Chances Cymbalta Will Work and How Cymbalta Compares with Other Meds:  As is typical with most antidepressants, the actual data from the clinical trials aren't in the PI sheet.  Instead you get the weasel wording, "Cymbalta demonstrated superiority over placebo as measured by improvement in the 17-item Hamilton Depression Rating Scale (HAMD-17) total score."  No indication as to how much improvement over placebo or if their depression stayed away.

However, I was able to find some hard data from a few studies and trials.  Note that these studies were all sponsored by Lilly in some way, and that tends to make the results more favorable for the med in question (on average 3.6 times more likely, according to a Yale study).  Eventually some independent studies will surface once Cymbalta has been around for awhile.  Also note that some of these studies had people taking dosages way over the current recommended maximum dosage of 60mg a day.  The PI sheet has the usually weasel wording that "There is no evidence that doses greater than 60 mg/day confer any additional benefits." 

Cymbalta vs. Prozac vs. placebo  120mg a day of Cymbalta improved the quality of life of 64% of the people taking it and the depression stayed away for 56%.  Compared to working for 52% of the people taking Prozac (fluoxetine) and it kept working for 30% of them, while 48% of the people taking a placebo got results and 32% of the people taking the miracle drug placebo stayed happy.

Cymbalta vs. placebo  60mg a day of Cymbalta improved the mood of 65% of the people taking it and kept the spirits up for 43% of them.  In this study placebo was able to score 42% and 28% on the effect and remission ratings.

Cymbalta vs. Paxil vs. Placebo 40mg a day of Cymbalta resulted in a net Hamilton Depression (HAMD-17) Rating change of 4.66 points.  80mg a day resulted in a HAMD-17 change of 5.86 points.  20mg a day of Paxil (paroxetine) resulted in a change of 3.56 points.  What do those numbers mean?  Enough of a change to write "significant" on your PI sheet when compared to what the placebo did.  At least for Cymbalta.  Had this been a clinical trial for Paxil (paroxetine) they would have buried these results and ramped up the dosage.  Of course 20mg of Paxil (paroxetine) is the minimum effective dosage for depression, while 80mg of Cymbalta (duloxetine hydrochloride) is above the current maximum dosage.  Hmmmm.

Cymbalta vs. Effexor vs. Placebo Unfortunately this is just a meta-analysis, where someone just pooled data from a buttload of existing randomized, placebo controlled studies.  The results of pure statistical analysis are that Effexor (venlafaxine) is somewhat more effective, but only a head-to-head grudge match will determine that for sure.  Otherwise they "comparably" sucky from a side effect standpoint, but that has little meaning in the real world.  What is an equivalent side effect to a researcher isn't equivalent to someone actually taking the medication.  Still, given all that, this Canadian study found that as the Effexor is cheaper, you may as well take the Effexor.  Just so long as you don't have to quit cold turkey, eh?

Long-term use of Cymbalta  Using Cymbalta in an open-label study (i.e. everyone knew what they were taking) for a year, over 500 patients took 80mg or 120mg a day.  This study showed that 70% of the people taking it felt better after two weeks, and after a year about 82% people stayed out of Depressionland.

Cymbalta vs. desipramine  Not a test of efficacy, but a measure of potency.  Cymbalta rates as potent an inhibitor of norepinephrine reuptake as desipramine, which is pretty potent.

Cymbalta to treat Depressed African Americans  On the plus side, at least someone did a study to ascertain if there could be a difference.  Alas, there was a ten to one ratio of white to black folks in this double-blind study.  Still, it was better than nothing in figuring out that Cymbalta works just as well or sucks equally for both groups.  Given the size of the groups taking Cymbalta (590 Caucasians to 59 African Americans) there's not enough statistical information to determine, for the latter part, if the side effects were a bit less bad for the African American community, or the Brothers and Sisters just complained less.  Oddly enough there was a pain scale used, so I don't know if neuropathic pain was tested as well, or if someone was really on the ball, and included the far too often physical effects of major depressive disorder.

See the page on a drugs' efficacy for an explanation of the tests used to evaluate if a medication is any good or not.

So you're stuck with the generic version regarding all Multiple Reuptake Inhibitors, Cymbalta (duloxetine hydrochloride) doesn't make you produce more serotonin and norepinephrine, rather it makes your neurons soak for a longer period of time in the serotonin and norepinephrine you already produce.  

But is that the same thing?
 
That depends on the person and the sensitivity of your 5Ht and alpha  1 & 2 receptors.  Sometimes it's the same effect, sometimes not.  When not it could be sub-par (to the point of being useless) or too much.

For depression they are two of the big three neurotransmitters, dopamine.  My wild-ass guess / rule of thumb puts 80% of depression issues somewhere in the realm of serotonin, norepinephrine and/or dopamine.  According to the PI sheet Cymbalta (duloxetine hydrochloride) also does a negligible amount of dopamine reuptake inhibition, so like Effexor (venlafaxine) you might be able to get a hit on all three.

Not all depression issues are solved by these particular neurotransmitters.  Presuming it's a biochemical problem in the first place, and not purely a psychological / situational issue (i.e. talking it out is only what you need to do and not take meds), some biochemical depression issues are in the realm of MAO, glutamate, GABA, opiods and/or voltage channels.

Remember, even if you do need meds, you still need to talk it out with a therapist and a support group.

As far as neuropathic pain goes, it's some combination of rebuilding nerve tissue and the additional marinating of your neurons in those neurotransmitters providing a form of pain relief.  I still need time to do some more research and give you some real answers backed up by real studies.  I still suck.

Cymbalta's Half-Life: An average of 12 hours in a range of 8-17 hours. 

Days to Reach a Steady State: Three days.