Disability Scoop has an article: CDC Set To Release New Autism Numbers. This was posted last week, and it starts:
The Centers for Disease Control and Prevention are expected to unveil a new autism prevalence estimate as early as this month.
The agency currently says that 1 in 110 children have autism, a figure first released in late 2009. Now, less than three years later, the CDC is set for an update.
I had heard rumors that the next prevalence estimate would be much higher. Not higher than the Korean prevalence numbers, but a significant jump. But–I heard these rumors together with the prediction that the estimate would come out last year. So, rumors are worth what you pay for them.
No point trusting something that has sold out to autism speaks. How many autistic people does it take to prove that it’s the diagnosis that cause the most harm some of the time?
Hello friends –
I’d be interested in hearing anyone’s take on the rationale for a complaint you see in some circles regarding the shifting of which states are used to comprise these estimates on a report by report basis; i.e., report with NJ numbers four years ago, but not two years ago (or whatever the timeframes / years are).
To my mind, I’m struggling to see how this is good practice; it would seem we could get better values by seeing every state, or the same states over time.
Maybe someone out there has some insight on why providing data from a subset of states, and a different subset of states is a good idea, or at least not a bad idea. (?)
– pD
pD–
I’ll try to find out. Keep in mind, the monitoring sites are not CDC people. The CDC partners with groups in the states. CDC can’t force groups or researchers to stay in the network.
Here is a list of members of the team from the 2009 report (looking at data from 2006, for kids born in 1998…it gets complicated fast!):
My guess it that they would like to have monitoring sites in all states, but that the cost is an issue. Even within each state they don’t survey everyone.
pD, I can’t comment as I am not familiar with the collection methods but would be interested.