4 responses to “Recommended Reading for August 24, 2010”

  1. Kaz

    I am highly interested in this autism diagnosed by MRI thing. People talk about how ASDs are omgsoeasilydiagnosed! *rolls eyes* but if that’s true at all, it’s really only the case for children. I’ve been through the mess that’s adult diagnosis and it took me ten months from first looking into diagnosis to actual diagnosis, and I was in the best possible position I could have been in many ways. (I was living in a city with an autism centre that has a branch specialising in adult diagnosis, and offers free assessment.) Things look decidedly different for a lot of adults looking for diagnosis. So although I worry it might become exclusive (if it identifies 90+% of autistics, what about the other 10%?) a test that’s relatively easy to administer (well, MRI) and objective (the symptom list for ASDs and various diagnostic criteria is biased towards kids because it’s usually diagnosed in childhood, there’s issues that it might be biased towards men and there’s undoubtedly other things here, and that’s not even getting into the objectivity of the professional doing the assessment…) would be good to have.

  2. n

    Yeah, there are a lot of concerns I have about MRI identification, like autism just being a name to describe certain stuff which doesn’t necessarily say anything about what’s at the root of things (or if any *one* thing even is) or if it’s the same thing for everyone, and what will happen if not everyone who falls under the header by observation ‘passes’ the MRI? But it might be good.

    Also everything Kaz said.

  3. Norah

    Actually that (above) was me. (I did something iffy when I was entering my name and email and such and closed the page or something). I had thought that the comment had just failed and was waiting until later when I’d be able to write something up again to try again. But I see it got through anyway.

  4. Ali

    The autism brain scan thing is…not awesome. It actually kind of sucks. Not because scanning brains for autism is a bad thing, but because the statistics of it means it’s pretty much useless.

    If you have 1000 people, about 10 have a diagnosable autism spectrum disorder. The scan is accurate for 90% of autistic people, and about 80% of non-autistic people. That means that 9 of those 10 will get accurately labelled as having autism. 1 will be mislabelled as non-autistic. Meanwhile, there are 990 people without autism. 792 will accurately be told they are non-autistic. 198 (!!) will get back false-positives. That means that the test has identified 207 people as having autism, and 793 as not having autism. Only 9 of those 207 people actually have autism! The test is only accurate in 4.34% of people who get a positive test result.

    This does a fairly good job explaining it: http://www.guardian.co.uk/science/blog/2010/aug/12/autism-brain-scan-statistics

Subscribe without commenting