XPC Hackathon Spotlight: Jung Hoon Son from EpiNet

Offcall recently sponsored the XPC Hackathon, an innovative digital challenge hosted by X=Primary Care and Dr. Paulius Mui designed to invent and showcase ideas that transform care workflows. The event brought together innovative teams of individuals determined to make care workflows smarter, faster, and more human, and the results were incredible! As part of our commitment to uplifting clinicians who are working to transform healthcare by putting fellow healthcare workers at the center, we are spotlighting the winners of this year's XPC Hackathon here. For the full list of winners and video descriptions of their companies and pitches, see here. In this article, we feature third place winner EpiNet led by Jung Hoon Son. For a profile of second place winner Shepherd Health, led by a team including Aylin Özpınar and Jack Talley, see here. For a profile of the first place winner Wellbeyond led by a team including Russell Stover, see here.

1. Jun Hoon, tell us about the problem you were trying to solve and the key elements to the solution you built? Both infectious disease reporting and quality metrics collection in healthcare are antiquated and messy. More often than not, it’s a crowd-sourced “sneaker-net” messy effort that borrows resources from the data sources, often involving many workers in health systems and public health departments to do mandatory reporting to centralized entities (e.g. local/state department of health, CDC, or hospital administration). Modern systems could easily alleviate that requirement by leveraging a seemingly simple rule: “If you see a positive COVID lab, add 1 to the counter.” Same for “A1c > 8” for monitoring diabetes.

The concept is simple: Everyone can create a “wallet” on a testnet of Bitcoin, Ethereum (I picked Polygon). Every time a lab/EHR receives a positive COVID test result, one token is transmitted to a “COVID_POSITIVE” wallet using a plug-in designed in Canvas Medical's EHR’s SDK. The amount of token the wallet has is equal to the total number of positive COVID tests. Imagine that for all the clinical labs in the U.S. for all the lab results. Expand that for quality metrics.

I wanted to showcase the plausible proof-of-concept that leverages public ledger (blockchain) to “count” the number of labs with positive / negative results, which can be used as a proxy for disease spread. Labcorp/Quest and larger systems can easily implement this, even today. Add an actual dashboard (effectively re-labeled “blockchain explorers”) that can read the wallet information, and then we have a complete decentralized solution.

2. Why is this such an acute pain point that you think is worth solving? This project was inspired by the fact that CDC was facing major resource cuts. The flu tracker, and surveillance mechanisms are all facing funding cuts. Also, I’ve seen first hand during residency how much paperwork is required for reporting things like transfusion reaction - all usually spills over to 10 mins of clinical resources at multiple levels. Those 10-30 minutes x 10 people involved in reporting a single case of infectious disease report seems like wasted patient care time, and I want to free up time.

This doesn’t directly help clinical workflows, but borrowing from the UI/UX world, “good workflow should be invisible.” If you can save FTE administrative work per case, that should trickle down hopefully to other areas of work, like clinical patient care workflow.

3. Where will you go from here? What are you planning to do with the solution you built? I do think there is a remote possibility that a solution like this would come in with proper incentives and funding, like the GPS infrastructure that’s funded by the U.S. for $5 million a day but the entire world can benefit from it (read about it!), without any subscription costs. A worldwide network for disease tracking would do wonders for all, especially if everyone has access to the public ledger.

Blockchain does not have good rep because of its association with cryptocurrency. But the “distributed ledger” portion of the technology is very unique and useful for keeping track of “counts” that need to be more public information.

If CDC, NCQA, or HHS can fund this, it should be very easily implementable. It admittedly doesn’t replace existing solutions because it can’t do individual-level tracking, but it’ll be close to weather forecast APIs. The data would be enough for public consumption, and not limited to U.S. jurisdiction, since anyone can “push” data.

4. What was it like to participate in an XPC hackathon? What was most memorable? I happened to have a good 30 hour chunk of the weekend free, along with Cursor credits to burn. My solution was far from actually clinical, but getting all the pieces. I felt bad for going solo but kinda had to because it’s more of an infrastructure-implementation. It was generally my first time programming something for blockchain but modern coding AI IDEs really helps when one needs to implement solutions that you know have been implemented in the past. I just knew I didn’t have to type out the code.

Canvas Medical to fire triggers whenever the COVID lab comes back with a positive result was the biggest challenge since I never used Canvas. Last-minute addition of “Zombie virus” LOINC code to trigger a separate hypothetical zombie outbreak was part of my goofiness and fun coming out.

The most memorable: I submitted the video that was cut off at the end because of the 5 minute time limit on Loom, just as I mentioned the zombie apocalypse happening. It had some cathartic humor in it, so I submitted it as is, Cloverfield movie style.

5. What is your #1 advice to fellow clinicians / physicians who are curious about entrepreneurship but may be hesitant to take the leap?

U.S. healthcare and medicine won’t change from the inside. It might not be comfortable, but if you can and have the means, I’d advise you to think outside the box. There’s no one who knows how healthcare will change, but being complacent with the system we have is the first step to becoming obsolete fast.

To learn more about the XPC Hackathon and read the entire breakdown of the winners and event, see here.