Data

PCORnet represents data from everyday healthcare encounters
with more than 30 million people annually across the U.S.

Data is the backbone of PCORnet, and the scale, quality, and security of PCORnet-accessible data is a differentiator for the Network. PCORnet® Network Partners perform rigorous work upfront that enables users to ask the same question to millions of people across the United States simultaneously, with fast answers delivered in a single, standardized format.

Data accessible via the PCORnet distributed network draw from millions of electronic health records (EHRs) with growing links to patient-reported and payor data to create a powerful, standard data set that facilitates large-scale, multi-site research.

PCORnet® Network Partners have developed policies and other critical documentation to ensure the quality, facilitate the accessibility, and govern the use of the PCORnet® Common Data Model and resources.

PCORnet Overview Deck Slide 2 Graphic

>30 Million in the Network

PCORnet represents data from everyday healthcare encounters with more than 30 million people across the U.S. each year.

Click to enlarge or download

All PCORnet-accessible data are rigorously screened in a two-stage process. The first stage examines data for conformance, or adherence to the standard organization and representation of data for PCORnet; completeness, including that diagnosis codes are aligned; plausibility to ensure that the values make logical sense; and persistence to ensure that records are not disappearing between refreshes.

The second stage of curation establishes the level of data quality for a specific study purpose. The Coordinating Center for PCORnet® examines accessible data for patterns to identify potential quality concerns for key variables within a given study or population. As a result, the strengths and weaknesses of data available via PCORnet for specific research use are well known and communicated with transparency at the outset of every research effort.

A key security feature of the PCORnet infrastructure is that the data stay with each network partner behind its firewall protected under HIPAA and are not amassed into a single data pool or data warehouse. Queries and responses are enabled via a secure Distributed Research Network Query Portal. The system includes strong governance, role-based access controls, and auditing. The PCORnet-enabled approach shares only the minimum necessary information needed to answer a question. Researchers’ queries are sent to the data — and answers, not data, are sent back to researchers.

 

View the Statement on Data Privacy.

Continuous, Open-Source Learning Model

The process for readying data from sites participating in PCORnet for research is one of continuous improvement. As the PCORnet® Network Partners build knowledge around EHR use for research purposes, they are constantly refining and evolving. Improving the nation’s public health is a core motivator in the development of PCORnet, which is why PCORnet data curation learnings are shared on GitHub, an open source project accessible to all.

Image of data graphics

Common Data Model

Multi-site Research, Simplified

Because every PCORnet® Network Partner operates under a shared PCORnet® Common Data Model, users can ask the same question to millions of patients simultaneously and receive a clear answer, even across disparate systems.

The strength of the PCORnet infrastructure is that it enables researchers to quickly capture data insights from a large network of clinical information systems at once, which has historically been challenging. That’s because each of these systems code and define data slightly differently from one another. A clinic might code birthdate as “Date_of_Birth,” a health system might call it “Birth_DT,” and a health registry might use the code “DOB.”

The PCORnet® Common Data Model, developed by the PCORnet® Network Partners, solves this challenge by standardizing millions of data points from the Network’s diverse clinical information systems into a common format. As a result, users of PCORnet can ask the same question simultaneously to hundreds of disparate systems and receive a clear, reliable answer.

Microsoft Excel Icon

PCORnet® Common Data Model v6.1 Parseable Spreadsheet Format (03 April 2023)

Adobe PDF Icon

PCORnet® Common Data Model v6.1 Specification (03 Apr 2023)

Adobe PDF Icon

PCORnet Data Checks v15 (Effective January 2024)

Microsoft Excel Icon

PCORnet® Common Data Model Value Set Reference File v1.13 (03 April 2023)