A Smarter Way to do Health Insurance
Have you ever had a bad experience with health insurance? Chances are you have. Perhaps a claim was denied, the pre-approval process resulted in a delay in medical care, or maybe with average premiums for employer-sponsored plans costing families $5,700 annually you’re one of the 9.8 percent of Americans forced to forgo health insurance altogether. Don’t forget about that time the uninsured were penalized!
It’s hardly an overstatement to say that everything about health insurance is simply awful. For most Americans, insurance is tied to their employer (a strange relic of World War II) which limits job choices and employment mobility. If instead they’re one of the 36 percent of workers that are freelancers, they’ll likely need to purchase a more expensive private health plan. With average annual single deductibles at $1,655, many shy away from necessary medical care to avoid paying out of pocket. This of course potentially increases the risk of developing larger complications down the road. Young adults need to remember that coverage under their parents’ insurance plan is probably going to be cut off when they turn 26 years-old. For physicians, the billing and reimbursement process is a cumbersome and agonizingly bureaucratic mess that results in payment delays and endless piles of paperwork. Billing complexity slows down the timely delivery of medical care, especially with regards to Medicaid (a program that 77.9 million Americans are enrolled in). This is a major reason why doctors spend more of their time writing medical charts to comply with employer and insurer billing requirements than they do face-to-face with their patients.
Of course, insurance companies also have an inherent conflict of interest in denying claims to maximize profits for shareholders. This puts patients and physicians alike in a constant battle with insurers who may end up exerting a fairly substantial influence over the plan of care. In many cases, insurers effectively end up practicing medicine de facto. Case in point:
And it’s not just private health insurance that’s the problem. Public plans are bad too. With regards to Medicare, under which 62.4 million Americans are enrolled, perverse incentives end up stifling innovation. It’s also fraught with fraud, accounting for an upwards estimate of 10 percent of all Medicare spending. Just as with Medicaid, reimbursement rates may be tinkered with but they rarely ever see a net increase since it’s a zero-sum game that requires a literal act of congress to change.
Ready to find out how much all this is costing us? Brace yourself.
U.S. healthcare spending grew 9.7 percent in 2020, reaching $4.1 trillion or $12,530 per person according to CMS. This represents 19.7 percent of our national gross GDP. Of that, administrative costs represent a whopping 34.2 percent of overall healthcare spending. We spend significantly more per capita compared to Canada with administrative costs the biggest contributor to the difference. Dividing this even further, we can see that billing and insurance-related costs are the largest component of administrative spending coming out to 18 percent of overall healthcare expenditure. Perhaps unbelievably, it’s quite possible your hospital employs more billing clerks than it has beds for patients! In the end, insurers are the ones really running the show.
It’s obvious we need to change something. But what could it be? We’ve had nine years of open Affordable Care Act marketplaces with their increasingly clear effect on stabilizing insurance premiums. This could be considered a good start although it’s not the paradigm shift we should be striving for. Some propose we enact a public option or even a single payer system like Medicare for All. There are many angles to this topic and it probably deserves its own post. Suffice to say, the sheer improbability of such a bill being passed in this country makes it a non-starter when searching for near-term solutions.
So What Can We Do?
Let’s go back to the drawing board and be idealistic for a second. Imagine we have a completely blank slate to build off of. What would a ‘realistically-perfect’ health insurance system look like? How much does it cost? What values do we imbue in it? If such a solution were available on the market today, would it be able to fundamentally outcompete existing private and public plans? Is there a way to reduce, if not eliminate, the need for rent-extracting administrators?
I’d argue an ideal system is one that is provably fair, underwritten and overseen by its policy holders, and uses automation as much as possible to minimize administrative bloat. It isn’t controlled by a centralized authority with unilateral power and corruptible motives. It is instead structured in a way that broadly decentralizes governance of its operations, is anti-fragile, and facilitates goal alignment. Among its member base, a spectrum of plan options and risk pool depths are available for entry and they are designed by fellow policy holders who share similar values. Each individual component to every single policy is written in plain language and easily understandable by anyone. Decisions related to claims approval are algorithmic, consistent, transparent, and ultimately equitable. Switching policies is a seamless user experience with an instantaneous transition and no need to notify doctors offices. The confluence of all these features results in a more accessible and dynamic entry point to the healthcare ecosystem. Such a system would represent a giant leap forward from where we are today compared to an incremental adjustment of the status quo. In effect, we’d be placing power into the hands of patients— which is exactly where it belongs.
You’re probably wondering how something like this could possibly work. In recent years, a key enabling technology has been developed that can help coordinate distributed teams as well as efficiently process insurance claims without a profit-seeking or bureaucratic intermediary. It’s called a smart contract. Smart contracts are essentially rules written in computer code that automatically execute when certain conditions are met. Smart contracts are stored on a blockchain, usually Ethereum, which means they are cryptographically secure and anyone can view and verify their execution. Once deployed, their actions are self-enforcing, can be predicted precisely given a known input, and cannot be censored or altered.
Despite being openly auditable, smart contracts applied to health insurance can be written in a way that completely preserves patient privacy (i.e. no personal data is kept in the contracts themselves— only unique, anonymous references). Their implementation in the insurance industry could eliminate large swaths of expensive administrative middlemen and result in substantial cost savings. An added benefit of living on the blockchain is that data from contract execution is forever preserved. So if someone were to attempt insurance fraud then evidence of their crime will be immutably locked into a public database. Anyone with some data analytics know-how could develop an algorithm to continuously scour transactions for anomalies. Perhaps these insurance contracts have a bounty system programmed into them that automatically rewards people that uncover fraud with a cut from the return. Smart, huh?
Useful applications of smart contracts in healthcare are actively being explored. They can help ensure the integrity of drug supply chains, securely store and query pharmacogenomics data, manage non-custodial medical records, service a decentralized medical image hosting platform, and streamline medical credentialing and licensing. Any number of different smart contracts can be linked together to create incredibly powerful services. For example, they could serve as a backend for the Internet of Medical Things (IoMT). In this case, real-time patient monitoring with automated responses is made possible through the secure management and analysis of medical sensors. These data can then autonomously interface with health records, physician apps, and third-party artificial intelligence agents.
Despite this progress, a comprehensive health insurance product constructed using smart contracts is yet to be designed. The scope of this endeavor along with unique design and integration challenges is admittedly daunting and likely why no one has attempted it yet. This does not mean, however, that it’s impossible to create or that we shouldn’t try. In fact, the necessary underlying smart contracts that define the logic of basic insurance operations have already been developed and open-sourced at etherisc.io. Example products built on top of their platform include flight and crop insurance. One of the beautiful properties of smart contracts is that innovators can easily plug them in to existing protocols like Etherisc which means they don’t have to reinvent the wheel.
With this foundation set, all that remains in the pursuit of an algorithmic, decentralized, member-operated health insurance product is the design of healthcare-specific requirements. So all that’s really left is assembling the right team to get it done. This team should come from a crypto-native community with key insights into how the healthcare industry works... and that's Medaverse!
Uncovering What’s Possible
Let’s take a look at how a relatively simple integration might work using medical malpractice insurance as an example.
For healthcare providers, insurance that covers malpractice is a must. Malpractice insurance is a form of professional liability protection, shielding policy holders from potentially significant financial damages in the event of alleged or actual malpractice. In the unfortunate circumstance a claim is filed, the insured’s policy will generally cover legal defense expenses up to an annual or lifetime limit. There are some organizations like Doctors Company and Applied Medico-Legal Solutions that are owned by their member physicians (a so-called ‘risk retention group’). They may be more value aligned than non-member owned companies but they are certainly not immune from the pitfalls of traditional centralized approaches that plague the industry.
Now imagine a malpractice insurance company that is also member-owned but instead it differentiates itself from competitors by utilizing smart contracts to streamline the claims process and to decentralize its operations. Much of the rest looks exactly the same from a user experience point of view: insureds would sign a relatively standardized policy contract determining coverage duration and type (claims-made vs occurrence, tail coverage, specific conditions and exclusions). Their premiums, the amount of which is determined by an open-source risk model, are then pooled together and used to cover claims payouts. Rather than hold the pooled funds inside a corporate-owned bank account, this company opts to lock them inside a smart contract— a battle-tested and commonplace practice in the world of decentralized finance. This means policy holders (and anyone else in the world, really) can view every single transfer of funds both in and out of the smart contract as well as calculate and verify the risk pool safety margins for themselves.
Relevant talk begins at 26:47.
In the above podcast, Mark Cuban describes how an insurance claims process might work using decentralized decision-making and smart contracts. Say a physician is sued and files a claim with their malpractice insurer. The data the physician submits is sent to a distributed network of people trained by the insurance company to verify that the policy holder has indeed been named in a court filing and that no exclusion criteria apply. To ensure no single person controls and gate-keeps this process, a unanimous decision is required by at least three verifiers. If there is a disagreement, or the decision is appealed by the claimant, or the outcome is questioned by someone acting as a challenger, then the claim is escalated to a broader audience of verifiers. These second-order verifiers are publicly elected to their position by demonstrating relevant qualification (e.g. subject matter experts, medico-legal attorneys) and they determine the company’s official decision. To maintain checks and balances, they have the power to slash the first-order verifiers and challengers rewards should they have acted in bad faith. This process ensures the ultimate decision reached by the insurance company is fair, accurate, and consistent.
When a payout is authorized, automatic and trustless execution of the contract terms by the underlying Etherisc protocol kicks into gear. The physician gets their legal defense paid for and the company avoids expensive overhead from processing the claims using humans. All of this means the malpractice insurance company is maximally transparent, consistently fair, and as structurally lean as possible. Competitors would struggle to match this level of efficiency.
P2P (Patient-to-Patient)
Malpractice insurance is relatively simple compared to what’s required for health insurance. There are many more layers of complexity including types of claims (ICD-10, CPT, HCPCS), plans (individual, family, employer-sponsored), provider networks (HMO, PPO), and levels of care (primary, specialty, emergency). Also in need of consideration is state and federal regulatory compliance, billing systems integration, and the unbiased calculation of individual health risks. To wrap our heads around all of this and see the potential for disruption, let’s again go through a simplified example. This time: dental insurance.
Let’s say people in Philadelphia are fed up with their dental insurance options and decide to take matters into their own hands. They start by connecting with each other pseudonymously through a dapp (decentralized application) built by Medaverse. Just as in the malpractice example, this dapp is powered underneath by Etherisc and facilitates the pooling of funds into a smart contract for claims payouts along with a decentralized verifier and challenger system. A new feature this time is the crowdsourced creation of provider networks through collective bargaining with local dentists. There are compounding network effects at play here: the more people that join, the greater the pool of funds, the wider the provider network grows. Using tools and insights provided by the dapp, Philadelphians begin crafting dental plans that take into consideration statistical risk models, safe capital pool margins, and payout configurations based on data oracles (which are essentially sources of real-world data). Remember, geographically-tied communities share similar health risks and so they will have to be systematic and unbiased with regards to collecting and computing all relevant variables. They can choose to restrict access to only those with an address in Philadelphia, widen it to some greater region, or keep it completely open for anyone to join. This flexibility in choice and method of finding product-market fit is only possible through decentralized decision making.
Once these dental insurance policies are codified on the blockchain, new users can then sign-up and select a plan based on whether or not their dentist is in the provider network. If there’s a specific plan the user likes but their provider isn’t in-network, the dapp can assist the user in lobbying their dentist to accept it. If no plans look appealing, the user can go through the process described earlier to design a different coverage plan or kickstart a new provider network using accessible tooling. Since this insurance collective is managed by a flat hierarchy, policy holders can also vote on what to do with revenue from premiums that exceed risk pool capital requirements. Do they reinvest in provider network growth? Subsidize low cost plans to increase dental care access? Sponsor public health campaigns? Decrease membership premiums? Ultimately, the future direction of the insurance collective is determined solely by the people that put money into the contracts in the first place. This is much better than just letting profits get siphoned off by an unelected cadre at the very top.
Linking it all Together
The principles of automation, checks and balances, decentralization, and trust-minimization described in the previous two examples can be applied to other aspects of the broader vision for a ‘realistically-perfect’ health insurance system. The goal here is to create an accessible and user-friendly platform that empowers communities to create their own medical insurance plans of all types— vision, dental, health. Viewed through the lens of a permissionless dapp powered by smart contracts, communities can now spontaneously coordinate over the internet to design and deploy value-aligned health plans, pool together the funds needed to ensure coverage of their collective risk, and then oversee their own governance.
Those added layers of complexity mentioned before— levels of care, billing systems, employer sponsorships, dynamic tooling, etc— can now be viewed as stackable features whose engineering solutions are only a matter of time before they get developed and integrated. In this instance, Medaverse guides development and acts as a parent foundation of sorts— it’s the platforms founding decentralized autonomous organization (DAO). Medaverse DAO develops the dapp, maintains the regulatory sandbox it operates within, protects the judicial independence of the verifier / challenger system shared amongst users of the platform, prevents and firewalls systematic risks, and plays an advisory role for specific policy plans and dispute cases.
There are definitely still aspects to this whole ‘completely redesign the entire insurance system’ thing for us to figure out. Anonymity on the blockchain needs to be accounted for and preserved in the event an insured reaches a private settlement. We need to figure out ways of automating the claims process on the provider side to reduce the burden of paperwork and free up countless hours for clinicians. Processes have to be in place to prevent and minimize the impact of any potentially unethical, inequitable, biased, or just plain wrong or broken ways users find to configure the system.
That being said, this space gets really interesting when we look into the future and consider the ways we might integrate Medaverse DAO and this as-yet unnamed health insurance dapp into the Internet of Medical Things. Here’s just one example for now to wrap things up: as more and more data from medical sensors stream into the IoMT pipeline, we will be able to securely and trustlessly compute and serve patients with personalized health recommendations. When these recommendations are implemented, which at times or most often will happen under the supervision of ones trusted physician, the returned data feeds can then be used to dynamically adjust individual health risk scores. Could we then incentivize, on a personalized basis and in near real-time, policy holders to implement healthy behaviors with financial rewards such as lower copays, deductibles, or premiums? What happens when we supercharge entire communities with this ability and find out how they optimize for their collective health? I certainly don’t know where all this leads us but it sure sounds exciting and I definitely want to find out.
If you want to support Medaverse in its endeavor to push forward projects like this, consider donating or joining our Discord server to help make them happen. To contribute your skills or to check on the status of our endeavor into smart insurance, head over to our project workspace on Notion.