Archives for 2024

In this week’s JAMA Internal Medicine, Michal Horny, myself and Mark Fendrick argue that insurers should do a better job at doing their job.

Currently, patient cost-sharing expenses are determined after care is delivered based on the realized treatment intensity, which may vary due to circumstantial factors such as the occurrence of midprocedure complications that require unanticipated interventions. We propose that payers prospectively set patient cost-sharing obligations based on expected treatment intensity (Figure). As is the case with episode-based or bundled payments already in place, insurers could use their actuarial expertise to quantify the expected mean expenditures in patients undergoing a specific procedure and require all patients to equally share the financial risk associated with unpredictable deviations in treatment intensity.

Insurers are in the business of aggregating individual level variance into a far less variant communal pool. Bad luck that produces devastating costs to one person is barely noticeable if that bad luck event happens once in a few dozen times. Insurers are supposed to be sophisticated actuarial engines and bad lack redistribution systems. Cost-sharing is intended to act as a hedge against private information that an individual has which is not revealed to the insurer until after the contract is signed.

Episode-based cost sharing would be especially suitable for care situations during which many patients receive low-intensity treatment but some—by virtue of chance—receive high-intensity treatment, such as cardiac pacemaker implantation. The implantation procedure comes with some clinical risks, namely, a 2.5% chance of intraprocedural pneumothorax. In that situation, patients receive additional health care services that would not have been provided otherwise. This increased treatment intensity then leads to increased cost of care and thus increased patient cost sharing.

No one who needs a pacemaker wants complications. No one who needs a pacemaker knows at 8pm the night before the surgery that they are going to have complications. There is no private information. At this point, cost-sharing that is conditional on good luck (no complications) or bad luck (unforeseen complications) is merely risk shifting from the entity that should be able to absorb risk and bad luck (insurers) to the individual.

We propose that for well defined bundles, cost-sharing is fixed irregardless of luck.

Yes, this will increase cost-sharing for the lucky but decrease it for the few but very unlucky. That is a challenge on the politics and likely welfare implications of this proposal, but insurers are in the business of absorbing individual level bad luck and variance so they should do that more often.

Proposing cost-sharing certainty for shoppable servicesPost + Comments (12)