Data associated with the publication: Taxonomy-based prompt engineering to generate synthetic drug-related patient portal messages

This dataset is a part of "Taxonomy-Based Prompt Engineering to Generate Synthetic Drug-Related Patient Portal Messages study". The dataset contains synthetic patient portal messages on drug related problems. The messages were generated using prompt engineering based on the taxonomy of patient-generated messages from the patient portal. This dataset aims to provide researchers with access to drug-related patient portal synthetic messages. The dataset contains 450 generated messages, which can be categorized into 2 groups. The first group was generated from all prompt options, derived from the taxonomy, with no specific race of the patient (200 messages). The second group was generated using subset of prompt options with specific race information. This group contains 50 messages for each race, including White, American Indian or Alaska Native, Black or African American, Asian, and Native Hawaiian or Other Pacific Islander (total 250 messages). All the messages are accompanied by prompt used for generating the message, taxonomy group, medication name, race, and urgency level.