Syntegra Synthetic EHR Data | Structured Healthcare Electronic Health Record Data

Organizations can license synthetic, structured data generated by Syntegra from electronic health record systems of community hospitals across the United States, reaching beyond just claims and Rx data. The synthetic data provides a detailed picture of the patient's journey throughout their hospital stay, including patient demographic information and payer type, as well as rich data not found in any other sources. Examples of this data include: drugs given (timing and dosing), patient location (e.g., ICU, floor, ER), lab results (timing by day and hour), physician roles (e.g., surgeon, attending), medications given, and vital signs. The participating community hospitals with bed sizes ranging from 25 to 532 provide unique visibility and assessment of variation in care outside of large academic medical centers and healthcare networks. Our synthetic data engine is trained on a broadly representative dataset made up of deep clinical information of approximately 6 million unique patient records and 18 million encounters over 5 years of history. Notably, synthetic data generation allows for the creation of any number of records needed to power your project. EHR data is available in the following formats: — Cleaned, analytics-ready (a layer of clean and normalized concepts in Tuva Health’s standard relational data model format — FHIR USCDI (labs, medications, vitals, encounters, patients, etc.) The synthetic data maintains full statistical accuracy, yet does not contain any actual patients, thus removing any patient privacy liability risk. Privacy is preserved in a way that goes beyond HIPAA or GDPR compliance. Our industry-leading metrics prove that both privacy and fidelity are fully maintained. — Generate the data needed for product development, testing, demo, or other needs — Access data at a scalable price point — Build your desired population, both in size and demographics — Scale up and down to fit specific needs, increasing efficiency and affordability Syntegra's synthetic data engine also has the ability to augment the original data: — Expand population sizes, rare cohorts, or outcomes of interest — Address algorithmic fairness by correcting bias or introducing intentional bias — Conditionally generate data to inform scenario planning — Impute missing value to minimize gaps in the data

搜集汇总

背景与挑战

背景概述

该数据集是由Syntegra生成的合成电子健康记录数据，基于美国社区医院约600万患者记录，涵盖药物、实验室结果、生命体征等临床细节，确保隐私且统计准确。它支持按需生成任意数量的记录，可用于产品开发、测试等场景，并能扩展人口规模或纠正数据偏差。