A friend of the blog raised a very interesting point about the current state reinsurance efforts. These efforts don’t modify risk adjustment coefficients. This leads to double counting of claims leading to risk based payments. This is a major problem with a couple of solutions.
We need to get deep into the weeds of chasing money like a rabbit at a greyhound track.
First, risk adjustment is zero sum at the state level. A dollar that Plan A receives is a dollar that Plan B gives up. Risk adjustment works by assigning a risk score to a collection of diagnoses. That risk score is then multiplied against average premium and a few standardization factors to calculate the individual’s risk value. The company’s risk values are summed and normalized against the state wide risk profile. If the company has less than state level risk, they pay into the risk adjustment fund which then pays their competitors who have more than state wide average risk.
Hemophilia receives a risk score of about 52 times average monthly premium to treat. In a state with an average premium of $500, that means this patient receives a risk value of $26,000 per month or $312,000 per year. A portion of that $312,000 covers ongoing maintenance medication that everyone receives. A portion covers catastrophic costs as a bad bleed can quickly run to a million dollar month.
Minnesota’s new reinsurance program pays a portion of claims that run from $50,000 to $250,000. That means a Hep-C cure will be paid for via risk adjustment and reinsurance. That means the maintenance prescriptions for a hemophiliac will be paid for by both risk adjustment and reinsurance. This means solid tumor cancers will be paid partially for by both risk adjustment and reinsurance.
Double counting is a problem on a market stabilization level and a political level. It needs to be fixed.
When I worked for UPMC Health Plan, the last three years had me focusing a significant chunk of my time on Medicaid risk adjustment. Pennsylvania Medicaid had revenue neutral risk adjustment like Exchange and a high cost risk sharing pool. These two elements of risk played nicely with each other. Risk adjustment was capped to be no more than approximately $75,000. Everything after $75,000 was put into a high cost risk sharing pool where money was spread around to cover unusually high expense cases. A dollar of spend was only ever going to be at risk for pool smoothing purposes once.
Minnesota or any other state that is thinking of using an aggressive reinsurance scheme to stabilize pricing needs to re-calibrate their risk adjustment co-coefficients to prevent double-dipping where the same dollar of spend will be credited to both risk adjustment values and reinsurance. They can also choose to apply risk adjustment corrections to the claims data that is used to trigger reinsurance payments. Both require local technical expertise, political will and time to implement.
As more states think about reinsurance, they also need to think about how they do risk adjustment. There are solutions, they just need to find the local solution that works.
And now the appropriate musical response to all of the above: