Before a medical device can be sold in the U.S., the manufacturer needs to demonstrate that it can be used safely and effectively. Do people hurt themselves by mistake when they try to use it for the first time? Can people pick it up and hold it so that it doesn’t slip and fall? Can people read its display and know what the display is telling them? Does the device resemble another device that is used differently, leading people to use the new device incorrectly? Do the instructions make sense to a wide range of people coming from different backgrounds?
Core Human Factors, Inc. specializes in testing new medical devices for just these sorts of issues, by having representative users come in and use new devices in a variety of scenarios designed to represent real life. This blog post is not intended as an advertisement for the firm. Rather, it is a “slice of life” from our niche in applied psychology, where careful simulations of reality are necessary to capture “use events”–instances in which people unknowingly use medical devices in ways that are not intended by the designer of the device, potentially leading to injury to themselves or others or an inability to benefit from the device.
Some etymology of the term “use event” is helpful in revealing the framework for medical device usability research. First, consider the term “user error.” If you are an engineer, the person using the device you built might use it the wrong way (i.e., in a way different than what you intended). You may believe that a “user error” occurred, and you as the engineer may shrug your shoulders and say, “oh well, people are stupid, but you can’t control that, and at least my device didn’t go wrong.” However, a more enlightened engineer might say, “Hm. I built a device that people are prone to using incorrectly and possibly hurting themselves. I’m going to call this a ‘use error’ (not user error) and not cast blame on the end-user in particular, for perhaps it was my device that was hard to use. Maybe I’ll go re-engineer the device so people won’t get frustrated or hurt themselves with it, even if they do use it in what I consider to be a wrong way.” This is the beauty of iterative design: one can go back and re-design with input from the end-user.
Just to broaden the net a bit, a use event can include an incident in which an end-user does something the designer never intended to be done with the device, but is not necessarily harmful. From the standpoint of understanding how people are going to use a new device, use events can be as informative as use errors. Using a hair dryer to de-fog a mirror is a use event, a use of a device that was not intended by the designer of the device. However, this use event is not in itself harmful, and is not exactly an error in use – perhaps it is a very positive creative new use — but it does open up a new range of downstream consequences for manufacturers to consider. Unexpected behaviors by end-users are, for me, the core of usability testing.
Why are people so unpredictable that usability testing is necessary to find out what they are going to do with a new injection device (besides inject themselves in the arm like the instructions say to)? Perhaps because, as people, our imaginations about what other people are going to do are ultimately limited. I would invoke here the argument that an entity cannot contain a complete representation of itself or even of another entity sufficiently like itself, and so an engineer – a human — can’t contain a complete representation of the minds of end-users (also humans). All of the fields of scientific psychology exist for basically this same reason: we are mysterious to ourselves, but careful empirical study can provide us with valuable new knowledge on the matter. The entirety of the innards of an injection device can be sufficiently mastered by engineers that the device will not fail in years of use, but the person who needs it might forget to prime it before injecting. To know when this will happen requires an understanding of forgetting, and this is an unsolved problem. To prevent the forgetting from happening requires not judicious application of oil to aligned gears, but to delve into educational psychology, an unsolved field (to say the least).
Engineers who first pass through our offices regularly facepalm. “How stupid people are,” they will say, as a study participant with Type 1 diabetes picks up the new diabetes device upside-down after reading the instructions twice. This is not the perspective human factors professionals take, which is more like “The user is not to blame” — see Richard Pak and Ann McLaughlin’s blog’s tagline, “Not blaming the user since 2007.”
Designing to a level where the device is intuitive to the vast majority of users is impossible in a vacuum. However, usability testing can be conducted early in development, and by the end of development, continued testing can give designers confidence in the success of a design that is intuitive not just to themselves, but to actual users. Reduce the facepalm.