Selective Serotonin Reuptake Inhibitor (SSRI) Drugs: More Risks Than Benefits? Joel M. Kauffman, Ph.D.
excerpts:
"Kirsch et al. concluded that the updated findings from 35 carefully vetted trials suggest that, compared with placebo, the four new-generation antidepressants (fluoxetine, venlafaxine, nefazodone, and paroxetine) do not produce clinically significant improvements in depression in patients who initially have moderate or even severe depression. They show statistically significant but clinically minor effects only in the most severely depressed patients. Moreover, the significance of the effect probably is based on a decreased responsiveness to placebo, rather than increased responsiveness to medication. Given these results, the researchers conclude that there is little reason to prescribe new generation antidepressant medications to any but the most severely depressed patients unless alternative treatments have been ineffective."
"Typical dropout rates in recent trials are claimed to be 5% (see below), but these must be short trials, or trials with a run-in period. In a meta-analysis of 62 earlier trials with a total of 6,000 subjects, the
mean total dropout rate and the proportion of dropouts due to side effects appear comparable to results in general practice: total dropout rates of between 30% and 70% have been reported by 6 weeks, of which some 30%–40% are attributed to side effects and the rest to failure of treatment."
"Early findings of severe adverse effects by SSRI makers came to light only after the class was established. Of 53 healthy volunteer studies on fluoxetine, the results of only 12 were openly reported. From 35 healthy volunteer studies on paroxetine, pre-launch, the results of only 14 appeared. Among the unpublished trials, there was one in which all volunteers dropped out because of agitation (akathisia). In published work on sertraline, data excluded material on behavioral toxicity, including at least one suicide of a healthy volunteer, and in a different trial, 2 of 20 volunteers became
intensely suicidal. This last is consistent with the dropout rate of 5% for agitation alone in actual trials. It is also consistent with Lilly’s animal studies, in which previously friendly cats treated with fluoxetine started growling and hissing—an unheeded warning."
Now why do you suppose so few of the trials results are published? 12 out of 53 that is less that a quarter, and 12 out of 35 is less than half---the drug companies are obviously trying to bury trial results which show "unfavorable results." The FDA is allowing drug companies to bury trial data relevant to a drug's safety and efficacy; it continues to be done...
more excerpts:
"Adverse effects, mostly dose-dependent, will appear in up to 75% of patients on normal doses. Of these, studies suggest that suicidality will be observed in an additional 2%–13% (1 in 50 to 1 in 8) of patients on normal doses, beyond what is seen on placebo or many non-SSRI antidepressant drugs. This is sufficiently frequent that a typical prescribing physician should observe examples in routine practice."
"The actual suicide rate could be about 123/100,000 (1 in 813) higher in patients on SSRIs than in those on tricyclics or placebo. Studies show that many more suicides are attempted on normal doses of SSRIs beyond what is seen on placebo or many non-SSRI antidepressant drugs."
"Available data suggest that actual murders may be committed at about the rate of 250/100,000 (1 in 400) SSRI-treated patients beyond what is seen on placebo or many non-SSRI antidepressant drugs, and that many more murders will be attempted on normal doses as well. While correlation does not prove causation, and results of court trials are not medical science, the data for suicide are solid, and the association of murder with simultaneous suicide is very suggestive."
WARNING:
It is extremely dangerous to abruptly stop or to quickly taper off of SSRI antidepressants. No one should never stop taking these or any psychotropic drug without medical supervision. The entire article is in the read more section below.
<!more!>
ABSTRACT
Anecdotal reports have suggested that selective serotonin
reuptake inhibitors (SSRIs) may cause suicidal or violent behavior
in some patients. Because of the publicity surrounding certain
events, and the numerous lawsuits that have been filed, a review of
benefits and risks is needed.
At most 30% of patients receive a benefit from SSRIs beyond the
large placebo effect in certain mental conditions, especially
depression, according to a recent meta-analysis of published trials.
An equally recent meta-analysis of all SSRI trials submitted to the
FDA showed a small benefit for the severely depressed patients
only. Many early unpublished trials did not show any benefit.
Adverse effects are common, occurring in up to 75% of
subjects. Severe adverse effects may be underreported. Metaanalyses
of controlled trials did not include any actual suicides or
murders, but only suicidality, some finding, in 1991 and 2007, no
evidence even of suicidality. Other meta-analyses using many of
the same trials found that suicidality doubled to 1 in 500 on SSRIs
compared with placebo or non-SSRI antidepressants, but did not
include any actual suicides or murders. The trial designs were
devised by SSRI makers to prevent reports of suicides, by
eliminating subjects with the slightest trace of suicidal tendencies.
Retrospective studies by others showed actual suicides on SSRIs
with a relative risk (RR) of 2–3 compared with non-SSRI
antidepressants, with an increased incidence of 123/100,000.
Lower doses than the smallest available ones were found to
maintain benefits in a majority of patients while reducing risks.
No causal connection between SSRIs and suicide and/or
violence has been proved; neither has it been ruled out. Physicians
need to be vigilant, and aware of legal precedents that may subject
them to enhanced liability when prescribing these drugs.
The Genesis of SSRIs
Fluoxetine (Prozac in the U.S., see Table 1), introduced in 1988
to combat depression, was the fourth selective serotonin reuptake
inhibitor (SSRI) on the U.S. market, after being seriously
considered by Eli Lilly as an antihypertensive drug. Unlike the
earlier “tricyclics” (amitripyline, clomipramine, dothiepin,
imipramine, etc.) and other drug classes, SSRIs acted on the brain
to raise levels of the neurotransmitter serotonin without raising the
levels of norepinephrine. This was thought to be a benefit in
treatment of depression, and later anxiety, panic, social phobia,
obsessive-compulsive disorder (OCD), and many other
conditions. The SSRIs listed in Table 1 are among the most
frequently prescribed in the U.S., and compete with the five non-
SSRIs shown, and others.
1
Benefits of SSRIs
A prominent recent meta-analysis of Bridge et al. included 27
trials of SSRIs for three defined mental conditions: major
depressive disorder (MDD),OCD, and non-OCD anxiety disorders.
Benefits, compared with placebo, were found to be highly
statistically significant. For MDD, data from 13 trials showed
benefit in 61% vs. 50% on placebo, a gain of 11% absolute
(NNT=10), <0.001 for all ages of participants. For OCD, data
from six trials showed benefit in 52% vs. 32% on placebo, a gain of
20% absolute (NNT=5), <0.001 for all ages. For non-OCD
anxiety, data from 6 trials showed benefit in 69% vs. 39% on placebo,
a gain of 30% absolute (NNT=3), <0.001 for all ages. These
results represent the maximum expectation of benefit from SSRIs
since 22 of the 27 trials were financially supported by SSRImakers,
and thus subject to the routinely positive bias of industry-sponsored
clinical trials.
Jay S. Cohen, M.D., author of the 2001 book
, wrote that half his patients did well on
fluoxetine, but he noted a high incidence (50%) with side-effects.
Cohen also cited a pre-approval study showing that the standard 20mg
per day starting dose helped 65% of patients, while 5mg helped 54%,
so Cohen became one of the pioneers in using lower doses before Lilly
made them available. The 1996
entry for paroxetine, at least, confirmed that the 17 most common
side-effects were dose-dependent.
In four observational cohort studies of four common SSRIs
reported by physicians as part of the prescription-event monitoring
program in the UK, with more than 10,000 patients in each drug
group, only 36% of the physicians reported fluvoxamine as effective,
compared with 60% for fluoxetine, sertraline, and paroxetine. These
possible benefit rates, which include the placebo effect, parallel the
percentage of patients remaining on the drug for 2months.
2
3
4
5
P
P
P
Over Dose: the Case
Against the Drug Companies
Physicians Desk Reference (PDR)
Joel M. Kauffman, Ph.D.
Table 1. Commonly Prescribed SSRIs and Other Antidepressants
Selective Serotonin
Reuptake Inhibitor (SSRI) Drugs:
More Risks Than Benefits?
USA Trade Name Generic Name
SSRIs
Celexa
Luvox
Paxil
Prozac
Zoloft
non-SSRIs
Effexor
Remeron
Serzone
Wellbutrin
(UK)
citalopram
fluvoxamine
paroxetine
fluoxetine
sertraline
venlafaxine
mirtazapine
nefazodone
bupropion
dothiepin
Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009 7
An old trial of placebo for anxious and depressed subjects
reduced distress in 43%. Three meta-analyses of the antidepressant
literature that appeared in the 1990s independently concluded that
two-thirds of the effectiveness attributed to SSRIs is actually
placebo effect. In a series of nine controlled studies on hospitalized
patients with depression, 57% of those given placebo showed
improvement in 2–6 weeks.
A1998 meta-analysis of 47 trials on antidepressant medication
including SSRIs indicated that 75% of the response to them was
duplicated by placebo. This meta-analysis was criticized on several
grounds. Therefore, Irving Kirsch, Ph.D., of the University of
Connecticut, with other authors, obtained data submitted to the
FDA on every placebo-controlled clinical trial on the six most
widely used SSRIs, and published a meta-analysis on 47 trials,
finding a small, clinically insignificant effect. This work was
updated in 2008:
Analyses of datasets including unpublished as well as
published clinical trials reveal smaller effects that fall well
below recommended criteria for clinical effectiveness.
Specifically, a meta-analysis of clinical trial data submitted
to the U.S. Food and DrugAdministration (FDA) revealed a
mean drug–placebo difference in improvement scores of
1.80 points on the Hamilton Rating Scale of Depression
(HRSD), whereas the National Institute for Clinical
Excellence (NICE) used a drug–placebo difference of three
points as a criterion for clinical significance when
establishing guidelines for the treatment of depression in the
United Kingdom.
Kirsch et al. concluded that the updated findings from 35
carefully vetted trials suggest that, compared with placebo, the four
new-generation antidepressants (fluoxetine, venlafaxine,
nefazodone, and paroxetine) do not produce clinically significant
improvements in depression in patients who initially have
moderate or even severe depression. They show statistically
significant but clinically minor effects only in the most severely
depressed patients. Moreover, the significance of the effect
probably is based on a decreased responsiveness to placebo, rather
than increased responsiveness to medication. Given these results,
the researchers conclude that there is little reason to prescribe newgeneration
antidepressant medications to any but the most severely
depressed patients unless alternative treatments have been
ineffective. In addition, they write that the decreased placebo
response in extremely depressed patients, combined with a
response to antidepressants comparable to that of less severely
depressed patients, is a potentially important insight that should be
investigated further.
Even these unimpressive findings exaggerated the benefits of
antidepressants. In three fluoxetine trials and in the three sertraline
trials for which data were reported, the protocol allowed
replacement of patients who, in the investigators’ judgment, were
not improving after 2 weeks. The trials also included a 1–2 week
washout period, during which patients were given a placebo prior to
randomization. Those whose scores improved 20% or more were
excluded from the study. In 25 trials, the use of other psychoactive
medication was reported. In most trials, a chloral hydrate sedative
was permitted in doses ranging from 500 mg to 2,000 mg per day.
Other psychoactive medication was usually prohibited but still
reported as having been taken in several trials.
6
1
7
8
9
9
Perhaps such considerations led David Healy, M.D., an SSRI
expert, to his conclusion that “...these drugs do not convincingly
work….” His evidence came from early unpublished clinical trials
whose results were revealed to him atFDAhearings. For fluoxetine,
Healy noted four trials with a positive result and four without. For
sertraline, only one of five early studies showed benefit.
Because of the huge placebo effect, 32–75%, most physicians
unfamiliar with the studies revealing this effect are likely, in my
opinion, to say that one-third to two-thirds of their patients are
improved on SSRIs. This would also explain Dr. Jay S. Cohen’s
findings on lower doses of fluoxetine.
SSRIs reportedly interact with 40 other drugs to cause
“serotonin syndrome.” This presents as twitching, tremors, rigidity,
fever, confusion, or agitation. Serotonin/norepinephrine reuptake
inhibitors (SNRIs) also may cause serotonin syndrome by
interactions. Most tricyclic depressants do not have these
interactions, with the exception of amitriptyline.
In a controlled trial of paroxetine vs. clomipramine sponsored
by GlaxoSmithKline, 75% of the subjects had an adverse effect on
paroxetine, 21% had a severe adverse effect, and 13% committed a
suicidal act (1 in 8). The 1996 ( )
entry for paroxetine lists 17 side-effects with an incidence of ≥5%
for approved doses. They are: asthenia, sweating, constipation,
decreased appetite, diarrhea (up to 15%), dry mouth (up to 21%),
nausea (up to 36%), anxiety, dizziness, nervousness, paresthesia,
somnolence (up to 22%), tremor (up to 15%), blurred vision,
abnormal ejaculation, impotence, and other male genital disorders.
Fully 31 additional side effects with an incidence at least1%greater
than placebo were listed, including uncontrollable yawning.
Murder, suicide, and suicidality were included. Nor were they
on comparable lists for fluvoxamine, or sertraline.
For fluvoxamine, suicide were separately listed as
“infrequent.” For fluoxetine, suicidal ideation was listed as a
voluntary report not proved to be drug related. For sertraline,
suicidal ideation and attempt were listed separately as “infrequent.”
The entry for venlafaxine was: “…the possibility of a suicide
attempt is inherent in depression.”
Not found in the was weight gain, which Cohen lists as a
serious side effect.
Typical dropout rates in recent trials are claimed to be 5% (see
below), but thesemust be short trials, or trials with a run-in period. In
a meta-analysis of 62 earlier trials with a total of 6,000 subjects, the
mean total dropout rate and the proportion of dropouts due to side
effects appear comparable to results in general practice: total dropout
rates of between 30% and 70% have been reported by 6 weeks, of
which some 30%–40% are attributed to side effects and the rest to
failure of treatment.
Early findings of severe adverse effects by SSRI makers came
to light only after the class was established. Of 53 healthy volunteer
studies on fluoxetine, the results of only 12 were openly reported.
From 35 healthy volunteer studies on paroxetine, pre-launch, the
results of only 14 appeared. From 35 pre-launch healthy volunteer
studies on sertraline, only seven appeared.Among the unpublished
trials, there was one in which all volunteers dropped out because of
agitation (akathisia). In published work on sertraline, data excluded
material on behavioral toxicity, including at least one suicide of a
10
11,12
13
14
4
15
Adverse Effects of SSRIs
Physicians Desk Reference PDR
not
attempts
PDR
8 Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009
healthy volunteer, and in a different trial, 2 of 20 volunteers became
intensely suicidal. This last is consistent with the dropout rate of5%
for agitation alone in actual trials. It is also consistent with Lilly’s
animal studies, in which previously friendly cats treated with
fluoxetine started growling and hissing—an unheeded warning.
Just a year after fluoxetine was introduced, Bill Forsyth of
Maui, Hawaii, had taken it for only 12 days when he committed one
of the first murder/suicides attributed to any SSRI. In the same year
Joseph Wesbecker killed eight others and himself in a Louisville,
Ky., printing plant where he worked, after 4 weeks on fluoxetine.
Yet as early as 1986, clinical trials showed a rate of 12.5 suicides per
1,000 subjects on fluoxetine vs. 3.8 on older non-SSRIs vs. 2.5 on
placebo!An internal 1985 Lilly document found even worse results
and said that benefits were less than risks. Such documents were
released into the public domain by Lilly as part of the settlement in
theWesbecker case. Fifteen more “anecdotes” of murder/suicide,
three with sertraline, were listed by DeGrandpre.
Lilly’s denials of a link to murder/suicide on national television
and elsewhere cited a sponsored meta-analysis in in 1991,
which exonerated fluoxetine as a cause of suicidal acts or thoughts
without even mentioning actual murder or suicide. This study
included only 3,067 patients of the 26,000 in the clinical trials it
utilized. None of the trials had a declared endpoint of suicidality.
Some of the trials had been rejected by the FDA. No mention was
made that Lilly had had benzodiazepines co-prescribed tominimize
the agitation that had been recognized with fluoxetine alone. The
5% dropout rate for anxiety and agitation (akathisia) would have
taken out the most likely candidates for suicide. Nevertheless, the
1991 study had its intended effect. For example, in 2006 a 900-page
tome entitled , which was aimed at attorneys, cited
this study, and failed lawsuits concerning SSRIs.
The 2007 meta-analysis by Bridge et al. may be influenced by
indirect conflicts of interest that are hard to prove based on the
financial disclosures. Their paper pooled excess risk above placebo
for “suicidal ideation/suicide attempt” from 27 trials. The excess
risk was said to be 0.7% and statistically significant across all
indications, but significant within each indication. Of the 27
trials, only five were sponsored by the drug maker, and one of
these, the 2004Treatment forAdolescents with Depression (TADS)
study of fluoxetine, had the highest rate of suicidality—7% above
placebo. Most of the same trials were used in a meta-analysis by the
FDA, which found a statistically significant excess risk of 2% (4%
vs. 2% on placebo, 1 in 50 more). Bridge et al. used a randomeffects
calculation, while the FDAused a fixed-effects calculation.
In commenting on the negative findings, Bridge et al. write: “No
study [in our meta-analysis] was designed to examine suicidal
ideation/suicide attempt as a study outcome, and in fact most trials
were conducted in patients who had been carefully screened to
exclude youths at risk.” No actual murders or suicides associated
with SSRI use were reported. Did the designs of the studies
preclude detection or reporting?
The Bridge meta-analysis was not just a vindication of SSRIs,
as communicated to the by Gilbert Ross, M.D.,
Medical Director of the American Council on Science & Health.
Ross went further, commenting that the FDA“Black Box warning”
(see below) was counterproductive because it was discouraging the
use of antidepressants! Ross speculated that the lethal rampage of
the Virginia Tech shooter might have resulted from premature
cessation of medications.
16
1
17
1
18
16
19
2
20
BMJ
Drug Injury only
only
not
not
Wall Street Journal
SSRIs in general have long lifetimes in the body. Fluoxetine and
its active metabolite in particular have a half-life of 16 days,
according to the 1996 . In a reexamination of trials in which
suicides or attempts during the inadequate washout period were not
blamed on the drug, it was shown that the relative risk (RR) of
suicidal acts ranged from 3 for sertraline to 10 for fluoxetine.
Aconcurrent meta-analysis of 24 trials by Kaizar et al. utilized
Bayesian statistics, a valid choice, in my opinion, because data do not
have to follow a Gaussian or normal curve to yield valid results, and
this method can be used to revise probabilities to determinewhether a
specific effect was due to a specific cause. They found an
association between SSRI use and suicidality with odds ratios of 2.3
(95% confidence interval [CI] 1.3-3.8), when the diagnosis was
MDD, not OCD, anxiety, nor ADHD. Non-SSRI antidepressants
were said to have no association with suicide. This supports the
FDA’s findings and requirement, as of October, 2004, for a Black
Box warning for all SSRIs, to monitor children and adolescents for
suicidality.Kaizar et al. were concerned that there were no completed
suicides among 4,487 subjects in the trials; that the trial times were
too short at median length of 8 weeks; and that in 10 of the 12 MDD
studies,
Again, there was no citation of actual suicides
associated with SSRIs and no citation ofHealy’s work.
Healy reviewed epidemiologic studies that have been cited to
exonerate SSRIs. One was analyzed by Healy to show a threefold
increase in suicidality compared with other antidepressants.
While “treatment-related activation” has been considered
primarily with regard to suicidality, it can lead to harm to others as
well as to self. Healy summarized data on “hostile episodes”
provided by GlaxoSmithKline from placebo-controlled trials with
paroxetine in subjects of all ages: 9,219 on paroxetine and 6,455 on
placebo. The rubric of “hostility” was used in the trial to code for
aggression and violence, including homicide, homicidal acts, and
homicidal ideation, as well as aggressive events and “conduct
disorders.” No homicides were reported from these trials. Overall,
during both therapy and withdrawal, the RR was 2.1 for hostile
events. In children with OCD the RR was 17. Separately, in healthy
volunteer studies, hostile events occurred in 3 of 271 subjects on
paroxetine vs. none of 138 on placebo. In trials of sertraline on
depressed children submitted by Pfizer, 8 of 189 subjects
discontinued for aggression, agitation, or hyperkinesis (a coding
term for akathisia), compared with 0 of 184 on placebo. In clinical
practice, the term akathisia has been restricted to demonstrable
motor restlessness, but if that is the only effect, it would have been
called dyskinesia according to Healy, who cites four studies linking
akathisia to both suicide and homicide.
Actual suicides were combined with suicide attempts in a 2005
meta-analysis of 702 trials of SSRIs vs. either placebo or an active
non-SSRI control. Studies were rejected if the citation was a
review, a result of duplicate publication, too short, crossover, or had
no reporting of actual or attempted suicide. The studies meeting the
criteria included 88,000 patients. For attempted suicide, the RR was
2.3 for SSRIs vs. placebo (95% CI, 1.14-4.55). The number needed
to treat to harm (sometimes called the “reverse NNT”) was 1 in 684.
There was no difference in actual suicide. Of the 702 trials, 104
failed to report adverse events below a certain pre-set limit of 3%,
5%, or 10% of patients. Only 493 trials reported dropout rates, with
a mean of 29%, and the mean follow-up time was only 11 weeks.
Thus, there was clearly gross underreporting of adverse effects.
PDR
children and adolescents with an elevated baseline risk of
suicide were excluded.
21
22
23
16,17,24
16
17
17
25
2
Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009 9
More importantly, because actual suicides are involved, Healy
cited a study by Donovan et al. that demonstrated a RR=3.4
( <0.01) for SSRIs compared with all non-SSRI antidepressants
involving 222 actual suicides, of which 41 were among patients
who had an SSRI within a month of their suicide. Also the British
Drug Safety Research Unit recorded more than 110 suicides in
50,000 patients taking an SSRI, an incidence of 219/100,000
compared with 96/100,000 for the non-SSRI mirtazepine
(Remeron), an increase of 123/100,000, or 1 in 813 (Table 2). Thus
theRRfor actual suicide in patients taking SSRIs was 2.3 (or 2.8 for
paroxetine). Even here, though, no murders were listed.
In another study cited by Healy, Jick et al. reported 143 actual
suicides among 172,598 patients taking antidepressants. The
relative risk of suicide in patients taking fluoxetine was 2.1,
compared with those taking the tricyclic antidepressant dothiepin.
The risk was not age-dependent.
SSRI makers keep insisting that there will be more suicides if
SSRIs are used as frequently as now. But theRRof 2–3 shown in
studies is a number that the number of suicides that may
have been prevented, so SSRI use is associated with more suicides,
not fewer.
The International Coalition for Drug Awareness in cooperation
with the Prozac Survivors Support Group has produced a website on
which about 1,600 violent incidents associated with SSRI use are
described (www.ssristories.com/index.php). The first column on the
type of incident (murder, school shooting, etc.) is a hot link to a publicly
available description of the incident, typically a local newspaper article.
Aselection of 10 entries (rows) is presented here asTable 3.
About 360 suicides are tallied as well as about 400 murder
incidents, many of which were multiple murders, each linked to
26
26
16
27
P
not
net includes
SSRIs Provide 1,600Anecdotes ofViolence
SSRI use (Rosie Meysenburg, personal communication, 2008 .As
the number of “anecdotes” exceeds 1,600—hardly a small
number—the association of SSRIs with murder/suicide, often
combined, must be taken seriously.
The SSRI website was searched to find combined
murder/suicide incidents attributed to a specific SSRI. There were
three for fluvoxamine, four for citalopram, 10 each for paroxetine
and sertraline, and 31 for fluoxetine. Where the studies above
substantiated suicide from SSRI use, the total on the SSRI website
of 48 simultaneous murder/suicide incidents associated with SSRI
use ties together SSRIs and murder. Since there were about two
murders per suicide, we may infer that the murder rate on SSRIs
could be about 250/100,000. Since no clinical trial involving
multiple homicides is ever likely to be run, no firmer evidence is
likely to be found. Healy noted that much of the evidence for suicide
andmurder came from the efforts of journalists and lawyers.
Note that thewebsite carries a prominentwarning that “withdrawal
can often be more dangerous than continuing on a medication.”
Nine violent events cited elsewhere—seven court cases of
homicide (one attempted) and two assaults—were associated with
specific SSRIs: three with paroxetine, three with sertraline, two
with fluoxetine, and one with venlafaxine.
Skeptics have cast doubt on whether the prescribed SSRIs were
actually taken, especially since many medical records of juveniles
were sealed. In the Columbine, Colo., shootings the toxicology
report showed “therapeutic” levels of fluvoxamine in one of the
shooters. The Red Lake, Minn., shooter had fluoxetine found,
according to news items referenced on the website.
A2004 editorial in by Simon Wessely, M.D., a spokesman
for Eli Lilly, and Robert Kerwin, Ph.D, cited only a single
paper byHealy as a source of claims of suicidality that have found a
receptive media audience. Tellingly, the only study described at
length is by Jick et al. on the correlation of SSRI use and “attempted
suicide,” in which the rates on dothiepin, amitriptyline, fluoxetine
and paroxetinewere not statistically different. Actual suicides in this
study (seven on SSRIs)were not mentioned byWessely andKerwin,
nor were the 143 suicides in Jick’s earlier paper. Jick et al. have
been supported partially by GlaxoSmithKline and Pfizer. No study
that reported actual suicides on SSRIs was described in detail, let
alone refuted. Wessely and Kerwin wrote: “The problem is that
depression is unequivocally and substantially associated with
suicide and self-harm.”True, but this not the truth.
)
12
17
28
24
29
27
JAMA
whole
12,692
10,983
13,741
12,734
50,150
13,554
31
20
37
22
110
13
244
183
269
173
219
96
Drug No.
Patients
No.
Suicides
Incidence/100,000
SSRIs
Fluoxetine
Fluvoxamine
Paroxetine
Sertraline
Total SSRI
Mirtazapine
Source: British Drug Safety Research Unit, adapted from Healy, 200316
Table 2. Suicides Related to SSRIs or Mirtazapine
Table 3. Selected Violent Incidents Associated with SSRI Use
Source: List of 1,600 violent incidents found at www.ssristories.com/index.php. Accessed Feb 21, 2009.
Prozac
Zoloft
Luvox/Zoloft
Prozac
Celexa
Paxil
Paxil
Prozac
Prozac
Effexor/Zoloft
Minnesota
South Carolina
Colorado
Oregon
North Carolina
Wyoming
England
Louisiana
Kentucky
Nebraska
10 dead, 7 wounded: dosage increased one week before rampage
15 year old shoots two teachers, killing one: then kills himself
Columbine High School: 15 dead, 24 wounded
Four dead, twenty injured after Prozac withdrawal
Teen shoots at two students: kills his father
Jury finds Paxil was cause of murder-suicide
Man cleared of charges due to Paxil withdrawal defense
Not guilty by reason of Prozac induced insanity: mother kills daughter
Nine dead, 12 wounded in workplace shooting
11 year old hangs himself: lawsuit
School Shooting
School Shooting
School Shooting
School Shooting
School Shooting
Murder-Suicide
Robbery/Armed
Murder
Murder-Suicide
Suicide
INCIDENT TYPE DRUG DATE LOCATION ADDITIONAL INFORMATION
March 24, 2005
October 12, 1995
April 20, 1999
November 5, 1999
August 30, 2006
August 9, 2001
August 10, 2003
July 11, 2003
September 14, 1989
October 6, 2004
10 Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009
The legal defense by Lilly, repeated by the media and others, is
that any suicides are caused by the condition, depression, not by
their drug—whether the violence is associated with short-term drug
use, long-term drug use, increased doses, withdrawal, or
rechallenge. There is no website, as far as I know, for violent acts
committed by persons who never received SSRIs, or for total
violent acts; hence the denominator for violent acts is not known.
Also unknown is the fraction of potentially violent persons who are
treated with SSRIs, or of persons treated with SSRIs who are
potentially violent. The published studies on actual suicide,
however, compare patients on SSRIs with similar patients on non-
SSRI antidepressants or placebo. Children diagnosed with OCD,
not depression, also became suicidal on SSRIs, as did healthy
volunteers. Actual two- to threefold increases in suicide rates have
been demonstrated as well as they could be. How else could
such effects be demonstrated? Who would submit, and what
institutional review board or human subjects committee would
approve a study explicitly designed to show whether assaultive,
homicidal, or other violent behavior increases in subjects
prescribed the study drug?
Denial by SSRI makers of culpability for these risks continues
to this day. Whether physicians’ acting on the Black Box warnings
of 2004 and 2007 for all SSRIs will diminish the incidence of
murders and suicides is not yet known.
Following the introduction of fluoxetine in 1988, only a year
passed before an early user committed multiple murders and suicide;
many other examples followed. More than 200 lawsuits have been
begun by users of SSRIs and victims’ families charging wrongful
death or failure to warn; these have had mixed outcomes.There is now
legal precedent for SSRIs as a cause of murder, and the maker of the
SSRI is potentially liable for damages, according to DavidHealy.
Eli Lilly responded with total denial to the lawsuits claiming a
link between fluoxetine and violence. Several claims were settled
out of court with secret details and no admission of guilt. The
Australian David Hawkins was freed from a murder charge by a
finding of temporary insanity caused by using sertraline. Tim Tobin
of Wyoming won $6.4 million from SmithKline Beecham when a
jury found that a murder/suicide committed by Donald Schell was
attributable to use of paroxetine. There are four other homicide
cases in which the SSRIwas deemed to have contributed, resulting in
a suspended sentence in one case and an insanity verdict in another.
One case of homicide, with a guilty verdict and a life sentence,
followed a judicial ruling that akathisia was associated with SSRI
use, but that a causal relationshipwith homicide could not be argued;
thus the link of an SSRI with homicide was disallowed. This was in
direct conflict with the findings of the four trials cited above.
The SSRI website was searched to find murders related to a
specific SSRI whose perpetrators were acquitted based on
temporary SSRI-induced insanity. There were two cases with
sertraline, four cases with paroxetine, and four cases with
fluoxetine. So a precedent has been established for legal
recognition that an SSRI can be a cause for murder, and that the
drug maker can be found liable for damages. The notices of
suicidality for the SSRIs found in thePDRor package inserts before
2004 did not really warn of actual suicide or murder.
1,11,17
10-12
1
17
200 SSRI-related Lawsuits
The Black Box warning of 2004 about possible suicide in
children under 18 years of age did not cover adults or murder at any
age, so potential liability for the SSRI makers still exists. In 2007
the warning was extended to persons under age 25 years. David
Healy was quoted as saying that the warning was overdue, and that
the risk was not likely to disappear above age 25. This was shown
by the trials from GlaxoSmithKline on paroxetine cited above.
Antidepressants are extraordinarily difficult to assess for risks
or benefits in trials.
At most, 11%–30% of patients with depression or related
conditions who take SSRIs actually benefited beyond the placebo
effect on normal doses. Of the perceived benefit, 32%–67% can be
attributed to the placebo effect.
Adverse effects, mostly dose-dependent, will appear in up to
75% of patients on normal doses. Of these, studies suggest that
suicidality will be observed in an additional2%–13%(1 in 50 to 1 in
8) of patients on normal doses, beyond what is seen on placebo or
many non-SSRI antidepressant drugs. This is sufficiently frequent
that a typical prescribing physician should observe examples in
routine practice.
The actual suicide rate could be about 123/100,000 (1 in 813)
higher in patients on SSRIs than in those on tricyclics or placebo.
Studies show that many more suicides are on normal
doses of SSRIs beyond what is seen on placebo or many non-SSRI
antidepressant drugs.
Available data suggest that actual murders may be committed at
about the rate of 250/100,000 (1 in 400) SSRI-treated patients
beyond what is seen on placebo or many non-SSRI antidepressant
drugs, and that many more murders will be attempted on normal
doses as well. While correlation does not prove causation, and
results of court trials are not medical science, the data for suicide are
solid, and the association of murder with suicide is
very suggestive.
Nowthat there is a stronger Black Box warning, physicians who
ignore it may be liable for damages; the warning primarily protects
the manufacturers of SSRIs.
There is obviously great peril in drawing conclusions about
causation from press reports or court decisions. While
manufacturers have a vested interest in exonerating their drugs,
plaintiffs have an interest in blaming it, and defendants in
exonerating themselves.We need careful, independent analysis of
existing study data. In addition to randomized controlled trials,
evidence from basic science (neuropharmacology) and
challenge/dechallenge/rechallenge investigations needs to be
sought. Both the public and individual patients are imperiled by an
incorrect answer to the pressing questions about these widely
prescribed drugs.
Future studies may show lower levels of murder and suicide
with close supervision, and with better matching of this drug type to
patient type.
30
Conclusions
attempted
simultaneous
Joel M. Kauffman, Ph.D.
Acknowledgements:
, is professor of chemistry emeritus at the
University of the Sciences, 600 S. 43rd St., Philadelphia, PA 19104-4495,
Contact: kauffman@bee.net.
Frances E. H. Pane edited the manuscript. David
Moncrief piqued my interest by providing a review copy of
by Richard DeGrandpre.
The Cult of
Pharmacology: How America Became the World’s Most Troubled Drug
Culture
Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009 11
Potential conflicts of interest: The author has neither a financial interest in
any drug mentioned, nor in any alternate treatments for treating any mental
illness.
REFERENCES
1 DeGrandpre R.
Durham, N.C.: Duke University
Press; 2006.
6,
The Cult of Pharmacology: How America Became the
World’s Most Troubled Drug Culture.
2
3
4
5
6
7
8
9
10
11
12
13
Bridge JA, Iyengar S, Salary CB, et al. Clinical response and risk for
reported suicidal ideation and suicide attempts in pediatric
antidepressant treatment. 2007;297:1683-1696.
Jørgensen AW, Hilden J, Gøtzsche PC. Cochrane reviews compared
with industry supported meta-analyses and other meta-analyses of
the same drugs: systematic review. doi:10.1136/bmj.38973.
444699.0B (publ Oct 2006).
Cohen JS. New
York, N.Y.: Tarcher/Putnam; 2001.
Mackay FJ, Dunn NR, Wilton LV, et al. A comparison of fluvoxamine,
fluoxetine, sertraline and paroxetine examined by observational
cohort studies. 1997;6:235-246.
Park L, Covi L. Nonblind placebo trial. 1965;336-345.
Cole JO. Therapeutic efficiency of antidepressant drugs: a review.
1964;190:124-131.
Kirsch I, Moore TJ, Scoboria A, et al. The emperor’s new drugs: an
analysis of antidepressant medication data submitted to the U. S. Food
and Drug Administration. 2002;5(1):23-33.
Kirsch I, Deacon BJ, Huedo-Medina TB, et al. Initial severity and
antidepressant benefits: a meta-analysis of data submitted to the
Food and Drug Administration. 2008;5(2):e45.
doi:10.1371/journal.pmed.0050045.
Healy D. One flew over the conflict of interest nest.
2007;6(1):26-27.
Healy D.
New York, N.Y.: New
York University Press; 2004.
Healy D. FDA Psychopharmacologic Drugs Advisory Committee
hearings. Available at:: www.healyprozac.com/PDAC. Accessed
May 13, 2007.
Wolfe SM, ed. SSRIs can have dangerous interactions with other
drugs. 2008;14(1):2-5. www.citizen.org/
hrg/. Accessed Feb 4, 2009.
JAMA
BMJ,
Over Dose: The Case Against the Drug Companies.
Pharmacoepidemiol Drug Safety
Arch Gen Psychiatry
JAMA
Prevention & Treatment
PLoS Medicine
World Psychiatry
Let Them Eat Prozac: The Unhealthy Relationship Between
the Pharmaceutical Industry and Depression.
Worst Pills Best Pills News
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Braconnier A, Le Coent R, Cohen D. Paroxetine versus clomipramine in
adolescents with severe major depression: a double-blind, randomized,
multicenter trial. 2003;42:22-29.
Anderson IM, Tomenson BM. Treatment discontinuation with
selective serotonin reuptake inhibitors compared with tricyclic
antidepressants: a meta-analysis. 1995;310:1433-1438.
Healy D. Lines of evidence on the risks of suicide with selective
serotonin reuptake inhibitors. 2003:72:71-79.
Healy D, Herxheimer A, Menkes DB. Antidepressants and violence:
problems at the interface of medicine and law.
2006;3(9):1478-1487.
Beasley CM, Dornseif BE, Bosomworth JC. Fluoxetine and suicide: a
meta-analysis of controlled trials of treatment for depression.
1991;303:685-692.
Cohen H. Antidepressants: clinical use and litigation. In:
2nd ed. O’Donnell JT, ed. Tucson,
Ariz.: Lawyers&Judges Publ.Co; 2006:379-390.
Ross G. Black Box backfire. Apr 21, 2007.
Donovan S, Clayton A, Beeharry M, et al. Deliberate self-harm and
antidepressant drugs. 2000;177:551-556.
Kaizar EE, Greenhouse JB, Seltman H, Kelleher K. Do
antidepressants cause suicidality in children? A Bayesian metaanalysis.
2006;3:73-98.
Berenson ML, Levine DM.
. 7th ed.Upper Saddle River,N.J.: Prentilee-Hall; 1998:213-217.
Healy D, Whitaker C. Antidepressants and suicide: risk-benefit
conundrums. 2003;28:331-337.
Fergusson D, Doucette S, Glass KC, et al. Association between
suicide attempts and selective serotonin reuptake inhibitors.
2005;330:396-402.
Donovan S, Kelleher MJ, Lambourn J, Foster T. The occurrence of
suicide following the prescription of antidepressant drugs.
1999;5:181-192.
Jick SS, Dean AD, Jick H. Antidepressants and suicide.
1995;310:215-218.
Wessely S, Kerwin R. Suicide risk and SSRIs. 2004;292:379-381.
Jick H, Kaye JA, Jick SS. Antidepressants and the risk of suicidal
behaviors. 2004;292:338-343.
Carey B. FDA expands suicide warning on drugs. ,
May 3, 2007:A17.
J AmAcadChild Psychiatry
BMJ
Psychother Psychosom
PLoS Med
BMJ
Drug Injury:
Liability, Analysis and Prevention.
Wall Street Journal,
Br J Psychiatry
Clinical Trials
Basic Business Statistics: Concepts and
Applications
J Psychiatry Neurosci
BMJ
Arch
Suicide Res
BMJ
JAMA
JAMA
New York Times
12 Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009
excerpts:
"Kirsch et al. concluded that the updated findings from 35 carefully vetted trials suggest that, compared with placebo, the four new-generation antidepressants (fluoxetine, venlafaxine, nefazodone, and paroxetine) do not produce clinically significant improvements in depression in patients who initially have moderate or even severe depression. They show statistically significant but clinically minor effects only in the most severely depressed patients. Moreover, the significance of the effect probably is based on a decreased responsiveness to placebo, rather than increased responsiveness to medication. Given these results, the researchers conclude that there is little reason to prescribe new generation antidepressant medications to any but the most severely depressed patients unless alternative treatments have been ineffective."
"Typical dropout rates in recent trials are claimed to be 5% (see below), but these must be short trials, or trials with a run-in period. In a meta-analysis of 62 earlier trials with a total of 6,000 subjects, the
mean total dropout rate and the proportion of dropouts due to side effects appear comparable to results in general practice: total dropout rates of between 30% and 70% have been reported by 6 weeks, of which some 30%–40% are attributed to side effects and the rest to failure of treatment."
"Early findings of severe adverse effects by SSRI makers came to light only after the class was established. Of 53 healthy volunteer studies on fluoxetine, the results of only 12 were openly reported. From 35 healthy volunteer studies on paroxetine, pre-launch, the results of only 14 appeared. Among the unpublished trials, there was one in which all volunteers dropped out because of agitation (akathisia). In published work on sertraline, data excluded material on behavioral toxicity, including at least one suicide of a healthy volunteer, and in a different trial, 2 of 20 volunteers became
intensely suicidal. This last is consistent with the dropout rate of 5% for agitation alone in actual trials. It is also consistent with Lilly’s animal studies, in which previously friendly cats treated with fluoxetine started growling and hissing—an unheeded warning."
Now why do you suppose so few of the trials results are published? 12 out of 53 that is less that a quarter, and 12 out of 35 is less than half---the drug companies are obviously trying to bury trial results which show "unfavorable results." The FDA is allowing drug companies to bury trial data relevant to a drug's safety and efficacy; it continues to be done...
more excerpts:
"Adverse effects, mostly dose-dependent, will appear in up to 75% of patients on normal doses. Of these, studies suggest that suicidality will be observed in an additional 2%–13% (1 in 50 to 1 in 8) of patients on normal doses, beyond what is seen on placebo or many non-SSRI antidepressant drugs. This is sufficiently frequent that a typical prescribing physician should observe examples in routine practice."
"The actual suicide rate could be about 123/100,000 (1 in 813) higher in patients on SSRIs than in those on tricyclics or placebo. Studies show that many more suicides are attempted on normal doses of SSRIs beyond what is seen on placebo or many non-SSRI antidepressant drugs."
"Available data suggest that actual murders may be committed at about the rate of 250/100,000 (1 in 400) SSRI-treated patients beyond what is seen on placebo or many non-SSRI antidepressant drugs, and that many more murders will be attempted on normal doses as well. While correlation does not prove causation, and results of court trials are not medical science, the data for suicide are solid, and the association of murder with simultaneous suicide is very suggestive."
WARNING:
It is extremely dangerous to abruptly stop or to quickly taper off of SSRI antidepressants. No one should never stop taking these or any psychotropic drug without medical supervision. The entire article is in the read more section below.
<!more!>
ABSTRACT
Anecdotal reports have suggested that selective serotonin
reuptake inhibitors (SSRIs) may cause suicidal or violent behavior
in some patients. Because of the publicity surrounding certain
events, and the numerous lawsuits that have been filed, a review of
benefits and risks is needed.
At most 30% of patients receive a benefit from SSRIs beyond the
large placebo effect in certain mental conditions, especially
depression, according to a recent meta-analysis of published trials.
An equally recent meta-analysis of all SSRI trials submitted to the
FDA showed a small benefit for the severely depressed patients
only. Many early unpublished trials did not show any benefit.
Adverse effects are common, occurring in up to 75% of
subjects. Severe adverse effects may be underreported. Metaanalyses
of controlled trials did not include any actual suicides or
murders, but only suicidality, some finding, in 1991 and 2007, no
evidence even of suicidality. Other meta-analyses using many of
the same trials found that suicidality doubled to 1 in 500 on SSRIs
compared with placebo or non-SSRI antidepressants, but did not
include any actual suicides or murders. The trial designs were
devised by SSRI makers to prevent reports of suicides, by
eliminating subjects with the slightest trace of suicidal tendencies.
Retrospective studies by others showed actual suicides on SSRIs
with a relative risk (RR) of 2–3 compared with non-SSRI
antidepressants, with an increased incidence of 123/100,000.
Lower doses than the smallest available ones were found to
maintain benefits in a majority of patients while reducing risks.
No causal connection between SSRIs and suicide and/or
violence has been proved; neither has it been ruled out. Physicians
need to be vigilant, and aware of legal precedents that may subject
them to enhanced liability when prescribing these drugs.
The Genesis of SSRIs
Fluoxetine (Prozac in the U.S., see Table 1), introduced in 1988
to combat depression, was the fourth selective serotonin reuptake
inhibitor (SSRI) on the U.S. market, after being seriously
considered by Eli Lilly as an antihypertensive drug. Unlike the
earlier “tricyclics” (amitripyline, clomipramine, dothiepin,
imipramine, etc.) and other drug classes, SSRIs acted on the brain
to raise levels of the neurotransmitter serotonin without raising the
levels of norepinephrine. This was thought to be a benefit in
treatment of depression, and later anxiety, panic, social phobia,
obsessive-compulsive disorder (OCD), and many other
conditions. The SSRIs listed in Table 1 are among the most
frequently prescribed in the U.S., and compete with the five non-
SSRIs shown, and others.
1
Benefits of SSRIs
A prominent recent meta-analysis of Bridge et al. included 27
trials of SSRIs for three defined mental conditions: major
depressive disorder (MDD),OCD, and non-OCD anxiety disorders.
Benefits, compared with placebo, were found to be highly
statistically significant. For MDD, data from 13 trials showed
benefit in 61% vs. 50% on placebo, a gain of 11% absolute
(NNT=10), <0.001 for all ages of participants. For OCD, data
from six trials showed benefit in 52% vs. 32% on placebo, a gain of
20% absolute (NNT=5), <0.001 for all ages. For non-OCD
anxiety, data from 6 trials showed benefit in 69% vs. 39% on placebo,
a gain of 30% absolute (NNT=3), <0.001 for all ages. These
results represent the maximum expectation of benefit from SSRIs
since 22 of the 27 trials were financially supported by SSRImakers,
and thus subject to the routinely positive bias of industry-sponsored
clinical trials.
Jay S. Cohen, M.D., author of the 2001 book
, wrote that half his patients did well on
fluoxetine, but he noted a high incidence (50%) with side-effects.
Cohen also cited a pre-approval study showing that the standard 20mg
per day starting dose helped 65% of patients, while 5mg helped 54%,
so Cohen became one of the pioneers in using lower doses before Lilly
made them available. The 1996
entry for paroxetine, at least, confirmed that the 17 most common
side-effects were dose-dependent.
In four observational cohort studies of four common SSRIs
reported by physicians as part of the prescription-event monitoring
program in the UK, with more than 10,000 patients in each drug
group, only 36% of the physicians reported fluvoxamine as effective,
compared with 60% for fluoxetine, sertraline, and paroxetine. These
possible benefit rates, which include the placebo effect, parallel the
percentage of patients remaining on the drug for 2months.
2
3
4
5
P
P
P
Over Dose: the Case
Against the Drug Companies
Physicians Desk Reference (PDR)
Joel M. Kauffman, Ph.D.
Table 1. Commonly Prescribed SSRIs and Other Antidepressants
Selective Serotonin
Reuptake Inhibitor (SSRI) Drugs:
More Risks Than Benefits?
USA Trade Name Generic Name
SSRIs
Celexa
Luvox
Paxil
Prozac
Zoloft
non-SSRIs
Effexor
Remeron
Serzone
Wellbutrin
(UK)
citalopram
fluvoxamine
paroxetine
fluoxetine
sertraline
venlafaxine
mirtazapine
nefazodone
bupropion
dothiepin
Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009 7
An old trial of placebo for anxious and depressed subjects
reduced distress in 43%. Three meta-analyses of the antidepressant
literature that appeared in the 1990s independently concluded that
two-thirds of the effectiveness attributed to SSRIs is actually
placebo effect. In a series of nine controlled studies on hospitalized
patients with depression, 57% of those given placebo showed
improvement in 2–6 weeks.
A1998 meta-analysis of 47 trials on antidepressant medication
including SSRIs indicated that 75% of the response to them was
duplicated by placebo. This meta-analysis was criticized on several
grounds. Therefore, Irving Kirsch, Ph.D., of the University of
Connecticut, with other authors, obtained data submitted to the
FDA on every placebo-controlled clinical trial on the six most
widely used SSRIs, and published a meta-analysis on 47 trials,
finding a small, clinically insignificant effect. This work was
updated in 2008:
Analyses of datasets including unpublished as well as
published clinical trials reveal smaller effects that fall well
below recommended criteria for clinical effectiveness.
Specifically, a meta-analysis of clinical trial data submitted
to the U.S. Food and DrugAdministration (FDA) revealed a
mean drug–placebo difference in improvement scores of
1.80 points on the Hamilton Rating Scale of Depression
(HRSD), whereas the National Institute for Clinical
Excellence (NICE) used a drug–placebo difference of three
points as a criterion for clinical significance when
establishing guidelines for the treatment of depression in the
United Kingdom.
Kirsch et al. concluded that the updated findings from 35
carefully vetted trials suggest that, compared with placebo, the four
new-generation antidepressants (fluoxetine, venlafaxine,
nefazodone, and paroxetine) do not produce clinically significant
improvements in depression in patients who initially have
moderate or even severe depression. They show statistically
significant but clinically minor effects only in the most severely
depressed patients. Moreover, the significance of the effect
probably is based on a decreased responsiveness to placebo, rather
than increased responsiveness to medication. Given these results,
the researchers conclude that there is little reason to prescribe newgeneration
antidepressant medications to any but the most severely
depressed patients unless alternative treatments have been
ineffective. In addition, they write that the decreased placebo
response in extremely depressed patients, combined with a
response to antidepressants comparable to that of less severely
depressed patients, is a potentially important insight that should be
investigated further.
Even these unimpressive findings exaggerated the benefits of
antidepressants. In three fluoxetine trials and in the three sertraline
trials for which data were reported, the protocol allowed
replacement of patients who, in the investigators’ judgment, were
not improving after 2 weeks. The trials also included a 1–2 week
washout period, during which patients were given a placebo prior to
randomization. Those whose scores improved 20% or more were
excluded from the study. In 25 trials, the use of other psychoactive
medication was reported. In most trials, a chloral hydrate sedative
was permitted in doses ranging from 500 mg to 2,000 mg per day.
Other psychoactive medication was usually prohibited but still
reported as having been taken in several trials.
6
1
7
8
9
9
Perhaps such considerations led David Healy, M.D., an SSRI
expert, to his conclusion that “...these drugs do not convincingly
work….” His evidence came from early unpublished clinical trials
whose results were revealed to him atFDAhearings. For fluoxetine,
Healy noted four trials with a positive result and four without. For
sertraline, only one of five early studies showed benefit.
Because of the huge placebo effect, 32–75%, most physicians
unfamiliar with the studies revealing this effect are likely, in my
opinion, to say that one-third to two-thirds of their patients are
improved on SSRIs. This would also explain Dr. Jay S. Cohen’s
findings on lower doses of fluoxetine.
SSRIs reportedly interact with 40 other drugs to cause
“serotonin syndrome.” This presents as twitching, tremors, rigidity,
fever, confusion, or agitation. Serotonin/norepinephrine reuptake
inhibitors (SNRIs) also may cause serotonin syndrome by
interactions. Most tricyclic depressants do not have these
interactions, with the exception of amitriptyline.
In a controlled trial of paroxetine vs. clomipramine sponsored
by GlaxoSmithKline, 75% of the subjects had an adverse effect on
paroxetine, 21% had a severe adverse effect, and 13% committed a
suicidal act (1 in 8). The 1996 ( )
entry for paroxetine lists 17 side-effects with an incidence of ≥5%
for approved doses. They are: asthenia, sweating, constipation,
decreased appetite, diarrhea (up to 15%), dry mouth (up to 21%),
nausea (up to 36%), anxiety, dizziness, nervousness, paresthesia,
somnolence (up to 22%), tremor (up to 15%), blurred vision,
abnormal ejaculation, impotence, and other male genital disorders.
Fully 31 additional side effects with an incidence at least1%greater
than placebo were listed, including uncontrollable yawning.
Murder, suicide, and suicidality were included. Nor were they
on comparable lists for fluvoxamine, or sertraline.
For fluvoxamine, suicide were separately listed as
“infrequent.” For fluoxetine, suicidal ideation was listed as a
voluntary report not proved to be drug related. For sertraline,
suicidal ideation and attempt were listed separately as “infrequent.”
The entry for venlafaxine was: “…the possibility of a suicide
attempt is inherent in depression.”
Not found in the was weight gain, which Cohen lists as a
serious side effect.
Typical dropout rates in recent trials are claimed to be 5% (see
below), but thesemust be short trials, or trials with a run-in period. In
a meta-analysis of 62 earlier trials with a total of 6,000 subjects, the
mean total dropout rate and the proportion of dropouts due to side
effects appear comparable to results in general practice: total dropout
rates of between 30% and 70% have been reported by 6 weeks, of
which some 30%–40% are attributed to side effects and the rest to
failure of treatment.
Early findings of severe adverse effects by SSRI makers came
to light only after the class was established. Of 53 healthy volunteer
studies on fluoxetine, the results of only 12 were openly reported.
From 35 healthy volunteer studies on paroxetine, pre-launch, the
results of only 14 appeared. From 35 pre-launch healthy volunteer
studies on sertraline, only seven appeared.Among the unpublished
trials, there was one in which all volunteers dropped out because of
agitation (akathisia). In published work on sertraline, data excluded
material on behavioral toxicity, including at least one suicide of a
10
11,12
13
14
4
15
Adverse Effects of SSRIs
Physicians Desk Reference PDR
not
attempts
PDR
8 Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009
healthy volunteer, and in a different trial, 2 of 20 volunteers became
intensely suicidal. This last is consistent with the dropout rate of5%
for agitation alone in actual trials. It is also consistent with Lilly’s
animal studies, in which previously friendly cats treated with
fluoxetine started growling and hissing—an unheeded warning.
Just a year after fluoxetine was introduced, Bill Forsyth of
Maui, Hawaii, had taken it for only 12 days when he committed one
of the first murder/suicides attributed to any SSRI. In the same year
Joseph Wesbecker killed eight others and himself in a Louisville,
Ky., printing plant where he worked, after 4 weeks on fluoxetine.
Yet as early as 1986, clinical trials showed a rate of 12.5 suicides per
1,000 subjects on fluoxetine vs. 3.8 on older non-SSRIs vs. 2.5 on
placebo!An internal 1985 Lilly document found even worse results
and said that benefits were less than risks. Such documents were
released into the public domain by Lilly as part of the settlement in
theWesbecker case. Fifteen more “anecdotes” of murder/suicide,
three with sertraline, were listed by DeGrandpre.
Lilly’s denials of a link to murder/suicide on national television
and elsewhere cited a sponsored meta-analysis in in 1991,
which exonerated fluoxetine as a cause of suicidal acts or thoughts
without even mentioning actual murder or suicide. This study
included only 3,067 patients of the 26,000 in the clinical trials it
utilized. None of the trials had a declared endpoint of suicidality.
Some of the trials had been rejected by the FDA. No mention was
made that Lilly had had benzodiazepines co-prescribed tominimize
the agitation that had been recognized with fluoxetine alone. The
5% dropout rate for anxiety and agitation (akathisia) would have
taken out the most likely candidates for suicide. Nevertheless, the
1991 study had its intended effect. For example, in 2006 a 900-page
tome entitled , which was aimed at attorneys, cited
this study, and failed lawsuits concerning SSRIs.
The 2007 meta-analysis by Bridge et al. may be influenced by
indirect conflicts of interest that are hard to prove based on the
financial disclosures. Their paper pooled excess risk above placebo
for “suicidal ideation/suicide attempt” from 27 trials. The excess
risk was said to be 0.7% and statistically significant across all
indications, but significant within each indication. Of the 27
trials, only five were sponsored by the drug maker, and one of
these, the 2004Treatment forAdolescents with Depression (TADS)
study of fluoxetine, had the highest rate of suicidality—7% above
placebo. Most of the same trials were used in a meta-analysis by the
FDA, which found a statistically significant excess risk of 2% (4%
vs. 2% on placebo, 1 in 50 more). Bridge et al. used a randomeffects
calculation, while the FDAused a fixed-effects calculation.
In commenting on the negative findings, Bridge et al. write: “No
study [in our meta-analysis] was designed to examine suicidal
ideation/suicide attempt as a study outcome, and in fact most trials
were conducted in patients who had been carefully screened to
exclude youths at risk.” No actual murders or suicides associated
with SSRI use were reported. Did the designs of the studies
preclude detection or reporting?
The Bridge meta-analysis was not just a vindication of SSRIs,
as communicated to the by Gilbert Ross, M.D.,
Medical Director of the American Council on Science & Health.
Ross went further, commenting that the FDA“Black Box warning”
(see below) was counterproductive because it was discouraging the
use of antidepressants! Ross speculated that the lethal rampage of
the Virginia Tech shooter might have resulted from premature
cessation of medications.
16
1
17
1
18
16
19
2
20
BMJ
Drug Injury only
only
not
not
Wall Street Journal
SSRIs in general have long lifetimes in the body. Fluoxetine and
its active metabolite in particular have a half-life of 16 days,
according to the 1996 . In a reexamination of trials in which
suicides or attempts during the inadequate washout period were not
blamed on the drug, it was shown that the relative risk (RR) of
suicidal acts ranged from 3 for sertraline to 10 for fluoxetine.
Aconcurrent meta-analysis of 24 trials by Kaizar et al. utilized
Bayesian statistics, a valid choice, in my opinion, because data do not
have to follow a Gaussian or normal curve to yield valid results, and
this method can be used to revise probabilities to determinewhether a
specific effect was due to a specific cause. They found an
association between SSRI use and suicidality with odds ratios of 2.3
(95% confidence interval [CI] 1.3-3.8), when the diagnosis was
MDD, not OCD, anxiety, nor ADHD. Non-SSRI antidepressants
were said to have no association with suicide. This supports the
FDA’s findings and requirement, as of October, 2004, for a Black
Box warning for all SSRIs, to monitor children and adolescents for
suicidality.Kaizar et al. were concerned that there were no completed
suicides among 4,487 subjects in the trials; that the trial times were
too short at median length of 8 weeks; and that in 10 of the 12 MDD
studies,
Again, there was no citation of actual suicides
associated with SSRIs and no citation ofHealy’s work.
Healy reviewed epidemiologic studies that have been cited to
exonerate SSRIs. One was analyzed by Healy to show a threefold
increase in suicidality compared with other antidepressants.
While “treatment-related activation” has been considered
primarily with regard to suicidality, it can lead to harm to others as
well as to self. Healy summarized data on “hostile episodes”
provided by GlaxoSmithKline from placebo-controlled trials with
paroxetine in subjects of all ages: 9,219 on paroxetine and 6,455 on
placebo. The rubric of “hostility” was used in the trial to code for
aggression and violence, including homicide, homicidal acts, and
homicidal ideation, as well as aggressive events and “conduct
disorders.” No homicides were reported from these trials. Overall,
during both therapy and withdrawal, the RR was 2.1 for hostile
events. In children with OCD the RR was 17. Separately, in healthy
volunteer studies, hostile events occurred in 3 of 271 subjects on
paroxetine vs. none of 138 on placebo. In trials of sertraline on
depressed children submitted by Pfizer, 8 of 189 subjects
discontinued for aggression, agitation, or hyperkinesis (a coding
term for akathisia), compared with 0 of 184 on placebo. In clinical
practice, the term akathisia has been restricted to demonstrable
motor restlessness, but if that is the only effect, it would have been
called dyskinesia according to Healy, who cites four studies linking
akathisia to both suicide and homicide.
Actual suicides were combined with suicide attempts in a 2005
meta-analysis of 702 trials of SSRIs vs. either placebo or an active
non-SSRI control. Studies were rejected if the citation was a
review, a result of duplicate publication, too short, crossover, or had
no reporting of actual or attempted suicide. The studies meeting the
criteria included 88,000 patients. For attempted suicide, the RR was
2.3 for SSRIs vs. placebo (95% CI, 1.14-4.55). The number needed
to treat to harm (sometimes called the “reverse NNT”) was 1 in 684.
There was no difference in actual suicide. Of the 702 trials, 104
failed to report adverse events below a certain pre-set limit of 3%,
5%, or 10% of patients. Only 493 trials reported dropout rates, with
a mean of 29%, and the mean follow-up time was only 11 weeks.
Thus, there was clearly gross underreporting of adverse effects.
PDR
children and adolescents with an elevated baseline risk of
suicide were excluded.
21
22
23
16,17,24
16
17
17
25
2
Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009 9
More importantly, because actual suicides are involved, Healy
cited a study by Donovan et al. that demonstrated a RR=3.4
( <0.01) for SSRIs compared with all non-SSRI antidepressants
involving 222 actual suicides, of which 41 were among patients
who had an SSRI within a month of their suicide. Also the British
Drug Safety Research Unit recorded more than 110 suicides in
50,000 patients taking an SSRI, an incidence of 219/100,000
compared with 96/100,000 for the non-SSRI mirtazepine
(Remeron), an increase of 123/100,000, or 1 in 813 (Table 2). Thus
theRRfor actual suicide in patients taking SSRIs was 2.3 (or 2.8 for
paroxetine). Even here, though, no murders were listed.
In another study cited by Healy, Jick et al. reported 143 actual
suicides among 172,598 patients taking antidepressants. The
relative risk of suicide in patients taking fluoxetine was 2.1,
compared with those taking the tricyclic antidepressant dothiepin.
The risk was not age-dependent.
SSRI makers keep insisting that there will be more suicides if
SSRIs are used as frequently as now. But theRRof 2–3 shown in
studies is a number that the number of suicides that may
have been prevented, so SSRI use is associated with more suicides,
not fewer.
The International Coalition for Drug Awareness in cooperation
with the Prozac Survivors Support Group has produced a website on
which about 1,600 violent incidents associated with SSRI use are
described (www.ssristories.com/index.php). The first column on the
type of incident (murder, school shooting, etc.) is a hot link to a publicly
available description of the incident, typically a local newspaper article.
Aselection of 10 entries (rows) is presented here asTable 3.
About 360 suicides are tallied as well as about 400 murder
incidents, many of which were multiple murders, each linked to
26
26
16
27
P
not
net includes
SSRIs Provide 1,600Anecdotes ofViolence
SSRI use (Rosie Meysenburg, personal communication, 2008 .As
the number of “anecdotes” exceeds 1,600—hardly a small
number—the association of SSRIs with murder/suicide, often
combined, must be taken seriously.
The SSRI website was searched to find combined
murder/suicide incidents attributed to a specific SSRI. There were
three for fluvoxamine, four for citalopram, 10 each for paroxetine
and sertraline, and 31 for fluoxetine. Where the studies above
substantiated suicide from SSRI use, the total on the SSRI website
of 48 simultaneous murder/suicide incidents associated with SSRI
use ties together SSRIs and murder. Since there were about two
murders per suicide, we may infer that the murder rate on SSRIs
could be about 250/100,000. Since no clinical trial involving
multiple homicides is ever likely to be run, no firmer evidence is
likely to be found. Healy noted that much of the evidence for suicide
andmurder came from the efforts of journalists and lawyers.
Note that thewebsite carries a prominentwarning that “withdrawal
can often be more dangerous than continuing on a medication.”
Nine violent events cited elsewhere—seven court cases of
homicide (one attempted) and two assaults—were associated with
specific SSRIs: three with paroxetine, three with sertraline, two
with fluoxetine, and one with venlafaxine.
Skeptics have cast doubt on whether the prescribed SSRIs were
actually taken, especially since many medical records of juveniles
were sealed. In the Columbine, Colo., shootings the toxicology
report showed “therapeutic” levels of fluvoxamine in one of the
shooters. The Red Lake, Minn., shooter had fluoxetine found,
according to news items referenced on the website.
A2004 editorial in by Simon Wessely, M.D., a spokesman
for Eli Lilly, and Robert Kerwin, Ph.D, cited only a single
paper byHealy as a source of claims of suicidality that have found a
receptive media audience. Tellingly, the only study described at
length is by Jick et al. on the correlation of SSRI use and “attempted
suicide,” in which the rates on dothiepin, amitriptyline, fluoxetine
and paroxetinewere not statistically different. Actual suicides in this
study (seven on SSRIs)were not mentioned byWessely andKerwin,
nor were the 143 suicides in Jick’s earlier paper. Jick et al. have
been supported partially by GlaxoSmithKline and Pfizer. No study
that reported actual suicides on SSRIs was described in detail, let
alone refuted. Wessely and Kerwin wrote: “The problem is that
depression is unequivocally and substantially associated with
suicide and self-harm.”True, but this not the truth.
)
12
17
28
24
29
27
JAMA
whole
12,692
10,983
13,741
12,734
50,150
13,554
31
20
37
22
110
13
244
183
269
173
219
96
Drug No.
Patients
No.
Suicides
Incidence/100,000
SSRIs
Fluoxetine
Fluvoxamine
Paroxetine
Sertraline
Total SSRI
Mirtazapine
Source: British Drug Safety Research Unit, adapted from Healy, 200316
Table 2. Suicides Related to SSRIs or Mirtazapine
Table 3. Selected Violent Incidents Associated with SSRI Use
Source: List of 1,600 violent incidents found at www.ssristories.com/index.php. Accessed Feb 21, 2009.
Prozac
Zoloft
Luvox/Zoloft
Prozac
Celexa
Paxil
Paxil
Prozac
Prozac
Effexor/Zoloft
Minnesota
South Carolina
Colorado
Oregon
North Carolina
Wyoming
England
Louisiana
Kentucky
Nebraska
10 dead, 7 wounded: dosage increased one week before rampage
15 year old shoots two teachers, killing one: then kills himself
Columbine High School: 15 dead, 24 wounded
Four dead, twenty injured after Prozac withdrawal
Teen shoots at two students: kills his father
Jury finds Paxil was cause of murder-suicide
Man cleared of charges due to Paxil withdrawal defense
Not guilty by reason of Prozac induced insanity: mother kills daughter
Nine dead, 12 wounded in workplace shooting
11 year old hangs himself: lawsuit
School Shooting
School Shooting
School Shooting
School Shooting
School Shooting
Murder-Suicide
Robbery/Armed
Murder
Murder-Suicide
Suicide
INCIDENT TYPE DRUG DATE LOCATION ADDITIONAL INFORMATION
March 24, 2005
October 12, 1995
April 20, 1999
November 5, 1999
August 30, 2006
August 9, 2001
August 10, 2003
July 11, 2003
September 14, 1989
October 6, 2004
10 Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009
The legal defense by Lilly, repeated by the media and others, is
that any suicides are caused by the condition, depression, not by
their drug—whether the violence is associated with short-term drug
use, long-term drug use, increased doses, withdrawal, or
rechallenge. There is no website, as far as I know, for violent acts
committed by persons who never received SSRIs, or for total
violent acts; hence the denominator for violent acts is not known.
Also unknown is the fraction of potentially violent persons who are
treated with SSRIs, or of persons treated with SSRIs who are
potentially violent. The published studies on actual suicide,
however, compare patients on SSRIs with similar patients on non-
SSRI antidepressants or placebo. Children diagnosed with OCD,
not depression, also became suicidal on SSRIs, as did healthy
volunteers. Actual two- to threefold increases in suicide rates have
been demonstrated as well as they could be. How else could
such effects be demonstrated? Who would submit, and what
institutional review board or human subjects committee would
approve a study explicitly designed to show whether assaultive,
homicidal, or other violent behavior increases in subjects
prescribed the study drug?
Denial by SSRI makers of culpability for these risks continues
to this day. Whether physicians’ acting on the Black Box warnings
of 2004 and 2007 for all SSRIs will diminish the incidence of
murders and suicides is not yet known.
Following the introduction of fluoxetine in 1988, only a year
passed before an early user committed multiple murders and suicide;
many other examples followed. More than 200 lawsuits have been
begun by users of SSRIs and victims’ families charging wrongful
death or failure to warn; these have had mixed outcomes.There is now
legal precedent for SSRIs as a cause of murder, and the maker of the
SSRI is potentially liable for damages, according to DavidHealy.
Eli Lilly responded with total denial to the lawsuits claiming a
link between fluoxetine and violence. Several claims were settled
out of court with secret details and no admission of guilt. The
Australian David Hawkins was freed from a murder charge by a
finding of temporary insanity caused by using sertraline. Tim Tobin
of Wyoming won $6.4 million from SmithKline Beecham when a
jury found that a murder/suicide committed by Donald Schell was
attributable to use of paroxetine. There are four other homicide
cases in which the SSRIwas deemed to have contributed, resulting in
a suspended sentence in one case and an insanity verdict in another.
One case of homicide, with a guilty verdict and a life sentence,
followed a judicial ruling that akathisia was associated with SSRI
use, but that a causal relationshipwith homicide could not be argued;
thus the link of an SSRI with homicide was disallowed. This was in
direct conflict with the findings of the four trials cited above.
The SSRI website was searched to find murders related to a
specific SSRI whose perpetrators were acquitted based on
temporary SSRI-induced insanity. There were two cases with
sertraline, four cases with paroxetine, and four cases with
fluoxetine. So a precedent has been established for legal
recognition that an SSRI can be a cause for murder, and that the
drug maker can be found liable for damages. The notices of
suicidality for the SSRIs found in thePDRor package inserts before
2004 did not really warn of actual suicide or murder.
1,11,17
10-12
1
17
200 SSRI-related Lawsuits
The Black Box warning of 2004 about possible suicide in
children under 18 years of age did not cover adults or murder at any
age, so potential liability for the SSRI makers still exists. In 2007
the warning was extended to persons under age 25 years. David
Healy was quoted as saying that the warning was overdue, and that
the risk was not likely to disappear above age 25. This was shown
by the trials from GlaxoSmithKline on paroxetine cited above.
Antidepressants are extraordinarily difficult to assess for risks
or benefits in trials.
At most, 11%–30% of patients with depression or related
conditions who take SSRIs actually benefited beyond the placebo
effect on normal doses. Of the perceived benefit, 32%–67% can be
attributed to the placebo effect.
Adverse effects, mostly dose-dependent, will appear in up to
75% of patients on normal doses. Of these, studies suggest that
suicidality will be observed in an additional2%–13%(1 in 50 to 1 in
8) of patients on normal doses, beyond what is seen on placebo or
many non-SSRI antidepressant drugs. This is sufficiently frequent
that a typical prescribing physician should observe examples in
routine practice.
The actual suicide rate could be about 123/100,000 (1 in 813)
higher in patients on SSRIs than in those on tricyclics or placebo.
Studies show that many more suicides are on normal
doses of SSRIs beyond what is seen on placebo or many non-SSRI
antidepressant drugs.
Available data suggest that actual murders may be committed at
about the rate of 250/100,000 (1 in 400) SSRI-treated patients
beyond what is seen on placebo or many non-SSRI antidepressant
drugs, and that many more murders will be attempted on normal
doses as well. While correlation does not prove causation, and
results of court trials are not medical science, the data for suicide are
solid, and the association of murder with suicide is
very suggestive.
Nowthat there is a stronger Black Box warning, physicians who
ignore it may be liable for damages; the warning primarily protects
the manufacturers of SSRIs.
There is obviously great peril in drawing conclusions about
causation from press reports or court decisions. While
manufacturers have a vested interest in exonerating their drugs,
plaintiffs have an interest in blaming it, and defendants in
exonerating themselves.We need careful, independent analysis of
existing study data. In addition to randomized controlled trials,
evidence from basic science (neuropharmacology) and
challenge/dechallenge/rechallenge investigations needs to be
sought. Both the public and individual patients are imperiled by an
incorrect answer to the pressing questions about these widely
prescribed drugs.
Future studies may show lower levels of murder and suicide
with close supervision, and with better matching of this drug type to
patient type.
30
Conclusions
attempted
simultaneous
Joel M. Kauffman, Ph.D.
Acknowledgements:
, is professor of chemistry emeritus at the
University of the Sciences, 600 S. 43rd St., Philadelphia, PA 19104-4495,
Contact: kauffman@bee.net.
Frances E. H. Pane edited the manuscript. David
Moncrief piqued my interest by providing a review copy of
by Richard DeGrandpre.
The Cult of
Pharmacology: How America Became the World’s Most Troubled Drug
Culture
Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009 11
Potential conflicts of interest: The author has neither a financial interest in
any drug mentioned, nor in any alternate treatments for treating any mental
illness.
REFERENCES
1 DeGrandpre R.
Durham, N.C.: Duke University
Press; 2006.
6,
The Cult of Pharmacology: How America Became the
World’s Most Troubled Drug Culture.
2
3
4
5
6
7
8
9
10
11
12
13
Bridge JA, Iyengar S, Salary CB, et al. Clinical response and risk for
reported suicidal ideation and suicide attempts in pediatric
antidepressant treatment. 2007;297:1683-1696.
Jørgensen AW, Hilden J, Gøtzsche PC. Cochrane reviews compared
with industry supported meta-analyses and other meta-analyses of
the same drugs: systematic review. doi:10.1136/bmj.38973.
444699.0B (publ Oct 2006).
Cohen JS. New
York, N.Y.: Tarcher/Putnam; 2001.
Mackay FJ, Dunn NR, Wilton LV, et al. A comparison of fluvoxamine,
fluoxetine, sertraline and paroxetine examined by observational
cohort studies. 1997;6:235-246.
Park L, Covi L. Nonblind placebo trial. 1965;336-345.
Cole JO. Therapeutic efficiency of antidepressant drugs: a review.
1964;190:124-131.
Kirsch I, Moore TJ, Scoboria A, et al. The emperor’s new drugs: an
analysis of antidepressant medication data submitted to the U. S. Food
and Drug Administration. 2002;5(1):23-33.
Kirsch I, Deacon BJ, Huedo-Medina TB, et al. Initial severity and
antidepressant benefits: a meta-analysis of data submitted to the
Food and Drug Administration. 2008;5(2):e45.
doi:10.1371/journal.pmed.0050045.
Healy D. One flew over the conflict of interest nest.
2007;6(1):26-27.
Healy D.
New York, N.Y.: New
York University Press; 2004.
Healy D. FDA Psychopharmacologic Drugs Advisory Committee
hearings. Available at:: www.healyprozac.com/PDAC. Accessed
May 13, 2007.
Wolfe SM, ed. SSRIs can have dangerous interactions with other
drugs. 2008;14(1):2-5. www.citizen.org/
hrg/. Accessed Feb 4, 2009.
JAMA
BMJ,
Over Dose: The Case Against the Drug Companies.
Pharmacoepidemiol Drug Safety
Arch Gen Psychiatry
JAMA
Prevention & Treatment
PLoS Medicine
World Psychiatry
Let Them Eat Prozac: The Unhealthy Relationship Between
the Pharmaceutical Industry and Depression.
Worst Pills Best Pills News
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Braconnier A, Le Coent R, Cohen D. Paroxetine versus clomipramine in
adolescents with severe major depression: a double-blind, randomized,
multicenter trial. 2003;42:22-29.
Anderson IM, Tomenson BM. Treatment discontinuation with
selective serotonin reuptake inhibitors compared with tricyclic
antidepressants: a meta-analysis. 1995;310:1433-1438.
Healy D. Lines of evidence on the risks of suicide with selective
serotonin reuptake inhibitors. 2003:72:71-79.
Healy D, Herxheimer A, Menkes DB. Antidepressants and violence:
problems at the interface of medicine and law.
2006;3(9):1478-1487.
Beasley CM, Dornseif BE, Bosomworth JC. Fluoxetine and suicide: a
meta-analysis of controlled trials of treatment for depression.
1991;303:685-692.
Cohen H. Antidepressants: clinical use and litigation. In:
2nd ed. O’Donnell JT, ed. Tucson,
Ariz.: Lawyers&Judges Publ.Co; 2006:379-390.
Ross G. Black Box backfire. Apr 21, 2007.
Donovan S, Clayton A, Beeharry M, et al. Deliberate self-harm and
antidepressant drugs. 2000;177:551-556.
Kaizar EE, Greenhouse JB, Seltman H, Kelleher K. Do
antidepressants cause suicidality in children? A Bayesian metaanalysis.
2006;3:73-98.
Berenson ML, Levine DM.
. 7th ed.Upper Saddle River,N.J.: Prentilee-Hall; 1998:213-217.
Healy D, Whitaker C. Antidepressants and suicide: risk-benefit
conundrums. 2003;28:331-337.
Fergusson D, Doucette S, Glass KC, et al. Association between
suicide attempts and selective serotonin reuptake inhibitors.
2005;330:396-402.
Donovan S, Kelleher MJ, Lambourn J, Foster T. The occurrence of
suicide following the prescription of antidepressant drugs.
1999;5:181-192.
Jick SS, Dean AD, Jick H. Antidepressants and suicide.
1995;310:215-218.
Wessely S, Kerwin R. Suicide risk and SSRIs. 2004;292:379-381.
Jick H, Kaye JA, Jick SS. Antidepressants and the risk of suicidal
behaviors. 2004;292:338-343.
Carey B. FDA expands suicide warning on drugs. ,
May 3, 2007:A17.
J AmAcadChild Psychiatry
BMJ
Psychother Psychosom
PLoS Med
BMJ
Drug Injury:
Liability, Analysis and Prevention.
Wall Street Journal,
Br J Psychiatry
Clinical Trials
Basic Business Statistics: Concepts and
Applications
J Psychiatry Neurosci
BMJ
Arch
Suicide Res
BMJ
JAMA
JAMA
New York Times
12 Journal of American Physicians and Surgeons Volume 14 Number 1 Spring 2009
Posts
No comments:
Post a Comment