Data Fusion for Quality Systems
Enhance Bio/Pharma Operations with
Data Fusion for Quality Systems
Data Fusion for Quality Systems enables you to answer difficult questions about your Pharma operations by unifying data from any database or file in order to analyze both quantitative and qualitative (text) information. Data Fusion for Qulity Systems is built on NoviSystems’ Fusion Analytics platform.
- Process deviation information and isolate specific terms and develop models of machine learning to determine if issues are becoming trends.
- Consolidate features from multiple systems to determine true root cause of process or testing failures. PCS, Deviation, and LIMS results can be linked.
- Easily match training data with personnel
- Quickly develop search criteria using NLP to explore development records for facts necessary to support investigations.
- Create custom dashboards to draw insights from your data.
- Aggregate data from multiple disparate data sources and make them act as one dataset.
- Interact with dynamic visualization tools to filter data to find the answer to your questions.
- Use powerful natural language processing and machine learning tools to create models to classify and predict outcomes from new data.
- Fusion Analytics is packaged and sold as a subscription.
- The subscription includes the software, hardware, consulting, and support you need to easily begin using Fusion Analytics in your environment.
- Fusion Analytics does not require any infrastructure or integration activities other than permissions for configuration.
“With Fusion Analytics, our organization has taken a more systematic and data driven approach to understanding our deviation trends. Fusion Analytics gives us the ability to quickly identify areas of weakness, and implement appropriate and meaningful actions to help reduce and eliminate deviations across our global manufacturing network. Fusion Analytics has been a real benefit for our quality teams.”
Data Fusion for Quality Systems Live Demo
FDA ViewFinder is a set of Data Fusion for Quality Systems dashboards that track FDA activities related to inspections and adverse events. These dashboards are available free of charge to illustrate the power of Data Fusion for Quality Systems.
These dashboards have the following capabilities:
- Using Natural Language Processing, the system scans PubMed for recent publications regarding adverse events as well as events posted in MedWatch.
- 483 Filings and Guidance Documents are automatically uploaded and added to the database. This data is evaluated using NLP and Machine Learning for topics applicable to Operations.
- If a Fusion Analytics subscription is purchased, ViewFinder is included. This allows the user to compare FDA findings with internal deviation reports or other quality system processes.
- The system also can connect to LIMS systems for in-depth analysis of trends.
My frustration at the inability to gather and analyze data from different development and production systems led to the formation of Fusion Analytics.
Click below to access FDA ViewFinder dashboards
The Citations and Classifications dashboard uses aggregated FDA data in an interactive dashboard that allows the user to filter and visualize content based on any combination of categories present in the data as well as using Natural Language Processing and Machine Learning. These data are valuable in establishing inspection trends.
The system contains Form 483’s obtained from the FDA regarding Drugs and Biologicals. This includes the commonly requested Form 483’s as well as those obtained by FOIA requests (selected based on OAI). The system uses Natural Language Processing a well Rules-Based and Machine Learning to select topics of interest with subsequent interactive visualizations. This dashboard can be filtered by the user using word searches.
The system allows for the visualization of the 483 as well as abstraction of key phrases and sentences, identification of Inspectors and the citations associated with various inspection teams.
Like the Form 483 dashboard, the system provides an interface for exploring all Drug and Biologics Warning Letters. The Dashboard was developed using NLP; the user can supplement the filters using term searches.
FDA ViewFinder refreshes its data nightly. When required, FOIA requests are filed on a monthly basis for all 483’s where OAI is indicated.
The Adverse Events dashboard allows the user to explore the FDA adverse events database in an interactive manner. Each attribute associated with adverse event reporting is linked so that the user can “drill down” and quickly visualize their results to spot trends.
In addition, the system uses Natural Language Processing to scan recent PubMed abstracts searching for reports of adverse events. Key phrases from the abstracts are displayed in the ViewFinder as well as links to the PubMed abstract.
This dashboard contains over 1600 Guidance documents that can be searched for topics of interest. This is of particular value when determining or establishing context behind as well as in formulating responses to inspection findings. The system is also useful when developing protocols or proposals for regulatory review.