International Journal of Emergency Mental Health 163
Adolescent Suicide and Suicide Contagion in Three
Soili Poijula, M.A., Karl-Erik Wahlberg, Ph.D., Atle Dyregrov, Ph.D.
ABSTRACT: This study investigated crisis intervention in three secondary schools after the suicides of
five students, focusing on the relation between crisis intervention and suicide contagion. The contagion
hypothesis was supported. Following a suicide, the number of suicides that occurred in secondary
schools in one year were markedly increased beyond chance. No new suicides took place at schools
where adequate first talk-throughs and psychological debriefing were conducted by a mental health
professional. Proper crisis intervention is recommended to prevent suicide contagion in schools.
[International Journal of Emergency Mental Health, 2001 3(3),pp 163-168].
KEY WORDS: suicide, contagion, intervention, psychological debriefing
Suicide is the third leading cause of death among
adolescents (Kochanek & Hudson, 1994). While suicide
research aims at prevention, the influence of preventive
strategies on suicide mortality has remained inconclusive
(Gunnell & Frankel, 1994; Lewis, Hawton & Jones, 1997).
Research on intervention to prevent suicide contagion is
Suicide contagion has been operationally defined as
when one persons suicide increases the likelihood of
anothers (Higgins & Range, 1996). Research indicates that,
although debatable, there is some evidence of suicide
contagion, but this occurs through modeling by friends when
crisis intervention has not been organized (Velting & Gould,
1997). The effect of suggestion on suicide is well established
(Taiminen, 1992). Adolescents are less likely to identify with
the deceased, and hence to commit suicide themselves, if
they learn that a suicide victim was psychiatrically disturbed,
was functioning in a psychopathological way, or was subject
to individual psychosocial stressors (Higgins & Range, 1996).
The majority of imitative suicides appears to take place within
Soili Poijula, Oy Synolon Ltd., Center for Trauma Psychology,
Oulu, Finland. Karl-Erik Wahlberg, Department of Psychiatry,
University of Oulu, Finland. Atle Dyregrov, Center for Crisis
Psychology, Bergen, Norway. Address correspondence concerning
this article to: Soili Poijula, Oy Synolon Ltd., Center for Trauma
Psychology, Valtatie 16 as 11, 90500 Oulu, Finland. Tel: 358 8 554
5814. Fax: 358 8 554 5801. E-mail: Soili.Poijula@netppl.fi.
the first ten days of exposure to suicide (Phillips & Carstensen,
1986). However, the latency period for imitative suicide has
been shown to be longer than six months (Väisänen &
Hägglund, 1981). The suicide risk of adolescents is higher if
they are exposed to suicide (Brent et al., 1989; Pelzer, Cherian,
& Cherian, 1998; Runeson, 1998). Brent et al. (1993) found
that seven months after the loss of an adolescent peer to
suicide, suicidal ideation with a plan or an attempt was elevated
in the suicide exposed. Bereavement can offer the adolescent
an opportunity to enhance values and coping skills or, by
contrast, the option to imitate a self-destructive response to
a crisis (Valente & Saunders, 1993)
After the suicide of a school-aged adolescent, the whole
school community, both students and teachers, suffers from
trauma and loss. Acknowledging the loss and reestablishing
normality within the school can be difficult. In many cases
school administrators are also very concerned about the
contagion aspects of suicide, and this is the single reason
why most school systems refrain from organizing crisis
intervention (Mauk & Weber, 1991). A community trauma,
such as suicide, requires that caretakers provide good
leadership to guide physical and emotional recovery (Williams,
Zinner, & Ellis, 1998). Cornell and Sheras (1998) conclude
that the results of school community crisis intervention
depend on the leaders ability to notice a crisis and
164 Poijula/ ADOLESCENT SUICIDE AND SUICIDE CONTAGION
acknowledge reactions. They claim that when leaders in some
school crises have not provided good leadership, they have
made the crisis worse. A nonsupportive, closed atmosphere
may result in students keeping traumatic grief internalized or
acting it out. The logical conclusion is that if the suicide is
publicly known and has caused reactions in students, it has
to be openly confronted.
In previous studies of bereaved adolescents it has been
recommended that a school experiencing such a sudden death
should organize crisis intervention (Brent et al., 1993;
Dyregrov, Gjestad, Bie Wikander & Vigerust, 1999; Mauk &
Weber, 1991). School crisis management may include
developing a contingency plan; informing students about a
death at school; conducting immediate talk-throughs, rituals,
and psychological debriefings among students; allowing
students to view the body, participate in the funeral and/or
meetings with parents; and following up of students over
time (Dyregrov, 1991; Leenaars & Wenckstern, 1998; Mitchell
& Every, 1996). In organizing interventions, the grief hierarchy
and exposure to the suicide are used in prioritizing to whom
help is offered, but everyone exposed to suicide requires
some level of care (Mauk & Weber, 1991).
Few studies to date have evaluated such interventions,
nor have there been studies on psychological debriefing and
suicide contagion prevention for adolescents.This study
concerned school crisis intervention and suicide contagion,
using the crisis intervention methods consisting of first talk-
throughs (FTT) and psychological debriefings (PD).
The first talk-through or defusing depicts a meeting of
those who were involved in or experienced a critical event.
The meeting is conversational in tone and takes place on the
same day as the event, usually within the first 8 hours
(Mitchell, 1995). It aims at providing emotional first aid to
people. To help students in crisis, the school creates a caring
climate and structures the chaos by providing the possibility
of ventilating thoughts and experiences. In schools, the
concept of grief hierarchy is used to prioritize to whom help
is offered. Usually this means focusing on the classmates.
The first talk-through is organized at school after informing
all students about the death. The conductor of the first talk-
through can be a teacher, a school nurse or a mental health
professional. Adults lead the adolescents in a conversation
where facts are shared and mutual support can be activated
(Dyregrov & Raundalen, 1995).
Psychological debriefing is a helpers group discussion
of a traumatic event designed to mitigate the impact of a
severe stress experience. It is conducted by trained
professionals (Mitchell, 1995). When psychological debriefing
is adapted and used with survivors of sudden death at a
school, there are both differences and similarities from the
original model of PD. At school, PD is a group discussion in
a class. It can last 1-2 lesson periods depending on the age of
the children.The conductor of the PD is a trained mental
health professional or a teacher. If the teacher is to be a
conductor, it is recommended that a mental health professional
co-leads the process. If a mental health professional is a
conductor, the teacher of the class can co-lead. The phases
of the psychological debriefing in schools are introduction,
facts, reactions, information and closure (Dyregrov, 1991).
In this study the schools crisis interventions following
different suicides are described and compared with the post
stress interventions described by Dyregrov (1991). The
suicide contagion hypothesis was also addressed with data
from a four-year follow-up period. Our hypotheses were that:
1) after a suicide of a student there will be an increased risk of
other suicides at the school, and 2) appropriate intervention
will reduce the risk of suicide contagion.
During the school year 1995-1996 five secondary school
students committed suicide in three schools of the Oulu area
in Northern Finland (Table 1). These three schools were all
located in small rural communities. Geographically they were
not neighboring communities. This cluster of unrelated
adolescent suicides was the starting-point for this study.
In August 1995, two weeks before the school term
started, a male 17 year-old ex-secondary school student (case
I) committed suicide by self-immolation in the schoolyard of
a secondary school (A). One month later, in September, a
male 15 year-old 9
grade student (case II) of this same school
(A) committed suicide with a firearm. He was a friend of the
first suicide victim (case I). In January 1996, 4 months later,
another male 15 year-old 9
grade student (case III) in school
(A) committed suicide with a firearm. He was a friend of the
second suicide victim (case II).
Again in August 1995, a 14 year-old female secondary
school (B) student (case IV) committed suicide with a firearm
one week before the school term started. In October 1995, a
14 year-old male student (case V) from the same school (B)
International Journal of Emergency Mental Health 165
committed suicide with a firearm. He was an acquaintance of
the first suicide victim of this school (case IV). In January
1996, a 13 year-old male secondary school (C) student (case
VI) committed suicide with a firearm. None of these
adolescents (cases I-VI) who committed suicide were known
to be psychiatrically disturbed or otherwise different from
School C had no contingency plan, but in case VI an
adequate crisis intervention (FTT and PD conducted by a
trained mental health professional) was organized and timed
appropriately. Here FTT was held during the first day after
the suicide and PD two days after the suicide. PD lasted for
Participants in this study were the students of the three
schools and homeroom classmates of the suicide victims.
All together there were 270 students in school A, 346 in
school B and 585 students in school C. Of these, 89 were
classmates (46 boys and 43 girls). From school A 15 boys
and 16 girls participated, from school B 15 boys and 17 girls
and from school C 16 boys and 10 girls. Participants were of
the same age (age 13 to 17) and had equal gender distribution.
Measures, Procedure, and Statistical Methods
The incidence of new suicides in the three schools was
followed for a four year period (August 1995- August 1999).
The first author had an agreement with three local school
psychologists of being informed on any new suicides in the
three schools during the follow up period (1995-1999).
The Poisson distribution was calculated for determining
if the number of suicides was increased beyond chance
(Rothman & Greenland, 1998). The statistical software used
was SPSS (9.0).
Interventions Carried Out in the Three Schools
In school A there was no contingency plan. After suicide
cases I and II no crisis intervention was organized. After
suicide case III an adequate crisis intervention, consisting of
a first talk-through (FTT) and a psychological debriefing (PD)
was organized, conducted by a trained mental health
professional (clinical psychologist ) and timed appropriately
so that FTT was held during the first day after the suicide
and PD (that lasted for two hours) was conducted during the
Although school B had a contingency plan, in case IV
only classroom meetings (an adapted version of PD, that
lasted for 1 hour conducted by a teacher) were organized in
all but one 8
grade class, where no intervention was
performed. The timing of the classroom meeting was late
(one week after the suicide). In case V an adequate crisis
intervention (FTT, and PD that lasted for one hour and was
conducted by a trained mental health professional) was
organized and timed appropriately (FTT was held during the
first day after the suicide and PD four days after the suicide).
CASE DATE METHOD AGE SEX EXPOSURE RELATIONSHIP
I 08/95 Self-immol. 17 M No
II 09/95 Firearm 15 M Heard Friend
III 01/96 Firearm 15 M Heard Friend
IV 08/95 Firearm 14 F No
V 10/95 Firearm 14 M Heard Acquaintance
VI 01/96 Firearm 13 M
Table 1. Characteristics of the Suicides of Students
166 Poijula/ ADOLESCENT SUICIDE AND SUICIDE CONTAGION
Contagion of the Suicides
Two suicides of 15-year-old male students occurred after
the first suicide within 4 months in school A (270 students).
Assuming a rate of 21.6 per 100,000 students per year (the
national suicide rate among 15- to 19-year-old males reported
by Statistics Finland, 1998), one would expect to see 0.0311968
suicides in this secondary school of 148 male students in
one year and 0.00324 in the group of 15 boys in homeroom
classrooms. The observed rate was 62.6 times the expected
rate in the whole school and 617.3 in the two homeroom
In school B (346 students), two suicides of 14-year-old
students occurred in one month. Assuming a rate of 1.9 per
100,000 students per year (the national suicide rate among
10- to 14-year-old males reported by the Statistics Finland,
1998), one would expect to see 0.003249 suicides of male
students in this secondary school of 171 males per year and
0.00285 in the homeroom classes group of 15 boys. The
observed rate was thus 307.8 times that expected for the whole
school for males and 3508.8 for homeroom class males. For
females (rate of 1.8 per 100,000) in a school of this size the
expected number is 0.00315 and for homeroom classes group
of 17 girls 0.00306. The observed rate is 317.5 times that
expected for the whole school for girls and 3268.0 for the
homeroom classes girls.
Using the Poisson distribution, the number of suicides
that occurred in all secondary schools in one year were
markedly increased beyond chance (p < 0.001). Contagion
did not appear in the first ten days, but 1 (two suicides), 1½,
2 and 4 months after the first suicide.
Intervention and Suicide Contagion
In cases III, V and VI but not in the others, the intervention
Case I none
Case II none
Case III First day
Case IV none
Case V First day
Case VI First day
Table 2. Schools and Interventions after Suicides
Timing of First
Duration and Conductor
Timing of debriefing
Mental Health Professional
Debriefing, 2 hours,
Exclusion of 1 class, 1 hour,
Debriefing, 1 hour
Mental Health Professional
Debriefing, 1 hour,
Mental Health Professional 2 days
International Journal of Emergency Mental Health 167
of the school was adequate (FTT and PD). In schools and
classes where a first talk-through and psychological
debriefing were conducted by a mental health professional
as the intervention, no new suicides appeared during the
four year follow up period (August 1995 August 1999, Table
2). In school B teachers conducted a classroom meeting in
all but one 8th grade class. In that school, a second suicide
was committed two months later by a student whose class
had not had the classroom meeting.
The contagion hypothesis was supported in this study.
A much higher probability for students to commit suicide
compared to the general population of the same age was
found. In contrast to a short delay between a death and a
new suicide found in a previous report (Phillips & Carstensen,
1986), the time interval from the death of the student before
a new suicide of a student took place in this study was from
1 to 4 months. The method of the suicides seemed also to be
an imitation from the first suicides of the students. All of the
victims of the second wave of suicides knew about the
method of suicide and had been friends of the student who
committed the initial suicide.
The second hypothesis about intervention reducing
suicide contagion was also supported. An appropriate
intervention (FTT and PD) by a trained mental health
professional seemed to be a factor in inhibiting new suicides
of the students. The intervention method varied in the
schools, from nothing to an adequate crisis intervention.
What we consider to be a complete intervention (well
conducted FTT and PD) was needed to prevent new suicides.
If an incomplete crisis intervention was conducted, a new
suicide was committed. The connection between crisis
intervention and the systemic effects of intervention, and
how to prevent possible negative effects of incomplete
intervention, demand further study. Early crisis intervention
fosters an open discussion of facts, thoughts and reactions.
This open and direct handling of death may prevent
fantasizing and idealization of death. It may prevent acting
out of unspoken trauma, and imitation of suicidal behavior. It
may also have long term effects in preventing suicidality of
potentially bereaved and depressed youth, exposed to
Limitations of the Study
The phenomenon of adolescent suicide contagion is rare.
This research is based on a small number of cases, not optimal
for statistical analysis, lacking sufficient statistical power.
Although problematic, the natural research design we used
was a way of developing new knowledge on the phenomenon
and the findings serve as hypotheses for further testing. The
conclusions should be considered as tentative, leading to
more research on preventive measures in this area.
Preliminary findings of this study show that early crisis
intervention and use of first talk-throughs and psychological
debriefing do not cause suicide contagion, but lack of
intervention may do so. In organizing interventions, leaders
in schools dont need to be afraid of interventing in a
supportive and open atmosphere, but should provide
systematic crisis intervention for exposed peers to guide
physical and emotional recovery.
Brent, D. A., Kerr, M. M., Goldstein, C., Bozigar, J., Wartella,
M., & Allan, M. J. (1989). An outbreak of suicide and
suicidal behavior in a high school. Journal of the
American Academy of Child & Adolescent Psychiatry,
Brent, D. A., Perper, J. A., Moritz, G., Allman, C., Schweers, J.,
Roth, C., Balach, L., Canobbio, R., & Liotus, L. (1993).
Psychiatric sequelae to the loss of an adolescent peer to
suicide. Journal of the American Academy of Child and
Adolescent Psychiatry, 32, 509-517.
Cornell, D. G. & Shears, P. L. (1998). Common errors in school
crisis response: Learning from our mistakes. Psychology
in the Schools, 35, 297-307.
Dyregrov, A. (1991). Grief in children. A handbook for
adults. London: Jessica Kingsley publishers.
168 Poijula/ ADOLESCENT SUICIDE AND SUICIDE CONTAGION
Dyregrov, A., & Raundalen, M. (1995) Sorg och omsorg.
Dyregrov, A., Gjestad, R., Bie Wikander, A. M., & Vigerust, S.
(1999). Reactions following the sudden death of a
classmate. Scandinavian Journal of Psychology, 40,
Gunnell, D., & Frankel, S. (1994). Prevention of suicide:
aspirations and evidence. British Medicical Journal,
Higgins, L., & Range, L. M. (1996). Does information that a
suicide victim was psychiatrically disturbed reduce the
likelihood of contagion? Journal of Applied Social
Psychology, 26, 781-785.
Kochanek, K. D., & Hudson, B. L. (1994). Advance report of
final mortality statistics, 1992. Monthly Vital Statistics
Report, 43(6), (Suppl.).
Leenaars, A. A., & Wenckstern, S. (1998). Principles of
postvention: Applications to suicide and trauma in
schools. Death Studies, 22, 357-91.
Lewis, G., Hawton, K., & Jones, P. (1997). Strategies for
preventing suicide. British Journal of Psychiatry, 171,
Mauk, G.W., & Weber, C. (1991). Peer Survivors of Adolescent
Suicide: Perspectives on Grieving and Postvention.
Journal of Adolescent Research, 6, 113-131.
Mitchell, J.T. (1995). Essential of Critical Incident Stress
Management. In George S. Everly, Jr. (Ed.) Innovations
in disaster and trauma psychology, Volume One:
Applications in emergency services and disaster
response, 68-89. Ellicott City, Md.: Chevron Publishing
Mitchell, J. T., & Everly, G. S. (1996). Critical Incident Stress
Debriefing: An operations manual for the prevention of
traumatic stress among emergency services and disaster
workers. Ellicott City, Md.: Chevron Publishing
Company (Second Edition, Revised).
Pelzer, K., Cherian, V.I., & Cherian L. (1998). Attitudes toward
suicide among South African secondary school pupils.
Psychological Reports, 83 (3 Pt 2): 1259-1265.
Phillips, D. P., & Carstensen, L. L. (1986). Clustering of teenage
suicides after television news stories about suicide. New
England Journal of Medicine, 315, 685-689.
Rothman, K., & Greenland, S. (1998). Modern epidemiology.
2 ed. (pp.247-248). Philadelphia (Pa.): Lippicott- Raven.
Runeson, B. S. (1998). History of suicidal behavior in the
families of young suicides. Acta Psychiatrica
Scandinavica, 6, 497-501.
Statistics Finland (1998). Causes of death 1996. Health, 3.
Taiminen, T. (1992). Projective identification and suicide
contagion. Acta Psychiarica Scandinavica, 85 (6), 449-
Valente, S. M., & Saunders, J. M. (1993). Adolescent Grief
After Suicide. Crisis, 14, 1.
Velting, D. M., & Gould, M. S. (1997). Suicide contagion. In
R. W. Maris, & M. M., Silverman (Eds), Review of
suicidology (pp. 96-147). New York, NY: The Guilford
Väisänen, E. J. & Hägglund, T.-B. (1981). Destruction
preceding suicide: An episode in the large series of
adolescents suicides. In J. P. Soubrier & J. Vedrinne
(Eds.), Depression et Suicide (pp. 670-675). Paris:
Williams, M. B., Zinner, E. S., & Ellis, R. R.(1998). The
connection between grief and trauma: an overview. In E.
S. Zinner, & M. B. Williams (Eds.), When a community
weeps (pp. 3-17). Philadelphia, PA: Brunner/Mazel.
International Journal of Emergency Mental Health 169
The Society of Clinical Psychology, Division 12 of the
American Psychological Association introduces a new
section for clinical emergencies and crises.
APA Division 12 has approved an innovative new Section that gives
recognition to the difficult clinical work that psychologists do with patients
or clients who engage in life threatening behaviors. The Section has been
established to advance the clinical and scientific understanding of
psychological/behavioral emergencies and crises as well as the clinical
abilities needed to evaluate and manage them. Emergencies include life
threatening behaviors such as acute suicidality, potential violence, and risk
to vulnerable victims of violence. The Section provides a forum for the
exchange of clinical information and research findings related to the
emergencies noted above and to the crises from which they so often develop.
It has the further purposes of fostering education and training in the
evaluation and management of these high risk clinical situations, as well as
understanding and assisting with the impact of such difficult and intense
work on the clinician. Membership is open to all members of the American
For information or an application contact: Philip M. Kleespies, Ph.D., Section VII Representative (Pro Team),
Psychology Service (116 B), VA Medical Center, 150 South Huntington Ave., Boston, MA 02130
Telephone: 617-232-9500 x4106 E-mail: Kleespies.Phillip_M_PHD@Boston.VA.GOV