <+> plus, key in another number, then press <=> equal, the market for desktop calculators exploded.

The HITECH Act and the Problem of Engineering Centered design vs. User Centered Design.

The Health Information Technology for Economic and Clinical Health Act (HITECH Act) is part of the American Recovery and Reinvestment Act of 2009 (ARRA). ARRA contains incentives related to health care information technology and contains specific incentives and penalties designed to accelerate the adoption of EHR systems among providers. The Office of the National Coordinator for Health Information Technology (ONC) released a set of Safety-enhanced Design §170.314(g)(3) certification and meaningful use requirements for Electronic Health Records (EHRs). Stage 2 of these certification required EHR vendors to include evidence of user-centered design and summative usability test results in their submission.

Summative usability testing for safety-enhanced design involves recruiting targeted users as test participants (Doctors, Nurses, and other medical practitioners) and asking these users to complete a set of pre-defined tasks. An expert test facilitator conducts the testing via an established test protocol while the test sessions are recorded and later analyzed.

The summative usability tests for ONC Meaningful Use Stage 2 certified EHRs are all made public on the CHPL site.

A big problem is that many of the vendors didn't work with medical professionals in their designs. They created what we call Engineering-centric designs, not User-centered Designs. They made HP Calculators. Their systems were designed and developed by software engineers with little input from medical professionals (the users). Complicating matters, a number of EHR vendors took serious end-runs around the regulations and did not conduct, nor report, on a proper summative usability test in order to become certified. Some of the Authorized Testing and Certification bodies seem to be rubber-stamping the usability reports perhaps without even looking at them.  

Think about this-If an EHR vendor took side steps in preparation of their usability evaluation, what other short-cuts did they take with development of their system? I’m frightened that someone may suffer serious injury because some EHR vendor ignored usability testing so that their clients can get ONC funding. 

The U.S. Food and Drug Administration has acknowledged getting hundreds of reports of problems involving health information technology including numerous patient injuries and deaths.

Some examples seen at hospitals across the country:

  • At Marin General Hospital in Northern California, RNs called on the Marin Healthcare District board to delay implementation of their EHR system. "Orders are being inadvertently passed to the wrong patients. People have gotten meds when they've been allergic to them. This is dangerous," Marin RN Barbara Ryan said in comments reported by the Marin Independent Journal.
  • In Chicago, the Chicago Tribune in 2011 reported on a patient death at Advocate Lutheran General hospital after an automated machine prepared an intravenous solution containing a massive overdose of sodium chloride — more than 60 times the amount ordered by a physician.
  • At Affinity Medical Center RNs in Massillon, Ohio, nurses raised multiple objections to the hurried introduction of an EHR system. Subsequently, they have cited medication errors, delays in care, problems with documentation, computers crashing, and other concerns.

For another example of why usability in healthcare is so important, see the article “How Bad UX Killed Jenny.” 

As noted earlier, Rep. Burgess, a physician, released a draft bill that is meant  to fix some of the serious shortcomings of the HITECH Act. Unfortunately, the draft bill completely ignores the problems with usability in healthcare IT and continues the policy of ignoring caregivers, patient safety and patient rights organizations, and other healthcare organizations, from playing an active role in the drafting of health IT policies.

On the other hand, on Friday March 20, 2015 the HHS released their proposed rules for Stage 3 of the meaningful use program.  Contained within these new rules was very significant change in the meaningful use program, that is an expansion of the Safety-enhanced Design (aka usability) testing portion.

See pages 190 to 196 of the proposed 2015 ONC certification document for the complete text of the changes to the Safety-enhanced Design program. A Quick summary of the enhancements includes:

  • ONC will require 17 instead on 7 areas to test
  • ONC recommends 15 participants, instead of no recommendation (we have seen many certified EHRs that only tested on two people!
  • ONC clarifies the User-centered Design reporting requirements.
  • ONC provides guidance on when an EHR needs to be retested due to changes in the UI

We welcome these changes to the usability testing portion of the Stage 3 criteria.
Many of these changes are a direct result of suggestions given as public comment on the 2014 certification program by those in the usability community.

What exactly is usability and user-centered design?

According to the ISO 9241-11 standard usability is defined as "The effectiveness, efficiency, and satisfaction with which specified users achieve specified goals in particular environments (ISO 9241-11)."

  • Effectiveness - The accuracy and completeness with which specified users can achieve specified goals in particular environments.
  • Efficiency - The resources expended in relation to the accuracy and completeness of goals achieved.
  • Satisfaction - The comfort and acceptability of the work system to its users and other people affected by its use.

Usability in healthcare can be difficult to achieve, but it is important to remember that it is not only based upon the aesthetics of the user interface.  Good Usability is also not determined by the number of clicks (see The Myth of Too Many Clicks).   

A useable healthcare system must be designed to match the mental models and workflow of its users.  A usable EHR needs to work (effective), work well (efficient), and not cause any unnecessary frustration (satisfying).

The big business interests of the Healthcare industry may cry wolf (and lobby hard) against enhancements to the usability program because they don’t want to spend the extra time and money to provide a healthcare system that truly follows a safety-enhanced design philosophy (or, acknowledge that low-cost, open source systems, like the VA’s VistA EHR, are highly ranked by users for their usability). 

These EHR players are no better than the automobile industry that fought hard against seat belts in the late 1960 and against The United States Intermodal Surface Transportation Efficiency Act of 1991 that required airbags in cars. With Congress working on legislation to fix major healthcare problems caused by the HITECH act, we hope that they will finally address the issue of lack of EHR usability.


Different perspective on

Different perspective on usability testing - I can tell you the the ONC-ACB body we used was very thorough in reviewing the usability testing. We took great pains to recruit different doctors, technicians and administrators to spend and hour or two each to go through the testing. The results were quite positive, but realize that the major design and testing were done long before this formal testing. Therefore, I don't see the value in having large numbers of testers at this point. Also realize that these testers are unpaid by requirement. They are only allowed to receive small tokens, such as gift cards for coffee, for their time. Therefore, recruiting larger numbers of people to do this testing would not only be burdensome, but also the improvements end users would see would not be much if anything at all.