About the PEDSpace Knowledge Bank

This collection catalogs research conducted using PEDSnet data, including both active projects and those initiated or completed since 2018. These studies reflect the wide-ranging capabilities of the PEDSnet Learning Health System, which integrates data from multiple pediatric institutions to support scalable, real-world evidence generation. Each study entry is linked to its corresponding peer-reviewed publications, as well as any concept sets, phenotype algorithms, or analytic tools developed during its course—enabling transparency, reproducibility, and reuse across future research efforts.

Studies in the collection are systematically organized by funding project, data source, and thematic area, making it easy to explore based on research focus or collaborative network. Studies conducted as part of large grant-funded projects such as the Glomerular Disease Learning Network (GLEAN) or Pediatric KIDney Stone (PKIDS) Care Improvement Network, or Preserving Kidney Function in Children with Chronic Kidney Disease (PRESERVE). Data source domains include PCORnet® Study and PEDSnet Study designation.

By showcasing this breadth of research, the collection not only serves as a valuable resource for current investigators but also highlights the ongoing contributions of PEDSnet to advancing learning health systems research, where continuous improvement in care is informed by systematically gathered and analyzed data.

Explore the PEDSnet study catalog →

This collection contains variable definitions based on phenotypic algorithms that identify cohorts of patients exhibiting specific clinical characteristics, conditions, or patterns of interest, as derived from electronic health records (EHRs). These phenotypes are defined using logic statements (algorithms) with combinations of diagnosis codes, procedures, medications, laboratory results, and other structured EHR data elements (concept sets) to represent real-world clinical presentations.

The collection is designed to support reproducibility and consistency across studies by clearly outlining the logic used to define each phenotype. It complements the PEDSnet concept set collection by offering a high-level overview of the variables defined in PEDSnet research, enabling users to explore how clinical features are operationalized across studies.

Explore Variable Definitions →

Concept sets, also known as code sets or value sets, are curated collections of standardized clinical codes used to represent specific clinical ideas, conditions, or treatments. These sets play a critical role in clinical research, healthcare analytics, and electronic health record (EHR) systems by enabling consistent identification and grouping of medical information across different data sources. Concept sets are typically built using standardized terminologies such as SNOMED CT, ICD-10, HCPCS, and RxNorm, which ensure semantic consistency and interoperability.

This particular collection includes concept sets developed as part of PEDSnet research initiatives. These sets are often used in combination with algorithmic logic to define and identify patient cohorts for observational studies and data analyses.

The concept sets are systematically organized into eight clinical domains:

• Devices
• Diagnoses
• Procedures
• Visits
• Laboratory results
• Physiological measurements
• Environmental and socioeconomic variables
• Medications

Each set includes individual codes representing approved, standardized concepts drawn from reputable clinical ontologies. All concept sets are available for download in CSV format via PEDSpace, making them easy to integrate into data analysis workflows or cohort-building tools.

Explore the Concept Set collection →

This collection offers access to an array of study-specific and network-level data quality analysis modules, designed to support robust and reliable clinical observational research within a learning health system like PEDSnet. High-quality data is foundational to generating valid, actionable insights in such systems where research and practice are tightly integrated.

Researchers can browse the collection to identify categories of data quality analyses or filter to select checks that align with specific stages of the research lifecycle—from initial cohort design to final analysis.

The Data Quality Collection comprises a comprehensive set of modules targeting key areas:

• Cohort Fitness: Evaluating whether a dataset can adequately represent the intended population.
• Cohort Identification: Validating inclusion/exclusion logic using real-world data signals.
• Concept Set Testing: Ensuring the accuracy and appropriateness of clinical concept definitions.
• Data Conformance: Checking for adherence to data model standards and vocabularies.
• Dataset Fitness: Assessing completeness, consistency, and temporal coverage of data.
• Variable Testing: Verifying the presence, distribution, and utility of analytic variables.

Explore data quality modules →

This collection contains the results of data quality analyses conducted using PEDSnet's Data Quality Modules. This collection includes analyses for the entire PEDSnet Network, conducted on a quarterly basis, as well as analyses performed during individual research studies utilizing PEDSnet data. It provides detailed insights into the accuracy, consistency, completeness, and reliability of data collected across multiple studies and institutions.

This collection serves as a centralized repository to support the operational, technical, and collaborative needs of PEDSnet's participating sites and research collaborators. This collection encompasses a diverse array of resources, ranging from technical documentation and data dictionaries to administrative guidelines and instructional content.

The collection is systematically organized into several key domains:

• Data Architecture: Contains table one metrics for current and historical versions of the PEDSnet database, ETL specifications, and comprehensive data dictionaries for OMOP, PEDSnet, and PCORnet.
• Electronic Resources: Features a catalog of web applications, providing tools and interfaces to interact with the network's data and resources.
• Documentation & Tutorials: Offers detailed documentation, including user guides, and instructional materials.
• Institutions: Provide data provenance records for each participating site.
• Reusable Code: Frequently used functions which may be readily customized per study needs.

Explore PEDSnet resources →