You don't just randomly put s*it in your mouth because it "looks innocent".
If I saw a colourless liquid on a table, and I didn't know where it was from, I wouldn't just randomly drink it.
Likewise with just randomly picking up plastic bits and eating them.
However, if you're talking about young kids who are liable to put anything in their mouth, then "looking innocent" doesn't even enter into the equation.
As a parent, you should be looking after your kids, and childproofing your house - you wouldn't put scissors on a table, or a kitchen knife within reach, or bleach or other household cleaning liquids, or a hammer, or nails.
And you wouldn't let them unattended near a swimming pool, or let them near the stove in the kitchen, or let them walk in a park without supervision.
All of these are basic parental responsibilities - I don't see how these magnets are any better or worse than letting your kid eat a staple, or a nail etc.
I mean, as a kid, I shoved a lego up my nose (yeah, I was a smart cookie), and had to be rushed to the doctors.
Another time, I nearly ate dog poo.
Did both of those things (lego and go poo) look "innocent"?
Kids do.
>The Centers for Disease Control reported 33 cases requiring surgery and one death.
Judging from this data, assuming the risk stays the same over time, you have to play with magnets for roughly about 1000 years for your risk to be one micromort[1]. Same one micromort risk is assumed by drinking two glasses of wine or walking for 17 miles. Playing with a magnet for a year is as dangerous as eating a banana (bananas are radioactive).
I'm not a stats guy, but it seems like this comparison would be more useful if we somehow account for the difference between the number of bathtubs installed and the number of neodymium magnet toys purchased among the population.