Managed by the National Library of Medicine (NLM), PubMed is widely recognized as one of the leading databases in the biomedical and life sciences fields, housing an impressive collection of over 27 million citations.1 This vibrant repository is constantly growing, with thousands of new citations added every day, reflecting the remarkable progress in research.

Traditionally, PubMed operated on a dual data management system made up of the PubMed system and the Data Creation and Maintenance System (DCMS).2 The PubMed system, overseen by the National Center for Biotechnology Information (NCBI), focused primarily on the storage, indexing, and retrieval of citations specifically for PubMed searches. On the other hand, DCMS was responsible for reviewing data and indexing select citations for MEDLINE, which serves as the definitive bibliographic database for life sciences and biomedical information. However, the distinct purposes of these systems meant that many citations occasionally struggled to fit into either system smoothly.

The data management process started with the submission of XML files from data providers, which would then go through an indexing phase before updating citations were exported for inclusion in MEDLINE. Unfortunately, this method led to delays and discrepancies, as synchronization between the two systems was often inconsistent. As a result, challenges such as data asynchronicity, loss or corruption of information, labor-intensive correction efforts, and inconsistencies within public datasets became more prevalent.

To address these systemic challenges, the PubMed team at NCBI collaborated with the Indexing Section at NLM Library Operations, embarking on an ambitious mission to revamp the data management approach entirely. Their goal was to create a much more streamlined and integrated workflow for adding and modifying citations, ultimately leading to the development and rollout of the ambitious PubMed Data Management (PMDM) system.3

The PMDM system was thoughtfully designed to automate and simplify data handling processes, significantly reducing the reliance on labor-intensive manual corrections throughout citation management. Before the system launched, data providers faced limitations regarding the submission of various publication types, as well as essential information like grant details, clinical trial identifiers, and interconnections between related citations. The implementation of PMDM has greatly broadened these capabilities, allowing data providers to submit a diverse range of information with ease.

One of the most exciting features of the PMDM is its automatic integration of all submitted information into PubMed’s live dataset, effectively removing the need for time-consuming manual reviews by NLM data reviewers. This automation not only enhances efficiency but also encourages data providers to adhere to the established editorial standards set forth by the International Committee of Medical Journal Editors.

Early results following the launch of the PMDM system are incredibly promising. By mid-March 2017, the system facilitated updates to over 12,100 citations by 175 external data providers, all achieved within just six months of becoming operational. This enthusiastic participation highlights the system’s effectiveness in overcoming earlier limitations and improving the citation process.

As the Chairman of the Information Management Committee of the Korean Association of Medical Journal Editors—one of PubMed’s prominent data contributors—I genuinely believe that these advancements will significantly enhance the timeliness, accuracy, and integrity of citation data available to the global research community. Ultimately, this will foster the quality of biomedical literature accessible to users worldwide, opening doors to richer and more informed research experiences.

https://doi.org/10.5051/jpis.2017.47.2.64