The challenges facing the Digital Patient initiative
Some important challenges to be overcome were brought up during the meeting. These challenges provide food for thought on how best to structure the research roadmap in order to tackle them.
Patient Consent and Data: It is imperative that medical data be used to drive the design of computational models, to provide for validation of models, and to allow the models to function in real-life. In order for this to happen, the legal issues surrounding patient data and the accessibility of data must be addressed.Validation: By far the most cited challenge at the meeting was, inevitably, validation! However, some important insights into the validation process were also given. For example the different meanings of validation in different applications from models for understanding (which may require less clinical validation) to models for decision-making (which clearly need to be the most robust). The question of who will fund (or be able to fund, given their long-term and expensive nature) clinical validation was raised.Communication between stakeholders: Understandably, members of different stakeholder groups have differing vocabularies and levels of understanding of other stakeholder groups. This can lead to divergence of, for example, the research direction from the clinical need. Greater communication of the possibilities and limitations of modelling is needed, as is communication of the specific clinical scenarios in which computational methods can assist.