53 - Jump Start: Creating a Query Library for Code Re(use) at an Academic Medical Center

Background: Data Librarians from NYU Health Sciences Library, in collaboration with a member from the Department of Population Health, are developing a query library for researchers to reuse code to search patient and medical databases (e.g., EPIC) more efficiently. The inception of this project took place during several cross-departmental conversations and library-led activities where data analysts expressed a need for more practical avenues to reuse and share code, especially for SQL queries. The Query Library avoids researchers having to “reinvent the wheel” while also promoting code reuse, collaboration, and conversation among data analysts and data librarians.

Description: The data librarians have developed a list of sixteen metadata fields for code queries. The current metadata list includes keywords, authorship, programming language, and data source, to list a few. This initial list also includes references to other resources, including ‘groupers,’ a function within Epic to store and share code related to clinical groupings. Metadata fields were then refined though preliminary user testing with the Medical Center’s Data Analysis Working Group. During user testing, data analysts brought snippets of code, such as a SQL query or reusable Python code, and documented their own code using a template spreadsheet with the developed metadata fields. Feedback generated during and after the activity was used to adjust metadata fields and provide notes on technical suggestions to ease the process of researcher description once the user interface is built.

Conclusion: Next steps for this project require further investigation into viable platforms for the Query Library's infrastructure. The Query Library team will also need to work with institutional IT to designate the Query Library as one of its official projects, allowing the team to gain access to developers and resources to create a user interface. Additional next steps include ensuring the project is properly integrated into researcher and analyst workflows, creating interoperability with other institutional systems where possible, and outreach and training with relevant researchers and analysts to help spread awareness and use of the Query Library.