Thursday, November 25, 2010

Pagoclone rumours

Here are good old rumours:
The Pagoclone study ended and they did not continue the study. They will be making a decision in December, I think, whether there was enough benefit to a sub-group of the trial study (meaning there was not a benefit to everyone in the study but definite benefit to some)  to justify continuing on with the drug.  My trial administrator in Atlanta said he saw definite benefit to a sub-group and he is still positive on the drug.

17 comments:

Gustaf said...

Good news. Long live Pagoclone. It's hardly surprising if a treatment works well for some but not others. To blame such discrepancies on the placebo effect is oversimplifying. I think it demonstrates that there are indeed more than one variant of stuttering, and that we can't expect to have one drug that works for everyone.

Dave Rowley said...

Any idea what identifies the sub-group, what characteristics do they have? Not just a response to Pagoclone, I hope?

Tom Weidig said...

WAIT, WAIT, Gustaf!

This is just one center. Why do you think they have multiple centers??

Again, placebo and natural fluctuations in stuttering should be considerable, and only a random control design can control for them.

I need to see the complete data set INCLUDING the placebo group. I would love to see the natural fluctuation of that group.

Pago.clone@gmail.com said...

I participated in the study. I feel my stutter reduced around 30-40 percent in frequency and intensity.

Why do I believe these were caused by pagoclone and are not explained by natural fluctuation (noise) or some other factors? Well, because I've stuttered long enough to be able to tell the difference, for my particular case. But I do agree with Tom that we need the full data before we can make any valid inferences.

However, anecdotal evidence, like mine, should not be dismissed offhandedly. Gustaf's basic point is right. If pagoclone works for some people and not others, even if the group it works for is relatively small, then this is an important finding. Even if we haven't specified the "causes" (or the "complex of interactive causes") of stuttering, then knowing that a particular drug reduces stuttering (without side effects) is significant. The next step should be identifying the characteristics of those for whom pagoclone was relatively successful. So I do think there is good scientific reason to pursue the pagoclone study (although there might not be the sufficient profit motive).

On the other hand, in addition to some of the concerns Tom has raised, there are measurement errors to take into account. The true natural fluctuation might be biased downward.

Tom Weidig said...

Hi Pago.clone,

do you know whether you were in the Pagoclone group or control group?

Are u still taking P?

You have noticed a correlation. I have also been quite fluent over the last months, and don't know why. but i was not on P.

I am not saying there is no subgroup effect, but i find it difficult to confirm it by just hearing a few individual cases.

what do you mean by biased downward?

pago.clone@gmail.com said...

I am not still on pagoclone. I stopped a month ago, after having been taking pagoclone in the "open label" phase for about a year.

I have no idea if I was in the treatment or control group during the off-label phase. I decided to stop speculating on that a while back. I am evaluating pagoclone's effect on me on the basis of the open label phase, where I was on pagoclone (1.2mg/day) for a year.

As I stated above, of course natural fluctuations could be the factor here. But throughout my entire life, the "period of fluctuation waves" in my case has always been a two or three months at most. In this case I felt my stutter had decreased consistently for about 10 months, hence my thinking that the effect was above and beyond any random change. I could be wrong of course. But there's another factor too. I feel that my stutter has increased I got off pagoclone (more precisely, about two weeks after I got off it). Is that random fluctuation too? Very well could be, but based my 10 month improvement, and my generally stable stuttering patterns, I don't think it is ALL random. I should say I've been in communication with about 15 other pagoclone study participants and I think most of them did not experience any substantial benefits. No one experienced side effects either (no one that I communicated with). And also, my claimed improvement was NOT that signifcant. I said 30-50% earlier. It may have been 20-50%. It's hard to assess these things.

I actually meant biased upward (or downward), just biased estimates, as well as imprecise estimates too boot (high standard errors). The reason I think there was measurement error (and it could bias the estimates in either direction) for the measurement of stuttering AND the various anxiety controls is they ask you how you're feeling on the particular day you go to the clinic, and that's subject to obvious errors. If for example people overestimated their improvement or underestimated it, then this would make the results imprecise. If participants misreported anxiety levels or give wrong answers to any of the 50+ questions on the questionnaire, then this would bias the estimates, but in an unclear direction. I think there probably was some measurement error in my case, at least.

Anyway, the bottom line is that if some delusional pagoclone participants really believe that there is a good chance their improvement was CAUSED by pagoclone, then for scientific purposes, it would be useful to understand what "type" of stutters these particular individuals are.

More generally, I think stuttering research needs more large N data on stutterers. With large datasets we can run cluster analyses to see if there are indeed "types of stutters." If there are, and I suspect there are, then treatments should be geared toward these clusters. This might resolve some of the tensions between behavioral speech therapists and neurologists/psychiatrists. Maybe people are just talking past each other. At least some of the smart researchers might be talking past one another. Most speech therapists and stuttering researchers are more or less clueless and have no idea how to produce sound research on stuttering. If they do know how, then for some strange reason they do not want to do good research. I know I'm not giving examples here. this part of this post is just a rant :) Anyway, the ones who choose not do good research are encouraged to do continue in their ways by the organizational/professinoal demands of their discipline (to keep on pumping out papers for publication). The "risk" is too high to explore more rigorous research. It also has to do with the fact that speech pathology probably attracts lots of less intelligent people, I'm sorry to say. People who gravitate toward the area may just be bad at science in general, and there are some good reasons to think that's the case. We need more research from medical schools, and we need more statistics courses taught in medical schools... AND CLUSTER ANALYSIS!

pago.clone@gmail.com said...

I typed up a response to this but I don't think it went through, perhaps because it was too long. Here's a shortened version.

I am not still taking pagoclone. My participation ended about a month ago (late october).

I do not know if I was in the treatment or control group. I speculated on this extensively! But I've concluded it's of no use. I'm just not sure.

My evaluation of pagoclone's effect/non-effect on me is based on the "open label" phase, which lasted about 12 months, and for which I'm sure I was pagoclone (1.2mg/day).

Of course, I noticed a correlation, but I still think there was some causation. This is because I've been stuttering since I could speak (mild to moderate) and while natural fluctuations do occur (within a day and across monhts), they have a pattern to them. Even error terms have patterns and distributions. When I would experience improvement over a sustained period, it would usually last around 3 months. While on pagoclone, I felt improvement for 10 or 11 months straight.

Of course, this could have been a placebo effect, or at least some of it could have been. But note that after a couple of months on pagoclone, I basically stopping thinking about being on it. I would just routinely take my pills in the morning and evening, and didn't think about pagoclone much (I did for the first couple months, but not the latter). So I don't think all of the effect can be explained as placebo.

I should say I'm in communication with about 15 other pagocloen participants, and most of them did not experience benefit. So I and a few others might be the anomaly, and this should obviously raise doubts like the ones many on here express. All I can do is give my honest account of my experience and offer some details about my personal speech patterns which lead me to think there was at least some substantial and positive speech effect CAUSED by pagoclone.

I actually meant biased downward or upward. On most pagoclone visits they make you fill out an extensive questionnaire and ask you if your stutter improved/worsened since the last visit. All of this is prone to measurement error. The answer categories are often not sufficiently exhaustive, and I always felt my answers were slightly inaccurate. Whether they overstated or understated things with regard to the anxiety questions might bias estimates and for the stuttering questions might make estimates imprecise (large standard errors). Given the questionnaires, I think there was some measurement error.

pago.clone@gmail.com said...

I typed up a response to this but I don't think it went through, perhaps because it was too long. Here's a shortened version.

I am not still taking pagoclone. My participation ended about a month ago (late october).

I do not know if I was in the treatment or control group. I speculated on this extensively! But I've concluded it's of no use. I'm just not sure.

My evaluation of pagoclone's effect/non-effect on me is based on the "open label" phase, which lasted about 12 months, and for which I'm sure I was pagoclone (1.2mg/day).

Of course, I noticed a correlation, but I still think there was some causation. This is because I've been stuttering since I could speak (mild to moderate) and while natural fluctuations do occur (within a day and across monhts), they have a pattern to them. Even error terms have patterns and distributions. When I would experience improvement over a sustained period, it would usually last around 3 months. While on pagoclone, I felt improvement for 10 or 11 months straight.

Of course, this could have been a placebo effect, or at least some of it could have been. But note that after a couple of months on pagoclone, I basically stopping thinking about being on it. I would just routinely take my pills in the morning and evening, and didn't think about pagoclone much (I did for the first couple months, but not the latter). So I don't think all of the effect can be explained as placebo.

I should say I'm in communication with about 15 other pagocloen participants, and most of them did not experience benefit. So I and a few others might be the anomaly, and this should obviously raise doubts like the ones many on here express. All I can do is give my honest account of my experience and offer some details about my personal speech patterns which lead me to think there was at least some substantial and positive speech effect CAUSED by pagoclone.

(cont.)

pago.clone@gmail.com said...

(cont.)

I actually meant biased downward or upward. On most pagoclone visits they make you fill out an extensive questionnaire and ask you if your stutter improved/worsened since the last visit. All of this is prone to measurement error. The answer categories are often not sufficiently exhaustive, and I always felt my answers were slightly inaccurate. Whether they overstated or understated things with regard to the anxiety questions might bias estimates and for the stuttering questions might make estimates imprecise (large standard errors). Given the questionnaires, I think there was some measurement error.

What needs to be done with the pagoclone (for scientific, not profit, motives, although they're usually aligned of course) is some cluster analysis to see if "types of stutterers" can be identified. It would be interesting if discernible patterns could be identified. We really need more large N data on stutterers. If "types" were observed, then this might bridge some of the gaps between behavioral speech therapists and neurologists/psychiatrists. It might be that behavioral treatments are very useful some people but useless for others, and it may be that pagoclone is helpful for some and not for others.

pago.clone@gmail.com said...

Testing? Seems my posts aren't going through.

pago.clone@gmail.com said...

Okay, they must have been too long.

Nope, I'm not sure if I was in the control or treatment group. I'm basing my comment on the 12-month open label phase through which i was on pagoclone.

i know my natural fluctuations pretty well. When my stutter does improve, it's usually for 3 months or so, and then it gets worse or levels off. During open label it was better consistently. That's why I think there's reason to believe pagoclone was CAUSING the effect.

i meant measurement bias upward or downward. People probably misreport answers to the questionnaires they give us. This makes estimates imprecise (for anxiety Qs) and bias (for stuttering Qs).

pago.clone@gmail.com said...

I think it would be very interesting to run some cluster analyses on the data we do have on pagoclone and from other studies. If there are "types of stutterers" then maybe we can start developing treatments catered to particular types. I think there's much promise in this. Obviously it's not a new idea, but are there any studies that focus specifically on identifying "clusters" of stutterers. The best data would of course be neurological, psychological, and involve life histories, but even without all that data, some clusters could still be identified.

Kerri Hogue said...

I was in the Pagoclone study for about 4 years. When I first started the study it was double-blind and I'm reasonably sure I was on the placebo. I felt no effects whatsoever and my speech did not improve. After several weeks I moved to the open label stage of the study. I did not expect Pagoclone to work as I had been self-medicating for several years without success. I had completely given up when a friend told me about the study. After I started taking Pagoclone my fluency improved noticeably, and it kept improving throughout the four years or so I took the drug. I got to the point where at times I even forgot I stutter. Everything was so much easier. Also, when I took Pagoclone, if I didn't eat enough first I would get a nice head buzz and would eventually feel tired. I didn't experience this effect with the placebo. I'd like to add something else. Without a doubt, my speaking is now WORLDS better than it was when I began the study. (Though it's not as good as when I was taking the drug.) So I don't know if Pagoclone had some permanent effect after taking it all those years, or maybe my brain finally figured things out on its own.

Giving up on this drug would be foolish.

pago.clone@gmail.com said...

Hey Kerri,

In many ways I feel like you do. If could you would like, send me an email at pago.clone@gmail.com. I wonder if we can get more people whom pagoclone did work for and perhaps write a letter to someone who could help.

It really depends on how much people there are who feel like you and me. If it's just us, there's probably little reason for them to continue the study (and with justification). But I suspect there are more, and we might be effective with some collective action. Email me if you would like. pago.clone@gmail.com

Scott said...

I was also in the trial. Pagoclone had the same effect on me. I felt like it improved my speech by around 40% or so. I have written the CEO of Endo twice asking for a continuation of the study or at least usage of the drug on a compassionate use.

pago.clone@gmail.com said...

Hey Scott, please send me an email at pago.clone@gmail.com. I'm preparing a petition.

Roy said...

Hello,
I was interviewed for the second phase, but was not asked to join due to a somewhat normal fluency interview. I attribute that to a depression drug which I was on because of the death of my son. I was taken "Sertraline" 50MG Tablet form. I noticed while I was on it my speech improved tremendously and I also noticed I was much more social when i was in group settings. Please feel free to e-mail me with any questions at Raptureairlines@aol.com