Synthetic Data Pack: 1000 patients with 5 years of longitudinal data

This data pack is a synthetic healthcare dataset comprised of 1000 patients with 5 years of longitudinal history.

Using a Monte Carlo simulation technique, each synthetic record is modeled to emulate clinically relevant treatment scenarios. During the generation synthetic patients progress through a series of healthcare encounters. It is these encounters and their events that are used to generate the dataset which is comprised of healthcare data messages across several HL7 messaging standards.

These records are highly realistic and even include gaps of information like a patient record in a real-world healthcare ecosystem.

Common conditions that may be contained in the data pack:

• Appendicitis

• Cancer

• Covid

• Deep venous thrombosis

• Diabetes

• Food Insecurity (SDOH)

• Hypertension

• Osteoporosis

• Pregnancy

• Pulmonary embolism

• STIs

• Zika

HL7 Message standards output that may be included for a synthetic patient record:

ADT

Admission, Discharge, Transfer (ADT) messages are used to communicate patient demographics, visit information and patient state at a healthcare facility.

This synthetic data set contains the following number of synthetic Admit, Discharge and Transfer (ADT) messages in HL7 messaging standard version 2.6 with the following event types:

• A01 - Admit / visit notification

• A03 - Discharge/end visit

• A04 - Register a patient

Message count in data set:

2,865 total A01

11 total A02

2,865 total A03

36,297 total A04

VXU

Unsolicited Vaccination Update (VXU) messages are used to receive and send patient’s vaccination information.

This synthetic data set contains synthetic Unsolicited Vaccination Record (VXU) messages in HL7 messaging standard version 2.5.1 with an event type of V04.

Message count in data set: 13,877

ORU

ORUs are unsolicited transmission of an observation message designed contain information about a patient's clinical observations and are used for transmitting patient’s laboratory results to other systems.

This synthetic data set contains synthetic Observation Result (ORU) messages in HL7 messaging standard version 2.5.1 with an event type of R01.

Message count in data set: 618

CCD

Continuity of Care Documents (CCD) are XML based markup standard built using HL7 Clinical Document Architecture (CDA) elements. CCD’s carry summary information about the patient within the broader context of the personal health record.

Current data fields in CCD’s:

• Patient demographics

• Medications

• Allergies

• Encounters

• Problem lists

• Diagnosis

• Lab results

• Immunization

• Social History

Message count in data set: 39,162

FHIR

Fast Healthcare Interoperability Resources (FHIR) is a modern standard for exchanging healthcare information electronically. FHIR leverages web standards like HTTP, RESTful APIs, and JSON to enable seamless communication between different healthcare systems, applications, and devices.

FHIR facilitates interoperability by providing a framework for representing and exchanging clinical data in a structured, standardized format, allowing healthcare stakeholders to easily access and share patient information across disparate systems, leading to improved care coordination, streamlined workflows, and enhanced patient outcomes.

The synthetic patient records generated by our statistic population health model generator are output in JSON FHIR version R4 resources.

Message count in data set: 14,238 bundles containing an average of 100 FHIR resources in each bundle (~1,423,000 total FHIR resources)