Why Caring About the User is as Important as Caring for the Patient: The Importance of UI/UX in Healthcare Clinical Decision Support Systems

By Sirj Goswami and Gitta Salomon

In computer science, an interface is the exposed ‘surface area’ of a system, presented to the outside world to mediate between what’s inside and outside. An optimal interface hides complexity and harmoniously interacts with the user. The disciplines referred to as User Interface (UI) Design and User Experience (UX) Design civilize technology, taking a messy, complicated system and coaxing it toward behaviors more suitable to user interaction.

Historically, the healthcare system has suffered from poor UI/UX design. In comparison to other industries (e.g., manufacturing, finance, retail, information and communication technology), ease of use and integration with typical user workflows has taken a back seat to implementation of raw functionality.

Ironically, the lack of priority given to UI/UX in healthcare is in direct contradiction to its importance; healthcare user experience is as important, if not more so, than in other industries because patient well-being is at stake. A well-designed UI/UX plays a critical role in hiding software complexity from end users. It emphasizes key features and functionality and crystalizes the clinical purpose of the software for the typical user, organically creating clear and compelling reasons to use the product. Equally important, a well thought out UI/UX reduces the likelihood of errors, minimizing patient risk.

There are several reasons why UI/UX has taken a back seat in the healthcare software development process. Before trying to fix the problem, it is important to understand why healthcare has been historically subject to poor software UI/UX. Understanding the challenges will inspire better design, increase product adoption, and ultimately improve patient care.

Electronic Medical Record (EMR) systems are old and designed for an outdated healthcare system. EHR systems like EPIC and CERNER own a majority of the market share in the U.S. and were originally designed to meet the requirements of fee-for-service systems and not the needs of clinicians or patients. EHR systems were built to manage claims, document basic patient information, and help hospitals get paid. For healthcare technology to take a leap forward, an evolution in the interoperability of existing EHR systems is required (a few companies have made progress in solving this issue). Smart health applications need to be able to extract and process data from EHR systems easily and present results to clinicians in a user-friendly way.

Clinical workflows are complicated. Let’s take precision dosing of complex drugs as an example. Clinicians are in charge of disease diagnosis and treatment selection. Clinical pharmacists undertake calculations and optimize dosing by looking at specific clinical lab data that is displayed in a format that is only understood by them. Finally, nurses are involved in administering drugs to the patients. An optimal UI/UX must pay careful attention to the entire workflow, and manage information transfer in a way that makes sense to all those involved. It requires great care to design a product experience and prioritize features when there are several cooks in the kitchen.

An optimal experience must take into consideration viewpoints of various user types who are involved in patient management. Clinicians have a different workflow than clinical pharmacists and nurses. To design a useable and useful software product, we must consider audience composition and be sure to include key players in the design process from the start of product definition.

Healthcare software typically follows a waterfall software development process – basically, products are developed using lengthy software development cycles. This occurs, in part, because Clinical Decision Support Software must be carefully developed to minimize the potential for any patient risk. The long development cycles are not conducive to validating product usability with typical users. At the end of a lengthy development cycle, developers often find that there are adoption and usability issues. Because they’ve invested significant development effort, they are reticent to make changes and often (inappropriately) attribute problems to user incompetence.

Complexity of data inputs/outputs and idiosyncratic terminology. Let’s look at precision dosing as an example again. Dosing history and drug concentration levels are complicated inputs, requiring domain expertise. Similarly, unless you are a domain expert, “Treatment Probabilities” and “harmacokinetic Estimates” are complicated outputs.  Idiosyncratic terminology is also an issue here. Harmonizing clinical/technical language is a difficult task – one that most companies do not anticipate needing to tackle.

To overcome these challenges, it is important to involve end users early and often in the development process. Lessons from user-centered design and design thinking –methodologies successfully adopted by other industries – can equally benefit the healthcare domain. For example, software developers should strive to develop empathy with their target users, and would benefit significantly from observing the context of use, and their needs and goals. Mockups, simulations or visualizations of the final product should be created (if not in actual code, using tools such as InVision or Marvel) to assess user engagement and usability long before the development team is fully committed to building a specific direction. Quantitative surveys can be undertaken to confirm hypotheses uncovered through qualitative work, or, conversely, to uncover trends that need to be further explored through qualitative research.

We cannot overemphasize the value and importance of putting product ideas in front of users well in advance of – as well as during – any significant development effort.

End user involvement need not be a massive undertaking. Conducting qualitative research with as little as five users can provide valuable and powerful insights. However, it is important to cover the spectrum of likely users. As mentioned earlier, users from different professional disciplines will ultimately interact with your product. In addition, both expert and ‘typical’ users’ perspectives are important. Expert users can contribute to the product vision by exposing features and functionalities that help revolutionize a field. These advanced users also appreciate specialized functionality and accept the complexity that comes with more power. On the other hand, typical users provide insight into the most essential features that will lead to widespread adoption. These users are less inclined to learn the nuances of the software and are most interested in minimizing the time required to learn and use the product.

During the early stages of software development, careful attention should be placed on the challenges discussed here. Doing so will increase the likelihood of product success, which in turn will facilitate a smoother translation of scientific research into clinical practice. Collaborating with specialists in the UI/UX domain will help provide the technical expertise to undertake these user-centered tasks, as well as minimize the influence of bias by bringing fresh eyes to your effort. UI/UX specialists are experts at developing empathy with your users, and can ensure that perspective is not forgotten.

Sirj Goswami is CEO and Co-Founder of InsightRX (www.insight-rx.com). Gitta Salomon is Principal of Swim Interaction Design Studio.

How to Make Your Company’s User Experience Effort Succeed

You’ve undoubtedly heard that user experience (UX) is a key competitive advantage for your product or service. You know you need a UX person in your organization. And even though you’ve never hired a UX person before, you’ve posted an opening that lists the job requirements you saw in another company’s UX posting. Fingers crossed, that results in a good hire and a positive UX outcome for your product.

Will hiring a single, hands-on, UX person solve it?
In a quick analysis of current UX/UI designer job postings, many of them include most or all the following requirements:

  • Conduct user research to uncover customer needs
  • Lead multidisciplinary design sessions to define product features
  • Create user flow wireframes
  • Produce pixel-perfect artwork
  • Conduct usability testing
  • Assist with front end coding

It’s my assessment – after addressing UX for hundreds of products and services – that you will be hard pressed to find one individual that meets all these distinct requirements well. And also based on experience, better UX designs arise from having more than one person undertaking UX activities. Two (or more) people can provide deeper coverage across the required skill sets. Discussion and give-and-take between UX peers leads to better outcomes, stronger design rationale and speedier results. A lone UX person is often at a significant disadvantage when it comes to fully exploring the design space and defending design decisions.

In the current marketplace, there are many individuals relatively new to UX who are looking for jobs. The ones who believe they meet the entire laundry list of requirements above are often lacking the depth of experience to be self-aware about their actual competencies.

That creates an awkward situation. You don’t know what you don’t know about who to hire for your UX position. The candidates don’t know what they don’t know about undertaking the work. The likelihood of creating a less-than-optimal UX situation at your company is high.

A lot of people mistake lovely graphics for good UX. Good UX is so much deeper than the visual skin job.

You need more than lipstick on the pig
Because many of the items on the laundry list above are subtle to communicate (except for front end coding, which I would argue isn’t truly a UX skill), UX candidates are often assessed by their visual portfolio. It’s easy for candidates to share visuals and easy for the hiring party to look at them and like them (or not).

Many times, visual design becomes the primary focus of the UX effort, and drives who gets hired as the sole UX person.

I’ve recently tried a few products that looked incredibly beautiful and professional. I was initially surprised at the level of refinement, knowing that they were created by early-stage startups. But that appearance was superficial. The screen designs belied the fact that it was difficult to understand where I was in the app or complete basic tasks. What I could accomplish was significantly more difficult than it should have been. Clearly, no one took the time to design a good flow for product interactions. Sensible behavior is an essential component of an interactive product’s experience.

Be careful not to fall for good visuals alone! A lot of people mistake lovely graphics for good UX. Good UX is so much deeper than the visual skin job.

What do you need?
Before you need a hands-on, individual contributor, you need a UX strategy. You need vision, direction, and big picture goals for the UX. You need user research to ensure you aren’t getting sidetracked by your own biases in defining your product and an initial framework for the product’s behavior that looks out ahead of near-term agile sprints.

You don’t necessarily need an in-house person for these strategic tasks. You do need someone who has significant experience applying design thinking and user-centered UX processes.

Once your strategy is in place, you’ll be better able to define who, specifically, you need to add to the team. You might already have a product manager that can undertake user testing. Or a developer who has good instincts around interface design coding. Or a marketing graphic designer that might be able to help with screen design.

If you don’t know what you don’t know about UX, you undoubtedly will not achieve deep success in UX with one relatively junior UX designer. Invest in figuring out your UX strategy first, and then advertise a position with a clear sense (and a more realistic list) of what you need that person or persons to do.

How does this perspective on UX capabilities relate to your situation?

Swim provides strategy, advising, user research and other valuable services that enable startups and companies lacking in UX skills to position themselves beneficially for growing an in-house UX effort. Principal Gitta Salomon serves as a fractional VP of UX to help put companies on the right track.

Feeling Blue

Last week I had two different user experiences of the same location, and the juxtaposition gave me pause. A beautiful new Blue Bottle coffee had opened in an old San Francisco office building. Passing by, I admired its clean, fresh, and inviting appearance. A few days later, I went to my dentist appointment in that same building, and experienced what the Blue Bottle architects had created from a completely different perspective.

It is important that we understand and consider the full context in which our designs will live.

I’ve been going to the same dentist in that downtown office building for over 15 years. And the same lobby receptionist greets me with a welcoming nod every time.

Last week as I entered the building for my appointment, something felt different. It was loud, and there were people in the foyer. A lot of them. Patrons were spilling out of the new Blue Bottle Coffee into the marble-clad corridor in that unmistakable, ever-present Blue Bottle customer line.

I headed for the elevator, but then I turned back to talk to the receptionist because I realized his environment was now so radically different. He told me it was noisy all day long. “Cacophony” was the word he used. And he remarked that because of the noise, he no longer could read at his desk.

While Blue Bottle (the Company) and its customers are experiencing a lovely new location, there is one person whose life is changed — for the worse — as long as he wants to stay in his job. The receptionist’s experience is peripheral to the one that was designed, but it should have been considered. A line at Blue Bottle is a given; the architects could have situated that line outside, or could have acoustically dampened the sound in the foyer.

While we often assume what we are designing is contained within set boundaries, looking into the places where a design spills over is always warranted. It is important that we understand and consider the full context in which our designs will live.

I hope that I still get a nod next time I go see my dentist, but I fear my receptionist could be gone by then. Will an entry kiosk replace him? Only a computer could ignore the aural difficulties of his current work environment.

5 Ways to Make UI Working Meetings Work

I was participating in a UI brainstorming meeting recently with four other designers and a product manager. The meeting was focused on “solving” the UI for a mobile app. I was struck by how many times I’ve been in this type of meeting — where the team hopes to emerge with the design for the app — but no particular process is in place to ensure it can happen.

Here are 5 techniques for making UI design meetings actually succeed.

1. Know exactly what features you are designing for
You can’t move a group conversation forward if you don’t have shared knowledge of the features you are going to provide. And no individual designer can possibly create a holistic UI framework without knowing all the pieces that need to be supported.

If there isn’t a well-defined list of features that everyone is privy to, then it’s too easy to get bogged down in “are we doing that?” feature wrangling. Inevitably, there is tension between what the MVP should contain and what ultimately could be delivered. It is beneficial to have that ultimate picture in mind so you don’t end up ‘band-aiding‘ additions at some later point. But if you don’t have justification that users would be interested in a specific feature, or engineering can’t possibly build it, or the marketplace doesn’t warrant your product containing it (because it’s peripheral to your offering and someone else is already doing it better), let that feature go. It is just creating noise.

What to do: Identify what your product is going to offer. Make a list, or better yet, construct a taxonomy — where larger features that encompass sub-features are clearly spelled out in an outline format. Then make sure that product management and engineering have signed off on it and that the designers are familiar with it before you head into a design meeting.

Continue reading 5 Ways to Make UI Working Meetings Work

Uncharted UI Hazards

On the night of November 29th, the 65-foot sailboat Team Vestas Wind slammed into a reef at 19 knots (a significant speed in a boat) in a remote part of the Indian Ocean.

They were on their way from Cape Town to Abu Dhabi, along with 6 other boats, in the second leg of the Volvo Ocean Race — a round-the-world sailboat race and one of the toughest sporting events in the world.

How did this happen? These were professionals — some of the best sailors in the world. The obvious answers are that either the charts were inaccurate or the boat’s crew committed “human error”. The consensus in the sailing world attributes the accident to human error. A recent New Yorker article sums it up:

“At a press conference on Monday, Team Vestas admitted that the crash appears to have been caused by ‘a simple human error’: the navigator, Wouter Verbraak, did not zoom in enough on his charts. Race organizers have pointed out that it was a high-stress situation caused by approaching bad weather.”

At the risk of reviving an age-old discussion, the navigator’s (in)action may have failed the crew, but the charting software’s user interface clearly contributed to the wreck.

Continue reading Uncharted UI Hazards

Is Cultural Insensitivity a Side Effect of Digital Calendars?

Why is this year different than all other years? Because so many people are holding events on Jewish High Holy Days – Rosh Hashanah and Yom Kippur (the most solemn day of the year) – that have nothing to do with those holidays. In fact, the scheduling of events on these days is likely to preclude Jewish participants.

In particular, I noted:

  • A business is holding its 30th anniversary party the evening Yom Kippur starts.
  • The University of California San Francisco is holding a neighborhood fitness festival on Rosh Hashanah.
  • A Jewish friend was invited to her good friend’s wedding shower on Yom Kippur.
  • A neighbor noted that his vintage car club decided to hold their special drive on Yom Kippur, thereby inadvertently excluding a bunch of members.

I have a hypothesis: might widespread adoption of digital calendars have fostered this?

Continue reading Is Cultural Insensitivity a Side Effect of Digital Calendars?

Why Good UI Designers Can Work in Any Domain

Many of our prospective clients don’t understand how an interaction design consultancy can possibly come up to speed on their complex domain. And understanding their domain is central to creating or improving their product’s user interface.

What these clients don’t often realize is that neither an interaction design consultancy, nor their users, need to know everything they know.

Interaction designers fill a knowledge gap between clients and their customers.

The reason complex product user interfaces are often hard to use is because they expose too much. The people creating these products assume their users will be as singularly focused on the product as they are. They expect their customers will equally value every last feature and function, and will devote significant time to mastering the product. That is akin to assuming every motorist wants to be a mechanic. In truth, most people just want to get somewhere.

Continue reading Why Good UI Designers Can Work in Any Domain

Invisibility is Our Superpower

When meeting new people, I’m often asked, “So, what do you do?”  I usually throw out various terms like Interaction Design, UI Design, UX — all the while scanning the questioner’s face for a hint of recognition. More often than not, the person politely smiles and nods…and then asks, “What’s that?”  Once I finish explaining the particulars of interaction design, I’m often met with “Oh, I didn’t know that was a specific job.”

A colleague recently pointed me to an article in the Wall Street Journal discussing David Zweig’s book (“Invisibles: The Power of Anonymous Work in an Age of Relentless Self-Promotion”) about professions that are mostly “invisible” — e.g., anesthesiologists, fact-checkers, translators — that is, their work, when done well, goes largely unnoticed by the public. This got me thinking that this is exactly what we strive to do as interaction designers. We design interactions that aren’t unnecessarily showy, but are so seamless and integral to the product that they just feel right. It’s a tenet that users notice interaction design only when it is done poorly, or when it wasn’t considered at all.

And yet — as the WSJ article points out — purposefully dedicating oneself to working in a largely unrecognized milieu is anachronistic in this age of relentless self-promotion. If our best work doesn’t loudly toot its own horn, how do we convince prospective clients that we are good at what we do?

Continue reading Invisibility is Our Superpower

Tripped Up

I have an interesting anecdote about the persistence of data, the ramifications of a poor UI, and the unintended consequences of things talking to other things on one’s behalf.

I recently traveled from the Bay Area to Arizona on Southwest Airlines. The day before my flight, I received a flight confirmation email – but it came from a completely unrelated airline, and for travel dates that were slightly different from the days I was supposed to be traveling.

I confess, I momentarily freaked out. I hadn’t made the reservations myself, but I was pretty sure this couldn’t be right. Upon closer inspection, it became clear that this was a reservation for “Mary” – someone I didn’t know – who was traveling between cities on the other side of the country.

So, I emailed Mary to let her know what happened (her email address was on the confirmation), and suggested that perhaps she should contact the airline.

My involvement in this bogus trip didn’t stop there. The information in the reservation email propagated to a number of services, and they all came rushing to help me:

  • TripIt automatically scanned my Gmail account and generated an itinerary for Mary’s flight – a flight I wasn’t taking – which it emailed to me.
  • TripIt also let my network know about the trip.
  • Google started sending me notifications when Mary’s flight was delayed.
  • While I was in Arizona, Google Now began providing me cards of interesting things to do – on the opposite side of the country – based on Mary’s destination. This seemed particularly odd. Google Now should have known where I was based on my phone’s location.

Continue reading Tripped Up

When Branding Falls Flat

A quick quiz: name the brands represented below.


No problem right? These are clear, distinct and easily identified.

Extending this a bit, the same should be possible with other design elements and imagery – specifically, a brand should be easily identified by its website as well.

Quiz #2: name the brands represented below by their websites.


Not so easy, is it?* Here’s the problem: when everyone adheres to the current flat design trend, distinguishing branding elements disappear. All brands tend to look very similar, essentially defeating the effort to create a lasting, memorable, and recognizable presence.

This critique usually applies when any design trend becomes adopted by millions. For example, the same thing happened with the Apple-originated aqua-like UI elements; once introduced, the style was widely adopted by many applications and platforms, resulting in a similar loss of brand differentiation.
Continue reading When Branding Falls Flat

UX strategies & observations