Tuesday, October 06, 2009

PEVOS: finally!

Finally, finally, finally. We (or at least I) have been waiting for years for PEVOS to complete their pilot study. And I have blogged here and here. As usual completely ignored by the English speaking world as is the large outcome study of the Kassel Stuttering Therapy with a sample size of 100s and long-term data.

What is PEVOS about? PEVOS looks at the outcome of different therapy approaches to see which ones are working. They have now published the pilot study data nearly 10 years later! See the PEVOS article in the German stuttering magazine Kieselstein. (Thanks to Norbert for the tip.) If someone wants to translate the article or make a summary of the key findings, send the translation or summary to me and I will post it. I don't have the time to also play translator of German.

And as usual, the same old story. 99 participated from different therapies and therapists, and only from 66 full data over the 2-year observation period are available despite massive pestering. The authors make a lot of correct and wise statements, for example early childhood treatment is much more difficult to measure and adult outcome study should be done first, and they are aware that recording of the phone is not ideal (I suggest that they should look into asking future participants to accept anonymous calls and recordings.)but the best they have. Further, they did not look at difference between therapies due to low sample size.

The graph shows the key finding on the pilot study. U1 is before therapy, U2 just after, and U4 after 2 years. As we have seen many times, fluency is very good immediately after therapy and gradually gets worse again as more and more people relapse. U4 still shows a treatment effectfrom 15% to 11%, and they claim that this shows treatment effect. Sorry, but I don't buy this statement. First of all, you only have 66 out of the 100 and it is likely that some downward bias exists. And then, you have the issue of the recordings themselves: the first recording is most stressful and a bit more than normal stuttering is expected, and the last recording is the least stressful and a bit less than normal stuttering is expected. So for me, this shows the depressing realization that in terms of fluency very very little has change in these 66 patients! :-( That's the truth we have to face up to. It is like diets or drug addiction...  But in terms of life quality I am convinced the 66 patients are better off than before therapy. So not really time wasted in some sense.

Unfortunately, they cannot find the funding to do the outcome study on a large scale. It is a scandal really. The German government spends billions and billions on research, but virtually none into stuttering. Of course, you can also blame this inability to get funding on those who want to get funding. On their political weakness. And especially on the German national stuttering association for a failure to lobby for such research.

But let me thank them in the name of us all for the hard work they have put into this. I wish it had been done much much faster, and I am sure they all agree with me that it was unacceptably long. I hope that they will find funding.


Dave Rowley said...

Thanks for highlighting this study. I wonder what the one third from whom they couldn't get data were like? Were they PWS who didn't improve and therefore did not want to participate any more? If so, then the final improvement scores are suspect and sadly the experiment shows that the therapy was not effective.

Tom Weidig said...

I will post a German translation.

Norbert @ BSA said...

I think the devil's in the detail. AS far as I can make out, the pilot study merely ascertained that in principle this is a workable way of proceeding if one were do do a large-scale study. Other than that it seemed that for the 2/3 of those who completed the data set, therapy in general has been a useful and beneficial exercise. We kind of knew the latter fact. Would be interesting to hear Tom comment on the full report (in German) which I believe is available from the BV? I haven't seen it yet.

Tom Weidig said...

Wait Norbert! First, the study does give us clues on the overall results after all we have 66 patients that is a big sample. Second, they did not want to look at between therapy differences, because they rightly say that the sample is too small. But I would not be surprised if they could make more qualitative or anectodal statements on which is better than another in private. Third, I am not convinced about positive results. They found that the 66% of patients improved marginally, and this could be due to the fact that they stuttered a bit more before treatment because the phone call was very stressful and after four calls they were a bit more relaxed. AND, we are missing 34% which might as Dave suggested not have benefited as much or got worse. SO overall, there is in my view no good evidence of good benefits for the overall population of people who attended treatment!!!! That is shocking!

Norbert @ BSA said...

I am sure there could be anecdotal statements about therapies but they'd not be valid or meaningful. I could certainly make anecdotal statements about therapy right now. But you know me, I won't! :-)

Re improvement - I have yet to hear of anyone who turned out to be worse off because of therapy. That's therapy with a capital 'T' from a qualified practitioner rather than a quack, of course.

I think there's enough evidence other than PEVOS that any therapy, i.e. looking at one's stammer, and one's own attitudes and emotions around stammering, on the whole is a beneficial exercise. Benefit might be marginal, as you say, and not long-lasting and I am not making any claims other than that.

I'm also not certain that the phone calls would have got any easier. They were too few, and with too much of a distance between them, to have any kind of habituation effect that would make the latter ones easier, I'd say.

Still, that leaves the 34% that could not be reached but you're just as entitled to assume that therapy was so wonderfully successful that stammering was no issue for them any more as to say that they were damaged by therapy. We just don't know. I personally don't think that 34% is an unreasonable drop out. I believe I read somewhere that about 15% of any personal address database will become invalid each year.

Tom Weidig said...

I agree Norbert that

1) treatment rarely has negative benefits, and if, mostly short-term.

2) treatment is nearly always increasing the quality of life and reducing the handicap of stuttering.

3) This is most true for first-timers with decreasing impact for further treatment.

4) Treatment can also be in form of self-help group, attendance of a stuttering conference, and reading about stuttering

5) I still maintain that there is little evidence that any treatment delivers good long-term fluency gain in a good number of patients. The gain are in life quality and reduction of handicap.