I love going to the HFES Healthcare Symposium to reconnect with old colleagues, meet new ones, and learn from each other’s experiences. And it’s always great to see such active participation from the FDA. They really show how much they care by committing to be an integral part of the community they serve.
Keynote speaker, Sorrel King (pictured below), told the crowd about Josie’s Story, her memoir about the medical error that led to her daughter’s death at eighteen months. Now a nationally-known patient advocate, King’s words provided much inspiration as she reminded us all why what we do is important and special. It was extremely motivating to hear that you often need little more than the will to make a real difference!
There were so many topics of note at the conference, it was hard to know what to feature in this post! I narrowed it down to three areas around which discussion was especially lively this year: 1) designing and testing for autonomous medical devices and systems; 2) AR/VR equipment for human factors research; and 3) designing and testing for global health solutions:
1. Designing and Testing for Autonomous Medical Devices and Systems
Devices that make adjustments to therapy based on the patient’s need without informing or gathering input from the clinician are referred to as ‘physiological closed loop systems.’ Currently, there are no specific guidelines for these, but don’t forget that the FDA’s human factors guidance published in February 2016 applies to these new systems and ecosystems — just as it has for myriad devices over the years!
The growth of such physiological closed loop systems brings many benefits regarding increased and improved healthcare. These systems can provide low cognitive load, fewer human errors, and minimized interactions with users. Of course, these benefits can also be viewed from the flip side — less interaction with clinicians could lead to increased risk in some cases. So, the potential for overlooking errors could increase.
Additional factors to consider while designing and testing for such systems are over-dependence on automated systems, over trust, over reliance, loss of situational awareness, and potential skill degradation that could result from fewer capabilities within these systems.
This is a new area of exploration, so it was interesting to hear about techniques and tools being used by user experience designers and researchers in this domain:
Ensure your designs acknowledge the user’s trust (or lack of trust) in the system. Do so even though there isn’t much guidance or standardization on how to ‘calibrate’ trust when designing for greater automation within the healthcare space. (Bill Horan suggests some guidelines in his post, “Who Knows Best? Designing User Control in the Age of AI.”)
Make sure you understand your users, their characteristics, and the ecosystem in which they work — any or all of these things might help or hinder the use of application or system. Not knowing how much trust is too much or too little, and how that impacts the tasks your users are doing, can be catastrophic. One way to approach this is to provide advanced settings within the system for additional controls.
Think about effective and efficient ways to provide on-boarding and refresher training to your users to ensure the system’s safe and effective use.
Think about borrowing practices from other industries, such as aviation or law enforcement. It’s even more important to structure and organize data in ways that are efficient for users when you have large quantities of it. With new simulation techniques, simulating for loss of situational awareness to evaluate for over-trust and reliance is a great way to test for scenarios that are less common yet critical. Another great option is to think about running longitudinal studies — those that run over a longer period of time. This will help provide great insights on learnability, trust, and user experience over the lifecycle of the product.
It’s even more important to structure and organize data in ways that are efficient for users when you have large quantities of it.
Remember to reach out to the FDA early in your development process. They can provide you with additional guidance that is specific to the nature of product and users you are designing for.
Think more broadly about designing for automation. Think about the right level of automation and where the automation really adds value in the work your users are performing. Where can the automation reduce the mental and/or physical workload of the user, and is it really going to provide value? Also, think about the overall human-system dynamic.
This is just the beginning of an extremely interesting era to be working in this space. Soon we’ll need to consider how and when we can automate decisions. Are there situations where the system needs special patient consent? Should medical devices and workflows make use of big data and artificial intelligence to make decisions? To what extent? When is tele-surgery appropriate? Are there domain- or country-specific policies or guidelines that might help or hinder the administration of care in these scenarios?
Issues will continue to need to be ironed out, including the integration of data sources and the large quantities of inflowing and outflowing data from a surgery, plus how this data would be used by the systems and the users interfacing with them.
2. Using AR/VR Technology for Human Factors Research
There was also a lot of buzz at the conference about leveraging AR/VR equipment for human factors research. We’ve begun to experiment with it here, too, for product development. Pictured above is user researcher, Satyajit Balial, operating our HTC Vive to view a true-to-scale Solidworks model. For those interested in AR/VR (and everyone working in human factors, in my opinion, should be!), here are some things to be aware of:
AR and VR applications and devices are currently being used in training for surgeons and medical students in a number of different ways. They can add dimensionality to anatomical images, which allows students to observe abnormalities. There are applications to help provide guided training for HCPs, including hands-on practice of procedures for surgeons. Physicians can also overlay scanned images onto a patient during reconstructive surgery to see beyond what the human eye is capable of today.
AR and VR methods and techniques can greatly help to increase realism in your testing while reducing the cost. If you are new to these technologies and have never used them before, start by playing around with some devices that are easy to get a hold of. There are a number of different applications already in use.
Some — features offered by Amazon and IKEA, for example — can be accessed from your phone. You can also use Google Cardboard (pictured above) for a more in-depth experience. For those who are already familiar with these, consider gaining more familiarity with the different emerging terms — reality, mixed reality (augmented reality and augmented virtuality), and virtual reality. Think about how you can use some of these to design your next product or test your early concepts and how you can realize them in more realistic simulations using devices like the Microsoft Hololens.
The use of augmented reality in user research is best suited for generative research. There are times when you’re trying to gauge look and feel with a looks-like model or trying to understand interaction of the device in the environment of use with a works-like model. AR environments lend themselves well to tangible design ideas (including the intended color, material, finish, size, scale, and depth) while still providing flexibility within the system to alter them.
Using these models, users can think beyond the look and feel and begin to interact with the device or system and get deeper into interactivity and context of use. AR models can be a lot faster to build and easy to change in-house compared to less interactive CAD printouts or more labor-intensive physical models that have a limited scope for modifications. Given the flexible nature of generative research, being able to modify and explore is extremely important.
The Microsoft Hololens is a great device for creating a more immersive experience. There are also options that are on the low-tech end of the spectrum such as using overlays on a display on a mobile device. However, as described by Laura Chang from Radius Innovation and Development, the Hololens (pictured above) is one of the best in market, easy to develop for (using Unity), and can stream video, which is extremely helpful to share with a larger team. Some of the drawbacks are the low resolution; and the registration, which may not be very accurate and may have limited field of view. Look out for the next gen which is coming soon!
What to keep in mind when designing an AR/VR-driven study? Much like other studies you have designed, you will need an interdisciplinary team consisting of designers, rapid prototypers, researchers, and subject matter experts. You will also do yourself well to have your note taker be a Hololens specialist!
When designing the study, consider including a ‘training’ task for users to familiarize themselves with the technology so when they start working on the tasks, they are comfortable enough with it to feel immersed in the task and inside the context of it.
You will need an interdisciplinary team consisting of designers, rapid prototypers, researchers, subject matter experts, and a notetaker who’s also a Hololens specialist!
Also, consider building in the capability to scale and manipulate the product or device to the extent that it makes sense to get feedback on that aspect of your project. For example, give users the ability the change the size (height, width, etc.) of your physical model or tool you are testing in order to fit into their hands or the environment of use. This is where this technology can really be leveraged to its full potential. And if you’re early enough in your project, consider designing a ‘build your own product’ study that can leverage all the benefits of a participatory research session and the AR/VR technology to support it.
3. Designing for Global Health Solutions
While we’re thinking about all these interesting areas of growth, it’s important for us to be aware that seven percent of the lesser developed nations in the world account for about eighty percent of the world’s population. These are countries that operate much differently and have very different needs from what we’re familiar with in more technologically developed nations.
These nations’ challenges and past experiences are a world apart from ours. They have a growing, aging population that is battling illnesses related to chronic diseases (compared to more communicable diseases in the past) while noticing a shift from spending out-of-pocket money to using government-supported services.
In these countries the focus is on value (not technology) and reusables (not disposables). There is often an intermittent power supply that’s not centrally integrated as well as a lack of well-defined hospital protocols. Once you’re aware of these challenges, it’s easy to see why nearly three-quarters of the systems designed and built for one part of the U.S. market are never used in other parts of the world and are relegated to the trash heap before they even get used!
The good news is while these systems do not hold up, the human factors design process does still hold up! So when designing for a more global solution, remember to think about the following:
Think about the environment and the context of use. Think about the constraints the system will be operating under. Think about the local systems and infrastructure you can leverage to make the system work to its full potential, and to provide your users with a powerful tool.
Think about your users. What skill sets do they possess? What are the cultural and personnel interactions that need to be taken into account when designing these systems?
Think about the tasks these users are trying to perform in these environments with the knowledge they bring.
Be creative in your approach to designing the solution and testing it, both in the generative and validation phases. Find ways to combine some generative research with concept validation of high level designs or prototypes. This will greatly help with keeping costs low on travel and timelines.
Integrate human factors early in your process. This can certainly help to identify and address a lot of the areas discussed above.
Leverage local resources and involve the local community in coming up with creative solutions. This will help to design products that are tailored to the specific needs of people in the area and will also be sustainable for the economy.
When you work through the design iteration process, you will be quickly humbled by the things we sometimes take for granted and exhilarated by how creative solutions can be born!
I hope this post is helpful for those who weren’t able to make it to the HFES symposium this year. I look forward to continuing to follow these topics and to learn how they’ve progressed in 2019!
(*Trust in technology map source: HSBC’s Trust in Technology report)
Learn about our Medical Product Design expertise.