Beruflich Dokumente
Kultur Dokumente
Objective: The purpose of this study was to analyze patterns of mental evaluated at 80% to 100% among sui-
health–related service utilization before and after hospitalization for at- cide completers (4,5). The most com-
tempting suicide. Methods: This retrospective cohort study included all mon diagnoses of completers include
persons 15 years or older with a clinical diagnosis of schizophrenia depression, substance abuse, person-
(N=195) or depression (N=330) hospitalized in Montreal, Quebec, from ality disorder, and a substance use
April 2003 to December 2004 for attempting suicide. Data on the pub- disorder that is comorbid with a men-
licly managed health and social services system were retrieved from the tal disorder (6,7). Lifetime risk of sui-
linked administrative databases of Montreal’s Health and Social Ser- cide has been estimated at 3% to 4%
vices Agency (April 2002 to March 2005). Twelve-month preattempt among people with a common mental
service utilization profile, health care contacts three months pre- and disorder such as depression (8,9) and
postattempt, and predictors of postattempt service utilization were an- at 4% to 5% among those with a se-
alyzed for two diagnostic groups (schizophrenia and depression). Re- vere and persistent mental disorder
sults: Specialized outpatient care and hospital emergency departments such as schizophrenia (10–12).
were the services most used by both groups before and after attempting Because previous suicide attempts
suicide. Use of hospital emergency services as a primary care service did are another key correlate of suicide
not adequately ensure aftercare, whereas prior contact with services (1,2), health care systems can play a
and concurrent substance use disorder predicted greater service uti- major role in suicide prevention (13).
lization postattempt among men but not women. Conclusions: The pub- Indeed, contacts with services are
licly managed health and social services system in Montreal seems to re- common in the year preceding sui-
spond rather well to severe suicide attempts, including those by men cide. A systematic review of suicide
with a concurrent substance use disorder known to be at high suicide cases showed that although nearly
risk. However, better coordination among hospital emergency depart- 78% of completers had made at least
ments, primary care, specialized mental health services, and addiction one visit to a general practitioner, only
services is needed in order to enhance continuity of care. (Psychiatric about 32% had seen a mental health
Services 63:364–369, 2012; doi: 10.1176/appi.ps.201000405) professional (14). Similar but lower
figures were reported for young adult
completers in Quebec in the early
W
orldwide, suicide is one of Men outnumber women 4 to 1 in this 1990s (15). Access to general practi-
the three leading causes of regard. Moreover, there are 20 times tioners is widespread in the publicly
death among 15- to 44- as many suicide attempts as suicide managed Canadian health care sys-
year-olds. It is responsible for nearly completions (1–3). tem. In 2006, 75% of the general pop-
half of all violent deaths and accounts Mental disorder is a key correlate ulation made at least one visit to a
for one million fatalities every year. of suicide. Its prevalence has been general practitioner (16). Among
these service users, one in five was di-
agnosed as having a mental disorder
Ms. Routhier is affiliated with the Psychosocial Research Division, Douglas Mental Health
(17). However, although contacts with
University Institute, Perry Pavillion (E-3311), 6875 LaSalle Blvd., Montreal, Quebec services are common, few suicide
H4H 1R3, Canada (e-mail: routhier.danielle@gmail.com). Dr. Leduc is with the Depart- completers seem to engage in ade-
ment of Health Administration, University of Montreal. Dr. Lesage is with Centre De quate treatment (14,18). Moreover,
recherche Fernand-Seguin, Hôpital L.-H. Lafontaine, University of Montreal. Dr. Be- poor coordination between primary
nigeri is with the Department of Social and Preventive Medicine, University of Montreal. care services, specialized medical
364 PSYCHIATRIC SERVICES o ps.psychiatryonline.org o April 2012 Vol. 63 No. 4
services, and addiction services is a million residents of Montreal, Canada’s Measures
barrier to suicide prevention (14). second largest city, located in the Mental health–related contact with
Although research has identified province of Quebec. As in the rest of services in the three months postat-
predisposing factors in service utiliza- Canada, the health and social services tempt was the primary outcome in-
tion, few studies have focused on system in Quebec is publicly funded vestigated. This was defined as con-
mental health service use in suicidal and offers citizens universal access to tact with any health care or psychoso-
populations. Clinical guidelines rec- medically required care, hospital cial resource for mental health rea-
ommend that individuals with a se- emergency services, hospitalization, sons (at least one contact versus
vere mental disorder such as schizo- outpatient general and specialized none). Mental health–related contact
phrenia should be in the care of spe- care, and general practitioners’ clinics. with services in the three months pre-
cialized services (19) and that those There are no private hospitals. In addi- ceding the attempt was defined simi-
with a common mental disorder such tion, in Quebec, psychosocial services larly to the definition for the depend-
as depression should be treated by are available in local community health ent variable (at least one contact ver-
primary care services, with more centers (CLSCs); these are considered sus none). Profiles of service use in
complex cases in shared care with primary care services, as are general the 12 months before the attempt
specialized services (20). These pat- practitioners’ clinics. Thirteen differ- were defined and classified hierarchi-
terns of mental health–related service ent general hospitals and four special- cally according to one of five levels of
use were observed in Montreal in the ized psychiatric hospitals are available treatment settings, listed in ascending
general population (21), but it is not to treat suicide attempters in Montreal. order (21): general practitioners’ clin-
known whether these patterns apply Individuals who visited hospital emer- ics, hospital emergency services,
also to the population of hospitalized gency departments and were treated CLSC, psychiatric outpatient care,
suicide attempters. Higher rates of on site without being hospitalized were and psychiatric inpatient care (hospi-
contact with services have been ob- not considered in this study. talization). Although profiles could
served among women compared with Of all the Montrealers 15 years old include services obtained at more
men, older adults compared with or older diagnosed by a physician as than one level, they were classified
youths, and self-poisoners compared having schizophrenia or depression according to the highest level
with those who attempt suicide by during the study period (April 2002 to reached. Other independent vari-
other means (18,22,23). However, in- March 2005), 525 were hospitalized ables included primary clinical diag-
dividuals with a substance use disor- for attempting suicide and discharged nosis (schizophrenia versus depres-
der tend to show lower rates of con- during the index period (April 2003 to sion), comorbid substance use disor-
tact with services (24–26). Previous December 2004). If an individual was der (yes versus no), method of at-
studies have examined contacts with hospitalized more than once for at- tempt (self-poisoning versus other),
services mostly in the general popula- tempting suicide during the index pe- sex (female versus male), and age.
tion or among suicide completers. riod, the first event was considered Clinical diagnoses were classified
Little attention has been paid to sui- and designated as the index attempt. according to the ICD-9: schizophre-
cide attempters with specific com- Data were retrieved from the linked nia, codes 295.0–295.9, 297.0–297.9,
mon or severe mental disorders. administrative databases of Montreal’s and 298.0–298.9; depression, 296.1,
Against this backdrop, we sought to Health and Social Services Agency. 296.3, 300.4, and 311.0–311.9; and
analyze patterns of mental health–re- Corresponding data from the public substance use disorders, 291.0–291.9,
lated service utilization among persons sector were previously retrieved from 292.0–292.9, 303.0–303.9, 304.0–
with a severe mental disorder (schizo- Quebec’s physician compensation 304.9, and 305.0–305.9. Given that
phrenia) or a common mental disorder database (RAMQ), the Quebec hospi- dual diagnoses were possible accord-
(depression) who were hospitalized in talization discharge database (MED- ing to primary diagnosis (schizophre-
Montreal after attempting suicide. ECHO), and the CLSC interventions nia or depression, plus a secondary
Our objectives were threefold: to com- database (I-CLSC). Medical insurance diagnosis of a substance use disor-
pare profiles of service use between numbers were used to link data for a der), we applied a hierarchical algo-
the two diagnostic groups in the 12 given individual from one database to rithm to ensure that schizophrenia
months before attempt, to identify fac- another. These were available only for did not preclude depression but that
tors influencing postattempt contact individuals in the Montreal area be- depression precluded schizophrenia.
with services, and to compare diagnos- cause the system had yet to be imple- We selected cases of hospitalized sui-
tic groups in terms of contact with mented in the rest of the province. cide attempters on the basis of self-
services three months pre- and postat- The agency was granted permission by inflicted external cause of injury E-
tempt, with a particular focus on serv- the Quebec Data Privacy Commission codes (self-poisoning, E950–E952,
ice utilization at the levels of primary to use the data for service planning versus other causes, E953–E959).
care and specialized medicine. and research purposes. The medical
insurance numbers were encrypted to Statistical analysis
Methods ensure confidentiality. The study was Chi square tests were carried out to
Study design and data sources approved by the Ethics Board of the compare service use profiles between
The sampling frame for this retrospec- Faculty of Medicine of the University the two diagnostic groups (schizo-
tive cohort study consisted of the 1.8 of Montreal. phrenia versus depression) and men-
PSYCHIATRIC SERVICES o ps.psychiatryonline.org o April 2012 Vol. 63 No. 4 365
Figure 1 stance use disorder. These charac-
Twelve-month preattempt service use profile of 525 hospitalized suicide teristics did not differ significantly
attempters with schizophrenia or depressiona between the two diagnostic groups.
Self-poisoning was more common in
100 the depression group (N=257, 78%)
90
than in the schizophrenia group
80
No contact (N=130, 67%, p=.005). Other meth-
70
Percentage