It did not take long for the data on prescriptions of antidepressants to
children and adolescents to reflect shifting patterns as physicians reacted
last spring to the potential connection between SSRIs and suicidal thoughts
and behaviors. Now, serious concerns are being expressed about the probable
impact of a "black-box" warning, jointly recommended last month by
two Food and Drug Administration (FDA) advisory committees (see
Data from Medco Health Solutions Inc. show that within roughly six to eight
weeks of the FDA's February 2 joint meeting of the Psychopharmacologic Drugs
Advisory Committee and Pediatric Advisory Committee investigating a possible
connection between antidepressants and increased suicidal thoughts and
behaviors in children and adolescents, there was a dramatic decrease in the
number of patients under age 18 who filled antidepressant prescriptions.
Medco is one of the nation's largest pharmacy benefit managers, providing
prescription benefits to some 60 million Americans a year, 3.5 million of whom
are under 18.
The data show that at the end of the first quarter of 2004 (March 31), the
number of patients under age 18 receiving antidepressants dropped by 18
percent, compared with the quarter ending December 31, 2003. That decline was
followed by an additional 5 percent decrease by the end of the second quarter
of 2004 (June 30).
These declines starkly contrast with the steady increases in prescriptions
for antidepressant (and other psychotropic) medications each quarter for the
last several years. In May Medco released another analysis showing a dramatic
77 percent increase in prescription-drug spending for behavioral conditions in
children and adolescents between 2000 and 2003. That analysis attributed the
rise to both increased costs and increased numbers of prescriptions.
The Medco analysis, commissioned and first reported by the New York
Times on September 21, "appears to support [the idea] that
information and education is a powerful tool to improve our health care
delivery system," said Robert Epstein, M.D., Medco's chief medical
officer, in a prepared statement.
The Psychopharmacologic Drugs and Pediatric advisory committees'
recommendation in favor of a black-box warning—the strongest possible
short of a contraindication—was the result of an often impassioned and,
by some descriptions theatrical, joint hearing September 13 and 14.
During that meeting, FDA officials presented data from IMS Health that
appeared to show not a decrease in antidepressant prescriptions, as do the
Medco data, but a 7 percent increase in the six months following the February
meeting and the subsequent warnings. Some committee members said they believed
the prior warnings had had no effect and wondered whether more drastic
warnings were needed.
In the end, the committees' vote on the black-box warning was 15-8 in favor
of adding it to the labels of all antidepressants marketed in the United
States. While the advisory committees did not discuss specific labeling, they
were asked by the agency whether they supported general phrases regarding
increased risk and need for closer monitoring and observation.
Members of the committees seemed to take to heart many of the concerns
expressed by those speaking during the public-comment portion of the hearing,
including several representatives of APA and the American Academy of Child and
Adolescent Psychiatry (AACAP). (See box for details on APA and AACAP efforts regarding SSRIs in children.) In a statement issued September 16, APA
expressed support for the advisory committees' recommendation that the FDA
maintain access to antidepressant medications and noted the hearing "was
inclusive and thorough and based on science and concern for patients and
APA said, "The advisory committees have established a strong case for
closer monitoring of patients with depression, especially vulnerable
Specifically, the advisory committees concluded that the finding of an
increased risk of suicidal thoughts and behaviors in pediatric patients
applied to each of the nine drugs studied so far, since there are no adequate
data available to exclude any single antidepressant from the risk. The nine
drugs are bupropion, citalopram, fluoxetine, fluvoxamine, mirtazapine,
nefazodone, paroxetine, sertraline, and venlafaxine. Some committee members
specifically worried that if only SSRIs carried the black-box warning, some
physicians and/or patients and their caregivers might be falsely led to
believe that tricyclics are safer, something committee members stated was not
At the same time, the advisory committees strongly urged the FDA to avoid
contraindication of the medications for pediatric depression, noting that"
access to these therapies is important for those who could
In addition, the committees recommended that all antidepressants be
dispensed with a mandatory patient-information sheet or "medication
guide" to ensure that not only prescribers are warned by the black-box
labeling, but also patients and their caregivers are appropriately advised of
the risks, as well as the benefits, of antidepressant therapy.
The committees also recommended that the results of controlled pediatric
trials of depression be included in the labeling for all antidepressant
At the core of the advisory committees' hearing and votes was the Columbia
University analysis of adverse events from 23 clinical trials of the nine
antidepressant medications for major depression, anxiety disorders, and ADHD
in children and adolescents. The FDA had asked Columbia last spring to
reanalyze patient-level data from the trials in an effort to better understand
the significance of the apparent signal seen in the summary data from the
The Columbia analysis, which took four months to complete, involved
reclassifying all of the reported serious adverse events noted in any of the
trials and was aimed at validating each adverse event as a "suicidal
event" or "not related to a suicidal event." After the
Columbia reclassification was forwarded to the FDA in July, the agency turned
it over to one of its internal experts for interpretation.
Tarek Hammad, M.D., Ph.D., a senior medical reviewer in the Division of
Neuropharmacological Drug Products at the FDA Center for Drug Evaluation and
Research, concluded that not only was the Columbia reclassification
methodologically and statistically sound, but also that the end results were
not very different from an original FDA review of antidepressant adverseevent
data by FDA medical reviewer Andrew Mosholder, M.D. (see
However, Hammad's analysis of the reclassified data from Columbia involved
much more sophisticated statistical analyses to deal with a complex and
heterogeneous database that originally had not been developed with any
intention of determining an association between treatment and suicidal
thoughts and/or behaviors. Some members of the advisory committees noted that
the analysis helped to clarify and support the credibility of the review
In all, Hammad's analysis of the data estimated "the absolute
increase in the risk of the event of interest [suicidal thoughts and
behaviors] due to treatment for all SSRIs in pediatric major depressive
disorder trials at 2 percent to 3 percent."
More simply put, Hammad reported, "Out of 100 patients treated, we
would expect two or three patients to have some increase in suicidality that
is due to shortterm treatment with the drug, beyond what would be expected as
a result of the disease itself."
Hammad calculated that the overall relative risk of suicidal thoughts
and/or behaviors fell between a low of 1.37 (citalopram compared with placebo)
to a high of 8.84 (venlafaxine compared with placebo).
Hammad and other FDA officials stressed to the advisory committees that any
firm conclusions from the available data are difficult to make because of the
extreme variability of the data from one drug to another and between
individual trials from the same medication.
In fact, when it came time for the committee to vote, the FDA director of
medical policy and acting director of the Office of Drug Evaluation I, Robert
Temple, M.D., noted, "We don't have the data to make drug-specific
informed decisions. We already know you don't have enough data. Our question
is, in the face of these limitations, what is your feeling?"
However, the committee members' "feeling" regarding the data
was not good enough for many.
"Notwithstanding the vote of the advisory committees," noted
APA Trustee-at-Large David Fassler, M.D., who delivered APA's testimony at the
advisory committees hearing, "the data still seem inconsistent and
inconclusive. If you look at `suicidal behavior' or `suicidal thinking' alone,
there are no significant differences. If you just look at the studies on
depression or the studies on anxiety disorders, there are also no significant
differences. Only when you combine suicidal behavior and thinking, and you
combine the studies on depression and anxiety disorders, do you find a
statistically significant increase. But even when you combine all the studies,
there is no significant increase in the `emergence' or `worsening' of suicidal
In addition to the inconsistency of the results, none of the clinical
trials on which the reanalysis was based was originally designed to detect the
events of interest, noted Peter Jensen, M.D., the Ruane Professor and director
of the Center for the Advancement of Children's Mental Health at Columbia
University. Jensen was not involved in the Columbia reanalysis project.
"So while there seems to be a `signal' for a potential increase
compared with placebo in some of the controlled studies," Jensen told
Psychiatric News, "none of these studies used groups where
comparisons were made to the natural course of depression itself and where the
person did not see a doctor or get a pill, a placebo, or any
Ironically, it was this highly variable and by some accounts inconsistent
and unreliable data that some used to bolster their points.
"Studies show that greater caution is needed when treating pediatric
and teen depression with medication, and it is clear that physicians are
heeding these warnings," said Medco Health's Epstein, referring to the
reported decrease in antidepressant prescriptions following the February
hearing of the advisory committees.
Many psychiatrists may not agree that the reported reductions in
prescribing are a good thing. In fact, a significant crosssection of the
mental health community is concerned about the potentially grave impact of the
proposed black-box warnings.
"I think the black box may be excessive and could do more harm than
good, if doctors start to withhold treatments where they may be indicated and,
in fact, in some instances lifesaving," Jensen said.
"I have had parents ask questions about the medications that their
children are on, especially those who are doing well and don't want any
changes made," said Paramjit Joshi, M.D., chief of child psychiatry at
Children's National Medical Center in Washington, D.C. "Some new
patients' parents are reluctant to have their child placed on an SSRI. The
pediatricians are definitely feeling more uncomfortable in prescribing these
medications, which raises the issue of access to care, given the low numbers
of child and adolescent psychiatrists."
"The sense of reluctance on the part of clinicians to prescribe SSRIs
to kids will undoubtedly grow," added Christopher Kratochvil, M.D., a
professor of child and adolescent psychiatry at the University of Nebraska
Medical Center. "Pediatricians have been reluctant already. Now this
will only serve to cement that feeling."
Information on the September 13 and 14 hearing is posted online at<www.fda.gov/ohrms/dockets/ac/cder04.html#PsychopharmacologicDrugs>.
The FDA's analysis of Columbia University's reclassification of adverse events
is posted online at<www.fda.gov/ohrms/dockets/ac/04/briefing/2004-4065b1-10-TAB08-Hammads-Review.pdf>.▪