If you develop an EpiPen and it costs you a bunch of money then in theory you can make it up by selling more volume. Except your competitor can just cut the price because they are already selling for an absurd amount of money, so how do you win now?
Maybe we should have a non-profit that does generic versions, and we can subsidize it by preferring them to companies like this (e.g. all medicare provided drugs now go through this company which sells generics). Unless the companies are willing to match prices.
It's not the insurance company, it's the drug company. The insurance companies are probably livid about the extra costs they're having to cover.
The answer to "how" is a variety of different techniques, including changing the formulation slightly (http://www.nytimes.com/2014/09/16/business/new-york-files-an...) and outright bribery of competitors (https://www.washingtonpost.com/opinions/pay-to-delay-pharmac...).
One of the alternatives was potentially better as they had additional features, but there were problems with the injector itself and it was recalled. That's what other companies have to contend with. It's bad if the injector fails to deliver the dosage and a person dies.
So, people making a 3D printed alternative can't really ensure that the injector works reliably without a lot of work.
That is incorrect -- devices can do that. Devices shown to be equivalent to an existing device can be put on the market with minimal regulatory overhead. The form to submit to the FDA for an equivalent device is a 510k Premarket Notification. Details are here:
http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidanc...
This is what an acceptance letter from the FDA looks like for a substantially equivalent device:
http://www.texasbiogene.com/upload/content/file/TBGFDAapprov...