PhD Scientific Days 2024

Poster Session S - Health Sciences 2.

Developing an application for calculating the MERIS score from real-world data

Text of the abstract

Introduction:
The calculation of the Danish medication risk score is a very effective and practical tool for identifying patients requiring intervention by a clinical pharmacist. The calculation requires the eGFR value, which indicates the patient's kidney function, and a list of drugs used simultaneously.
Aims:
The aim of the project was to create a web application that is easy to use and accessible to a wide range of users, through which a calculator based on the MERIS algorithm can be tested on live data.
Method:
We used the Python programming language to develop the MERIS calculator, which was integrated into an application created in the Streamlit framework. The user is asked for an eGFR value and a list of medicines in free text form. For the processing of the text, we have developed a hybrid technology (using local strict matching drug name recognition and GPT technology), which allows us to effectively convert the unstructured information into data that can be processed by the calculator. The app was launched on a public Streamlit server, and then we asked test users to try it out. Based on the feedback from the testers, we have made corrections to the program and evaluated the results of the calculations performed by the tests.
Results of
The testers performed 79 calculations using live data from their own or their practice. The average response time was 2.66 seconds. The average MERIS value is 5.89 points. On average, three medicines were used simultaneously.
Conclusion
In the experiment, we successfully created a hybrid text processing model that combines the speed and accuracy of the rule-based algorithm with the flexibility of the GPT-based technology and reliably processes the data entered by the user in free text form to perform the MERIS calculation. Furthermore, we have made the calculator widely available to users on the Internet.
Funding
Project No. 2023-2.1.2-KDP-2023-00016 has been implemented with the support provided by the Ministry of Culture and Innovation of Hungary from the National Research, Development, and Innovation Fund.