Aims Paediatric drug development is an example of a high-dimensional complex problem that must interrelate a paucity of diverse data, much of which is subjective and fragmented. Using the operational research framework of General Morphological Analysis (GMA), this abstract describes the construction of a decision-support tool that is able to capture clinical, commercial, and formulation factors in the development of viable paediatric formulations.

Methods GMA, a method developed in astrophysics,¹ is a framework for structuring and investigating complex relationships contained in a multi-dimensional problem field.² A morphological field was developed by modelling the problem space of the paediatric drug development field by a subject-matter specialist team. The problem space is a set of critical parameters (aka factors or dimensions) where each parameter consists of possible states/values as identified by the working group. The workshop team comprised of an industrial pharmacist, a paediatric pharmacist, a Qualified Person and technical scientist from an international contract manufacturing organisation. The solution space was synthesised by excluding logically impossible and empirically improbable pair-wise combinations of parameter values achieved by a process known as the Cross-Consistency Assessment (CCA).³

Results The morphological field analysis isolated a 7-dimensional field with 55 values, giving a total set of 560 000 possible configurations for the five age groups as defined by the European Medicines Agency. A configuration is a point in a high-dimensional space that can be represented as a line cutting each parameter at the appropriate value.

Dimensions included age of the paediatric patient, dosage form type, problematic excipients, drug characteristics, delivery constraints, types of hurdles (eg, clinical, manufacturing) and the regulatory strategy (using the Ansoff typology). Facilitation of this vast problem space by the CCA process dramatically reduced the viable configurations to 57 457, a reduction close to 90%—reduction greater than 99% is not uncommon in such exercises. Deeper analysis revealed that only 78 schemes (ie, configurations) were available for the pre-term age group and 760 for the new born category (0–27 days).

As many of the values involved are not meaningfully quantifiable, containing strong social, regulatory and clinical concerns, GMA facilitated (i) a collective judgmental process to be placed on a grounded methodological basis, and (ii) a process of constructive dialogue amongst the diverse participants to develop shared concepts, terminology and ownership of the problem formulation.

Conclusions Transdisciplinary approaches, such as GMA, allow a holistic picture of a fragmented and complex landscape to be developed by the relevant stakeholders. This not only gives decision support under genuine uncertainty but allows incorporation of emerging data derived from pure and applied (translational) research in early paediatric drug development programmes.