Nov 2020: EDM Data Dictionary and Release Notes

COVID Updates

We’ve updated the COVID-19 related section of the Knowledge Center with data and trends recent through the end of September 2020.

Structured Data

We identified and have since updated a small number of cancelled orders in our database that were previously appearing as non-cancelled orders. Approximately 1.5% of patients across EDMs experienced some change in line of therapy, with about 1% experiencing a change in line of therapy (LOT) line name and about 1% experiencing a difference in time to treatment discontinuation (TTD) at the cohort level.

The median difference in line start / end date was 0 days, and the median impact on time to TTD at the cohort level was 0 days. The affected orders have not been shown to be associated with specific treatment patterns or therapies.

Methodology and Analytic Guidance

Flatiron has published a new position paper to the Knowledge Center describing the supporting evidence for several of Flatiron’s recommended approaches to considering the various testing-related dates in the EnhancedBiomarker table when assigning patient-level biomarker status for applications such as study cohort selection. This article supplements the Assigning Biomarker Status for Cohort Selection Methodology Tutorial.

Flatiron R Packages

We’ve released a new package - rwendpoints (0.1.2) - that helps users prepare data for and run overall survival and time-to-treatment discontinuation time-to-event analyses. The package is designed to help implement the guidance found in our methodology tutorials, while also providing enough flexibility to support your specific use cases. Functionality includes:

• derive_duration_and_status: takes a single data table with patient IDs, index dates, censor dates, and event dates, and derives the time at risk and the binary event status for each patient.
• prepare_overall_survival: prepare time at risk and binary event status for overall survival time-to-event analysis
• prepare_time_to_treatment_discontinuation: prepare time at risk and binary event status for time to treatment discontinuation time-to-event analysis

If you do not have access to this or any of our other packages, please submit the form on this webpage and Flatiron will follow up with further information.

Network Changes

Two new practices, Practice IDs SADA451D4B1FB and S4A783FFFDBC9, have been added to EDM sampling eligibility. Patients from these practices will begin appearing in EDMs and other Flatiron datasets (e.g., Spotlights) starting with October deliveries.

Additionally, as part of ongoing checks to ensure the quality of our data and compliance with our practice requirements, Flatiron identified two practices for removal from EDMs due to stale data. As of this delivery, Flatiron has removed all records associated with Practice IDs S2C7623FA7BFE and SC2A1560A6818 from our EDMs. These records have been replaced by organic cohort growth, and there are no decreases expected in any dataset sizes.

Due to two in-network acquisitions, patient data associated with two Practice IDs will be newly associated with new parent Practice IDs with a 30 day recency. Patient IDs associated with these practices will change across EDMs and all other research deliverables with this migration.

• Patient data associated with Practice ID SBED7B7197691 is now associated with the parent Practice ID S32D7D3854A0D
• Patient data associated with Practice ID S415080450E6E is now associated with the parent Practice ID SBA2549C29E0D.

Disease	Total EDM Patients (November 2020)	Academic Patients (November 2020)
Chronic Lymphocytic Leukemia	12236	895