LETTERS
ters
TO THE
as part
biology
EDITOR
of the
NIMH
Collaborative
of Depression
Discrepancies
have more ing and
gave
in rates
to say about
local
referral
treatments
than
in persons
with
practices
of the
of suicide
of attempts
differences
across
clinical
in clinical
or panic
disorders
intensity of thanatophobia, ing associated with fear death (4). Thanatophobia,
samples
protocol
screen-
into
the true rate ofsuicide
that the findings further exploration
Psycho-
attempts.
or “self-selection”
they do about mood
We conclude tainby warrant
Study
a history
specific
attempts
in the community.
are not anomalous on their nature
and cerand cause.
REFERENCES I . Beck AT, Steer RA, Sanderson WC, Skeie TM: Panic disorder and suicidal ideation and behavior: discrepant findings in psychiatric outpatients. Am J Psychiatry 1991; 148:1195-1199 2.
Weissman
MM,
Kierman
GL,
Markowitz
JS,
Ouellette
R: Suici-
dal ideation and suicide attempts in panic disorder and attacks. N EnglJ Med 1989; 321:1209-1214 3. Noyes R: Suicide and panic disorder: a review. J Affective Disord 1991; 22:1-11 4. LepineJP, ChignonJM, Teherani M: Suicidal behavior and onset of panic disorder. Arch Gen Psychiatry 1991; 48:668-669 S. Petronis KR, Samuels JF, Moscicki EK, Anthony JC: An epide-
miologic
investigation
of potential
risk factors
for suicide
tempts. Soc Psychiatry Psychiatr Epidemiol 1990; 25:193-1 6. AnthonyJC, Petronis KR: Panic attacks and suicide attempts ter). Arch Gen Psychiatry 1991; 48:1114 7.
FawcertJ,
Schefrner
W, Clark
W: Clinical predictors disorders: a controlled 144:35-40
D, Hedeker
of suicide prospective
D, Gibbons
at99 (let-
R, Coryell
in patients with major affective study. Am J Psychiatry 1987;
MYRNA M. WEISSMAN, PH.D. GERALD L. KLERMAN, M.D. JIM JOHNSON, PH.D. New York, N.Y.
the unfounded conviction of dyof news that provokes thoughts of
as measured by Kellner’s Illness Attitude Scales, was found to be significantly higher in patients with agoraphobia associated with panic attacks, compared to healthy control subjects, and to wane when agoraphobia was treated by behavioral methods (5). Thirty-eight patients (54.3%) of the 70 here reported scored 4 or above for thanatophobia on the phobia item of the Clinical Interview for Depression. It is rather common in various research interviews to start exploring suicidal ideations by asking whether there has been a period when the patient thought a lot about death. The majority of patients with panic disorder and agoraphobia were found to be obsessed by the idea of death, and yet these thoughts are part of hypochondriacal concerns (4). Unlike experienced clinicians, lay interviewers in epidemiobogical studies may be misled by these thoughts and interpret them as part of suicidal tendencies.
REFERENCES I . Noyes R: Suicide and panic disorder: a review. J Affective Disord 1991; 22:1-11 2. Paykel ES: The Clinical Interview for Depression. J Affective Disord 1985; 9:85-96 3. Fava GA, Keliner R, Lisansky J, Park 5, Perini GI, Zielezny M: Rating depression in normals and depressives. J Affective Disord 1986; 11:29-33 4. Fava GA, Grandi 5: Differential diagnosis of hypochondriacal fears and beliefs. Psychother Psychosom 1991; 55:114-119 S. Fava GA, Kellner R, Zielezny M, Grandi 5: Hypochondriacal fears and beliefs in agoraphobia. J Affective Disord 1988; 14: 239-244 GIOVANNI
SILVANA GIANNI
SIR: Dr. Beck and associates reported on the lack of association between panic disorder and suicidal ideation and behavion in clinical populations, in contrast with previous research evidence mainly based on epidemiologicab studies (1). We investigated the occurrence of suicidal ideations and attempts in a consecutive series of 70 outpatients referred to our Affec-
Dr.
tive Disorders Program (48 women and 22 men; mean age=34.2 years, SD=9.S years), five of whom satisfied the DSM-III-R criteria for panic disorder with agoraphobia. Pa-
we
tients
with
concurrent
major
depressive
disorder
were
cx-
eluded. Suicidal tendencies were evaluated by the 7-point scale of Paykel’s Clinical Interview for Depression, a semistructured research interview (2). This scale consists of specific anchor points, from “no suicidal ideations” (score 1 ) through “patient
has
thoughts
of taking
no plans”
(score
4) to “suicidal
kind” 75%
life,
control disorder
subjects and
but
attempt
(score 7). The cut-off point of patients with DSM-III
healthy panic
his
would
not,
of 4 was
found
tant regarding suggest another ing
some
1412
is a rather
cases,
has
minor
to apply
to
melancholia and to 5% of (3). Only one of the 70 patients with
agoraphobia
(1 .4%)
reported
suicidal
dencies, as identified, in the previous month. These thus support those by Dr. Beck and colleagues. Dr. Beck and colleagues discuss several potential tions for the epidemiobogical vious history
and
of any but most
ten-
findings explana-
discrepancies between their clinical study and findings ( 1 ). We agree that undetected or preof mood disorders may be particularly impor-
suicidal behavior. We have data, explanation of suicidal ideations. common it may occur
however, to Fear of dy-
symptom during panic attacks. also at other times and reach
In the
A. FAVA,
M.D.
GRANDI,
M.D.
SAVRON,
M.D.
SANDRA CONTI, CHIARA RAFANELLI, Bologna,
Beck
and Colleagues
M.D. M.D. Italy
Reply
SIR: Although our article was published in September 1991, wrote it shortly after reading Dr. Weissman and associates’ I 989 article ( I ). Therefore, we are indebted to them for describing more recent reviews and studies. We concur that the original
findings
We address
can
no
longer
be
considered
“anomalous.”
their
points below. 1 . Replication. A critical review of the literature is beyond the scope of this better. However, many of the studies cited by Dr. Weissman and colleagues are not so supportive as suggested. For example, the 1 6 studies reviewed in the article by Noyes consist of eight retrospective studies of the relation of neuroses (not panic disorder) to suicide attempts and eight retrospective studies ofthe relation of anxiety states or anxiety neuroses to suicide attempts. Several of these samples included patients whose diagnoses antedated by several decades the inelusion of panic disorder as a separate entity in DSM-III. Also included in this review were six studies of completed suicides, none of which showed a relationship between panic disorder and suicide. 2. Comorbidity and Multiple Diagnoses. To test the issues about comorbidity raised by Dr. Weissman and associates, we identified a sample of 559 (31.2%) of 1,794 outpatients consecutively evaluated according to DSM-III-R criteria at the Cen-
Am
]
Psychiatry
149:1
0, October
1992
LETTERS
ter for Cognitive Therapy between 1986 and August 1991 who represented the types ofdiagnostic combinations that Dr. Weissman and colleagues suggested might be evaluated with respect to past suicide attempts. We excluded patients from 1985 because some of them were diagnosed according to DSM-III entenia. All of the current patients were diagnosed with the SCID, and the information about a past suicide attempt was gathered with the Scale for Suicide Ideation as in our previous study. Instead of focusing on just secondary comorbid diagnoses, we
extended
our scope
None
of the
disorder
for
to tertiary
53
patients
whom
diagnoses. diagnosed
there
was
no
suicide attempt. Five (3.6%) with a primary panic disorder tiary mood disorder described
(6.4%)
of the 310
patients
with
a primary
panic
described
a past
comorbidity
question
EDITOR
of recent suicide attempters, c) examination of individuals currently diagnosed as having panic disorder, d) “psy-
chobogical gitudinal affective
autopsies”
of completed
prospective studies disorder patients.
suicides, of
We appreciate seeing the letter from who point out several possible sources
panic
and
e) Ion-
disorder
and
Dr. Fava and associates of error (fear of death,
hypochondriacal fixations, etc.) in categorizing death-related thoughts as suicidal ideation. Their overall finding that suicidal ideation is not specifically associated with panic disorder raises further questions regarding the issue of panic disorder as a risk factor in suicidal behavior.
of the
140 patients diagnosed either a secondary or tera past suicide attempt. Twenty
and
diagnosed
with
a primary
mood
disorder for whom there was no comorbidity described having made a past suicide attempt, while four (7A%) of the 56 patients diagnosed with a primary mood disorder and either a secondary or tertiary panic disorder described having made a past suicide attempt. In summary, the presence of a mood disorder, whether primary, secondary, or tertiary, was the consistent correlate of previous suicide attempts. 3. Referrals to the Centerfor Cognitive Therapy. Our finding of 7% for lifetime suicide attempts may not be low for an outpatient sample. In the study by Fawcett (cited by Dr. Weissman and colleagues), 80% of the sample consisted of inpatients, many of whom presumably had been hospitalized because of recent suicide attempts.
A major
TO THE
is the ascertainment
of suicide
REFERENCE 1 . Weissman MM, Klerman GL, Markowitz JS, Ouellette R: Suicidal ideation and suicide attempts in panic disorder and attacks. NEnglJMed 1989; 321:1209-1214 AARON T. BECK, M.D. ROBERT A. STEER, ED.D. WILLIAM C. SANDERSON, PH.D. Philadelphia, Pa.
With
Seizure
and that
Low
Clozapine
of
SIR: I refer to a letter to the Editor by Pierre Thomas, M.D., Michel Goudemand, M.D. ( 1 ). From the letter it appears the patient in reference was prescribed clozapine after he
did not respond
attempts.
Doses
to treatment
have determined stantial degree patients cidal
“filtering
hypothesis,”
that our entire sample of patients of suicidality. In fact, 587 (32.7%)
in all diagnostic
categories
presented
with
we
has a subof 1,794 current
sui-
United
States,
the use of cbozapine would have been premature. Use of a medication with the potential for complications that is carried by clozapine should be reserved for cases where alternative
most
treatments
patients
would
have
been
respond
reasonably
exhausted.
to habopenidob,
While
20 mg/day,
some
ideation.
We recommend
that
further
research,
particularly
prospec-
tive studies, should address a number ofmethodological, clinical, and conceptual issues before we can arrive at any definitive conclusions. Our recommendations include the following: 1 . Ascertainment ofsuicide attempts. A more refined definition that restricts “suicide attempts” to instances in which
there
“deliberate
is an
intent
to
self-harm”
or
die
and
classifies
“parasuicide”
other
acts
ditions
that
fective
disorders,
combination,
can
be
associated
multiple personality
with
suicide
attempts:
af-
depression-anxiety (especially
borderline
personality disorder), and substance use disorders. 5. Comprehensive study ofhypotheses. Convergent validity of the hypothesized association between panic and suicide should be established on the basis of a variety of related studies: a) community surveys (as in the studies by Dr. Weissman and colleagues), b) clinical examination
]
Psychiatry
149;1
0, October
I . Thomas (letter).
P, Goudemand AmJ Psychiatry
M: Seizure with low dose 1992; 149:138-139
of clozapine
JOSEPH MORE, Middletown,
M.D. Conn.
is suggested.
diagnoses, disorders
REFERENCE
as
2. Definition ofpanic disorder. The policy of equating diagnoses of “anxiety states” with the diagnosis of “panic disorder” appears to be unjustified. 3. Sequence. Some kind of program needs to be laid out to address the questions of sequence and of proximal versus distal relationships of clinical diagnoses and suicidal attempt. 4. Comorbidity. Data should include all of the clinical con-
Am
in the
If
patients may require higher doses, and if that failed, a trial with another “conventional” neurobeptic would have been in order. I hope that the letter of Drs. Thomas and Goudemand will serve to encourage us to maintain these standards.
the
prevalent
20 mg/day.
suicide
Regarding
standards
habopenidol,
this
attempts.
is so, by current
with
Many individuals report having made “suicide attempts” but acknowledge not having had any wish to die. It is possible that our stringent criteria for ascertainment of lifetime suicidal attempts may account for what appears to be a low rate. In any event, there is no discernible basis for the suggestion that patients referred to our clinic were screened out for previous
1992
Dr. Goudemand
and Dr. Thomas
SIR: We agree with Dr. precautions to be taken treatment.
More’s before
Reply advice about implementation
However, our purpose was to mention effect (seizure with low doses of clozapine) ticular precautions ing dosage. Thus,
our patient idol
changed,
an unusual adverse and consequent par-
(i.e., during EEG) to be taken while increaswe did not go deeper into the reasons why
was prescribed
treatment,
the reserves and of clozapine
psychotic
moreover,
clozapine features
our patient
treatment. and
delusions
experienced
With
haloper-
persisted
incapacitating
un-
and
incorrigible extrapyramidal symptoms. In our clinical expenience such a combination required an increase in dosage, as webb as the use of another “conventional” neuroleptic. In our ward, since 1989 only 10 patients have been pre-
1413