Do you still use fever as a minor criterion when applying the Duke–ISCVID criteria for infective endocarditis given data suggesting diagnostic accuracy may improve when it is omitted?