Standardized and Decentralized — Redefining Medical Databases
Fragmented… Inefficient… Sporadic…
These are all terms that describe what current healthcare records look like right now. Every medical practice has different Electronic Medical Record (EMR) systems that are unable to communicate with each other.
Want to send healthcare medical data to another hospital? Great… export the data from this EMR system, figure out a way to standardize it to fit within the bounds and data points of the new system (combat variation in the type of data points), and import it into the new system. Totally easy, simple, and seamless — no problem whatsoever 🤣
TL;DR: These systems are incompatible and don’t communicate with each other.
Medical record systems currently are chaotic as HECK! The amount of moving parts in this system is insane. It almost seems dumb how these systems don’t interact with each other.
Note: This article is based on North American-oriented solutions as opposed to global solutions.
Why is this important?
Healthcare systems run on data. Diagnoses are made from data. Medications are prescribed by analyzing data. Not having efficient systems for patient data is essentially our healthcare system shooting itself in the foot… multiple times.
If this hasn’t convinced you enough, having accurate data within the healthcare industry is everything — but having it FAST is just as important. Every time we go to new hospitals with different EMR systems, we’re dealing with more and more data collection to rebuild this medical history, or painfully transferring old medical history into a new one. Our current systems are:
- Inefficient
- Unoptimized
- Problematic
These systems aren’t just painful or inefficient, they’re disastrous!
“Well, why don’t we just make these systems communicate with each other?”
I’m glad you asked — doing this means standardizing every piece of data from every EMR system in every medical practice... and THEN proceeding to attempt to figure out compatibility solutions for every system to interact with one another… If it was that simple, it would’ve been done already 😢.
BUTTT… what if we created a Universal, Decentralized, and Standardized Electronic Medical Records system, or a UDS-EMR as I like to call it? Without a centralized authority, our system itself is more universal without an innate imbalance, and standardization comes with the fact that it’s a single, standardized system as opposed to 15 scattered ones.
That’s exactly what I did.
I created a database (decentralized application) on the Blockchain in the form of a Smart Contract that does exactly this… In other words, I created a USD-EMR.
Why Blockchain? 🤨
Note: If you don’t know what the Blockchain is or what a Smart Contract is, I highly recommend you check out this article. It’s not necessary but it is preferred.
Blockchain has two 🔑 aspects that are super important when it comes to this problem:
- Blockchain is IMMUTABLE — data can be added but not deleted
- Blockchain is DECENTRALIZED — all copies of the database are the same and distributed across everybody
Immutability
Beginning with immutability, info in this database cannot be changed once added. This means that data accuracy remains intact, transactions are secure cryptographically, and trust is no longer placed within a centralized system. Instead, we’re trusting in the blockchain technology itself, a technology worthy of our trust.
Decentralized Nature
Blockchain is decentralized. Unlike current systems, there’s no central system that runs it like the government or a specific certain medical practice. This lends itself to its standardized nature of both data and communicative ability. There’s actually some super cool Artificial Intelligence development impact of having standardized data which I’ll bring up later.
Another thing to note is that there’s no longer a single point of failure. Since there’s no centralized system, it becomes much more difficult to hack into these databases without a 51% attack (an attack in which hackers take control of at least 51% of the network in order to take it over). The blockchain itself can be distributed among all users, ensuring that everybody is on the same page about what’s going on… LITERALLY on the same page (do you get the joke? same page, same webpage?… it was funnier in my head 😭).
This IS a Blockchain Project!
I’ve heard the argument that Univeral EMR’s shouldn’t be a blockchain project or that this isn’t a great application for this industry… That being said, blockchain isn’t just a great application for this problem, it’s the future.
Let’s think about this for a second. We’re looking for a standardized system that’s secure (screams blockchain), is decentralized (screams blockchain), is the exact same no matter who’s looking at it (screams blockchain), allows for secure transfer between medical facilities (screams blockchain), and is distributed so everybody has access to it and authorized users can access patient data (screams blockchain).
If this doesn’t scream blockchain, I don’t know what does.
This isn’t just a change to the status quo, it’s a redefinition.
Privacy: Comparison with Status Quo
While working on the project, the biggest thing that I tried doing was disproving my theory — the theory that this decentralized solution is actually BETTER than the status quo. This meant research on current tactics to keep EMR systems secure from a privacy aspect.
Sooooo what were the results of the research? 🤔
Well, I’m so happy you asked — and if you didn’t, pretend you did 🔪
Current methods for safety and privacy are literally regular cryptographic methods and strong login systems.
These systems are not BAD at all. They’re actually pretty good privacy measures. But that’s the thing… our new UDS-EMR system can use the exact same encryption methods.
For example, the backend of this database can be made on the blockchain. Private keys can be used as an authorization method to make sure only doctors working on these cases have access to this information, acting as an incredibly secure measure due to the nature of the blockchain.
Quick Definition: Private keys are essentially the identifiers for our cryptocurrency wallets. Think of it as proof of ownership for the wallet.
On top of that, the EXACT SAME encryption and cryptographic methods can be used as protection!
In terms of privacy, we’re essentially taking the current status quo of data privacy for EMR systems… and literally one-upping it. This means we genuinely have no privacy-related incentive NOT to implement this solution.
That doesn’t mean this solution has no problems. In fact, there are two major problems that come to mind initially.
Problems and Roadblocks 🥲
This article isn’t meant to market this solution. It’s meant to explain what this solution is from a holistic approach — both the good and the bad. This means describing its problems as well 😔.
Before going into that, let’s understand exactly how our database works!
Logistics
First, let’s take a look at what data points we can access:
If you take a look on the left, this is a base prototype for potential data points that we can access when adding new patients to the system. HOWEVER, before complete deployment, we can add as many of these data points as we want. However, there IS a caveat to this:
Once the smart contract is deployed, it CANNOT be changed
We cannot add new sets of data points that don’t fit the bounds of the smart contract. However, we have access to current data points for healthcare records and can compile all of them together to cater to all these EMR systems. This also fosters STANDARDIZED DATA POINTS, a potential massive innovation within the healthcare industry (you’ll see why soon enough 😏).
We can search for candidates through any of these data points (name, ohip #, etc., id, etc.) For the purposes of this article, replace ohip # with a general healthcare identifier. I only put in OHIP as a placeholder since I’m familiar with it, coming from Ontario. However, all provinces in Canada DO have unique healthcare identifiers which can all be added to this system to make it universal.
This is an example of how we can utilize this search function to get data from the database using these unique healthcare identifiers, functioning in the same way that a normal database would.
NOTE: not everybody will have access to this patient data. All this data is stored on the blockchain, but won’t be accessible without authorization techniques discussed in the privacy section. The privacy concerns for data getting into the wrong hands are the same, if not lower than the status quo.
Immutability: A Double-Edged Sword ⚔️
We’ve already discussed why immutability is a great thing for this system. That being said, it’s also a double-edged sword.
Immutability means every single time the database is to be updated, a NEW block needs to be added to the blockchain. Now, this isn’t necessarily a horrible thing, but it means that there are gonna be a lot of redundant blocks in the database. This can be combated in one of two ways:
- Database search only shows NEWEST entry in the database
- All old databases don’t show up on the front end of the application — a front end is what the user sees, such as a website or application
Unfortunately, there are various repercussions to systems like these. For example, old entries that might need to be accessed may not have an easy mode of access. However, this CAN be offset in development through version-searches (search different versions of entries).
For example, in an entry for the name Aayush Grover, we can show version history on a front-end application, offsetting this problem. This means that I can see every iteration of the data under my profile.
Implementation Tactic — or Lackthereof 😬
The healthcare system is NOTORIOUS for having horrible implementation tactics. Currently, we have a system that works. We don’t have a system that’s efficient, hence why this project exists in the first place. This means that the widescale implementation of a project at this scale, a universal, decentralized, and standardized electronic medical records system, could potentially take a LONG time (5–10 years). The fact is that the healthcare industry innovates at a much slower pace than any other industry. Unlike other industries, healthcare plays with people’s lives. Innovation in this industry cannot afford to be rushed.
HOWEVER, this doesn’t take away the fact that this IS the future. We’re moving towards a future where the entire internet is decentralized. This isn’t a possibility, it’s an inevitability.
Potential AI Implications 🙀
As somebody who’s attempted to breach the healthcare industry with AI applications, I know all too well how painful it is to get medical data that are standardized. It would be an absolute shame if I didn’t talk about how crazy this would be for our healthcare system.
If you want to learn about the role and implications of AI in the healthcare industry, read this article.
There’s no doubt that AI will be a major disruptor within the healthcare industry.
AI needs data. It runs on the fundamentals of working with data, manipulating it, analyzing it, and leveraging it.
However, there are two primary issues when it comes to this industry….
- Confidentiality of patient data prevents the ability to develop AI algorithms due to the difficulty of getting your hands on any type of data
- DATA ISN’T STANDARDIZED! You can get one set of data from one healthcare facility and another set of data from another which is what normally happens, especially since the required data for AI can be in the tens of thousands of data points. However, you may be entirely unable to leverage them in an Artificial Intelligence algorithm due to a lack of standardization.
Band-Aid to Confidentiality
Recall back to the logistics section of this article. You may have noticed a “dataUsageAgreement” datapoint. We can essentially compile a dataset in which patients can agree to have their data used for the purposes of these algorithms.
In our prototype, our datatype was of type “string,” aka text. However, the Ethereum blockchain actually allows us to store digital signatures through the validation of our Ethereum address. If you haven’t connected the dots yet, we have an incredibly secure way to validate whether or not these patients actually agree to utilize this data for these algorithms.
However, I must admit.. this is not a solution, it’s a band-aid. This that we’re not entirely solving this problem. However, still progresses this problem in relation to the status quo.
Data Standardization
Here’s where things become crazy. We’re now dealing with one LARGE database of decentralized patient information that’s standardized in terms of data collection techniques. This means that this ISN’T just a band-aid, it’s a potential solution.
This is a massive consideration we need to make, opening the doors for potential enormous AI disruptions that weren’t even close to possible before. It’s definitely something to look forwards to and be excited about!
Next Steps… The future of Healthcare
We’ve identified a potential high-impact solution to the problem of fragmented healthcare data. The implementation of USD-EMR could potentially disrupt this industry. It’s important to understand however that there are major gaps in a project as such, mainly with implementation.
Regardless of technical gaps, there’s great potential that an idea as such could be rejected by the industry itself and certain hospitals refuse to implement this. This can still allow for an impact to be made but defeats the entire purpose of a universal system.
The fact is that this is how the world of startups works. A really cool, legit idea can mean nothing if we identify things like implementational challenges, and that’s one of the biggest realities within the healthcare industry.
Before ending of this article, I want to mention something important. Google was the 10th search engine to exist. It wasn’t the first. It took 10 iterations of a GREAT idea for it to become the standardized norm.
The biggest thing that I’ve learnt working in both AI and Blockchain within the healthcare industry is that this industry in particular is nearly impossible to innovate in. Every person I’ve met has told me to my face that the idea of working in this industry from an innovative perspective is satirical. But this is an industry in which innovation doesn’t just save money, it saves lives.
Let’s impact this industry… let’s impact the world.
Check out my video on this project here:
If you liked this article, please check out my other articles here, or check out my article on the ERC-20 technical standard here.
Feel free to check out my other socials: LinkedIn, Twitter
Consider subscribing to my newsletter here! I talk a lot about my progress, experiences, and my struggles. My January issue just came out where I discuss my relationship with confidence and what that means to me on a fundamental level.
