Comments: Researcher Assails Anti-Depressants
Dear FS,
This is an interesting article. Thank you for taking your time to do these postings so early this morning :)
In my opinion, the PURE remission rates as a primary outcome in depression treatment RCTs, unadjusted for possible confounding factors is not valid. After people drop out of RCTs, the groups are no longer balanced. Propensity for non-response weights NEED to be used. I know this is GEEK talk. But this is very important. It means simply, that most of the time, the characteristics of those who do and do not "comply" with the treatment protocol or drop out are very different from those who were initially randomized AND those who were initially approached for consent to be in a study. AND OFTEN TIMES people in the different arms of the study drop out for different reasons, thus, these groups are also imbalanced and need statistical bandaids all over the place. Ask yourself.. what is the number of actual data points they collected. In some of the biggest trials this is VERY hidden. It is as high as 56% of data missing... where do they dig up the missing data? Ask me sometime.
In addition, the TYPE of treatment (modality, cost), the SIDE EFFECT PROFILE (mild to deadly), the NUMBER of people who would potentially benefit from this treatment globally (generalizability),and the DISSEMINATABILITY of the depression treatments NEED to be taken into account in any evaluation of a remission rate.
The Star*D and Pathways, and Impact and PEARLS and Rush's disaster, and many other studies are NOT doing this. It is sorta very sub par. These ideas of mine, which I have taught to my students for the past 18 years are going to become increasingly important as the NIMH representatives on several committees are pressuring me strongly to write a new kind of evaluation for their granting decisions. They are also forcing me (like they can REALLY do that) to write a book for THEM. No way, guys.. If I do a book, IT IS FOR ME.
I hope bloggers and commenters will feel free to give me input or their suggestions for how to evaluate research grants on treatment for depression.
I think the traditional methods of politics (think - who you know and who you ...), power and sample size estimates (totally manipulatable) , and fake pilot data are numbered.
I just don't know if I am the right person for the job, as I am so psychiatrically vulnerable and so angry at psychiatry and my associates right now.
Any suggestions, thoughts, positive wishes will be accepted :)
Have a great weekend.
Peace and Love,
BK
Posted by Dr. Black Kitty at April 27, 2007 06:07 AM
I appreciate the t-shirt ready slogan "The mission for the condition is remission".
I found this article interesting.
"Treating the insomnia pharmacologically or behaviorally can improve outcomes in depression," said Christopher L. Drake, Ph.D., of the Henry Ford Hospital Sleep Disorders and Research Center in Detroit.
One of things that I am grateful for is that I found a psychiatrist that actively treats my insomnia and my crazy. It's made a world of difference.
Posted by Chloe at April 27, 2007 09:15 PM
I just know there can't be a placebo groups to compare any drug to. All psych meds have a noticable physical effect, be it dry mouth, constipation, or the munchies, you know when you are on something or not. Then you have the question of the placebo control group, how many want to beleive they are being given a drug? How many have been convinced they are getting a powerful new drug? Are the optimists and pessimist in the two groups in equal numbers? Ya that is real science there.
Posted by Mark(p.s.2) at April 27, 2007 11:54 PM
That's a good article Chloe.
I find this interesting: "Poor sleep quality is also predictive of suicidal behavior."
Does anyone agree with that? that it's predictive? That is alarming.
Posted by Stephany at April 28, 2007 03:20 AM
Dr. BK,
My hat's off to you. I wouldn't even begin to know how to control for all of the potential confounding variables in studies dealing with mental health. If a study found that a particular Rx lessened the depression of a significant % of people in the study, I would always wonder if people are getting better because of the Rx or maybe they are better because they were a part of a study and were given some hope. Conversely, maybe some people get worse because of problems at home (philandering spouse, kids leaving home for college, whatever) that have nothing to do with the medication. I just don't know how researchers can control for these sorts of things.
While I'm quick to blame the worsening of my own depression on psych meds(perhaps some of you have noticed this? lol), I would imagine that having a crazy therapist who wrote me poetry and left odd little gifts on my doorstep and a psychiatrist who also had/has some serious mental problems likely helped push me straight over the edge. The ambience of the psych hospital didn't exactly lift my spirits, either. Maybe it was the medication that worsened my problems, maybe it was having mental health professionals with serious mental problems...I suspect it was a combination of a lot of things. Just as I suspect it is a combination of a lot of things that have helped me get better. I give the personal example for a couple of reasons. 1. I was officially diagnosed with cluster B traits so I would be remiss if I didn't somehow relate this all back to me. 2. So much of a person's daily life affects their depression. How does one control for daily life stuff when examining whether a treatment worked or not?
Posted by Lisa at April 28, 2007 07:04 AM
Dear Lisa,
I was also failed by several different psychiatrists. I could not get help here in the U.S. I had to take matters into my own hands. I feel that people have to have the opportunities to try many different treatments which appeal to them at a gut level.
We are not children, or mentally challenged, rather, consumers of mental health services are in my experience generally above average in intelligence. If this is the case, then, for people who choose to be a partner in their OWN care, all treatments, both those suggested by the consumer and the provider of care, should be discussed, until a consensus can be reached.
If you are told you are a difficult patient, because YOU want a say in YOUR treatment. FIND A NEW PROVIDER. Of course, this is not always possible, especially for people in rural areas, small towns, on public assistance or have no financial resources..
How do we control for all the variables that can confound a given depression treatment trial? We have enough participants, that ALLOWS us to do so. The rule of thumb for folks to remember is you need approximately 10 participants for every factor you want to control for. For example, if you want to control for age (kid - adult), gender, ethnicity and baseline severity level, you now need at least 40 people for these tests. Now you need to test the treatment and its time interaction, that adds 30 more folks.This wold imply that the researcher would really need at least 71 people to complete the trial. How many factors should you control for? Life stressors, certainly, attitudes toward treatments? number of prior treatments... it is an infinite list.. really.
This is why we need GLOBAL studies of the effectiveness of our "efficacious" treatments..
These are rules of thumb (its not always true.. don't snark at me)... Also due to manipulation of pilot data, big effect sizes are used to justify numbers smaller than what I quoted above...
Peace and Love,
Dr. BK
Posted by Dr. Black Kitty at April 28, 2007 07:29 PM
I appreciate your response. It took a while but I finally got to the point where I quit accepting crappy treatment. There are practioners out there who do respect patients' opinions. My 3rd therapist has a PhD in Marriage and Family Therapy. She's not hung up on the DSM, and for me it's proved to be a good match. I don't need any more labels. I don't need any more diagnoses. I just needed somebody to help me find my way through this mess. Medication isn't for everyone, and it sure would be nice if organizations like NAMI would recognize that.
Posted by Lisa at May 2, 2007 10:50 AM