Administrative costs in the health-care system are a classic public good. Payers and providers may together agree that standardizing billing codes and quality reporting would be valuable, but no single actor has an incentive to pay for standardization when others will benefit as well. For example, if insurer A chooses to harmonize its policies with insurer B, that lowers administrative costs across the board and thus fees that all insurers collectively need to pay. However, insurer A will not take these cost savings to other insurers into account. As a result, insurer A will be discouraged from investing in harmonization.
Pointer from Tyler Cowen.
As if there were no incentives anywhere for the private sector to solve this problem. But let me re-frame this from the perspective of an entrepreneur making a pitch to a venture capitalist.
“Doctors and hospitals have a big pain point in that their staff needs to fill out different claim forms for different insurance companies. CutlerMedForms has the solution. We will provide a software application that allows administrative staff to fill out a single, easy-to-understand on-line form. They simply check which insurance payer to whom to submit the bill, and our software fills out that insurance company’s form with the proper insurance codes. We estimate that providers can save $X billion of dollars in administrative costs using our software, making this a large profit opportunity for CutlerMedForms.”
Someone reading this might be skeptical that the profit opportunity actually exists. By the same token,, one should also be skeptical that the “classic public good” really exists.
https://www.availity.com/
This ‘clearinghouse’ exists. I’m sure there are others.
One of the commenters to Scott Alexander’s recent book review of Hoover: An Extraordinary Life in Extraordinary Times, linked to a fascinating article about Hoover’s campaign as Commerce Secretary for “specification, standardization, and simplification”. Hoover’s idea was that there was a lot of waste that the feds could eliminate by facilitating surveys, making reports, organizing conferences, etc. The beginning of the article is a little pretentious but it gets better 🙂
We’re talking doctors, right? See Atwul Gawande’s “Checklist Manifesto”.
These things often fail because the mechanism for supporting the engineering simply doesn’t exist. Aviation has checklists because the alternative was viscerally horrible. Medicine doesn’t because “it’s not how we roll.” I’d frame this as quite path-dependent. Nobody has time to learn everything, and process design is a thing.
If you do manage to make it to a doctor again this year, count who does what in the office. You have one actual doctor and multiple clerical staff. It’s all about charging rents on the privilege afforded by that fancy degree.
The big problem here is that coding maps to a decision to pay or not. There is no way to gain efficiency in coding without insurers agreeing to insure a common set of codes. Otherwise, gaming will be required.
We usually don’t like hypocrisy, but sometimes the way society works is that we need to support the strongest expression of values and rules, but which don’t have the right amount of nuance or flexibility to deal with the messy, complex real world, and then also retain a little leeway for small violations to vent the excess pressure that builds up.
For that to work, we need a tacit social consensus that violators are discrete in their violation and continue to pay lip service to the values (the tribute paid to virtue), and that enforcers and investigators don’t get carried away, tolerate a lot of suspected violation, and only go after the most flagrant, public, and harmful instances.
This pops up in more places than you think. There are a lot of occasions when someone has to make and apply judgments about other people which are socially undesirable (as when reality clashes with sacred beliefs) and thus look really bad on paper, and when they are terrified of being penalized under vague, harsh legal rules, they simply cannot afford to ever put down on paper, and need some socially acceptable excuse as the alternative ‘explanation’ (however absurd, but, like currency, good so long as it is accepted) to rationalize away the decision.
The trouble is, software is terrible for this. Law and bureaucracy and zero tolerance policies are all bad enough, but in human systems, occasionally one encounters a human with some wisdom, judgment, and compassion who knows when to bend the rules. Obviously this opens the door to corruption, to bend the rules when they shouldn’t be bent, and minimizing both types of errors is a hard problem.
But software totally eliminates the possibility. The machine is totally inflexible and stupid and doesn’t understand this problem. The people who code the machine may understand it, but they can’t fix the problem, because they have to stick to the formal rules, in an environment where everything is recorded forever with attribution.
I have a friend who does Business Operations Enterprise Software stuff in Eastern Europe. Yes, corruption is bad, but in the race-to-the-bottom equilibrium, if you aren’t bending the rules as much as your competitors, you will go out of business, and selection favors the worst.
So, the business owner trying to get the benefits of the latest, most sophisticated business operations software will like the new usefulness it provides him, but will also want it to do the things he used to do without software, like move revenues around and create a bunch of phony transactions selling as losses with other shell companies he owns to minimize tax liabilities. Big multinational corporations are pretty good at doing this in technical obedience of the rules, and also pretty good at lobbying for rules that only big multinationals can game in terms of technical obedience.
And my friend has to deliver the bad news that the software doesn’t allow for such games, and like automatic law enforcement, catches everything and interprets them as errors which can’t be validated, certified, etc.
Now, from the perspective of a member of a high-trust-equilibrium society where Gerson’s Law doesn’t operate with such rigor, one is tempted to say, “Good!” for this particular example.
But I suspect in medicine there are many times a patient comes in and the Doctor comes to the honest conclusion that the individual is lonely or crazy or malingering or saying the “Typical Addict BS” in trying to get drugs they aren’t legally allowed to get.
I’m sure most doctors have clever ways to try and determine the true nature of the situation and the patient, and then wise ways to handle those people in the most reasonable, least costly, least harmful ways, which don’t necessarily involve making awkward and uncomfortable accusations to their faces, or writing those sentiments down in the notes.
And this is probably for the best. Humans are messy and complex, and health care situations even more so. Doctors probably say they need room for human judgments and so forth, but part of that is the rationalization for what they really need, which is the flexibility to take the above approaches without making a federal case about it.
But software and forms don’t like or allow any of that. And I think on some level most doctors have a real terror of a world in which they are effectively shackled by software with a Puritanical zeal to eradicate anything that doesn’t fit in the rules.
I posted just before you, and got quite a chuckle out of this because of how we said essentially the same thing in our own ways.
Your pitch seems enticing, but if anything is wrong with it it’s that it is essentially identical to the one the existing healthcare IT/electronic healthcare records providers use. Companies like EPIC/Cerner already control a large share of the market and base their proposed value add on that fact: there’s a network effect and we are the network!
HHS recently finalized rules on “interoperability” of patient data (https://www.hhs.gov/about/news/2020/03/09/hhs-finalizes-historic-rules-to-provide-patients-more-control-of-their-health-data.html) pursuant to the 21st Century Cures Act (https://en.wikipedia.org/wiki/21st_Century_Cures_Act) which are intended to establish required, non-proprietary data standards to lay on top of existing Health IT systems as a means of facilitating exchange of data to software/app developers and spur innovation in these areas.
I offer these up without judgment of whether they will work as intended or not.