Now Dann works as the community education coordinator at the Epilepsy Foundation of Massachusetts and Rhode Island. She's in a better place mentally and physically. Despite surgery, she still needs medications.
"Because I've gone through at least half a dozen different anti-seizure medications and different things have happened with me, I've just come to the determination that everyone, every [person's] body is different.
"Doctors can only give a general statement," she said. "Doctors can't say 'yes' or 'no' this can happen to you."
On the opposite end of that conversation — when a doctor reviews medication with a patient — epilepsy specialists and neurologists wonder what exactly the black box label would do.
"You have to wonder what the doctor or patient is supposed to do with the information that a drug causes an increased risk of suicide," said Dr. Clifford B. Saper, the James Jackson Putnam professor of neurology and neuroscience at Harvard Medical School.
"There is no evidence that I know of that providing this information to the patient, who needs to take the drug anyway, will reduce that risk," Saper said.
Even the definition of "that risk," or the 0.37 percent chance of suicidal behavior, fell under doctors' criticism. The analysis of 199 studies included 11 different drugs and then gave one final figure for all of them.
Dr. Mark Stacy, a neurologist at Duke University, explained that a pooled analysis of this type, called a meta analysis, can be tricky. Did all the studies show a slight increase of suicide, or did many studies find nothing and only a few studies show a drastic increase? If so, for what drug and what type of people were taking it?
"The devil is in these details," Stacy said. "I believe it would also be important to review these data by drug; this would allow assessment of individual compounds, rather than by class."
In another concern, Walson wonders whether including a low percent risk, such as the 0.37 percent risk, may actually interfere with the purpose of the black box warning.
"I think the information belongs in the labeling but not as a black box," said Walson. "Overuse of the black box will dilute its effects — it should be reserved for more clear-cut and higher-risk problems."
"The FDA, in my opinion, is making an honest attempt to react appropriately to new information as it comes available," Walson said. "Unfortunately the FDA has limited ability to influence medical practice."