Unified Under a Universal Language: The Pistoia Alliance’s IDMP Ontology

June 25, 2024

By Irene Yeh 

June 25, 2024 | Identification of Medicinal Products (IDMP) implementation greatly varies across organizations and regulatory jurisdictions, which impacts drug safety and pharmacovigilance. To ensure that organizations around the world can use a universal semantic interoperability based on FAIR data principles, the Pistoia Alliance has built an IDMP Ontology (IDMP-O Release 1), augmenting existing ISO IDMP standards set by the European Medical Agency. This project earned them the Innovative Practices Award at April’s Bio-IT World Conference & Expo. 

During Bio-IT World, Christian Baber, chief portfolio officer at Pistoia Alliance, and Sheila Elz, master data manager at Bayer AG, presented the IDMP Ontology in detail and elaborated on how the Pistoia Alliance brought together 11 pharmaceutical companies—Bayer, Novartis, GSK, Roche, Merck KGaA, Boehringer Ingelheim, Johnson & Johnson, AstraZeneca, Amgen, AbbVie, and Pfizer, as well as representative from EDM Council, ACCURIDS, OSTHUS, and Chemantics—to collaborate on this project. The project is intended to help everybody benefit from IDMP standardization, including improving pharmacovigilance, enabling cross-border prescriptions, and helping the prevention of medication shortages through interoperability with manufacturers. 

IDMP Standardization 

IDMP standards play a critical role in giving healthcare providers and manufacturers reliable information about medicinal products, regardless of their brand name. This is to ensure that medications are made to the correct specifications and can help healthcare providers with a more efficient way to identify and prescribe safe alternatives if one brand becomes unavailable. Despite this, IDMP implementations are not fully standardized, which has resulted in inconsistencies of interpretations amongst the industry, caused misunderstandings, slowed development, and hindered progress.  

A key issue behind the lack of standardization comes from different countries having different drug information and medicinal product dictionaries. The large volume of information is also tedious and complex to sift through for pharmaceutical companies, drug manufacturers, and regulators.  

“We need [a] shared data language like a Rosetta Stone that deciphers what it is we’re talking about,” explains Elz. “If we think we're talking about the same thing, we might be, but it feels like a thousand different identifiers for that one object.” 

How the IDMP Ontology Works

The Pistoia Alliance’s IDMP Ontology aims to facilitate cross-functional and industry-wide unified implementations of the IDMP standard. By serving as a common language, the IDMP Ontology enables effective communication and common understanding among stakeholders, as well as provides a robust data backbone for organizations to automate IDMP data management and standardize data aggregation from different sources. The IDMP Ontology transcends functional silos, providing answers to product-related inquiries that bridge across clinical development, manufacturing, supply chain management, and data analytics.  

This was only possible with the Pistoia Alliance’s framework, which operates much like a membership that provides the resources and services for organizations to work together on joint projects. “By having this framework, by having a sort of minimal level of staffing, we actually enable that collaborative innovation just by getting rid of the problems associated with large companies working together,” elaborates Baber.   

Results and Case Studies

The IDMP standards are a set of five PDF documents, which transforms them into a machine-readable format. Because alignment within the industry, regulatory authorities, and standards development organizations are paramount, an iterative development approach was adopted with the objective of demonstrating tangible business value through real-world examples implemented by the project team.  

Initially, the project focused on the substance and active ingredient, the core elements of a product. To drive ontology development, the project team put together competency questions and use cases that covered previously unanswerable queries and provided substantial value to businesses and organizations. Then, the ontology was developed based on these real-world examples contributed by pharmaceutical company members. Finally, through the pilot implementations, the ontology was created, allowing real-world challenges to be addressed in the early stages of progress. 

With this approach, the development team of expert semantic engineers learned more about the requirements and improved the ontology step-by-step and case-by-case. As a result, the ontology was shaped by specific needs rather than evolved theoretically. More importantly, the real-world business cases that mattered most for the industry were identified, and the team had tangible examples to showcase value. The IDMP Ontology demonstrated that it could help pharma organizations with proving to management that it is a viable and dependable resource to use when carrying out projects.   

In their entry, the authors identify six high-priority real-world case studies for which the IDMP Ontology has the potential to deliver significant value. They include: 

  1. Connecting regulatory and manufacturing processes with harmonized terminologies and seamlessly integrated data to increase efficiency, reduce errors, and accelerate time-to-market for new products.  

  1. Providing a comprehensive product 360, substance, and ingredient view to streamline data management, ensuring regulatory compliance while supporting better decision-making at every stage of the product lifecycle. 

  1. Simplifying the Falsified Medicines Directive (FMD) Report through a standardized framework, enabling expedited reporting while guaranteeing compliance and bolstering patient safety.  

  1. More effective portfolio management and out-licensing report automation resulting in faster decision-making and seamless collaboration with prospective partners.  

  1. A better understanding of shortages in pharmaceutical ingredients or other materials can be gained through the ontology and bottlenecks can be identified more quickly by removing data silos.  

  1. Transforming knowledge into corporate memory by capturing and organizing it within a standardized   framework resulting in improved decision-making, operational efficiency and more innovative practices. 

“I would say we’ve already put it in the water, but now we want to really start gliding across the water,” said Elz. As the IDMP Ontology continues to develop, the goal of a universal codex of IDMP standards gets closer. 

The Next Steps

“What’s happening this year in Phase Three, we’re working together with the FDA and with EMA on use cases of how we can really finally implement these standards and use them as best we can,” said Elz about the Pistoia Alliance’s next steps. But this also means bringing on stakeholders and working with them. Elz used an image of a knot tied with multiple strings of different colors, representing the different perspectives of all the companies and stakeholders that create something bigger and better. 

IDMP-O knowledge graphs are currently being established in pharma companies and 2024 will see IDMP-O data products built on top of these knowledge graphs, generating greater insights for pharma organizations, improving submission data quality, and saving valuable time.