As you probably know, the United States Food and Drug Administration (FDA) has placed a warning on antidepressants when used in children and adolescents, indicating that these medications may
increase suicidal thoughts and actions in a small number of patients. But what you may not know is that the FDA is actively considering a similar warning for adults.
Meanwhile, the maker of Paxil (paroxetine; GlaxoSmithKline) has sent a letter to doctors indicating that GSK's own numbers suggest possible risk of increased suicidal action in adults taking Paxil. Although viewed in one way these numbers were not "statistically significant", the increased risk also appeared in patients
not taking Paxil for depression. This fact was mentioned specifically in their letter, as though GSK itself regards it meaningful -- that appearance of suicidality in such patients suggests that antidepressants can actually
cause this shift in thinking and behavior. Here is
their letter to doctors.
The FDA is in a very tricky position. Data on a medication's
effectiveness arrive in their hands long before accumulated data, in thousands of patients, regarding a medication's
safety. If a medication appears to be very helpful, but then later it begins to appear that the medication is perhaps unsafe in some way, the FDA may have to withdraw an approval they already granted, or add a significant warning -- as in the case of suicidality in children and adolescents taking antidepressants.
Think about the dilemma the FDA faces at that point: if they judge a medication wrongly, and add a warning when none is really needed, they can cause people to avoid using a medication that could really help. For example, in the case of the suicide warning, the FDA could have caused direct harm by making people afraid to use medications that could actually
lower the risk of suicide.
On the other hand, if they wait to long to issue a warning about a medication's risks, then they will have allowed people to be harmed whose injuries -- or death -- could have been avoided. If I was in their position, I think I'd have difficulty sleeping at night worrying about whether to act (issue a warning) or wait for more data.
One of the results of this dilemma is a "double standard" for evidence about whether a medication
works, versus whether it is
safe. As long as a few research studies show a medication to be superior to placebo, the FDA concludes that it works. This is because showing that one treatment is truly better than another, in rigorous statistical terms, is difficult. It's actually pretty easy for a drug to be truly better than placebo and yet not appear so in a research study. The FDA knows this. That's why most antidepressants have many more studies showing it was no better than placebo, than they have studies showing clear evidence of effectiveness. That doesn't mean the medications don't work. It means that showing they work is difficult (but not impossible; if there are no studies showing it is better than placebo, we can much more confidently conclude that the medication in question really
doesn't work).
But when it comes to safety, much more rigorous standards must be met to conclude that a medication is unsafe, e.g. causes suicidal thinking and actions. The FDA has been waiting for enough data to conclude that there really is a risk, but in this case studies which fail to show evidence for risk are not simply disregarded, as they are when judging effectiveness. Rather, such studies are given very substantial weight. This might sound very unfair, as though the FDA is deliberately tilting the scales of justice toward the drug companies when making their decisions. And frankly, watching from the sidelines, it sure looks like that's true sometimes. Yet at the same time, it
could be that the FDA is just trying to make sure they don't interfere with medical practice until they're relatively sure a problem exists -- and arriving at that conclusion is difficult, slow, and reliant on a different standard of evidence.
I don't think I'm just being naive there (or speaking from the back pocket of the drug industry; for more information on how I handle that relationship, see
Funding). I think that it's possible that the FDA is really trying to do the best they can, without too much direct pharmaceutical company influence. I'm thinking of a few particular FDA officials, in this respect, e.g. Drs. Mosholder and Laughren. A correspondent who keeps me up to date on all this, who has attended the FDA hearings, might offer a different conclusion. She sent me a detailed analysis of this problem by a very skeptical -- but smart -- psychiatrist, an essay by Dr. David Healy in the current British Medical Journal: "Did Regulators Fail Over Selective Serotonin Reuptake Inhibitors (SSRI's)? It's
worth a look if you can follow the language.
The bottom line: the FDA is considering a warning about antidepressants and suicidality in
adults. They are dealing with the same dilemma they faced over issuing the warning for children, and that will make them slow to act. If they wait much longer, and we later see clearer evidence that antidepressants
cause suicidal thoughts and actions, we'll blame them for not acting sooner. If they're wrong by issuing a warning later found to be unnecessary, we'll blame them for bureaucratic interference with medical practice (and there have been
many such criticisms, already, over the kids' warning).
What to do in the meantime? Two steps are clearly warranted: first, make sure before using an antidepressant that you don't have "depression
plus" -- a version of depression with some degree of bipolar disorder mixed in, which can be subtle (Bipolar II, and "soft" bipolarity). Almost everyone agrees that antidepressants pose more risk of suicidality in people with bipolar depression. How are you supposed to find out whether you have some degree of bipolarity? Start by reading my
website about Bipolar II, the version which looks just like depression but often does not get better on antidepressants and may actually get worse. Note that I'm not making any money when you go there. You might get interested and buy my book. That's one of the reasons for writing this blog, I confess.
The other step: learn about the
9 antidepressants that aren't antidepressants. It's nice to know that there are so many alternatives, especially if you conclude that you have some degree of bipolarity. Many of the 9 are good for
any depression, while some should probably only be used if you have bipolar depression. Good luck with your learning about all this. Thanks for reading.
Dr. Phelps.