A Universe of Data that Drives Evidence-Based Research and Individualized Patient Care
The Cosmos data set is unlike any other used in health research today. Cosmos combines billions of clinical data points in a way that forms a high quality, representative, and integrated data set that can be used to change the health and lives of people everywhere.
Cosmos has records from over 190 million patient records from over 7.5 billion encounters, representing patients in all 50 states.
Cosmos is built on a data platform optimized to answer clinical questions over time in minutes. Clinicians and other scholarly explorers at organizations using Cosmos can ask questions on-demand without needing to put a request in for data.
Because Cosmos is built by health systems from across the country, it aligns closely to population metrics of the United States Census. Cosmos has a representative sample of patients across all races, sexes, ages, rural-urban locations, and types of insurance.
Epic integrates both inpatient and outpatient charts into a single record, including all clinical specialties to form a more comprehensive and detailed picture of a patient’s health. For example, a cancer patient’s record contains detailed oncology visits, cancer staging, advanced lab results, hospitalizations, and standard outpatient visits.
Cosmos goes well beyond diagnoses and medications found on claims transactions, and includes patient-generated health data (PGHD), birth records, vitals, and social determinants like transportation and financial security assessments.
The Epic Community has already seen great success using the universe of Cosmos data for research and are continuing to expand the impact of this unique data set.
Cosmos is the data source for studies in academic journals. These include:
Vaccine Effectiveness of JYNNEOS against Mpox Disease in the United States published in The New England Journal of Medicine
A study on neonatal hospitalizations and readmissions, published in Pediatrics
Findings on antibiotic prescribing and testing practices in NEJM Catalyst
Insights on COVID reinfections as seen in CDC's Morbidity and Mortality Weekly Report
Diabetic patient outcomes during the COVID pandemic in Diabetes Care
A study investigating the link between COVID-19 and type 1 diabetes in Diabetes
Racial and ethnic differences in early diagnosis of autism spectrum disorder in JAMA Network Open
Racial and ethnic disparities in the prescription of Nirmatrelvir (Paxlovid) for COVID-19 in Journal of General Internal Medicine
While traditional research continues to serve as a primary source of information to inform clinical practice, we identified a need for a new kind of analysis focused on getting good data out quickly to clinicians and decision-makers where it can make a difference. To address this need, we launched Epic Research. Using our dual-team process, Epic Research teams evaluate timely, topical, and relevant questions, and make findings available to the public at no-cost at EpicResearch.org and to Epic Research subscribers.
Several Epic Research briefs have gained media and policymaker attention, such as:
Two in three Mpox cases prevented in patients fully vaccinated with JYNNEOS. Even one dose of JYNNEOS provided some protection against Mpox infection.
Despite significant drops in 2020 and low rates in 2021 through the beginning of 2022, strep throat infections exceeded their 2017 peak by 30% in February 2023.
Since USPSTF updated its guidelines in 2021, there has been an increase in colorectal cancer screening rates, most significantly in 45–49-year-olds. There has also been an increase in polyp and colorectal cancer diagnoses in 45–49-year-olds.
After an initial myocardial infarction, patients prescribed a beta blocker did not show improvement in the rate of subsequent myocardial infarction within the next year.
The vision for Cosmos has always included a way for providers to make an immediate impact with their patients. With our ongoing development of Best Care for My Patient and Look-Alikes, providers will be able to use attributes from the patient in front of them to review real-time insights they can use alongside their own expertise and experience to partner with their patient on shared decision-making.
Participating organizations agree to a legal structure that establishes shared trust.
Patient identifiers are removed before data moves from a healthcare organization.
From the start, no unique patient identifiers (names, phone numbers, etc) are sent or stored in Cosmos. Additionally, Cosmos is architected to allow each organization the ability to design their own local patient opt out policies that meet their communities' specific rules and regulations.
208 Epic organizations and counting are contributing to the Cosmos data set