Lowering Barriers, Fostering Discovery: How NLM is Supporting A Year of Open Science

Guest post by colleagues at NLM’s Lister Hill National Center for Biomedical Communications (LHNCBC).

Did you know that 2023 is a Year of Open Science? To honor this, NIH bolstered its commitment to open science, and NLM is excited to announce two new projects in the LHNCBC Applied Clinical Informatics Branch (ACIB) that will enable access to scientific data: the Center for Clinical Observational Investigations (CCOI) and Leveraging Large-Scale Clinical Datasets for HIV outcomes projects.

The Center for Clinical Observational Investigations will remove barriers to accessing data for researchers. NLM will curate a list of available clinical datasets and catalog these datasets with the relevant metadata. This metadata will be accessible to researchers from universities, industry and around the world. NLM will provide informatics, data science, and statistical analysis support to query the datasets for research questions posed by NIH researchers.

The CCOI will lower the barrier for adoption of research based on clinical observational data (data from electronic health record systems and insurance claims), making it possible to efficiently test hypotheses that arise from basic or pre-clinical research on real clinical data. The CCOI will also foster collaboration between and provide training to NIH researchers with various degrees of clinical data science and clinical or biological expertise.

The Leveraging Large-Scale Clinical Datasets for HIV outcomes project, funded by the NIH Office of AIDS Research, will use reproducible HIV cohort definitions found in datasets from HIV clinical records and claims data to automate dataset characterization, and then will use data science, artificial intelligence, and machine-learning methods to improve identification and classification of patients with HIV. Instead of a hundred or fewer participants for normal-scale datasets, large-scale clinical datasets include thousands to hundreds of millions of participants across multiple demographics; this larger and more diverse set makes it possible for HIV researchers to detect subtle patterns and relationships and make international comparisons.

Through both initiatives, the Center for Clinical Observational Investigations and Leveraging Large-Scale Clinical Datasets for HIV outcomes project, NLM will leverage records from multiple international datasets, perform data science, artificial intelligence, and machine-learning methods to analyze the data, and generate findings based on clinical data that aid in further research in their respective fields. Data from these projects have been mapped to the Observational Medical Outcomes Partnership (OMOP) Common Data Model. Early progress in the OMOP model shows promise for the transformation of Medicaid data to a common data model that could support the generation of evidence, with special attention to low-income patients on public insurance. There is further evidence to demonstrate the value of using the OMOP model to support harmonization of data collection across human clinical studies through an analysis of the All of Us Research Program dataset.

NLM is excited by these new projects and the opportunities they offer to advance open science by lowering barriers to access and supporting discovery from clinical datasets through analytics to improve public health.

Featured image attachment page