WEBINAR
Access Solutions to Amplify Your Clinical Real-World Data Research Webinar Recording
Enrich your data and expand your research opportunities as part of a global community advancing safer and more effective therapies for better patient care.
Join us as we examine how rich and timely clinical data, unified in a research-ready environment, can provide clinical researchers with the best opportunity to accelerate published research and advance outcomes.
Presenters:
Dr. Uazman Alam
Honorary Consultant Physician, University of Liverpool, United Kingdom
Professor Gregory Lip MD FRCP (London Edinburgh Glasgow) DFM FACC FESC FEHRA
Price-Evans Chair of Cardiovascular Medicine, University of Liverpool, United Kingdom; Distinguished Professor, Aalborg University, Denmark
Complete the form below to access the webinar recording.
Sample of Webinar Attendee Questions & Answers
We are just starting to learn the basics, so any tips or tricks would be great.
Every month, we offer five virtual training sessions to help you gain mastery of all that the TriNetX LIVE™ platform has to offer. Registration links and recorded sessions of TriNetX training materials can be found by logging into the TriNetX LIVE™ platform Help Center here: https://support.trinetx.com/hc/en-us/articles/1260801205269-Upcoming-Live-Training-Webinars
How much control does the TriNetX LIVE™ platform give you over the parameters for the propensity matching? Can you specify calipers, mandatory matches, etc.?
You can select any parameters that exist within TriNetX standardized terminology, including demographics, for matching. Caliper is set to 0.1 and cannot currently be edited on the TriNetX LIVE™ platform.
Can you speak in a bit more detail on how propensity score matching is accomplished when retrieving and analyzing data from the TriNetX LIVE™ platform interface?
Details can be found by logging into the TriNetX LIVE™ platform Help Center here: https://support.trinetx.com/hc/en-us/articles/360011978033-In-compare-outcomes-how-are-patients-matched-when-balancing-cohorts
Does the TriNetX LIVE™ platform perform all statistical analyses regardless, or, for example, would it not perform Cox regression if the hazards were not proportional?
Yes, the Advanced Analytics will run if the hazards are not proportional; however, the system will test for proportionality on a hazard ratio and present the Χ², degrees of freedom, and p-value.
We currently do not perform a cox model analytic on the TriNetX LIVE™ platform, this will be available later this year.
Details on all our statistical methods can be found by logging into the TriNetX LIVE™ platform Help Center here: https://support.trinetx.com/hc/en-us/sections/115001119073-Outcomes-Analytics-Compare-Outcomes-and-Analyze-Outcomes
Do you have any recommendations on validating the sensitivity and specificity of the phenotypes you are constructing in TriNetX?
We recommend leveraging an ICD-10 coding resource like icd10data.com to confirm the terms that are possible synonyms or exclusions. We would also recommend performing a literature search to find publications that provide the specific ICD-10 codes for defining a concept. Users can always reach out to Query Design Assistance by clicking here for querying guidance.
Please can you give a brief live demo of the TriNetX LIVE™ platform?
Please click here to Request a Demo or reach out to your TriNetX Account Manager or Healthcare Partnership Manager.
What is the current policy on using patient level healthcare data outside the source country in EU? Did you control the difference among the countries when conduct propensity score matching? How long is the lag time for collecting the death data, especially who died at home or outside the hospital in general? For the medication taken at home, do you have information on the prescription fill?
Patient-level data cannot be downloaded from the TriNetX application; however, ex-USA data can be analyzed on either the Advanced Analytics feature of the TriNetX LIVE™ platform or in TriNetX’s LUCID notebook environment. Users cannot match on country in the propensity score matching built into the Advanced Analytics.
Mortality data is refreshed at the rate that individual healthcare organizations are refreshing their data, with the average healthcare organization refreshes their data monthly.
Medications in the TriNetX data come from patient history, prescriptions, orders, and electronic medication administration systems (eMAR). Any medication documented at the healthcare organization in one of these ways will be queryable in the data.
Could you tell me how to set up the exposure duration of intervention in the TriNetX LIVE™ platform? For example, amiodarone 6 months at least. How can we set up it in the database?
Users can leverage Related Groups in the Query Builder to require patients to have two instances of a treatment, at least six months apart.
Thanks for a great session. Please could you expand for the audience on 1) which country provides the source of the 130 million patients and 2) How does an organization get access to TriNetX solutions?
Countries on the TriNetX network include: Australia, Belgium, Brazil, Bulgaria, Colombia, Estonia, France, Georgia, Germany, Ghana, India, Israel, Italy, Japan, Lithuania, Malaysia, Poland, Singapore, Spain, Taiwan, United Arab Emirates, United Kingdom, and USA.
Healthcare organizations that are a part of the TriNetX network can have access to the TriNetX LIVE™ platform.
Interested in joining the TriNetX network? Click here to get in touch with us and explore the possibilities.
Very slick and good presentations from both groups. Were there any challenges in the ethics/IRAS application, - Using the dataset to power future studies is a fantastic use of the data, is there data on medical device use (i.e. stents, ventricular assist devices, is there data on different makes of these devices? Also, on exhaled breath condensates assaying devices? Is there genetic variant data on various cardiovascular and respiratory diseases, and is there comprehensive data linking one patient's data to various risk factors, drug-interactions, etc.?
Device data can be queryed for using diagnostic and procedural terms in the data model. Brand and SKU information for devices is not queryable in the data if specific procedural (e.g., CPT, HCPCS) terms do not exist for specific devices. Genomics data are queryable, including genotypes (i.e., wildtype or specific point mutations). TriNetX has a Linked Network that links electronic health record (EHR) data to a source of closed claims data for comprehensive real-world evidence generation.
Is TriNetX data available as a csv for training artificial intelligence/machine learning (AI/ML) models?
Yes, users can download patient-level datasets for training artificial intelligence/machine learning (AI/ML) models. For more information on TriNetX Downloadable Data Sets, please click here.