• Hospital-insurer integration

    Sarah Kliff’s post on a Massachusetts chain of hospitals offering a health insurance product is a must read.

    Right now, health insurers and hospitals are separate entities. This is something different: a hospital selling coverage, with the restriction that  you only seek care at its locations. The hospital chain is Steward, a Massachusetts-based business that has been buying up community-based hospitals for more than a year now. And the gamble it made was rolling out a health insurance plan that almost-exclusively serve its hospitals (there are a few exceptions). The plan, administered through Tenet Health Plans, is called Steward Community Choice.

    Why would a hospital chain integrate with an insurer? Some possible answers, several speculative:

    • To control referrals, capturing a larger share of patients than it would otherwise
    • To raise the costs of other insurers by charging them more than the integrated product for access to the hospital chain
    • To offer a more price-competitive product by cutting out the administrative costs of hospital-insurer negotiation and streamlining coordination
    • To avoid the challenging negotiations with other large and powerful insurers (like Blue Cross/Blue Shield)

    The rationale given in Kliff’s post is to cut premiums by restricting choice. This doesn’t make sense to me. If that worked, why aren’t insurers doing it already? Put another way, severely restricted networks can exist without hospital-insurer integration. So why integrate? There’s got to be another reason.

    Of course the premium reductions suggested by Steward–20 to 30% below other products–could be an attempt to gain market share with unsustainably low prices. If that’s the case, they won’t last.

    Am I missing anything?

    • Insurers don’t restrict because patients hate it. Hospitals stand a better chance of convincing customers go accept restriction because they do it indirectly by referral, rather than the insurers option of baltantly and obviously limiting choice.

    • “The rationale given in Kliff’s post is to cut premiums by restricting choice. This doesn’t make sense to me. If that worked, why aren’t insurers doing it already?”

      Some insurers have tried it, but it is tough to market a product that doesn’t give you access to all of the providers in the area. The dominant providers are the expensive ones, and those are the ones people want access to, so it’s a bit of a catch 22. Restrict access to them and no one wants to buy your product, restrict access to providers that people are willing to go without and you haven’t saved much money.

      I’m working on something like this for an insurance client right now, it’s hard to find the right balance when meaningful savings and a marketable product are competing with each other.

    • Selling insurance might improve the hospital’s cash flow, a minor advantage, unless there is something the hospital wants to do that requires the financials look better on paper.

    • There’s also the point that if insurance companies actually makes money, then doing the insurance in-house means that that revenue stream now accrues to the hospital, not the insurer. (This is my objection to the very existence of insurance companies in the game at all showing (Japan gets this one right).) Also, how about streamlined paperwork: apparently paperwork is a royal pain that costs doctors significant amount of time (and thus money) and every insurer has different forms, making it really hard. (This is my Japan comparison thing showing (again); over here there’s only one payer, so the forms are uniform, although still a major pain, my in-laws tell me.)

      These are very close to your 2nd and 3rd points, so you aren’t really missing them.

    • Austin
      You covered the bases.

      As I read the deal on Friday, the premium cuts struck me as Draconian and signaled a loss leader type of strategy. Whether its Partners or Tufts or whomever, there will be 3.2 outpatient visits, 4 prescriptions, 0.5 hospital stays per capita, etc. You don’t get rich with a 30% discount (or stay in biz for that matter) when your competitor down the street doles out the same actuarial service equivalent.

      Cerebrus has pumped in several hundred million dollars to this chain. You need to build volume and solidify you lines of business—bypass, ortho, ob, etc.

      This is the protypical new candy shop on the corner giving away free Milky Way bars to bring the customers in.

      Build your team, buff your stats, and then move from the double A to triple A ball. The major leagues are in the viewfinder for sure.


    • Yes, you are missing something.

      By having a Hospital Group offer insurance which requires the insured to use the Group’s resources, it removes the incentive for health care providers to increase costs. Assuming that the health care services are those largely employed by the Hospital Group, there is no incremental cost to the insuance company when the insured seek services. Furthermore the incentive is to reduce costs, since cost savings remain with the Hospital Group/Insurer.

      This is very good trend, (see Pittsburgh’s similar move) and can ultimately lead to cost control by the health care system. We need to get health care providers out of a system of being fee based entrepreneurs and into a system where they are fixed cost employees.

      There is the problem of choice, but Americans are going to have to accept the fact that choosing fee based health care providers simply encourages the system to raise costs, and such a system is not viable in the long term. Heck it might not even be viable in the short term.

      Think about Medicare. If Medicare gave a hospital group that had all the resources to provide health care a fixed fee to cover Medicare patients in that area, and there was competition then health care costs would stabilize and might even be reduced.

      It’s called basic economics and finance, get the incentives right and the market, with regulatory oversight, will take care of itself.

    • A lot of the rationale would be found in the extent of control the hospital has over its service providers. If a large percentage of Doctors are direct employees, then there can be restricted spending that would not be noticed in care delivery. Simply set normal protocols and algorithms for each diagnosis, which means costs are known in advance.

      Being both the provider and the payer of health care gives a lot of ways to save money. Accounting and billing are enormously simplified. There can be a real budget, with an actual amount of money to draw upon. Instead of the current vague guesses.

      Suspect this makes more sense to accounting and finance people than to others.

    • @dalea

      You are correct, I am a finance person, and health care economics is a finance and economics issue. Treatment is a medical issue.

      Much of this is a question of incentives. If health care providers can make more money by having a high cost health care delivery system, then the health care delivery system will be a high cost one. This is what we have now. HIgher costs are higher income to providers.

      If you have a situation where the health care providers are also the insurers things change. Insurers get a fixed payment. This means that if the providers are the insurers, providers get a fixed payment. In this situation more money can be made only by decreasing the costs of providing care. This is what the industry seems to be slowly moving towards and what government policy should encourage.

      There is no free lunch. There are disadvantages to such a system. One is that care may be deficient. This problem can be overcome by competition and regulation. A second disadvantage is the issue of choice. This problem can be ameliorated to some degree by education, by choice within a large organization and by acceptance by the public that choice may just be too expensive.

      Those who provide health care are healers, they should not be entrepreneurs. The system should incentivize them to provide the highest possible care at the lowest possible price. The action in Massachusetts is step in that direction.

      • Thanks, David R. I always enjoy reading about how incentives affect the market. My understanding is that part of the VA’s success is due to the fact that it only has a certain level of funding, so the incentive is to keep patients as healthy as possible rather than being incentivized to see patients as often as possible for treatments that bring in the highest reimbursements. And yes, the doctors are salaried.

        I haven’t done the reading, and I’m not a professional, but isn’t there a risk of adverse selection? I think it would be lower, the longer health care in the state is “mandatory”, but it seems like a risk.

        It will be very interesting to see how these experiments play out in the real world.

    • How is this different from Kaiser? I worked for Sharp in San Diego for a while and they were trying to do the same thing with a 300+ multiple-specialty group and a couple of hospitals. They offered their own HMO product. I’m sure there are other similar examples.

    • Between Austin and the commentors, I think all the major bases have been covered. However, I think there is a misunderstanding on one point. Steward can’t be seriously proposing to be a staff model HMO, with providers in the network exclusively working with the in-house insurer, can it? I assume Steward’s hospitals will continue to accept other insurers.

      That means not just that comparisons to the VA and Kaiser are a stretch, but that administrative savings are likely to come only from marginally improved cash flow and transfer of net income that would have gone to the external insurer for that fraction of their business. There won’t be additional savings for administrative simplification, because the hospitals will still participate with many insurers and insurance products for the foreseeable future.

      I do think this trend of hospitals forming insurers and insurers buying hospitals is going to continue.

    • This is very close to the Kaiser structure. And my understanding is that that Kaiser has been successful.

      One needs to understand the basics of insurance. Insurance is risk shifting. In health insurance it changes an unknown and unlimited variable cost into a known fixed cost for the insured. A family of four that pays, say $10,000 per year for health insurance knows that their health care costs will be $10,000.00 per year (assume away deductibles, co pays, incidental expenses for the momment or consider them small enough to be ignored for purposes of argument).

      The problem is to whom is the risk shifted. With conventional insurance it is shifted onto an insurance company that hopes through covering a large, statistically independent base that it can diversify away the risk (statistical variance). The problem is that the providers have an incentive to constantly increase costs to the insurer, because that is the only way in which they can increase their income.

      By combining the insurer with the provider you cause the insurer to set up a structure whereby the incentive to increases costs is replaced by an incentive to reduce costs.

      Really, it’s that simple.

      We know what to do, we simply lack the political will, the education and the ability to change an outdated system.

      • There are lots of hospital systems that own all or part of an insurer. The majority, perhaps a large majority of Steward hospital patients will be getting their insurance from other insurers. Also, Steward probably isn’t a true integrated delivery system, but relies on lots of independent community docs to feed patients to its hospitals. This is more like Geisinger than Kaiser.

        These levels of provider and insurer integration are on a continuum. You compare the Steward move to Kaiser because that is your vantage point from a distance, but as you get closer the difference become more salient.

        Another point on the continuum: health care systems that accept global capitated payments from insurers and become responsible for all care (preventive, chronic and acute). They don’t own an insurer, yet they take on risk just as surely as if they did and have very similar incentives.

        • Do you, or anyone reading this, know how I can find all (or many) of the hospitals that own all or part of an insurer? Or, conversely, insurers that own all or part of hospital systems? How would I go about finding other examples and learning about them?

          • I don’t know of a listing specific to hospital ownership of a health plan. Is that what you want, or is it organizations in which an insurer and hospital are wholly or partly under the same ownership?

            Cases where a hospital (system) existed first and then created or bought a health plan are pretty limited. Mayo has done it. Geisinger may have started that way. More generally, a lot of hospital systems in the 90s tried it in response to HMOs, but almost all the commercial plans failed. Some Medicaid plans have survived. In New York, for example, there are Metroplus, Healthplus and Healthfirst, all owned by hospitals and together insuring over a million New Yorkers in Medicaid, Medicare and other government programs.

            Almost all of these cases have an in-house insurer that is only a fraction of the total managed care reimbursement for the hospital.

            You may also want to look at old-school HMOs like Healthpartners in MN and Group Health in WA. I believe these started as physician groups that offered insurance, then purchased or built hospitals.

            You can find a bigger list by going to the ACHP website- America’s Community Health Plans.

          • Oops, I mangled the name. There are a few more of the organizations you’re looking for on this page. No idea where to go from there, other than strategic Googling.

    • Is it possible Steward can qualify as an ACO and benefit from provisions in the ACA? I used to belong to Kaiser Permanente when I lived in their area, and it seems Steward is doing something similar.