The Bureau of Vital Statistics at the New York City Department of Health conducts extensive data sharing with academic and policy partners. The bureau wanted a system for monitoring policy actions resulting from data sharing.

This work entailed building an NLP pipeline that drew data from the pubmed API and integrated it with the bureau’s client management system. This closed the loop on data sharing projects that could take years to complete, by identifying data sharing partners when they published research of the data.

Integration of the research clearinghouse data with operational data is a strategy that has been used by the pharmaceutical industry and others for some time to make the most of relevant insights from research. It was a novel approach in the context of tracking policy actions that resulted from analysis of birth and death data.