Designing user interfaces for worst-case scenarios: Lessons from the COVID-19 pandemic

It’s too early to draw conclusions about the ongoing COVID-19 outbreak. Months or years will pass before we understand the virus’ root causes, its full impact, and the effectiveness of the various attempts to control it. That said, current events continue to provide lessons about the way in which we go about our personal and professional lives, during both normal and exceptional times. This is perhaps especially true for people who design, engineer, and manufacture medical devices. 

As medical teams around the world deal with the healthcare crisis caused by the virus, medical device manufacturers face parallel challenges as they attempt to manufacture and supply scarce products essential to treatment of COVID-19 patients. You might have heard about shortages of COVID-19 testing kits, personal protective equipment (PPE), and ventilators. 

While many medical device manufacturers are appropriately focusing on production and supply chain issues at the moment, current events can also teach them important lessons about medical device design, and specifically user interface (UI) and user experience (UX) design, that might be easy to overlook in more “normal” times.

The best time to think of the worst-cases

International standards such as IEC 14971 (Medical devices — Application of risk management to medical devices) and IEC 62366 (Medical devices — Part 1: Application of usability engineering to medical devices) require medical device manufacturers to identify, analyze, and mitigate the various risks that might result from the use of their products. Such activities typically focus on the most common use scenarios associated with a given medical device.  

Well-intentioned medical device designers might be eager to exceed regulatory requirements and produce medical devices that protect against worst-case scenarios, such as those that involve use by untrained or unauthorized individuals or use in unintended use environments. But, even the most well-intentioned manufacturers face practical constraints such as finite development schedules and budgets. Ultimately, manufacturers must strike the right balance between taking a pragmatic development approach and mitigating the widest range of potential worst-case scenarios. 

Having helped many medical device manufacturers design their devices’ user interfaces, I have seen well-intentioned designers and engineers forced by project practicalities to underestimate the likelihood of certain use scenarios. When trying to resolve hundreds of product development issues on a tight timeline it can be tempting to classify worst-case scenarios as too unlikely to consider and focus instead on mitigating the most common risks. 

However, the current COVID-19 crisis reminds us that worst-case scenarios and “black swan” events do occur and that their consequences can be catastrophic. This reminder provides a beneficial case of recency bias: medical device manufacturers currently working on risk analyses might be more likely to consider and account for the types of worst-case healthcare scenarios that are now occurring around the world. 

This was the case with the GE Healthcare design team that worked on the Engstrom Ventilator in the mid 2000s. As Paul Micheli, a product manager and UX designer, recounts on his site and in a recent LinkedIn post: the design team worked on the ventilator shortly after the “Avian Flu” outbreak of 2006 and was inspired by prevailing events to consider worst-case scenarios in which untrained users might be required to operate a ventilator. 

Paul’s team ended up implementing the ventilator’s “Basic UI” – a simplified version of the ventilator’s user interface that could be used by clinicians lacking respiratory therapy experience. 

Had the design team been working on the ventilator at a different time, when the Avian Flu outbreak was not on their minds, they might have been inclined to consider use by untrained individuals as too unlikely to merit a mitigation. But, the recency bias resulting from the Avian Flu outbreak led the team to implement a specialized feature that might actually be helping to save lives in today’s healthcare crisis. 

Expecting the unexpected

In addition to highlighting the importance of considering untrained users, the Engstrom Ventilator story illustrates another facet of accounting for worst-case use scenarios: the importance of attempting to identify unexpected or “abnormal” use scenarios.  

While it is appropriate and necessary to identify the most typical and common use cases, focusing on them exclusively means that a medical device could end up being used in unexpected circumstances that introduce new risks or increase the potential for identified risks to occur. 

The ongoing crisis gives us several such examples. The highly infectious nature of COVID-19 means that many medical devices and even consumer products must be handled in ways that minimize potential for spreading the virus.

Consider the following cases:

• Medical devices that might previously have been used without gloves should now be used with gloves and other protective equipment to prevent the virus’ spread. Touchscreen displays and physical buttons that were easily accessible without gloves might become more difficult to use and more vulnerable to use errors when operated with gloves.
• Devices that were previously used with multiple patients might now require thorough cleaning or disinfection between uses. Devices with complex enclosure shapes and tough to clean crevices might not be properly disinfected between uses and could become a potential carrier of infectious diseases.
• Devices that are typically used in a modern, well-equipped intensive care unit, might need to be re-purposed for use in a minimally-stocked field hospital. Resources that are normally taken for granted, such as electrical power, air and gas supply, and physical space, might be more limited when used in unexpected or unintended use environments.
• Despite their use in non-medical environments, certain consumer products, such as ATMs and credit card readers, become potential sources of infection as well and might be vulnerable to the same risks and use cases described above. 

One might think that such use cases only become apparent once a crisis actually occurs. But, using various a mix of risk identification techniques during a product development effort can help identify plenty of potential use scenarios that are likely to occur during worst-case scenarios. For example, using a mix of brainstorming, user interviews, and usability tests can help designers and researchers identify unexpected product uses that are unlikely to be discovered through more conventional engineering methods. 

Can regulations help?

The aforementioned standards have succeeded in getting manufacturers to utilize risk management techniques and have arguably made contemporary medical devices safer than their predecessors. But, the current standards focused by and large on typical, “normal” use. 

While it is unrealistic to expect manufacturers to mitigate against every possible risk in every possible scenario, including worst-case and unexpected scenarios, perhaps future versions of such standards could guide manufacturers to address the types of risks that occur during worst-case use scenarios. For example, future versions of these standards could identify specific types of emergency scenarios that lie outside the realm of typical use but should not be considered abnormal use, such as use during a pandemic, natural disaster, or war. 

The paradox of preparedness

The average medical device development effort is already relatively challenging and expensive, and medical device regulations are already seen by some as too demanding. So, I suspect that most medical device manufacturers are not eager to take on the added challenge of identifying and mitigating against worst-case scenarios that lie outside the realm of typical use. 

But, it’s important to remember that making medical devices more resilient to worst-case scenarios will likely also make them safer and more usable in day-to-day use and in the smaller scale emergencies and unexpected uses that happen more frequently than true worst-case scenarios. In fact, that’s why more extensive mitigations are often underappreciated – if they are effective at preventing dangerous scenarios from occurring, then people assume they are unnecessary. This “paradox of preparation” explains why effective mitigations are often misjudged as being excessive. 

In retrospect, we can judge GE Healthcare’s decision to include the Basic UI in the Engstrom Ventilator as clever and progressive. But at the time, it probably seemed to some at GE Healthcare to be an excessive response. Fortunately, the design team had the insight, and perhaps the luck of good timing, to recognize the feature’s true value. 

Let’s hope that today’s healthcare crisis inspires medical device manufacturers to implement safety features and risk mitigations that will help save lives in, or even prevent, future healthcare crises.