How Bad UX Killed Jenny(medium.com)

6 years ago from Jonathan Shariat, Designer @ Google | Author @ O'Reilly | Podcaster @ DesignReview

  • Mitch Malone, 6 years ago

    This article hits home for me. I work on an EHR, a system very similar to Epic. We talk about this stuff all the time. There are a lot of breakdowns that can result in a situation like the one described in the article.

    Simply redesigning the UI can be a fool's errand when you don't know the system architecture and business goals. A lot of these systems (including the one I work on) are legacy systems with layers and layers of different code stacks, APIs, databases, etc. that results in a house of cards. Or you have a tech stack that's so old, all the people who worked on it are gone and newer engineers don't even want to touch it in fear of breaking everything.

    For enterprise healthcare systems, like Epic, the UI can become an amalgam of shit because competing functional teams at the hospital vie for competing needs. And then the people who actually use it (i.e., nurses and doctors) are stuck trying to use a broken UI.

    For ambulatory systems, like the one I work on, the perception from customers and companies is that the product must bundle with 100+ functions and features that no one will use. The more features, the better. This makes sense intuitively but it usually results in common usability problems with navigation and discoverability.

    Health systems are starting to wake up though. I'm inspired by ListRunner which helps doctors track their patients. It's a focused, single-purpose mobile app that doctors actually like using. I'm hoping we start to see more of these unbundled style healthcare apps. There's clarity and power in a focused experience that's separate from the main EMR system.

    13 points
    • Jonathan ShariatJonathan Shariat, 6 years ago

      I was hoping someone with real experience would share about what its like. I thought it might be a lot of red tape and business issues.

      So do you think theres any hope of fixing the current systems, or is it going to take a brand new company to step in?

      What do you suggest can be done?

      2 points
      • Mitch Malone, 6 years ago

        There's a lot to be done. Sometimes, I think maybe the best thing to do is just blow up the entire ecosystem and start over.

        Aside from just solid UI design, I think there needs to be innovations in business models for healthcare systems. We're starting to see these emerge. One model that I'm inspired by is direct primary care which removes a lot of the administrative costs to healthcare and changes incentives from volume-based care to patient-based care. With this model, it doesn't matter how many patients you see in a day (which leads to 10min appointments and poor care). What matters is how well are your patients doing.

        Sometimes, the best thing to do for a patient is to do nothing. But that won't get a doctor paid, even if they know it's in the patient's best interest. So they order unnecessary tests and procedures they can bill for. Direct care removes this kind of incentive.

        Iora Health is a company attempting to get direct care models in place.

        One Medical is also disrupting the traditional care model.

        From a UX perspective, I think unbundled, single-purpose product experiences that are so common in consumer apps today will start to permeate in the healthcare space. As I mentioned before, ListRunner is doing this.

        In general, healthcare is getting a lot of VC recently so I wouldn't be surprised if we start seeing these things sooner than later.

        Hale Health has a really cool telemedicine product that I'm excited to try out.

        1 point
    • Rob GormleyRob Gormley, 6 years ago

      In addition to working in Operations here at LayerVault, I spend a lot of time on an ambulance, delivering patients to an Emergency Department that runs on Epic, and it's quite ... horrific. Absolutely horrible UI, and there would be very little surprise that information could easily get lost, possibly critical. The UI is amateur, ugly, and complex.

      I should submit some screenshots from Tiburon (who supplies our Fire Deparments with mobile data terminal software). We're talking an embedded IE7 engine in an ugly app with quite possibly the world's worst map UI (picture working on a ToughBook in a moving fire engine with a map that works in only a single context at once - pan, zoom, scroll, select), for one.

      1 point
      • Jonathan ShariatJonathan Shariat, almost 6 years ago

        Thank you so much for sharing. I'd love to see those screenshots.

        As someone who uses these types of software, what do you think can/needs to be done?

        0 points