STAR*D "controversy" ?

[Armadillos]

Quick medstudent question, after asking a FP preceptor why he always used citalopram as his first SSRI he said that he just does it because thats what they did in the STAR*D trial so figures he can't go wrong. I didn't know what the STAR*D trial was (havent done psych yet), so I started doing some googling. In addition to finding the study I also found tons of articles/websites about how flawed/biased/dishonest/etc the STAR*D trial is.

So my question is- Is STAR*D actually a controversial/dishonest trial, or are these articles just the psych equivalents of "young earthers" or climate change deniers, etc.?

---

Members don't see this ad.

 

[billypilgrim37]

So my question is- Is STAR*D actually a controversial/dishonest trial, or are these articles just the psych equivalents of "young earthers" or climate change deniers, etc.?

Nailed it. There's nothing controversial about STAR*D, except that it showed that treatment-resistant depression is difficult to treat (surprise!), and people typically ignore that bit.

 

[whopper]

I didn't know what the STAR*D trial was (havent done psych yet), so I started doing some googling. In addition to finding the study I also found tons of articles/websites about how flawed/biased/dishonest/etc the STAR*D trial is.

No study is perfect, every study has something you can nitpick, but to my knowledge I don't know of anything that puts it in the dishonest category.

I'm not saying you're wrong. Maybe you know something I ought to know, and don't let the fact that I'm an attending and you're s student make you feel you shouldn't teach me something.

But I can say this. Several people harp on everything on the Internet. That doesn't mean it's true. Scientology among several other organizations and individuals like Kevin Trudeau have presented misleading if not completely false data concerning several medical treatments and their information has been published or on a nice looking website giving a false impression to the public that it's credible data.

http://en.wikipedia.org/wiki/Kevin_Trudeau

If anything, I think the study showed the limitations of SSRIs, pointed to several generics as very good treatment options, and it had very good methodology overall. All of these things point to a good study without much detectable commercial bias. Feel free to add anything to possibly change my mind.

 

[hamstergang]

Quick medstudent question, after asking a FP preceptor why he always used citalopram as his first SSRI he said that he just does it because thats what they did in the STAR*D trial so figures he can't go wrong.

So like others have said, there doesn't seem to be anything terribly dishonest about STAR*D, but I don't get this preceptor. The study started with citalopram, but it's not like it showed it to be better than other SSRIs or even to be useful at all (there was no placebo control).

Citalopram is fine to start with, but there are certainly reasons why I choose different SSRIs to start with most of the time.

 

[shan564]

Citalopram is cheap (it's on the $4 list in my area) and has fewer drug interactions than the other two comparably-priced SSRIs, fluoxetine and paroxetine. I'll only use something other than citalopram if I have a specific reason to do so.

 

[whopper]

Here's my own personal algorithm.

Citalopram first. Exceptions are if the person already tried it or Escitalopram without success at the maximum dosage, or if they already successfully tried something else and want that something else, or if the person wants a benefit that could be provided by another antidepressant. E.g. Bupropion (they want to lose weight, stop smoking, and or have ADHD), Mirtazapine (they can't sleep well and want to gain weight) etc.

Next along the line is Sertraline.

My reasoning is you go with less interactions and the cheaper med cause in the end they're all about equally efficacious. Another advantage with Citalopram is 1-you can get to the maximum dosage in a week. Fluoxetine? 4 weeks. Sertralne? 4 weeks. 2-Comes in a tablet that's easily cut in half. You could start them on 40 mg pills, cut in half, and then tell them to go to 40 without a new prescription.

Fluoxetine? The tablets are expensive, only come in 20 mg, and the capsules, while cheap, require you to call in a new one every week if you want them to go up their dosage as you'd likely have to do given that antidepressants usually don't work well until you get to the highest dosages.

 

[splik]

citalopram is the most commonly prescribed SRI. This was the case before STAR*D, and as a result of STAR*D it has continued to be the case. citalopram started to fall out of favor a little bit as a result of the FDA black box warning about doses higher than 40mg. this was a great shame, and more recent data have suggested the risk of QT prolongation has been overblown.

Personally, I don't recall ever having initiating citalopram (though I have continued or increased it). In general I don't really prescribe SRIs for depression. I am more inclined to do so for anxiety disorders as my reading of the evidence, and the neuropsychological theories for how these drugs exert their effect, support their use for neurotic disorders more than depressive illness.

There are a lot of shoddy things about the STAR*D trial, but as most psychiatrists have been taught to accept it as gospel (along with CATIE and STEP-BD), its results have not be critically analyzed as much as they should be. In residency also, it is just taken as a given that you can believe everything in STAR*D. Personally, I think that the results themselves are pretty damning - only 1/3 of patients will get better with an SRI, and if you switch to another SRI, bupropion, add buspirone, they are all as bad as each other, less than 1/4 will get better with that, and with 3rd line agents, only 1/6 will get better. an all this is not even taking into consideration the significant placebo effect here which is probably doing most of the work, and explaining most of why remission decreases with each additional phase of the study (you are selecting the non-placebo responders)

 

[Armadillos]

Personally, I think that the results themselves are pretty damning - only 1/3 of patients will get better with an SRI, and if you switch to another SRI, bupropion, add buspirone, they are all as bad as each other, less than 1/4 will get better with that, and with 3rd line agents, only 1/6 will get better

I guess it depends on how you look at it, considering the low cost and relatively tolerable side effects of SSRIs it seems pretty amazing to me. Especially if you compare it to the cost benefit/ratios of other interventions in other medical fields.

 

[FrasierMD]

I am more inclined to do so for anxiety disorders as my reading of the evidence, and the neuropsychological theories for how these drugs exert their effect, support their use for neurotic disorders more than depressive illness.

Hey splik, can you elaborate on this? There is good data for anxiety and depressive disorders from the evidence i've read.

 

[MacDonaldTriad]

I participated in STAR*D and it was NIMH, not industry sponsored. Trust me no one profited, it was hard to break even given the costs and the pay schedule. There were a lot of sites and our input was asked for. It was never intended to be a drug vs drug study. The attempt was to give us guidance as to what to do when plan A fails, and what to do when plan B fails. A cynic may say that we proved our antidepressants were not very effective. Unfortunately, the augmentation arms were more similar to each other than different in outcome. It showed that meds alone are often not enough. Nothing shocking really, but all study designs can be torn apart.

Prime the pump, be optimistic, keep trying things methodically, and take credit when something finally works (we never know when remissions are a result of treatment or just a natural course, so why reinforce helplessness if we don’t really know). Rain dances always work when you keep dancing until it rains.

 

[whopper]

Regression to the mean.

http://en.wikipedia.org/wiki/Regression_toward_the_mean

Some patients just get better because they would've gotten better anyway no matter what you did.

So while you tried them on 4 meds, and they happen to get better just a few days to weeks after med #4 was started, that doesn't mean med #4 caused the improvement.

It's weird. My former PD is a statistical genius and knows it better than some Ph.D. mathematicians. You tell him something going on and he thinks about things in terms of charts and statistics. It's simply regression to the mean, as he pointed out, but most docs, including myself, don't think like he does.

 

[whopper]

ut as most psychiatrists have been taught to accept it as gospel (along with CATIE and STEP-BD), its results have not be critically analyzed as much as they should be. In

Agree, but at the same time, IMHO, a good psychiatrist has read these studies cause most of the ones that haven't read these studies are the ones that don't know what they're doing.

Those studies attempted to answer fundamental and serious clinically practical questions.

I've written this in another thread. Despite that I believed the STEP-BD is a very well done study, I don't agree with it. I've had patients with bipolar depression, get better on an antidepressant, then because of STEP-BD, I took them off of it and they all got worse within a time-frame where you'd expect it to be from the med change. Then when we put them back on it, they got better.

I have, however, seen pretty much everything from CATIE and STAR*D stand up in real life clinical work.

Most of the docs I know who read CATIE, STEP-BD, and STAR*D actually have an evidenced-based frame of reference when they talk about which meds to use in which situations, where as the ones that haven't read it don't. Reading them isn't the mark of a good psychiatrist, but I still think they were the some of the best studies done in our field in the last 10 years. You can't simply base your opinion on a study, even a well done one, but a well done one deserves it's place for consideration.