The Medical Loss Ratio (MLR) in PPACA is supposed to keep insurers honest. They are expected to spend either 80% (small group and individual markets) or 85% (large group) of premium dollars on medical claims and quality improvement activities. The goal was to keep insurers honest as they could not just raise rates to make more money if they weren’t paying more out in claims. MLR can be gamed aggressively by integrated delivery systems. If an insurer has a business segment that goes under the MLR, the insurer sends out rebate checks to drop net premiums to a point where the MLR is met.
This logic works very well for entities that are solely insurers such as Aetna or UHC or Cigna. It is a fairly easy calculation as claim dollars plus certain quality improvement dollars can be added up and divided by premium revenue. There are ways to game the system on the quality improvement side but it is a fairly reliable metric that gives useful information to consumers and regulators.
The logic is fundamentally irrelevant for integrated payer providers that operate behind a closed gate like Kaiser in California. Kaiser gets money in, and pays it staff to do stuff. The medical expenses are almost all internal expenses (limited claims dollars go out of Kaiser for out of area emergency care, transplants and a few other high end services). Kaiser’s MLR is calculated as the sum of external claims spend plus internal claims spend that is priced by internal methods plus quality improvement divided by total revenue. The internal claims spend is fuzzy as the transfer pricing is variable.
However Kaiser is not the issue that I am thinking about as they are odd in that they are mostly a closed garden in some areas.
Most integrated payer-provider systems are not like Kaiser in California. Instead they are an insurer, a hospital system, primary care and specialty providers and a lot of ancillary services. The insurance network also has numerous high volume, high claim value external providers. The MLR is calculated as internal claim costs plus external claim costs plus quality improvement divided by net premiums. Internal provider costs from the point of view of the insurance side of the integrated system should be calculated as if it was an arms length transaction, but the potential for mischief is out there.
Let’s take the following scenario.
The insurance company realizes that its spend is projected to be $5 million dollars below the needed level to hit the desired non-rebate paying MLR. What could an insurer do to minimize rebates?
A company that is only an insurer might look at that problem and figure they’ll pour some money into a quality improvement initiative that had been on the wish list for a couple of years but never funded in order to max out the allowable QI allowance. They could buy a little bit of goodwill from a provider by not fighting as hard on a disputed billing practice or the Hep-C cure authorization process could loosen up to cover F-2 patients as well as the current F-3 and F-4 patients. Finally, the insurer could write refund checks.
Open integrated payer provider systems have all of the same options plus one more; fooling around with internal transfer pricing models so that the provider side of the business charges the payer side of the business $5 million dollars more. This could be done as easily by increasing the contracted rate due to some “unforeseen” shock to the pricing structure. If the integrated payer-provider system is based on a high end hospital, it could be done by diverting low complexity surgeries from non-owned community hospitals to higher cost owned specialty hospitals. If the goal is to move money on the individual provider side instead of the facility side, gain sharing agreements could be created where the target MER where bonuses start to get paid is inflated to a point where a $5 million bonus is guaranteed.
These are just some of the easy ways to use internal transfer pricing to game MLR for an integrated payer-provider system that has an open network. I’ve not had my coffee yet this morning, so I am sure there are others. These games don’t matter at insurance companies that are only insurers, and they are fundamentally irrelevant at closed garden integrated payer-providers, but the quality of information conveyed by the MLR at open network integrated payer providers is very low as it is too easy to marginally game.
David Fud
Are you referring to Kaiser Permanente and the like? Does it matter that they are theoretically non-profit?
Big R
Just finished reading Baumgartner and Jones’ book on entropic versus expert information, and saw something I recognized from your work – HHI! In this case, used as a measure of information diversity, rather than a measure of market concentration. B&J make the point that HHI began life as you use it – a measure of market concentration. They note that an alternative measure (Shannon’s H) does a better job of distinguishing between situations where there is low concentration.
My question is this: have you considered looking at your 2×2 payer/provider model using Shannon’s H to break out that bottom right box (low/low) more granularly? Could it change your results?
Mathguy
Richard–Thanks for all of these posts on health insurance. I worked as a life actuary for a few years and know a few of the ins and outs, but your series has been superb. Gaining some understanding of the complexities of health insurance has been very beneficial.
Also, moar sexy time in the futbol stuff!
Richard Mayhew
@Big R: i have no idea, let me do some research on Shannon’s H as I don’t know enough (really anything) about that tool
Richard Mayhew
@Big R: Doing a fast read on the Shannon index, I don’t think it provides too much insight at a 30,000 foot level when both the payers and providers are very fragmented. There are definitely different flavors of split markets and if I was a network development guy for an insurance company or a contract management fellow for a provider group, I think the Shannon H could provide some insight/actionable information, but I am not running scenarios through my head that have significantly different large policy differences between simple HHI and Shannon H for this quadrant.
I think Shannon H would be much more useful at the High-Lo and Lo-High quadrants as there the granularity matters and it could have significant impact on what solutions anti-trust regulators pursue.
Richard Mayhew
@David Fud: Yep, Kaiser in California is a special snowflake in the US healthcare system.
It really does not matter that they are non-profit, they still don’t like sending money out the door when they can avoid it.
lawtalkinguy
I am an attorney who has handled a lot of cases against Kaiser for bad faith. They are unique. But when talking about Kaiser, it should be understood that there are three separate Kaiser entities (actually four technically, but three that are involved with a given members care), that make up the Kaiser system. There is Kaiser Foundation Health Plan (Health Plan). That is the insurer, and it is a non-profit. Health Plan then contracts with The Permanente Medical Group (TPMG). They are the physicians. There are actually two medical groups, one that covers all of northern California, and one that covers southern California. Then the Health Plan contracts with Kaiser Hospitals and Clinics to provide all the medical facilities and staff.
While the Health Plan and, I believe, the Hospitals, are non-profit, TPMG is very much for profit. The agreement between Health Plan and TPMG provides payment on a capitated basis. There is a formula based upon the total number of members, and that is the base compensation. However, where the “non-profit” Health Plan has money in excess of its operating expenses, what is otherwise known as profit, that money gets shared with the TPMG, who disperses is as bonuses to the physicians.
I should note that the agreements I have seen and the information I have precedes the ACA. However, the system as it was before was already designed to provide a mechanism for maintaining certain level of spending on health care, because any money transferred to TPMG would be considered health care spending, even if just used to provide doctors bonuses.
Apart from whether this arrangement is gaming the ACA’s MLR system, it definitely affects the quality of health care, and creates perverse incentives for physicians. The less medical services the physicians provide, the more money they make. Further, the Health Plan contracts with TPMG to provide utilization review services, so TPMG is actually not only serving as the physician, but also the insurance adjuster. But your doctor should not be your insurance adjuster. I understand that in today’s world, there has to be some cost control, and some of that has to come from the doctor in how they provide treatment, but the doctor’s ultimate recommendation must be what is medically necessary, without the concern of how it affects their bonus.
Richard Mayhew
@lawtalkinguy: And that is the key:
Or better hookers and more blow — the entire Kaiser model is ripe for internal transfer pricing shenanigans for MLR, they don’t need to play games because their structure already does it without thought.
Cheap Jim, formerly Cheap Jim
I live in Maryland and am enrolled with Kaiser. Man, those trips to California just to see the doctor are tiring.
mclaren
More incomprehensible alphabet-soup gobbledygook from Richard Mayhew.
Once again, people, Richard Mayhew is lying to you.
“Many Say High Deductibles Make Their Health Law Insurance All but Useless,” The New York Times, 14 November 2015.