Tuesday, December 23, 2014

Lidcombe is dead: long live any treatment?

It took Peter years to get Marie-Christine Franken on StutterTalk. So enjoy the podcast! And it took her years to have the 2007 comparative study between Lidcombe and Demands&Capacities analysed and published: a true source of frustration to me as this is such a key piece of research to dispel claims of supremacy by Lidcombe. Already in 2012, they had the data but it took more than 2 years to write it up and get it published.

The listener will note that Marie-Christen is a bit scared of misspeaking and of her findings... She might be an overly cautious person by nature, but if a friend of a friend told you that she actually got bullied from someone in a distant land and was concerned about her job security, you can understand why she behaves the way she behaves.

The key statements are
  • The large-scale study started in 2007 did not find a significant overall difference between the two treatment approaches: Lidcombe and Demand and Capacities treatment.
  • She expects the study to be published within months.
  • She refutes that it is only ethical to use Lidcombe.
  • Not sure about subgroups, i.e. one treatment better for some children.
  • She believes that Lidcombe and DC are better than no treatment and really make a difference and contribute to recovery, and can speed up recovery, but as far as I understood this is not based on hard evidence. The study only looks at the relative difference between the two treatments not the absolute.
  • If not in terms of fluency but to keep the problem to a milder stuttering and good psychosocial adaptation.
  • A couple of children did not work well with Lidcombe. Some children are not fluent on a word level. Some children can be very fluent on short sentences. Kid does not like contingencies.
  • They talk about an interesting therapy component, namely to practise nonsensical words.
In conclusion: Her results fits my predictions (see past posts), which is also based on the knowledge that, in general, psychotherapy research has shown than no treatment is clearly better than any other. Both approaches perform similarly. They did not discuss whether the quality of patient-client relationship is important: in psychotherapy it is key.

I do not believe that the study has shown that ANY treatment is actually better than no treatment. And I actually believe that no treatment can achieve full recovery much better than natural recovery, because it is a neurobiological phenomenon. They have not talked about neurobiological limits to recovery. This is very important, because if something is neurobiologically messed up, you can only do so much by shaping behaviours, especially for very complex brain processes as speaking as opposed to simple motor skills.

But I do think that treatment helps to get a good psychosocial adaptation and to cater for the parents' needs.


Don T. said...

Sad to see her pull the "we know that early intervention works, so we can't do the study that would test whether early intervention works" business. Because WE KNOW! So much for scientific rigor. She does some hand-waving about different populations - those represented by parents who bring their children into therapy early and those who don't. If she tried to pull that one at a conference, I'd read her the basic scientific competency riot act. The results don't surprise me at all. Both methods 'do something.' Neither has any support I'm aware of to show what it is they do that actually 'works.'

Anonymous said...

Hey Tom,

just want to say thank you four your hompage and your blog about stuttering!
There not many sources on the Internet that talk about stuttering in such an open way and in such a detailled fashion !!

thank you!

Tom Weidig said...

Hi Don,

"we know that early intervention works, so we can't do the study that would test whether early intervention works"

here a few thoughts:

1) Onslow et al did try to keep a control group but the parents' pressure was too great. But as far I remember, they found that the control group had less fluency gains. This could well have been from kids who would have recovered anyway.

2) We have to distinguish between reducing dysfluency, eliminating dysfluency, getting a better psychosocial adaptation, and helping parents to cope better. I am skeptical about fluency but not about psychosocial improvements.

3) A conflict of interest exist because practising clinicians are doing the trials.

4) Could we not argue that at the very least it is not hurting the kids, so we should try it?

Don T. said...

The claim that intervention shouldn't be delayed came BEFORE any controlled studies were done. They simply realized that the old Wendell Johnson 'ignore the child's stuttering' practice had no basis whatsoever.

Onslow's work has always been self-serving. Obviously, he has a personal interest in being right, not wrong. When OTHERS test his approach and support it with data, then I'll start considering the 'evidence' in favor of Lidcombe.

Anonymous said...

Our experience was that we tried using the Demands and Capacities model to treat our child who had a severe stutter for around 18 months, with no reduction in severity or the percentage of syllables stuttered. We then tried Lidcombe. I was skeptical about whether it would work too but with a lot of hard work and problem solving my child slowly became more and more fluent until he was stutter free; he has remained stutter free for many years. You will argue that it was natural recovery, but from research that looks into those children who are likely to be persistent stutterers (sex of child, family history, time post onset, type of stutters eg prolongations, blocks etc and number of iterations of repetitions), the odds of him becoming stutter free naturally were stacked against him. To me it is too much of a coincidence that when we started Lidcombe there was a very gradual reduction in the percenatge of syllables stuttered and severity ratings until he was stutter free. You will argue that 80% recover naturally anyway. That is true if you are measuring all children from onset, but not when you are considering a child 18 months after onset with other risk factors. If I was deciding again what treatment to try, I would do the same again; I would try the demands and capacities model for the first twelve months and then if my child was still stuttering, I would use Lidcombe. Although the demands and capacities model didn't work for us I'm not saying it wouldn't for another child. There may be different aetiologies therfore different approaches may work for different child.
From our first hand experience I am confident that my child became stutter free as a result of Lidcombe; the results are too much of a coincidence to say otherwise. I know you will argue otherwise. I worry that you are giving a rather one sided argument against Lidcombe that could put a parent off trying the approach with their child. If that child was one of the 20% who didnt recover naturally then I think that would be a real shame. The British Medical Journal wouldn't have published studies on the efficacy of Lidcombe if they didn't think they were pretty robust; I would say it's definitely worth trying.

Don T. said...


Publication of scientific papers is not a seal of approval. Most of what is published in scientific journals is junk. A survey of breakthrough publications in the top journals - Science and Nature - in the medical field showed that half of them had been proved wrong within five years. And that's your best case. Scientific publication is not the end of scientific discussion - it's the beginning. I've read the Lidcombe group's papers - their sample sizes are typically tiny - and thus worthless - and they rely on natural recovery to get their results. Which is more likely - that your child was one of the 80% that naturally recovers, or one of the 20% that doesn't?

Anonymous said...

You ask, 'What is more likely - that your child was one of the 80% that naturally recovers, or one of the 20% that doesn't?'

A well respected author in this field, Yairi, University of Illinois, in an article First Year of Stuttering in The Stuttering Foundation says this:

'For example, a child who is a boy, has a family history of persistent stuttering, and shows no improvement over several months, presents with a high risk for persistence.'

My child is a boy, with a family history of persistent stuttering and showed no signs of improvement over 18 months, so according to Yairi 'presents with a high risk for persistence.' This would put him as most likely to be in the 20%.

Jane said...

Franken says Lidcombe cannot be used with a child who is not fluent at single word level.

This worries me; my son was not fluent at the single word level and using strategies, we were able to work through the Lidcombe Programme successfully.

I would hope that therapists doing Lidcombe, especially those conducting research studies, would be familiar with strategies to use or could work out what to do from their knowledge of stuttering. It makes me question how well the Lidcombe Programme is being implemented in clinics in general and by those in this study.