A Review of Evidence-Based Follow-Up Care forSuicide Prevention

Where Do We Go From Here?Gregory K. Brown, PhD, Kelly L. Green, PhD

Context: Follow-up services are an important component of a comprehensive, national strategy forsuicide prevention. Increasing our knowledge of effective follow-up care has been identified as anAspirational Goal by The National Action Alliance for Suicide Prevention’s Research PrioritizationTask Force.

Evidence acquisition: Several recent comprehensive reviews informed the selection of studiesincluded in this brief review. Studies of follow-up services that reported significant effects for theoutcomes of death by suicide, suicide attempts, or suicidal ideation were included.

Evidence synthesis: Although there is a paucity of research in this area, promising paradigms thathave demonstrated effectiveness in preventing suicide and suicide attempts and reducing suicidal ideationwill be discussed. The major limitations of the literature in this area include numerous methodologicalflaws in the design and analyses of such studies and the lack of replication of studies with positive findings.

Conclusions: This paper identifies several breakthroughs that would be helpful for advancing thisarea of research and describes a comprehensive research pathway for achieving both short- andlong-term research objectives.(Am J Prev Med 2014;47(3S2):S209–S215) & 2014 American Journal of Preventive Medicine

Introduction

The development and implementation of effectivefollow-up care for individuals at risk for suicide isimportant for reducing rates of suicide andrelated behaviors. In response to the ongoing need foreffective treatments aimed at preventing suicide, theNational Action Alliance for Suicide Prevention’s(Action Alliance) Research Prioritization Task Force(RPTF) developed a comprehensive set of goals.1

Specifically, Aspirational Goal 6 aims to “ensure thatpeople who have attempted suicide can get effectiveinterventions to prevent further attempts.” Follow-up careis defined as services interventions that aim to bothincrease access to and engagement in care, as well as toprevent suicide and related behaviors, as opposed to moreacute care interventions, such as psychotherapy.The aims of this article are to (1) briefly review the state

of the science for follow-up care; (2) summarize limitations

of the current research and needed breakthroughs; and (3)describe both short- and long-term research objectives aswell as a step-by-step research pathway to advance the fieldof providing follow-up care for suicide prevention.

Evidence AcquisitionAs a comprehensive review was beyond the scope of this paper,several recent comprehensive systematic reviews2–5 were used toidentify studies to include in this brief review. Those studies withsignificant findings for the outcomes of death by suicide, suicideattempts, or suicidal ideation were selected for inclusion. There areadditional studies2–5 that have examined the effectiveness offollow-up approaches, primarily on the outcome of suicideattempts or self-injury behavior, that failed to report significanteffects and are not included in this brief review. Table 1 providesmore detailed descriptions of the intervention and comparisonconditions evaluated in each study, as well as the assessedoutcomes and results.

Evidence SynthesisThe primary finding noted from these reviews is thatonly two RCTs have examined the effect of follow-upcare on death by suicide. The first study6 followedpatients who had attempted suicide and refused or

From the Perelman School of Medicine of the University of Pennsylvania,Philadelphia, Pennsylvania

Address correspondence to: Gregory K. Brown, PhD, Department ofPsychiatry, University of Pennsylvania, 3535 Market Street, Room 2030,Philadelphia PA 19104-3309. E-mail: gregbrow@mail.med.upenn.edu.

0749-3797/$36.00http://dx.doi.org/10.1016/j.amepre.2014.06.006

& 2014 American Journal of Preventive Medicine � Published by Elsevier Inc. Am J Prev Med 2014;47(3S2):S209–S215 S209

Table1.Stud

iesreview

edwith

cond

ition

descrip

tions,a

ssessedou

tcom

es,a

ndresults

Stud

yRCT

nFo

llow-upservicede

scrip

tion

Com

paris

oncond

ition

descrip

tion

Prim

ary

outcom

e(s)

Results

Motto

and

Bostrom

,20016

Yes

389Interven

tion

454Con

trol

Subjects

weresent

lettersexpressing

care

andconcernby

research

staff(24letters

over

5years);letters

wereno

n-de

man

ding

(i.e.,sub

jectswereinvitedto

respon

difthey

wishe

d,bu

tthiswas

notrequ

ired)

Subjects

received

noad

ditio

nalcon

tact

from

stud

ystaff

Dea

thby

suicide

Kap

lan–

Meier

survival

prob

abilitie

sbe

twee

ngrou

psweresign

ificantly

differen

tforthefirst2yearsof

follow-up

(p¼0

.043);M

(SE):

Year

1:

Trea

tmen

t,0.990(0.005)

Con

trol,0

.978(0.007)

Year

2:

Trea

tmen

t,0.983(0.006)

Con

trol,0

.964(0.009)

Fleischm

ann

etal.,20087

Yes

922Interven

tion

945Con

trol

Subjects

received

abrief1-hou

rpsycho

educationa

linterventioncloseto

discha

rgean

d9follow-upcontacts

(eith

erby

phon

eor

in-person)

over

18mon

ths

Subjects

received

TAU

Dea

thby

suicide

At18-m

onth

follow-up,

sign

ificantlymore

subjects

died

bysuicidein

theTA

Ucond

ition

than

theinterven

tioncond

ition

:χ²¼1

3.83,p

o0.001

Welu,

19778

Yes

62Interven

tion

57Con

trol

Subjectsreceived

trea

tmen

twith

inthecontext

ofaspecialo

utreachprogram,inwhich

they

werecontactedby

amen

talhea

lthclinicianas

soon

aspo

ssible

followingdischa

rge,

and

follow-upcontacts

includ

edaho

mevisitan

dwee

klyor

biwee

klycontactover

a4-mon

thfollow-uppe

riod

Subjects

received

TAU

Suicide

attempt

At4-m

onth

follow-up,

Fisher’sexacttest

indicatedthat

fewer

subjects

inthe

interven

tioncond

ition

repo

rted

asuicide

attempt

compa

redto

theTA

Ucond

ition

:p¼

0.04573

Carteret

al.,

20059an

d200710

Yes

378Interven

tion

394Con

trol

Subjects

received

8po

stcardsexpressing

care

andconcernover

a12-m

onth

perio

dSu

bjects

received

nopo

stcards

Intentiona

lself-po

ison

ing

At12-m

onth

follow-up,

therewereno

sign

ificant

differen

cesintheprop

ortio

nof

subjectsinea

chgrou

pwho

repe

ated

self-

poison

ing;

however,the

numbe

rof

repe

titions

was

sign

ificantlylower

inthe

interven

tiongrou

pcompa

redto

the

controlg

roup

:IRR¼0

.55,p

¼0.01,9

5%

CI¼

0.35,0

.87

At24-m

onth

follow-up,

therewas

nosign

ificant

differen

cein

theprop

ortio

nof

subjects

who

repe

ated

self-po

ison

ing;

however,the

numbe

rof

repe

titions

was

sign

ificantlylower

intheinterven

tion

grou

pcompa

redto

thecontrolg

roup

for

wom

enon

ly:IRR¼0

.49,p

¼0.004,9

5%

CI¼

0.30,0

.80

Hassanian

-Mogha

ddam

etal.,201111

Yes

1,150Interven

tion

1,150Con

trol

Subjects

received

9po

stcardsexpressing

care

andconcernover

a12-m

onth

perio

d.Su

bjects

received

TAU

Suicide

attempt,

suicidal

At12-m

onth

follow-up,

theinterven

tion

grou

pde

mon

stratedless

suicidal

idea

tion

(con

tinue

don

next

page

)

Brown and Green / Am J Prev Med 2014;47(3S2):S209–S215S210

www.ajpmonline.org

Table1.S

tudies

review

edwith

cond

ition

descrip

tions,a

ssessedou

tcom

es,a

ndresults

(con

tinue

d)

Stud

yRCT

nFo

llow-upservicede

scrip

tion

Com

paris

oncond

ition

descrip

tion

Prim

ary

outcom

e(s)

Results

idea

tion

(RRR¼0

.31,9

5%

CI¼

0.22,0

.38),an

dlower

rate

ofsuicideattempt

(RRR¼0

.42,

95%

CI¼

0.11,0

.63)c

ompa

redwith

the

controlcon

ditio

nAd

ditiona

lly,the

numberofsuicide

attempts

was

also

redu

cedin

theinterven

tion

cond

ition

comparedto

thecontrolcon

dition

(IRR¼0

.64,

95%CI¼0

.42,

0.97

)Va

ivaet

al.,

200612

Yes

147

Phon

ecalls

after

1mon

th145

Phon

ecalls

after

3mon

ths

312Con

trol

Subjects

received

follow-upph

onecalls

either

1or

3mon

thsafterthesuicide

attempt

from

apsychiatris

tdu

ringwhich

the

psychiatris

treview

edthetrea

tmen

trecommen

dedby

theED

andsuggesteda

newtrea

tmen

tplan

iftheoriginal

was

too

difficultforthepa

tient

tofollow;a

nurge

ntap

pointm

entw

asalso

sche

duledat

theED

ifthepa

tient

was

considered

athigh

riskfor

suicide

Subjects

received

TAU

Suicide

attempt

At13-m

onth

follow-up,

thenu

mbe

rof

subjects

who

attempted

suicidewas

sign

ificantlylower

over

the6mon

ths

post-con

tact

forthosethat

received

contactafter1mon

thcompa

redto

the

TAUgrou

p:χ²¼4

.7,p

¼0.03,

differen

ce¼1

0%,9

5%

CI¼

2%,1

8%;

therewereno

sign

ificant

differen

ces

betwee

nthe3-m

onth

contactgrou

pan

dtheTA

Ugrou

p

Term

ansen

andByw

ater,

197513

No

57Interven

tionwith

samemen

tal

health

worke

r57Interven

tionwith

crisiscenter

voluntee

r50

Assessmen

tinED

38Iden

tificationfrom

EDad

mission

records

Subjects

inthefollow-upcare

cond

ition

sreceived

either:(1)a

ssessm

entin

theED

andfollow-upfor3mon

thsby

samemen

tal

health

worke

rwho

assessed

thepa

tient

intheED

;or(2)a

ssessm

entin

theED

and

follow-upfor3mon

thswith

acrisiscenter

voluntee

r;contactoccurred

with

tape

ring

freq

uencyover

thecourse

of12wee

ks

Subjects

inthecontrol

cond

ition

sreceived

either:(1)a

ssessm

entin

theED

andno

follow-up;

or(2)ide

ntificatio

nfrom

emerge

ncyad

mission

recordson

ly

Suicide

attempt

At3-m

onth

follow-up,

subjects

inthefirst

grou

pwho

received

follow-upby

thesame

men

talh

ealth

cliniciande

mon

strated

sign

ificantlyfewer

suicideattempts

compa

redto

theothe

rthreegrou

ps(no

test

statistic

repo

rted

,p¼0

.05)

Torhorst

etal.,198714

Yes

68Trea

tmen

tfrom

sametherap

ist

85Rou

tinereferral

tolocala

gency

73Trea

tmen

tat

suicide

preven

tioncenter

Subjects

received

trea

tmen

tfrom

thesame

therap

istthey

sawin

theho

spita

lSu

bjectsinthecomparison

cond

itionsreceived

either

(1)a

routinereferralto

alocalagency;or

(2)

treatm

entfro

madiffe

rent

therapistat

aspecialized

suicidepreven

tioncenter

Suicide

attempt

At12-m

onth

follow-up,

subjects

who

received

trea

tmen

tfrom

adifferen

ttherap

istat

asuicidepreven

tioncenter

hadalower

suicideattempt

rate

than

thosein

theexpe

rimen

talg

roup

(χ²¼

5.363,d

f¼2,p

o0.1)

Kinget

al.,

200115

No

600Con

trol

300Decea

sed

Thisstud

ywas

aretrospe

ctivechartreview

stud

y;subjects

wereeither

discha

rged

individu

alswho

subseq

uentlydied

bysuicide

ormatched

psychiatric

controls;the

presen

ceof

continuityof

care

andcontactwith

the

sameprofession

alwereexam

ined

NA

Dea

thby

suicide

Bothcontinuo

uscare

(OR¼0

.57,9

5%

CI¼

0.37,0

.87,p

¼0.01)a

ndcontactw

iththesameprofession

al(OR¼1

8.45,9

5%

CI¼

4.46,7

6.32,p

o0.001)p

redicted

decrea

sedris

kof

deathby

suicide

ED,e

merge

ncyde

partmen

t;IRR,inciden

ceris

kratio

;RRR,relativeris

kredu

ction;

TAU,treatmen

tas

usua

l

Brown and Green / Am J Prev Med 2014;47(3S2):S209–S215 S211

September 2014

discontinued outpatient treatment in the month afterdischarge from the hospital, and then randomized themto receive either a caring letters intervention or no follow-up. The study found that the rate of suicide for theintervention condition was significantly lower than thatfor the control group for the first 2 years of follow-up.The second study7 enrolled suicide attempters from

eight emergency departments (EDs) in five low- tomiddle-income countries and randomized them toreceive either treatment as usual (TAU) or a briefintervention with follow-up contact. Follow-up over an18-month period revealed that individuals in the inter-vention condition had a significantly lower rate of suicidethan those receiving TAU.More attention has been given to investigating the

effect of follow-up care on preventing or reducing suicideattempts and self-directed violence (i.e., some studiesreported one outcome that combined suicide attemptsand non-suicidal self-injury) than has been given to theoutcome of death by suicide. For example, one study8

found that fewer participants who were assigned toreceive an intensive follow-up contact interventionexperienced a repeat suicide attempt over a 4-monthfollow-up period relative to those assigned to TAU.Three studies have examined less time-intensive fol-

low-up services. An Australian study9,10 recruitedpatients from toxicology units following intentionalself-poisoning and randomly assigned them to receiveeither follow-up postcards or no intervention. This studyfound that participants assigned to receive the postcardshad fewer numbers of intentional self-poisoning behav-iors than controls over a 24-month follow-up period.A similar study11 recruited individuals who intention-

ally self-poisoned and randomized participants to receiveeither follow-up postcards or TAU. Results indicated thatthose in the intervention condition demonstrated fewerinstances of suicidal ideation and suicide attempts (bothin terms of rate and total numbers) than those in TAU.A third study12 involving patients discharged from the

ED following an intentional overdose randomized par-ticipants to receive a follow-up call 1-month post-discharge, a call at 3 months post-discharge, or TAU.Participants in the intervention condition that receivedthe 1-month call were less likely to make subsequentsuicide attempts than those in TAU over the first 6months of the 13-month follow-up period.Three other studies have found significant results for

follow-up interventions, depending on the specific indi-vidual who performed the follow-up contact. Onecompared13 follow-up by a mental health worker,follow-up by a crisis volunteer, and no follow-up forpatients discharged from a hospital after a suicideattempt. The study found a significant reduction in

repeat suicide attempts for follow-up by a mental healthworker compared to follow-up by a crisis center volun-teer or no follow-up.Torhorst and colleagues14 reported that the rate of

suicide attempts in the group of patients who saw a differenttherapist for treatment following discharge from thehospital was lower than that of patients who saw the sameclinician who treated them in the hospital. A retrospectivechart review study,15 on the other hand, found that bothcontinuity of care alone and contact with the sameprofessional predicted reduced suicide risk in dischargedpatients who had died by suicide and matched controls.In summary, there are several studies with promising

initial findings concerning the efficacy of follow-up careand suicide prevention. Specifically, research suggeststhat clinicians who reach out to patients (especially thosepatients not engaged in treatment) using caring letters toexpress concern and support may help to reduce the rateof suicide following discharge from a psychiatric hospital.Additionally, low-cost follow-up interventions (e.g.,

phone calls, postcards) may be effective and particularlyimportant for reducing death by suicide and repeatsuicide attempts, especially in areas with limited resour-ces. Outreach programs that provide comprehensivemental health treatment and emphasize follow-up andcontinuity of care following discharge from the hospitalmay also help to prevent repeat suicide attempts.

Gaps and Limitations of the Current State ofthe ScienceAlthough findings from these studies warrant optimismthat follow-up services can ultimately be an effectivestrategy for suicide prevention, there are several gaps inour current knowledge, as well as major limitations (i.e.,methodological flaws) of the work that has been donethus far.With regard to gaps in the literature, the first major

limitation is the paucity of RCTs, especially thoseinvestigating effects of follow-up services on death bysuicide.4 Specifically, only two studies6,7 have demon-strated efficacy for preventing suicide. Although severalstudies have demonstrated efficacious follow-up servicesfor preventing suicide attempts and self-directed vio-lence, these outcomes are only proxies for death bysuicide and may not generalize to services that willactually prevent suicide. Additionally, the studies thathave found positive results have not investigated themechanisms by which the follow-up services affectedoutcomes (e.g., greater engagement in care).Further, our knowledge of effective services for specific

subpopulations, particularly those at high risk relative tothe general population, is severely limited. For example,

Brown and Green / Am J Prev Med 2014;47(3S2):S209–S215S212

www.ajpmonline.org

there are no RCTs of follow-up services that havedemonstrated efficacy to prevent suicide or relatedbehaviors for adolescents, older adults, and other minor-ity groups.Additionally, existing studies have recruited patients

mostly from acute treatment settings (e.g., hospitals,EDs). Research16 has found that most individuals whoattempt suicide seek no treatment following theirattempt. Thus, it is unclear whether findings from studiesof follow-up services conducted to date can be general-ized to other settings, such as primary care, outpatientmental health, or other community settings.Finally, the failure to replicate studies that have found

significant effects is a major gap in the literature.Although developing novel interventions is important,there has been less emphasis placed on replicating studieswith positive results or improving existing interventionsthat have been found to be effective.With regard to methodological problems, there are

many major flaws in the RCTs that have been conductedthus far that have been described in previous reviews.2,4

Many of these methodological problems also apply toacute intervention research and were discussed in moredetail in Brown and Jager-Hyman’s psychotherapyreview17 in this issue.Those problems discussed previously that also apply to

follow-up services research include (1) failure to provideoperational definitions or use a standardized nomencla-ture for assessing suicide, suicide attempts, suicidalideation, and other related behaviors; (2) failure toinclude reliable and validated outcome measures; and(3) failure to control for sources of bias. Methodologicalproblems such as those outlined here led to the followingconclusion in the Veterans Affairs systematic review:“Overall, these intervention trials had methodologicallimitations that resulted in their providing only lowstrength and insufficient evidence to properly drawconclusions on the effectiveness of the various treatmentinterventions and follow-up strategies.”4

DiscussionFuture research should seek to achieve breakthroughs,which are needed to address these limitations andincrease our knowledge about effective follow-up servicesfor suicide prevention. These needs include (1) improv-ing methodological rigor in future studies; (2) developingadditional follow-up services and paradigms that arecost-effective and innovative; (3) expanding research toadditional settings and subpopulations; and (4) replicat-ing and disseminating evidence-based follow-up services.Improving the methodological rigor in designing

future RCTs and other studies is of paramount

importance. There are several short-term research goalsthat can achieve this aim. First, it is important thatstudies use standardized assessments that have beenfound to be valid and reliable, and it is important thatsuch measures correspond to standardized nomenclatureof suicide ideation and behavior such as the CDC’s Self-Directed Violence Classification System (SDVCS).18

Second, future research should be devoted to develop-ing novel assessment methods, such as ecologicalmomentary assessment, to more accurately track suicidalideation and behavior over time. Third, future researchshould include methods to address ambivalent suicidalbehavior (e.g., suicide adjudication boards).Fourth, future studies should include methods for

controlling sources of bias, such as performing intent-to-treat analyses, identifying and measuring non-study co-interventions, and blinding research staff and/or researchparticipants and assessing any breaks in blinding. Finally,future studies should develop innovative methods forretaining participants in studies and monitoring long-term outcomes.Developing and testing novel follow-up services for

suicide prevention is also especially warranted. In orderto improve the feasibility of conducting adequatelypowered studies to detect the treatment effects on deathby suicide, it would be beneficial to develop interventionsof minimal economic cost as a short-term research goal.Studies of these approaches should determine whetherfollow-up care actually facilitates treatment engagementand reduces rates of suicide, suicide attempts, or suicidalideation. Cost-effectiveness studies should also be con-ducted alongside efficacy and effectiveness trials of testedinterventions.Additionally, the development of follow-up services that

use innovative electronic health technologies (e.g., chatrooms, texting, smartphone apps, and other web-basedapplications) as stand-alone or adjunctive services is alsoneeded and achievable over the short term. These tech-nologies have the potential to reach a larger segment of thepopulation at a low cost. Thus far, one small pilot test19 oftext messaging over 4 weeks following discharge foundthis intervention to be feasible and acceptable to patientswho attempted suicide. To date, however, no study hasbeen conducted to evaluate the impact of electronicservices on suicide, suicide attempts, or suicidal ideation.Ultimately, identifying and developing evidence-based

follow-up services that can be delivered following dis-charge from acute care settings for the prevention ofsuicide is especially needed. This long-term goal can beattained by conducting large-scale, adequately poweredRCTs. These studies should determine whether theeffects of an intervention are partially mediated byengagement in mental health care or whether there is a

Brown and Green / Am J Prev Med 2014;47(3S2):S209–S215 S213

September 2014

direct effect on outcomes. Such studies should alsoexplore whether there are other moderating or mediatingeffects of the intervention by identifying and testingpotential mechanisms of action in effective interventionsand developing valid and reliable measures of suchmechanisms.Once effective follow-up services are identified,

expanding the research into new settings and populationsis also needed in order to investigate the generalizabilityof these interventions. Thus, over the long term,researchers should continue to develop novel methodsto recruit and screen at-risk individuals both in acute caresettings such as EDs as well as in the community at large.Schools, community centers, primary care settings, andworkplaces are also potential areas to target in order toobtain more representative samples and reach individu-als at risk for suicide who may not present to mentalhealth facilities. Future research should also examine therelative efficacy of evidence-based follow-up services forspecific subpopulations that are at an increased risk forsuicide, such as adolescents, older adults, and otherminority populations, as warranted by empirical data.Finally, studies with positive findings of follow-up

services should be replicated by independent researchgroups to ensure that robust effects are generalizableacross locations and populations. An especially impor-tant long-term objective is for researchers to develop andtest models to efficiently disseminate evidence-basedfollow-up services so that they can be widely availableand become the standard of care for facilitating engage-ment in treatment and ultimately preventing suicide.Figure 1 illustrates a proposed step-by-step research

pathway that can serve as a model for future studies thattest the effectiveness of follow-up services to reduce

suicide risk. Briefly, following this pathway, researchparticipants should be screened using standardizedmeasures. Following screening, enrolled participantsshould be randomized to either TAU alone or in additionto the study intervention. Potential mechanisms of actionshould then be assessed over the course of care todetermine what aspects of an intervention lead toreductions in suicide-related outcomes. Increasedengagement in care as a result of the study interventionshould also be evaluated as a potential mediator of therelationship between the intervention and outcome.

ConclusionsAlthough promising initial findings on follow-up careand services for suicide prevention exist in the literature,there are significant research gaps. Thus, additionalresearch is warranted to both improve the quality ofthe research in this area and expand current knowledge.A major research goal involves the rigorous study ofnovel, cost-effective approaches to follow-up care acrossa variety of populations and settings. Ultimately, suchstudies may result in the improvement of the standardof care for individuals who are at risk for suicide bydisseminating evidence-based strategies to preventsuicide.

Publication of this article was supported by the Centers forDisease Control and Prevention, the National Institutes ofHealth Office of Behavioral and Social Sciences, and theNational Institutes of Health Office of Disease Prevention.This support was provided as part of the National Institute ofMental Health-staffed Research Prioritization Task Force ofthe National Action Alliance for Suicide Prevention.

Figure 1. Proposed step-by-step research pathway for future RCTs

Brown and Green / Am J Prev Med 2014;47(3S2):S209–S215S214

www.ajpmonline.org

No financial disclosures were reported by the authors ofthis paper.

References1. USDHHS Office of the Surgeon General and National Action Alliance

for Suicide Prevention. 2012 National Strategy for Suicide Prevention:Goals and Objectives for Action. Washington, DC: HHS, September2012. http://www.surgeongeneral.gov/library/reports/national-strategy-suicide-prevention/full-report.pdf.

2. Self-harm: The NICE guideline on longer-term management. London:The British Psychological Society and The Royal College of Psychia-trists, 2012.

3. Mann JJ, Apter A, Bertolote J, et al. Suicide prevention strategies: asystematic review. JAMA 2005;294(16):2064–74.

4. O’Neil M, Peterson K, Low A, et al. Suicide prevention interventionsand referral/follow-up services: a systematic review. VA-ESP Project#05-225 2012.

5. O’Connor E, Gaynes BN, Burda BU, Soh C,Whitlock EP. Screening forand treatment of suicide risk relevant to primary care: a systematicreview for the U.S. Preventive Services Task Force. Ann Intern Med2013;158(10):741–54.

6. Motto JA, Bostrom AG. A randomized controlled trial of postcrisissuicide prevention. Psychiatr Serv 2001;52(6):828–33.

7. Fleischmann A, Bertolote JM, Wasserman D, et al. Effectiveness ofbrief intervention and contact for suicide attempters: a randomizedcontrolled trial in five countries. Bull World Health Organ 2008;86(9):703–9.

8. Welu TC. A follow-up program for suicide attempters: evaluation ofeffectiveness. Suicide Life Threat Behav 1977;7(1):17–29.

9. Carter GL, Clover K, Whyte IM, Dawson AH, D’Este C. Postcardsfrom the EDge project: randomised controlled trial of an intervention

using postcards to reduce repetition of hospital treated deliberate selfpoisoning. Br Med J 2005;331(7520):805–7.

10. Carter GL, Clover K,Whyte IM, Dawson AH, D’Este C. Postcards fromthe EDge: 24-month outcomes of a randomised controlled trial forhospital-treated self-poisoning. Br J Psychiatry 2007;191(6):548–53.

11. Hassanian-MoghaddamH, Sarjami S, Kolahi AA, Carter GL. Postcardsin Persia: randomised controlled trial to reduce suicidal behaviours 12months after hospital-treated self-poisoning. Br J Psychiatry 2011;198(4):309–16.

12. Vaiva G, Ducrocq F, Meyer P, et al. Effect of telephone contact on furthersuicide attempts in patients discharged from an emergency department:randomised controlled study. Br Med J 2006;332(7552):1241–5.

13. Termansen PE, Bywater C. S.A.F.E.R.: a follow-up service for attemptedsuicide in Vancouver. Can Psychiatr Assoc J 1975;20(1):29–34.

14. Torhorst A, Möller HJ, Bürk F, Kurz A. The psychiatric managementof parasuicide patients: a controlled clinical study comparing differentstrategies of outpatient treatment. Crisis 1987;8(1):53–61.

15. King EA, Baldwin DS, Sinclair JM, Baker NG, Campbell MJ,Thompson C. The Wessex Recent In-Patient Suicide Study, 1. Case-control study of 234 recently discharged psychiatric patient suicides. BrJ Psychiatry 2001;178:531–6.

16. Bruffaerts R, Demyttenaere K, Hwang I, et al. Treatment of suicidalpeople around the world. Br J Psychiatry 2011;199(1):64–70.

17. Brown GK, Jager-Hyman S. Evidence-based psychotherapies forsuicide prevention: future directions. Am J Prev Med 2014;47(3S2):S186–S194.

18. Crosby AE, Ortega L, Melanson C. Self-directed violence surveillance:uniform definitions and recommended data elements, version 1.0.Atlanta GA: CDC, 2011 www.cdc.gov/violenceprevention/pdf/Self-Directed-Violence-a.pdf.

19. Chen H, Mishara BL, Liu XX. A pilot study of mobile telephonemessage interventions with suicide attempters in China. Crisis 2010;31(2):109–12.

Brown and Green / Am J Prev Med 2014;47(3S2):S209–S215 S215

September 2014