UHL CSAR-AI Study

This portal provides access to the UHL CSAR-AI Study Human Healthcare Worker Survey, which runs on a custom instance of the open-source application LimeSurvey .

The aim of this study is to assess the performance of a GDPR-compliant AI model, which we have trained to read and interpret handwritten clinical notes. We fine-tuned this model by exposing it to thousands of anonymised excerpts from clinical notes written by healthcare workers at UHL.

We would like to compare the accuracy of this model against that of human healthcare workers, in order to evaluate whether this type of AI-assisted interpretation may have a role in improving the readability, interpretation, and safe handling of handwritten clinical documentation.

Participant information

The participant information leaflet (PIL) explains the purpose of the study, what participation involves, the specific data that will be collected and how this data will be used.

Privacy and data

All respondent data will remain inside the secured LimeSurvey hosting environment. Only data within the scope of the current study will be collected. Any export, backup, retention, or deletion of participant data will be conducted in a manner compliant with European GDPR legislation and local data retention policies .

This landing page is hosted via a publicically-accessible GitHub repository ; which is used to serve the website frontend only. This repository will not be used to host respondent data, participant email lists, exported survey responses, database credentials, or private configuration files.

Support

If you have questions about the study, or experience issues with the LimeSurvey platform, please feel free to contact the research team using the details provided in the participant information document.

Survey

When you are ready, use the button below to open the LimeSurvey questionnaire.