How do different approaches to the economic modelling of genetic and genomic tests drive differences in results?