The inspiration behind Hepius was to change the patient-intake experience by: A) decreasing physical traffic and overcrowding in ER lobbies, B) allowing low-severity ER patients to queue virtually from home, C) route low-severity patients to more appropriate health facilities (walk-in clinics, urgent care centres as opposed to ER departments in hospitals), and D) improve patient intake experience. A couple of our teammates was born in Alberta, Canada, the idea came from the fact that the average ER wait time for a low severity patient is over 3.5 hours. Often times, prospective patients unsure of the severity of their conditions show up at a hospital seeking expedient care, playing the situation safer rather than “sorry”. This leads to the high presence of low-severity patients in traditionally high-severity facilities, leading to the extraneous costs of simple treatments. This situation is incredibly undesirable given today’s COVID-19 environment, where overcrowding areas can have potentially disastrous outcomes. Our solution seeks to provide these patients with a centralized, digital entry point for various facility treatment options, and would route patients to the most appropriate facility with the shortest wait.
What it does
Our system is developed to separate lower-severity patients from higher-severity patients, who require more immediate assistance from emergency room services. Imagine that a person was playing soccer and they rolled their ankle, the person is unsure if they had possibly broken a bone or torn a ligament. This person is indecisive about what their next options should be. Why not have a chatbot help make that decision for you? With the Project: Hepius system, users can first download the Hepius app on their IOS device, a virtual chatbot application. Users will be introduced to a live chatbot named Hepius, who will collect their user information and their self-described symptoms. Hepius will then place the user request into the emergency room queue and forward the information to tele-nurses. The tele-nurses manage the Hepius Dashboard, a web application that receives and processes this information, allowing for medical professionals to make a decision on if the user requests are low-severity or high-severity. If the user is considered low-severity, the tele-nurse will send the request to a nearby urgent care clinic. If the user is considered high-severity, the tele-nurse will send the request to the nearest emergency room.
How we built it
We first created a project roadmap on a spreadsheet and delegated roles to each teammate. We then used Figma to wireframe the dashboard UI/UX and eventually built the Hepius Dashboard through HTML and CSS. The Hepius IOS chatbot app was built through Swift, XCode, DialogFlow, and Kommunicate. The Hepius Dashboard and Hepius IOS chatbot are connected through Google Firebase.
Challenges we ran into
The challenges encountered by our team primarily focused on integration between consumer-facing UIs and backend services. Our team sought to pull user-symptom data from a mobile app (with the help of our chatbot Hepius), who would then structure that data into a usable triage intake form. Hepius would then stay with the patient and advise them until a tele-nurse (publicly available in Alberta) was available to electronically assess the patient. Unfortunately, because we could not connect Dialogflow to Firebase to send the symptom data, our nurse-side dashboard app could not fetch aggregated patient-queue data for consumption.
Accomplishments that we are proud of
We were able to successfully train the Hepius chatbot to receive user inputs and output responses accurately. We were also able to create a web application for the Hepius Dashboard with little experience with HTML and CSS.
What we learned
We learned that having a good plan doesn't always guarantee success. We ended up having a lot of unexpected technical bugs that ended up taking up all of the time in our project development to fix. The bugs would have taken more than 24 hours to fix, so we ended up not being able to properly deploy a successful demo.
What's next for Project: Hepius
With more time and preparation in the future, Project: Hepius has a lot of potential to improve in features. The main task to improve this system is to figure out how to properly connect Dialogflow to Firebase to get the Hepius chatbot to give real-time updates on user requests to the Hepius Dashboard. We also want to possibly add video calling capabilities between the user and the tele-nurse in order to more accurately diagnose the possible patient.