Content uploaded by Alcion Sponholz Junior
Author content
All content in this area was uploaded by Alcion Sponholz Junior on Feb 05, 2016
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=usui20
Download by: [Universidade de Sao Paulo], [Alcion Sponholz Junior] Date: 04 February 2016, At: 16:38
Archives of Suicide Research
ISSN: 1381-1118 (Print) 1543-6136 (Online) Journal homepage: http://www.tandfonline.com/loi/usui20
Clinical Features, Psychiatric Assessment, and
Longitudinal Outcome of Suicide Attempters
Admitted to a Tertiary Emergency Hospital
Alcinéia Donizeti Ferreira, Alcion Sponholz Jr., Célia Mantovani, Antônio
Pazin-Filho, Afonso Dinis Costa Passos, Neury José Botega & Cristina Marta
Del-Ben
To cite this article: Alcinéia Donizeti Ferreira, Alcion Sponholz Jr., Célia Mantovani, Antônio
Pazin-Filho, Afonso Dinis Costa Passos, Neury José Botega & Cristina Marta Del-Ben
(2015): Clinical Features, Psychiatric Assessment, and Longitudinal Outcome of Suicide
Attempters Admitted to a Tertiary Emergency Hospital, Archives of Suicide Research, DOI:
10.1080/13811118.2015.1004491
To link to this article: http://dx.doi.org/10.1080/13811118.2015.1004491
Accepted author version posted online: 11
May 2015.
Published online: 11 May 2015.
Submit your article to this journal
Article views: 65
View related articles
View Crossmark data
Clinical Features,
Psychiatric Assessment,
and Longitudinal Outcome
of Suicide Attempters
Admitted to a Tertiary
Emergency Hospital
Alcine
´ia Donizeti Ferreira, Alcion Sponholz Jr., Ce
´lia Mantovani,
Anto
ˆnio Pazin-Filho, Afonso Dinis Costa Passos, Neury Jose
´Botega,
and Cristina Marta Del-Ben
The objective of this study was to characterize admissions to an emergency hospital
due to suicide attempts and verify outcomes in 2 years. Data were collected from
medical records and were analyzed using descriptive statistics and logistic regression.
The sample consisted of 412 patients (58.7%women; mean age ¼32.6 years old,
SD ¼14.3). Self-poisoning was the most frequent method (84.0%), and they were
diagnosed mainly as depressive (40.3%) and borderline personality disorders
(19.1%). Previous suicide attempts and current psychiatric treatment were reported
by, respectively, 32.0%and 28.4%. Fifteen patients (3.6%, 9 males) died during
hospitalization. At discharge, 79.3%were referred to community-based psychiatric
services. Being male (OR ¼2.11; 95%CI ¼1.25–3.55), using violent methods
(i.e., hanging, firearms, and knives) (OR ¼1.96; 95%CI ¼1.02–3.75) and
psychiatric treatment history (OR ¼2.58; 95%CI ¼1.53–4.36) were predictors
for psychiatric hospitalization. Of 258 patients followed for 2 years, 10 (3.9%) died
(3 suicide), and 24 (9.3%) undertook new suicide attempts. Patients with a history
of psychiatric treatment had higher risks of new suicide attempts (OR ¼2.46, 95%
CI ¼1.07–5.65). Suicide attempters admitted to emergency hospitals exhibit severe
psychiatric disorders, and despite interventions, they continue to present high risks for
suicide attempts and death.
Keywords emergency psychiatry, emergency room, suicide attempt, suicide
INTRODUCTION
Suicide is associated with multiple risk
factors, and a prior history of suicide
attempts is a strong predictor of death by
suicide (Gunnell & Frankel, 1994; Haukka,
Suominen, Partonen, & Lo
¨nnqvist, 2008).
After an episode of deliberate self-harm,
the risk of death by suicide in the sub-
sequent year varies from 0.5%to 2.0%,
Archives of Suicide Research, 0:1–14, 2016
Copyright #International Academy for Suicide Research
ISSN: 1381-1118 print=1543-6136 online
DOI: 10.1080/13811118.2015.1004491
1
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
and it increases to more than 5%after 9
years (Owens, Horrocks, & House, 2002).
The World Health Organization (WHO)
estimates that by 2020, more than 1.5
million people will commit suicide
(WHO, 2003), and a significantly greater
number of individuals will most likely
attempt suicide because attempts can be
up to 40 times more frequent than the act
itself (Schmidtke et al., 1996).
Many people who attempt suicide
require medical attention, and contact with
health services is an opportunity for suicide
risk detection and a referral for specialized
treatment (Fleischmann et al., 2008). This is
particularly relevant in emergency settings,
in which a significant portion of suicide
attempters receives first care aid (Jo et al.,
2011). However, the routines regarding
the psychiatric assessment of suicide
attempters can vary considerably from
one emergency center to another. For
instance, a survey of 223 Californian
emergency departments found that less
than half of the services reported the avail-
ability of at least one mental health pro-
fessional for the assessment of suicide
attempters, and only one-quarter had the
possibility of a psychiatric admission to
their own hospitals (Baraff, Janowicz, &
Asarnow, 2006). In low- and middle-
income countries (LAMIC), the situation
can be even worse. A study conducted in
emergency rooms in eight LAMIC showed
that in half of the emergency services
studied less than one-third of the suicide
attempters were referred to health services
for further evaluation and follow-up
(Fleischmann et al., 2005).
Until recently, suicide was not viewed
in Brazil as a public health problem, as it
was most likely obscured by the high
mortality rates for homicide and traffic
accidents in the country. However, a recent
study showed that from 1998 to 2008, the
total number of deaths due to suicide in
Brazil rose from 6.985 to 9.328, represent-
ing an increase of 33.5%(Waiselfisz,
2011). These numbers might be even
greater in light of under-reporting and the
poor quality of information contained in
the death certificates found in the some
areas of the country (Lovisi, Santos, Legay,
Abelha, & Valencia, 2009). Despite these
understatements, suicide is the third leading
external cause of death in Brazil. Moreover,
nearly 20%of deaths due to external causes
(a total of 143.256 in 2010, http://tabnet.
datasus.gov.br/cgi/idb2012/matriz.htm)are
recorded based only on the mechanism of
death and not on the intention (i.e., ‘‘falls
and drowning’’), making it even more diffi-
cult to ascertain the real numbers of deaths
by suicide in the country (Bertolote, Botega,
& de Leo, 2011).
Considering the paucity of data regard-
ing suicide in LAMIC and the heterogeneity
of the health care of suicide attempters, we
sought a better understanding of the profile
of suicide attempters in LAMIC, as well as
of the medical care provided for this con-
dition. Therefore, the aims of this study
were to characterize the persons admitted
to a tertiary referral emergency hospital in
the southeast of Brazil due to suicide
attempts and to verify the outcomes of
these patients in the years following their
admission. Specifically, we sought to: 1)
delineate the sociodemographic and clinical
profiles of suicide attempters admitted to
referral emergency hospitals over a period
of two years, considering eventual differ-
ences between genders, 2) verify the exist-
ence of possible predictors of decision-
making regarding the discharge disposition
after a suicide attempt, and 3) assess the
possible predictors of the occurrence of
new serious suicide attempts in the 2 years
following the suicide attempts.
METHODS
Study Design and Setting
This is a retrospective cohort study
conducted at the Emergency Hospital of
Features, Assessment, and Outcome in a Tertiary Emergency Hospital
2 VOLUME 0 NUMBER 0 2016
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
the Ribeira
˜o Preto Medical School of the
University of Sa
˜o Paulo, Brazil. It is the
main tertiary emergency hospital of a catch-
ment area composed of 26 municipalities,
containing an estimated population of
1,300,000 inhabitants and located in the
southeast of the country. Ribeira
˜o Preto
is the main city of the region, with
600,000 inhabitants and a per capita gross
domestic product of approximately
US$14,000 (IBGE, 2010). This hospital
accounts for 51%of the emergency visits
in the region, receiving mainly cases of high
complexity (Adolfi Ju
´nior et al., 2010).
Referrals to this emergency hospital follow
local and regional emergency medical
regulations, which take into account the
principles of hierarchy and the complexity
of the service network of the Brazilian
Health System (Sistema U
´nico de Sau
´de,
SUS). This emergency unit has an emerg-
ency room (ER) where the patients are
initially admitted mainly by the internal
medicine team or the surgical team, accord-
ing to their supposed diagnosis established
by the regulatory system. The patients are
evaluated and if they need to be observed
or start treatment, this emergency unit
has 169 beds, which work as inter-
mediate beds until the patient needs are
confirmed and a treatment plan can be
determined.
Several medical specialties work
together in the emergency room, among
them is psychiatry. According to the clini-
cal routines of the hospital, all patients
admitted due to a suicide attempt, regard-
less of the clinic of admission, should be
submitted to a psychiatric assessment.
The psychiatric evaluation basically
consists of verifying the maintenance of
suicidal ideation, the identification of any
psychiatric conditions underlying the
suicide attempt, early intervention in the
emergency context, and a referral to psy-
chiatric care at one of the outpatient or
inpatient mental health services available
in the region.
Selection of Participants
The study population consisted of all
patients admitted to the emergency unit,
from January 1, 2006 to December 31,
2007, with a diagnosis of ‘‘intentional self-
inflicted injury’’ and coded as X60 to X84,
according to the tenth edition of Inter-
national Classification of Diseases (ICD
10) (WHO, 1993). We chose to use these
diagnostic categories as a way of sorting
the cases, as there is not a specific code for
a suicide attempt in the ICD 10. This strat-
egy also allowed for a broad variety of cases,
avoiding losses due to a lack of recording of
suicide attempts in the medical files.
The local ethical committee approved
the study (protocol number: 13437=2008).
Data Collection and Outcomes
Data were collected from medical
records, according to protocols already
applied in the Brazilian population, as
part of the SUPRE-MISS study (Bertolote
et al., 2005; Fleischmann et al., 2005;
Fleischmann et al., 2008). The variables of
interest were: demographic data; methods
of suicide attempts; specialty of admission;
length of stay; completion of the psychiatric
assessment; psychiatric diagnosis; occur-
rence of previous suicide attempts; current
psychiatric community-based care; and psy-
chiatric referral at the time of discharge
from the emergency unit (community-based
treatment vs. partial or full psychiatric hos-
pitalization). Data were extracted by a social
worker familiarized with the routines of the
referral emergency hospital, who was trained
by two senior psychiatrists regarding the
possibilities of register in the medical
records of the relevant information for the
collection of the variable of interest.
The psychiatric diagnosis was defined
as the diagnosis recorded in the medical
files as the main diagnosis, not including
information on co-morbidities. According
Alcine
´ia Donizeti Ferreira et al.
ARCHIVES OF SUICIDE RESEARCH 3
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
to the routines of the service, senior
psychiatrists, based on the clinical history
elaborated by medical residents, defined
the psychiatric diagnosis. These senior
psychiatrists had a systematic training in
the application of operational diagnostic
criteria, and can achieve good reliability
indexes of the diagnosis performed under
indirect supervision (Del-Ben et al., 2005).
The discharge disposition from the
emergency hospital was categorized
according to the need for continuity of care
in a protected environment immediately
after discharge from the emergency room,
including both partial and full psychiatric
hospitalization. All patients referred to par-
tial hospitalization and the majority of the
patients referred to full hospitalization
agreed with this proposal. According to
Brazilian laws, in the case of involuntary
admission the consent was obtained from
a relative who became legally responsible
for the admission.
To analyze the outcomes after dis-
charge from the emergency unit, we con-
sidered only patients living in the main city
of the region (Ribeira
˜o Preto). This strategy
was adopted for two reasons: 1) the differ-
ences between the main city in the region
and other cities regarding the availability
of mental health services, and 2) the unfea-
sibility of acquiring information concerning
the possible occurrence of death among
patients from other municipalities. New
admissions to the emergency hospital
because of a new suicide attempt in the 2
years following the index admission were
assessed by querying the hospital database.
The occurrence of deaths was checked in
the Mortality Information System of the
municipality of Ribeira
˜o Preto.
Statistical Analysis
Data were analyzed with the SPSS soft-
ware package, version 16.0. For the
exploratory analysis and comparison
between genders regarding demographic
features, qualitative variables were analyzed
using the chi-square test or Fisher’s exact
test, and quantitative variables were evalu-
ated using the t test. Values of p <0.05
were considered statistically significant.
To explore the decision-making
process with regards to discharge dispo-
sition, an immediate referral to psychiatric
hospitalization (partial or full) at the time
of discharge from the emergency hospital
was established as the main outcome, and
sex, age, ethnicity, education, marital status,
occupational status, methods used in the
attempted suicide, previous suicidal
attempts, current psychiatric treatment,
main psychiatric diagnosis, and length of
stay were the predictor variables. Initially,
we calculated odds ratios (ORs) and confi-
dence intervals (95%CI) of the variables of
interest, and logistic regression analyses
were then conducted. The variables that
reached statistical significance (p <0.05)
in the univariate analyses were analyzed
jointly by the forward stepwise method.
A second logistic regression analysis
was performed, aiming to establish possible
independent predictors of the occurrence
of further suicide attempts or death due
suicide in the two-year period following
admission to the emergency hospital. The
predictor variables were the same as
described above and included referral to
psychiatric hospitalization.
RESULTS
From January 1, 2006 to December 31,
2007, we identified 534 patients with the
diagnosis of ‘‘intentionally self-inflicted
injury,’’ whose medical records were con-
sulted to characterize a suicide attempt. In
122 (22.8%) records, we found that the
self-induced harm was not associated with
clear suicidal intent and was instead charac-
terized as impulsive, disruptive, or aggressive
behavior, such as punching a wall or a glass
door during a quarrel. Suicidal intent was
Features, Assessment, and Outcome in a Tertiary Emergency Hospital
4 VOLUME 0 NUMBER 0 2016
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
confirmed in 412 medical records, for which
the variables of interest were extracted.
Figure 1shows the flowchart of the
patients included in the study.
Demographic and Clinical Features
The sample consisted of 412 suicide
attempters, with a significantly higher
proportion of women (n ¼242, 58.7%)
than men (n ¼170, 41.3%)(p¼0.001).
Their ages ranged from 12 to 83 years old
(mean ¼32.6, SD ¼14.3), but the women
(mean ¼31.5, SD ¼13.6) were significantly
younger than the men (mean ¼34.3, SD ¼
15.0) (p ¼0.046). As observed in Table 1,
most of the patients reported themselves
as unmarried (64.8%), Caucasian (76.5%)
and with up to 8 years of schooling
(81.8%), without significant differences
occurring between the sexes (p >0.05).
There was a higher proportion of men
(54.7%), in comparison with women
(22.3%), who were employed (p <0.001)
at the time of admission to the emergency
room.
Table 1also shows that the majority of
patients used self-poisoning as the method
of suicide attempt (84.0%), more often by
women (91.3%) than men (73.5%). Men
were more likely to use violent methods,
such as hanging, firearms, and knives. As
a consequence, there was a higher pro-
portion of women (76.4%vs. 65.3%)
admitted by clinical specialties and a greater
proportion of men (20.0%vs. 11.2%)
admitted by surgical specialties (p ¼0.025).
Regardless of sex (p ¼0.106), approxi-
mately half of the patients (52.4%) were
discharged within 24 hours after admission,
while 18.9%remained hospitalized for 4
days or more. The length of hospital stay
ranged from 1 to 150 days (mean ¼3.9,
SD ¼10.0).
Psychiatric assessment was performed
in 380 (92.2%) of the patients admitted
for suicide attempts. The psychiatric assess-
ment was not completed in 13 patients who
died due to their suicide attempts
(described below), in 9 patients who left
the hospital without medical consent, in 7
patients who were discharged a few hours
FIGURE 1. Flowchart of the patients included in the study.
Alcine
´ia Donizeti Ferreira et al.
ARCHIVES OF SUICIDE RESEARCH 5
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
TABLE 1. Distribution of Patients Admitted to an Emergency Hospital From January 2006 to
December 2007 due to a Suicide Attempt, According to Sex And Demographic and
Clinical Variables
Women (242) Men (170) Total (412)
P%%%
Demographic features
Single=divorced=widowed 64.0 65.3 64.8 0.862
Caucasians 74.4 79.4 76.5 0.236
Up to eight years of schooling 81.0 82.9 81.8 0.215
Employed 22.3 54.7 35.7 <0.001
Methods (%)<0.001
Self-poisoning 91.3 73.5 84.0
Violent methods 8.7 26.5 16.0
Admission (%) 0.025
Internal medicine 76.4 65.3 71.8
Surgical specialties 11.2 20.0 14.8
Psychiatry 12.4 14.7 13.3
Length of stay (%)
Up to 24 hours 55.8 47.6 52.4 0.106
25 to 72 hours 28.5 28.8 28.6
More than 72 hours 15.7 23.5 18.9
Psychiatric assessment (%) 0.296
Yes 93.4 90.6 92.2
No 6.6 9.4 7.8
Psychiatric diagnosis (%)<0.001
Depressive disorders 44.9 33.5 40.3
Psychotic disorders 5.1 9.9 7.1
Personality disorders 23.3 13.0 19.1
Psychoactive-related disorders 1.3 14.9 6.9
Reactive disorders 11.4 11.2 11.3
Not recorded 14.0 17.4 15.4
Previous suicidal attempts (%) 0.811
Yes 32.2 31.8 32.0
No 27.7 25.3 26.7
Not recorded 40.1 42.9 41.3
Psychiatric treatment (%) 0.053
Yes 31.8 23.5 28.4
No 35.1 32.4 34.0
Not recorded 33.1 44.1 37.6
Discharge disposition (%) 0.004
Community-based services 82.2 68.2. 76.5
Partial or full hospitalization 15.3 26.5 19.9
Death 2.5 5.3 3.6
Features, Assessment, and Outcome in a Tertiary Emergency Hospital
6 VOLUME 0 NUMBER 0 2016
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
after the emergency admission or in 3
patients who remained hospitalized longer.
As shown in Table 1, diagnoses of
depressive disorders (40.3%) and person-
ality disorders (19.1%) were the most fre-
quent and were more common in women
than in men (44.9%vs. 33.5%; 23.3%vs.
13.0%, respectively). In contrast, there was
a higher proportion of men diagnosed with
psychotic disorders compared to women
(9.9%vs. 5.1%), as well as psychoactive sub-
stance-related disorders (14.9%vs. 1.3%).
Previous suicide attempts and current
psychiatric treatment were reported,
respectively, in 32.0%and 28.4%of the
medical records. There were no differences
between the sexes regarding previous sui-
cide attempts (p ¼0.811), but there was a
trend toward a higher proportion of
women (31.8%vs. 23.5%) receiving psychi-
atric care at the time of admission to the
emergency hospital (p ¼0.053). In approxi-
mately one-third of the medical files there
were no records indicating the occurrence
or the absence of previous suicide attempts
or of previous psychiatric treatment.
Fifteen patients (3.6%, 9 males) died
during hospitalization due to their suicide
attempts. There were no statistically signifi-
cant differences between the patients who
died and the survivors regarding sociode-
mographic or clinical variables, except for
the use of violent methods by 11 (73.3%)
of the patients who died (p <0.001) (data
not presented in the table).
At the time of the emergency hospital
discharge, most of the patients were
referred to community-based psychiatric
services (76.5%), but a higher proportion
of men (26.5%) compared to women
(15.3%) were referred to full or partial hos-
pitalization (p ¼0004).
Discharge Disposition
Of the 397 patients who survived their
suicide attempts, 82 (20.7%) were referred
to partial (n ¼36) or full (n ¼46) psychi-
atric hospitalization immediately after
discharge from the emergency hospital,
whereas 315 (79.3%) were referred to
community-based services (Figure 1). As
observed in Table 2, in the univariate
analysis sex, the suicide attempt method,
previous suicide attempts, psychiatric treat-
ment, and diagnoses of psychotic or per-
sonality disorders were associated with a
greater likelihood of referral for psychiatric
hospitalization at the time of discharge
from the emergency hospital. After adjust-
ing for possible confounders, male sex
(OR ¼2.11, 95%CI ¼1.25–3.55), violent
methods (OR ¼1.96, 95%CI ¼1.02–3.75)
and a history of psychiatric treatment
(OR ¼2.58, 95%CI ¼1.53–4.36) remained
predictors of referral for full or partial psy-
chiatric hospitalization.
Outcomes Over Two Years
The outcomes over the subsequent 2
years after discharge from the emergency
hospital were examined in 258 (65.0%) resi-
dents of the main city in the region
(Ribeira
˜o Preto). No significant differences
between those included in this assessment
(patients living in Ribeira
˜o Preto) and those
not included (patients from other munici-
palities of the region) were found regarding
sex (p ¼0.612), age (p ¼0.705), ethnicity
(p ¼0.924), schooling (p ¼0.092), occu-
pational status (p ¼0.147), psychiatric diag-
nosis (p ¼0.979), previous suicide attempts
(p ¼0.700) or current psychiatric treatment
(p ¼0.585). However, the proportion of
patients from other municipalities who
used violent methods in the index suicide
attempt (20.1%) was higher than among
patients from the main city of Ribeira
˜o
Preto (10.5%)(p¼0.008).
Of the 258 patients followed, 29
(11.2%) were readmitted to the emergency
department within 2 years following the
index admission; among the readmissions,
Alcine
´ia Donizeti Ferreira et al.
ARCHIVES OF SUICIDE RESEARCH 7
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
TABLE 2. Regression Analyses Examining the Demographic and Clinical Features of Patients
Admitted to an Emergency Hospital due to a Suicide Attempt (N ¼397), as Predictors for
the Discharge Disposition to Partial or Full Psychiatric Hospitalization
Hospitalization Crude OR
(95%CI) P
Adjusted OR
(95%CI)No (315) %Yes (82) %
Sex
Male 36.8 54.9 2.09 (1.28–2.62) 0.003 2.11
(1.25–3.55)
Female 63.2 45.1
Age
Up to 29 years old 50.6 47.6 0.88 (0.54–1.44) 0.620
More than 30 years old 49.4 52.4
Ethnicity
Caucasian 76.5 73.2 0.84 (0.48–1.46) 0.530
African descent 23.5 26.6
Education
Up to 8 years 91.1 95.1 1.90 (0.65–5.58) 0.243
More than 8 years 8.9 4.9
Marital status
Single=Divorced=Widowed 62.5 72.0 1.34 (0.78–2.39) 0.286
Married=Partnered 37.5 28.0
Occupational status
Employed=Homemaker=Student 84.1 78.0 0.67 (0.37–1.23) 0.195
Unemployed 15.9 22.0
Methods
Violent methods 11.4 23.2 2.33 (1.26–4.34) 0.007 1.96
(1.02–3.75)
Self-poisoning 88.6 76.8
Previous suicidal attempts
Yes 30.2 45.1 1.90 (1.16–3.13) 0.011
No or not confirmed 69.8 54.9
Psychiatric treatment
Yes 25.4 43.9 2.30 (1.39–3.81) 0.001 2.58
(1.53–4.36)
No or not confirmed 74.6 56.1
Diagnosis
Depressive disorders 41.6 35.9 1.18 (0.64–2.18) 0.597
Psychotic disorders 5.4 13.4 3.45 (1.42–8.40) 0.006
Personality disorders 17.5 25.6 2.04 (1.03–4.04) 0.042
Others=Unknown 35.6 25.6
Length of stay
Up to 24 hours 52.7 53.7 1.03 (0.64–1.69) 0.877
More than 24 hours 47.3 46.3
Note.
In bold Odds Ratios, 95%Confidence Interval, p <0.05.
Features, Assessment, and Outcome in a Tertiary Emergency Hospital
8 VOLUME 0 NUMBER 0 2016
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
TABLE 3. Regression Analyses Examining Demographic and Clinical Features of Patients Admitted
to an Emergency Hospital Due to a Suicide Attempt (N ¼251), as Predictors for the
Occurrence of a New Suicide Attempt or Confirmed Death due to Suicide in the 2 Years
Following the Index Admission
New suicide attempt Crude OR
(95%CI)
Adjusted OR
(95%CI)No (224)%Yes (27)%
Sex
Female 58.5 63.0 1.21 (0.53–2.76) 0.655
Male 41.5 37.0
Age
Up to 29 years 51.1 44.4 1.17 (0.24–5.59) 0.846
30 to 49 years 38.9 48.1 1.67 (0.35–7.92) 0.523
More than 50 years 10.0 7.4
Ethnicity
Caucasian 76.8 77.8 1.06 (0.41–2.76) 0.908
Afro descendent 23.2 22.2
Education
None=unknown 9.4 7.4 0.40 (0.07–2.30) 0.304
Up to 8 years 81.3 79.1 0.46 (0.16–1.36) 0.160
More than 8 years 9.4 18.5
Marital status
Single 67.9 63.0 0.81 (0.35–1.85) 0.609
Married 32.1 37.3
Occupational status
Active 90.6 92.6 1.29 (0.29–5.85) 0.738
Inactive 9.4 7.4
Methods
Self-poisoning 85.7 91.2 1.09 (0.31–3.91) 0.892
Violent methods 14.3 8.8
Psychiatric assessment
Yes 95.1 92.6 0.65 (0.14–3.08) 0.583
No 4.9 7.4
Suicidal attempts prior
to the index admission
Yes 30.8 48.1 2.09 (0.93–4.67) 0.074 1.48
(0.62–3.53)
No or not confirmed 69.2 51.9
Previous psychiatric
treatment
Yes 25.0 48.1 2.79 (1.24–6.28) 0.014 2.46
(1.07–5.65)
No or not confirmed 75.0 51.9
(Continued )
Alcine
´ia Donizeti Ferreira et al.
ARCHIVES OF SUICIDE RESEARCH 9
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
24 (82.8%) were due to new suicide
attempts. The time between admission
and readmission ranged from 9 to 695 days
(mean ¼222.3, SD ¼190.0, median ¼194).
During the period of this study (from
2006 to 2009), 123 individuals from Ribeira
˜o
Preto, 72.3%male, had suicide as the cause
of death recorded in the Mortality Infor-
mation System. This corresponds to 0.6%
of the total deaths in the county and a mor-
tality rate of 5.1 suicides per 100,000 inhabi-
tants, with 7.8 per 100,000 for males and 2.6
per 100,000 for females.
Of the 258 patients discharged from
the emergency hospital, three cases of
death by suicide were confirmed in the
Mortality Information System. Through
the same system, the deaths of other
patients were confirmed, including 2 due
to accidents, 1 due to homicide and 4 due
to clinical causes. These seven individuals
whose cause of death was other than sui-
cide were excluded from the final analysis,
leaving 251 patients (63.2%of the original
sample) for whom we performed searches
for independent predictors of new suicide
attempts or death due to suicide.
In Table 3, a univariate analysis asso-
ciated psychiatric treatment and referral to
psychiatric hospitalization at discharge
from the emergency hospital with a new
suicide attempt in the subsequent 2 years.
There was also a trend toward an associ-
ation with a history of previous suicide
attempts (p ¼0.074). However, after an
adjustment by logistic regression, only
psychiatric treatment was confirmed as a
predictor of a new suicide attempt or death
in the 2 years after the suicide attempt
(OR ¼2.46, 95%CI ¼1.07 5.65).
DISCUSSION
In this study, we intended to characterize
the profile of suicide attempters admitted
to an emergency hospital of a LAMIC,
considering the paucity of information
regarding the managing of suicide behavior
in emergency settings and also in LAMIC.
We demonstrated that patients with suicide
attempts serious enough to warrant
treatment in a tertiary referral emergency
hospital of LAMIC present a similar profile
TABLE 3. Continued
New suicide attempt Crude OR
(95%CI)
Adjusted OR
(95%CI)No (224)%Yes (27)%
Diagnosis
Depressive disorders 41.5 37,0 1.36 (0.47–3.92) 0.567
Psychotic disorders 7.1 11.1 2.38 (0.54–10.51) 0.254
Personality disorders 17.4 29.6 2.60 (0.84–8.02) 0.097
Others=Unknown 33.9 22.2
Length of stay
Up to 24 hours 58.0 51.9 1.62 (0.35–7.49) 0.540
25 to 72 hours 28.6 40.7 2.58 (0.54–12.37) 0.236
More than 72 hours 13.4 7.4
Referral to hospitalization
Yes 21.9 40.7 2.46 (1.07–5.63) 0.034 2.08 (0.89–4.89)
No or not confirmed 78.1 59.3
Note.
In bold Odds Ratios, 95%Confidence Interval, p <0.05.
Features, Assessment, and Outcome in a Tertiary Emergency Hospital
10 VOLUME 0 NUMBER 0 2016
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
to those observed in other countries
(Mehlum, Jørgensen, Diep, & Nrugham,
2010; Schmidtke et al., 1996): they were
young, with a predominance of women,
diagnosed with severe psychiatric disorders,
already under psychiatric treatment, and
with previous suicide attempts. The great
majority of the patients received a formal
psychiatric diagnosis, mainly depressive
and personality disorders, confirming the
significant association between psychiatric
diagnosis and deaths by suicide (Harris &
Barraclough, 1997).
Additionally, similar to other emerg-
ency departments of LAMIC (Fleischmann
et al., 2005), self-poisoning was the most
common method used in the suicide
attempts, and the use of self-poisoning pre-
dominated in women compared to men,
while men more often chose violent meth-
ods, such as hanging, gunshots, and stab
wounds. Although women are likely to
attempt suicide more often, men make
more serious suicide attempts, with a
consequent higher risk of death by suicide
(Hawton & van Heeringen, 2009). We did
not find a significant difference between
women and men regarding the occurrence
of previous suicide attempts, but this can
be related to the absence of this infor-
mation in the medical records.
The great majority of suicide attemp-
ters admitted to our emergency hospital
underwent a psychiatric assessment, which
is in line with the recommendation of
systematic assessment of suicidal risk at dis-
charge of psychiatric patients (Qin &
Nordentoft, 2005). This rate of psychiatric
assessment was higher than the data from
emergency departments in the United
States, where approximately 50%of
patients admitted for deliberate self-harm
received mental health evaluations (Olfson,
Marcus, & Bridge, 2012). These discrepan-
cies might have occurred because the ser-
vice in our study is that of an academic
hospital, which includes a psychiatric team
among its staff, and an important part of
local network of the mental health services.
Another possible reason is referral bias
because only more severe cases are referred
to our institution. Moreover, the study was
conducted at an emergency general hospital
where more severely injured patients could
be admitted at intermediate beds for a
longer stay, besides the initial evaluation
at the emergency room.
Although the great majority of patients
had undergone a psychiatric evaluation,
there was a lack of recording of relevant
clinical information, such as the occurrence
or not of previous suicide attempts, in a
significant number of medical records. This
is actually an interesting result of this study,
i.e., that despite knowing the importance of
this information, it is frequently missing
from the assessment of suicidal patients.
This finding points to the constant need
for staff training and for the use of specific
protocols to standardize and systematize
care practices.
We found the discharge dispositions to
be similar to those described in the United
States, where most patients admitted to
emergency units due to deliberate self-harm
were discharged to the community (Olfson
et al., 2012). Referral for psychiatric hospi-
talization after the emergency psychiatric
evaluation was determined by a combi-
nation of factors, including the severity of
the suicide attempt (the use of violent meth-
ods) and psychiatric morbidity because
patients with a history of psychiatric treat-
ment were more likely to be referred to full
or partial hospitalization, which is consistent
with the reported data (Miret et al., 2011;
Olfson et al., 2012). The higher proportion
of men referred to inpatient units might have
been due to the presence of psychotic disor-
ders,aswellastheuseofviolentmethods.
Studiesregardingtheriskofdeathdue
to suicide attempts have reported percen-
tages ranging from 2%to 14%(Loas,
2007). Suicide attempt survivors are at high
risk for recurrence because up to 12%of
survivors end up committing suicide, a risk
Alcine
´ia Donizeti Ferreira et al.
ARCHIVES OF SUICIDE RESEARCH 11
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
that is 38 times greater than in the general
population and is greater than any other
psychiatric disorder (Harris & Barraclough,
1997; ten Have et al., 2009)
.
In the present
investigation, 3.6%of the patients died dur-
ing admission to the emergency hospital,
and, among those who survived their suicide
attempts, 3.9%died in the 2 subsequent
years, 3 of them with suicide as the con-
firmed cause of death. Moreover, 9.3%of
the survivors had at least one serious new
suicide attempt that justified a new referral
to an emergency hospital. Taking together,
these results underscore the seriousness of
the suicide attempts treated in this referral
emergency. We should note, however, that
referral to an emergency hospital is most
likely not the case for a significant number
of people who make suicide attempts in
the country. Illustrating this fact, a study
conducted in Campinas, Brazil, found that
only one-third of those who attempted sui-
cide received medical care (Diekstra, 1989).
A history of psychiatric treatment was
significantly associated with the risk of a
new suicide attempt, confirming that ser-
ious mental illness that requires hospitaliza-
tion is among the strongest risk factors for
suicide (Black, Warrack, & Winokur, 1985).
The higher risk of a new suicide attempt
among patients referred to psychiatric hos-
pitalization, observed in the univariate
analysis can be related to the presence of
a severe psychiatric disorder. Although
only a trend toward statistical significance
was observed in the univariate analysis,
patients with prior suicide attempts had a
higher proportion of new suicide attempts
or deaths (47.1%vs. 30.8%) within 2 years.
These results are in agreement with retro-
spective studies, indicating that up to 25%
of all suicides are preceded by non-fatal
attempts and that a history of one or more
suicide attempts is an important predictor
of suicide (Crandall, Fullerton-Gleason,
Aguero, & LaValley, 2006; Diekstra, 1989).
The ratio of death by suicide in the city
of Ribeira
˜o Preto was similar to those
described in the whole country (Lovisi
et al., 2009) and lower than those from sev-
eral countries around the world (Schmidtke
et al., 1996; WHO, 2003). However,
because Brazil is predominantly a Roman
Catholic country, where significant social
stigma is likely to be attached to suicide
and suicidal behavior, we cannot exclude
the possibility of a significant under coding
of the diagnosis of deliberate self-harm and
of suicide as the primary cause of death.
The results of this study should be inter-
preted with caution due to some methodolo-
gical limitations. The data collection was
retrospective and was based on data from
medical records or from the hospital data-
base. Although strategies were employed to
expand the searches for confirmed suicide
attempts, through the use of broader and
nonspecific diagnostic codes, such as inten-
tional self-harm, we cannot completely
exclude the possibility of the absence of
proper registration of suicidal intention in
patients admitted due to self-inflicted injuries.
Inthesamedirection,welostvaluableinfor-
mation for the suicide risk assessment such as
the occurrence of previous suicide attempt.
Another limitation was that for patients
who repeated suicide attempts, we did not
have information on adherence to treatment
as prescribed at their discharge after the first
attempt. This limitation could have weak-
ened our multivariate logistic analysis
because previous or further adherence to
psychiatric treatment was not reported. On
the other hand, we did check for rates of
mortality and cause of death in the sub-
sequent years after the suicide attempt, what
can be considered as strength of this study.
In summary, this study demonstrated
that patients admitted to a referral emerg-
ency hospital due to suicide attempts had
severe psychiatric disorders and, despite
interventions at the time of an emergency
evaluation, these patients continued to
present a high risk of further suicide
attempts or even death. These results
highlight the need for specific prevention
Features, Assessment, and Outcome in a Tertiary Emergency Hospital
12 VOLUME 0 NUMBER 0 2016
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
strategies through more effective care,
according to the individual risk of each
patient. Suicide prevention must be the task
of health professionals from all areas,
including emergency settings, and should
not be restricted to mental health services.
AUTHOR NOTE
Alcine´ia Donizeti Ferreira, Alcion
Sponholz-Jr., Ce´lia Mantovani, and
Anto
ˆnio Pazin-Filho Emergency Hospital
of the Medical School of Ribeira
˜o Preto,
Division of Psychiatry, Department of
Neuroscience and Behavior, Medical
School of Ribeira
˜o Preto, University of
Sa
˜o Paulo, Sa
˜o Paulo, Brazil.
Afonso Dinis Costa Passos, Medical
School of Ribeira
˜o Preto, University of
Sa
˜o Paulo, Brazil, Department of Social
Medicine, Sa
˜o Paulo, Brazil.
Neury Jose´ Botega, Faculty of Medical
Sciences, University of Campinas, Brazil,
Department of Medical Psychology and
Psychiatry, Campinas, Brazil.
Cristina Marta Del-Ben, Emergency
Hospital of the Medical School of Ribeira
˜o
Preto, Division of Psychiatry, Department
of Neuroscience and Behavior, Medical
School of Ribeira
˜o Preto, University of
Sa
˜o Paulo, Sa
˜o Paulo, Brazil.
CM received grants from Cristalia and
Eurofarma for articles in specific publica-
tions, funds from Servier for participation
in a scientific meeting; CMDB received
grants from Janssen Cilag for an Investiga-
tor’s Initiative Study
Correspondence concerning this article
should be addressed to Alcion Sponholz-
Jr., Emergency Hospital of the Medical
School of Ribeira
˜o Preto, Division of
Psychiatry, Department of Neuroscience
and Behavior, Medical School of Ribeira
˜o
Preto, University of Sa
˜o Paulo, Brazil,
Avenida Bandeirantes 3900, Ribeira
˜o Preto,
14048-900 Brazil. E-mail: sponholz@uol.
com.br
FUNDING
CMD-B is supported by research
fellowships from the ‘‘Conselho Nacional
de Desenvolvimento Cientı
´fico e Tecnolo´-
gico’’ (CNPq).
REFERENCES
Adolfi Ju
´nior, M. S., Pallini, F. M., Pessotti, H., Wolf,
C. M., Patelli, H. T., Capeli, R. D., ... Pazin-Filho,
A. (2010). Emergency medical coordination using
a web platform: A pilot study. Rev Saude Publica,44,
1063–1071.
Baraff, L. J., Janowicz, N., & Asarnow, J. R. (2006).
Survey of California emergency departments about
practices for management of suicidal patients and
resources available for their care. Annals of Emerger-
gency Medicine,48, 452–458, 458.e451–452.
doi:10.1016=j.annemergmed.2006.06.026
Bertolote, J. M., Botega, N., & de Leo, D. (2011).
Inequities in suicide prevention in Brazil.
Lancet,378, 1137. doi:10.1016=S0140-6736(11)
61502-X
Bertolote, J. M., Fleischmann, A., De Leo, D., Bolhari,
J., Botega, N., De Silva, D., ... Wasserman, D.
(2005). Suicide attempts, plans, and ideation in
culturally diverse sites: the WHOSUPRE-MISS
community survey. Psychological Medicine,35, 1457–
1465. doi:10.1017=s0033291705005404
Black, D. W., Warrack, G., & Winokur, G. (1985).
The Iowa record-linkage study. I. Suicides and
accidental deaths among psychiatric patients.
Archives of General Psychiatry,42, 71–75.
Crandall, C., Fullerton-Gleason, L., Aguero, R., &
LaValley, J. (2006). Subsequent suicide mortality
among emergency department patients seen for
suicidal behavior. Academy of Emergency Medicine,
13, 435–442. doi:10.1197=j.aem.2005.11.072
Del-Ben, C., Hallak, J., Sponholz, A. J., Marques,
J., Labate, C., Contel, J., & Zuardi, A. (2005).
Accuracy of psychiatric diagnosis performed
under indirect supervision. Rev Bras Psiquiatr,
27, 58–62. doi:S1516-44462005000100013 [pii]=
S1516-44462005000100013
Diekstra, R. F. (1989). Suicide and the attempted sui-
cide: An international perspective. Acta Psychiatrica
Scandinavica Supplement,354, 1–24.
Fleischmann, A., Bertolote, J. M., De Leo, D.,
Botega, N., Phillips, M., Sisask, M., ...Wasserman,
Alcine
´ia Donizeti Ferreira et al.
ARCHIVES OF SUICIDE RESEARCH 13
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
D. (2005). Characteristics of attempted suicides
seen in emergency-care settings of general hospi-
tals in eight low- and middle-income countries.
Psychological Medicine,35, 1467–1474. doi:10.1017=
s0033291705005416
Fleischmann, A., Bertolote, J. M., Wasserman, D.,
De Leo, D., Bolhari, J., Botega, N. J., ... Thanh,
H. T. T. (2008). Effectiveness of brief intervention
and contact for suicide attempters: A randomixed
controlled trial in five countries. Bulletin of the World
Health Organization,86, 703–709. doi:10.2471=
blt.07.046995
Gunnell, D., & Frankel, S. (1994). Prevention of
suicide: aspirations and evidence. BMJ,308,
1227–1233.
Harris, E. C., & Barraclough, B. (1997). Suicide as an
outcome for mental disorders. A meta-analysis.
British Journal of Psychiatry,170, 205–228.
Haukka, J., Suominen, K., Partonen, T., &
Lo
¨nnqvist, J. (2008). Determinants and outcomes
of serious attempted suicide: a nationwide study in
Finland, 1996–2003. American Journal of Epidemi-
ology,167, 1155–1163. doi:10.1093=aje=kwn017
Hawton, K., & van Heeringen, K. (2009). Suicide.
Lancet,373, 1372–1381. doi:10.1016=S0140-
6736(09)60372-X
IBGE. (2010). Instituto Brasileiro de Geografia e
Estatı
´stica (IBGE) Cidades. Retrieved 01=08=
2012, from http://www.ibge.gov.br/cidadesat
Jo, S. J., Lee, M. S., Yim, H. W., Kim, H. J., Lee, K.,
Chung, H. S., ... Seo, Y. M. (2011). Factors asso-
ciated with referral to mental health services
among suicide attempters visiting emergency
centers of general hospitals in Korea: Does history
of suicide attempts predict referral? General
Hospital Psychiatry,33, 294–299. doi:10.1016=j.
genhosppsych.2011.01.008
Loas, G. (2007). Anhedonia and suicide: A 6.5-yr.
follow-up study of patients hospitalised for a sui-
cide attempt. Psychological Report,100, 183–190.
Lovisi, G. M., Santos, S. A., Legay, L., Abelha, L., &
Valencia, E. (2009). [Epidemiological analysis of sui-
cide in Brazil from 1980 to 2006]. Rev Bras Psiquiatr,
31, S86–94. doi:S1516-44462009000600007 [pii]
Mehlum, L., Jørgensen, T., Diep, L. M., & Nrugham,
L. (2010). Is organizational change associated with
increased rates of readmission to general hospital
in suicide attempters? A 10-year prospective
catchment area study. Archives of Suicide Research,
14, 171–181. doi: 10.1080=13811111003704811
Miret, M., Nuevo, R., Morant, C., Sainz-Corto´n, E.,
Jime´nez-Arriero, M., Lo´ pez-Ibor, J. J., ...
Ayuso-Mateos, J. L. (2011). The role of suicide
risk in the decision for psychiatric hospitalization
after a suicide attempt. Crisis,32, 65–73. doi:
10.1027=0227-5910=a000050
Olfson, M., Marcus, S. C., & Bridge, J. A. (2012).
Emergency treatment of deliberate self-harm.
Archives of General Psychiatry,69, 80–88.
doi:10.1001=archgenpsychiatry.2011.108
Owens, D., Horrocks, J., & House, A. (2002). Fatal
and non-fatal repetition of self-harm. Systematic
review. British Journal of Psychiatry,181, 193–199.
Qin, P., & Nordentoft, M. (2005). Suicide risk in
relation to psychiatric hospitalization: Evidence
based on longitudinal registers. Archives of General
Psychiatry,62, 427–432. doi:10.1001=archpsyc.
62.4.427
Schmidtke, A., Bille-Brahe, U., DeLeo, D., Kerkhof,
A., Bjerke, T., Crepet, P., ... Sampaio-Faria, J. G.
(1996). Attempted suicide in Europe: Rates, trends
and sociodemographic characteristics of suicide
attempters during the period 1989–1992. Results
of the WHO=EURO Multicentre Study on Para-
suicide. Acta Psychiatrica Scandavica,93, 327–338.
ten Have, M., de Graaf, R., van Dorsselaer, S.,
Verdurmen, J., van’t Land, H., Vollebergh, W.,
& Beekman, A. (2009). Incidence and course of
suicidal ideation and suicide attempts in the
general population. Canadian Journal of Psychiatry,
54, 824–833.
Waiselfisz, J. J. (2011). Mapa da viole
ˆ
ncia 2011: Os
jovens no Brasil.Sa
˜o Paulo, SP Brasilia, DF: Insti-
tuto Sangari Ministe´rio da Justic
¸a.
WHO. (1993). ICD-10 Classification of mental and beha-
vioural disorders. Geneva: World Health Organization.
WHO. (2003). The World Health Report 2003:
Shaping the future.
Features, Assessment, and Outcome in a Tertiary Emergency Hospital
14 VOLUME 0 NUMBER 0 2016
Downloaded by [Universidade de Sao Paulo], [Alcion Sponholz Junior] at 16:38 04 February 2016
