I think this is applying a "car insurance model" (sorry in advance for the car analogy) to healthcare. Ignore third-party insurance for the moment as it's not really applicable to health. You don't generally make an insurance claim for something like an oil change, 30/60/90k mile checkup, or something like that. On the other hand, if a tree fell on your car and smashed the engine, that's when insurance kicks in: for the catastrophes.

I think it's the exact inverse of what you describe. The article describes something like a subscription model for the oil changes that also satisfies the legal requirement for health insurance under 26 U.S.C. § 5000A. The enormous liability and collision part of auto insurance in this analogy is not covered by the plan described in the article.

On one hand, it seems like a great idea for those whose employers don't provide subsidized insurance coverage. If it's truly affordable, it should allow more people to get routine care. Especially for women, who generally need more routine doctor visits than men. OTOH the things that bankrupt people are hospitalizations and real insurance plans are generally there to help mitigate that.

The last paragraph of the linked blog post mentions that it should be combined with an HDHP to cover things like hospitalizations.

I understand what would happen for routine check-ups and serious accidents.

How would such a model handle predictable-but-expensive medical needs, like hip replacements for the elderly and monthly prescriptions for HIV and cancer patients?

The hip replacement can be covered under another policy like a home warranty. At the same time, without coverage, the prices have to come down. If they don't, all of those expensive doctors are bringing in near $0. Such an economic position is untenable.

The other two are probably going to be fixed through direct market forces. If the people of America can only pay $X, then the makers of the drugs can only charge $X + small premium to have a market. Since they would face a similar price pinch abroad, they'd have to adjust their prices.

We've seen the inverse of the price pinch with the epi-pen scandal. The makers knew the price could go up since insurance would absorb the cost.

Hiv cost over a lifetime is only 340k and 1:300 in the US has HIV, so the subscription would only have to cover a little over a 1k in lifetime risk; actually less, since HIV is overrepresented in the homeless and poor segments.

In either case, two components that could be separated: a service plan (preventative care) and catastrophic coverage.

I think the premise is that the cost savings for maintenance and low complexity care make the risk of paying for emergency care worth it? Which I'd love to see numbers on.

I imagine any chronic condition would almost certainly rule this out as presented.

Models where a provider network gets $X a year to handle a chronic condition patient are one of the few bright spots in recent health care policies.

Ten years back we may be looking at high deductable plans as a disaster -- if you have the money to fill an HSA they are great, but for most people in the U.S. an HDCP means that diabetics will not take metformin or insulin but instead they will get their feet amputated at much greater cost.

One clear example is asthma. Steroid inhalers can often enable an asthma patient to breathe freely and avoid hospital admissions for about $200 a month. That looks like a crazy high price, but a hospital admission is going to cost $8000 or more. In the fee-for-service model the hospital profits from amputations, asthma exacerbations, etc. In a flat fee model they make a profit from practing good medicine.

According to tfa you're still supposed to have high deductible insurance (thus defeating the purpose?)

It depends on what your purpose is. It is at least debatable that one of the major problems with our current system rarely addressed by anyone on any side is that it conflates a health care plan with a health insurance plan. The two things may with some justification be the same thing inside of your head, but in monetary terms, they have radically different statistical distributions of monetary payouts required, one being a fairly reliable steady drip of relatively small amounts of funds, one being a highly irregular payout of large amounts of funds. Only the latter is something that "insurance" is a solution to; if you're getting a steady drip of low-payout events the only thing insurance can do for you is add a middleman and raise net prices as a result. (Some individuals may win and some may lose, but it has to come out net more expensive for the system.) Trying to bind the two together into one thing may be part of, though probably not the entire, problem.

Edit. I more meant that almost all insurance is hoghbdeductible insurance these days and it's extremely expensive for what you get.

If you could buy catastrophic insurance with a plan like this then That would be awesome.