Inspiration

The idea for this project was inspired by the recurring challenges that we observed in clinical practice:

  1. Patients often struggle to keep track of their medications especially if the list is long and complicated.
  2. Patients often miss out on sharing about their use of supplements, traditional medications, and over-the-counter medications.
  3. NEHR may not have the complete list of medications, especially if they obtain them from private clinics, hospitals or overseas. The use of supplements, traditional medications, and over-the-counter medications may not be recorded too.
  4. Patient may not be compliant to the medications prescribed or dispensed.
  5. Healthcare providers face difficulties during medication reconciliation, especially at the point of hospital admission, as patients may not recall their medications especially when they ain't feeling well.
  6. Inaccuracy in medication reconciliation can pose a danger for patients as it affects continuity of care and can cause drug related problems.

We realised that empowering patients with a simple, AI-driven solution could address these issues. This drove us to design a chatbot-powered app that would help patients easily build and maintain a personal medication list, either by:

  1. Retrieving medication record from HealthHub / NEHR and confirming their compliance to the prescribed instructions.
  2. Taking photos of their prescriptions, pill packs, or medications.
  3. Manually keying in their medications with guided input.

Ultimately, we wanted to bridge the gap between patients and healthcare providers, ensuring medication lists were accurate, up-to-date, and shareable.

What We Learned

Throughout this project, we deepened our understanding of patient-centered design, the importance of usability, accessibility, and privacy in healthcare apps.

How We Built It

The development flow consisted of several key components:

  1. Front-End (Mobile App): 2 Photo mode → medication recognition pipeline.
  2. Manual entry mode → chatbot-assisted text input.
  3. Validation & Sharing:
  4. Patients can digitally share a structured medication list with their providers.
  5. Providers then reconcile this list with hospital orders.

Challenges that We Faced

  1. Coming up with application features that would make it more user friendly, especially for those who are less IT savvy users
  2. How to ensure accuracy of the AI, which will affect the trust and usability of our application. We believe that building an intuitive user interface and ensuring clear AI workflow logic is essential to the application's accuracy. It is also our intention for this application to support, but not replace, the importance of healthcare providers in verifying the information, as medication reconciliation is a complex process.

Through overcoming these, we ensured that our app is not only functional but clinically reliable.

Conclusion This project reinforced our belief that technology can empower patients, giving them more control over their own health and improving safety for everyone. By using AI to simplify medication tracking, while still leaving final verification to providers, we struck a balance between innovation, usability, and safety.

We hope this solution contributes to reducing medication errors, and improving transitions of care.

Built With

Share this project:

Updates