"Despite recent advancements in medical science, the mortality rate remains stubbornly fixed at 100%"

If a person has diabetes, and is given proper preventative treatment, the argument here is that that person can be removed from the list of individuals who will eventually need major medical care at some point.

Unfortunately, this is not the case. Ignoring accidental death, homicide, and suicide, everyone will eventually get sick and die. Thus the individuals receiving preventative care can only be removed from the list of individuals likely to have major complications stemming directly from their diabetes. Thus, one day when the big C rolls around, or heart disease, or whatever else strikes the unfortunate patient, major medical bills will eventually ensue. This is the case for EVERY individual! The major medical bills are never avoided.

So the real question is this: Given that everyone will eventually have major medical expenses, is it profitable for the HMO to provide preventative care in the interim? In other words, can the HMO still profit from that individual while paying for preventative care? If the answer is no, then the solution for the HMO is to withhold that care and either force them to switch or let them die sooner rather than later. If the answer is yes, then it's a dilemma indeed.

The data point you are missing is that 80% of all medical expenses occur in the last year of life. Combine this with the fact that MOST health insurance plans COST MORE than just straight out-of-pocket expensing of healthcare and you find your profit motive

Thus as long as an insurer can reduce, or more importantly DELAY the onset of that last year of expenditure, they can get that much more years of ROI and premiums out o fthat individual. Furthermore, since Medicare kicks in at 65, if an HMO can push that onset to past 65, they are even in better shape on ROI

There is another analytic flaw in the Diabetes example. Assuming (which is a bit of a stretch here but useful statistically) that every plan paid for insulin pumps - then it would NOT mattter if Plan A paid for Person W's pump only to have Person W move to Plan B in 2 years. Why? Because assuming Plan A and Plan B are competitive in their offerings, their market shares will stay relatively stable. That means for every Person W with diabetes moving from Plan A to Plan B, there is another person moving from Plan B to Plan A.

If both Plan A and Plan B pay for insulin pumps, the each gets to recoup the savings on the investment by the other.

Now of course in real life it isn't that simple, the flows aren't always equal, market share isn't always stable and not all plans cover diabetes.

Yet another reason for Single Payer.