A
 
naturalistic
 
study
 
ofsuicidaladolescents
 
treated
 
with
 
an
 
SSRI:Suicidal
 
ideation
 
andbehavior
 
during
 
3-monthpost-hospitalizationperiod
§
NeeraGhaziuddin
a,
*,ChristopherMerchant
b,1
,
 
RichardDopp
a
,CherylKing
a
a
University
 
of 
 
Michigan,
 
 Ann
 
 Arbor,
 
United
 
States
b
Northeastern
 
Illinois
 
University,
 
Chicago,
 
United
 
States
1.
 
Introduction
The
 
overall
 
aimofthe
 
present
 
study
 
was
 
to
 
describe
 
suicidalideation
 
and
 
suicide-related
 
and
 
other
 
emergencies
 
during
 
a
 
3-month
 
post-hospitalization
 
period
 
among
 
depressed
 
and
 
suicidaladolescents
 
who
 
were
 
receiving
 
SSRIs.
 
A
 
secondary
 
aim
 
ofthestudy
 
was
 
to
 
describe
 
variability
 
inoutcome
 
among
 
subgroups
 
of participants
 
receiving
 
SSRIs
 
alone
 
or
 
in
 
combination
 
withantipsychotic
 
and/or
 
an
 
anti-epileptic
 
mood
 
stabilizer
 
medication.Theclinical
 
significance
 
ofthis
 
study
 
is
 
considerable
 
becausesuicidal
 
thoughts
 
and/or
 
behaviors
 
are
 
common
 
amongdepressed
 
adolescents
 
2009),and
 
treatmentofthese
 
adolescents
 
has
 
become
 
increasingly
 
challenging
 
duetoconcerns
 
of 
 
increased
 
suicidal
 
ideation
 
and/or
 
behavior
 
inassociation
 
with
 
SSRIs.
 
Published
 
literature
 
to
 
date
 
hasaddressed
 
several
 
important
 
questions
 
related
 
to
 
theassociation
 
between
 
suicidal
 
ideation
 
and/or
 
behaviors
 
and
 
SSRIs.However,
 
few
 
studies
 
have
 
examined
 
associations
 
betweenSSRIs
 
and
 
suicidal
 
thoughts
 
and
 
behaviors
 
among
 
adolescentswhoare
 
already
 
suicidal.
 
Furthermore,
 
there
 
is
 
scantinformation
 
regarding
 
these
 
outcomes
 
when
 
SSRIs
 
are
 
concur-rentlyprescribed
 
with
 
antipsychotics
 
and/or
 
anti-epilepticmoodstabilizers
 
to
 
depressed
 
and
 
simultaneously
 
suicidaladolescents.
A
 
 
T
 
I
 
C
 
LE
 
I
 
N
 
F
 
O
 Articlehistory:
Received
 
11
 
November
 
2013
Receivedinrevisedform22March2014
Accepted
 
31
 
March
 
2014
Availableonlinexxx
Keywords:
AdolescentsSuicideSSRI
A
 
BS
 
T
 
RA
 
C
 
T
Objective:
Describe
 
suicidalideationandsuiciderelated/otheremergencies
 
(SRE),amongdepressedandacutelysuicidaladolescentsduringa3-monthperiodfollowingpsychiatrichospitalization.
Methods:
Onehundred
 
twentyadolescents,whowereboth
 
depressedand
 
suicidal,
 
were
 
receivinganSSRIeitheralone
 
orin
 
combinationwithothermedications,remainedonaconsistent
 
medicationregimenbetweenbaselineandat3-monthsandtheir3-monthoutcomedatawereavailable.Theparticipantswere
 
dividedinto
 
fourmedicationgroups:SSRIantidepressant
 
only
 
(
n
=71);SSRI
 
plusmoodstabilizer
 
(
n
 
=
 
17);SSRIplusantipsychotic(
n
 
=20);andSSRIplus
 
antipsychoticandmoodstabilizer(
n
=
 
12).
 
Standardizedinstrumentswereused.
Results:
Meanage=
 
15.5
 
1.3,
 
Caucasian=
 
80.8%,female=
 
74.2%,
 
meanCDRS-R 
 
=
 
61.7
12.1,
 
suicideattempt
 
duringmonth
 
prior
 
tohospitalization
 
=
 
58.6%.
 
During
 
the
 
3-month
 
post-hospitalization
 
period:
 
(1)there
 
wereno
 
suicides,sixparticipants
 
(5%)attempted
 
suicide
 
and
 
21
 
(17.5%)experienced
 
anSRE;
 
(2)declinein
 
suicidalideation
 
anddepressionseveritywasnoted;
 
(3)
 
SSRI
 
plus
 
anantipsychoticgroupreported
 
thehighest
 
number
 
of 
 
SREs;
 
(4)higher
 
baselinehopelessness
 
andaggression
 
scoreswereassociatedwith
 
greaterreduction
 
in
 
suicidal
 
ideation
 
at
 
3-months.
Conclusion:
Declinesinsuicidalideation,depressionseverity,and
 
suicide
 
attemptswere
 
noted,irrespectiveofpsychotropic-combinationreceived.A
 
higherrateof 
 
SREswasassociatedwith
 
receivinganantipsychoticagentincombinationwithanSSRI.Givennaturalisticdesignofstudy,cause-effectconclusionscannotbedrawn.Thelackofan
 
objectivemeasureto
 
identifymedicationadherenceisastudylimitation.
 
2014ElsevierB.V.All
 
rightsreserved.
§
The
 
study
 
was
 
supported
 
by
 
theGrant
 
NIMH
 
R01
 
MH63881
 
andK24MH077705.* Corresponding
 
author.
 
Tel.:
 
+1
 
734
 
764
 
7562;
 
fax:
 
+1
 
734
 
936
 
8907.
E-mail
 
addresses:
 
(N.
 
Ghaziuddin),C-Merchant@neiu.edu
(C.Merchant),
 
(R.
 
(C.King).
1
He
 
was
 
adoctoral
 
student
 
at
 
the
 
University
 
ofMichigan
 
when
 
this
 
study
 
wasconducted.
G
 
Model
AJP-582;
 
No.
 
of 
 
Pages
 
7
Please
 
cite
 
this
 
article
 
in
 
press
 
as:
 
Ghaziuddin,
 
N.,
 
et
 
al.,
 
Anaturalistic
 
study
 
of 
 
suicidal
 
adolescents
 
treated
 
with
 
an
 
SSRI:
 
Suicidalideation
 
and
 
behavior
 
during
 
3-month
 
post-hospitalization
 
period.
 
Asian
 
 J.Psychiatry
 
(2014),
 
Contents
 
lists
 
available
 
at
 
AsianJournalo
 
Psychiatry
jou
 
rn
 
al
 
h
 
om
 
epage:
 
 
2014
 
Elsevier
 
B.V.
 
All
 
rights
 
reserved.
 
Suicidality
 
(completed
 
suicide,
 
suicide
 
attempt,
 
preparationtoward
 
imminent
 
suicidal
 
behavior,
 
suicide
 
ideation)
 
is
 
aproblemintimately
 
associated
 
with
 
adolescent
 
depression
 
and
 
is
 
aleadingcause
 
of 
 
death
 
in
 
this
 
age
 
group,
 
ranked
 
third
 
in
 
the
 
United
 
Statesbehind
 
accidents
 
and
 
malignancies
 
2006).Treatment
 
of 
 
adolescent
 
depression
 
has
 
beencomplicated
 
by
 
aconcern
 
raised
 
in
 
2003
 
regarding
 
a
 
possible
 
association
 
betweensuicidality
 
and
 
SSRIs,
 
which
 
was
 
followed
 
by
 
ablack-box
 
warningissued
 
in
 
2004
 
by
 
the
 
Food
 
and
 
DrugAdministration
 
(FDA).Thewarning
 
states
 
that
 
‘‘antidepressants
 
increased
 
the
 
risk
 
forsuicidal
 
thinking
 
and
 
behavior
 
in
 
children,
 
adolescents,
 
andyoung
 
adults
 
in
 
short-term
 
studies
 
of 
 
major
 
depressive
 
 
This
 
regulatory
 
decision
 
by
 
the
 
FDA
 
was
 
based
 
onameta-analysis
 
of 
 
24trials
 
involving
 
over
 
4500
 
patients,which
 
found
 
agreater
 
risk
 
ofsuicidality
 
(risk
 
ratio
 
=
 
1.95;
 
95%CI=
 
1.28–2.98)
 
during
 
the
 
first
 
few
 
months
 
of 
 
initiating
 
antide-pressant
 
treatment
 
2006).A
 
summary
 
of 
 
relevant
 
publications
 
related
 
to
 
this
 
suspectedassociation
 
between
 
antidepressants
 
and
 
suicidality
 
is
 
as
 
follows:children
 
and
 
adolescents
 
treated
 
with
 
an
 
SSRI
 
may
 
experience
 
amodest
 
increase
 
in
 
suicidal
 
ideation
 
and
 
behaviors
 
 
noted
 
in
 
ameta-analysis
 
of 
 
24
 
trials
 
(the
 
majority
 
sponsored
 
by
 
pharmaceu-ticalcompanies
 
with
 
the
 
exception
 
of 
 
the
 
TADS
 
multicenter
 
trial)(Hammad
2006).However,
 
another
 
meta-analysis
 
of 
 
27
 
trialsinvolving
 
adolescents
 
with
 
MDD,
 
non-OCD
 
anxiety
 
or
 
OCD(trialsconducted
 
between
 
1988
 
and
 
2006),
 
found
 
only
 
a
 
non-significantincreased
 
risk
 
in
 
suicidal
 
ideation
 
and
 
behaviors
 
among
 
those
 
whohad
 
received
 
an
 
SSRI
 
compared
 
with
 
placebo,
 
irrespective
 
of diagnosis
 
2007).
 
A
 
case-control
 
study
 
of 
 
medicaidrecipients
 
who
 
were
 
depressed,
 
found
 
that
 
children
 
and
 
adoles-cents
 
receiving
 
an
 
SSRI,
 
but
 
notadults,
 
are
 
more
 
likely
 
to
 
attemptsuicide
 
(OR 
 
1.52;
 
95%
 
CI)
 
or
 
die
 
by
 
suicide
 
(OR 
 
15.62;
 
95%CI).Several
 
explanations
 
havebeen
 
offered
 
about
 
thisassociation.
 
Oneof 
 
these
 
was
 
the
 
result
 
of 
 
a
 
re-analysis
 
of 
 
randomized
 
controlledstudies
 
of 
 
fluoxetine
 
in
 
youth,
 
which
 
found
 
that
 
althoughdepressive
 
symptoms
 
improved
 
following
 
fluoxetine,
 
suicidalthoughts
 
or
 
behaviors
 
did
 
notimprove
 
in
 
thisage
 
group
 
2004).Another
 
study
 
suggested
 
that
 
adolescents
 
who
 
engage
 
insuicidal
 
behaviors
 
had,
 
in
 
fact,
 
not
 
received
 
adequate
 
treatmentwith
 
antidepressants
 
2010).Along
 
similar
 
lines,
 
oneother
 
study
 
found
 
that
 
pre-treatment
 
patient
 
characteristicslargelyaccounted
 
for
 
increased
 
suicidality
 
2011),while
 
others
 
have
 
stated
 
that
 
the
 
studies
 
which
 
formed
 
the
 
basis
 
of the
 
FDA
 
regulatory
 
decision
 
were
 
limited
 
due
 
to
 
the
 
exclusion
 
of suicidal
 
adolescents
 
in
 
clinical
 
trials
 
2009)Inaddition
 
to
 
the
 
concern
 
regarding
 
the
 
safety
 
of 
 
antidepres-sants,
 
the
 
FDA
 
recently
 
introduced
 
a
 
warning
 
regarding
 
anincreased
 
risk
 
of 
 
suicidal
 
thoughts
 
and
 
behaviors
 
associated
 
withseveral
 
anti-epileptic
 
drugs
 
(AEDs).
 
Thisrisk
 
involving
 
AEDs
 
ispertinent
 
to
 
psychiatrists
 
because
 
these
 
agents
 
are
 
often
 
pre-scribed
 
as
 
mood
 
stabilizers.
 
Although,
 
thiswarning
 
does
 
notspecifically
 
involve
 
adolescents,
 
it
 
adds
 
to
 
the
 
concern
 
due
 
to
 
theoverall
 
higher
 
risk
 
of 
 
suicidality
 
among
 
adolescents.
 
The
 
FDA’swarning
 
regarding
 
AEDs
 
was
 
based
 
on
 
a
 
review
 
of199trialsinvolving
 
eleven
 
different
 
agents;
 
four
 
participants
 
receiving
 
anactive
 
agent
 
had
 
died
 
by
 
suicide.
 
Because
 
of 
 
concerns
 
involvingboth
 
SSRIs
 
and
 
AEDs,
 
it
 
is
 
pertinent
 
to
 
examine
 
the
 
effect
 
of 
 
AED-moodstabilizers
 
on
 
suicidal
 
ideation
 
and
 
behavior
 
and
 
whetherthese
 
agents
 
may
 
add
 
to
 
the
 
risk
 
ofsuicidality,
 
when
 
received
 
inconjunction
 
with
 
SSRIs.
 
Thisclass
 
of 
 
agents
 
(valproate,
 
carbamaz-epine,
 
oxycarbmazepine
 
(to
 
name
 
a
 
few)
 
is
 
often
 
prescribed
 
formood
 
stabilization
 
to
 
adolescents
 
who
 
either
 
do
 
not
 
respond
 
toSSRIs
 
alone,
 
or
 
present
 
with
 
symptoms
 
that
 
may
 
be
 
responsive
 
tothese
 
agents.Are-examination
 
of 
 
the
 
suicide
 
rates
 
during
 
the
 
past
 
decade
 
hasshown
 
apuzzling
 
increase
 
with
 
the
 
United
 
States
 
(US)
 
experienc-ing
 
its
 
largest
 
ever
 
year-to-year
 
increase
 
(14%
 
between
 
2003
 
and2004)
 
since
 
the
 
Center
 
for
 
Disease
 
Control
 
(CDC)
 
firststartedsystematic
 
collection
 
ofsuicide
 
data
 
in
 
1979.
 
Similarly,
 
theNetherlands
 
had
 
experienced
 
adramatic
 
increase
 
of 
 
49%for
 
thesame
 
period
 
2007).Further,
 
independent
 
research-ersfound
 
that
 
the
 
rise
 
in
 
adolescent
 
suicide
 
rate
 
for
 
period
 
2003–2004
 
remained
 
largely
 
unchanged
 
for
 
2005
 
2008),which
 
bolstered
 
the
 
initial
 
concerns
 
that
 
the
 
increased
 
suicidalitynoted
 
during
 
2003–2004
 
was
 
possibly
 
associated
 
with
 
achange
 
inantidepressant-prescribing
 
practice.
 
However,
 
adecline
 
in
 
thesuicide
 
rate
 
in
 
2006
 
and
 
2007
 
to
 
6.9
 
per
 
100,000
 
has
 
beenreassuring.
 
Nevertheless,
 
longer
 
term
 
data
 
are
 
necessary
 
tounderstand
 
these
 
fluctuations
 
and
 
whether
 
the
 
FDA
 
warninghad
 
indeed
 
resulted
 
in
 
atemporary
 
change
 
in
 
clinical
 
practice
 
thatwasassociated
 
with
 
an
 
increased
 
suicide
 
rate
 
in
 
the
 
US
 
and
 
abroad.Clearly,
 
the
 
pharmacotherapy
 
of 
 
adolescent
 
depression
 
continuestobe
 
fraught
 
with
 
concerns.Although,
 
it
 
is
 
beyond
 
the
 
scope
 
of 
 
this
 
study
 
to
 
clarify
 
theoverall
 
association
 
between
 
suicidality
 
and
 
SSRIs,
 
we
 
describe
 
heresuicide-related
 
events
 
(SREs)
 
among
 
depressed
 
adolescents
 
whowere
 
acutely
 
suicidal
 
at
 
study
 
entry
 
and
 
were
 
receiving
 
aSSRIduring
 
a
 
3-month
 
period
 
following
 
apsychiatric
 
hospitalization.Additionally,
 
in
 
keeping
 
with
 
our
 
secondary
 
objective,
 
we
 
examinevariability,
 
if 
 
any,
 
in
 
SRE-rates
 
among
 
subgroups
 
ofadolescentsreceiving
 
an
 
SSRI
 
alone,
 
or
 
in
 
combination
 
with
 
other
 
psychotropicmedications.
2.Methods
 2.1.
 
Participants
 
and
 
study
 
design
Data
 
were
 
extractedfromhospitalizedsubjectswho
 
hadparticipated
 
in
 
an
 
IRBapprovedpsychosocialinterventionstudy(YouthNominatedSupport
 
Team
 
II,fundedbyNationalInstitute
 
ofMental
 
Health
 
[R01MH63881];
n
=
 
448).Theinclusioncriteria
 
fortheparent
 
study
 
were:
 
(1)arecentsuicideattempt(during
 
the
 
pastmonth),orsevere,unrelentingsuicidalideation,
 
(2)
 
ages13–17years,
 
and
 
(3)parent
 
or
 
guardianinformedconsent
 
andadolescentassent.
 
Exclusion
 
criteriawere:
 
(1)severe
 
cognitive
 
impairmentand/oractivepsychosis,(2)direct
 
transferto
 
medical
 
unit,
 
(3)transfertoresidentialplacement,(4)lived
>
40
 
miles
 
oronehour
 
driving
 
time
 
fromhospital(King
al.,2009).Fromthislarger
 
grouprecruitedfortheparent
 
study,additionaldata
 
were
 
extractedfor
 
asubgroup(
n
=
 
120)that
 
metthecriteria
 
forthepresentstudy:
 
(1)diagnosisof 
 
anydepression,
 
based
 
on
 
theSchedulefor
 
AffectiveDisordersandSchizophreniaforSchool-AgeChildren
 
(definite,partiallyremittedor
 
probable
 
diagnoses
 
formajordepression,dysthymia,or
 
depressivedisorder
 
NOS)(KSADS-PL;Kaufmanetal.,1997);
 
(2)receiving
 
an
 
SSRI(either
 
aloneor
 
incombinationwith
 
other
 
psychotropic
 
agents,
 
butnotanothertypeo
 
antidepressant;i.e.,
 
tricyclic
 
antidepressant
 
and/orMAOI);(3)receiving
 
a
 
consistentmedication
 
regimen
 
duringthe3-month
 
follow-upstudyperiod.Although
 
220participantsinitiallymetthe
 
inclusioncriteria
 
for
 
thepresentstudy,
 
75(34.1%)
 
were
 
excluded
 
from
 
thefollow-up
 
analysesbecause
 
of lackof3-month
 
data,while25
 
(11.4%)
 
were
 
examinedseparatelybecause
 
of 
 
changesmade
 
to
 
theirmedicationregimen.
 
Reasonswhy
 
or
 
thetimingwhen
 
the
 
SSRI,moodstabilizeror
 
an
 
antipsychotic
 
wereinitiated
 
were
 
notrecordedduring
 
thedata
 
collection.Alsonotavailable
 
were
 
theage
 
of onset
 
ofthedepressivedisorder.
N.Ghaziuddin
 
et 
 
al.
 
 / 
 
 Asian
 
 Journal
 
of 
 
Psychiatry
 
 xxx
 
(2014)
 
 xxx–xxx
2
GModel
AJP-582;
 
No.
 
of 
 
Pages
 
7
Please
 
cite
 
this
 
article
 
in
 
press
 
as:
 
Ghaziuddin,
 
N.,
 
etal.,
 
Anaturalistic
 
study
 
of 
 
suicidal
 
adolescents
 
treated
 
with
 
an
 
SSRI:Suicidalideation
 
and
 
behavior
 
during
 
3-month
 
post-hospitalization
 
period.
 
Asian
 
 J.Psychiatry
 
(2014),
 
 
The75participantslost
 
tofollow-upat
 
3monthswerecomparedto
 
the145adolescents
 
with
 
3-monthfollow-updata.Nosignificantdifferenceswere
 
foundbetween
 
the
 
retainedandthe
 
lost-to-follow-up
 
groupsinterms
 
ofdemographics,
 
thetypeor
 
theseverityo
 
psychiatricsymptoms.Of 
 
the145adolescents
 
forwhom
 
wehadfollow
 
up
 
data,25
 
didnotremainin
 
thesame
 
medicationgroupthattheyhadstartedin.Therefore,thefinalgroup
 
consistedof 
 
120
 
who
 
remainedonaconsistent
 
treatmentregimen,
 
in
 
additiontomeetingotherstudyrelatedcriteria.These
 
final
 
120
 
adolescents
 
were
 
grouped
 
on
 
the
 
basis
 
of 
 
otherpsychotropic
 
medications
 
received
 
atstudy
 
entry,
 
in
 
addition
 
toanSSRI:
 
SSRIs
 
alone
 
(SSRI;
 
n
=
 
71),
 
SSRI
 
combined
 
with
 
amoodstabilizer
 
(SSRI
 
+
 
MS;
 
n
 
=
 
17),
 
SSRI
 
combined
 
with
 
an
 
antipsychotic(SSRI
 
+AP;
 
n
 
=
 
20)
 
and
 
SSRI
 
combined
 
with
 
amood
 
stabilizer
 
andanantipsychotic
 
(SSRI
 
+
 
BOTH;
 
n
 
=12).
 
Adolescents
 
were
 
notexcluded
 
based
 
on
 
their
 
usage
 
of 
 
ADHD
 
medication
 
(stimulant
 
ornon-stimulant)
 
or
 
a
 
benzodiazepine.
 2.2.
 
Standardized
 
measures
All
 
standardized
 
instruments
 
described
 
below,
 
except
 
forthe
 
Services
 
Assessment
 
Record
 
(SARS;
 
andtheCritical
 
Incident
 
Check
 
(CIC;
 
 
unpublished)
 
werecompleted
 
at
 
study
 
entry.
 
These
 
two
 
instruments
 
(CIC
 
and
 
theSARS)
 
were
 
only
 
completed
 
at
 
3months.
 
Instruments
 
that
 
wererepeated
 
at
 
3
 
months
 
included
 
the
 
Children’s
 
Depression
 
Scale-Revised
 
 
and
 
the
 
Suicidal
 
IdeationQuestionnaire
 
1988).The
 
KSADS-PL 
 
 
isa
 
semi-structuredclinical
 
interview,
 
designed
 
to
 
assess
 
the
 
presence
 
of 
 
DSM-IVmental
 
disorders.
 
Interviewers
 
were
 
trained
 
mental
 
healthprofessionals
 
who
 
completed
 
20
 
hours
 
of 
 
KSADS-PL 
 
training
 
andestablished
 
inter-rater
 
reliability
 
with
 
asenior
 
diagnostician
 
onfour
 
consecutive
 
interviews
 
prior
 
to
 
study
 
onset.The
 
Children’s
 
Depression
 
Rating
 
Scale-Revised
 
(CDRS-R;Poznanski
1996)is
 
asemi-structured
 
interviewconducted
 
with
 
adolescents
 
to
 
assess
 
the
 
severity
 
of 
 
depressivesymptoms
 
within
 
the
 
previous
 
two
 
weeks.
 
Inter-rater
 
reliability
 
inthe
 
study
 
was
 
high
 
(mean
 
alpha
 
across
 
raters
 
=
 
98).TheBeck
 
Hopelessness
 
Scale
 
(BHS;
 
1988):
 
This
 
is
 
aself-report
 
measure
 
that
 
has
 
been
 
shown
 
to
 
have
 
good
 
internalconsistency
 
(with
 
a
 
KR-20
 
coefficient
 
of 
 
0.86),
 
strong
 
concurrentvalidity
 
(indicated
 
by
 
apositive
 
correlation
 
with
 
depression[
 
=0.53],
 
and
 
a
 
negative
 
correlation
 
with
 
reasons
 
for
 
living[
=
0.65]
 
among
 
adolescent
 
psychiatric
 
inpatients.The
 
Suicidal
 
Ideation
 
Questionnaire-Junior
 
(SIQ-JR;
 
 
This
 
is
 
a15-item
 
self-report
 
questionnaire
 
for
 
assessing
 
thefrequency
 
of 
 
suicidal
 
thoughts.
 
Higher
 
total
 
scores
 
indicate
 
higherseverity
 
of 
 
suicidal
 
ideation.
 
Total
 
scores
 
on
 
the
 
SIQ-JR 
 
haveshownhigh
 
internal
 
consistency
 
(
a
 =
 
0.91)
 
and
 
test–retest
 
reliability(
tt 
 =0.91)
 
1999).TheYouth
 
Self-Report(YSR;Achenbach,
1991):
 
Thisisa
 
112-itemscalethat
 
assesses
 
a
 
broad
 
range
 
ofpsychopathology.
 
Ithastwo
 
broadbandscales(internalizing
 
and
 
externalizing)thathavedemonstratedgood
 
construct
 
validityandcorrelationswithrelatedself-report
 
 
Forthesample
 
fromwhich
 
theseparticipants
 
wereselected(YST–IIstudy),internalconsistencywas
 
0.89fortheinternalizingscale
 
and0.86fortheexternalizing.Additionally,thereareeightnarrowbandsubscales(social
 
problems,thoughtproblems,attention
 
problems,
 
withdrawn,somaticcomplaints,anxious/depressed,delinquentbehavior,andaggres-sivebehavior).The
 
Multidimensional
 
Anxiety
 
Scale
 
for
 
Children
 
(MASC;
 
1997):
 
This
 
is
 
a
 
39-item
 
rating
 
scale,
 
completed
 
by
 
achild
 
oran
 
adolescent
 
toassess
 
abroad
 
spectrum
 
ofanxiety
 
symptoms.Each
 
item
 
is
 
scored
 
on
 
afour-point
 
Likert
 
scale.
 
The
 
MASC
 
hasdemonstrated
 
excellent
 
test–retest
 
reliability,
 
with
 
correlationcoefficients
 
among
 
adolescents
 
ranging
 
from
 
0.75
 
for
 
thePerfectionism
 
subscale
 
to
 
0.90
 
for
 
the
 
tense
 
subscale
 
1999).The
 
Childand
 
Adolescent
 
SocialandAdaptive
 
FunctioningScale(CASAFS;
 
 
a24
 
item
 
self-reportinventory
 
developedspecificallyto
 
examine
 
socialandadaptivefunctioning.Findingssupport
 
thescale’sreliability,
 
validity,
 
andsensitivitytothe
 
impactofa
 
depressivedisorder.
 
Subscalesincludeschool
 
performance,peerrelationships,family
 
relation-ship,andhome
 
duties/self-care.
 
Higherscores
 
indicatebetterfunctioning.The
 
Services
 
Assessment
 
Record
 
Parent
 
Interview
 
(SAR;Preuss
 
and
 
King,
 
unpublished)
 
was
 
specifically
 
developed
 
to
 
assesstreatment
 
adherence
 
during
 
intervention
 
and
 
follow-up
 
periods.Its
 
structure
 
and
 
format
 
was
 
adapted
 
from
 
the
 
Services
 
Assessmentfor
 
Children
 
and
 
Adolescents
 
 
Parent
 
Interview
 
al.,2004).The
 
SARS
 
was
 
used
 
to
 
obtain
 
information
 
at
 
3-monthfollow-up
 
assessment
 
regarding
 
medications,
 
medication
 
changes,psychosocial
 
treatments,
 
use
 
of 
 
intensive
 
services
 
(e.g.,
 
psychiatrichospitalizations),
 
and
 
number
 
of 
 
appointments
 
attended
 
formedication
 
and/or
 
psychosocial
 
treatment.The
 
Critical
 
Incidents
 
Checklist
 
 
unpublished)
 
isa15-item
 
self-report
 
measure
 
about
 
the
 
presence
 
and
 
number
 
of 
 
avariety
 
of 
 
adverse
 
events
 
for
 
the
 
interval
 
since
 
the
 
participant’s
 
lastcontact
 
with
 
the
 
study
 
team.
 
This
 
measure
 
includes
 
questionspertaining
 
to
 
history
 
of 
 
abuse
 
with
 
responses
 
coded
 
as
 
a
 
‘‘yes/no’’.The
 
Diagnostic
 
Interview
 
Schedule
 
for
 
Children-IV
 
(DISC-IV;Shaffer
 
records
 
suicide
 
attempts
 
in
 
the
 
past
 
year
 
andtheadolescent’s
 
lifetime.Suicide
 
Related
 
and
 
other
 
Emergency
 
(SRE)
 
is
 
a
 
compositecritical
 
incident
 
variable,
 
which
 
was
 
created
 
for
 
assessment
 
at
 
3-month
 
follow-up.
 
This
 
variable
 
is
 
based
 
on
 
aparent’s
 
or
 
theadolescent’s
 
report
 
about
 
asuicide
 
attempt,
 
psychiatric
 
hospitali-zation
 
(forany
 
reason),
 
or
 
suicide-related
 
visit
 
to
 
the
 
emergencydepartment.
 
This
 
variable
 
was
 
created
 
from
 
items
 
on
 
the
 
CICandtheDISC-IV.
 
It
 
is
 
important
 
to
 
note
 
that
 
the
 
YST-II
 
intervention(parent
 
study)
 
was
 
not
 
associated
 
with
 
areduction
 
in
 
adolescentsuicide
 
attempts
 
2009).
 2.3.
 
Data
 
analyses
 
and
 
results
Analysesofbaselineclinicalfeatureswere
 
conducted
 
forthe220participantswho
 
hadmetthe
 
studycriteriainitially(regardless
 
of3-month
 
outcomedataor
 
consistency
 
ofmedica-tioncombination).Withinthis
 
sample,themedicationgroupsdifferedsignicantlyonbaseline
 
aggressivebehavior
 
(
[3]=
 
2.713,
 
 p
 
=0.046)andbaselinethoughtproblems
 
(
[3]=
 
5.631,
 
 p
 
=0.001)identified
 
on
 
YSR 
 
subscales.Participantswhowere
 
prescribedSSRI+
 
AP
 
reportedsignificantlygreaterthoughtproblemsthanthoseprescribedSSRI
 
alone
 
or
 
SSRI
 
+
 
MS;however,thisgroup(SSRI
 
+
 
AP)
 
didnotsignificantly
 
differfromSSRI
 
+
 
BOTH.
 
Problems
 
with
 
aggressivebehavior
 
were
 
signifi-cantlymore
 
common
 
among
 
thoseprescribed
 
SSRI+
 
BOTH,thanamong
 
thoseprescribedan
 
SSRI
 
alone.There
 
were
 
no
 
other
 
groupdifferenceson
 
theYSR.Additionally,therewere
 
no
 
medicationgroupdifferences
 
for
 
theMASC
 
subscales,BHS,SIQ-JR,
 
CASAFS
 
orCDRS-R.Itshouldbe
 
noted
 
thatthesegroupdifferences(relatedtothoughtproblemsoraggression)were
 
notdetectedamong
 
the120whowere
 
furtherselectedforthe3-monthfollow
 
upanalyses;thiswas
 
possiblyrelatedtothesmallernumberoparticipants
 
ineach
 
group
 
andreducedstatisticalpower.
 
Forthepurposeof 
 
thisstudy,thisgroupo
 
120is
 
designatedas
 
the‘‘consistentmedication
 
sample.’’
N.Ghaziuddin
 
et 
 
al.
 
 / 
 
 Asian
 
 Journal
 
of 
 
Psychiatry
 
 xxx(2014)
 
 xxx–xxx
 
3
G
 
Model
AJP-582;
 
No.
 
of 
 
Pages
 
7
Please
 
cite
 
this
 
article
 
in
 
press
 
as:
 
Ghaziuddin,
 
N.,
 
et
 
al.,
 
Anaturalistic
 
study
 
of 
 
suicidal
 
adolescents
 
treated
 
with
 
an
 
SSRI:
 
Suicidalideation
 
and
 
behavior
 
during
 
3-month
 
post-hospitalization
 
period.
 
Asian
 
 J.Psychiatry
 
(2014),
 
 
 2.4.
 
Consistent 
 
medication
 
sample
There
 
were
 
120
 
participants
 
who
 
continued
 
to
 
be
 
prescribedthesame
 
medication
 
combinations
 
during
 
the
 
3-month
 
post-hospitalization
 
period
 
(consistent
 
medication
 
groups).
 
Chi-squareand
 
ANOVA
 
analyses
 
indicated
 
no
 
difference
 
by
 
medication
 
groupfor
 
any
 
demographic
 
variables.
 
Among
 
this
 
group,
 
the
 
majority(90.8%)
 
was
 
participating
 
in
 
individual,
 
family,
 
or
 
group
 
therapy
 
atbaseline;
 
with
 
an
 
average
 
of 
 
5.2
 
sessions
 
between
 
baseline
 
and
 
3-month
 
follow-up.
 
There
 
were
 
no
 
differences
 
between
 
themedication
 
groups
 
in
 
either
 
the
 
proportion
 
of 
 
adolescents
 
intherapy
 
or
 
inthe
 
average
 
number
 
of 
 
sessions
 
received.
 
See
 
1for
 
racial/ethnic
 
breakdown,
 
parental
 
income,
 
education
 
andpsychotherapy
 
information
 
for
 
the
 
sample
 
and
 
by
 
medicationgroup.
-Test
 
and
 
chi-square
 
analyses
 
revealed
 
no
 
differences
 
inage,sex,
 
or
 
race
 
between
 
the
 
120
 
(consistent
 
medication
 
group)adolescents
 
and
 
the
 
25
 
who
 
were
 
inconsistent
 
(non-consistentmedication
 
group;
 
medication
 
changes
 
weremade
 
at
 
request
 
of the
 
patient,
 
the
 
family
 
or
 
were
 
physician-directed).
 
Additionally,these
 
groups
 
did
 
not
 
differ
 
on
 
past
 
history
 
of 
 
physical
 
[
x
2
(3)
 
=
 
2.77,
 p
=
 
0.429]
 
or
 
sexual
 
abuse
 
[
x
2
(3)
 
=
 
6.13,
 
 p
 
=
 
0.106]
 
orbypsychiatric
 
hospitalization
 
in
 
either
 
parent.
 
However,
 
participantsin
 
the
 
non-consistent
 
medication
 
group
 
were
 
more
 
likely
 
to
 
be
 
infamilies
 
receiving
 
public
 
assistance
 
(16.7%
 
vs.
 
3.4%;
 
x
2
[1]
 
=
 
6.53,
 p
=
 
0.011)
 
and
 
scored
 
significantly
 
higher
 
on
 
MASC
 
total
 
scores
 
of adolescent
 
anxiety
 
(
 
[142]
 
=
 
2.32,
 
 p
 
=
 
0.022).
 
The
 
mean
 
MASCscore
 
for
 
the
 
non-consistent
 
group
 
was
 
55.2
 
compared
 
with
 
45.7forthe
 
consistent
 
group.
 
There
 
were
 
no
 
other
 
differences
 
onany
 
other
 
measures
 
ofpsychopathology
 
(SIQ-JR,
 
CDRS-R,
 
BHS
 
andYSR-INT,
 
YSR-EXT).
 
Further,
 
there
 
were
 
no
 
differences
 
between
 
theconsistent
 
and
 
non-consistent
 
groups
 
in
 
terms
 
of 
 
the
 
presence
 
of SREs
 
(attempts
 
or
 
hospitalizations)
 
at
 
the
 
3-month
 
time-point.
 2.5.Differences
 
within
 
consistent-medication
 
 groups
2presents
 
baseline
 
clinical
 
features
 
for
 
the
 
entireconsistent
 
medication
 
sample,
 
including
 
past
 
history
 
of 
 
suicideattempts,
 
and
 
features
 
for
 
eachmedication
 
group
 
within
 
thislargergroup.
 
No
 
significant
 
group
 
differences
 
were
 
found
 
in
 
baselineclinical
 
features
 
or
 
suicide
 
attempt
 
history.
 
3presents
 
co-morbid
 
diagnoses
 
for
 
the
 
overall
 
consistent
 
medication
 
group
 
andfor
 
each
 
medication
 
sub-group.
 
Specifically,
 
we
 
report
 
on
 
thepresence
 
of 
 
any
 
anxiety
 
disorder,
 
PTSD
 
or
 
acute
 
stress
 
disorder,any
 
disruptive
 
behavior
 
disorder,
 
and
 
anysubstance
 
use
 
disorder.Chi-square
 
analyses
 
revealed
 
no
 
differences
 
in
 
the
 
presentation
 
of comorbid
 
diagnoses
 
between
 
the
 
medication
 
groups.
 2.6.
 
 3-Month
 
critical
 
incidents
 
and
 
outcomes
4
presents
 
descriptive
 
statistics
 
for
 
Suicide
 
RelatedEmergencies
 
(SREs),
 
3-month
 
SIQ-JR 
 
scores,
 
and
 
the
 
3-monthchange
 
in
 
SIQ-JR 
 
scores
 
for
 
eachof 
 
the
 
consistent
 
medication
 
subgroups
 
and
 
the
 
overall
 
group.
 
For
 
the
 
overall
 
group
 
and
 
within
 
eachgroup,
 
there
 
was
 
a
 
decline
 
in
 
the
 
mean
 
3-month
 
SIQ-JR 
 
scoresrelative
 
to
 
atstudy
 
entry
 
(SIQ-JR 
 
for
 
overall
 
group
 
=47.15
 
 
20.04and
 
post
 
SIQ-JR 
 
at
 
3
 
months
 
foroverall
 
group
 
=21.75
 
16.25;
 
seeTables
2
4
for
 
details).
 
Similarly
 
suicide
 
attempts
 
reported
 
by
 
the
 Table
 
1
Demographic
 
characteristics
 
of 
 
sample:
 
overall
 
and
 
bymedication
 
group.Overall
n
 
=
 
120SSRI
n
 
=
 
71MS
n
 
=17AP
n
 
=20BOTH
n
 
=12Sex(%female)
 
74.2
 
74.682.4
 
60.0
 
83.3Race(%)White
 
(non-Hispanic)
 
80.8
 
78.976.5
 
85.0
 
91.7Black
 
(non-Hispanic)
 
5.05.6
 
11.8
 
0.0
 
0.0Other(non-Hispanic)
 
7.5
 
9.9
 
5.9
 
5.0
 
8.3Hispanic
 
5.8
 
5.6
 
5.9
 
10.0
 
0.0Age(mean
 
(SD))
 
15.51
 
(1.26)
 
15.60
 
(1.08)
 
15.72
 
(1.58)
 
15.12
 
(1.62)
 
15.30
 
(1.06)Income
 
(%)
<
$40,000
 
17.9
 
23.0
 
11.7
 
15.8
 
0.0$40–$80,000
 
35.7
 
36.935.3
 
31.6
 
36.4
>
$80,000
 
46.4
 
40.0
 
52.9
 
52.7
 
63.6Family
 
on
 
public
 
assistance?
 
(%yes)
 
3.4
 
4.3
 
0.05.0
 
0.0Youthin
 
individual,
 
group,
 
or
 
family
 
therapy?
 
(%
 
yes
 
[mean
 
number
 
ofsessions])
 
95.4
 
(4.77)
 
93.8(4.35)
 
88.2
 
(6.53)
 
90.0
 
(4.59)
 
90.9
 
(4.91)
Notes
.SSRI
 
=
 
SSRI
 
alone,
 
MS=
 
SSRI
 
+mood
 
stabilizer,
 
AP
 
=SSRI
 
+
 
antipsychotic,
 
BOTH
 
=SSRI
 
+mood
 
stabilizer
 
and
 
aantipsychotic.
 Table
 
2
Baseline
 
psychopathology
 
variables
 
for
 
consistent
 
medication
 
sample:
 
overall
 
and
 
bygroup.Overall
n
 
=
 
120Mean
 
(SD)SSRI
n
=71Mean
 
(SD)MS
n
 
=17Mean
 
(SD)AP
n
 
=
 
20Mean
 
(SD)BOTH
n
=12Mean
 
(SD)SIQ-JR 
 
47.15
 
(20.04)
 
46.65
 
(21.10)
 
44.53
 
(15.32)
 
50.65
 
(20.18)
 
48.00
 
(20.78)CDRS-R61.70
 
(12.08)
 
62.08
 
(12.62)
 
58.41
 
(10.83)
 
62.35
 
(12.08)
 
63.00
 
(11.10)MASC45.65
 
(18.87)
 
45.04
 
(18.76)
 
42.06
 
(17.34)
 
48.45
 
(21.97)
 
49.58
 
(16.95)BHS9.19(5.75)
 
9.30
 
(5.69)
 
6.53
 
(5.63)
 
10.65
 
(5.27)
 
9.92
 
(9.46)YSRinternalizing
 
28.30
 
(10.30)
 
27.88
 
(10.87)
 
26.12
 
(9.92)
 
31.15
 
(9.16)
 
29.00
 
(9.32)YSRthought
 
disruption
 
4.14(3.12)
 
3.84
 
(3.13)
 
3.53
 
(2.81)
 
5.65
 
(3.17)
 
4.25
 
(2.99)YSRexternalizing
 
19.78
 
(9.31)
 
18.63
 
(9.20)
 
19.76
 
(10.69)
 
21.65
 
(9.43)
 
23.42
 
(7.03)YSRaggression
 
problems
 
12.92
 
(6.49)
 
11.94
 
(6.13)
 
13.29
 
(8.05)
 
14.35
 
(7.03)
 
15.75
 
(4.29)Hasthe
 
youth
 
attempted
 
suicide
 
in
 
the
 
past
 
month?
 
(%
 
yes)
 
58.663.1
 
58.8
 
50.0
 
45.5Suicideattempts
 
lifetimeNo
 
attempt
 
(%)
 
25.825.4
 
17.6
 
40.0
 
16.7Single
 
attempt
 
(%)
 
35.038.0
 
35.3
 
20.0
 
41.7Multiple
 
attempts
 
(%)
 
39.236.6
 
47.1
 
40.0
 
41.7
Notes
.SSRI
 
=
 
SSRI
 
alone,
 
MS=SSRI
 
+
 
mood
 
stabilizer,
 
AP
 
=SSRI+
 
antipsychotic,
 
BOTH
 
=SSRI+
 
mood
 
stabilizer
 
and
 
aneuroleptic;
 
suicide
 
attempt
 
questions
 
are
 
based
 
on
 
childreporton
 
the
 
DISC.
N.Ghaziuddin
 
et 
 
al.
 
 / 
 
 Asian
 
 Journal
 
of 
 
Psychiatry
 
 xxx
 
(2014)
 
 xxx–xxx
4
GModel
AJP-582;
 
No.
 
of 
 
Pages
 
7
Please
 
cite
 
this
 
article
 
in
 
press
 
as:
 
Ghaziuddin,
 
N.,
 
etal.,
 
Anaturalistic
 
study
 
of 
 
suicidal
 
adolescents
 
treated
 
with
 
an
 
SSRI:Suicidalideation
 
and
 
behavior
 
during
 
3-month
 
post-hospitalization
 
period.
 
Asian
 
 J.Psychiatry
 
(2014),
 
 
entire
 
group
 
during
 
the
 
month
 
prior
 
tostudy
 
entry
 
declined
 
from58.6%
 
to
 
5%
 
reported
 
at3-month
 
follow
 
up
 
2
4).
 
Chi-squared
 
analyses
 
revealed
 
significant
 
differences
 
between
 
medica-tion
 
groups
 
in
 
the
 
number
 
of 
 
adolescent
 
or
 
parent
 
reportedpsychiatric
 
hospitalizations
 
(
x
2
[3]
 
=
 
16.41,
 
 p
 
=0.001)
 
and
 
thenumber
 
of 
 
overall
 
SREs
 
(
x
2
[3]
 
=
 
10.32,
 
 p
 
=
 
0.016).
 
A
 
series
 
of 
 
2
 
2Chi-square
 
tests
 
were
 
conducted
 
posthoc
 
todetermine
 
the
 
specificbetween-group
 
differences.
 
All
 
medication
 
augmentation
 
groups(SSRI
 
+
 
MS,
 
SSRI
 
+
 
AP,
 
SSRI
 
+
 
BOTH)
 
reported
 
significantly
 
elevatedlikelihood
 
of 
 
apsychiatric
 
hospitalization
 
compared
 
with
 
thegroupreceiving
 
an
 
SSRI
 
alone,
 
(17.6%
 
vs.1.4%
 
forthe
 
mood
 
stabilizer
 
group,
x
2
[1]
 
=
 
8.34,
 
 p
 
=0.004;
 
30.0%
 
vs.
 
1.4%
 
for
 
the
 
antipsychotic
 
group,
 
x
2
[1]
 
=
 
17.97,
 
 p
 
=
 
0.000;
 
16.7%
 
vs.1.4%
 
for
 
the
 
group
 
onBOTH,
 
x
2
[1]
 
=
 
6.86,
 
 p
 
=
 
0.009).
 
Additionally,
 
those
 
receiving
 
an
 
SSRI
 
with
 
anantipsychotic
 
were
 
morelikely
 
to
 
report
 
an
 
SRE
 
than
 
those
 
receivingan
 
SSRI
 
alone
 
(40.0%
 
vs.
 
9.9%,
 
x
2
[1]
 
=
 
10.30,
 
 p
 
=
 
0.001).
 
Therewere
 
noother
 
statistically
 
significant
 
differences
 
between
 
the
 
groups
 
intermsof 
 
SREs,
 
hospitalizations,
 
or
 
suicide
 
attempts.
Multiple
 
linear
 
regression
 
was
 
used
 
to
 
identify
 
which
 
variableshadinfluenced
 
the
 
suicidality
 
related
 
outcomes
 
among
 
themedication
 
groups.
 
The
 
dependent
 
variable
 
in
 
these
 
analyseswas
 
change
 
in
 
SIQ-JR 
 
from
 
baseline
 
to
 
3
 
months
 
(a
 
negative
 
changevalue
 
for
 
an
 
individual
 
indicated
 
a
 
decrease
 
in
 
SIQ-JR 
 
score).Covariates
 
were
 
scores
 
on
 
measures
 
that
 
differed
 
betweenmedication
 
groups
 
at
 
baseline
 
and
 
scores
 
on
 
measures
 
with
 
eitherconceptual
 
or
 
empirical
 
association
 
with
 
suicidality.
 
The
 
covariatesfor
 
the
 
test
 
of 
 
suicidal
 
ideation
 
change
 
weremedication
 
group,
 
YSR aggression,
 
YSR 
 
thought
 
problems,
 
hopelessness,
 
and
 
baselinereport
 
of 
 
a
 
suicide
 
attempt.
 
The
 
overall
 
model
 
(
[5,117]
 
=3.27,
 p
 
<
 
0.01)
 
accounted
 
for
 
12.7%
 
of 
 
the
 
variance
 
in
 
change
 
scores.However,
 
only
 
two
 
predictors
 
individually
 
accounted
 
for
 
signifi-cantvariance
 
in
 
change
 
scores.
 
Higher
 
baseline
 
hopelessness
 
andaggression
 
scores
 
were
 
associated
 
with
 
greater
 
reduction
 
insuicidal
 
ideation
 
at3-months
 
(hopelessness:
 
b
 
=
 
1.13,
 p
<
 
0.001;
 
aggressive
 
behavior:
 
b
 
=
 
0.67,
 
 p
 
=
 
0.003).
3.
 
Discussion
The
 
main
 
findings
 
of 
 
the
 
present
 
study
 
are:
 
(1)
 
adolescentsreported
 
adecline
 
in
 
suicidal
 
ideation
 
and
 
suicidal
 
behaviorsduring
 
a3-month
 
post-hospitalization
 
period;
 
(2)
 
decrease
 
insuicidal
 
ideation
 
at
 
3-month
 
follow
 
up,
 
irrespective
 
ofmedicationgroup,
 
was
 
associated
 
with
 
higher
 
baseline
 
hopelessness
 
andaggression;
 
(3)
 
statistically
 
higher
 
number
 
of 
 
psychiatric
 
hospi-talizations
 
were
 
observed
 
in
 
groups
 
SSRI
 
+
 
MS
 
and
 
SSRI
 
+
 
BOTH,relativeto
 
SSRI
 
alone
 
group;
 
and
 
(4)
 
and
 
statistically
 
highernumber
 
of 
 
SREs
 
were
 
found
 
in
 
group
 
SSRI
 
+
 
AP,
 
relative
 
toSSRIalonegroup.
 
Based
 
on
 
these
 
data,
 
although
 
we
 
cannot
 
shed
 
light
 
ontheFDAblack
 
box
 
warning,
 
we
 
concluded
 
that
 
suicidality
 
was
 
notexacerbated
 
during
 
treatment
 
with
 
an
 
SSRI
 
in
 
a
 
group
 
of adolescents
 
who
 
were
 
simultaneously
 
depressed
 
and
 
acutelysuicidal.
 
Increased
 
suicidal
 
related
 
outcomes
 
noted
 
in
 
those
 
whoreceived
 
combination
 
medication,
 
in
 
particular
 
those
 
who
 
receivedan
 
antipsychotic
 
agent
 
in
 
conjunction
 
with
 
an
 
SSRI,
 
most
 
likelysuggests
 
agreater
 
illness
 
severity
 
at
 
baseline.
 
Our
 
findings
 
alsounderscore
 
aconcern
 
that
 
has
 
been
 
raised,
 
thatit
 
is
 
important
 
toinclude
 
clinical
 
and
 
treatment
 
variables,
 
or
 
their
 
interaction,
 
whichare
 
often
 
excluded
 
from
 
the
 
assessment
 
of 
 
suicide
 
relatedoutcomes
 
2009).A
 
major
 
strength
 
of 
 
the
 
presentstudy
 
is
 
that
 
depressed
 
and
 
simultaneously
 
suicidal
 
adolescents,whoare
 
often
 
excluded
 
from
 
clinical
 
trials,
 
but
 
frequently
 
presentin
 
‘‘real
 
life’’
 
clinic
 
settings,
 
were
 
included.Decline
 
in
 
suicide
 
attempts
 
reported
 
at
 
3-month
 
follow
 
uprelative
 
to
 
study
 
entry
 
should
 
be
 
viewed
 
in
 
context
 
of 
 
the
 
studydesign.
 
Because
 
our
 
sample
 
comprises
 
adolescents
 
who
 
werepsychiatrically
 
hospitalized
 
with
 
suicide
 
risk
 
concerns,
 
all
 
medi-cationgroups
 
include
 
high
 
percentages
 
of 
 
adolescents
 
who
 
madesuicide
 
attempts
 
in
 
the
 
past
 
month
 
(ranging
 
from
 
45.5%
 
to
 
63.1%).Ashas
 
been
 
shown
 
in
 
multiple
 
published
 
follow-up
 
studies
 
of suicidal
 
and
 
psychiatrically
 
hospitalized
 
adolescents
 
duringtheimmediate
 
post-hospitalization
 
period,
 
the
 
percentage
 
of 
 Table
 
3
Comorbid
 
diagnoses
 
for
 
consistent
 
medication
 
sample:
 
overall
 
andbymedication
 
group.Overall
n
=
 
120SSRI
n
 
=71MS
n
 
=17AP
n
 
=
 
20BOTH
n
 
=
 
12Presenceof 
 
any
 
anxiety
 
disorder
a
27.526.8
 
23.5
 
30.0
 
33.3Presenceof 
 
PTSD
 
or
 
acute
 
stress
 
disorder
 
25.021.1
 
41.2
 
25.0
 
25.0Presenceof 
 
any
 
disruptive
 
behavior
 
disorder
b
37.535.2
 
29.4
 
60.0
 
25.0Presenceof 
 
any
 
substance
 
usedisorder?
c
16.718.3
 
17.6
 
5.0
 
25.0
Notes
.Percentages
 
presented
 
aredefinite,
 
probable,
 
or
 
partially
 
remitted
 
presence
 
of 
 
diagnosis,
 
all
 
diagnoses
 
are
 
based
 
on
 
K-SADD
 
interview.
a
Any
 
anxiety
 
disorder
 
excludes
 
PTSD
 
and
 
acute
 
stress
 
disorder;
 
includes
 
separation
 
anxiety,
 
panic
 
disorder,
 
avoidant
 
disorder,
 
simple
 
phobia,
 
social
 
phobia,
 
agoraphobia,generalized
 
anxiety
 
disorder,
 
andobsessive-compulsive
 
disorder.
b
Includes
 
attention
 
deficit/hyperactivity
 
disorder,
 
conduct
 
disorder,
 
oroppositional
 
defiant
 
disorder.
c
Includes
 
alcohol
 
andother
 
substances;
 
abuse
 
anddependence.
 Table
 
4
Three-month
 
outcome.3-Month
 
analysis
 
by
 
groupSSRIMS
 
AP
 
BOTH
 
OverallYouthor
 
parent
 
reported
 
attempt
 
(%
 
yes)
 
5.60.0
 
10.0
 
0.0
 
5.0Youthor
 
parent
 
reported
 
psychiatric
 
hospitalization
 
(%
 
yes)
 
1.417.6
*
30.0
 
16.7
*
10.0SRE:Youth
 
or
 
parent
 
reported
 
suicide-related
 
emergency
 
(%
 
yes)
 
9.923.5
 
40.0
16.7
 
17.53month
 
SIQ-JR 
 
total
 
score
 
[mean
 
(SD)]
 
22.16
 
(17.21)
 
19.29
 
(7.78)
 
22.35
 
(17.39)
 
22.17
 
(18.74)
 
21.78
 
(16.25)3month
 
SIQ-JR 
 
change
 
score
 
[mean
 
(SD)]
 
24.66
 
(23.85)
 
25.24
 
(16.82)
 
28.30
 
(17.35)
 
25.83
 
(19.94)
 
25.47
 
(21.41)
Notes
.SSRI
 
=SSRI
 
alone,
 
MS=SSRI
 
+mood
 
stabilizer,
 
AP
 
=SSRI+antipsychotic,
 
BOTH
 
=
 
SSRI
 
+
 
mood
 
stabilizer
 
andanantipsychotic,
 
suicide
 
attempt,
 
hospitalization,
 
andemergency
 
items
 
taken
 
from
 
the
 
CIC
 
andDISC;
 
negative
 
SIQ-JR 
 
change
 
score
 
indicates
 
adecrease
 
insuicidal
 
ideation.
*
Significantly
 
different
 
from
 
SSRI
 
alone
 
(
 p
 
<
 
0.01).
**
Significantly
 
different
 
from
 
SSRI
 
alone
 
(
 p
<
 
0.001).
N.
 
Ghaziuddin
 
et 
 
al.
 
 / 
 
 Asian
 
 Journal
 
of 
 
Psychiatry
 
 xxx(2014)
 
 xxx–xxx
 
5
G
 
Model
AJP-582;
 
No.
 
of 
 
Pages
 
7
Please
 
cite
 
this
 
article
 
in
 
press
 
as:
 
Ghaziuddin,
 
N.,
 
et
 
al.,
 
Anaturalistic
 
study
 
of 
 
suicidal
 
adolescents
 
treated
 
with
 
an
 
SSRI:
 
Suicidalideation
 
and
 
behavior
 
during
 
3-month
 
post-hospitalization
 
period.
 
Asian
 
 J.Psychiatry
 
(2014),
 
of 7