nonobviousness in device design

I always have reservations when I open an email on a Sunday from my work address. That being said, this FDA presentation showed up in my inbox today from our VP of R&D about medical devices being connected to the wrong attachment, resulting in some pretty nasty consequences:

(as a product of the US Government, none of this material is copyrighted - its use and reproduction is encouraged)

At first glance, as a medical device engineer with a significant clinical background, this seems ridiculous. Connecting a nasogastric tube (used to feed patients when they aren't capable of eating and need significant nutrition) to a Foley catheter (used to drain the bladder of urine)? A feeding tube connected to a tracheostomy port? By a trained nurse?


But as I look closer, some of these aren't obvious to me (are they obvious to you?). And if they're not obvious to someone who makes the stuff, I can't imagine anyone that would have an easy time with knowing how to place each and every component that surrounds a patient in a hospital bed. In the hubbub of a hospital room, with patients, doctors, family, and the scores of other people that are involved in healing, it's not always easy for someone to glance at something and know exactly where it goes and how it connects. And that is terrifying. Some of these actions have killed patients (well, when you're pumping milk into an infant's lungs...). Others have led to gross deformity and reduced quality of life.


The onus is on us, as designers, to make things as obvious as possible. Don't assume that a nurse is trained on your device. Don't assume that a doctor has read the Instructions for Use. They might have been trained on a different procedure, or be familiar with a different device. Assume nothing, and make it as difficult as possible for someone to use your device incorrectly. Don't make standardized luer connections on the end of everything just because it's cheaper to manufacture - it's also easy to connect a standard luer to pretty much anything... including things you don't want to connect to. A second molded component is worth the cost if you can save a patient's life.


In general, it's a good idea as a designer to never rely on your instruction manual. Most people (myself being the exception as I am a user's manual disciple) don't take a second look at the instructions that came with a product. And why should they? Instruction manuals are famous for being wordy, detailing features that we'll never use (but cited in our decision to buy the product), and overly complicated. They're written in technical (and legally manipulated) language. If you're in a developing or foreign country, the translations may not be perfect, and nor do you really care about reading through them. And when you have a critical patient in front of you, an instruction manual is about the last thing you have time for.


A design should limit people from doing what they shouldn't do - "designed incompatibility." It should be difficult to do things with a device that lead to nondesired outcomes. It should be damn well near impossible to do things that would cause someone harm.


Oh, and don't go walking around the hospital connecting random tubes together.