Designing for Generations – A Look at UX in Healthcare

“Well, you know, I’m older. I’m not good with technology.”

How often do you hear users describe themselves this way? Because we hear it a lot, particularly when we are conducting usability research with nurses. In fact, we can barely get through a usability study with a nurse over 45 without her or him (usually her since nearly 90% of nurses are women) stating it. And what’s perhaps most remarkable, and most telling, is that the participants are volunteering this information; we’re not asking them how they feel about their tech competency. We’re simply asking them to complete tasks with software that they might use one day to help care for patients.

The strength of this pattern had us wondering…

• Is there truth to what they say?
• Are there ways that we as UX designers can improve nurses’ confidence when faced with new technology?
• What are the ramifications for this self-described insecurity in an environment that can be life-or-death?

The truth about generational differences in tech

Popular opinion, at least in the United States, is that younger people are better at technology than older people. But what does “better” mean? That they are faster at accomplishing a task? Can do more things with technology? Are earlier adopters?

Nielsen Norman Group conducted a global study to try to understand if, and how, users from multiple generations use technology differently. NNG concluded, Compared to older users, young adults tend to be extremely confident in their own ability to navigate digital interfaces, even when encountering radically new design patterns.

We have seen this firsthand while conducting user research across verticals. Young adults will trust themselves to figure something out. They proceed through a digital experience with little concern that they could make a mistake. The fascinating contradiction (though it makes sense when presented) is that this confidence means that “young adults are error prone when using interfaces.”

Meanwhile, we observe older adults (those older than Millennials, aka born before 1980) say things like “[I am] worried about what might happen if I click on this, can you tell me?… I don’t want to screw up.” This is even within the controlled context of a usability study where there are no patients and no colleagues, and the participants have been expressly told that there are no right or wrong answers.

If nurses are this hesitant to take action in a safe context, imagine how hesitant and concerned they must feel when there is a patient relying on their decisions. Mistakes are not allowed in their profession. Software shouldn’t be adding to the existing burden of perfection required of nurses, pharmacists, doctors, and other clinicians. These are moments that as UX practitioners we need to be aware of and manage.

Mistakes are not allowed in a nurse’s profession. Photo credit

While many digital experiences serve broad age ranges (Facebook, Starbucks, Netflix) most do not have the safety implications of the software and digital devices used in hospitals. Within nursing, professionals range from their mid-twenties to their mid-sixties. The average age of nurses is 43.9 and over one third of all nurses are 50 or over. In essence, there is a 40+ year age gap amongst healthcare providers using and sharing software systems.

Meanwhile, the demand for nursing is high; the profession is expected to grow at 14.8% (versus the national average job growth rate of 7.4%). Thus the challenge to UX designers: prevent user errors that younger generations might make out of false confidence AND make the software less intimidating to older generations.

One might think, isn’t the answer just to have nurses better trained with their software tools? It isn’t quite so simple.

Barriers to learning software on the job

Time

Nurses are super busy. We cannot stress this enough. They have many duties and depending on the type of environment/department they work in, they may be responsible for many patients and/or patients who need care multiple times within an hour. Their job performance may be measured partly in terms of time (for example, the % of time that they successfully complete charting while taking a patient’s vital signs). Prioritizing tasks is not necessarily a solution for this time-shortage either; competing urgent priorities are often the reality. For example, a patient out of surgery needing a bed ASAP vs. a patient needing medication before the top of the hour.

Availability of computers and software

Most software in use in hospitals is expensive. As a result, the number of copies may be limited. And nurses may not have personal “work” computers; they tend to use desktop computers that are in locations designed for sharing (like a main desk) and/or laptops on wheeled carts. The demand for the computers may make it impossible to take one for an hour of self-training. Additionally, nurses may want to learn in a more private environment rather than in a space where they might be seen by colleagues, patients, and patients’ families.

“Safe” testing mode

Think back to how you learned to use Photoshop, Excel, Word, or PowerPoint. Now imagine that the software didn’t have an “undo” option, the files were all shared with other people, there were no tutorial videos publicly available, and, oh, making a mistake could have serious consequences. This is the case for nurses using clinical software. There’s no such thing as a “test” or imaginary patient in these systems. With all this, it comes as no surprise that nurses aren’t going to click around unless they have absolute confidence in the effects their actions will have.

Doing better for nurses

Our work has taken us to many facets of healthcare: workflow software related to payments; software for predictive analytics in healthcare that tracks provider performance, and stand-alone medical devices that are regulated by the FDA. This last group, which includes things from a giant CT machine to incubators for babies, is where we’ve had the opportunity to interact with dozens of nurses, as they are the primary users of products we’ve helped design.

The good news is that the FDA recognizes that usability is critical to the safety of these devices and requires rigorous usability testing; the bad news is that the software we’ve seen in clinical settings does not have consistently high learnability, efficiency, or desirability. Overall, products in hospitals seem well behind consumer technology. Here are some early insights we have gleaned that might help address these deficiencies:

Research recommendations

• Use every opportunity to interact and research with nurses, especially in their actual work environment. While we conduct usability testing in a lab environment that simulates a hospital room (complete with a life-sized dummy in the bed), there is no way to accurately recreate the scale or complexity of a hospital environment. Ensure you are observing the real thing.
• Remind nurses that there’s no right and wrong; not just at the start of the study but whenever they discount their own feedback.
• Remind nurses that they are a representative user and that you are trying to design something that is great for all nurses. Data is important to clinicians, so (if their age was part of the recruiting process) you can remind them that X percentage of nurses are within 5 years of their age, so their experience and insights are highly valid.

Design recommendations

• Ensure as many actions as possible are “undo-able.” Many softwares we have observed still rely on users confirming an action, rather than allowing undo. This is A) annoying, since most of the time nurses do, in fact, want to do the action and B) frustrating should a nurse make an actual mistake since they cannot undo it.
• Shift from tech-centric to human-centric language and iconography. One of the great paradoxes and design dilemmas we’ve seen within healthcare is that because digital products are updated much more slowly than they are in the consumer world (reasons include cost of upgrades, complexity to upgrading, and time/cost to get FDA approval), the use of “normal” lingo lags behind other softwares. For example, we’ve seen an icon of a server (a three-layered cylinder) used to indicate a patient’s previous data within the application. This presumably was intuitive to an engineer 10 years ago, but doesn’t align with any current convention and isn’t intuitive to a new nurse. The dilemma, however, is that for a nurse with 20 years of experience, they may have become accustomed to that icon. We highly recommend sticking with the best practice of icon plus a text label whenever possible to help bridge these gaps.
• Choose readable fonts and font sizes. Nurses’ job is intense, and their time is scarce, so use fonts that are highly readable and easy to scan. Consider the distance the text needs to be legible from—much of the time nurses are looking at screens across a room or while they walk by. Pick colors palettes that reduce eye strain and, if appropriate, offer night modes.
• Don’t be a slave to trends. One of the advantages of clinical software not being a consumer product is that the design doesn’t need to reflect the latest UI trends. The screen should look pleasing, but there is no pressure to sacrifice usability for aesthetics (you know the trends we’re talking about). Electronic health records, patient monitors, MRI imaging applications, etc, are all very information-dense and so having obvious cues is a good thing.

Training recommendations

While the goal is always to have perfectly learnable, intuitive UX, the reality is that healthcare software is very specialized and often very complex. We can support nurses, and other healthcare providers, by providing better learning tools for the applications they use.

• Make training videos publicly available. While software companies often choose not to do this for proprietary reasons, making training accessible for nurses outside their shift hours is crucial. These videos could simply require an employee login to view. Alternatively, a company could choose to buck this trend and gain many fans if they provided short tutorials on YouTube. And, of course, if other software is any indication, clinical users will take it upon themselves to create tutorials and offer helpful tips. The goal is to provide a mix of comprehensive training videos and super short how-to videos that nurses can view on demand.
• Offer a training mode within software. Some healthcare software already features a “demo” mode that sales people use for demonstrating the product features. This mode simulates patients and scenarios. It would be great to modify this mode so that clinicians can use it as a psychologically safe space to try out features, learn about different actions, and play around a bit.
• (Print) manuals are not helpful. The absolute last thing a nurse has time to do is hunt around for a giant manual, then attempt to locate the relevant information in a table of contents or index. In our opinion (and we’ve never once heard a healthcare provider mention a manual), a print manual has zero value to nurses. A digital manual is slightly better, but only if it is searchable, includes screenshots, and can be easily accessed (ideally both from within the application and online).
• Make software training courses part of the continuing education requirements. Nurses are required to complete continuing education as part of their professional licensing. While requirements vary by state, software training hours could be added to the ways nurses fulfill these requirements. This is possibly one of the best methods to build confidence in nurses of any age while destigmatizing the need for training. It’s a great way of acknowledging that people might need some help using technology and to reduce embarrassment around that knowledge gap.
• Don’t rely on coach marks or onboarding overlays. Coach marks (UI hints often shown as a transparent layer) are generally designed to appear on a specific trigger, such as the first time a user interacts with a button or visits a page. This system is based on the assumption that there is only one user of the application. However, nurses often operate on shared computers, and some software does not require authentication. Thus there is no way for the system to know when, and for whom, to display the hint.

We can support nurses by providing better learning tools. Photo credit

Shifting the definition of ‘user’ in healthcare

As the points above make clear, traditional business goals and current user goals are not always aligned. Healthcare is hardly the only sector struggling with this notion. So much of healthcare software is determined by business strategy that it isn’t serving the nurse or doctor who is spending all their time with patients. When companies allow business goals to eclipse the real needs of their users, eventually they lose out to the competition or disruption.

As UX researchers and designers working in healthcare, it’s our job to advocate for the user. Involving nurses in innovation and research is critical in advancing the digitization of healthcare. As you plan research and design, consider the varied ages and tech approaches of clinicians and beware of the complex ecosystem in which your designs will live. Given the growth of voice interfaces, telemedicine, AR/VR, etc, we are witnessing a wave of new technology in healthcare and with it comes an equally large learning curve.

Moreover, it is the job of everyone to remember who the ultimate end user is: the patient. They are the benefactors of software that is easy, intuitive, and delightful to use. Great UX means nurses have more time to spend with patients, waste less of their day being frustrated, and make patients’ healthcare experiences feel safer. The core journey is the patient-healthcare provider journey. It is an experience that is complex and anxiety-inducing under the best of circumstances. Our responsibility as UX practitioners is to enable nurses to use all the tools available to them confidently and effectively, so they can get back to being superheroes.