Technology vs. usability: DIY medicine

November 08, 2016

Access to information and technology has advanced remarkably.

One of the more interesting areas enabled by these advancements appears to be comprised of equal parts initiative, innovation and daring: DIY medicine.

Becoming our own health advocates is increasingly important, so access to information is more important than ever. But taking responsibility for our health is not always simple or straightforward. If you're not feeling well and you start browsing the internet you will, invariably, come to the conclusion that you have either Ebola, stage IV brain cancer or lycanthropy.

But the internet isn’t just for self-diagnosis anymore. There’s a guy who straightened his teeth out by 3D printing his own braces. Night Scout is another example: a group of people developed open-source IoT solutions for continuous glucose monitors before the manufacturers did.

But the one that really caught my attention is the EpiPencil. Rightly outraged by the inexplicably steep and steady price climb of the EpiPen, this group figured out a way to assemble an epinephrine auto-injector from inexpensive and readily available medical supplies.

Near the end of the instructional and demonstration video, the spokesperson for the group says that the EpiPencil is equivalent in function to an EpiPen and “with no special training, anybody can use it.”

That claim is outrageous. It’s not necessarily wrong, mind you, just outrageous. I really, really, really wish he hadn’t said that.

An important part of using a medical device is preparing it for use: unpacking it, assembling it, loading it, etc. Not everyone is going to be able to get through that assembly process correctly. Some people, believe it or not, won’t even be able to get through the assembly process at all. People with diabetes (especially Type I) have to manipulate, draw and deliver insulin on a daily basis, so over time they develop mad skillz with a syringe. People with severe allergies? Not so much. They are different user groups and both would have to be represented in a usability study if the device were to be indicated for both uses.

If you're designing and building the device, you aren't a user. You can’t get rid of the biases you have.

The video instructs the viewer how to assemble the parts and draw the syringe, but those Instructions for Use (IFU) are a fundamental part of being able to use the device safely and effectively and therefore need to be tested for usability as well. If a user can’t understand the instructions, then it’s unlikely they’ll be able to correctly assemble and prepare the device.

But let’s assume for a moment that anybody and everybody is able to correctly perform all of the steps to build their own EpiPencil. Then it’s a simple matter of just using it, right? Except that usability testing typically reveals use errors that were not anticipated during the design — one of the fundamental reasons that the FDA requires usability testing in the first place. Use errors can range from minor to extreme to bizarre. They are usually something of a surprise.

And that’s the thing: it’s hard to fathom and impossible to predict all of the different mistakes people will make. Until you actually watch them forget to use all of the parts. Until you see them grab the wrong end of the device and try to use it backwards. Until you see them fail completely, it just doesn’t seem possible that anybody could get it wrong. Then you realize it’s not about you. You realize that you really do need to find a bunch of total strangers who would actually need to use this thing and test them, one by one, and find out which part of the design needs to be changed. Then you need to repeat that — maybe a few times. Then, when you’re sure you’ve seen everything and designed out the difficulties, you might be ready for one last usability study.

If you're designing and building the device, you aren't a user. You can’t get rid of the biases you have. Coming up with the idea, developing it, figuring out what parts to use and how to use them, doing functional tests and making an instructional video: all of these are necessary steps in developing a medical device. But they are not sufficient. You simply can't un-know things about the device you just built and approach it like a novice user.

If you build an EpiPencil for your spouse or child, it doesn't matter that you think they will be able to use it. It is my studied and professional opinion that people simply don't have the ability to make those kind of predictions. If that were possible, I would do it myself. My job would be easy and I could just fiddle around with each medical device and tell you which ones are usable and which aren't. I've observed countless hours of usability studies and been surprised nearly every time at the things people struggle with.

Using information is not the same as simply finding it on the internet — you have to know when and how to apply it. By the same token, medical devices — DIY, OTC or prescription — pose use error risks: you have to be able to set them up and use them.

Still feeling in a DIY mood? The FDA has published human factors guidelines for ensuring the safety and effectiveness of medical devices, which you can read here. Knowledge is power.