Failing doesn’t hurt much

A valued commenter sent in a good question via e-mail:

Here in my state, we have a number of new names in the market, and I have no idea how to evaluate their offerings, or even if they will be around next year.  What happens to people whose insurer fails?

Actually this is two questions, and I’ll deal with the second one first right now and the first one later on this week.

What happens when an insurance company fails, especially a new insurance company?

There are a couple of types of failure.  The most common is a soft failure.  In other circumstances, this would just be an indication to get a prescription for a FYWP word, but in the insurance world, it means the company is still able to cover its medical expenses but can’t cover its overhead and administrative back-end.  In this case, the most common solution is for the failing company to sell its operating profitable businesses to a larger insurer and close out. 

In this case, members would see their policies transfer along with all of their pre-exisiting deductibles, co-pays and co-insurances.  The individual would be no worse off for a mid-year move to a new insurer than they would have been if their insurer had not changed.  Often times, there is a decent probability of some members being made slightly better off because of the conversion as translating histories between systems is a royal pain in the ass, and the default assumption is “don’t screw the member” as the state regulators really don’t like that, so ties go to members.

For instance, several, pre-PPACA, years ago, Mayhew Insurance absorbed a small regional insurer.  SRI had an odd set of plumbing regarding urgent care clinics.  Members paid 100% of the contracted rate for urgent care visits and this did not apply to deductible or out of pocket.  The goal was to drive members to PCP offices.  The SRI did not apply urgent care claims to deductibles and they way they did that was by matching Taxpayer ID, Place of Service code, NPI Type 2 and NPI taxonomy on the submitted claims to a white list that SRI maintained.    Mayhew Insurance does not require NPI taxonomy to be on the claim so we can’t tell if Dr. Bob at 123 Sesame Street is billing as a PCP office during regular business hours or if Dr. Bob is working at 123 Sesame Street Urgent Care during off-hours and weekends.  The decision was to take the backwards looking claims reconciliation and apply all potential urgent care claims to regular processing.  Members were seeing significant refund checks during the run-out period as it was cheaper for us to write checks than rebuild our claims logic. 

That is the easy type of failure.  It is a smooth transition and individuals should be at least made whole if not slightly better off. 

Now let’s look at a harder fail. 

The insurance company can’t cover its medical expenses with ongoing premium and investment revenue and there is no easy rescue via buy-out in place.  In this case, the state regulator is supposed to order the insurance company to stop writing new policies and begin a reasonably fast wind-down of operations.  Long term contracts may be sold to other insurers, and policies that are ending in the next couple of months will not be renewed when they expire.  Members are supposed to be made whole because insurance companies are supposed to maintain massive reserves to cover extremely low probability outcomes even if the insurance company is not taking in any premium revenue. 

Harvard Pilgrim maintains half a billion dollars in reserves in 2012. Kaiser Permante as a combined entity has $14 billion in net assets.  Aetna has $14 billion in net assets.  Maine Community Health Options, a new co-op, has $14 million in reserves.  The goal is to have sufficient assets on hand to pay out all claims if no further premiums are received for three to six months.  And if this fails, there are state level guarantee associations which is a combined group of all insurance companies in a state.  The responsibility of the association is to cover any losses from the failed health insurance company. [see Update #1]

So as long as someone goes to an insurer that is actually registered and regulated as an insurer, their claims will eventually get paid by someone even if the original insurer fails.

Update #1 from comments:

a Guarantee Association is a entity created by statue that assesses solvent carriers to cover claims of an insolvent carrier. There is also a Liquidator appointed to represent the Commissioner of Insurance of the State that the carrier is domiciled. Once a judge signs an Order of Insolvency, the Liquidator marshalls assets, manages the day to day operations, mails Proof of Claim notices to any party doing business (suppliers, reinsurers, etc) with the failed insurer, etc. Proof of claims that are returned to the Liquidator by the Bar Date (in the Order of Insolvency), the Liquidators team analyzes the claims, sorts them by priority class, and then send of Notices of Determination, informing the claimant of the allowed amount of their claim. This amount may not be what the claimant will actually receive, which is usually a percentage of the assets recovered by the Liquidator. The claimant has the right to object to the amount determined by the Liquidator. Eventually, the remaining assets are divided among the Guarantee Association and other claimants as determined in the Notice of Determination.

Four is more than three

I know this is almost nutpicking, but the Wall Street Journal editorial page either can’t do basic subtraction or is suffering from amnesia:

The health-care law was generated by an administration promoting government as the solution to inequality, yet the greatest irony of ObamaCare is what will undoubtedly follow as a long-term, unintended consequence of the law: a decidedly unequal, two-tiered health system. One will be for the poor and middle class, and a separate system will be for those with the money or power to circumvent ObamaCare.

With the Affordable Care Act, the government has dramatically expanded its authority as final arbiter over health insurance and consequently over access to medical care. After the law’s Medicaid expansion and with the population aging into Medicare eligibility, the 107 million under Medicaid or Medicare in 2013 will skyrocket to 135 million five years later, growing far faster than the ranks of the privately insured.

Besides the fact that I love how he elides the aging of the population into Medicare as a bad thing, the thing I love about this excerpt is the assumption that the United States had a one tier system of healthcare pre-Obamacare. 


We had a four tier system with some caveats and carveouts in 2009.  We are moving towards a three tier system with some caveats and carve-outs under Obamacare.

In 2009 and 2014, the first tier was the tier for the rich and very well insured.  Senator Ted Cruz’s $40,000/year family policy that his wife is the primary contract holder for is an example of this tier.  He can go to whatever provider he wants without worry, and his wait times will be minimal.  If he blows out his elbow while pulling his head out of his ass, the distinguished Senator from Texas can go to Dr. James Andrew, the Tommy John specialist for a repair.  If his kids get cancer, they can go to whatever clinic they want to in the United States and get top line treatments that cost more than my family’s annual income for an extra three or four months of life.  This type of insurance is fundamentally the same between 2009 and 2014.  The big difference is that some of the premiums will be taxed in the near future due to Obamacare instead of being entirely tax free.

The second tier of 2009 coverage was solid employer provided group insurance and solid individual coverage.  It was possible to have solid individual coverage, you just had to be lucky.  This tier of coverage has some limitations, it has some deductibles and co-insurance and it is the most common tier of coverage.  Employer provided insurance also has massive explicit (employer provider) and implicit (tax advantages) subsidies.  In 2009, this tier was a shrinking share of coverage even as it is the dominant share of coverage for people under 65.  Now the Exchanges and the threat of the employer mandate is growing this tier of solid, private market coverage.

The third tier in 2009 was government insurance provided through a variety of programs. The big programs are Medicare, Medicaid, CHIP and the VA.  Medicaid was income, asset, and “deserving” poor status limited.  Working adults without insurance and low incomes were out of luck in most states for Medicaid. CHIP had expanded but was not all inclusive for all kids.  Medicare and VA were functioning reasonably well. 

Now, Medicaid in half the states is neither asset nor “deserving” poor status dependent; it is just income dependent.  CHIP was expanded in the winter of 2009.  VA has not been significantly altered.  Medicare’s drug benefit has been enhanced, and Obamacare is equalizing the risk adjusted payment rates for traditional fee for service Medicare and Medicare Advantage as well as engaging in massive experimentation on new payment models.

The fourth tier in 2009 was the “You’re on your own” tier.  This was for people who either had no insurance or had insurance that was so skimpy it could not protect people from financial ruin from a moderate size medical event much less a major medical problem. 

The fourth tier is being phased out in half the states.  The long run goal is for most of the people in this tier to move to either the third tier via Medicaid or CHIP or the second tier by enabling community rated non-medically underwritten policies to be sold on the Exchanges. The first tier will shrink due to the Cadillac excise tax, and the fourth tier is larger than it should be due to the sadists on the Supreme Court and sociopaths in the Republican Party, but the long term goal and program design is to move towards a three tier instead of four tier system. 


About that Express-Scripts data

I just want to highlight and raise some data caveats about the Express Scripts data set concerning pharmacy/prescription costs for Exchange patients.  There are some interesting nuggets in the data.

Population inferences:

Our early analysis reveals that, in January and February, use of specialty medications was greater among Exchange enrollees versus patients enrolled in a commercial health plan. Approximately 1.1% of total prescriptions in Exchange plans were for specialty medications, compared to 0.75% in commercial health plans, a 47% difference.


More than six in every 1,000 prescriptions in the Exchange plans were for a medication to treat HIV. This proportion is nearly four times higher in Exchange plans than in commercial health plans.


  • The proportion of contraceptives was 31% lower in Exchange plans

What can we infer from these data points?  There are a couple of inferences.  The first is that early Exchange population is sicker than the typical person covered in an employer sponsored health plan.  We can infer this from the higher use of specialty drugs, and the higher use of HIV drugs.  This was expected. 

The second is that the average age of the early Exchange utilization pool is older than the average age of the typical person covered by an employer sponsored health plan.  The tell here is the much lower use of contraceptives.  Contraceptives are heavily used by women between their late teens to early 40s.  Prescription contraceptive useage spikes in the twenties and steadily declines as women either have children, have permanent birth control, enter menopause or use barrier methods.  If prescription contraceptive usage has cliff-dived, it is probably because there are fewer 53 year olds who need or want to use it.

So what does this mean?

Does the data suggest a pharmaceutical death spiral?

No, as there is one other massive caveat in the dat set, and that is the population examined. 

The analysis is based on a national sample of more than 650,000 de-identified pharmacy claims from Jan. 1, 2014 through Feb. 28, 2014  [my emphasis] for patients enrolled in a Public Health Insurance Exchange

We know one critical thing about the population being analyzed.  

The people who were eligible for this analysis were the first adapters of Exchange insurance.  The January cohort being studied are  are the two million or so people who signed up on the Exchange in October, November and December.  They are the people who were so desperate and self-identifying as needing insurance that they were willing to keep on coming back again and again until the fucked up website worked well enough for them.  Of course, we should expect this group to have higher than expected claim utilization.  The February cohort is the January cohort plus the people who signed up on the Exchanges by January 15th.  This is another million or so individuals.  This cohort should be slightly healthier and slightly younger than the January only cohort. 

The Express Scripts data is reflecting a limited universe of claims utilization by people who self-identified as needing insurance now.  We know from enrollment data that January eligible individuals skewed old and skewed female.  We knowfrom  the February data the risk and utilization pool was slightly younger.  The March enrollment surge pushed the average age of the entire Exchange risk pool down further still, and the April extended blue box enrollment will probably reduce the average age again. 

Express Scripts data is confirming what the enrollment demographic data suggests.  The older and sicker signed up first and got care first.  The younger and healthier are coming in late.  This is not surprising.

Doin it, doin it, doin it well

Sarah Kliff at Vox highlights the problem of readmissions after kidney stones are busted:

That makes the removal of a kidney stone one of the most basic, most common procedures in the United States – not the type of thing that you’d think would land you in the emergency room a few weeks later.

Except it turns out that, after the extremely routine surgery, one in seven patients actually do make an unexpected trip back to the doctor – and that can cost upwards of $30,000….

 how many of these unplanned visits are preventable….Scales, for his part, thinks a good number might be preventable. He points to the fact that people seen at really high volume facilities – places that do lots of kidney stone removals – tend to have 20 percent fewer unplanned follow-up visits than people seen at places that do fewer procedures. That suggests having some kind of expertise in the procedure can lead to better outcomes.

This makes sense.  As someone or a team does a task more often,they get better.  Their technique gets better, their recognition of that odd little detail which could lead to problems gets better, their supplies are better suited for a particular task and their ability to recognize and recover from an error gets better.  It is a simple division of labor and skill.

Right now, the scope of practice for a specialist can be fairly wide.  A general surgeon can go to town from neck to ankles.  She might have a gallbladder operation on Tuesday, a bile duct repair on Wednesday, a hernia repair on Thursday.  An orthopedic surgeon might rebuild on ankle on Monday morning, insert pins into the elbow after lunch, and replace a hip before dinner. 

This could change in the future.  The broad scope of practice is a necessary skill base, and it can be a necessity in more remote areas, but insurance companies may modify their payment models for better and in the long run, cheaper care in urban areas.

Read more

Rural Hospitals, DSH payments, and Vegas hotels

A commenter asked why so many rural hospitals in Georgia are closing.  It is a good question, and a decent chunk of the explanation is PPACA via the asshole Chief Justice et al and the remainder of the explanation is the economics of running a hospital or a Vegas hotel.

Let’s start with the Vegas hotel.  I mentioned in September that hospitals have very high costs to open up the doors.

 Time Magazine in August had a good piece on Las Vegas’s hotel and gambling industry that has an interesting nugget of explanation for hospital pricing:

A 5,000-room casino hotel that runs 24/7 has high operating costs, and it’s the gambling action that has covered them. The magic of a casino hotel is that once the costs are covered, profit mounts prodigiously–in accounting jargon, this is a business with very high operating leverage.

Hospitals and most other medical practices are the same way. Just opening the doors is extremely expensive as the fixed costs are very high. However, the marginal cost of treating the next patient for most situations (high end drug treatments excluded) are not that high. Hospitals with high census or heads in beds counts are able to use the high usage of their facilities to cover fixed costs and then operating costs.

A recent article on a hospital in Georgia closing illustrates this point:

Lower Oconee Community Hospital in southeast Georgia has closed due to financial problems, becoming the state’s fourth rural hospital to do so in the past two years.

The 25-bed “critical access” hospital in Glenwood, in Wheeler County, is looking to restructure, its CEO said in a statement….

The Wheeler County area had a 23 percent uninsured rate, and 10 percent of citizens are unemployed, according to the County Health Rankings from the University of Wisconsin and the Robert Wood Johnson Foundation.

Forty-one percent of the county’s children live in poverty.

“We just did not have sufficient volume to support the expenses,” O’Neal told WMAZ. “It’s a terrible situation, and it’s tragic, the loss of jobs and the economic impact.”

So how did this hospital survive so long despite serving a very poor and underinsured area? Read more

Tiering and Steering

One of the major legitimate complaints about Obamacare Exchange policies has been the prevelance of narrow networks.  Narrow networks exclude hospitals and providers so that a policy holder may not be able to go to the physician or hospital that is down the block from their house.  Insurers like narrow networks because it gives them a bit more cost control and negoatiating leverage with common service providers.  Narrow networks allow insurers to say no to providers.

Insurers that want to avoid public backlash for having networks that are too narrow may embrace the tiering model of network and benefit design.  This is an alternative that allows an insurance company to say it is offering its full network to potential members while still giving the insurance company significant cost control.  So how does this dessert topping and floor wax work?

 A traditional single tier PPO network will have an in-network benefit design and an out of network benefit design.  If a member sees an in-network provider, they’ll have a single deductible, co-pay and co-insurance.  Their “slash line” might be $1,500, $20/50, 20%,$4,000 which means they have a $1,500 deductible, a $20 PCP co-pay, and a $50 specialist co-pay and 20% co-insurance and a $4,000 out of pocket maximum.  Their out of network benefits may be $3,000, $100, $50, $8,000.  From a member point of view, it is simple — see an in-network provider and get a much better deal.  The providers who contract for the in-network rate know that they are going to see a lot of patients and get paid quickly. 

A tier and steer benefit design makes this more complicated.  The in-network provider network is split into two or more groups or tiers.  The preferred group of providers are the “narrow” part of the tier and steer.  These are the providers the insurance company wants its policy holders to go see.  The way that members are sent to the preferred providers is by changing the slash line.  The preferred tier slash line might be $500, $10/25, 15%, $2,000.  That looks like a good deal as the deductible is low, and total potential exposure is low.  This is bought at the cost of a limited network of providers. 

The second but still in-network group of providers will trigger  different slash line.  The slash line for these providers may be $2,000, $20/$50, 20% $4,000.  These providers are in-network but they are expensive to see.  This is a patina of plausibility that the product is a broad network product.  The much higher slash line means these providers won’t see many patients, and they will either get out of network in the near future, or drop their rates to get into the preferred tier.  However, if a member wants to see the doctor who fixed Aunt Maria’s hip really well last year, they can without being denied or ran around. 

The final tier is still the traditional out of network tier. 

Tiering and steering is an effective cost control system. My personal insurance is a tiered product.  I get the good rate at every provider group except those that belong to Big City Academic Medical Group.  BCAMG has a brand of having BCAMG in everything, so the information and search cost of figuring out who I should see and who I should not is fairly low.  I save 15% or so compared to the broad, single tier product, although the truly narrow network is slightly cheaper than the tiered network. 

However not all tier and steer networks are designed to solely get low cost, high quality providers on the priority usage list.  Some tiered networks are designed by integrated payer-providers to keep almost all activity in the network.  Other times, the tiered networks are part of a core business strategy designed to leverage synergistic opportunities while fucking over competitors and consumers.  Other times, a tiered network is a one-off exception for a large group that wants to be a special snowflake. 

I think tiered networks will be more common on the Exchanges this fall as well as more common for employee sponsored plans as they produce cost savings while at least maintaining the illusion of unlimited choice.

Cost Control in the UK

Via the Guardian:

A Herceptin-style drug that can offer some women with advanced breast cancer nearly six months of extra life has been turned down for use in the NHS because of its high cost.

In draft guidance now open to consultation, the National Institute for Health and Care Excellence (Nice) blames the manufacturers, Roche, who are asking for more than £90,000 per patient, which is far more than any comparable treatment….

It is not a cure, but in trials it extended life by a median of 5.8 months, compared with the current combination of lapatinib plus capecitabine

This drug will be available to British patients through a special fund but it will not be a first line, readily available course of treatment.  NICE, the NHS cost effectiveness research body is charged with determining what things should be paid for by the National Health Service and at what price.  It has the ability to say no in a very public and very enforceable manner.  NICE’s decision framework is fundamentally simple. 

NICE uses cost-effectiveness thresholds that value a perfectly healthy year of life to be worth spending 20,000 to 30,000 pounds sterling or roughly $33,000 to $49,000.

Having used the QALY measurement to compare how much someone’s life can be extended and improved, we then consider cost effectiveness – that is, how much the drug or treatment costs per QALY. This is the cost of using the drugs to provide a year of the best quality of life available – it could be one person receiving one QALY, but is more likely to be a number of people receiving a proportion of a QALY – for example 20 people receiving 0.05 of a QALY.

Cost effectiveness is expressed as ‘£ per QALY’.

Each drug is considered on a case-by-case basis. Generally, however, if a treatment costs more than £20,000-30,000 per QALY, then it would not be considered cost effective.

Using NICE’s methodology, the drug, Kadcyla, supplies less than half a year of life, and once adjusted for quality, significantly less than half a QALY for at least six and potentially twelve times the cost that NICE and NHS is willing to pay for that type of improvement in health. 

There are two possible scenarios going forward.  The first is that Roche, the drug manufacturer, dramatically drops the price of the drug to get it underneath the cost effectiveness threshold.  The second is that Roche does not move its price point by much if at all, and NHS spends its limited budget on more effecient interventions for other people.