If you're interested in this kind of thing, try reading the book Hacking Healthcare. From what I gather so far (haven't quite finished), there are large costs

If you're interested in this kind of thing, try reading the book Hacking Healthcare. From what I gather so far (haven't quite finished), there are large costs that are incurred by dealing with insurance. The USA's billing system -- matching diagnoses to codes, and codes to billable amounts -- seems like it makes sense on the face of it, but once you start digging in, its a rat's nest. There are several disparate coding ontologies, many of which are proprietary. Different insurance companies accept different ontologies, meaning clinics need to pay for access several times over. The ontologies don't necessarily match up one to one with diagnoses, and clinicians often don't care to match them up themselves. So, clinicians diagnose a patient, and then back-office staff need to go through and match those diagnoses to codes. The insurance companies will reject anything that doesn't match their own data. So, if an payment request comes through for John Q. Bloggs, they won't issue payment if the name they have on file is Jonathan Q. Bloggs. So, you need more back-office staff to deal with the rejections, marry the data up properly, and send it again (in order to save time and money, there are apparently clearing houses now; middle-men between the clinics and the insurance company who can advise staff more cheaply if a given request will be rejected for invalid data, or unbillable codes). Then, they can reject certain codes that a patient doesn't have insurance for, even if its covered by some other code. So, again, the back-office staff need to go through and retry the payment request with more relevant codes (which are now further from the diagnosis, but hey, whatever). Finally, once it's all been hashed out, insurance companies often won't pay the full amount anyway -- so the patients sometimes need to be chased up for the remaining payment. Wrapped around all that are all the requirements for dealing with fraud, auditing, etc. which make it all the more difficult. It's massively inefficient, and somewhat self-fulfilling: the complexity of it requires more staff, which drives costs up. There's no direct way to bill for staff time, so charges to patients go up (hence the $10 aspirin or whatever the famous example is). Which leads to more complexity in billing, which leads to more staff, and on and on.