We've watched from a distance as the FDA has considered whether certain drugs indicated for epilepsy are associated with increased suicidality.
Here, for example, is a post from Pharmalot on that issue, which in turn links to the FDA's raw data.
Please don't confuse us with biostatisticians; we don't even play 'em on TV.
But we looked at the evidence that epilepsy drugs are associated with suicidality, and we're scratching our heads.
First, the FDA looked at a vanishingly small number of incidents -- 142 suicide-related events (FDA report at 22) out of 43,892 patients enrolled in the clinical trials. Id. at 17.
In that tiny data set, the FDA then saw a miniscule difference in suicidality between the drug and control groups -- .37% suicidality in the drug group versus .24% in the placebo patients. Id. at 1.
If you were reading with a skeptical eye, you'd then note a bunch of curious results -- such as there not being increased suicidality in trials conducted in North America (id. at 6) or in females. Id. at 36.
Throw in the likely ascertainment bias in clinical trials, about which we've posted previously.
And then think about the fact that the supposed "signal" for suicidality appears to be driven almost entirely by data from just two of the eleven drugs that the FDA evaluated. Id. at 23.
We're rooting for the FDA on this issue. We sure hope the Agency interprets the data correctly and requires the right warning.
Because if the FDA gets this wrong, there will be a price to pay in public health.
If the FDA requires a black box warning about suicidality on these drugs, then prescriptions are sure to plummet. Which means that people being treated for epilepsy (and others being treated off-label for bipolar disorder, which is itself a significant risk factor for suicide) will be deprived of necessary treatment, and we may once again see suicide rates increase in the wake of over-warnings.
We're on your side, FDA. Think hard, be smart, and protect the public health.