Knowing What Not to Do

“The essence of strategy is choosing what not to do.”  Michael Porter.

It is so often the case that organizations try to do things they should not do.  Call it irrational exuberance; getting out in front of the curve; or a bridge too far.  Hospital systems are examples of that.  Already large, complex organizations doing incredibly challenging things with billions of dollars flowing through their systemic blood vessels, they are understandably tempted to do more.  They always are.  That is the inevitable urge of active hospital board members and ambitious executives.  Do more; not do less.  After all, who arrives to such an exalted position to do less?

Their collective corporate eye is cast toward health insurers who have been called bloated and inefficient; dinosaurs; dim witted at best.  The President of the United States, no less, disparaged insurers while promoting the ACA, labelling them the “villains” of the healthcare system.  Speaker Pelosi called them “immoral.”  How difficult can it be to do health insurance better than the insurers have done it? Should be easy for people as smart as those who run complex healthcare delivery systems.

“Hospitals think this is a way to cut out the middle person, tailor care more closely and save a lot of extra money, but there’s a history to this and it generally doesn’t work,” said Howard Berliner, a visiting professor of health policy at NYU. “It sounds easy, but it winds up being incredibly complicated.”

Well, it isn’t easy, and although there are exceptions such as Kaiser Permanente based in California and Geisinger Health System in Pennsylvania, the road is littered with the detritus of hospital systems which have tried and failed. 

The December 16, 2016 edition of the Wall Street Journal carries an article in its Business News Section on this subject.  Titled Hospital Firm Retreats as Insurer, it is about Catholic Health Initiatives, a 103-hospital system located in 18 states.  CHI bought a managed-care plan looking to market coverage using its network of hospitals and physicians for care.  However, this nascent insurance business apparently lost $110 Million for the fiscal year ended in June, and CHI has declined to say whether it was continuing with insurance or not.  We know what that usually means.

The article goes on to mention similar challenges faced by Tenet Healthcare based in Dallas which expects to exit the insurance business, and Piedmont Healthcare which, after less than two years and steep losses, shut its insurance arm down.  The article also describes the less than rosy experiences of Ascension Health in St. Louis, and Northwell Health in Great Neck, NY. 

There are others.  Closer to my home is Partners in Boston, which in 2012 acquired the HMO Neighborhood Health Plans (NHP).  NHP uses Community Health Centers to provide care primarily to Medicaid beneficiaries and the indigent via the MassHealth Connector (the Massachusetts Exchange or so-called Romney Care).  One can imagine the visions at Partners about rescuing this financially challenged but highly rated insurer (which to its credit it did), establishing itself in the ever-growing Medicaid-via-Exchange world, and slowly branching out with a little traffic-flow-directing on the way (which it also did).  However, since 2014 NHP has lost $241 Million, and announced it was no longer accepting new MassHealth subscribers, at least for now.

UPMC (University of Pittsburgh Medical Center) has reportedly done well with its insuring arm, insuring 2.6 million members mostly in Western Pennsylvania, right in Highmark Blue Cross’ back yard.  [An aside:  Two of the most prominent downtown Pittsburgh buildings are those of UPMC and Highmark, which tells us something.]   UPMC’s insurer trajectory, however, has been a long one, starting back in the early 1990s with success coming only recently as the significant protracted legal and other battles with Highmark start to wind down.

There are entirely too many other cautionary tales of hospital systems venturing into the health insurance arena and exiting battered and beaten.  George Santayana, in Vol. I in Life of Reason (1905-06), which is aptly titled Reason in Common Sense, wrote “Those who cannot remember the past are condemned to repeat it.” And indeed, this all has happened before as late as in the 1990’s, when a number of large hospital systems made disastrous forays into the insurance world via HMOs.

So, what is going on here?  Why are these very bright capable executives continuing to repeat past decisions which so often have led to financial catastrophes for hospital systems? 

First of all, it is understandable that hospital systems might think they can do insurance.  After all, they already are taking on more and more risk, whether through the various CMS ACO programs or via risk-based or value-based contracting with private insurers.  The thinking goes like this: how much more difficult can it be to insure if we are already taking on substantial risk?  Isn’t insurance all about the assumption of risk?  Let’s cut out the useless middleman (commercial insurers) and capture the margin for ourselves.  And there are commentators out there actually encouraging them to do just that.

Yes, but…

Modern Healthcare in a 2015 article outlined the challenges hospitals faced doing this, and their track record.  It highlighted Premier Health’s activities (a $1.9 Billion Catholic Health Initiatives 5 hospitals system in Dayton, Ohio), quoting one of its executives:

“For us, the insurance business is just a vehicle to cover as many lives as we can in our service area with our population health initiatives…. We’re not out to be one of the large national players in the insurance market.”

Uh huh.  “Just a vehicle.”  The Modern Healthcare article goes on to quote a consultant as saying that once hospitals “get into product design and pricing and all of the nuanced areas, you can get into trouble reasonably quickly.”   Indeed.  And the recent WSJ article about CHI seems to bear this out.

Hospital systems often start out self-insuring and self-administering their own employees, just as Premier did.  Then, after just a few years with nothing remarkable happening, some executive looking to add value suggests “modest” expansions.  And they get sucked in.  “Just a vehicle.”

Although current thinking, largely by providers, is that health insurers are not a value-add in the delivery of care world, they lose sight of what insurers do and how complex and necessary it is to do it.  First of all, health insurers were first created in the 1920’s and 1930’s by hospitals (the first Blue Crosses) and physicians (the first Blue Shields) to increase the odds of actually getting paid for services rendered during the Great Depression.  And for many years, the Blues were creatures of those providers who dominated their boards of directors until anti-trust concerns changed that in the late 1970’s.

Health insurers are pulled in many directions, owing obligations to many constituencies:  the insured members whose care is covered; the employers who pay most of the freight; the providers which demand “fair” reimbursement; the healthcare “system;” and the public on many nuanced levels.  If they are for profit, their fiduciary duty is, of course, to their shareholders.  If nonprofit, their fiduciary duty, while more diffuse, includes financial stability via strong reserves. 

Claims processing used to require many, many employees with decades of hands-on experience.  Now it mostly requires legacy core processing systems that are incredibly expensive and nightmares to upgrade and replace. 

Insurers create and maintain networks of providers which require a host of activities and oversight.  Negotiating fee discounts is merely one such.  Insurers manage thousands of contracts, moving money in myriad ways; contracting is challenging, nasty, complex, and risky; the actuarial and underwriting activities are not for amateurs or the faint of heart; and it goes on. 

Moody’s Investors Services suggests that the biggest operational challenge hospital systems face in undertaking insurance functions is the lack of the specific expertise and skills needed to effectively run an insurance plan, including actuarial and underwriting practices, marketing, customer service, IT, and the maze of compliance needed. 

As Moody’s states, the skills needed for insurance are profoundly different from those needed for the delivery of acute care.  “Not-for-profit hospitals with a health insurance business … tend to operate at noticeably lower operating cash flow margins than similar health systems without insurance,” says Moody’s Analyst Mark Pascaris. “Entering the insurance business inevitably suppresses hospital system margins from the beginning.”

Why is that?  Because hospital financial success in today’s reimbursement world is based on maximizing treatment volume, while insurer success is based on just the opposite.  This is an inherent conflict that perhaps someday can be mitigated by payment reform and ACO-type models.  But not yet; not today.

Then there is scale.  Without it, insuring is almost always doomed to failure.  While I am sure there are exceptions, in my judgment, the lowest member population a commercial insurer can reasonably scale to is about 200,000.  And they should never, ever try to build their own claims processing module from scratch.  That way lies madness and costs hundreds of millions of dollars.  If starting out relatively small, they must use a very efficient, bare bones TPA claims processor which charges by member month.

AHIP (America’s Health Insurance Plans) tells us that if providers want to get into the insurance business, they have to be willing to deal with many complex government regulatory requirements including maintaining hefty reserves and paying the ACA’s health insurance tax.  This ties up huge amounts of available hospital funds, usually a scarce resource.

So this is all about scale and skill sets.   If a delivery system wishes to go this route, I would suggest that it has to be prepared for a large-scale merger with a significantly sized insurer with existing adequate reserves, a large employer-based book of business, and existing expertise.  Those sorts of insurers usually are not in play precisely because they are reasonably healthy.  And, yes, for that very reason, acquisition or merger will be very expensive; and there may be some question of who controls the resulting merged entity.

How about buying a failing insurer on the cheap?  That can work, but it seldom does.  A bad medical loss ratio is wicked difficult to turn around, and the Partners/NHP example discussed above suggests that it is not the optimum model.

The point of this article?  I think some day it will be easier to combine the delivery and financing elements under one roof.  But that can happen only when reimbursement methodologies have changed to such a paradigm extent that the delivery organization succeeds financially ONLY when it reduces the rate of utilization of care per covered life by making them healthier.  At that point, the care and financing arms are financially aligned.

So, the cautionary tale is obvious.  And yet, I have little doubt that even after reading this article, there will be a hospital executive or two that will be unable to resist the siren song temptation to dip their toes into these waters, thinking they can do it better.  Maybe they can, but the odds are against them.

Categories: Uncategorized

23 replies »

  1. You are right, Jim. The word ration creates all types of hysteria. But we should not become confused. The hysteria is not over the fact that rationing takes place, rather who does the rationing. Everyone rations their time , money and even love.

    Do you know who knows how to ration? Sure you do. Physicians. I have had to ration care for ICU beds, Blood products etc. and even had to look the patients in the face when doing so. That was back in the old days when these things were in short supply. However, when I rationed an ICU bed or blood product they went directly to another patient more suited for the treatment. They weren’t banked as profits to be divided later between profit and care for other patients. The profits were and are for those few that made the decisions, a most definite conflict of interest. Such are the new ethics of medicine that sometimes create heinous reasons for denying care.

  2. Do you understand that by using the word “ration” you have already radicalized the conversation? We have to, ummmm…limit care. The burning issue is how to do that.

  3. Not important, but perhaps you lost your place.

    I said “Wow, do you really think anyone on this blog believes CEO’s shouldn’t be well paid?” You seem to be responding to a different posting.

  4. Actually, I’ve seen numerous people on this blog attribute high insurance premiums to high CEO compensation at both insurers and hospitals. It’s wrong of course. I’ve heard others opine that hospital CEO’s shouldn’t make more than $1 million per year and some have suggested as little as $500K. That’s not where the market is. There are single payer Medicare for all advocates who comment here though not necessarily on this particular thread.

  5. “The trend toward value based contracts could push providers in a more cost-conscious direction”

    I assume you understand that is one way to ration care. I also assume you recognize that when you push physicians there has to be something behind the ‘push’. What is it that you advocate should be behind the ‘push’ assuming you don’t believe a ‘push’ happens all by itself. You do realize that pushing can lead to physicians doing the wrong things for their patients,right? That is why I like Jim as the insurance CEO, or his employees, dealing directly with the patients. That way the patient might have a say in what gets pushed into or out of his life.

  6. Thank you for the correction. I am sure Jim was as good a CEO of an insurer as he might have been as a CEO for a hospital.

    ” resulted in numerous scribes being hired”

    Duh, well yes, those scribes add administrative costs, but do the programs improve quality, provide better access, and provide lower costs ?

    “have to compete for executive talent”

    Wow, do you really think anyone on this blog believes CEO’s shouldn’t be well paid?

    The graph demonstrates something even if you don’t want to avail yourself of that information. There is barely any growth of physicians, but a wild growth of administrators. Many of them aren’t, as you say, twiddling their thumbs, rather dealing with the ACA that you strongly supported (along with other rules and regulations), so by definition they are working , but there is very little utility in a lot of that work. That is where a lot of problems lie.

  7. I wonder about the role of physician practice patterns and the extent to which those could move in a more conservative direction as doctors become more aware of the cost of the treatment options they recommend even when the patient doesn’t bring up cost as an issue important to him or her. The trend toward value based contracts could push providers in a more cost-conscious direction, especially with respect to being more aware of who the most cost-effective high quality providers are and then referring more of their patients to them.

  8. Jim was the CEO of an insurer — Rhode Island Blue Cross, not of a hospital.

    The increasing documentation requirements that doctors have to adhere to because of government regulations and other actions resulted in numerous scribes being hired at around $50K per year to assume some of that administrative burden now associated with treating patients. That adds to administrative costs.

    There are lots of other things that have driven up hospital costs over the years from stricter building codes to the trend toward single rooms to house inpatients to investments in new technology like proton beam equipment to treat prostate cancer.

    There aren’t armies of executives twiddling their thumbs or shuffling papers in the C-Suite of hospitals. Hospital CEO’s are paid pretty well because hospitals have to compete for executive talent in a national market and if they want to attract competent talent, they have to pay the going rate. That’s the free market at work at least with respect to executive compensation.

  9. Our assumptions as to what constitutes an administrator vary. I assumed a wider group for if it were narrower I believe the graph would portend even a worse environment.

    We need doctors and hospitals that require a certain amount of administration to carry out their tasks. The quesiton becomes, how much of the rest are needed and how such an increase is able to give an increased dollar value for services? Based upon my own andecdotal experience we are experiencing the same thing as other large bureaucracies. Higher costs and lower dollar value.

    My overhead when I started practice was around 33% and increased to almost 75% just prior to retirement which is difficult to evaluate since I was cutting down. Somewhere in the 60’s would probably be a better estimate. Capital investment had ceased so I surmise that most of that increase was due to administrative aspects of the practice and government.

    I don’t know when you were the CEO of your hospital, but it would be interesting to compare the number of adminstators then and the number that exist now unless of course you recently retired from that job.

  10. Fifty years ago, the leadership of the Tennessee Valley Authority learned that the leadership and governance for a fossil fuel/hydroelectric power plant wasn’t applicable to a nuclear energy power plant. It still isn’t. The Omaha insurance company, Mutual of Omaha, learned the same lesson about 20 years ago: indemnity and health insurance are not the same.

  11. I assumed “administrators” referred to the non care delivery executives in institutional providers–not insurers. Could be wrong. On the insurer side, there are far fewer insurers today than in 1970. The Blues went from over 100 plans to less than 40. Consolidation has reduced the number of insurance company executives. With hospitals, consolidation does NOT seem to reduce the number of executives. There are, of course, exceptions, but typically even after an independent hospital is acquired by a system, it’s administrative staff and executives remain in place. The other reason there are many more administrators is that insurers and governments require providers to do so many more, sometimes inane, things requiring a huge back office. That’s my guess.

  12. The rate of general inflation, CPI increases, measure price increases in goods and services. Dozen eggs, quart of milk, gallon of gas, etc. Health insurance increases say 3-4 times the rate of inflation because it has three components. Its price component is fee increases given to docs and other providers. Then there is “mix” or the relative expensiveness of services and procedures (MRIs rather than X-rays; miracle drugs rather than aspirin). Then, and this is the killer, there is the rate of use of services per person per year. That is called “utilization.” That during the last two decades probably has averaged close to 10%. When you combine all three, it’s not surprising that increases are double digit.

    Increases have moderated over the last five years. No one knows why. Some pols claim it’s because of the ACA. Nonsense. The ACA increased costs. My view is that the cost of healthcare is almost like a living organism with cycles of ups and downs. When it was “threatened” by Hillarycare in the 90’s, increases moderated. Might have been the HMO movement too. Absent a change in structure, increases will increase again.

    We do have an unsustainable product, particularly when there is the mindset that we cover everything at any cost. And live increasingly unhealthy lifestyles. There will, at some point, have to be a very disruptive systemic change in what we cover and how we cover it.

  13. Terrific post, Jim.

    One burning question that I have for an ex-insurance CEO is this: Even the best health insurers have had to increase premiums over the rate of general inflation since time immemorial (at least, it seems that way—am I correct?).

    Doesn’t that indicate that health insurance is fundamentally an unsustainable business model that has managed to stay in existence only because it’s an addictive and self-promoting service? In other words, once health insurance was eased into existence on a widespread basis by the Wage Stabilization Act of 1942, it became nearly impossible to keep things under control. Demand for healthcare skyrocketed (due to moral hazard) putting relentless upward pressure on prices and premiums.

  14. “My own perception is that most of these people are probably in jobs related to IT and regulatory compliance but I don’t know for sure.”

    I would bet that is true, but are they treating patients? No, so perhaps we are bloated with administrators as one might surmise just looking at the graph. But, perhaps one should look even further and see what caused this possible bloat. Where does this increased so called need of administrators come from? We know some are needed to run doctor’s offices and hospitals, but so many? Then again there are a lot of administrators working for the government. Perhaps the root of the problem is not the hospital, but government which has created a bureaucratic nightmare that hospitals, insurers and doctors are forced to live with. Good hospital administrators want to make money, not lose it so I doubt they are intentionally bloating the system with administrators though I am sure in many instances that the administrators at the top are paid extermely well..

  15. Since Jim is a former CEO of Rhode Island Blue Cross, I think he probably has a better understanding of how administrators are defined, what they do in hospitals and why the number has increased so sharply, at least in percentage terms, since 1970. My own perception is that most of these people are probably in jobs related to IT and regulatory compliance but I don’t know for sure. I would love to hear Jim’s insights on this subject including what impact the sharp increase in the number of administrators since 1970 had on the total cost of running a hospital as measured by cost per licensed or occupied inpatient bed.

    As for insurers themselves, they have been working hard to drive their administrative costs down as a percentage of premium revenue primarily through automation. In terms of the medical cost trend, United estimates it at 6% plus or minus 50 basis points for both this year and next with two percentage points of that attributable to increased utilization of services and four percentage points due to higher prices for services, tests, procedures and drugs with outpatient hospital services and prescription drug prices increasing at a high single digit percentage rate while physician fees are growing at a comparatively modest 3%.

  16. In his response to Barry, I think Jim was correct in his change of focus from physician fees to the rate of utilization of services But, to that cost center I think we have to add still another, administrators. I’ll let Thomas Sowell explain.

    “It’s amazing that people who think we cannot afford to pay doctors, hospitals, and medication, somehow think that we can afford to pay for doctors, hospitals, medication and a government bureaucracy to administer it”

    There is a graph that shows the growth of physicians and administrators from 1970 to 2009. It is a shocker.


    Dr. Kevin explains further “While the numbers of physicians that are entering the workforce have trended toward a constant number (with little or no growth) the numbers of administrators has risen nearly 3000 percent over the last 30 years.”


    I don’t want Jim to assume that I find no value in administrators. I do and am sure that as an administrator himself, he was absolutely necessary and worth every penny he was paid. But, look at that graph. Administrators don’t diagnose or treat disease. Doctors do, yet the growth of administrators is phenomenal while the physician workforce is almost constant.

  17. Excellent post, Jim. Simply great. I’ve worked mostly on the insurance side of things in New York for over 12 years, and have long thought about the barriers to success on the creation of truly integrated health systems that combine all the pieces of insurance, primary care, specialty care, hospital, and various ancillary services and facilities. The problems are no less severe when you start from the perspective of an insurer than a hospital or IPA.

    In addition to all the good points you make, I would emphasize that a critical mass is needed in a regional area. It’s not just that you need a minimum of 200,000 lives, but that you need the provider locations grouped tightly so that the network created by the merged entity is adequate enough to be attractive as an insurance product. Partners did that, but arguably the Catholics did not. Starting from a health system is going to be very hard in most major markets for scale and network adequacy reasons.

    But on the other side, starting from a health plan, you’ve got the problem that only a fraction of your members were visiting the hospitals and physicians that you merge with. What happens to the other 70% (to make up a number)? At worst their preferred providers are out of network now, and those members are going to have a terrible retention rate, eating directly into core revenue. Or, there will be a tiered network and the 70% are going to be unhappy paying higher out of pocket costs and also leave, though probably at a slower rate. Or, the new plan/provider entity will eat higher costs on the 70% in order to not lose membership, but that’s not a tenable situation for long. It’s very hard to turn the corner and have enough overlap between the merged provider patient base and the health plan member base to not experience huge disruption at the point of merger. And that’s putting aside all the ego conflicts and culture conflicts between the merging entities.

  18. Great article, Jim.

    We all believe we need health insurance because the yearly variance in our medical costs can easily wipe out our bank accounts and cash. If this were not true we wouldn’t need insurance.

    But it seems also true that the great defect in health care financing is that of the third party payer. No one knows what real prices are. No one knows what real marginal costs are. We all feel that we are dealing with a money-bags Santa insurer and everyone charges too much.

    Thus, we need the health care providers believing that their fees are coming from the patient and not from some rich insurance company. And the patient and the provider and the doctor and everyone need to know prices.

    And yet, we must be able to be altruistic to our unfortunate folks; this has to mean money transfers from better-off people to our less-better-off people. Viz. taxes.

    But some of us would never trust a federally run system with the ability to control 1/6 the economy.

    Accordingly, I cannot see any way to satisfy all these apparently contradictory needs other than to use some kind of voucher or medi-buck system and have it run as a utility by the county or state or hospital district.

    You could start with hospital non-ambulatory care first. Make hospital health care a public good and forget the insurance aspect of all current care.

    Just think of the relief it would be to eliminate all billing functions from the entire non-ambulatory health care sector.

    At first, we pay no attention to ambulatory care. Leave it up to the people to do with anyway they wish. Disorders that might kill us, or disable us, or cause us bankruptcy are best handled as public goods. And these are almost always hospital managed diseases.

  19. Barry, I’ve not been “inside the tent” of a staff model HMO or combined insurer/provider, so the quick answer is I don’t know. I’ve come to believe that the focus on fees is misplaced. It’s the rate of utilization of services and which services that determines the cost most. So, I’m expecting that the fee levels themselves aren’t all that important. That’s my surmise.

  20. Excellent. The major problem we face with insurers is that they are forced to satisfy governmental whims which prevent them from adequately performing a task in a competitive environment that insurers have been doing for millenia.

  21. Great essay Jim. Estimating actuarial risk isn’t so easy. On the other hand, losing lots of money like Medicare does is a breeze. Maybe that’s what Nancy Pelosi had in mind. Just sustain the guarantee of unlimited benefits and modest beneficiary premiums and whatever it costs it costs and taxpayers can be made to pay for it.

    One thing I always wondered about with the hospital owned insurer model is how does the hospital determine reimbursement rates to be paid by its in-house insurer vs. what it can extract from other insurers for the same services? In theory, it can hold insurer claims costs down by charging it less but that will hurt the hospital side revenue holding volume constant. It’s a tough balancing act any way you slice it. I’m glad to see the hospitals learning this lesson the hard way.

  22. Anybody know how to add an upvote button to WordPress?

    We may have to build one specifically for this post.

    This should be required reading — not just for hospital executives — but for government healthcare folks, IT folks, new administration folks, lots of folks

    I have a meeting, but more from me on this later.

    / j