During my clinical rotation in a hospital, a covid patient was being prescribed tocilizumab. But there is a rare and dangerous interaction between the medication and the patient's disease multiple sclerosis. What if when the doctor prescribes a drug in the software, the software warns the doctor about the interaction? Currently, their software only checks drug-to-drug interactions. It doesn't consider the patient's health conditions.
What it does
This app checks the user's dugs and health condition interactions.
How we built it
We built it with front-end react.js. The backend was with flask API. We web scraped healthcare websites to generate a list of medication names, diseases, and dynamically generated drug to disease interactions based on the users' input.
Challenges we ran into
The data we retrieved was inconsistent because drugs have many brand names. Furthermore, diseases have many similar names and medical names. This complicated our data structures and logic
Accomplishments that we're proud of
We were proud to be able to dynamically generate disease and drug interactions through web scraping different healthcare websites
What we learned
It was tough building this app due to inconsistencies in medical terminologies. Sometimes acronyms were used or brand names were used. Other websites used combo generic names and old medical terminology.
What's next for Drugs and You
We need to improve our algorithm and data structure. An improved data structure would be a list of objects with attributes of the brand and a generic name. This would allow an association between the brand and generic drugs. The same could be applied to diseases. We would have 1 disease name and relate to its other similar names.