Audit Site Selection
Goal
In the Annual Audit Plan (AAP), each planned audit has a target year and estimated duration, but it does not have a specific date to start the planned audit. So the objective of this audit scheduling workstream is to build an algorithm to optimize the timing of individual audits given certain factors, including but not limited to: (1) Timing factors: In general, we don’t want to schedule an audit when the site is super busy, so we will try to avoid conducting audits near the time of important study milestones (FPI, FSA, etc.) or near the holiday. (2) Resource factors: factor in the work schedule and workload of the audit manager, auditor, and other individuals involved in the audit; (3) Travel factors: optimize the travel distance and cost for planned audits.
Expected Outcomes
Our expected outcome from this workstream is an audit scheduling tool, that can provide insights for audit managers to decide when to start a planned audit. The specific output format could be either a dashboard or a flat-file (.xlsx, .csv, etc.) generated by a Python script. The operational plan of this tool will be: When the AAP is finalized and/or modified, our audit planning tool will pull all planned audits from the AAP, and for each audit, the tool will propose some recommended audit start dates given the optimization of all relevant factors
Anomaly Detection in Clinical Data
Goal
Across the industry, the integrity of data’s has been a focused area for numerous years. While in certain GxP area such as GMP, the definition of high level of data integrity can be monitored through a common ground of understanding (e.g human error whether intended or not, ALCOA+ principle), in other GxP area such as GCP and PV, defining Data integrity remains rather ambiguous and subject to interpretation of each involved parties.
The purpose of this Work Product is therefore to establish a shared view across the pharmaceutical industry on how to define and evaluate the integrity of data for GCP area and in the Quality processes associated to it.
To do so we will:
- Share the current practices/ or lack of across our different organizations
- Identify the pain point and HA expectations based on observations history
- Align on a minimum requirements for evaluation of data integrity in areas in scope
Expected Outcomes
Set of metrics and trending guidelines that establish the minimum evaluation level that should be performed to monitor the trustworthiness of our data.
Quality Briefs
Goal
A change in the Quality paradigm to reduce the burden of retrospective, time-consuming traditional QA activities, and ultimately accelerate approval and patient access to innovative products.
- To ensure quality maintains pace with technological advances, using an outcomes-based approach enabled by Quality Briefs* (QBs), supporting innovation to support risk-based assessment and decision making to get new products to patients without delays. To create QBs as a living entity, serving from molecule, filing, through to post-approval and inspections across the pipeline.
- To influence & engage Health Authorities through building trust and confidence via proactive provision of QBs with clear Quality Position conclusions based on risk-assessed Critical to Quality (CtQ) Factors ; ultimately reducing compliance delays on route to approval.
- To create a future of quality professionals acting as enablers of innovation rather than traditional gatekeepers.
Expected Outcomes
- Establish agreed Quality Brief methodology and template available for IMPALA members to pilot
- Establish implementation toolkit & guidance for QBs for IMPALA members to pilot implementation
- Pilot use of QBs with selected Health Authorities to engage, influence & ensure fit for purpose: building trust & confidence.
- White paper on QB methodology, including risk assessment of Critical to Quality Factors, associated Quality Strategy, evidence-based Quality Position Conclusions and outcomes-focus published for open-source availability to industry and Health Authorities
- Long term: Influence Policy so QBs are understood by Health Authorities in both a filing & inspection context
*Terminology may change as the Consortium brings together one-industry approach
Education
Goal
To support the mission of IMPALA to transform the biopharmaceutical Quality Assurance process in GCP & GVP using data analytics, it is imperative to define the framework, including skills, capabilities and requirements needed to develop an analytics domain within biopharma quality organizations.
The Education Work Product Team’s goal is to deliver a framework enabling biopharma organizations to develop GCP/GVP QA data analytics capabilities.
Expected Outcomes
The Education Work Product team’s expected outcome is a framework with the following components:
- Framework including but not limited to
- Guide for developing analytics capabilities
- Recommended competencies
- Lessons Learned & Best Practices (e.g., Adoption Engagement, Implementation and Quality Analytics Champions, etc.)
- Communication
- Whitepapers
- Conferences
- Other forms of media such as Podcasts, etc.
Submit your information to view the completed Framework here.
AI Powered Regulatory Intelligence
Goal
In order to comply with evolving regulation guidelines, pharmaceutical companies are challenged with the labor-intensive process of monitoring and implementing regulatory changes, along with the risk of overlooking critical updates due to human error. This Work Product Team aims to use AI technology to streamline how regulatory updates are handled. By automatically analyzing new regulatory changes and comparing them with internal documents, the AI system will help to pinpoint which documents might be affected and identify any discrepancies or omissions within internal documentation to ensure compliance.
Expected Outcomes
A detailed playbook that outlines the entire process of developing and utilizing an AI-powered Regulatory Intelligence tool. This will include specifications for input formats, output details, AI modeling techniques, infrastructure requirements, and additional information.