The study was further extended, in the third part, with the goal of implementing and scaling the attributes globally. It investigated whether the global dose form attributes could be applied to a broader range of diverse substances and dose forms beyond those initially studied i.e. more focus on variety than quantity regarding medicinal products. The study also examined medicinal products with multiple or ambiguous terms and assessed the scalability of included dose forms and efficiency of the validation process for assigning these attributes.

The collection of medicinal product data was expanded, with approximately 350 substances selected, allowing for an increased number of dose forms to be evaluated. To limit the number of medicinal products, 3,200 were selected—comprising three products per substance and dose form—from eight countries: Brazil, Canada, the European Union (represented by Croatia, France, and Greece), Norway, Switzerland, and the US. The medicinal product data was validated by matching the global dose form attributes to the information provided in each product's SmPC or label. The global scope was further broadened by investigating products on the Japanese market with a subset of the substances. Of the 350 substances investigated seventeen substances were selected. Using the same approach as for the other countries, selecting three medicinal products for each substance and dose form, a total of 148 corresponding Japanese medicinal products were validated. To retrieve comparable results during the study, changes in the Standard Terms [15] due to updates from EDQM were not implemented.

The validation process was further refined during this study. Findings and issues were continuously collected, summarized, categorized, and escalated internally or referred to the GIDWG expert group [20] for guidance and decision-making (Fig. 5). The decisions were used to refine and formalize the GIDWG business rules, which formed the basis for assigning global dose form attributes. For example, the business rules specify the allowed number of terms for each attribute—one term for basic dose form and release characteristic, and multiple terms for intended site and administration method.

Validation process for selecting global dose form attributes for medicinal products.

Throughout the study, over 99% of the 3,200 medicinal products were successfully assigned all four global dose form attributes. The 1% that could not be assigned the global dose form attributes were due to a lack of information in the SmPC. Sixty-three unique combination sets of global dose form attributes were assigned to one or more medicinal products. Of these 63 sets, 47 corresponded to a single EDQM pharmaceutical dose form, indicating the same level of granularity for the two systems. For example, the EDQM term “eye drops, emulsion” was assigned the global dose form attributes “emulsion, instillation, ocular, conventional”. Another example was the term “prolonged-release solution for injection” which was assigned the attributes “solution, injection, parenteral, prolonged”. For the Japanese subset it was possible to assign global dose form attributes to all included medicinal products. This resulted in fifteen sets of global dose form attributes of which ten corresponded to a single EDQM pharmaceutical dose form.

Some of the EDQM pharmaceutical dose forms include more granular information than the corresponding dose form attributes. For example, the EDQM terms “capsule, hard” and “capsule, soft” were aggregated to the same set of global dose form attributes: capsule, swallowing, oral, conventional. Another example was the four terms “coated tablet, film-coated tablet, tablet, tablet with sensor” which all were aggregated to the attributes “tablet, swallowing, oral, conventional”. During the study, 16 unique combinations of global dose form attributes corresponded to two or more EDQM pharmaceutical dose forms. For the Japanese subset five combinations of global dose form attributes corresponded to two or more EDQM pharmaceutical dose forms. All 15 sets of global dose form attributes in the Japanese subset overlapped with the 63 unique global attributes sets.

Five sets of global dose form attributes assigned to medicinal products did not correspond to any EDQM pharmaceutical dose form, demonstrating the flexibility of the global dose form attributes. One example was the U.S. local dose form “tablet, orally disintegrating, delayed release,” which was assigned the global dose form attributes “tablet, orodispersion, oral, delayed” which did not correspond to an EDQM pharmaceutical dose form. For the Japanese data, all global dose form attributes corresponded to an EDQM pharmaceutical dose form.

There were challenges when some SmPCs and labels lacked necessary information for accurate assignment of the global dose form attributes. For example, the manufactured dose form effervescent tablets sometimes lacked information on which administrable dose form it corresponded to—”solution” or “suspension”. This issue was addressed by a business rule assigning the term “solution” in such cases.

To improve global aggregation, rules for harmonization of global dose form attributes were implemented (see Table 1). For example, only the term “injection” was used for the administration method, excluding “infusion”. In another example, the term “syrup” was not used; instead, the appropriate basic dose form “solution” or “suspension” was selected, which required manual validation of each SmPC or label.

For intended site, several terms were often described in the SmPC or label. The following business rule was applied for harmonization purposes: if a medicinal product could be used in more than one way, the intended site was assigned based on the primary use, or on the term with the strictest microbiological requirements. For example, a solution for injection that could also be administered by a nasogastric tube was assigned the intended site “parenteral” as the primary use due to the strictest microbiological requirements. Consequently, the intended site “gastric” was not assigned. In cases where it was not possible to identify a primary use or determine the strictest microbiological requirements, several characteristics were used to indicate two or more uses.

During the process of selecting release characteristics, it was observed that this information could be found in different sections of the SmPC or label, such as the product name, dose form name, or the pharmacokinetics sections. It was also noted that terms were used inconsistently, and this led to an increased amount of manual work, required to identify and assign a correct release characteristic term to the medicinal products.

For example, when harmonizing the release characteristics of patches, the term “conventional” was assigned for local effects, while “prolonged” was used for systemic effects (Table 2). If no information about release characteristics was found, the term “conventional” was assigned. Additionally, the EDQM terminology suggests using the term “modified release” specifically for pulsatile release. However, it was observed that this term was used to describe prolonged or delayed release characteristics for some medicinal products.

Defining the release characteristics for certain products also posed challenges. To address this, the EDQM definition was complemented with detailed clarifications in the GIDWG Business Rules [20], providing guidance for assigning release characteristics. The important element to consider was when the drug becomes available in the body, meaning the time from injection/ingestion/application to the pharmacological response of the drug as described in the SmPC. This does not specifically refer to the time of release but may also consider delayed/prolonged/modified absorption. Often, a comparison was made to a conventional release product with the same active ingredient(s).

Another finding was that products with combinations of release characteristics were described differently in different regions. To address this, it was decided to harmonize these medicinal products according to the USP nomenclature guidelines [21] (Table 2).