Sections in This Article
About Signal Data Sources
Vault Signal supports the following data sources for Signal detection: FAERS (FDA Adverse Event Reporting System) data, Vault Safety (Pharmacovigilance) data, and EVDAS (EudraVigilance Data Analysis System) data.
FAERS data is ingested quarterly, after the FDA release. FAERS data is also cleaned to remove invalid data entry inputs and more. See Data Cleaning for more information.
The PV data is copied in real-time as each Case (ICSR) reaches the Completed state.
The EVDAS system is used to access Adverse Events in the EMA regulatory database. EVDAS data must be manually loaded into Signal in the form of eRMRs (electronic Reaction Monitoring Reports).
Vault Signal retains all retrieved FAERS, PV, and EVDAS data displayed on the Signal Detection Dashboard.
About the FAERS Data Source
The FDA Adverse Event Reporting System (FAERS) is a data source used by pharmaceutical companies to report, extract, and view adverse event data on post-market drugs submitted to the FDA. Healthcare professionals (such as physicians, nurses, pharmacists, and others), consumers, and manufacturers submit reports directly to FAERS. It includes over 15 million ICSRs and over 60,000 medicinal products.
The structure of the FAERS database adheres to the international safety reporting guidance issued by the International Conference on Harmonisation (ICH).
Signal ingests FAERS data quarterly, shortly after the FDA release.
When you initiate Signal data mining with the FAERS data source, Vault Signal analyzes the FAERS database for information on adverse events in relation to post-market drugs. Signal compares the disproportionate reporting of an Adverse Event associated with a particular Product with the same Adverse Event for any other Product in the database. If an Adverse Event for a particular Product is unexpectedly high, it displays a Product-Event pair which may indicate a causal relationship between the Product and the Adverse Event. Product-Event pairs are displayed on the Signal Detection Dashboard through the Distribution chart, the MedDRA Heat Map, and the Product-Event Combinations grid.
See Use Signal Detection to learn more.
Note FAERS does not yet include Submission data on Adverse Events reported during clinical trials of drugs or biologic medications.
MedDRA Coding Signal Data
Vault Signal auto-codes the FAERS Adverse Event for each report against the latest MedDRA Preferred Term (PT) version. If the term can’t be matched, Vault Signal attempts to auto-match against the MedDRA version on the Individual Case Safety Report (ICSR). Unmatched terms are manually reviewed and coded by a Vault Signal Data Steward.
Vault Signal uses the MedDRA PTs on the Signal Detection Dashboard. Product-Event analysis is performed at the PT level.
Data Cleaning
Vault Signal performs data cleaning at the time of quarterly ingestion. Data is cleaned using the cleaning methodologies and metrics, and by transforming and cleaning invalid entries upon ingestion.
Data Cleansing Metrics
Vault Signal uses the following techniques to automatically clean entries wherever possible:
- MedDRA PT Term Coding
- Product National Drug Code (NDC) Matching
- Generic Name / (Registration) / (Other Product Fields) Imputation
- Field mismatches (Medicinal Product / Brand Name corresponds to Generic Name)
Data Transformation
Vault Signal uses an ETL (extract, transform, load) pipeline to transfer the reports from the FAERS package to the Signal database. As the data is being copied to the Signal database, Signal automatically prescribes a set of transformations and normalizations. This includes normalizing values to a standard, usable format or cleaning invalid inputs (for example, removing invalid characters and matching Products or reactions to a standard form in NDC or MedDRA, respectively).
About the PV (Pharmacoviligance) Data Source
The PV data source uses Cases from Vault Safety for Product-Events analysis.
Before Vault Safety Cases can be included in Signal detection, they should be copied to the Signal database. Cases that are already in the Closed, Superseded, Voided, Nullified, or Deleted State Type lifecycle state can be copied to the Signal database in bulk using the Send to Signal action. Individual Cases are automatically copied to the Signal database once they reach the Closed, Voided, Nullified, or Deleted State Type lifecycle state. See How Vault Safety Copies Cases to Signal for more information.
Once the copy is completed, Vault Safety Cases can be filtered and organized on the Signal Detection Dashboard and the Signal Workbench by selecting the PV Database data source. The latest version of the Case for the selected reporting period is retrieved from the Signal database, processed, and calculated for Product-Event pairs.
Prerequisites
An administrator must first enable PV Cases on Signal to use the PV data source.
Link PV Cases to FAERS Cases
You can display PV and FAERS Cases as a combined data source for a more accurate Product-Events analysis. Vault Signal automatically links your Vault Safety Cases to the FAERS Cases based on the Case unique identifier fields. Combining these two data sources provides a larger data set and number of Cases. For example, the combined data source includes Cases from Vault Safety that were not submitted to the FDA as well as Cases submitted to the FDA that were not have been received by the Marketing Authorization Holder (MAH).
To display linked PV and FAERS Cases, select both data sources on the Detection Summary panel on the Signal Detection Dashboard or the Workbench Summary panel on the Signal Workbench.
Vault Signal calculates the Case Counts and scores, and updates the page with the parameters of the combined data set.
About the EVDAS Data Source
EVDAS, the EudraVigilance Data Analysis System, is used by Market Authorization Holders (MAHs) to access the summary of Adverse Events in the EMA database.
EVDAS data is contained in a eRMR reports (electronic reaction monitoring reports). The eRMR is a summary report that includes the active Substance, the MedDRA Adverse Events, and some Signal scores such as the ROR.
For EVDAS Signal detection, you can load eRMR reports into Vault Signal using Vault Loader. Once the reports have been loaded, you can access and export this data from the Signal Detection Dashboard.
For more information, see EVDAS Signal Detection.
