While we value the efforts made in terms of visualizing the spread of COVID19 and other data related projects, we wanted to solve a simpler, more practical problem.
In the case of an overshoot of hospital capacity, field hospitals will have to be set up. Think of those as beds lined up in conference centers for first-world countries (best scenario), and as tents on the floor for third-world countries (worst scenario.
We talked to many experts that insured us that even in first-world countries, managing 20 beds on field after a fire is a huge pain, and that no health information systems that can be quickly setup and is easy to use exists yet. Worse, they also mentioned that it is mostly impossible to scale to more than roughly 100 beds.
Healthdesk wants to solve that problem by providing a simple interface and workflow.
The prototype is limited to COVID19 symptoms but could easily be extended to other diseases or situations.
What it does
- Patients can pre-register at home with their name and symptoms (easing up the admission process and the planing of resources)
- Nurses can admit and triage patients using a simple interface, including symptom assessment, healthcheck, personal data, & more (easing up the triage and further patient tracking)
- Bar code generation for patient tracking (easing up assignment to a bed and healtcheck)
- Bed management, including bed types, bed state, & more (easing up the bed rotation and patient assignment)
- Healthcheck creation: daily blood pressure, GCS etc. can be submitted in the system and are always assigned to a patient using their bar code (easing up the overview of health evolution)
- Patient overview: see how many patients currently are assigned a given severity, what is the discharge rate etc. (easing up the allocation and planning of resources)
How we built it
The backend is a REST API built on Python/Django, thus allowing further integrations with other health systems, mobile applications & more.
The deployment has been done on Heroku and Netlify.
Challenges we ran into
- Clearly structuring the patient flow is difficult
- The flow varies depending on the situation/region
- Tying healthcare specific concepts together without having healthcare knowledge
Accomplishments that we're proud of
- The usability of the forms
- Having built so many features (form, triage, bed management, healthcheck) with so little developers
- Solving a real problem! (we got good feedback from docs)
What we learned
It would have been easier if we had had a chance to talk to a doctor in the earlier stages of the Hackathon.
What's next for Healthdesk
- Improve it during the next hackathon
- Get in touch with more medical staff in order to get feedback
- Clean up the data structure and flow
- Build up the frontend interface to include all modules
- Work on some machine learning algorithms that could ease up processes based on the submitted data
Some functionality is only available in the Django Admin (bed management and healtcheck) and is not exposed for public usage in the demo link below since it could break stuff (it's a prototype). If you know how to run code, you can run the prototype locally and try out all the features.
A demo of two features is available here:
As stated above, more features are available in the Django admin but can't be exposed publicly yet.
Make sure to check out the yellow help button in the lower right corner if you get lost!
The Healthdesk Team