ISPE GAMP^® 5 Second Edition

Key changes

• Recognizing the need for software development in large or custom systems to allow for iterative, incremental and exploratory methods, the agile model is promoted as an alternative to traditional linear waterfall methodology

• The importance of critical thinking has been highlighted and applied to relevant areas throughout the document along with Case for Quality principles, to enable efficient risk-based assurance of software throughout its lifecycle

• User requirement, functional and design specification details in the development appendices have been updated to allow for an agile software approach

• The appendix on Electronic Production Records has been updated to reflect cloud-based technology and blockchain, defining expectations for electronic records, electronic signatures and audit trails

• Operational appendices now reflect the modern Quality Management System (QMS) IT Infrastructure Library (ITIL) approach to software development, as well as updates on topics such as incident and problem management and operational change and configuration management

New appendices

• IT Infrastructure – reflecting modern GxP infrastructure and infrastructure management, replacing paper based quality and compliance records with digital technologies, automation tools and AI and how these should be applied

• Critical Thinking – to support a risk-based approach

• Agile Software Development – focusing on the agile approach to project management and software development in the GxP regulated environment

• Software Tools – as part of the risk-based approach, traditional documents such as installation, operational and performance qualifications (IQs, OQs and PQs) are transitioning to risk-based records, this appendix addresses appropriate tools for modern software development lifecycles

• Distributed Ledger Systems (blockchain) – allowing for the use of distributed ledger technology (DLT)

• Artificial Intelligence and Machine Learning (AI/ML) – recognizing the increasing use of AI and ML, and applying these to GAMP^®