Insurance incentive design in the tail

I am assuming that these are charges and not contracted rates. But my god, this is a lot of money. Even assuming a net 75% discount to the contracted rate, it is still a lot of money. And it is not the entire cost of the episode as professional fees that go to clinicians are a separate set of bills, outpatient rehab is a distinct set of claims and all the ancillary costs are elsewhere too.

This episode of care is an outlier at whatever contracted rate is applied. It is wicked expensive. And it destroys any cost sharing incentive designs. A deductible only model has the out of pocket maximum met in the first seventy seven seconds of the episode. A copay model that is heavily charges inpatient stays maxes out in a few days at the hospital. A coinsurance model of say 20% coinsurance on surgery charges is maxed out after two or three days in the ICU.

There is no benefit design structure that changes marginal costs or incentives. In this case, open heart surgery is most likely not a shoppable service, so it is irrelevant but there are enough other cases where something is both wicked expensive and somewhat shoppable. Someone with a cancer diagnosis in Chapel Hill can go to Duke or UNC hospitals. Right now the only insurance steering mechanism for that decision is network design. My personal insurance steers me to Duke Hospitals while some of the folks at my gym would be steered to UNC hospitals.

I think that the outlier cases are where the money is and rejiggering cost sharing and steering structures is a possible way to get a little more value and competition.

North Carolina, Medicaid Expansion, Premiums and Silver Gapping

North Carolina Republicans have introduced a Medicaid Expansion bill. It is not a non-waiver expansion. Instead, it wants work requirements, wellness requirements and premium payments of 2% of income for beneficiaries who earn between 50% and 138% Federal Poverty Level. We know that general wellness programs do nothing but add administratively complexity when applied to the employer sponsored universe. We know that work requirements are paperwork requirements that increase compliance costs and reduce enrollment without actually advancing the core purpose of Medicaid.

I want to focus on premiums. Here is the language from the bill:

SECTION 4. Participant contributions. – NC Health Care for Working Families
39 program participants shall pay an annual premium, billed monthly, that is set at two percent (2%) of the participant’s household income. Participant contributions shall be utilized to fund the program as required by Section 7 of this act. Failure of a program participant to make a premium contribution within 90 days of its due date shall result in the suspension of the program participant from the program unless that program participant shows that he or she is exempt from the premium requirements prior to the expiration of that 90-day period. An individual who was suspended from the program for nonpayment of the monthly premium may reactivate coverage if that individual meets the eligibility requirements and pays the total amount in previously unpaid premiums owed by the individual.

Collecting these premiums are expensive. 2% of 50% FPL is $10-11 per month. The administrative cost of mailing bills, processing checks and sending reminder notices and chasing back premiums will eat up a significant amount of administrative expenses. Arkansas tried a complex monthly individual collection process and they ended up spending twenty times as much money as they collected. Here is evidence from 2017:

Collecting small dollar premiums that are unlikely to be universally tied to a credit card or an automated electronic withdrawal is expensive. Chasing people down for these payments is expensive. Retroactively repaying claims is expensive. This is an administrative nightmare.

It is also likely to make many people objectively worse off due to Silver gapping on the Exchange.

A single individual earning $16750 (approximately 138% FPL) is expected to pay $48 per month for the benchmark Silver plan. This is approximately 3.4% of income. This buys a CSR-94 plan with a $200-$500 deductible. However, as we all know, people can take their ACA subsidy and buy a less expensive plan. They save every incremental dollar.

Monthly premiums for healthy people who are not network sensitive and only price sensitive are cheaper if there is at least a $20 gap between the benchmark and the least expensive silver plan. Below is a Tableau of the Silver Gap for a single 40 year old in 2019. Every county except Wake County has at least a $20 Silver Gap. Even in Wake County, married couples and adults with children earning 138% FPL will likely qualify for less expensive ACA plans as premiums increase faster than income for a given FPL level when more people are added to a family unit so therefore the subsidy amount increases faster than income. Almost all counties have zero premium Silver plans for 40 year old individuals earning just over 100% FPL.

Collecting and tracking premiums will be an administrative nightmare.

Collecting premiums will make many people who are currently on Exchange in a CSR-Silver plan worse off.

Collecting premiums will deter long term enrollment of the reasonably healthy.

But I’m at a point where a sub-optimal Medicaid expansion with known and significant problems is better than no expansion.

Oklahoma Medicaid expansion is on the ballot

Oklahoma activists are going the same route as Utah, Idaho and Nebraska activists successfully used in the 2018 election cycle: They are trying to get enough signatures to get Medicaid expansion on the ballot.



If you live in Oklahoma, this question needs slightly more than 177,000 valid signatures to appear on the 2020 ballot.

Odds are that even if it passes, there will be follow-on shenanigans as we have seen in Utah, Nebraska and Idaho to either delay or water down the expansions. In my opinion, a bad expansion is better than a perfect non-expansion. I assess the counterfactual as no expansion instead of a full expansion so people with different reasonable counterfactuals will vehemently disagree with me.

The ballot box is not the only way that Medicaid expansion of some sort may come to Oklahoma. There is a bananpants county level expansion proposal floating out there.

Here the scheme would be two or more bordering counties could expand Medicaid. The state share of the expansion (10% of costs) would be funded by local taxes. This would be wonderful for health and public finance economists and a complete cluster for everyone else.

Centers of excellence and network adequacy

David Cutler at JAMA Forum lays out four fairly technocratic changes to the US healthcare system to make it better.  I want to engage on one of them as there is a significant implementation challenge in the form of network adequacy regulation.  It is fixable but it needs to be addressed.

First, systems can encourage their clinicians to specialize in particular types of patients or procedures. It has long been known that volume matters for outcomes—for example, a surgeon who does 100 operations a year typically has better outcomes than one who does 20. In addition, a recent study found that specialization matters.  For several common procedures, including coronary artery bypass graft surgery and valve replacement, physicians in the top quartile of specializing in a given procedure have mortality rates that are 15% to 46% lower than those in the lowest quartile of physicians specializing in that procedure.

The evidence is strong that volume matters. Insurers that have a long shadow of the future should want to send as many of their covered lives to hospitals and doctors that do a lot of a particular procedure.  This is probably easier to do in urban areas rather than rural areas for two reasons.  First, urban areas have far more people within a given radius who need a particular service.  Secondly, urban areas are more likely to have multiple hospitals or doctors so specialization is far easier to support economically.   A clinical oncologist at Duke University Hospital can specialize in lower GI tract cancers while a clinical oncologist on the Outer Banks treats everything from skin cancer to pancreatic cancer to brain cancers on any given afternoon.

An insurance company constructing networks that prioritize high volume and specialization runs into a major regulatory challenge; network adequacy regulations are means of state regulators and private accreditation agencies to determine if a network is good enough.    Most of these regulations are based on either a specific percentage of the covered population has to be within a particular distance to a particular type of provider or there must be one provider of a given type for every so many thousand people.

The underlying assumption behind either a radius method or a ratio method of network adequacy is that any provider in a particular specialty is close enough to identical to another provider.  These methods are quality agnostic.

This is problematic is an insurer thinks that it can build a high quality network based on centers of excellence and a few specialists who do a lot of repeated procedures in order to maximize their learning by doing as well as maximizing the division of very skilled and expensive labor.  Such a network could have a large, dense and widespread network of primary care physicians and first line specialists.  However it would be a network that routinely has patients driving past half a dozen hospitals to get to a local center of excellence.  Or it would be a network that has patients flying halfway across the country for a back surgery consult.

These types of network designs happen in self-insured corporate plans.  Walmart will only pay for spine surgery at a small number of specialty hospitals across the country nowThese are high end hospitals like Mayo Clinic and Geissenger.  The idea is to minimize variation in treatment and seek out the best care possible.  The cost savings comes from reducing re-treatment and properly steering people who don’t need surgery out of the surgical pathway.

Self-insured plans are very loosely regulated.  A center of excellence model can easily pass regulatory muster there.  However fully insured plans where the insurance company has the full risk of an OMG claim, or government plans tend to have ratio or radii network adequacy rules.  These rules make a center of excellence model far harder to implement.  There could be value in short term waivers to pilot demonstration models that exempt some plans from network adequacy requirements on a few deferrable services and specialties to see if there is a way to hold cost constant and improve quality or decrease costs while holding quality constant.




Actual good news

This is interesting.


Treating all forms of tobacco like alcohol would be a major public health improvement.  The initial reporting says that this will include all products, including vape pens.

Speaking as a former under-age drinker,  the barrier to access to booze when I was sixteen was me and my boys getting one of their older brothers to buy for us.  He would do it with the understanding that we were drinking at his apartment or in other reasonable(ish) safe places where there was no chance of driving.  This was not a frequent request as we could only bug him so much.

As a sixteen year old, if I was a smoker, getting a quasi-competent fake ID or having a slightly older friend go into the Store 24 for a pack was a much less socially expensive ask.  The universe of people who would be willing to supply a near peer instead of a younger brother’s friend with an illicit substance was far larger.

This seems to be a good idea.