MediGuide

Inspiration 🎯

Medication non-persistence is a critical issue affecting patient safety in the United States and beyond. According to studies, "between 40% and 75% of adults fail to take their prescribed medications on time each day" [1], leading to worsening health outcomes, increased hospitalizations, and preventable deaths. Non-adherence can lead to serious complications, including "delirium, hypoglycemia, and acute kidney injury" [2]. Existing digital reminder solutions often suffer from high onboarding friction, making them less accessible to those who need them most—particularly senior citizens and less privileged members of society who may have limited access to or knowledge of modern technology like smartphones or the internet. We were inspired to create MediGuide to address these gaps, providing an inclusive, user-friendly solution that enhances medication adherence and improves patient safety for everyone.

What It Does 💊

MediGuide is an AI-powered reminder and monitoring system designed to enhance patient safety by ensuring adherence to medication regimens. Here's how it works: 1️⃣ Seamless Onboarding with Healthcare Professionals: Patients are onboarded through a consultation session with a healthcare professional, either in-person or via telehealth. The doctor understands the patient's needs, writes a prescription, and enters it into the MediGuide platform, specifying the patient's preferred method of reminder. 2️⃣ Personalized Reminder Delivery: MediGuide's intelligent AI system identifies the medication and dosage, setting up personalized reminders for the patient via their chosen method—automated phone calls or text messages. Each reminder includes clear, simple instructions tailored to the patient's specific medication regimen. 3️⃣ Monitoring and Feedback: After each reminder, the system prompts the patient to confirm medication intake. It periodically checks in about their experiences, detecting early signs of adverse reactions and prompting timely interventions. 4️⃣ AI-Powered Adaptation: The AI continuously learns from patient behavior, using predictive analytics to adapt and optimize reminders for each individual, improving adherence over time. By integrating this data-driven approach, MediGuide aims to minimize medication-related risks, reduce hospital readmissions, and ultimately improve overall patient safety.

How We Built It 🛠️

We built MediGuide using Next.js and Flask for the backend, providing a robust and scalable infrastructure. The frontend was developed with React.js, TypeScript, and shadcn, ensuring a responsive and user-friendly interface. To understand doctors' prescriptions and generate smart, personalized reminders, we utilized top-performing Large Language Models (LLMs). For the phone call reminders, we leveraged advanced conversational voice AI technologies to deliver engaging and accessible interactions with patients who may not be familiar with modern digital devices.

Challenges We Ran Into 🚧

One of the main challenges we faced was integrating advanced AI technologies within a tight time frame. Developing an AI system that accurately understands and processes doctors' prescriptions required careful tuning of Large Language Models. Ensuring that our platform is accessible to patients with limited technological proficiency was another challenge, as we needed to design interfaces and interaction methods (like phone calls) that are user-friendly for all age groups and technological backgrounds. Additionally, we had to consider compliance with healthcare regulations and patient privacy laws, which added complexity to our development process.

Accomplishments That We're Proud Of 🎉

We are proud of being able to create MediGuide in under 36 hours for a noble cause. Developing a functional AI-powered system capable of understanding prescriptions, generating personalized reminders, and interacting with patients through conversational voice AI within such a short time frame was a significant achievement. We hope that this creation can eventually evolve into a product that helps more patients take their medications timely and recover faster, improving patient safety for millions across the world.

What We Learned 📚

Throughout this project, we learned about the complexities surrounding medication adherence and the significant impact it has on patient safety and healthcare costs. We gained valuable experience in integrating advanced AI technologies, such as Large Language Models and conversational voice AI, into a practical application. Additionally, we learned the importance of designing solutions that are accessible and user-friendly for all patients, regardless of their technological proficiency. This project reinforced the value of interdisciplinary collaboration between technology and healthcare to address real-world problems.

What's Next for MediGuide 🚀

We're excited to hear feedback from our peers and mentors and to speak with actual healthcare practitioners. Our next steps involve engaging with medical professionals to gather insights and validate our solution in real-world settings. We aim to explore avenues of collaboration with hospitals and clinics to integrate MediGuide into existing healthcare systems, ultimately improving patient safety on a larger scale. We're also considering further development to comply with healthcare regulations, enhance AI capabilities, and expand accessibility features to reach more patients. Citations

Built With

• flask
• nextjs
• openai
• prisma
• python
• react
• shadcn
• typescript
• vapi