Auto-enrollment is a nifty idea where a governing entity would use administrative data to put people into zero net premium plans that they qualify for even if the individual does not actively sign-up for the plan. Administratively, it is massive challenge, especially in the individual market where there are several distinctive populations: long term individual market enrollees who are there for business reasons, individuals who aren’t covered through work but are working in fields where employer sponsored coverage or Medicaid is likely to occur, and people who are uninsurable in an underwritten world at a reasonable premium.
The individual market for the second group can be thought of as a holding area until something better comes along. We live in an economy where jobs and incomes change fairly frequently. There is a lot of eligibility churn in the status of insured vs. uninsured. There is even more variance in income which is what would drive the availability of a zero premium plan.
The Brookings Institute’s Sobin Lee and Christen Linke Young look at the challenge of using administrative records for auto-enrollment. There are big pragmatic challenges as the data gets stale fast.
This analysis shows that when using insurance status information that was just one month old, 5% of the group designated for auto-enrollment would actually have obtained other coverage, and, conversely, 5% of the “true” uninsured would not be auto-enrolled. If the insurance status information was 5 months old, 20% of those auto-enrolled would in fact have other coverage, while 20% of the truly uninsured would not be auto-enrolled.
If we assume fragmented insurance markets and insurance programs (ESI, exchange, Medicare, Medicaid, CHIP etc) with different and mutually exclusive eligibility requirements, auto-enrollment runs into massive data problems every time it is implemented without a real time eligibility check. Auto-enrollment works well when it is the second step of a process such as when an individual qualifies for Medicaid in a Managed Care state and does not make an active selection; they will be sent to an insurer on some pre-specified system that operates silently behind the scenes.
Another Scott
OT – In case you missed it, Rep. Don Beyer is paying attention to your ACA sabotage paper.
Cheers,
Scott.
smintheus
OT, will there be a WWC thread?
As I feared, the shakiness of the US defense is being exposed. And as expected, Tobin Heath is being her fantastic self.
Eric
When healthcare is actually practiced, there is the expectation that some fraction of people treated don’t really need the treatment and some fraction of people are not receiving the treatment that they’d need. I often find that most of my patients are on one or more medications that might not be helpful — the alternative if you stop them and they are important, however, is permanent disability.
Why would anyone expect that other healthcare decision are any more precise. I’m actually surprised that its as good as misclassifying only 5-10% based on month-old information. Is there an interpretation of this study similar to the number-needed-to-treat used in medicine? How many people need to be auto-enrolled to prevent one catastrophic life-changing medical bill? Or how many people need to be auto-enrolled to erroneously enroll one non-qualifying person who ends up costing more to fix than benefits their peers?
In fact, doctors can be penalized financially for not prescribing a treatment (or intervention) even when its applicability is questionable (meaningful use metrics jump to mind). For reimbursement, auto-rejection by insurance companies and auto-penalization by Medicare doesn’t seem to be the “second step in the process” — it often seems like the default. So if treatment targets and meaningful use is considered good for an individual’s care, what about applying this trigger-happy penalty more broadly to Medicare & Medicaid spending — penalize states that have too many residents un-insured?