Persist Health

• 

  Doctors view with a sponsored section of Pharma companies on the side with clear complaint notice along with sponsored tags

• 

  Persist's role between a doctor and a patient

• 

  A Doctor's view of a patient's EMA data

• 

  A Doctor's view of a patient's EMA data

• 

  A 1-page-summary option for clinicians to review a patient's history after they left their clinic last session

• 

  AI analysis of the data, to make sure important details are not left out

• 

  Patient view where they see options to log their mood and a journaling place; While the realtime view of their wearable

• 

  Patient's view of their journals and mood logs

• 

  Patient Notification center for automatic triggers/Manually sent by the clinician

Inspiration

We realized that a good portion of time in an appointment with a Healthcare Practicitioner is spent asking patients questions like "Can you tell me what you did this week?", "Did you getting enough rest?", and "When did you start feeling pain?" Question about events that happened most of the time more than a week ago and thus relying on flawed human memory and the cognitive bias that comes along with presenting oneself as more "normal".

Ecological Momentary Assessment (EMA) is a technique that uses repeated collection of people's ecologocal data at the time of the event. This removes recall bias; instead of asking people how they slept a few weeks ago, EMA reports it at the moment. EMA can range from breathing rate, heart rate, o2 levels etc.

This issue inspired us to create Persist Health, an AI integrated tool that persists between your sessions, during your session and after your session with a Healthcare practitioner.

What it does

Persist Health is an AI integrated tool that helps doctors understand the patients' health better while improving efficiency for the patient so they get most out of their appointments and in-betwen support. Persist Health is provided pro-bono to the patient and the doctor, while sustaining its cost from the Pharma companies who are willing to promote their medical methods/drugs. By complying with regulations, Having clear sponsored tags along with other drugs in the market, A sidebar with no patient identifiable information supplied to it can be found when a doctor looks at a report of a patient.

For patients, Persist Health lets them log daily health observations in real-time—mood, sleep, pain, energy, and journal entries. The barrier of entry is really low, as we integrate with the wearables that they might already be using; It takes two minutes. Because data is captured when it happens, not weeks later, it's accurate and complete. Over time, patterns emerge that patients don't realize. The app tracks trends, spots correlations, and alerts patients to schedule follow-ups or seek urgent care when needed

Doctors deal with unintuitive UI all time time. Having to observe a CSV file or an excel file full of numbers is not very ideal when the stakes are high. Our product provides a beautiful user interface with AI features that makes identifying trends and patterns very easy.

How we built it

Persist Health uses React to handle the front end of our application. We have express as our server. We used MySQL for the database. MySQL database and the Express server are both deployed on Vultr. We made use of the $100 credit we got. We got the domain persist.health from godaddy which we connected to the app using Cloudflare. We used openrouter for any AI api's for the prototyping.

Challenges we ran into

We wanted to build something that beared no cost to the end user, so us matching drug options based on doctor's input and patient's data which helped with the whole process seemed amazing, but we realized healthcare compliance is complex. The Anti-Kickback Statute prohibits any payment or remuneration tied to referrals, and we needed to ensure our matches were only discriminatory based on evidence. We discussed with alot of representatives from great companies and got to the conclusion that there is however a way to do this in a legal way. Which we implemented.

Accomplishments that we're proud of

We are proud of coming up with a novel idea and actual producing a working prototype thats working. We also a bucketlist of plans that we can work on for this project if it were to get attention and attrack interest.

What we learned

We learned quite a lot in this hackathon. For Chris, this was his first hackathon, he was new to React and learned how create a web application in React. For Nikas, he learned more about HIPPA compliance, FDA regulations. We visited virtually every workshop we could attend. We also learned to leverage Hokie AI, we build a sidekick for our application there. We both also learned alot about developing for healthcare as a whole.

What's next for Persist Health

There are a lot of thing in store for the future of Persist Health. For one, a more fleshed out version with stronger guard rails, better security are a must. Anything that involves AI is really vulnerable to attacks these days, so keeping up with the CVE's is something we look forward to. We have a go-to market strategy as well, If this makes out of the hackathon, we would love to work with experts and come up with a more polished version of Persist Health.

Built With

• express.js
• godaddy
• mysql
• openrouter
• react
• vultr