Building a Health Insurance Company that WORKS

Everything you need to know to save health insurance in the United States and the world

Introduction

We have all seen the headlines: “Cigna Defrauds Medicare,” “UHC Denies More Claims,” “BCBS Settles Antitrust Case for $2.7B” or “Aetna settles suit alleging claim-denying medical director never read patient’s records.” We know we are not getting good value for our money. Today we are going to discuss a new way to provide health insurance to the people of the United States. It will be about half the cost of traditional health insurance, and on par with the least expensive care in the developed world (Organisation for Economic Co-operation and Development (OECD)) while maintaining our leadership in the best care available.

The History

The first documented thought of insurance goes back thousands of years to ancient Babylon where King Hammurabi created a code of laws that included the earliest known insurance policy. Insurance as it’s known today was invented at Lloyd’s Coffeehouse in London (yes that Lloyd’s), established in 1648, because it was a gathering place for business to be conducted. There, a merchant might have his trading voyage underwritten by a wealthy financier for a small fee called a “Premium”. If the voyage went according to plan, the financier kept the premium. If the cargo and ship were lost, the financier would pay the face value of the contract. That means that the only money the financier kept was earned premiums. We’ll talk more about this in the next section.

One Major Innovation in the Insurance Industry

There has only been one major innovation to this system and the way it was run in the 17th century: float. Float is the time between when a valid claim is made and the time when that claim is paid. In the early 2000s, with the rise of PayPal, insurers realized they too, like PayPal, could delay payment on claims and keep the money in investments and earn extra billions by delaying payments by days or even weeks. This of course meant they were quite literally killing people, but they don’t really care about anything but profits.

What we have, then, is a several thousand year old idea with a several hundred year old execution, the only innovation in all those hundreds of years since 1648 is one that makes insurance even more slow, and evil.

Here is what we do to fix it.

The Solution

Usually, in these articles, we kind of gloss over the minutiae of the plan to completely replace traditional, coffee house style health insurance, but today we are going to dive deeper and get into the meat of the subject.

The Concept

After many sleepless nights and hundreds of ideas, we have settled on a concept that will allow us to completely automate health insurance. If we, as the insurance company, were to provide our own Electronic Medical Records System (EMR) to practices and hospitals, free of charge, then we could pluck out procedures performed, and pay for them in real time. For the privilege of maintaining the patient’s medical records, we charge a flat $10 per month fee to the insured, plus, of course, the actual cost of the risk.

That’s it.

This eliminates medical coding, insurance networks, adjudication, rate negotiating, the cost of a third party EMR, all the big insurance buildings in every city in the US, the millions and millions of people that you, as the insured, pay for. In short, we eliminate everything that isn’t “pay for your healthcare.”

The Execution: How it all works

This is going to get a tiny bit technical, but plow through it, it is worth it, and I'll be brief. This is just a tiny overview of what makes us different.

The Technology

We use Microsoft products for everything that faces the client. There are a couple of open source things we use in the background for monitoring and logging, but everything you see is Microsoft.

Blazor

This is the User Interface (UI) that replaces archaic codein the browser and inserts modern, compiled, object oriented language in the browser for fast and clean code execution. JavaScript is gone, for the most part.

SQL Server

SQL Server is the world’s premier database engine and I would state it is the only engine left worth pursuing. There are arguments for that outside the scope of this article. If you have a different opinion I welcome discussion.

The Architecture

We designed and built a suite of microservices broken up along usage lines.

Master Data Management (MDM)

This is the repository of all things enterprise. All information that would be useful to more than one application is housed here, like people and companies and phones, email and web addresses.

Single Sign On (SSO)

This is the repository of all things Authentication and Authorization. Here we keep the usernames, passwords and things like whether a particular user can see or edit patient data. Of note, this is also the basis for our multitenancy, which is also outside the scope of this document, but is a table based multitenancy, for those of you familiar or that can use a search engine.

Backend For Frontend (BFF)

This is a broker layer in between the various layers of the application that adds value by coordinating the calls to Application Programming Interfaces (APIs) for each application. This was developed by Phil Calçado in the 20- teens for SoundCloud. This is basically the one stop shopping for one of our client facing applications so we don’t have to maintain many disparate connections between various APIs just connect to the one BFF and find everything you need.

Application Generation Tool (AGT)

Once the architecture was designed, I had to find a way to automate the production of it so that the cats we call developers could be herded into doing the same thing at the same time in the same way. I updated an old tool with a similar purpose to produce first .NET core code, then .NET 6 and now .NET 8 C# code. This tool looks at an existing database and generates stored procedures and API methods for Add, Update, Delete, Search, Search Exact, and by both foreign key and many-to-many relationships. This allows us to achieve the same results in the same amount of time, and a more uniform and maintainable application, as literally a hundred developers, with less than ten.

Starting Anew

If we look at any of the big EMR vendors, Epic, Oracle Cerner, Athena et ad nauseum, we find they are all doing it wrong. They almost literally develop a new piece of software for every specialty. There are 135+ specialties. If I went to the Ferrari store and saw an engine documentation system and a transmission documentation system and a cooling documentation system, I would leave my Ferrari right there for dead and go get a Porsche. They are generally better track cars anyway.

Unified Medical Language System (UMLS)

The alphabet soup is getting thick, but it makes referring to things much easier. The UMLS is a clever implementation of everything you need to know or can medically do to a human body. There are 14 million rows and you can download it for free from the National Institutes of Health (NIH). We stuck this in our relational database engine, SQL Server, so we can search it quickly. The UMLS includes SNOMED_CT, VSAC, RXNorm and several of the nomenclatures for the various other EMR vendors in the world. The genius part is that every search is done the exact same way. All the data is in the exact same tables with references to all the other information you may need. You can use the exact same queries with different parameters to get wildly disparate pieces of data. You can translate from SNOMED to ICD or CPT or even EPIC nomenclatures all with the same procedure. Again this is a survey of what we are doing, the exercise of figuring out what is in the UMLS and how to use it left to the student.

Our new EMR

The basis for our system, then, is the new EMR built on top of the UMLS that eliminates (and translates to and from) all other code sets and is the only thing available for actually documenting a patient encounter front to back, top to bottom. What this UMLS allows us to do, then, is to have one universal EMR, with no specialties, that is capable of documenting anything that can happen in medicine. This allows us to write one program. This program is maintainable by the small team that produced it, since we automated the production of new software just like we are automating health insurance. This not only allows us to produce and maintain one small lightweight program but also to be responsive to requests for enhancements. One program, one enhancement is doable and maintainable.

I can hear all the doctors screaming now, “MY PRACTICE IS DIFFERENT, YOU CAN'T HAVE A UNIVERSAL EMR!” Yes, doctor, I know, everyone is different, everyone is smart and everyone is a good driver in Lake Woebegone. Look at the statement though, Your PRACTICE is different. The UMLS is up to the task, but maybe you have some kind of procedure in place to avoid litigation or for patient safety. One of the Ambulatory Surgical Centers (ASCs) we work with has a survey with heart rate, blood oxygen, blood pressure and the like, that is filled out on the hand off from surgery to recovery. We have no interest in putting this survey into the EMR, it isn’t relevant to anything that isn't surgical, for example. We produced a survey tool that would allow a practice (or a specialty) to design and publish their own surveys. Since this isn’t really a part of the record, just a check for wellbeing, the survey is a way for the practice to modify the application for its own purposes.

Next Steps

There are going to be several problems we have yet to solve to make this all work. We can and have written the software, but that is only the first part. Below we will discuss what it will take to solve the healthcare crisis in the US and put us on par with South Korea on the costs, about 50% savings, while maintaining the best care on the planet.

Policies

Policies depend on incurred risk and as far as I can tell, either nobody really knows what this risk is or they aren’t telling me. I’m pretty sure this should be public domain, but then we could call the existing legacy insurance companies on the difference between the incurred risk and what they actually charge. That would allow us to audit their Profit and Loss statements without the accounting obfuscation and then put them out of business when we find that they keep (or waste with unnecessary work) over half or your premiums instead of returning them as benefits. Consequently, however, that means that I'll have to recalculate the risk or find someone that knows how to lay hands on it.

Support Staff.

While AI can handle some routine requests for password resets and the like. I would prefer to write code for that and have real, live, English (German, French, Swahili, whatever) speakers running the support desk. The survey tool is very suitable for creating training documents, just have to add a “How To” video to it and then the questions and answers become tests for understanding of the material. But this is all expensive.

Marketing and Advertising

This is not what you think. We have to have a way to reach not only the patients, but the front office staff and the doctors as well. The patients should be the easy part. We sell the service to self insured businesses at a 66% discount and there you are. We post the 50% discounted policies on the Health Insurance Exchange and they sell like hotcakes. The part we need marketing for is convincing front office staff that ”yes you do take our policy” and the doctors that “yes, our EMR really is better, switch to it” is a whole different problem. Marketing should be able to solve this for us, but it will be expensive.

Counsel

It costs about $1 million to become authorized to sell health insurance in any given state. We are probably going to want to get started in Texas and California, but we need attorneys to manage that process for us. We have retained Layna Cook-Rush of Baker Donelson to help us out with this. If subsequent states don’t fall in line and accept us as legitimate after the first two big states, this could run $50MM.

Regulatory Reserves

Each state has laws regulating the amount of reserve cash an insurance company must have on hand, in case there are unexpected, or unexpectedly large claims. Our insurance startup advisor, Michael King, estimates this at somewhere around $50 million average per state. Once we are established in the big markets, however, this will become self-sustaining and we can move into the smaller markets without needing additional capital.

Data Science

Since there are no denials, we have to watch for practices and practitioners who may want to pad their books a little bit. There are well known ways to detect fraud with data science. We have tapped Matthew Tichenor as our Chief Data Science Officer.

Next Next Steps

We really need a complete, end to end solution to finally fix everything that is wrong in medicine including the waste at the practice and hospital level.

Health and Wellness

84% of healthcare expenditures in the US are for behavior based, chronic disease. $4.7 trillion is what was spent in the US last year on healthcare. The average OECD country has 225 deaths per 100,000 population. The US has 335 deaths per 100,000. While that is not a dollar figure on how much we spend, like the 84% is, we can ballpark that if we were as healthy as the average OECD country we could chop about a third of the cost (225/335) out of that 84%. So we could say

$4.7T x 0.84 x 0.33 = $1.3Trillion in savings

That $1.3 trillion is about an additional 25% reduction in the cost of healthcare in the US. Yes we know that violence is about that other ⅓ but there is a mental health survey inherent in the application that will help us identify and treat at risk patients.

Practice Management (PM)

I have had a couple of arguments about what exactly a Practice Management system is. Administrators seem to think it is a billing system. If that were the case, then it would be called a billing or accounting system, and that is not what we are talking about. No, Practice Management is a fully fledged Enterprise Resource Planning (ERP) system. Larger companies use these to manage basically everything from staff, to inventory, to capital assets, to processes. I attend several conferences per year to keep my finger on the pulse of the industry, and at every one, there is a breakout session detailing how to find your “money leak.” Some accounting or business type sits in an office with a spreadsheet and does the stare and compare for days or weeks and may or may not find what they are looking for, if it is even there at all. If you have all the information in the same place at the same time. Finding this is a one click report.

We do have a design for this system, but it is not yet complete. The base code has been generated with our code generation tool, but we have to go in and make sure it is ‘pretty’ and has all the functionality and logic that we can’t automate the production of.

Other Applications

Already designed and built, are a secure ‘email’ system, since you can’t email medical information, a telemedicine application, a DICOM viewer for x-rays and MRIs, a way to draw on the DICOM image so that structures can be highlighted for emphasis, a self scheduling tool, the questionnaire tool we talked about earlier, the reporting tool we discussed earlier, and of course as events warrant, the compliance and efficacy tool to automate the reports we need, using the reporting tool as a vehicle. There is also a built in research tool that can search for any piece of information and search for its associations. For example, if your patient presents with a list of symptoms, you can type those into the research tool and it will give you back a list of diagnoses that have been associated with those symptoms. You can also go the other way with a differential diagnosis and then look for symptoms you may have missed. This suite of applications should give us everything we need to automate as much of medicine as possible and therefore, give us all the tooling we need to save the most lives possible at the least cost.

Conclusions

We have shown a way to save more than 75% from the cost of health insurance and have addressed both of the big problems with health insurance in the US: cost and education. We have all of this written and deployed in a prototype application. The only thing we really need to get this all started is to clean up that application and turn it into an enterprise system with logging, administration and redundancies in hardware. We will need funding, probably about $10 million over the first year. For comparison, United Healthcare had revenue of $371.6 billion and net earnings of $22.3 billion in 2024. With about 50% upfront savings we should service and retain 90% or more of the 330 million insured people in the US. That gives us a revenue of $36 billion.

This figure shows that this is a viable business proposition.

We have shown a way to make patients healthier by educating them on the consequences of their behavior, and a way to capitalize on that to the sum of $1.2 trillion or about 25%. If we add that to the process automation savings of our solution, we are in the ballpark of more than 75% savings in total. We already have the best doctors and the best equipment; we just need to implement the above detailed framework to give them all the tools necessary for success.

We have this system in prototype now, fully functioning.

If you liked what you read, please like and subscribe, click on the notification icon, subscribe to our newsletter, and follow us on all our social media and blog sites.

We have built a comprehensive health information system to keep the patient healthy and on the right track with the ability to incentivize healthy living. Implementing this system should be fairly simple and will completely revolutionize the way healthcare is paid for, saving countless lives. We have shown a way to use this system to make the best healthcare system in the world also the most efficacious and the most affordable.