Inspiration

We were inspired by how difficult a prescription issue can become when an HCP needs to quickly find an appropriate alternative for a specific patient. A generic medication may be a valid alternative in general, but factors like allergies, existing medications, and patient history can make certain options inappropriate. We wanted to build a tool that turns this process into a faster, more personalized workflow for HCPs.

What it does

RxRescue helps HCPs find patient-specific generic alternatives when a prescription encounters an issue. It analyzes the patient's medical information, checks potential medications against relevant warnings and contraindications, and ranks eligible alternatives based on patient-specific factors. HCPs can review why each option was recommended, select an alternative, and send an updated prescription back through the workflow.

How we built it

We built the frontend with React and Vite and the backend with FastAPI. We created a patient portal, prescription workflow, medication warning database, and recommendation engine that connects patient information with potential generic alternatives. The backend processes the patient and medication data, filters out unsafe options, and returns ranked recommendations to the HCP interface.

Challenges we ran into

One of our biggest challenges was determining how to personalize recommendations without making the system overly complicated. We had to distinguish between factors that should exclude a medication entirely, such as a known allergy, and factors that should influence the ranking of otherwise eligible options. We also had to design the system so that recommendations were transparent and the HCP remained in control of the final decision.

Accomplishments that we're proud of

We are proud of building an end-to-end workflow that connects the patient, pharmacy, HCP, and prescription update process. Instead of simply displaying a list of generic alternatives, RxRescue explains why different options may be more or less appropriate for a particular patient.

What we learned

We learned how important context is when building healthcare applications. A recommendation that looks reasonable in isolation can change significantly when patient-specific information is considered. We also learned how to design a system where AI and automated decision support can assist an HCP without replacing their clinical judgment.

What's next for RxRescue

We want to expand RxRescue with more comprehensive medication data, additional patient-specific factors, real-time pharmacy and prescription integrations, and stronger explanations for each recommendation. We also want to explore how RxRescue could integrate directly into existing HCP workflows so that finding an appropriate alternative requires as little friction as possible.