Você está na página 1de 7

Research

JAMA Psychiatry | Original Investigation

Association of an Early Intervention Service for Psychosis


With Suicide Rate Among Patients With First-Episode
Schizophrenia-Spectrum Disorders
Sherry Kit Wa Chan, MRCPsych; Stephanie Wing Yan Chan, BSc; Herbert H. Pang, PhD; Kang K. Yan, MSc;
Christy Lai Ming Hui, PhD; Wing Chung Chang, MRCPsych; Edwin Ho Ming Lee, MRCPsych; Eric Yu Hai Chen, MD

Editorial
IMPORTANCE Patients with schizophrenia have a substantially higher suicide rate than the Supplemental content
general public. Early intervention (EI) services improve short-term outcomes. However, little
is known about the association of EI with suicide reduction in the long term.

OBJECTIVE To examine the association of a 2-year EI service with suicide reduction in patients
with first-episode schizophrenia-spectrum (FES) disorders during 12 years and the risk factors
for early and late suicide.

DESIGN, SETTING, AND PARTICIPANTS This historical control study compared 617 consecutive
patients with FES who received the 2-year EI service in Hong Kong between July 1, 2001, and
June 30, 2003, with 617 patients with FES who received standard care (SC) between July 1,
1998, and June 30, 2001, matched individually. Clinical information was systematically
retrieved for the first 3 years of clinical care for both groups. The details of death were
collected up to 12 years from presentation to the services. Data analysis was performed from
October 30, 2016, to August 18, 2017.

MAIN OUTCOMES AND MEASURES Suicide rate during 12 years was the primary measure. The
association of the EI service with the suicide rates during years 1 through 3 and years 4
through 12 were explored separately.

RESULTS The main analysis included 1234 patients, with 617 in each group (mean [SD] age at
baseline, 21.2 [3.4] years in the EI group and 21.3 [3.4] years in the SC group; 318 male [51.5%]
in the EI group and 322 [52.2%] in the SC group). The suicide rates were 7.5% in the SC group
and 4.4% in the EI group (McNemar χ2 = 5.55, P = .02). Patients in the EI group had
significantly better survival (propensity score–adjusted hazard ratio, 0.57; 95% CI, 0.36-0.91;
P = .02), with the maximum association observed in the first 3 years. The number of suicide
attempts was an indicator of early suicide (1-3 years). Premorbid occupational impairment,
number of relapses, and poor adherence during the initial 3 years were indicators of late
suicide (4-12 years).

CONCLUSIONS AND RELEVANCE This study suggests that the EI service may be associated with
reductions in the long-term suicide rate. Suicide at different stages of schizophrenia was Author Affiliations: Department of
associated with unique risk factors, highlighting the importance of a phase-specific service. Psychiatry, Li Ka Shing Faculty of
Medicine, The University of Hong
Kong, Hong Kong (S. K. W. Chan,
S. W. Y. Chan, Hui, Chang, Lee, Chen);
The State Key Laboratory of Brain
and Cognitive Sciences, The
University of Hong Kong, Hong Kong
(S. K. W. Chan, Chang, Chen); School
of Public Health, Li Ka Shing Faculty
of Medicine, The University of Hong
Kong, Hong Kong (Pang, Yan).
Corresponding Author: Sherry Kit
Wa Chan, MRCPsych, Department of
Psychiatry, Li Ka Shing Faculty of
Medicine, The University of Hong
Kong, New Clinical Building, Room
JAMA Psychiatry. doi:10.1001/jamapsychiatry.2018.0185 219, 102 Pokfulam Rd, Hong Kong
Published online April 4, 2018. (kwsherry@hku.hk).

(Reprinted) E1
© 2018 American Medical Association. All rights reserved.

Downloaded From: by Clinica Alemana de Santiago S.A., Juan Undurraga on 04/04/2018


Research Original Investigation Early Intervention Service for Psychosis and Suicide Among Patients With First-Episode Schizophrenia-Spectrum Disorders

P
atients with schizophrenia have a substantially higher
mortality rate and suicide rate in particular than the gen- Key Points
eral population.1,2 A systematic review1 suggested that
Questions Is an early intervention service associated with a
the median standardized mortality ratio (SMR) for all-cause reduction in the long-term suicide rate in patients with
mortality among patients with schizophrenia was 2.58 (mean, schizophrenia-spectrum disorders?
2.98) and the median SMR for suicide was 12.86 (mean, 43.47).
Findings In this historical control study of 1234 patients with
The lifetime risk of suicide among individuals with schizo-
first-episode schizophrenia-spectrum disorders (617 each in the
phrenia was 4% to 6%,3,4 and the population-attributable risk early intervention and standard care groups), patients receiving a
of suicide was estimated to be 8.9% among those with 2-year early intervention service had a significantly lower suicide
schizophrenia.5 The most accurate method to study the sui- rate during 12 years, with the main difference observed during the
cide rate would be to follow up patients with first-episode first 3 years.
schizophrenia (FES) to avoid the selection bias of the survi- Meaning An early intervention service may be associated
vors of the high-risk period immediately after the first admis- with reductions in the suicide rate among patients with
sion. Such studies6-8 are relatively limited. schizophrenia-spectrum disorders during their most vulnerable
Many studies6,8-14 have investigated the risk factors for sui- period, and the benefits may persist in the long term.
cide and suicide attempts in patients with schizophrenia. Pre-
vious systematic reviews suggested that past suicidal ide- was 87%.28 The current study aimed to evaluate the associa-
ation, prior self-harm, previous depressive episode, substance tion of the EI service with the suicide rate among patients with
misuse, poor drug adherence, and a high number of hospital- FES disorders during a 12-year period with a large representa-
izations were risk factors for suicide attempts9 and suicides10 tive sample in Hong Kong. Both short- and long-term associa-
in patients with schizophrenia. Studies6,8,11-14 on the risk fac- tions were evaluated. Risk factors associated with early suicide
tors for suicide in first-episode psychosis (FEP) have reported (years 1-3) and late suicide (years 4-12) were explored sepa-
that previous suicide attempts, substance misuse, young age, rately. The results of the current study can enrich our under-
male sex, hopelessness, and depressive and psychotic symp- standing of the association of an EI service with the suicide rate
toms at the first- and second-year follow-ups were signifi- and provide a direction for further service development.
cantly associated with subsequent suicide.
Knowledge about risk factors can inform the service de-
velopment to reduce the suicide rate. There has been an ac-
tive development of the early intervention (EI) service for FEP Methods
worldwide, with an aim to improve the outcomes for patients Study Design and Sample Identification
with psychosis.15 Despite the variation in the content of the EI A historical control design was adopted because the EASY pro-
services, most adopt an intensive integrated service model. gram was implemented as a regionwide service in Hong Kong
Many studies have reported the benefits of an EI service in im- in 2001.26 The samples were selected with close temporal prox-
proving short-term clinical and functional outcomes, includ- imity to minimize the potential cohort effect. Furthermore, 617
ing randomized clinical trials 16-20 and historical control consecutive patients enrolled in the EASY program for the first
studies,21-23 when the services were introduced as a region- time in the entire region between July 1, 2001, and June 30,
wide program. Many studies24,25 found that an EI program re- 2003 (the first 2 years of the program), with a diagnosis of
duced suicidal attempts, but few studies6,19,20 reported a posi- schizophrenia-spectrum disorders were identified from the
tive outcome on reduction of completed suicide. This finding centralized hospital database (Clinical Management System
was mainly because mortality and suicide are rare events, and [CMS]). Then, 617 patients who received the standard general
therefore a considerably large sample size and long follow-up psychiatric service between July 1, 1998, and June 30, 2001,
are required. With a 5-year follow-up period, the OPUS study, in the entire region matched individually by sex, diagnosis, and
a randomized clinical trial with the largest sample size for a age were identified from the same system as the comparison
randomized clinical trial comparing outcomes of EI and stan- group.7,8,14,22 Details of the matching process are reported in
dard care (SC) (n = 547), did not find any statistically signifi- the eMethods in the Supplement. Patients with organic con-
cant difference in the mortality or suicide rate between the EI ditions, drug-induced psychosis, or intellectual disability (men-
and SC groups.6 A well-designed historical control study evalu- tal retardation) were excluded from the sample. Patients who
ating the long-term outcomes of a regionwide EI service can had received prior psychiatric treatment for more than 1 month
perhaps obtain a sufficient sample size to study the potential were also excluded. Details on the number of excluded indi-
association of an EI service with the suicide rate. viduals are reported in eTable 1 in the Supplement. Data analy-
Hong Kong had a population of 6.7 million in 2001. The EI sis was performed from October 30, 2016, to August 18, 2017.
service in Hong Kong, Early Assessment Service for Young The requirement of informed consent from individual pa-
People with Psychosis (EASY), was established as a regionwide tients was waived by the institutional review boards and eth-
service in 2001 to provide 2-year phase-specific intervention to ics committees of the considered sites that provided institu-
patients with FEP in the age range of 15 to 25 years. It is an as- tional review board approval (Hong Kong East Cluster Research
sertive intervention provided by the key workers based on a Ethics Committee, Institutional Review Board of the Univer-
specifically developed protocol, the Psychological Interven- sity of Hong Kong/Hospital Authority Hong Kong West Clus-
tion Program in Early Psychosis.22,26,27 The service retention rate ter, Research Ethics Committee [Kowloon Central/Kowloon

E2 JAMA Psychiatry Published online April 4, 2018 (Reprinted) jamapsychiatry.com

© 2018 American Medical Association. All rights reserved.

Downloaded From: by Clinica Alemana de Santiago S.A., Juan Undurraga on 04/04/2018


Early Intervention Service for Psychosis and Suicide Among Patients With First-Episode Schizophrenia-Spectrum Disorders Original Investigation Research

East], Kowloon West Cluster Research Ethics Committee quality assurance. Validity and interrater reliability for major
[Kowloon East Cluster], Research Ethics Committee Joint variables, including DUP and the CGI-SCH positive and nega-
Chinese University of Hong Kong–New Territories East Clus- tive scales, were evaluated using records of 12 patients with
ter Clinical & Research Ethics Committee, and New Territo- the intraclass correlation coefficient (ICC) comparing the cli-
ries West Cluster Research Ethics Committee). nicians and research staff. The test results (DUP: ICC = 0.78,
CGI-SCH positive: ICC = 0.89, and CGI-SCH negative:
Data Collection ICC = 0.77) revealed a satisfactory level of concordance.
Mortality Information
All deaths within the cohort during the 12-year follow-up Statistical Analysis
period were identified via the computerized CMS. The CMS is The SMRs and 95% CIs for all-cause mortality and suicide were
the centralized clinical database system for the public health calculated. Details of the calculation of the SMR are reported
care service, which is responsible for more than 90% of the hos- in the eMethods in the Supplement. The McNemar χ2 test was
pital health care in Hong Kong. All clinical information, in- used to detect significant difference in suicides between groups.
cluding mortality status, is recorded in the CMS. However, P < .05 (2-sided) was considered to be statistically significant.
those who had emigrated and died abroad were not able to be Overall suicide survival was defined as the time between
identified. For all identified deaths, details, including the date, the entry point and the date of suicide death. The date of first
reason, and location of death, were obtained from the coro- presentation to the service was used as the entry point. Par-
ner’s court report. Deaths caused by intentional self-harm and ticipants were censored if they did not die at the date of 12 years
events of undetermined intent were coded as suicides.29 after the first presentation or if they died of causes other than
suicide. Multivariable Cox proportional hazards regression
Demographic and Clinical Information accounting for correlated groups of observations33 was used
All demographic and clinical information of 1234 patients dur- to model the overall suicide survival in the 2 treatment groups
ing the initial 3 years was obtained from the CMS and written with a change point at year 3 to allow for different hazards be-
clinical records by trained researchers. Data were systemati- fore and after this point while adjusting for sex, age, age at on-
cally retrieved each month for the first 3 years using a standard- set, and years of education. To compare the survival differ-
ized data entry form based on operationalized definitions. For ences between the 2 treatment groups, the propensity score
the patients in the EI group, this covered the 2-year EI service approach with inverse probability weighting was used to re-
and the transitional year after the EI service facilitated the duce the bias of nonconcurrent samples. The details of the
gradual transfer of patients to the general psychiatric service. unweighted and propensity score–reweighted patient charac-
The demographic variables included were sex, age, and teristics reported according to the brief guidelines34 are given
years of education. The other baseline data obtained in- in eTable 2 in the Supplement. The propensity score–
cluded diagnosis, premorbid occupational functioning, age at adjusted hazard ratios (HRs) for year 3 or earlier and later than
onset, duration of untreated psychosis (DUP), number of sui- year 3 for suicide were reported. The number needed to treat
cide attempts during DUP, and positive, negative, and affec- was calculated using a parallel group approach.35 SPSS statis-
tive symptoms at onset measured with the Clinical Global tical software, version 23.0 (SPSS Inc) and R, version 3.3.2 (The
Impressions–Schizophrenia (CGI-SCH) scale.30 Duration of R Foundation) were used for the statistical analysis.
untreated psychosis was defined as the period (in days) be- Cox proportional hazards regression accounting for cor-
tween the first appearance of psychotic symptoms and the use related group analysis was used to explore the association be-
of successful psychiatric treatment as assessed by clinicians. tween the demographic and clinical information and the time
Clinical information during the first 3 years included the num- to suicide for years 1 through 3, years 4 through 12, and the en-
ber of hospitalizations, substance misuse, number of suicide tire 12-year period. Cox proportional hazards regression con-
attempts, number of relapses, medication adherence, clini- siders not only time to event information but also event cen-
cal symptoms measured with the CGI-SCH scale, and mean de- soring. Patients who died during the first 3 years were excluded
fined daily dose31 of antipsychotic medication. The baseline from the analysis of the indicators of time to suicide during
diagnosis was determined by the clinician on the basis of the years 4 through 12. Last digit carry forward was used to man-
International Statistical Classification of Diseases and Related age any missing variables. All statistically significant vari-
Health Problems, Tenth Revision criteria using all available clini- ables were used for the multivariable Cox proportional haz-
cal information. Adherence to medication for each patient was ards regression with backward selection accounting for
given a score of 1 to 3 for each month, and a mean of 3 years correlated group analysis to determine the model signifi-
was generated, with 1 indicating good adherence and 3 indi- cance with adjustment for treatment group.
cating poor adherence. Relapse was operationally defined as
a change in the CGI-SCH scores from 1 to 3 or from 4 through
6 to 5 through 7, followed by hospitalization or adjustment of
antipsychotic medication.32
Results
The main analysis included 1234 patients, with 617 in each
Validity and Reliability group (mean [SD] age at baseline, 21.2 [3.4] years in the EI group
Consensus meetings were conducted among the clinicians and and 21.3 [3.4] years in the SC group; 318 male [51.5%] in the EI
researchers every 2 weeks during the data collection period for group and 322 [52.2%] in the SC group). Baseline demo-

jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online April 4, 2018 E3

© 2018 American Medical Association. All rights reserved.

Downloaded From: by Clinica Alemana de Santiago S.A., Juan Undurraga on 04/04/2018


Research Original Investigation Early Intervention Service for Psychosis and Suicide Among Patients With First-Episode Schizophrenia-Spectrum Disorders

Table 1. Baseline Characteristics of the EI and SC Groupsa

Characteristic EI (n = 617) SC (n = 617) χ2 or zb P Value


Age, mean (SD), y 21.2 (3.4) 21.3 (3.4) NA NA
Male sex 318 (51.5) 322 (52.2) NA NA
Diagnosis
Schizophrenia and schizoaffective disorder 484 (78.4) 486 (78.8) NA NA
Acute and transient psychotic disorder 87 (14.1) 87 (14.1) NA NA
Psychosis not otherwise specified 46 (7.5) 44 (7.1) NA NA
Educational level, mean (SD), y 10.9 (2.37) 10.5 (2.44) −3.186 .001
Age at onset, mean (SD), y 20.46 (3.38) 20.64 (3.59) −2.746 .006
DUP, median, d 92 61 −2.588 .01
Baseline positive symptoms 4.18 (0.72) 4.59 (0.96) −7.439 <.001
Baseline negative symptoms 2.68 (1.29) 2.68 (1.31) −0.025 .98
Baseline affective symptoms 2.13 (1.18) 2.02 (1.21) −1.936 .05
Suicide attempts during DUP 0.13 (0.54) 0.26 (0.74) −3.163 .002
Premorbid occupational functioningc
Impaired 47 (7.6) 48 (7.8) 0.00 >.99
Not impaired 570 (92.4) 569 (92.2) NA NA
Abbreviations: DUP, duration of untreated psychosis; EI, early intervention; patients with a diagnosis of schizophrenia.
NA, not applicable; SC, standard care. b
The z values are derived from the Wilcoxon test for paired samples; the χ2
a
Data are presented as number (percentage) of study participants unless values are derived from the McNemar test.
otherwise indicated. The EI and SC patients were individually matched based c
Impaired premorbid occupational functioning indicates unemployment or
on age (±2 years), sex, and diagnosis; 548 pairs were successfully matched. prolonged educational stagnation before the onset of illness.
For unmatched EI patients, diagnosis was matched first and then sex. Two EI
patients with psychosis not otherwise specified were matched with SC

Figure. Number Needed to Treat for Each Year During the 12 Study Years Table 2. Propensity Score–Adjusted Cox Proportional Hazards
Regression Full Model for Survival
70
Variable HR (95% CI) P Value
60 Intervention
Standard care (control) 1 [Reference] NA
50
No. Needed to Treat

Early intervention 0.57 (0.36-0.91) .02


40 Sex
Male 1.56 (0.95-2.56) .08
30
Female 1 [Reference] NA
20 Age (1-y increase) 0.99 (0.83-1.17) .87
Age at onset (1-y increase) 1.03 (0.87-1.21) .76
10
Educational level (1-y increase) 0.93 (0.85-1.01) .10
0
1 2 3 4 5 6 7 8 9 10 11 12 Abbreviations: HR, hazard ratio; NA, not applicable.
Year

(Table 2). Kaplan-Meier plots for suicide survival are shown


graphic and clinical comparisons between the groups are re- in the eFigure in the Supplement. The propensity score–
ported in Table 1. In all, 77 patients (6.2%) died during the 12- adjusted HRs for comparing overall suicide survival between
year period, and 73 patients (5.9%) were considered to have the EI and SC groups were 0.32 (95% CI, 0.14-0.73; P = .007)
completed suicide, including 27 (4.4%) in the EI group and 46 for year 3 or earlier and 0.87 (95% CI, 0.48-1.58; P = .65) for later
(7.5%) in the SC group (McNemar’s χ2 = 5.55, P = .02). Details than year 3 (eTable 4 in the Supplement).
of cause of death and suicide are given in eTable 3 in the Supple-
ment. The SMRs for all-cause mortality were 5.67 (95% CI, 3.84- Risk Factors for Suicide
8.09) in the EI group and 8.38 (95% CI, 6.27-10.99) in the SC Patients with a high number of suicide attempts during DUP and
group. The SMRs for suicide were 28.01 (95% CI, 18.84-40.19) the first 3 years of the intervention were more likely to complete
in the EI group and 44.66 (95% CI, 33.08-59.06) in the SC group. suicide during years 1 through 3 (Table 3). The multivariable
The number needed to treat was 61.7 at year 1 and 29.4 at year analysis revealed that the model with these 2 variables was sig-
12, with a sharp decrease between years 1 and 3 (Figure). nificant (likelihood ratio test χ2 = 20.06, P < .001) (Table 4).
Patients with impaired premorbid occupational function-
Survival Analysis ing and a relatively high incidence of relapse and poor adher-
A significant association of EI with suicide survival was ob- ence during the first 3 years were more likely to complete sui-
served during the 12 years (HR, 0.57; 95% CI, 0.36-0.91; P = .02) cide during years 4 through 12 (Table 3). The model was

E4 JAMA Psychiatry Published online April 4, 2018 (Reprinted) jamapsychiatry.com

© 2018 American Medical Association. All rights reserved.

Downloaded From: by Clinica Alemana de Santiago S.A., Juan Undurraga on 04/04/2018


Early Intervention Service for Psychosis and Suicide Among Patients With First-Episode Schizophrenia-Spectrum Disorders Original Investigation Research

Table 3. Univariate Association of Demographic and Clinical Variables With Suicide Deatha

Suicide Death, HR (95% CI)


Years 1-3 Years 4-12 12 Years
Demographic (n = 34) (n = 39) (n = 73)
Age at onset 0.99 (0.90-1.09) 1.02 (0.92-1.13) 1.01 (0.94-1.08)
Educational level 0.94 (0.83-1.06) 0.89 (0.78-1.00)b 0.91 (0.84-0.99)b
Sex (reference group: female) 1.34 (0.69-2.63) 1.70 (0.88-3.31) 1.52 (0.94-2.47)
Baseline variables
DUP, d 1.09 (0.86-1.38) 1.00 (0.83-1.20) 1.04 (0.89-1.21)
Baseline positive symptoms 1.03 (0.65-1.64) 0.97 (0.70-1.35) 1.00 (0.76-1.31)
Baseline negative symptoms 0.96 (0.74-1.24) 1.10 (0.85-1.41) 1.03 (0.86-1.23)
Baseline affective symptoms 1.11 (0.83-1.47) 0.94 (0.70-1.28) 1.02 (0.83-1.26)
Premorbid occupational impairment 0.76 (0.18-3.17) 2.69 (1.17-6.13)b 1.72 (0.85-3.45)
(reference group: not impaired)
No. of suicide attempts during DUP 1.64 (1.24-2.17)c 1.32 (0.89-1.96) 1.48 (1.18-1.86)c
Variables of the first 3 y Abbreviations: DDD, defined daily
No. of suicide attempts 1.95 (1.39-2.73)c 1.50 (0.96-2.34) 1.72 (1.30-2.78) dose; DUP, duration of untreated
psychosis; HR, hazard ratio.
No. of hospitalization 1.00 (0.96-1.05) 1.03 (0.98-1.08) 1.02 (0.98-1.05)
a
Cox proportional hazards regression
No. of relapse 0.95 (0.70-1.28) 1.34 (1.03-1.75)a 1.17 (0.95-1.43) was used accounting for correlated
d d
Adherence 1.63 (0.79-3.34) 2.55 (1.32-4.94) 2.11 (1.31-3.40) groups of observations.
b
Presence of substance abuse (reference 0.56 (0.19-1.59) 0.89 (0.28-2.80) 0.70 (0.32-1.53) P < .05.
group: no substance abuse) c
P < .001.
Mean DDD 0.85 (0.44-1.67) 1.10 (0.70-1.72) 0.98 (0.67-1.45) d
P < .01.

Table 4. Multivariable Analysis of Suicidea

Suicide Death
Years 1-3 Years 4-12 12 Years
Variable HR (95% CI) P Value HR (95% CI) P Value HR (95% CI) P Value
Baseline variables
Premorbid occupational impairmentb NA NA 2.76 (1.20-6.33) .02 NA NA
No. of suicide attempts during DUP 1.47 (1.10-1.96) .009 NA NA 1.39 (1.08-1.74) .009
Variables during first 3 y
No. of suicide attempts 1.77 (1.22-2.55) .002 NA NA 1.58 (1.17-2.14) .003
No. of relapses NA NA 1.32 (1.01-1.72) .04 NA NA
Adherence NA NA 2.77 (1.32-5.80) .007 2.40 (1.44-4.01) .001
Abbreviations: DUP, duration of untreated psychosis; EI, early intervention; accounting for correlated group of observations with adjustment for
HR, hazard ratio; NA, not applicable; SC, standard care. treatment group.
a b
Backward stepwise Cox proportional hazards regression model was used Reference group: not impaired.

statistically significant (likelihood ratio test χ 2 = 14.60, through 12. Jumping from height was found to be the main sui-
P = .006) (Table 4). cide method, similar to that of the general population in Hong
For suicide during the entire 12 years, patients with a rela- Kong.36 The number of suicide attempts during the DUP and
tively high number of suicide attempts during DUP and the first the initial 3 years were indicators of early suicide (years 1-3).
3 years and with poor adherence during the first 3 years were Premorbid occupational impairment, number of relapses, and
significantly more likely to complete suicide during 12 years poor adherence during the initial 3 years of treatment were in-
(Table 3). The model was statistically significant (likelihood dicators of late suicide (years 4-12). Overall, patients who had
ratio test χ2 = 25.56, P < .001) (Table 4). more suicide attempts during DUP and the initial 3 years and
had poor adherence were more likely to complete suicide
during the entire 12 years.
This is the first study, to our knowledge, to report the as-
Discussion sociation of an EI service with the long-term suicide rate among
The results of the current study suggest that patients who un- patients with schizophrenia-spectrum disorders in a repre-
derwent a 2-year EI service had significantly fewer suicides and sentative sample. Many studies15,37,38 have reported that the
longer survival rates during 12 years than those who did not. risk of suicide was the highest during the early stage of psy-
The main association was observed during the initial 3 years. chosis. The suicide rate among patients with FEP was re-
The suicide rates of the 2 groups were similar during years 4 ported to be 2.7 times higher than that among patients with

jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online April 4, 2018 E5

© 2018 American Medical Association. All rights reserved.

Downloaded From: by Clinica Alemana de Santiago S.A., Juan Undurraga on 04/04/2018


Research Original Investigation Early Intervention Service for Psychosis and Suicide Among Patients With First-Episode Schizophrenia-Spectrum Disorders

chronic conditions.39 The 2-year SMR of suicide among the pa- find a positive effect on the reduction of relapses. This find-
tients with FEP in a recent study40 was reported to be as high ing highlights the importance of strengthening the EI service
as 81.91. The results of our study provide evidence that pa- with a specific focus on relapse reduction.
tients with schizophrenia-spectrum disorders who received EI
service had a significantly lower suicide rate than those who Strengths and Limitations
did not during the most vulnerable period, with the effect The main strength of this study is a relatively large represen-
maintained for 12 years. Suicide is a serious global public health tative sample studied for a long period, which allowed the study
concern. The overall suicide rate can be reduced by almost 50% of both the short- and long-term association of the EI service
if the suicide rate among individuals with mental illness were with suicide. Because the EI service was a regionwide ser-
to be reduced to the level of that in the general population.41 vice, no concurrent comparison group was possible. How-
Because schizophrenia was found to have the second highest ever, measures to ensure the compatibility of the 2 groups were
population-attributable risk for suicide among patients with implemented in participant matching and statistical analy-
major psychiatric conditions, the implementation of an EI ser- sis. The reliability of the clinical information was limited by
vice can potentially make a significant contribution to the the quality of the clinical records and was likely to be differ-
global effort of suicide reduction. Although the SMR of the sui- ent between the SC and EI groups, which may limit the inter-
cide rate in the EI group was almost half of that in the SC group pretation of the findings. The retrospective nature of the study
in the current study, it was still higher than the median re- precluded the determination of causality of the EI on suicide
ported in another study.1 This finding reflects the need for reduction. Patients who had emigrated and died abroad could
further improvement of the local EI service. not be identified in the current study. Although 1070 patients
In the current study, we found that a different set of risk (86.7%) had clinical activities in the recent year or future clini-
factors contribute to suicides at different stages of illness. This cal appointments at the time of obtaining mortality informa-
finding highlights the importance of the different stages of tion, suggesting that patients were likely to be residing lo-
schizophrenia as suggested in a previous study.42 A history of cally, the number of suicides in the current study could be
suicide attempts during the pretreatment and the early treat- underestimated. The generalizability of the results of the study
ment periods was found to be a significant risk factor for early may be limited by the age range of the patients in the EI ser-
suicide (years 1-3) and suicide during 12 years but not late sui- vice and the restriction of the diagnosis and duration of treat-
cide (years 4-12). A history of suicide attempts has been con- ment of patients before entering the study.
sistently reported as an indicator of suicide in patients with
FEP37 and chronic illness.8-11 The significance of suicide at-
tempts during the DUP period highlights the importance of
shortening the DUP.
Conclusions
The number of relapses and adherence during the first 3 This study provides evidence to support the potential asso-
years of treatment and premorbid occupational functioning ciation of of an EI service with reduction in suicide during a
were found to be significant indicators of suicide in the late 12-year period, with the maximum association during the first
but not the early stage. This finding agrees with the existing 3 years. Relapse and adherence during the early stage of the
literature.37,43,44 The results further emphasize that these fac- illness indicated late suicide. This finding highlights the im-
tors during the early stage of the illness could specifically affect portance of the further improvement of the EI model to effec-
suicide at a later stage. Although the EI services have been tively reduce the relapse rate. Different sets of risk factors were
implemented worldwide and improve the short-term out- associated with suicide at different stages of illness, suggest-
comes of patients with psychosis, most studies16,18,19,45 did not ing the importance of phase-specific intervention.

ARTICLE INFORMATION Study supervision: S. K. W. Chan, Pang, Chen. approval of the manuscript; and decision to submit
Published Online: April 4, 2018. Conflict of Interest Disclosures: Dr Chen reported the manuscript for publication.
doi:10.1001/jamapsychiatry.2018.0185 being the convener of the Early Assessment Service
for Young People with Psychosis (EASY) service at REFERENCES
Author Contributions: Dr S.K.W. Chan had full
access to all the data in the study and takes the Hospital Authority EASY Working Group on a 1. Saha S, Chant D, McGrath J. A systematic review
responsibility for the integrity of the data and the voluntary basis. Drs S. K. W. Chan, Chang, and Lee of mortality in schizophrenia: is the differential
accuracy of the data analysis. reported being members of the working group in a mortality gap worsening over time? Arch Gen
Study concept and design: S. K. W. Chan, Chen. voluntary position. No other disclosures were Psychiatry. 2007;64(10):1123-1131.
Acquisition, analysis, or interpretation of data: All reported. 2. Walker ER, McGee RE, Druss BG. Mortality in
authors. Funding/Support: This study was partly funded by mental disorders and global disease burden
Drafting of the manuscript: S. K. W. Chan, grant 11121721 from the Health and Medical implications. JAMA Psychiatry. 2015;72(4):334-341.
S. W. Y. Chan, Pang, Yan, Hui. Research Fund Hong Kong Health (Dr S. K. W. Chan) 3. Palmer BA, Pankratz VS, Bostwick JM. The
Critical revision of the manuscript for important and grant 03040141 from the Health Services lifetime risk of suicide in schizophrenia:
intellectual content: S. K. W. Chan, Pang, Yan, Hui, Research Fund of Hong Kong (Dr Chen). a reexamination. Arch Gen Psychiatry. 2005;62(3):
Chang, Lee, Chen. Role of the Funder/Sponsor: The funding sources 247-253.
Statistical analysis: S. K. W. Chan, S. W. Y. Chan, had no role in the design and conduct of the study;
Pang, Yan. 4. Inskip HM, Harris EC, Barraclough B. Lifetime
collection, management, analysis, and risk of suicide for affective disorder, alcoholism and
Obtained funding: S. K. W. Chan, Chen. interpretation of the data; preparation, review, or
Administrative, technical, or material support: schizophrenia. Br J Psychiatry. 1998;172:35-37.
S. K. W. Chan, S. W. Y. Chan, Chen.

E6 JAMA Psychiatry Published online April 4, 2018 (Reprinted) jamapsychiatry.com

© 2018 American Medical Association. All rights reserved.

Downloaded From: by Clinica Alemana de Santiago S.A., Juan Undurraga on 04/04/2018


Early Intervention Service for Psychosis and Suicide Among Patients With First-Episode Schizophrenia-Spectrum Disorders Original Investigation Research

5. Heilä H, Haukka J, Suvisaari J, Lönnqvist J. trial of the effectiveness of specialised care for early and functional remission rates: Worldwide
Mortality among patients with schizophrenia and psychosis. BMJ. 2004;329(7474):1067. Schizophrenia Outpatient Health Outcomes
reduced psychiatric hospital care. Psychol Med. 20. Grawe RW, Falloon IRH, Widen JH, Skogvoll E. (W-SOHO) study. Br J Psychiatry. 2011;199(3):194-
2005;35(5):725-732. Two years of continued early treatment for 201.
6. Bertelsen M, Jeppesen P, Petersen L, et al. recent-onset schizophrenia. Acta Psychiatr Scand. 33. Lee EW, Wei LJ, Amato DA. Cox-type regression
Suicidal behaviour and mortality in first-episode 2006;114(5):328-336. analysis for large number of small groups correlated
psychosis: the OPUS trial. Br J Psychiatry Suppl. 21. Cullberg J, Levander S, Holmqvist R, Mattsson failure time observations. In: Klein JP, Goel P, eds.
2007;51:s140-s146. M, Wieselgren IM. One-year outcome in first Survival Analysis: State of the Art. Amsterdam, the
7. Dutta R, Murray RM, Allardyce J, Jones PB, episode psychosis patients in the Swedish Netherland: Springer; 1992.
Boydell JE. Mortality in first-contact psychosis Parachute project. Acta Psychiatr Scand. 2002;106 34. Yao XI, Wang X, Speicher PJ, et al. Reporting
patients in the U.K. Psychol Med. 2012;42(8):1649- (4):276-285. and guidelines in propensity score analysis:
1661. 22. Chen EYH, Tang JYM, Hui CLM, et al. a systematic review of cancer and cancer surgical
8. Yuen K, Harrigan SM, Mackinnon AJ, et al. Three-year outcome of phase-specific early studies. J Natl Cancer Inst. 2017;109(8).
Long-term follow-up of all-cause and unnatural intervention for first-episode psychosis. Early Interv 35. Bath P, Hogg C, Tracy M, Pocock S; Optimising
death in young people with first-episode psychosis. Psychiatry. 2011;5(4):315-323. the Analysis of Stroke Trials Collaboration with the
Schizophr Res. 2014;159(1):70-75. 23. Malla A, Norman R, McLean T, Scholten D, Writing Committee. Calculation of
9. Haw C, Hawton K, Sutton L, Sinclair J, Deeks J. Townsend L. A Canadian programme for early numbers-needed-to-treat in parallel group trials
Schizophrenia and deliberate self-harm. Suicide Life intervention in non-affective psychotic disorders. assessing ordinal outcomes. Int J Stroke. 2011;6(6):
Threat Behav. 2005;35(1):50-62. Aust N Z J Psychiatry. 2003;37(4):407-413. 472-479.

10. Hawton K, Sutton L, Haw C, Sinclair J, Deeks JJ. 24. Park AL, McCrone P, Knapp M. Early 36. Yip PS, Caine E, Yousuf S, Chang SS, Wu KC,
Schizophrenia and suicide: systematic review of risk intervention for first-episode psychosis: broadening Chen YY. Means restriction for suicide prevention.
factors. Br J Psychiatry. 2005;187:9-20. the scope of economic estimates. Early Interv Lancet. 2012;379(9834):2393-2399.

11. Björkenstam C, Björkenstam E, Hjern A, Bodén Psychiatry. 2016;10(2):144-151. 37. Pompili M, Serafini G, Innamorati M, et al.
R, Reutfors J. Suicide in first episode psychosis. 25. Melle I, Johannesen JO, Friis S, et al. Early Suicide risk in first episode psychosis: a selective
Schizophr Res. 2014;157(1-3):1-7. detection of the first episode of schizophrenia and review of the current literature. Schizophr Res. 2011;
suicidal behavior. Am J Psychiatry. 2006;163(5): 129(1):1-11.
12. Robinson J, Harris MG, Harrigan SM, et al.
Suicide attempt in first-episode psychosis. 800-804. 38. Nordentoft M, Laursen TM, Agerbo E, Qin P,
Schizophr Res. 2010;116(1):1-8. 26. Tang JYM, Wong GHY, Hui CLM, et al. Early Høyer EH, Mortensen PB. Change in suicide rates
intervention for psychosis in Hong Kong—the EASY for patients with schizophrenia in Denmark,
13. Chang WC, Chen ESM, Hui CLM, Chan SKW, Lee 1981-97. BMJ. 2004;329(7460):261.
EHM, Chen EYH. Prevalence and risk factors for programme. Early Interv Psychiatry. 2010;4(3):214-
suicidal behavior in young people presenting with 219. 39. Brown S. Excess mortality of schizophrenia:
first-episode psychosis in Hong Kong. Soc 27. Chan SKW, So HC, Hui CLM, et al. 10-Year a meta-analysis. Br J Psychiatry. 1997;171:502-508.
Psychiatry Psychiatr Epidemiol. 2015;50(2):219-226. outcome study of an early intervention program for 40. Melle I, Olav Johannesen J, Haahr UH, et al.
14. Dutta R, Murray RM, Allardyce J, Jones PB, psychosis compared with standard care service. Causes and predictors of premature death in
Boydell J. Early risk factors for suicide in an Psychol Med. 2015;45(6):1181-1193. first-episode schizophrenia spectrum disorders.
epidemiological first episode psychosis cohort. 28. Chan TC, Chang WC, Hui CL, Chan SK, Lee EH, World Psychiatry. 2017;16(2):217-218.
Schizophr Res. 2011;126(1-3):11-19. Chen EY. Rate and predictors of disengagement 41. Mortensen PB, Agerbo E, Erikson T, Qin P,
15. Nordentoft M, Madsen T, Fedyszyn I. Suicidal from a 2-year early intervention program for Westergaard-Nielsen N. Psychiatric illness and risk
behavior and mortality in first-episode psychosis. psychosis in Hong Kong. Schizophr Res. 2014;153 factors for suicide in Denmark. Lancet. 2000;355
J Nerv Ment Dis. 2015;203(5):387-392. (1-3):204-208. (9197):9-12.

16. Kane JM, Robinson DG, Schooler NR, et al. 29. Reininghaus U, Dutta R, Dazzan P, et al. 42. Scott J, Leboyer M, Hickie I, et al. Clinical
Comprehensive versus usual community care for Mortality in schizophrenia and other psychoses: staging in psychiatry: a cross-cutting model of
first episode psychosis: 2-year outcomes from the a 10-year follow-up of the ÆSOP first-episode diagnosis with heuristic and practical value. Br J
NIMH RAISE early treatment program. Am J cohort. Schizophr Bull. 2015;41(3):664-673. Psychiatry. 2013;202(4):243-245.
Psychiatry. 2016;173(4):362-372. 30. Haro JM, Kamath SA, Ochoa S, et al; SOHO 43. De Hert M, McKenzie K, Peuskens J. Risk
17. Kuipers E, Holloway F, Rabe-Hesketh S, Study Group. The Clinical Global Impression- factors for suicide in young people suffering from
Tennakoon L; Croydon Outreach and Assertive Schizophrenia scale: a simple instrument to schizophrenia. Schizophr Res. 2001;47(2-3):127-134.
Support Team (COAST). An RCT of early measure the diversity of symptoms present in 44. Berger G, Fraser R, Carbone S, McGorry P.
intervention in psychosis: Croydon Outreach and schizophrenia. Acta Psychiatr Scand Suppl. 2003; Emerging psychosis in young people, part 1: key
Assertive Support Team (COAST). Soc Psychiatry 107(416):16-23. issues for detection and assessment. Aust Fam
Psychiatr Epidemiol. 2004;39(5):358-363. 31. Nosè M, Tansella M, Thornicroft G, et al. Is the Physician. 2006;35(5):315-321.
18. Petersen L, Jeppesen P, Thorup A, et al. Defined Daily Dose system a reliable tool for 45. Nordentoft M, Rasmussen JO, Melau M,
A randomised multicentre trial of integrated versus standardizing antipsychotic dosages? Int Clin Hjorthøj CR, Thorup AAE. How successful are first
standard treatment for patients with a first episode Psychopharmacol. 2008;23(5):287-290. episode programs? a review of the evidence for
of psychotic illness. BMJ. 2005;331(7517):602. 32. Haro JM, Novick D, Bertsch J, Karagianis J, specialized assertive early intervention. Curr Opin
19. Craig TKJ, Garety P, Power P, et al. The Lambeth Dossenbach M, Jones PB. Cross-national clinical Psychiatry. 2014;27(3):167-172.
Early Onset (LEO) Team: randomised controlled

jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online April 4, 2018 E7

© 2018 American Medical Association. All rights reserved.

Downloaded From: by Clinica Alemana de Santiago S.A., Juan Undurraga on 04/04/2018

Você também pode gostar