April 28, 2011
"Test Flags Babies With Autism, But Also Feeds False Alarms."
It's a 5-minute questionnaire, to be answered by the parents of 1-year-olds. The point would be to begin treatments earlier, when they might be more effective. I'm not sure what the treatments are... but perhaps there are exercises that could beneficially be done with all babies, so that it would not be crucial to know early on if a child is autistic.
Subscribe to:
Post Comments (Atom)
Posts
21 comments:
Parents, will you forgive me if I say that I can't shake the feeling inside me, that autism is a crock?
Not ALL of it, but a LOT of it.
I don't see anything on Journal of Pediatrics' home page regarding autism or ASD. Nor do I see anything in the table of contents for the May issue [PDF]. What issue is this supposed to be in?
Here's a breakdown of the test results:
10,000 kids were tested.
184 had a 'positive' test. 32 of these were later diagnosed with autism. The positive predictive value of the test is 32/184 = 17.3%. Meaning that a positive test has relatively weak predictive value and will need to be followed up with more accurate assessment. This is in part due to the relatively low prevalence of autism in the general population of kids (1%). The predictive value would be improved if the test was used to screen a population with a higher risk for autism.
Imagine if tornado warnings had an 83% false alarm rate like that Autism test. Who would heed such a test?
Here is the questionnaire used:
http://tinyurl.com/3wtxxzo
Sometimes I think the American Academy of Pediatrics won't be happy until every American baby is diagnosed with autism. We already diagnose 7 times as many children with an autism spectrum disorder as other developed countries. With your help, we can increase this number.
Clarifying Triangle Man's numbers. Out of 184 flagged children:
32 were later diagnosed with ASD
56 with language delay
9 with developmental delay
So, out of 184 children flagged, 97 (over half) had some developmental issue that needed to be addressed, even though only 17% had ASD.
Parents, will you forgive me if I say that I can't shake the feeling inside me, that autism is a crock?
Not ALL of it, but a LOT of it.
I don’t know if it’s a “crock” but I do tend to be somewhat skeptical of claims that we’ve seem a large increase in autism. I suspect that what’s going on is largely a combination of (a) better diagnostics and (b) children now being labeled as “autistic” who years ago would have been labeled as something else. That being said, I wouldn’t rule out the possibility that doctors are over-diagnosing children as autistic just like I wouldn’t rule out any patient being over diagnosed as something that is on the forefront of people’s minds because it’s received a lot of media attention.
As far as the test itself, a lot of the success of our health care system comes from early detection (e.g. cancer survival rates) and it makes sense to me that all things being equal, an autistic child who is treated earlier is more likely to have a more successful treatment (such as it is). It also means that we’re probably going to have a lot of initial false positives as we learn more and get more sophisticated in how we detect and treat autism (just as we have with things like cancer). As such this questionnaire is probably just part of the early stages of refining a process that will hopefully lead to more successful early detection and treatments down the road.
I have a daughter "on the spectrum." I'm pretty sure in the past she wouldn't have been diagnosed with anything, just with being a quirky kid. I started out very suspicious of the whole autism thing. But I think over the years I've learned
1) there really is a there there. Kids diagnosed on the spectrum really do, in my experience, have enough in common to put a name on their characteristics. One thing they do seem to have in common is the need to be taught things that other kids just pick up on their own. It isn't a matter of low intelligence, if by low intelligence you mean an inability to learn, even when taught. It's more like a kind of cluelessness, especially about interpersonal relationships.
2) intervention seems to work. Basically, a lot of autism spectrum intervention consists of teaching these kids precisely the things other kids just pick up on their own. It does seem to draw them out of their inward looking selves and it does seem to work not just for the one thing being taught, but also in teaching them to be more like other kids in regard to learning things on their own.
"teaching these kids precisely the things other kids just pick up on their own"
Well said.
The disorders are overdiagnosed, but these problems can be ameliorated by coaching.
It needs to be de-medicalized.
But that's our culture, ain't it?
We loves us some health care.
I skimmed the validity research for the instrument and it is really sound for picking up verbal delays with a good rater, but there are myriad reasons for verbal delays.
I agree that more kids are being diagnosed with the spectrum, and that can be good if it is used to help parents get taught how to provide a very emotionaly warm and verbal environment for the children.
Of course, all children need an emotionally warm and verbal environment.
Trey
I suspect a lot of kids who are being diagnosed with autism now would have just been treated like normal kids in the past and been described as "a little odd, but basically all right". That was probably a better way of handling it than medicalizing the issue.
If you don't mind false alarms, just say every baby has autism.
Speech therapy, ABA therapy, floor time are all extremely beneficial for very young children with autism.
HOWEVER, there is a bigger problem with this. The false positives can have horrific consequences if the child is not thoroughly evaluated to rule out all other problems, including a hearing test.
Finally, I firmly believe based upon personal experience that autism is being over diagnosed.
@Triangle Man and MadisonMan:
You guys both make a very good point, though MadisonMan seems to think it's a flaw with THIS particular test. Actually ANY test, even if it is 99.99% accurate, will give MOSTLY false positives if it tests for something rare in the population. That is a simple statistical fact.
Some simple numbers:
We test for something that's 1 in a million, we test a hundred million people, the test is accurate to 99.99%.
This is what we get:
True positive: 100
True negative: 99,989,900
False positive: 10,000
False negative: 0 (population not big enough to show false negatives)
In this example, if you get a positive result, 99% of the time it is a false positive.
It's not a defect in the test; it is a defect in our understanding of what the test does. The accuracy of the test is simply not enough to judge the result, you also must consider the rarity of the condition.
Suppose that we are testing for a condition that's one in a thousand with a 99.99% accurate test:
True positive: 99,990
True negative: 99,890,010
False positive: 9990
False negative: 10
In this case, a positive result has a 9% chance of being false.
Our pediatrician already does a questionnaire like this. It doesn't say that it's for autism though; you don't find that out until you go into the exam room, and he says, "That was a questionnaire for autism." "Oh."
Has anybody considered the possibility that autism is something a normal child grows out of?
Substantial lacks in comments on Autistic Spectrum Disorders(ASDs) seem the rule so far. I have Asperger Syndrome, an ASD. I have had it all my life, and am now looking at my 60th birthday in 2 months,...so, no, you don't grow out of it.
It is already demonstrated that ASDs are associated with particular differences in the brain from differences in the phasing of growth of different parts of the brain. In many of us, the neo-cortex grows faster, which will link cortex cells together while waiting longer for the Thalamic System that processes emotional communications to mature. This means when the Thalamic System send axons towards the Neo-cortex to connect, there are fewer connections possible, and those axons die back. The result is a delay in receiving and expressing emotional communications from the thalamic system. These are the majority of social communication. So, Aspies famously lack social nimbleness.
ASDs were shown to be genetically caused as early as WW2, but political roadblocks inside and outside psychiatry kept this from being acceptable doctrine till the 1990s. Now we have MRI studies and genetic studies showing the physical differences and some of the several dozen genes involved.
ASDs swelled recently for 2 reasons. First, neurotypicals stopped killing us so fast, after the industrial revolution provided niches where we could be of use. Second, many of us got jobs in computer industry cubicles next to other Aspies, got married, with better incomes, and the genes of each often added up to someone further down the Spectrum than either parent.
ASDs are not mysterious, their real causes are simply very uncomfortable for both neurotypicals and those on the Spectrum. We don't like thinking how close we are to being killed. Neurotypicals don't like to think their ancestors could have "done things like that." We are, and they did. Get used to it.
Nice blog about babies feed and theirs health I agree that more kids are being diagnosed with the spectrum, and that can be good if it is used to help parents get taught how to provide a very emotionaly warm and verbal environment for the children.
diaper coupons
its nice blog This is in part due to the relatively low prevalence of autism in the general population of kids (1%)
braille numbers
Is the treatment over at present.. These test flag babies are entirely new for me..
pampers gifts to grow
Post a Comment