Thanks for this summary comment. This is backed in the literature by people who are systems / usability engineers:
tldr: Reliable design is HUGELY important. You can't just put these things together haphazardly.
"The fact that less than 50% of participants across all devices could follow the labeled instructions without committing a single error provides confirmation that the need for training on the use of epinephrine autoinjectors is still important...The user-centered device design may have a significant impact on correct epinephrine autoinjector use and patient preference."
In the study below, moms of kids with severe allergies were trained to use the EpiPen. Six weeks later in a simulated test, only about two out of five were able to successfully deliver a lifesaving dose of adrenaline.
Shouldn't the FDA require that manufacturers provide evidence that consumer medical devices can be used reliably for their intended purpose?
You're right. With some Googling, I learned that the FDA does indeed have a "Human Factors Pre-Market Review Process."[1]
But for EpiPen, it seems that the process didn't work, because 60% of consumers apparently can't use the device reliably. I hope that EpiPen is an anomaly, and not typical for FDA-approved lifesaving consumer medical devices.
They should allow reasonably safe products that are among the best available options.
So if people can use an Epipen more reliably than they can prepare an injection, allow it to be marketed. If someone with a history of anaphylaxis is better off with an Epipen than they are with no treatment at all, allow it to be marketed.
tldr: Reliable design is HUGELY important. You can't just put these things together haphazardly.
"The fact that less than 50% of participants across all devices could follow the labeled instructions without committing a single error provides confirmation that the need for training on the use of epinephrine autoinjectors is still important...The user-centered device design may have a significant impact on correct epinephrine autoinjector use and patient preference."
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2892620/