St. Paul's Co-educational College (POON Clarisse)
Items | Content |
---|---|
Year | 2024 |
Awards |
|
Name of Entry | Medisafe |
Description |
Medisafe is an AI-powered web application that cross-checks prescriptions with patient data to detect potential drug prescription errors and enhance medication safety.
|
Description
Blunders of drug prescription by doctors are common in Hong Kong, with rising incidence due to over-worked doctors, especially in public hospitals. In fact, it’s a world-wide problem causing harm or even death to patients and a high cost. Doctors face medicolegal or even criminal consequences because of medication errors. Common errors include prescription of drug with known allergy, undesirable drug-drug interactions, incorrect dosage without consideration of patient’s renal or liver function and special clinical conditions. Current system in hospital depends on a pharmacist to cross-check doctors’ prescriptions, but this is prone to human errors. Elimination of such medication errors are important to avoid hazards to patients. With the advent of AI, it is conceivable that a software capable of automatic checking of drug prescription using AI can be a solution to the problem of medication errors. Our invention is a software using artificial intelligence to automatically cross-check doctors’ prescription. Alert will be triggered if there is potential error in the prescription after matching patient’s drug history and clinical conditions. To our knowledge, this is the first software aimed to safeguard patients by automatic checking of drug prescriptions. The project aims to develop a web-based software using artificial intelligence to automatically screen doctors’ prescription for possible errors based on patient’s clinical condition, and to provide recommendations to doctors on correct prescription.
Comments from Judging Panel
This project originated from the observation and concern of social problems, starting from an excellent idea, based on detailed data collection, full understanding of LLM, SQL, Vector, and in-depth medical knowledge, and making good use of the drug database, we have successfully constructed a very excellent and easy-to-use system. It has been proved by doctors and a large number of clinical patients that the system is effective in assisting doctors in prescribing medication, reducing unnecessary medical financial burden and playing an important role in ensuring medical safety. This project is not only concerned about the user experience, but also the privacy of the patients. It is at a commercial level of completion and has a promising future.