The Food and Drug Administration (FDA) warnings about risks of suicidal
thoughts associated with antidepressants have indeed had discernible effects
on the prescribing trends of these medications, especially for younger
patients, a new study in the January Archives of General Psychiatry
shows. Opinions differ, however, on what these trends mean.
Using prescription drug claims from the database of Medco, a large pharmacy
benefit management company, Mark Olfson, M.D., M.P.H., a professor of clinical
psychiatry at New York State Psychiatric Institute and Columbia University
College of Physicians and Surgeons, and two coauthors analyzed prescribing
patterns for paroxetine and other antidepressants from 2002 to 2005 to
investigate whether two major warnings by the FDA, announced in June 2003 and
October 2004, had any measurable effect. This study was funded by the Agency
for Healthcare Research and Quality and the Carmel Hill Fund.
The researchers chose three time periods for comparison: (1) from May 1,
2002 to June 19, 2003, the date that the FDA announced its suspicion that
paroxetine may increase suicidal thoughts and behaviors in youth; (2) from
June 20, 2003, to October 15, 2004, the date that the agency issued a
black-box warning for all antidepressants on the suicide risk in children and
adolescents; and (3) from October 26, 2004, to December 31, 2005.
Within each time period, according to the described methods, the database
contained more than 2 million Medco members with continuous eligibility. The
authors drew a 2 percent random sample from the member pool in each period.
They further broke down the subjects into three age groups: youth aged 6 to 17
years, adults 18 to 64, and older adults 65 or
FDA Warnings Found to Have Impact on Antidepressant Trends
The actual number of patients in each category who were included in the
analyses was not provided in the article.
The trends for the use of all antidepressants, regardless of what they were
prescribed for, in youth changed from an increase (at an annualized rate of
36.0 percent) before June 2003 to flat (annualized rate of -0.8 percent) in
the second period, to a decrease (-9.6 percent) after the black-box warning
(see chart). The declining trend after October 2004, however, was not
statistically significant within the period. The reverse prescribing trend was
more pronounced for paroxetine, with trends changing from growth before June
2003 to a sharp decline in the second period and further decline after October
The prescribing trends in adults were less affected by the regulatory
warnings. The annualized trends for all antidepressant use"
decelerated" from a pre-2003 growth rate of 10.9 percent to an
essentially flat 0.8 percent after the black-box warning was issued in October
2004. The increase in prescribing for all antidepressants for older patients
resumed the upward trend after the FDA's black-box warning, which defines the
population at risk as pediatric patients and young adults.
The authors also analyzed new antidepressant use, defined as patients with
no antidepressant use in the preceding 120 days before the first documented
antidepressant claim. The annualized trends for all antidepressants for youth
also changed from growth to flat, followed by a decline that was statistically
nonsignificant within the final period (see chart). The sample sizes of new
use and the actual percentages of use in these groups were not reported in the
"When we separated out the new antidepressant use and total
prescriptions," Olfson told Psychiatric News, "the trends
were largely similar."
As the prescription declines were primarily seen in young patients under 18
and more prominent for paroxetine, the authors concluded that the FDA's
warnings appeared to "have had a modest and reasonably targeted effect
on antidepressant treatment patterns" and that "these changes were
broadly consistent with the FDA warnings and the scientific
These conclusions contrast with those from other studies on
antidepressant-use trends that were published in recent issues of the
American Journal of Psychiatry (AJP). Those studies examined how the
FDA's public health advisory, issued in October 2003, paralleled changing
trends in antidepressant prescribing patterns in patients diagnosed with
depression (Psychiatric News, June 15, September 7, and December 21,
2007). A group of researchers from the University of Colorado and Harvard
Medical School found, in one of a series of studies, that significantly fewer
patients between 5 and 18 years old with newly diagnosed episodes of
depression were prescribed antidepressant treatment following diagnosis in the
time period after the FDA advisory. These authors, along with other experts
and patient advocates in psychiatry, expressed concerns that the decline in
prescribing meant that many patients were not getting the pharmacologic
treatment they needed.
Anne Libby, Ph.D., an assistant professor of psychiatry at the University
of Colorado, told Psychiatric News, "Our studies had some
methodological differences in terms of population, time-point selection, and
sampling from Dr. Olfson's." Both groups of authors collected
prescription claims data from large health care insurance databases and used
time-series analyses to plot trends before and after significant warnings.
Both groups identified the reversal of upward trends of antidepressant
prescriptions after the highly publicized regulatory actions in 2003 and 2004.
Libby and her colleagues focused their data analyses on patients with a
diagnosis of a new episode of depression and the medications they were
prescribed, arguing that those who were already taking medications were less
likely to be taken off of the drugs.
Olfson and his colleagues, meanwhile, focused primarily on total
antidepressant prescriptions regardless of indication in three age groups
(youth, adults, and elderly).
"Dr. Olfson's study is a cross-sectional and broader look at overall
antidepressant use compared with our studies, and paroxetine in
particular," said Robert Valuck, Ph.D., an associate professor of
pharmacy at the University of Colorado and a coauthor of the three
AJP studies with Libby. "We tried to find out the changes in
the care of patients with depression."
In addition to decreased prescriptions for antidepressants, Libby, Valuck,
and colleagues observed declining trends in the diagnosis of depression after
the FDA's advisory in both young and adult patients. For those who were
diagnosed with depression, they found little or no compensatory increases in
nonantidepressant psychoactive medications or psychotherapy claims. Nor did
they observe increased frequency in physician office visits by patients who
were started on antidepressant prescriptions despite the agency's
recommendation on intense monitoring for adverse events and suicidal ideation.
In the Olfson study, the authors acknowledged that their data source did not
allow them to assess whether or how the advisories influenced use of
psychotherapy or antidepressant treatment of specific disorders.
"We did not find that there was a significant absolute decrease in
the rate of antidepressant use by young people," Olfson commented."
Instead, we found that the rate of growth in antidepressant use
significantly slowed following release of the warnings. This pattern is
consistent with increased clinical caution in the use of these
Valuck and Libby, however, expressed concerns about the prescribing
patterns gleaned from their own studies. "It is hard to determine
whether the prescribing patterns reflect appropriate care based on these
available data, which are subject to different interpretations," said
Libby. "We need more data and analyses to understand what is really
going on in the context of actual patient care. We are far from
An abstract of "Effects of Food and Drug Administration
Warnings on Antidepressant Use in a National Sample" is posted at<archpsyc.ama-assn.org/cgi/content/abstract/65/1/94>.▪