Is 1.7% significant when diagnosis on the spectrum is so subjective? How could they track variations in diagnosis? Did you read my link above?