Breaking Down Pharma's Silos For The Sake Of Data Integrity
By Petter Moree, Life Sciences Industry Principal, OSIsoft
The importance of data in pharmaceutical manufacturing cannot be overstated. The data captured during the process development is the lifeblood of any company, as it ensures the most critical elements used, such as raw materials, formulations, and workflows, are meeting the specifications set for a drug product. With the application of advanced drug production and testing techniques in pharmaceutical manufacturing over the last several years, regulators have become more vigilant in their oversight of data integrity, as unreliable and inaccurate data can mask serious quality and safety issues.
Specifically, data collected throughout process development as well as the manufacturing process must meet the data integrity principles of being attributable, legible, contemporaneous, original, and accurate. Also known as ALCOA, these standards are part of a framework defined by the U.S. FDA to ensure compliance of paper and electronic data.1 Regulators want to make sure manufacturers have a system that not only provides access to data but also that the data itself is reliable, searchable, and interpretable. As the pharmaceutical industry strives to make a digital transformation through connected data, companies must figure out how to effectively use and share their insight with numerous other stakeholders across their network.
Connecting Islands Of Data
Every stage of the process, from R&D to clinical manufacturing and, later, commercial production, generates an immense amount of data, all essential to the quality, safety, and success of a drug substance and drug product. However, traditional manufacturing approaches often isolate the information collected from each unit operation into “islands” of data that are managed through different systems and databases and, often, in different computer “languages,” even across different sites and countries. With the success of a product dependent on the collective efforts of a wide range of experts, not sharing this information will likely have a major impact on efficiency and quality. The complexity of data sharing increases if you are working with multiple partners, with each one collecting critical pieces of information about a product’s life cycle.
Working in silos like this, whether internally or with external stakeholders, is a destructive business practice that limits the ability to efficiently capture, share, and analyze process data and weakens the data integrity for a pharmaceutical drug product. Adopting a system that can deliver decision-ready data where and when it is needed, no matter the vendor, source, format, or destination, increases transparency and creates opportunities to identify issues with your process before they create costly problems, such as delays or, worse, lost product. Essentially, it turns data into information by putting it into a format that can be understood, trusted, and identified by various individuals, regardless of its origin. It also allows for proactive problem solving, where you can identify potential problems before they occur and then investigate and troubleshoot them.
Through the unification of data, you can break down the silos in your organization and ecosystem, thereby protecting your investment and providing a product with the intended level of quality and efficacy.
Readying For Pharma 4.0
Integrating digital communities in order to share data in real time is often heard of when discussing “digital twins” in the pharmaceutical industry. For example, leveraging digital twin technology to create 3D views of operations in the drug manufacturing process provides insight into how a process or drug will behave in real time, allowing for improved process optimization and reduced risks. The same can be said for shared data, which is the foundation for the digital revolution of Pharma 4.0, as it leans on the utilization of automation, data, and analytics to reduce errors and improve efficiency and quality.
Moving into this next manufacturing revolution requires more than just the use of automation, advanced analytics, or artificial intelligence; you must first identify the business opportunities you can reap from a Pharma 4.0 structure before you implement any new technologies. For example, having a strong data integrity foundation allows you to integrate other solutions, such as a manufacturing execution system for recipe management, genealogy, and recording or an enterprise resource planning process for material and maintenance management. In addition, any analytics project using machine learning is dependent on data integrity and data governance. Therefore, once you define your overall goals, you can create a framework for success that is supplemented by the right tools, partners, and resources, ultimately driving your business values and building a long-term solution for data integrity requirements enabling business success.
- FDA. (April 2016). Data Integrity and Compliance with CGMP, Guidance for Industry. Retrieved from https://www.fda.gov/media/97005/download.
About The Author
Petter Moree is Global Industry Principal at OSIsoft, specializing in the Life Sciences and Pharmaceutical industries. He has a M.Sc. in technical chemistry with a specialization towards chemometrics and data science from Umea University in Sweden, and has over 20 years of experience helping leading global enterprises obtain optimal value using best practices in data science.
About OSIsoft
For over 40 years, OSIsoft has been dedicated to helping people transform their world through data. Our data infrastructure software turns critical operations data streams generated by utilities, manufacturers and other industrial customers into rich, real-time insights for saving money, making critical decisions or developing new products.
You’ll find the PI System in oil refineries, mining sites, wind farms, national labs, pharmaceutical manufacturing facilities, distilleries, data centers and even stadiums helping people save energy, increase productivity and make better decisions. Worldwide, the PI System handles more than 2 billion sensor-based data streams. Founded in 1980, OSIsoft has over 1,300 employees and is headquartered in San Leandro, California. To learn more, please visit www.osisoft.com.