As part of the consultation meeting, representatives from two of the important stakeholder groups discussed within their groups about what they thought were the most important issues going forward.
Clinicians: The focus of research was a key theme in the
discussion, in particular, how computational models must be relevant, and that
a highly developed model ill-suited to the clinical decision-making pipeline
may not be very useful. Also, the ability to feedback long-term data on patients
was seen as important for model development, and in particular to focus on
disease prevention. The interaction
between the clinician and the model is crucial, in particular, how the
information on prediction (as well as uncertainty estimation) is conveyed to
the user. It was suggested that the scope for the model in decision-making
should be clearly defined, the analogy given was that of the autopilot. While
models can provide valuable information throughout the decision-making
pipeline, clinicians will need to make final decisions.
No comments:
Post a Comment