Alternative payment models and precision medicine
Another approach is to try to target treatments to the specific condition the patient has to see what works. For example, when a cancer patient doesn't respond to the normal course of treatment, one thing that has been done in a few cases has been to take samples of that cancer and give it to lab mice, thus creating a mouse model of that specific patient's cancer on which they can test a variety of different drugs to see which, if any, the cancer will respond to. In some cases, we get a hit; one of the drugs that is not normally standard for the type of cancer does work in the mouse model, and with a little luck, also works on the patient. This wouldn't work without the animal model--you can't just administer a whole bunch of highly toxic drugs to a patient until something works. Not only would the patient be harmed, but lack of proper randomization and controls means you can't really be sure what if anything worked.
So there's a promising future in using precision medicine to help patients that don't respond to standard treatments, and potentially to help reduce the risks to patients as well. However, creating an animal model is pretty expensive and may not even yield any useful results. Studying the animal model doesn't exactly count as rendering treatment to the patient, and is inherently experimental since it literally involves running an experiment for each specific patient. The question is, who pays? This is beyond the scope of traditional insurance reimbursement. So far all examples of this technique have been funded entirely through research grants and institutional sources. If it does turn out to be as effective as initial results suggested, how will we scale this?
Traditional insurance has no incentive to fund such a procedure. Not only is it experimental, possibly won't even yield a result that would affect the course of treatment, and expensive, but it also could extend the lives of specifically those cancer patients who are most likely to relapse. I hate to be so dismal, but financially, that actually penalizes the insurer that tries it.
It seems to me that precision medicine requires an alternative payment model, one that rewards insurers for outcomes and encourages them to take risks on funding the development of individualized treatments for each of their most challenging patients. I'm not sure ACOs fit the bill. ACO payment models are designed so that providers and insurers share the benefits of any cost savings achieved while caring for a given population--but precision medicine techniques like the one above, while they could reap huge benefits, almost certainly increase costs, and quite substantially so.
With out some fresh ideas on how to reform insurer reimbursement models, I doubt many of the best precision medicine ideas will ever survive outside the research centers that dreamt them up.
I'm quibbling over a tangential detail, but the mouse model doesn't sound inherently "experimental," at least not in the insurers-being-unwilling-to-pay-for-unproven-treatments sense. The procedure that you described sounds almost exactly like standard of care bacterial susceptibility testing. If these mouse cancer models were established as effective, I don't see why insurers wouldn't be expected to pay for them.
I'm quibbling over a tangential detail, but the mouse model doesn't sound inherently "experimental," at least not in the insurers-being-unwilling-to-pay-for-unproven-treatments sense. The procedure that you described sounds almost exactly like standard of care bacterial susceptibility testing. If these mouse cancer models were established as effective, I don't see why insurers wouldn't be expected to pay for them.