Good stuff, D.A. Health care financing needs to move to a defined contribution model, away from a defined benefit model – in the same way pensions made the move. Your paper envisions one way this could happen and includes ideas that bear serious consideration. The three goals of your vision are clear: every American should get satisfactory healthcare; care should be evidence based; and we need a payment model that drives competition, efficiency and innovation.
Using bundled payments for episodes of care, as a foundational element of your model, also makes complete sense. I have built HMOs, ACOs and every type of payment model, and bundled payments are unquestionably the killer app. They align the payment with the patient and provider experience and outperform other strategies on virtually every measure.
As you surely know, there is growing support for a ‘premium support model’ for Medicare, to shift Medicare spending to a defined contribution model. Supported by groups like the American Action Forum, American Enterprise Institute and Manhattan Institute, the Premium Support model also gives people a subsidy to then select a health plan who receives the premium payment. That model fails to address your many criticisms of existing health plan models and likely further the hegemony of health plans and health systems. Your approach, on the other hand, tears down the entire paradigm of health insurance as we know it today. Pretty audacious!
I wonder, however, if what you propose really needs a blockchain infrastructure or cryptocurrency to accomplish your goals. Smart contracts alone would not handle the complexity of bundled payments without robust off-chain oracles and components. And as much as we all hate regulation, it is not likely that regulatory oversight, contract amendments or dispute resolutions can be fully automated in a smart contract. A further problem is that smart contracts lack efficient storage and processing power for large datasets, which we would have in your vision. Combined with excessive computation and gas costs, I fear that the brilliance of your core ideas may be quickly dismissed if they require a new blockchain protocol and a new cryptocurrency from the start. A hybrid model would seem to be a more practical way to introduce your ideas, with the intent to evolve towards more on-chain components over time.
As a final note……recent work using the open-source PACES episode grouper - which overcomes the known problems of the Symmetry and Prometheus episode definitions -- has shown that prices for many episodes of care vary by over 500% within three digit zip code areas. And we still can’t seem to find any correlation between price of an episode and the patient health outcome. So, you are right in your thesis – great healthcare is obtainable at much lower costs!
Your voice is a welcome addition to this important dialogue.
I wonder how this could be piloted successfully. Despite some of the elegance here, I think we can all agree that on the national scale in our country it is virtually impossible to imagine the stakeholders (including voters/citizens) being able to appreciate, comprehend, and then support such a "disruptive" model.
The irony of course is that this is not nearly as disruptive or unfamiliar as a pure nationalized/socialized model is - though it may seem so on the surface. Apologies if oversimplified, but it seems more or less a gamified HMO+HSA with some fun new bells and whistles (and currency) to improve/streamline decision-making and incentives. But, the fact that it involves so many new system-level ideas (including departing from USD), this of course begs the question (like Peter seemed to imply): maybe it's just simpler to go single payor?
Perhaps someone can design an in silico pilot that can be designed that is a literal game simulation that tasks players with amassing as much time+money as possible, but they come at it from various perspectives: 1) the patient, 2) the doctor/provider, 3) the payor (proxy for either taxpayer or commercial entity), 4) the innovator/biopharma, 5) the mutual fund, etc.
There could 3 basic worlds/maps: 1) status quo US-based world with hybrid employer/taxpayer based payors and semi-market-driven pricing, 2) single payer/social model, 3) zero toll. In which world are we maximizing healthspans, efficiency, access, innovation, economic activity, etc.? Maybe there's an AI game dev tool that can quickly prototype this?
As broken as our US system is today, no doubt it has buttressed the vast majority of capital deployment and therapeutic advancement we see today. But that dynamic is shifting on the world stage, whether talking about China or access to advanced therapies in interesting decentralized ways in parts of Europe... Maybe zero toll is an optimal amalgamation. But how to best prove or demonstrate it? Hard to imagine even a Kaiser type system would just flip the switch to this in one fell swoop, let alone the entire federal electorate/government, without some kind of pilot model proving the concept...
The main value of doing it in crypto is programmability of the payments. Using smart contracts, payments can be made spendable in very specific and constrained ways. And they can expire if unspent and cash flows can be tracked / utilized to automate tax deductions and so on. With regular money in regular bank accounts, you would have a hard time doing any of that.
I suppose, although I don’t think it would be too hard to program the dispersal system to take the existing balance into account, and to limit the places the account can spend money with regular bank money. I guess it is more of a question of the optimal method of database construction
I like these ideas. In order to get started though there are at least two major components in my view that I find to be essential
1) Right now, how much access do you have, in a consolidated and interpretable way, to your medical records. I can say as a patient of Sutter and Kaiser in SF, it's possible but cumbersome to get some, and probably insurmountable to get all, and neither would be comprehensive.
2) For the average person, how much do they think about, or want to think about their health? The discrepancy between our physical relationship to our bodies and intellectual desire to know anything about them is very high. This is a gradient of course, but by and large, the degree to which people understand (or even desire to understand) their health, is a function of how much it immediately impacts them.
The economic and care provision issues are one domain, medical record access and synthesis issues are another, but the biggest is human behavior and interest.
It is both idealistic and desirable to believe that this possible, but the issues are rooted in behavior with the systemic parts making that more or less easy to influence.
This is really interesting! Do you know of any companies or research organizations that are working on something like this? I am a senior smart contract engineer and working on something like this sounds fascinating to me!
Also, speaking of blockchains that have the speed and decentralization required for something like this, Flow (https://flow.com/) is doing some amazing things and could be a perfect place for this!
This is a very interesting and creative exercise but the complexity is a problem. I have a graduate degree and can barely follow all the details of how this is supposed to work. If the system depends on your average patient understanding it, it's dead on arrival. Even if you assume the average patient normally has the wherewithal and intelligence to manage the system under normal cases, what happens in abnormal/emergency situations? Is someone in the ER with a broken leg supposed to calmly evaluate the various provider rankings and cryptocoins to decide on the optimal way of setting his bones? Seems there is a limit to what you can leave up to patients or even their designated guardians/caregivers -- some paternalism likely needs to be baked into the system.
Good stuff, D.A. Health care financing needs to move to a defined contribution model, away from a defined benefit model – in the same way pensions made the move. Your paper envisions one way this could happen and includes ideas that bear serious consideration. The three goals of your vision are clear: every American should get satisfactory healthcare; care should be evidence based; and we need a payment model that drives competition, efficiency and innovation.
Using bundled payments for episodes of care, as a foundational element of your model, also makes complete sense. I have built HMOs, ACOs and every type of payment model, and bundled payments are unquestionably the killer app. They align the payment with the patient and provider experience and outperform other strategies on virtually every measure.
As you surely know, there is growing support for a ‘premium support model’ for Medicare, to shift Medicare spending to a defined contribution model. Supported by groups like the American Action Forum, American Enterprise Institute and Manhattan Institute, the Premium Support model also gives people a subsidy to then select a health plan who receives the premium payment. That model fails to address your many criticisms of existing health plan models and likely further the hegemony of health plans and health systems. Your approach, on the other hand, tears down the entire paradigm of health insurance as we know it today. Pretty audacious!
I wonder, however, if what you propose really needs a blockchain infrastructure or cryptocurrency to accomplish your goals. Smart contracts alone would not handle the complexity of bundled payments without robust off-chain oracles and components. And as much as we all hate regulation, it is not likely that regulatory oversight, contract amendments or dispute resolutions can be fully automated in a smart contract. A further problem is that smart contracts lack efficient storage and processing power for large datasets, which we would have in your vision. Combined with excessive computation and gas costs, I fear that the brilliance of your core ideas may be quickly dismissed if they require a new blockchain protocol and a new cryptocurrency from the start. A hybrid model would seem to be a more practical way to introduce your ideas, with the intent to evolve towards more on-chain components over time.
As a final note……recent work using the open-source PACES episode grouper - which overcomes the known problems of the Symmetry and Prometheus episode definitions -- has shown that prices for many episodes of care vary by over 500% within three digit zip code areas. And we still can’t seem to find any correlation between price of an episode and the patient health outcome. So, you are right in your thesis – great healthcare is obtainable at much lower costs!
Your voice is a welcome addition to this important dialogue.
I wonder how this could be piloted successfully. Despite some of the elegance here, I think we can all agree that on the national scale in our country it is virtually impossible to imagine the stakeholders (including voters/citizens) being able to appreciate, comprehend, and then support such a "disruptive" model.
The irony of course is that this is not nearly as disruptive or unfamiliar as a pure nationalized/socialized model is - though it may seem so on the surface. Apologies if oversimplified, but it seems more or less a gamified HMO+HSA with some fun new bells and whistles (and currency) to improve/streamline decision-making and incentives. But, the fact that it involves so many new system-level ideas (including departing from USD), this of course begs the question (like Peter seemed to imply): maybe it's just simpler to go single payor?
Perhaps someone can design an in silico pilot that can be designed that is a literal game simulation that tasks players with amassing as much time+money as possible, but they come at it from various perspectives: 1) the patient, 2) the doctor/provider, 3) the payor (proxy for either taxpayer or commercial entity), 4) the innovator/biopharma, 5) the mutual fund, etc.
There could 3 basic worlds/maps: 1) status quo US-based world with hybrid employer/taxpayer based payors and semi-market-driven pricing, 2) single payer/social model, 3) zero toll. In which world are we maximizing healthspans, efficiency, access, innovation, economic activity, etc.? Maybe there's an AI game dev tool that can quickly prototype this?
As broken as our US system is today, no doubt it has buttressed the vast majority of capital deployment and therapeutic advancement we see today. But that dynamic is shifting on the world stage, whether talking about China or access to advanced therapies in interesting decentralized ways in parts of Europe... Maybe zero toll is an optimal amalgamation. But how to best prove or demonstrate it? Hard to imagine even a Kaiser type system would just flip the switch to this in one fell swoop, let alone the entire federal electorate/government, without some kind of pilot model proving the concept...
I don’t understand how this requires the crypto aspect. Surely the stablecoin could be dollars in a HSA type vehicle, with a UBI type payment to them.
The main value of doing it in crypto is programmability of the payments. Using smart contracts, payments can be made spendable in very specific and constrained ways. And they can expire if unspent and cash flows can be tracked / utilized to automate tax deductions and so on. With regular money in regular bank accounts, you would have a hard time doing any of that.
I suppose, although I don’t think it would be too hard to program the dispersal system to take the existing balance into account, and to limit the places the account can spend money with regular bank money. I guess it is more of a question of the optimal method of database construction
I like these ideas. In order to get started though there are at least two major components in my view that I find to be essential
1) Right now, how much access do you have, in a consolidated and interpretable way, to your medical records. I can say as a patient of Sutter and Kaiser in SF, it's possible but cumbersome to get some, and probably insurmountable to get all, and neither would be comprehensive.
2) For the average person, how much do they think about, or want to think about their health? The discrepancy between our physical relationship to our bodies and intellectual desire to know anything about them is very high. This is a gradient of course, but by and large, the degree to which people understand (or even desire to understand) their health, is a function of how much it immediately impacts them.
The economic and care provision issues are one domain, medical record access and synthesis issues are another, but the biggest is human behavior and interest.
It is both idealistic and desirable to believe that this possible, but the issues are rooted in behavior with the systemic parts making that more or less easy to influence.
This is really interesting! Do you know of any companies or research organizations that are working on something like this? I am a senior smart contract engineer and working on something like this sounds fascinating to me!
Also, speaking of blockchains that have the speed and decentralization required for something like this, Flow (https://flow.com/) is doing some amazing things and could be a perfect place for this!
This is a very interesting and creative exercise but the complexity is a problem. I have a graduate degree and can barely follow all the details of how this is supposed to work. If the system depends on your average patient understanding it, it's dead on arrival. Even if you assume the average patient normally has the wherewithal and intelligence to manage the system under normal cases, what happens in abnormal/emergency situations? Is someone in the ER with a broken leg supposed to calmly evaluate the various provider rankings and cryptocoins to decide on the optimal way of setting his bones? Seems there is a limit to what you can leave up to patients or even their designated guardians/caregivers -- some paternalism likely needs to be baked into the system.
That's what the Medical Accountant role is for. They can hone recommendations for emergency situations, and keep them updated for their clients.