Dear Colleagues,

We are delighted to announce the launch of the ENACT research platform! Beginning on July 17, ENACT is available for research using aggregated de-identified counts on over 142 million lives across more than 50 CTSA sites. This is the largest EHR dataset available for research using your desktop computer. As an investigator-centered tool, the user interface is intuitive and queries can be built without programming knowledge.

The data use agreement, IRB guidance documents and governance documents have been updated to enable research activities and are available on the ENACT website here. Strict quality assurance procedures and metrics are also in place to improve the data quality for research.

Aggregated de-identified counts can still be used for cohort discovery and assist with study design and feasibility. The types of research are limited only by the imagination of the investigator and can include population science, in silico clinical trials, outcomes research, epidemiology and more. Answers to critical questions can often be obtained in minutes. Three recent use cases include:

1.    COVID is associated with myocarditis in patients less than 20 yrs old. A simple query looking at the prevalence of individuals with COVID and the ICD10 diagnosis of viral myocarditis using ENACT revealed an incidence of 0.038%. The reported incidence from multiple studies was 0.045% (J Am Coll Cardiol. 2022;79:1717-1756).

2.    The SGLT-2 inhibitor empaglifozin is not associated with increased risk of lower extremity amputation. Empaglifozin was compared with canaglifozin and was not associated with increased risk (odd ratios 0.93 and 1.51, respectively, for amputation compared to other SGLT-2 inhibitors). These conclusions are similar to a large meta-analysis (PLoS One 2020 Jun 5;15(6):e0234065).

3.    Hydroxychloroquine does not prevent COVID. No difference was noted in the risk of patients with rheumatoid arthritis who were being treated or not treated with concomitant hydroxychloroquine (14-15% rates for each). This compares to a controlled trial in non-rheumatoid arthritis individuals with 12 and 14% rates, respectively (N Engl J Med 2020 Aug 6;383(6):517-525).

Each of these queries took under 5 minutes to generate results or preliminary data. Of course, more complex queries can also be built to answer critical health questions. As limited datasets are added, the number and types of research questions will continue to grow.

Thank you to everyone at our participating CTSA sites and NCATS for remaining engaged and active in our planning processes. Additional information on ontology and software updates will be provided as needed, and we encourage each site to stay current. We look forward to working with you and your investigators!

Warmly,

Steve Reis (sreis@pitt.edu)

Gary Firestein (gfirestein@health.ucsd.edu)

Bob Toto (robert.toto@utsouthwestern.edu)

Shyam Visweswaran (shv3@pitt.edu)