Ian Turner comments on How much (more) data do we need to claim extreme cost-effectiveness?