Monday, September 21, 2009

Mark Onslow's talk at BSA London 2009

Here is a summary of Mark Onslow's talk at BSA.

One aspect of his talk was social anxiety and fluency improvements after therapy. They found that the social anxiety scores went down significantly after a cognitive behavioural therapy but not the fluency score. I am not surprised after having been to many therapies: it is very difficult to stay more fluent over a long time (well actually until forever), but it is much easier to change someone's social anxiety levels and attitudes. It is most noticable in intensive therapy or self-help groups, newcomers are really scared to attend the self-help group or group therapy, but after a few sessions they loose up. I guess their body realizes that it's not going to die! The reason behind this is mostly neurobiological. Beliefs we have stored in our brain are just that statement. They are were different to individually learned memories like associations, motor code or episodes of our life. I can change some of your beliefs in a second, for example: show you a picture of your cheating wife. But I cannot change your individually learned memories like your fear of spider, your way of speaking fast, or your childhood memories easily. These need to be changed permanently if you want to have lasting fluency gains. Social anxiety is fuelled by beliefs that we are holding. Change them and your anxiety is gone or reduced. He also mentioned that a colleague has compiled a list of 66 beliefs on stuttering that contribute towards social anxiety.

Mark also talked about his study led by Reilly et al. that, according to his words, even Tom Weidig is agreeing with. Well, of course that doesn't mean that the study is correct or does not need to be carefully studied to be interpreted as their numbes are relatively high. I have been talking about the study here, and I discussed the study on StutterTalk here with Peter and Eric. He reported that a follow-on study is out to the reviewers. The first observation period stopped at age 3, and 8.5% of the kids had shown dsyfluent speech. And assuming 1% of the population stutters, you get to about 90% recovery of kids. Mark showed data from the one-year extension period; so they observation stops at 4. And the curve of kids starting to stutter climbed further to 12.5% or 13.5% (I cant remember), and then plateaued (i.e. decreased only very slowly). Clearly, most kids start stuttering before age of 3.5. The data further suggests a high recovery rate; now you have to divide 1% (adults) by 12.5% (kids), and you have more than 90% recovery. Of course, the 1% of adult population might be too low; covert stutterers might be missed. Still the recovery is very high. He also said that in this new observation period only 1 in 7 kids recovered from stuttering! That's very low. So I am looking forward to reading the article. And if you are one of the reviewers, review it now! I don't want to wait for months and months.

He also talked about telehealth, and on doing therapy over Skype. In principle a good idea but Skype is still too unstable to be very effective in my view. And I am not sure it is easily solved in the future as more users come onto the net and more data is pushed around.

And again he showed this meta-analyis graph of Lidcombe. Competely flawed graph. He should have shown the long-term outcome of the random control trial that is not so glorious but the only one coming close to a good outcome trial, except of course the low sample size. And he could have mentioned that Marie-Christen Franken is working on a large-scale trial.


Norbert @ BSA said...

At his study day in London mark also showed research that suggest that maintenance of fluency after therapy is much higher if the social anxiety is addressed through a programme like CBT. Again, makes sense when you think about it...

However, I have concerns about the Reilly study. What are we to do with research the findings of which fly in the face of our daily experiences? There are about 30 children born after me in my family (with a history of stammering and language delay) and not one of them has ever gone through a dysfluent phase. If I got the findings right, the research states that one in eight children should stammer noticeably for more than a year in their development by the time they are four. I can't find those children and I have spoken to a number of people and they can't find them either. One in eight?

Tom Weidig said...

I don't know about the high numbers. But what I know is that 1) the sample is very big, 2) they seem to well organized and apply criteria in a very systematic way, and 3) the article was well written and balanced.

They might have a lower threshold for stuttering. Or it might be that most experts simply have a biased view because they never see some stuttering kids whereas here they see the whole population.

The estimated recovery rate might be too high, because the 1% adult population taken to compute the recovery rate is too low. There might be more adults who are stuttering, but not in ways that show significant handicap to them. Or the 1% does not pick up much covert stuttering. Maybe some of the 12.5% of the kids are "going into hiding" and become covert stutterers.

There was some open issue in the article that the authors to their credit discussed. Stuttering seems to be related to the socioeconomic status of the mother. Maybe these mothers are more likely to detect and report stuttering. Or maybe there is a certain bias in the data. OR, interestingly it might show a higher environmental demand on the child which pushes some borderline kids into early childhood stuttering.

But what they certainly have done is to show that stuttering starts with a sudden onset in most cases.

Anonymous said...

"But what they certainly have done is to show that stuttering starts with a sudden onset in most cases."

Good. The sudden onset finding.

It is a total myth that total recovery is impossible after childhood, it is possible, with and without the help of speech therapists. SLPs like to use scare tactics....

Who said 1% of the population stutters, how do we really know for sure? It is an estimate. The trouble is, researchers have an incentive to over-inflate, because it does no one any good to prove that less people stutter than it is widely believed.

The measurement of stuttering needs to be further refined in the 21st century. How you measure stuttering is very important.

And we need to consider the fact that treatment programs/SLPs "bully" their clients/PWS into showing less stuttering after treatment.

Before Treatment: Oh, please, show us how you stutter....let your stuttering come out. 25% stuttering.

After Treatment: Now, we are going to tape record you reading a passage. 1-2% stuttering.

If you had spent $2000 going to a treatment program and the treatment director calls you for a follow up after treatment, wouldn't the PWS be afraid to show too much stuttering and just pick words carefully.

It is laughable that people who have a vested interest conduct their own research. Like auditing your own taxes. How do you police yourself.

The Lidcombe program needs a thorough independent evaluation study from a third party. All the outcome studies of Lidcombe came from the same Australian Stuttering Research Centre.