In the past ten days, a few different blog entries have been written calling our attention to a meta-analysis that suggests perhaps Cymbalta (duloxotine) — a newer antidepressant — isn’t as good as the company claims it is for the physical symptoms of pain associated with depression.
But, as regular readers of World of Psychology know, a single study does not a conclusion make. Not even a meta-analysis.
The new meta-analysis, conducted by Glen Spielmans, included only five of the published studies that have examined the question of pain, depression and duloxetine (Cymbalta). Why only five, when there have been dozens of studies that have been published examining depression and duloxetine with a pain measure in them?
One key to a good meta-analysis is how is it constructed. An author can choose any set of criteria to include a study in her or his meta-analysis. The criteria can be biased to include only a certain subset of studies that will produce the desired result in the meta-analysis. So a “meta analysis” in and of itself isn’t an indicator of any better science (or quality) than a single study, especially a poorly constructed meta analysis.
I’m not saying this is the case in this particular study. But the new meta-analysis does contradict another meta analysis, which found that some people who took Cymbalta did have reduced pain.
So who’s meta-analysis are we supposed to believe?
Well, until questions like the above are answered, both. Clearly there is a fair amount of evidence that Cymbalta has a positive effect in helping reduce people’s physical pain symptoms (we readily found over a dozen articles in Medline supporting this conclusion). Perhaps that effect isn’t as robust as initially thought (which seems to be so often the case with clinical effects of pharmaceutical drugs), but it seems to be an effect nonetheless (which even the current meta analysis acknowledges).
In the meantime, we’ll let one of the previous meta-analysis findings’ speak for itself:
Improvements in quality of life were significantly greater for both men (p = .006) and women (p = .001) receiving duloxetine than placebo and showed no significant difference by gender.
At the end of the day, isn’t this what’s important about an antidepressant prescribed for depression? That it helps a person relieve their depressive symptoms and improve their quality of life?
(2 votes, average: 5 out of 5)
Loading ...You can leave a response, or trackback from your own site.
Links to This Article
Bookmarks Tagged Blog (12/29/2007)
Cymbalta » Blog Archive » Cymbalta (1/21/2008)
5 Comments to
“Cymbalta Still Good For Pain?”
I have read the meta-analysis. I also did my own quick Medline search and there are not “dozens” of trials using Cymbalta for pain in depressed patients. There are not many actual clinical trials.
There might have been a couple studies not included in the meta-analysis — I wonder if they might have been published after the meta-analysis was in-press? Also, the meta-analysis noted that two of Lilly’s own studies did not provide sufficient statistical information to be included in the meta-analysis. If they didn’t provide enough information, could that have been because the data were not in favor of Cymbalta?
The Lilly meta-analysis cited in this post did not examine Cymbalta vs. placebo in treating pain in depressed patients, so I don’t think it is a comparable analysis to the meta-analysis that actually examined pain in depression. They are two separate analyses that examined different measures. Pain measures and quality of life measures are not synonymous.
When is a meta-analysis not a meta-analysis might be a good title for this topic! Sadly I have to agree that studies that don’t use specific pain outcome measures (and not just visual analogue, but also measures of function, reduction of distress, depression etc), and that don’t include a GOOD placebo arm just don’t cut it for me. And once again, medication on its own doesn’t do a lot of good if the fear and anxiety associated with pain aren’t also addressed.
Speaking as someone who has had chronic pain for the last 25 years, Cymbalta does help. At first. I’ve been on it for about 15 months now at 90mg daily. It does help, I went from a daily pain avg of 7 to a 5, but it’s starting to wear off. In addition I’ve noticed an alarming side effect that is not mentioned anywhere in the literature, but can be found on forums and bulletin boards online. I’m craving alcohol. And I don’t mean like one would crave a steak or ice cream, I mean like I can’t look at a commercial or ad for it, I can’t eat out at a place that serves it. I become emotionally unstable when challenged. It’s scary and I think I’d rather be depressed and in pain.
Are you experieincing symptoms of hypomania?
I was on this drug for quite a while. That’s when things spiraled out of control. I had an uncotrolable compulsion to drink alcohol. I thought I was an alcoholic, but it didn’t make any sense. I tried everything to stop drinking, and I mean everything! I have lost everything in my life because of this. I never new it was cymbalta until I lost my job thus losing my insurance. I had to stop taking this drug cause I could not afford it. To my amazement as soon as I stopped the last dose of it ALL compulsion for alcohol stopped. It’s been many months now dealing with the withdrawl symptoms (which are brutal). With out a doubt this drug made me drink alcohol.
Join the Conversation! Post a Comment:
Last reviewed: By John M. Grohol, Psy.D. on 2 Jan 2008
