Diagnosis

I was astounded while reading this article from the New York Times, about the revised estimate of the prevalence of Autism, when I came to this passage:

The new estimate is about the same as one from a study published in October, which found a rate of slightly more than 1 in 100 children who received a diagnosis. Yet that study, based on a phone survey of 78,000 households, also found that almost 40 percent of the children who had received an autism spectrum diagnosis grew out of it or no longer had the diagnosis.

(I tried to find more information on the study, but only a summary was available, and there was nothing about children who were no longer diagnosed.)

Understand, I'm not dismissing the results of the Centers for Disease Control and Prevention study (which was based on an analysis of medical records of 400,000 children), but I can't help but wonder how valid a conclusion can be reached based on a medical study conducted by phone. More importantly, though, what criteria do you use to determine who "grew out" of the diagnosis? Is that a medical opinion — or a parent's hope? Are parents really having their children re-diagnosed?