guidance on action to be taken at suicide hotspots - sprc

36
Guidance on action to be taken at suicide hotspots

Upload: vodan

Post on 14-Feb-2017

218 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Guidance on action to be taken at suicide hotspots - SPRC

Guidance on action to be taken at suicide hotspots

Page 2: Guidance on action to be taken at suicide hotspots - SPRC

Acknowledgements

This guidance was developed by the Research and DevelopmentDepartment, Devon Partnership NHS Trust, in partnership with thePeninsula Medical School.

Project team:

Dr Peter AitkenDr Christabel OwensSally Lloyd-Tomlins Vita FitzSimonsTobit EmmensHelen Mattacott Maria SheppardIan Pearson (Devon and Torbay Local Implementation Team)

Others who have assisted in the development of the guidance include:

David Hess, Peninsula Medical SchoolDr Elizabeth King, University of Southampton Joe Ferns, SamaritansDr Elizabeth Earland, HM Coroner for Exeter & Greater DevonIan Arrow, HM Coroner for TorbayNigel Meadows, HM Coroner for PlymouthMembers of the Devon Inter-agency Forum on Self-Harm and Suicide Mike Francis, South Devon Healthcare NHS TrustDevon & Cornwall Constabulary Malcolm Dobson, British Transport PoliceColin McNicol, Rail Fatalities Management GroupMaurice Wilsdon, Rail Safety and Standards BoardTim Wood, Cornwall County CouncilProfessor David Gunnell, University of BristolProfessor Annette Beautrais, Christchurch School of Medicine, New ZealandPaul Wong, University of Hong Kong, Centre for Suicide Research & Prevention

Special thanks to:

Carrie Morgan, CSIP/NIMHE South West Region

Page 3: Guidance on action to be taken at suicide hotspots - SPRC

1

Contents

Brief outline and aims of the guidance 2

Part 1: BackgroundIntroduction 31. What is a suicide hotspot? 32. How will action at hotspots help to reduce the suicide rate? 43. What types of location are likely to be hotspots? 54. How many suicides are needed to make a hotspot? 65. Why is inter-agency collaboration important? 7Interventions at hotspots: a review of the evidence 81. Physical barriers 82. Signs and telephone hotlines 93. Suicide patrols 104. Training for staff of non-health agencies working at or near hotspots 105. Restrictions on media reporting 116. Deciding between available options 12Summary of available measures: pros and cons 12

Part 2: Identifying and managing suicide hotspots.A practical guide to inter-agency collaboration

Organisation, planning and personnel 131. The role of an Inter-agency Forum on Self-Harm & Suicide 132. Planning the hotspots programme 142a Programme personnel 142b Key early tasks 14Identifying suicide hotspots 161. Data collection 161a Who is already collecting it? 161b Where to find the data 161c What data to collect 171d Additional sources of data 181e Ethics 192. Data analysis 192a Quantitative analysis 192b Mapping suicides using GIS 20Managing hotspots: reducing risk and opportunity for suicide 211. Agreeing local priorities: the role of a stakeholder conference 212. The role of project teams 212a Forming effective alliances 212b Assessing site-specific risks 222c Considering the options 222d Drawing up local arrangements 222e Financial planning 222f Implementation 22Audit and evaluation 231. Audit 232. Evaluation 233. Reporting back 23Examples of good practice in identifying and managing suicide hotspots 24

References 25Appendix 1: Resources associated with programme 27Appendix 2: List of potential stakeholders and partner agencies 27Appendix 3: List of variables to include 28Appendix 4: Example of local suicide data mapped using GIS software 29Appendix 5: Stakeholder conference: pre-conference questionnaire and

outline of consensus method for use in priority setting 30Appendix 6: Useful websites 32

Page 4: Guidance on action to be taken at suicide hotspots - SPRC

2

Brief outline andaims of the guidance This best practice guide has been developed to:

1. support the development of effective collaborationby local multi-agency suicide prevention groups;

2. assist such groups to identify particular placeswithin their local area that are 'hotspots' forsuicide and to take appropriate steps to improve safety and deter acts of suicide at those locations;

3. contribute to the implementation of the NationalSuicide Prevention Strategy for England and toachieve an overall reduction in suicides, in linewith the target set out in the White Paper SavingLives: Our Healthier Nation (Department of Health,1999).

There are two parts to the guidance. The first part deals with thedefinition of 'suicide hotspot', outlines the range of measures that canbe taken to improve safety at such locations and summarises theevidence of effectiveness.

Part two describes a process for identifying and managing suicidehotspots at local level based on interagency collaboration. The modelhas been developed by means of 'action learning', using the countyof Devon as a pilot site.

Page 5: Guidance on action to be taken at suicide hotspots - SPRC

INTRODUCTION

1. What is a suicidehotspot?

The term 'suicide hotspot' hastwo possible meanings. It isfrequently used to refer to both:a) a geographical area with a

relatively high rate of suicideamong its resident population(e.g. a town, borough,county or country), and

b) a specific, usually public,site which is frequently usedas a location for suicide andwhich provides either meansor opportunity for suicide(e.g. a particular bridge fromwhich individuals frequentlyjump to their deaths).

This guidance deals withhotspots in the second senseand the term will be used in thissense throughout.

Many well-known locationsthroughout the world havebecome associated with suicidalacts. They include both man-made structures and naturalsites, some of which have iconicstatus or significance. TheGolden Gate Bridge in SanFrancisco, the Sydney HarbourBridge, the Empire StateBuilding and Niagara Falls areamong the top suicide sitesworldwide. Such places seem toact as magnets, drawing suicidalindividuals to them.

In the UK, Beachy Head cliffs inSussex and the CliftonSuspension Bridge in Bristol arenotorious as suicide sites.However, there are also manyless well-known locations, andevery local area will have sitesand structures that lendthemselves to suicide attempts.

In many cases, the place itselfprovides the means of suicide.The cliffs at Beachy Head, forinstance, supply the means ofsuicide by jumping, in the sameway that a bottle of tabletssupplies means of suicide bypoisoning or overdose.

3

PART 1BACKGROUND

Page 6: Guidance on action to be taken at suicide hotspots - SPRC

4

2. How will action at hotspots helpto reduce thesuicide rate?

In 1999, in its White PaperSaving Lives: Our HealthierNation, the government set atarget to reduce the suiciderate by at least one-fifth by theyear 2010 (Department ofHealth, 1999). The NationalSuicide Prevention Strategy forEngland was launched in 2002,in order to guide and co-ordinate efforts to achieve thattarget. Goal 3 of the Strategy isto reduce the availability andlethality of suicide methods(Department of Health, 2002).

Three conditions are necessaryin order for a suicide attempt totake place. The individual must:a) resolve to die or to give up

on life;b) decide on a method (e.g.

hanging, overdose, jumping);c) obtain the means by which to

carry out the plan (e.g. rope,tablets, jumping site).

There is general agreement thatit is possible to interrupt thesuicidal process by making itdifficult for people to obtain themeans by which to killthemselves. Restricting accessto means of suicide isrecognised as having thepotential to save lives (Cantor &Baume, 1998; Gunnell,Middleton, Frankel, 2000). Themost compelling evidence forthis comes from the dramaticreduction in suicides thatfollowed the withdrawal of toxiccoal gas from British homesduring the 1960s and early 70s(Kreitman, 1976).

Falls in suicide rates have alsobeen shown to be associatedwith the introduction of catalyticconverters in cars (Amos,Appleby, Kiernan, 2001),changes in firearms legislation(Cantor & Slater, 1995;Beautrais, Fergusson, Horwood,2006) and the introduction oflimits on sales of paracetamol(Hawton, Townsend, Deeks etal, 2001; Hawton, Simkin, Deekset al, 2004).

Some 'method substitution'inevitably occurs. If one meansof suicide is made unavailable,there will always be people whoare determined enough to seekout an alternative means ofkilling themselves. Measures tolimit the availability of means areaimed mainly at reducing thosesuicidal acts that are impulsiveor are the result of an acute ortemporary crisis. Making itdifficult to access the means ofsuicide is a way of 'buying time'and giving the individual achance to reconsider. It does notsolve the problems that gaverise to the suicidal impulse, norlessen the mental suffering ofthe individual, and is therefore afairly crude approach toprevention (Gunnell, Middleton,Frankel, 2000). Nevertheless, itis recognised as effective andhas a place in the suicideprevention strategy of everynation that has one (Simkin,Hawton, Sutton et al, 2005).

Restricting access to lethalmeans is an important elementin an overall suicide preventionstrategy because it targets thewhole population and provides away of reaching the many at-riskindividuals who are not incontact with health and socialcare services. Identifying andmanaging frequently usedlocations is one way ofrestricting access to the meansof suicide. It removes thespotlight from high-risk peopleand focuses on high-risk places.

Page 7: Guidance on action to be taken at suicide hotspots - SPRC

5

3. What types oflocation are likelyto be hotspots?

High-risk places are those thatprovide opportunities forsuicide by:• jumping from a height • placing oneself in front of a

moving vehicle • other methods, particularly

car exhaust poisoning

Suicide by jumping from a heightLocations offering opportunitiesfor suicide by jumping includebridges (vehicle and pedestrian),viaducts, high-rise hotels, multi-storey car parks and other tallbuildings, cliffs and othertopographical features.

Jumping from a high place is arelatively uncommon method ofsuicide. In 2004, there were just154 cases in England andWales, accounting for 3% of allsuicides and open verdicts(Office for National Statistics,2005). In some other countriesof the world, the proportion isvery much higher (Gunnell &Nowers, 1997).

However, suicidal jumps almostinevitably occur in publiclocations, have a high fatalityrate (Spicer & Miller, 2004) andare highly traumatic forwitnesses and people livingbelow the jump site (Reisch &Michel, 2005). Jumps also tendto attract media attention, whichhelps places to gain macabrereputations and can lead tofurther copycat suicides. All theworld's leading suicide hotspotsare in fact jumping sites.

Suicide by jumping or lying in front of a moving vehicleSuicide by jumping or lying inf ront of a moving object is againfairly uncommon in this country.The Office for National Statistics(ONS) re c o rded 150 suchdeaths in England and Wales in2004, occurring on road and railnetworks combined. Again, thisequates to 3% of all suicidesand open verdicts in Englandand Wales for that year.

The Rail Safety and StandardsBoard (RSSB) collects its owndata on railway suicides and,using slightly different criteriafrom the ONS, recorded 181such deaths in the same year onthe railways alone. Of these,approximately 50-60% occurredon open track, 30% at stationsand 10% at level crossings.

These figures are very lowc o m p a red with other Euro p e a ncountries that have much denserrail networks (Kerkhof, 2003).H o w e v e r, the emotional andpsychological damage caused todrivers and other witnesses ofsuicides on the transport networksis immense (Williams, Miller,Watson et al, 1994). Services aredisrupted and, in the event that atrain is derailed, further fatalitiesand serious injuries may result.

The RSSB is taking the problemof railway suicide very seriouslyand is working hard to achievethe target of a 20% reduction inrailway suicides set out in theNational Suicide PreventionStrategy. A major report onSuicides and Open Verdicts onthe Railway Network (SOVRN) wascommissioned in 1999 to identifyways of reducing the incidenceand impact of railway suicides.

Following publication of thereport in 2003, the former RailFatalities Management Group(disbanded in April 2006) wasset up, and as part of this worksponsored a series of visits tostation operators by RSSB andSamaritans. During these visits,the issues were discussed andvarious counter measuresexamined. As a result, somestation operators are planning todisplay Samaritans posters, offerdirect phone links to Samaritansand send staff on sessions withthe Samaritans providing suicideawareness training for railwaystaff and counselling formembers of staff affected by arailway suicide. Network Rail isalso carrying out an extensiveprogramme of fencing to restrictpublic access to rail tracks.

London Underground Limited(LUL) has also recentlycompleted a pilot initiative at one of its stations in an effort to reduce suicides and suicidalbehaviour on the LondonUnderground. LUL hasimplemented measuresincluding: staff training,increasing the number of CCTVcameras in strategic positions,Samaritans posters andSamaritans phones.

The Highways Agency, which isresponsible for the constructionand maintenance of motorwaysand major trunk roads inEngland, is also keen to take aproactive stance towardssuicides and is currentlyintroducing preventativemeasures at a number ofmotorway bridges from whichsuicidal jumps have been made.

Page 8: Guidance on action to be taken at suicide hotspots - SPRC

6

Suicide by other methodsNo figures are available forsuicides by other methods thatoccur at high-risk publiclocations. However, two studies(Pearson, 1993; King & Frost,2005) have highlighted theopportunity for suicide presentedby secluded car parks or otherisolated rural locations, where anindividual can sit undisturbed ina car for as long as is necessaryin order to die. Such locationsare associated mainly with carexhaust poisoning. They do notsupply the means of suicide assuch, but do provide peace andseclusion, which is a vitalingredient in many suicides.

The rate of car exhaustpoisoning is steadily declining inthe UK, thanks to modificationsin vehicle design. In 2004, theONS recorded 225 cases of carexhaust poisoning in Englandand Wales, accounting for 4.5%of all suicides and open verdicts.Almost nothing is known aboutwhere these take place andresearch is difficult because HMCoroners are currently notrequired to record the location ofa suicidal act. However, it islikely that public car parks andlaybys are used in a significantproportion of cases.

Analysis of pilot data collected inDevon revealed that 76% of allcar exhaust poisonings occurredin public places, and theseincluded car parks at beautyspots, in well-known areas ofwood and moorland, oncommons and in other rurallocations. The pattern for eachcounty is likely to be differentand will be determined by localgeography. Analysis of local datais essential in order to identifyfrequently used locations.

Any site at which a suicide hasoccurred can achieve notorietyand quickly become a hotspot,particularly if the death isreported extensively in the newsmedia. A prime example of thisis Cheung Chau Island, apopular 'getaway' destination offthe coast of Hong Kong. Afternewspapers publicised the factthat a visitor to the island hadbrought about his own death byburning charcoal in a rentedholiday flat, the island becamepopular as a venue for suicideand many more visitorssubsequently killed themselvesusing the same method (Yip,2005). It is well known thatmedia reporting of suicides canresult in copycat behaviour, andsuicidal acts carried out in publicplaces are more likely to attractmedia attention than thosecarried out in private homes.

The contribution to theachievement of the nationalsuicide reduction target that willbe made by introducing safetymeasures at specific high-risksites is likely to be relativelysmall. Nevertheless, suicidesthat occur in public places havefar-reaching consequences forthe health of others and therebycontribute to the overall burdenof mental illness andpsychological distress.

4. How manysuicides areneeded to make a hotspot?

More than one suicide at aparticular site, in any period forwhich there are records, shouldgive cause for concern. This issufficient to demonstrate thatthe site has appeal for suicidalindividuals and offers eithermeans or opportunity forsuicide. However, findings fromthe SOVRN report into suicideson the railways suggest thathotspots may be transient,shifting rapidly from one part ofthe network to another.

There are varying degrees of'hot'. Decisions on what actionto take will depend on thenumber and nature of suicidalacts, the frequency with whichthey occur and thefatality/serious injury rate, as wellas on site-specific factors.

High-risk locations may vary insize. A single car park on acommon may have been thevenue for more than one suicide,and would therefore beconsidered a hotspot. However,there may be a number of carparks on a common that havehad one suicide each. In thiscase, it would be sensible totreat the common as a whole asa hotspot for suicide. Similarly, ashort stretch of motorway or ofcliffs might be designated ahotspot if there has been aspate of jumping incidents, eventhough these may haveoccurred at different points.

Page 9: Guidance on action to be taken at suicide hotspots - SPRC

7

A suicide pact, in which two ormore individuals die together byarrangement, should be countedas a single incident.

Local suicide prevention groupswill need to exercise judgmentand make their own decisions asto the size and boundaries ofeach site, based oninterpretation of local data andknowledge of local geography.

5. Why is inter-agencycollaborationimportant?

Managing suicide hotspotsinvolves complex questions ofownership, responsibility andresources (Kerkhof, 2003).

Partnership is increasingly beingrecognised as vital to effectivedelivery of public services. Thegovernment's new strategy forlocal government includes thedevelopment of Local AreaAgreements (LAAs), which aredesigned to strengthenpartnership working at local leveland facilitate more co-ordinatedservice delivery. LAAs provide amechanism for setting localpriorities and developing localsolutions and delivery plans, inhealth as well as in other keypolicy areas. Multi-agency LocalStrategic Partnerships, includingpublic services, privatecompanies, voluntaryorganisations and communitygroups, are intended to play acentral role in drawing up theagreements.

The White Paper Our Health,Our Care, Our Say (Departmentof Health, 2006a) sets out asimilar vision to deliver joined-upcommunity services that areresponsive to local patient needsand are prevention-focused. Itassigns joint responsibility toDirectors of Public Health andDirectors of Adult SocialServices for undertaking regularstrategic needs assessments todetermine local priorities foraction (see also Department ofHealth, 2006b).

The railway industry is alsosetting up, under thechairmanship of Network Rail, a series of Community SafetyPartnership Groups, whose remitis to agree local strategies formanaging a range of risks,including suicide, vandalism and assault.

No single agency is responsiblefor suicide prevention. Healthand social care services have astatutory responsibility to identifyand manage at-risk individualswith whom they come intocontact. Police and otheremergency services have a keyrole to play, as do a number ofvoluntary organisations. Localauthorities, agencies thatmanage and maintain thetransport networks, owners ofhigh-rise buildings, privatelandowners and bodies such asthe National Trust that ownslong stretches of coastline allhave a responsibility for ensuringpublic safety.

At the same time, in consideringwhat action to take at suicidehotspots, the interests ofenvironmentalists, ramblers andmany local community groupswill need to be taken intoaccount. Both the economiccost and the environmentalimpact of interventions need tobe considered.

For these reasons, werecommend wide inter-agencyco-operation in the identificationand management of suicidehotspots. This should form partof the local arrangements fordelivery of interventions towardsachieving the national suicidereduction target and should beseen as part of a whole systemsapproach to suicide prevention.

Page 10: Guidance on action to be taken at suicide hotspots - SPRC

8

Although the numbers involvedwill be relatively small, there isgood evidence that lives canbe saved by either impeding ordeterring suicide attempts athigh-risk locations. This sectionprovides a brief summary ofthe evidence from bothpublished studies and othersources on the measuresavailable and theireffectiveness.

1. Physical barriersThe most effective form ofprevention at jumping sites is aphysical barrier, which literallyrestricts access to the drop.Safety nets serve a similarpurpose but rescue from a net may be difficult should ajump occur. A study in Bern,Switzerland showed thatsuicides at the MuensterTerrace, a well-known jumpingsite in the old city, ceasedcompletely following theinstallation of a safety net.Furthermore, there was nochange at other nearbyjumping sites, suggesting thatwould-be jumpers did notsimply go elsewhere (Reisch & Michel, 2005).

Further compelling evidencecomes from a New Zealandstudy of the effect of removingbarriers from a city bridge.Safety barriers that had been in place for 60 years weredismantled as a result ofpressure from communitymembers, who were concernedthat they were unsightly.The study showed that theremoval of barriers led to a five-fold rise in the number ofsuicides from this particularbridge, while suicidal jumps atother nearby sites decreased(Beautrais, 2001).

In the UK, the Clifton SuspensionBridge in Bristol erected barriersin 1998 and a study of their eff e c ton local patterns of suicide isc u r rently in pro g ress. Many of the most popular jumping sitesa round the world have installedbarriers of some sort and in every case the authorities claimthat significant reductions insuicide rates have beenachieved. These include theBloor Street Viaduct in To ro n t o ,the Jacques Cartier Bridge inM o n t real, The Sydney HarbourBridge, the Gateway Bridge inBrisbane, the Empire StateBuilding and the Eiffel To w e r.

Finally, in a small but persuasivestudy based on interviews withindividuals who survived suicidaljumps from the Golden Gate andSan Francisco-Oakland BayBridges, all of the survivorscalled for the construction ofsuicide barriers (Rosen, 1975).

Any form of obstruction at ahigh-risk site not only gives theindividual time to reconsider but,by delaying the suicidal act, mayalso increase the chances ofintervention (Lindqvist Jonsson,Eriksson et al, 2004).

The issue of erecting suicidep revention barriers or nets atparticular sites is invariably a highlycontentious one. There is oftens t rong public opposition onaesthetic grounds, particularly ifthe site or structure is a famouslandmark. However, a studyconducted by engineeringundergraduates at the University ofC a l i f o rnia in Berkeley show thatbarriers can be both effective andaesthetic. After years of fierc ecampaigning, the installation ofbarriers at the Golden Gate Bridgein California is still being re s i s t e don grounds of aesthetics alone,despite a steady toll of around 25suicides per year (one every twoweeks on average). However, in2006 the Board of Directors of theGolden Gate Bridge Highways andTransportation District voted toa p p rove a 2 year study onpossible alterations to the bridge.The first phase would review paststudies and review suicide barrierdesigns such as a net under thebridge, a fence added to existingrailing, or total replacement ofexisting railings. The second phasewould conduct a more thoro u g hengineering, environmental andcost study of the designs. Onlarger structures, such as majorriver crossings, they can alsop resent complex engineeringchallenges. Wind resistance isalways a major issue on bridges,but there will be other diff i c u l t i e sthat are peculiar to each structure ,due to the fact that bridges are allbuilt to individual designs.

INTERVENTIONS AT HOTSPOTS:A REVIEW OF THE EVIDENCE

Page 11: Guidance on action to be taken at suicide hotspots - SPRC

9

Physical barriers are likely to bethe most costly of availablemeasures, and resolving theissue of financial responsibilitycan be difficult (Kerkhof, 2003).Barriers are also permanent andtherefore may not be warrantedunless it is clear that a locationwill remain attractive as a suicidesite for the foreseeable future.Particular sites may only behotspots for a short time.

Nevertheless, raising the heightof an existing parapet orinstalling a further barrier ofsome sort is a serious option toconsider, if a site has been usedfor suicidal jumps on two ormore occasions or if there havebeen acts of vandalism or otherincidents that endanger thepublic, such as objects beingthrown from a bridge onto aroad or railway below.

T h e re are numerous options asre g a rds materials and designs. Themain condition is that the materialshould not provide any footholdand the barrier itself should be asd i fficult as possible to scale. A totalheight of between six and nine feetis considered to be eff e c t i v e(Berman et al, 1990).

While the cost of installing safetybarriers may be high, warn i n g shave been issued in the USA thatlegal action could be taken byrelatives, either of a suicide victimor of a person killed or injured by afalling body, if a site is well knownfor suicide and the authoritieshave not taken action to impro v esafety (Berman et al, 1990).

2. Signs andtelephone hotlines

Signs encouraging distressed orsuicidal individuals to seek helpand displaying a contactnumber for the Samaritans, arein place at a number of locationsin the UK that have beenidentified as hotspots, includingBeachy Head cliffs, CliftonSuspension Bridge in Bristol andmany less famous sites.

Signs displaying the Samaritans'national help-line number and thelocation of the nearest publictelephone were positioned inselected car parks in the NewF o rest in Hampshire as part of amulti-agency suicide pre v e n t i o ninitiative, after it was discovere dthat they were associated withhigh numbers of car exhaustsuicides. A 3-year evaluation of thescheme showed a significant dro pnot only in the number of car parksuicides, but also in the totalnumber of suicides in the NewF o rest district (King & Frost, 2005).A further 3-year evaluation(unpublished) has shown that thenumber of car park suicides hasremained low, and that most of thesuicides that have subsequentlyo c c u r red in the New Forest havebeen in car parks without signs.

On the Mid-Hudson Bridge in theUS, dedicated suicide pre v e n t i o nhotlines are linked directly to a24-hour Psychiatric EmergencyService. A two-year evaluation ofthe scheme showed that, out of39 would-be jumpers, 30 usedthe phone to call for help and, ofthese, only one went on to makea fatal jump, whilst 5 of thosewho did not use the phonejumped to their deaths (Glatt,1987). Crisis hotlines are alsoinstalled on the Golden GateBridge and at many otherjumping sites worldwide.

Signs promoting help seekingand advertising appropriatesources of help are almostinvariably the best initial step atany location that is causingconcern. They are low cost andcapitalise on existing services,both voluntary and statutory,such as Samaritans and NHSDirect. Samaritans' nationaloffice is able to advise on thewording and design of signs.Analysis of local patterns ofsuicide will determine whether ornot there is a need for signs inminority languages and contactnumbers of organisationsoffering support to specificminority groups.

Another major advantage ofsigns is that they are notmethod-specific. Analysis of pilotdata collected in Devon revealedthat a number of local hotspotswere associated with more thanone method of suicide, e.g.jumping and hanging, or jumpingand carbon monoxide.Strategically placed and carefullyworded signs have a chance ofspeaking to all individualscontemplating suicide,regardless of their chosenmethod, whereas physicalbarriers only prevent jumping.Again, close scrutiny of localpatterns is essential.

Objections to Samaritans signshave been raised at someknown hotspots, on the groundsthat they may 'promote' thelocation as a venue for suicide.There is no evidence to supportthese fears.

The main limitation of both signsand telephone hotlines is thatthey rely on the individual beingambivalent enough aboutsuicide to make the call.

Page 12: Guidance on action to be taken at suicide hotspots - SPRC

10

3. Suicide patrolsDedicated suicide patrolsrepresent a more proactiveapproach than signs andtelephones, and have beentried at a number of locationsworldwide. There are nopublished studies, however,and the evidence ofeffectiveness is weak. TheGolden Gate Bridge inCalifornia is patrolled every dayduring daylight hours by paidsuicide prevention officers. It isclaimed that there has been asignificant reduction in fataljumps since the patrols wereintroduced in 1996, and thathuman contact is morepowerful in preventing suicidaljumps than a physical barrier.

Other sites attracting highnumbers of suicides arepatrolled by committedvolunteers. At Beachy Head cliffsin Sussex, a team of volunteercounsellors is on duty everyevening. The scheme is run by acharity that was set up by alocal man whose wife jumped toher death from the cliffs, and iscredited with having reduced thenumber of fatal jumps.Depending on the length of thebridge or size of site, a dutyteam may need to be supportedby CCTV cameras in order tospot distressed individuals.

Paid suicide patrols are a highlycostly option, which is unlikely tobe justified unless a locationattracts very high numbers ofsuicides and there is a very clearpattern of use at certain peaktimes. The alternative is to relyon volunteers. Either way,patterns of suicides at the sitewill need to be studied closely toidentify times of the day, week,month or year at which patrolsor counsellors are likely to bemost effective.

There is some concern, basedon anecdotal evidence, that anill-timed intervention mayprecipitate a suicidal act(Berman et al, 1990).

4. Training for staffof non-healthagencies workingat or near hotspots

If a dedicated suicide patrol isnot a realistic option, theremay be other staff working ator in the vicinity of a high-risklocation and who can play arole in identifying individuals in distress, alerting emergencyservices and intervening if necessary.

A telephone survey of 10 majorUK toll bridges found that allbridge authorities expected theirstaff to be alert to the possibilityof suicide attempts, and all hadclear protocols for staff to followin the event of a person actingsuspiciously. However, only 1out of the 10 provided specifictraining in suicide awareness orsuicide prevention. Two othershad had occasional staffbriefings given by either theSamaritans or the local police.

Samaritans can providepackages of suicide awarenesstraining tailored to the needs of individual organisations. They are currently working withthe Train Operating Companiesto deliver a training programmeto station staff, in order toincrease staff confidence inidentifying and responding to at-risk individuals.

Page 13: Guidance on action to be taken at suicide hotspots - SPRC

11

Samaritans' training for localauthority and ForestryCommission staff was includedas part of the New ForestSuicide Prevention Initiative (King & Frost, 2005), but wasnot evaluated.

Highways Agency Traffic Officers (HATOs) who patrol the motorways, car parkattendants, countryside rangersand staff of many other non-health agencies whose workregularly takes them near knownhotspots may benefit fromreceiving basic training in suicideawareness and responding topeople in distress.

5. Restrictions onmedia reporting

It is known that news reports ofsuicides are associated with asubsequent increase in suicides,and that the greater the mediacoverage, the greater thesubsequent increase in numbers(Pirkis & Blood, 2001).Conversely, restrictions onreporting of suicide have beenshown to be associated withsustained reduction insubsequent suicides (Sonneck,Etzersdorfer, Nagel-Kuess,1994). Detailed reporting of themethods used is known toencourage imitation (Sonneck,Etzersdorfer, Nagel-Kuess,1994; Yip, 2005).

Suicides that occur in publicplaces and involve 'spectacular'acts such as jumping fromlandmark structures or sites aremore likely to attract mediaattention than those that occurin private homes.

Negotiating with the news mediato limit reporting is therefore avital element in the managementof suicide hotspots. Followingrecent revision, the PressComplaints Commission Codeof Practice now deals specificallywith reporting of suicide.

A new clause, introduced toprevent copycat suicides,demands that care be taken toavoid excessive detail about themethod used(http://www.pcc.org.uk/cop/practice.html). This should provide abasis for discussions betweenlocal stakeholders and themedia. In addition, Samaritanshave published media guidelinesin downloadable format(http://www.samaritans.org/know/pdf/media.pdf).

Agreement should be securedfrom local news editors toabstain not only from reportingon actual cases of suicide athigh-risk sites, but also fromreporting on any preventativemeasures being introduced atthe site, since this too may draw attention to the site'spotential as a suicide spot (King & Frost, 2005).

Page 14: Guidance on action to be taken at suicide hotspots - SPRC

12

6. Deciding betweenavailable options

Each site under considerationwill be an individual case andsubject to local conditions. Thechoice of measures will dependlargely on the size of the problem.

The number of suicidal acts,their nature (method of suicide),the frequency with which theyoccur and the fatality/seriousinjury rate will determine whatlevel of intervention is deemednecessary at a particular site.Engineering, environmentalconsiderations and pressure

from local interest groups willalso influence the decision.

Other options to consider arei m p roved CCTV and lighting/visibility at particular sites. There islittle evidence that these measure sused in isolation prevent suicidesunless considered as part of apackage of measure s .

SUMMARY OF AVAILABLE MEASURES: PROS AND CONS

Physical barriers

Signs and telephone hotline

Suicide patrols

Training for staff of non-health agenciesworking at or near hotspots

Restrictions on mediareporting of suicides athotspots

Pros

• good evidence ofeffectiveness;

• i n c rease chances of interventionby delaying the jump;

• recommended by survivors ofsuicidal jumps;

• prevent other acts ofvandalism that endangerpublic, e.g. throwing things.

• good evidence of effectivenessfor signs alone;

• not method-specific;• low cost;• use existing voluntary services,

e.g. Samaritans.

• human contact may beimportant.

• i n c reased likelihood ofidentifying individuals in distress;

• i n c reased confidence in re s p o n d -ing to individuals in distress;

• increased likelihood ofemergency services beingalerted in time.

• good evidence ofeffectiveness;

• cost-free.

Cons

• aesthetic considerations,particularly at famouslandmarks;

• in some cases may presentcomplex engineeringchallenges;

• high cost in some cases ;• permanent;• method-specific.

• rely on suicidal individual tomake the call.

• weak evidence of effectiveness; • paid patrols costly; • may need to rely on volunteers;• ill-judged intervention may

precipitate suicidal act;• peak high-risk times need to

be identified.

• likelihood of any staff memberencountering a potentialsuicide may be small;

• no evidence of effectiveness(untested).

Page 15: Guidance on action to be taken at suicide hotspots - SPRC

13

1. The role of anInter-agencyForum on Self-Harm & Suicide

As a first step, we recommendthe setting up of a countywideInter-agency Forum on Self-Harm & Suicide to improve theintegration of services, to shareevidence on best practice andto own a portfolio of workaimed at reducing suicide andself-harm.

The Forum may wish to meet atleast annually and will draw itsmembership from seniorexecutive level stakeholders in awide range of statutory andvoluntary agencies, including:• primary and secondary health

care trusts• adult social services • children's trusts• school health services• police

• criminal justice agencies(prisons and probation service)

• ambulance services• relevant voluntary

organisations (e.g. Samaritans, Mind)

• drug action teams • service users and carers• academic partners

The lead in convening and co-ordinating the Forum is likely tocome from within the health andsocial care community. TheForum should ideally have thecapacity and authority to set upand co-ordinate operational-levelworking groups, and to carry outtheir recommendations usingexisting executive channels in itsmember organisations.

PART 2IDENTIFYING AND MANAGING SUICIDEHOTSPOTS: A PRACTICAL GUIDE TOINTER-AGENCY COLLABORATION

ORGANISATION, PLANNING AND PERSONNEL

Page 16: Guidance on action to be taken at suicide hotspots - SPRC

14

2. Planning thehotspotsprogramme

The identification andmanagement of suicidehotspots will constitute oneprogramme of work, possiblysitting alongside others in abroad portfolio of local suicideand self-harm prevention work,to be carried out under theoverall leadership of the Forum.

2a Programme personnelThe hotspots programme willideally have; a lead agency; anexecutive sponsor; a clinicalchampion and a steering group,in addition to a dedicatedprogramme team.

Lead agency

The programme will need to beled by a health or social careagency (PCT or specialist MentalHealth Trust) and to be locatedwithin an existing departmentwith a track record for gettingthings done, possibly withinResearch & Development orClinical Effectiveness.

Where there is an establishedlocal Suicide Audit Group, it maymake sense for the lead tocome from this group, since itwill already have mechanisms inplace for collecting andanalysing data relating to localsuicides, feeding back findingsto primary and secondary care services and coordinatinglocal arrangements.

The energy and commitment ofthe lead agency will be a criticalfactor in securing theengagement of key stakeholders,maintaining momentum andcarrying the programme throughto completion.

Executive sponsor

This will be someone in a keystrategic role, such as Chair ofthe Local Implementation Team(LIT), with responsibility fordevelopment and delivery oflocal services, a track record ofcommunication and engagementwith a wide range of agenciesand authority to ensure thatdecisions are translated intopractice.

Clinical champion

Ideally this will be a senior fro n t l i n ehealth or social care professionalwho has credibility with peersand is able to liaise between thehotspots programme andservice planners and providers.

Programme team

In addition to the above, thefollowing skills are needed withinthe team:• quantitative and

qualitative data collection and data analysis;

• familiarity with Excel and/orSPSS for handling data;

• use of GIS mapping software; • effective networking

and communication with key stakeholders and partner agencies;

• conference organisation;• project management; • secretarial and

administrative skills;• Steering group.

Ideally, the programme teamshould be able to draw on theexpertise of a wider steeringgroup, with service user andcarer representation and, wherepossible, an external expert insuicide prevention.

2b Key early tasksKey early tasks include: drawinga boundary; identifying andprocuring necessary resources,and engaging key stakeholders.

Drawing a boundary

A key early task will be to set the boundary of the area to beincluded within the programme.If the programme is beingconducted under leadership ofan Inter-agency Forum whoseconstituency is countywide, itwill make sense for theprogramme to identify and tacklehotspots across the county.Pilot work demonstrated that it is both appropriate andfeasible to operate within acounty framework.

Identifying and procuringnecessary resources

It will be necessary to cost theprogramme fully and to exploreexisting shared resource,including personnel,administrative support, ITfacilities and software andconference facilities, beforeseeking additional budget withinpartner agencies. Appendix 1provides a guide to the possiible resource implicationsassociated with the programme.

Page 17: Guidance on action to be taken at suicide hotspots - SPRC

15

Stakeholder engagement

Identifying and involving a widerange of stakeholders andpartner agencies is critical to the success of the programme.The programme team wouldneed to consider:a) identifying key agencies and

g roups with maximumintelligence in relation topotential suicide hotspots, and

b) identifying a relevant contactwithin each agency andsecure their commitment tothe programme.

Key agencies to involve will bethose who own, manage ormaintain high-risk locations.Their representatives will need tobe at an executive level, withauthority to allocate budget andtake action. Representation mayneed to come from outside thecounty (from regional or nationalboards) in order to carry thenecessary authority for action.

A list of potential stakeholdersand partner agencies is given inAppendix 2. Local areacharacteristics and the presenceof particular sites and structureswill determine the membership,and the full range of relevantstakeholders will not emergeuntil local data have beenanalysed and high-risk locations identified.

A stakeholder conferenceprovides a mechanism forbringing interested partiestogether, capturing localintelligence and specialistknowledge relating to particulartypes of site or structure,agreeing local priorities andforming planning groups. A suggested format for theconference is outlined below(see: Managing hotspots 1:Agreeing local priorities andAppendix 5). This should nottake place, however, untilcollection and analysis of localdata are complete.

Page 18: Guidance on action to be taken at suicide hotspots - SPRC

16

1. Data collectionThe key to success inidentifying and responding tosuicide hotspots is thesystematic and ongoingcollection of local data onsuicides and open verdicts.

Continual real-time collectionand interrogation of local dataare essential in order to monitortrends and to recognise andrespond quickly to any suddenescalation in use of a particularsite. Findings from the SOVRNreport into suicides on therailways suggest that hotspotsmay be transient, shifting rapidly from one part of thenetwork to another.

1a Who is alreadycollecting it?

The first step is to establish whatdata are already being collectedand by whom, in order toprevent duplication of effort.PCT information analysts andclinical audit managers insecondary care trusts should beconsulted to find out whetherthey are collecting data fromcoroners as part of their suicideaudit process.

S t a n d a rd 7 of the NSF for Mental Health (1999) includeddevelopment of local systems forsuicide audit by local health andsocial care communities. If so,data-sharing arrangementsshould be established that allowfor frequent review of hotspots.The NIMHE Primary Care SuicideAudit Tool (available ath t t p : / / w w w. e a s t m i d l a n d s . c s i p . o r g. u k / s u i c i d e _ d b / i n d e x . h t m l )p rovides compre h e n s i v eguidance, together with as t a n d a rd data collection pro f o r m aand electronic database. Thebenefit of the NIMHE tool is that ite n s u res that all localities arecollecting the same minimumdata set, thereby permittingcomparative analysis at local,regional and national levels.

If no system is in place for suicideaudit, or if the data collected donot include 'Location of act' (seebelow: 2c What data to collect),the team will need to collect therelevant data themselves.

1b Where to find the dataThe main sources of data oncompleted suicides and deathsby undetermined injury (openverdicts) are the local coronersand the County Records Office.

All violent, unnatural or suddendeaths are subject to aninvestigation or inquest by thecoroner in whose district thedeath occurs. Coroners' recordscontain the information requiredfor identifying the location ofacts resulting in suicide orundetermined death.

Permission will need to besought from each coroner withinwhose jurisdiction the countyfalls in order to access therecords. Most coroners arewilling to facilitate research oraudit that is clearly in the publicinterest, but they are under noobligation to do so. A largecounty may be served by two ormore coroners. Contact detailsof all coroners in England andWales are available from theCoroners' Society(http://www.coroner.org.uk/public/search.asp), or from CountyCouncil web pages.

Searching coroners' records is atime-consuming task and thereis unfortunately no shortcut. Allcoroners report on an annualbasis to the Office for NationalStatistics (ONS), which istherefore able to supplyaggregated data on suicides andundetermined deaths. However,the ONS does not receive anyinformation relating to location ortiming of the suicidal act, whichare the key variables here.

Once permission to access acoroner's records has beenreceived, the Register of Deathsfor each year of interest willneed to be studied in order tocompile a list of suicides andopen verdicts. The next step isto request the file relating toeach suicide or open verdict andto extract the necessaryinformation. Coroners' practicesvary widely, as will the contentsof their files.

IDENTIFYING SUICIDE HOT SPOTS

Page 19: Guidance on action to be taken at suicide hotspots - SPRC

17

Typically, an individual file willcontain an initial police report,completed at the scene of anysudden or unexplained death,reports of post-mortemexamination and anytoxicological analyses, reportsfrom medical practitioner(s),witness statements, originalsuicide notes and any otherevidence presented at inquest.

We suggest setting up an Excelor SPSS spreadsheet inadvance of the visit and enteringrelevant data directly from thefiles using a laptop computer.This removes the need forphotocopying and ensures thatonly such information as isstrictly necessary leaves thecoroner's office.

After one or two years(depending on volume),coroners' files are moved to theCounty Records Office forarchiving. Access to these willbe required for retrospectivedata collection and will need tobe authorised by the coroner,but thereafter the procedure willbe exactly the same as for thecoroners' offices.

The content of coroners’ files is frequently highly distressingand whoever is collecting thedata should be offered someform of debriefing.

1c What data to collectFor the purposes of suicide auditand research, it is customary toinclude open verdicts along withsuicides. An open verdictcategory will include caseswhere suicide was suspectedbut could not be proven.

R e s e a rch has shown that somecauses of death, includingjumping or falling from a height,a re particularly difficult to establishas suicides under the coro n e rsystem and are more likely toreceive an open than a suicidev e rdict (Cooper & Milro y, 1995;Sampson & Rutty, 1999). Therecommended practice is to re a dthe contents of each open-verd i c tfile care f u l l y, make a judgement asto the likelihood of suicide andinclude those of moderate to highlikelihood (Hawton, Appleby, Plattet al, 1998).

I n i t i a l l y, we recommend collectionof five years’ re t rospective data ino rder to establish a baseline.T h e re a f t e r, arrangements will needto be made for updating of thedatabase at 3-6 monthly intervals.

Appendix 3 provides a list ofvariables to include. Somepersonal data are needed inorder to check that all suicidesand open verdicts have beenincluded and that none hasbeen duplicated, and toestablish whether or not thesuicide took place at theindividual's home address.Some demographic informationmay also be helpful inestablishing a profile of users ofparticular locations. Ethnicity willbe important in order toestablish whether signs shouldbe provided in minoritylanguages, with contactnumbers of organisationsoffering support to specificminority groups. However,personal details should be keptto a strict minimum (see below:1e Ethics). The essentialvariables for identification ofhotspots are the location anddate of the suicidal act.

Location of act

The location of the act resultingin death by suicide orundetermined injury may not beimmediately apparent. Coronersare required to record place ofdeath, which may notnecessarily be where the suicidalact occurred. If the individualwas still alive when found andsubsequently died in hospital,the hospital will be given as theplace of death. Identifying theprecise location of the suicidalact will involve readinghandwritten statements or freetext entries.

If the act occurred in a publicplace (see below: Public orprivate location), as muchinformation as possible shouldbe captured in order for theexact location to be pinpointed.The location should be enteredin a free text field using placenames and as much narrativedetail as is available, e.g. “Foundin vehicle parked in gateway tofield on unclassified roadbetween Foxbridge andHareswell, just on brow ofCrows Hill.” For subsequentmapping using GeographicalInformation System (GIS)software, a postcode will beneeded. For public locations,this is unlikely to have beenrecorded. However, a text fieldshould be set up so that it canbe ascertained and entered later.If by any chance an OrdnanceSurvey grid reference has beenrecorded, this should becaptured, since this will enablethe location to be identified withthe greatest precision.

Page 20: Guidance on action to be taken at suicide hotspots - SPRC

18

Public or private location

Suicides and open verdictsshould be classified according tothe status of the location inwhich the act occurred. Relevantdeaths for the identification ofhotspots are those resulting fromacts carried out in public places.Pilot findings suggest thatapproximately one third of allsuicides and open verdicts fallinto this category, whilst theremaining two thirds occur atprivate addresses.

Definitions of public and privatelocations will need to be agre e dwithin the team. For purposes ofthe pilot, a private location wasdefined as any private homea d d ress, including a farmer'sown land, but excluding hotelsand guest houses unless thepermanent residence of thedeceased. The definition of apublic location included all openland not owned by the deceased,transport networks, publicbuildings, and hotels and guesthouses in which the deceasedwas a temporary re s i d e n t .

Local issues and areacharacteristics will determine the value of including particulartypes of location in the publiccategory. The pilot wasconducted in a county in whichthe tourist industry plays a majorrole, and the inclusion of hotelsand guest houses in the analysiswas considered valuable inorder to gauge the extent of'suicide tourism'.

Psychiatric in-patient units,prisons and probation hostelsare known to house high-riskindividuals and will already havemeasures in place to managesuicide risk. These are bestclassified as private andexcluded from the analysis, inorder to keep the hotspotsprogramme sharply focused.

Date and time of act

The date of death, as recordedon the coroner's certificate afterinquest, may not necessarily bethe date on which the suicidalact occurred. However, in mostcases it is difficult, if notimpossible, to ascertain thelatter. Police reports and witnessstatements will give details ofwhen the body was found, andthis will have to serve as a proxymeasure. It is in any caseunlikely that a suicide thatoccurred in a public location willhave gone undiscovered for along time.

1d Additional sources of data

Coroners' records will onlysupply information on completedsuicides. There will be manymore 'near misses' or serioussuicide attempts that did notresult in death. The followingadditional sources of data maybe useful for supplementing the picture.

Police and emergencyservices

Police operational logs containdetails of every incident to whicha police officer is called out.These will include someattempted or threatenedsuicides, where death may havebeen prevented by theintervention of police officers ormembers of the public. Localforce information officers may bewilling to assist in identifying thelocations of such incidents.

Hospital admissions data

Numbers of suicide attemptsthat are not fatal but result inserious injury may be useful indetermining what level ofintervention is necessary at aparticular site. Hospital EpisodeStatistics data relating tointentional self-harm areavailable from public healthobservatories, but will not help inidentifying locations. It may bepossible to obtain thisinformation from accident andemergency departments, but itwill involve a time-consumingsearch of individual records, notonly of those presenting withself-harm per se but also allcases of major bodily trauma, inorder to isolate admissionsresulting from self-inflicted injuryoccurring in a public place.

Railway fatalities database

The Office of Rail Regulationmaintains a database containinginformation about all fatalities,including suicides, associatedwith the mainline railway since1994. The database records alocation for each fatality andsuicides can be extracted on acounty by county basis. TheOffice is willing to shareanonymised data withappropriate organisationswishing to look in more detail atrailway hotspots. Details athttp://www.rail-reg.gov.uk/server/show/ConWebDoc.8070.

Page 21: Guidance on action to be taken at suicide hotspots - SPRC

19

Local stakeholder conference

Many other agencies willpossess a wealth of professionalknowledge regarding frequentlyused locations. A stakeholderconference provides the meansof capturing local intelligenceand personal stories. This willflesh out the picture built upfrom coroners' records andother sources of raw data. Abrief questionnaire sent out todelegates prior to theconference and brought alongon the day is helpful instimulating thinking in advance.A suggested format for theconference is given below (seebelow: Managing hotspots 1:Agreeing local priorities andAppendix 5).

1e EthicsCollection of data for thepurposes of audit, monitoringand service planning does notnormally require Research EthicsCommittee approval. However,all members of the programmeteam must be mindful of theneed for strict confidentiality inrelation to individual data andagreement must be reachedwithin the team as to how thiswill be maintained.

HM Coroners and CountyArchivists will require whoever iscollecting the data to sign anundertaking of confidentiality andto anonymise all data.Identification of hotspots focuseson places rather than people, soa minimum of personal data isrequired, but there isunfortunately no reliable way toway to obtain information onlocations except throughindividual records.

It is also worth considering thatsuicidal acts that occur in publicplaces will generally have beensubject to media reporting, sosome knowledge will already bein the public domain. Some mayhave become high profile casesand will be easily identifiable tolocal audiences. Whenpresenting data, for example ata local stakeholder conference,particular care should be takento maintain the focus on placesrather than people and to revealas little information as possibleabout individual cases.

Any further use of the data for purposes other than theidentification and management of local hotspots (e.g. for research with a view to publication) will be subject to Research EthicsCommittee approval.

2. Data analysisFor the purposes of identifyinghotspots, only those actscarried out in public locationsshould be included in theanalysis (see above: 1c Whatdata to collect).

2a Quantitative analysisQuantitative data can be analysedusing a number of differentsoftware packages. Basicfrequencies and tabulations canbe carried out using Excel. Astatistical package such asSPSS facilitates cross-tabulationof different variables and morecomplex analyses. The numberof deaths being analysed is likely to be relatively small, sothis may not be warrantedunless it can be supplementedby 'near miss' data.

Cross-tabulating locations withdates and times will identify anytimes of the year, month, weekor day at which particular sitesmight be especially high-risk,and will therefore help intargeting interventions. For thispurpose, locations will need tobe entered in a form that permitsquantification, rather than aslong strings of text. The field'Location type' is used for this.

Page 22: Guidance on action to be taken at suicide hotspots - SPRC

20

2b Mapping suicidesusing GIS software

By far the most graphic andeffective way of representinglocations of suicidal acts isthrough the use of GeographicalInformation System (GIS)software. A GIS packageenables any data that has ageographical or spatial elementto be linked to an OrdnanceSurvey (OS) map and displayedvisually: literally 'put on themap'. This lends itself well to theidentification of suicide hotspots.

Mapping can also highlight theproximity of suicide sites to other relevant locations such as psychiatric hospitals, prisons and probation hostels,where at-risk population groups are concentrated.

Appendix 4 provides an exampleof mapped suicide data,showing the pattern of suicidesin public locations over fiveyears. The locations shown arebased on fictitious datagenerated for demonstrationpurposes only.

There are a number of GISpackages available. Two of mostcommonly used are ArcGIS andMapInfo. There are no majordifferences between these two.The pilot was conducted usingthe Environmental SystemsResearch Institute (ESRI) ArcGIS9 system, supported byOrdnance Survey maps providedunder a pilot agreement with theNHS (England). This pilotagreement makes a wide rangeof Ordnance Survey data anddigital mapping productsavailable to NHS organisations inEngland until March 2007.Ordnance Survey has adedicated NHS helpdesk thatprovides valuable information onGIS and all products availableunder the pilot agreement (Tel:0845 458 0650, or e-mail:[email protected]).

A single user licence for ArcGIScosts approximately £1,500.One or two days are required toinstall the software, import mapsand create various libraries andlayers of information.

Each of the major softwareproducers offers training in theuse of its packages. This cancost up to £400 a day. Localuniversities may run introductoryGIS training sessions atreasonable cost. For example,the University of Bristol offers aone-day introduction to ArcGISfor £50 per person.

Many large public serviceorganisations, such as countycouncils, police forces andpublic health observatories,regularly use GIS and haveskilled analysts who may be able to assist in mappingsuicides. University departments of geography will also be able to offer adviceand practical assistance.

Page 23: Guidance on action to be taken at suicide hotspots - SPRC

21

1. Agreeing localpriorities: the role of astakeholder conference

The primary purpose of astakeholder conference is toreach agreement on where thelocal hotspots are and to setpriorities for action.

Following a presentation offindings from the data collectionand analysis phase, werecommend the use of round-table discussions and aconsensus method. A modifiednominal group technique(Gallagher, Hares, Spencer et al,1993) is appropriate for thispurpose and a suggested outlineis provided in Appendix 5.

Consensus will need to bereached on the followingquestions:

• does the county/constituencyhave any hotspots, and if sowhere are they?

• which should be consideredpriorities for action?Discussions should take intoaccount: numbers of incidents,fatality/serious injury rate andthe possibilities forintervention.

• what type of measures arelikely to be most appropriateand effective at each location?

The conference shouldculminate in the identification ofkey individuals with sufficientinfluence and skill to direct theplanning around prioritylocations. A small core team willbe assigned to each prioritylocation with a mandate to:• form a multi-agency alliance

consisting of all keystakeholders in respect of theassigned location;

• investigate and assess therisks at the assigned location;

• assess the feasibility,acceptability, cost and likelyeffectiveness of differentinterventions, consulting allrelevant local interest groups;

• develop risk-management andoverall local arrangements;

• secure agreement on financialresponsibility and negotiatebudget where necessary.

It is recommended that atimescale is set for reportingback to the overall hotspotsprogramme lead.

2. The role of projectteams

Each priority location wouldneed to be managed as adiscrete project. The key tasks for each project team are outlined below.

2a Forming effectivealliances

Additional stakeholderengagement may need to benegotiated at this stage. A keytask for each group will be tosecure the commitment to theproject of agencies or individualswho own and manage thesite/structure and to co-optthose in positions of authority.Statutory and voluntary andbodies with responsibility forsuicide prevention in the areashould also be represented.

Each team will need to appointan executive sponsor, a clinicallead and a project lead. Theexecutive sponsor will need tobe a senior executive in theagency that owns the location.The clinical lead must be local tothe hotspot, with access rightsto local information and powerto influence decision makingwithin the health and social carecommunity. The project lead willco-ordinate the work and liaiseacross agencies.

MANAGING HOT SPOTS: REDUCINGRISK AND OPPORTUNITY FOR SUICIDE

Page 24: Guidance on action to be taken at suicide hotspots - SPRC

22

2b Assessing site-specific risks

Factors to consider whenassessing site-specific risksinclude:• size and remoteness of area

under consideration • existing rights of access and

barriers to access• proximity to and ease of

access from establishmentshousing high-risk populations(psychiatric in-patient units,prisons, probation hostels andhostels for the homeless)

• existing surveillancearrangements: CCTV, securitypatrols, likelihood of adistressed individual beingspotted by existing patrols orstaff working in the vicinity

• transport and communicationlinks; ease of access foremergency services.

Discussions with staff who work on or near the location and who may have witnessedsuicide attempts may supplyvaluable information. Furtherinterrogation of data fromcoroners' records may benecessary to ascertain the detailof suicidal acts (e.g. whetherone side of a bridge is morefavoured for jumping than theother). Where possible, times ofday, week, month or year shouldbe studied in order to identifypeak periods for suicide risk.

2 c Considering the options All available options should be considered. Teams will need to assess the feasibility,cost and likely effectiveness of a range of interventions, as wellas their acceptability to localinterest groups.

Public consultation at this stageis vital. At beauty spots, naturereserves, Sites of SpecialScientific Interest (SSSIs) andnational parks particular care willneed to be taken to limitenvironmental damage. Theimpact of any intervention bothon the local environment and onthe local economy will need tobe widely debated.

Expert economic help may needto be sought in order to carryout cost-benefit analyses.

2d Drawing up localarrangements

Each site-specific risk factorshould be addressed by a risk-management plan. This may consist of relatively simpleand low-cost measures, such as relocating staffed points(e.g. ticket booths) to improve visibility.

The risk-management plansshould be ranked in order ofpriority and together will form theoverall action plan for thelocation. The overall locationplan should include a timetablefor implementation, together withcontingency plans for non-completion of critical phases.

Media restrictions to coverreporting of the initiative itselfand any subsequent suicides atthe location should be includedas a matter of course.

2e Financial planning The team may wish to draw upa financial plan to support thevarious elements of the localarrangements. Economicconstraints may require the localarrangements to be phased;hence the need to prioritise risk-management plans (see above:2d Drawing up localarrangements).

Financial responsibility will needto be explored and agreementreached between the 'owner' ofthe location and other bodieswith responsibility for publichealth and safety. Budget willneed to be negotiated withinthose organisations.

2f ImplementationDepending on the size of theoverall local arrangements forthe location, the project leadmay take responsibility forimplementation. Alternatively, ifthe size of project demands it, adedicated project managershould be appointed.

The project manager should setout key events and timescales inthe form of a project plan, drawdown the agreed financialcontributions from partneragencies and commission thework. Completion of key stagesand any complications orsetbacks should be reportedback to the stakeholders.

Page 25: Guidance on action to be taken at suicide hotspots - SPRC

23

Responsibility for audit andevaluation will reside with theoverall hotspots programme lead,to whom site-specific projectteams will be accountable.

1. AuditThe programme lead shouldaudit all priority locations todetermine:a) whether the locally agreed

arrangements have beenimplemented;

b) whether it has beencompleted in accordancewith the agreed timetable.

Periodic follow-up may also benecessary in order to ensure thatinterventions are being sustainedor that agreed measures remainin place.

2. EvaluationIt is considered that amonitoring period of 3 years would be necessary in order to determine whetheror not the measures introducedat each site have led to areduction in numbers ofsuicidal acts, both at the targetsite and at similar sites nearby(e.g. other bridges) todetermine whether suicidalindividuals have simply gone elsewhere.

The contribution of each site-specific project to the overallhotspots programme shouldalso be evaluated throughongoing collection andinterrogation of countywide data.

Because the numbers involvedat any location are likely to bevery small, it may be difficult toidentify an effect with anycertainty. It may, however, bepossible, if mechanisms havebeen established for collecting'near miss' data (see above:Data collection 1d: Additionalsources of data), to show areduction in rates of fatalityand/or serious injury resultingfrom acts carried out at the site.

Even if no retrospective data areavailable, the agency or bodythat manages the locationshould be encouraged to recorddetails of all future 'near misses'.These can be analysed todetermine the factors thatoperated to prevent a fataloutcome. Survivors of 'nearmisses', witnesses and rescuers may be willing to beinterviewed. Such interviews will generate rich qualitative datathat will enable real learning totake place.

Local universities or NHSResearch and DevelopmentSupport Units (RDSUs) will beable to assist with evaluation.

3. Reporting backThe findings from each site andfrom the hotspots programmeas a whole should be reportedto the Inter-agency Forum forSelf Harm & Suicide in order toshare the learning with the wholesuicide prevention community.

AUDIT AND EVALUATION

Page 26: Guidance on action to be taken at suicide hotspots - SPRC

24

Examples of good practice of identifyingand managing suicide hotspotsSouth Devon Healthcare Trust has a long-standing arrangementwith the local coroner's office, which routinely forwards copies ofinitial police reports of potential suicides to the clinical audit team.A large database, the design of which is based on the NationalConfidential Inquiry into Suicide and Homicide by People withMental Illness, has been built up over the last 15 years. Thedatabase is regularly interrogated with a view to identifying localtrends and comparing these with the national picture.

This enabled 2 local hotspots to be identified. Findings werepresented at the locality audit meeting and taken up with the localcouncil and other relevant organisations. Safety measures weresubsequently introduced, including fencing in of all upper levels of aprominent multi-storey car park.

For further details, contact: Clinical Audit Team Leader, South DevonHealthcare NHS Trust. Tel: 01803 655770.

The Wessex Suicide Audit, based in the University ofSouthampton/Royal South Hants Hospital, began in 1988 andis still regularly updated using information from coroners' files.

This large database of all deaths that have been subject to aninquest is widely used as a resource by researchers, local NHSTrusts, Health Authorities and other organisations, such as thepolice and Samaritans.

Routine analysis revealed that car parks in the New Forest wereacting as a magnet for suicidal individuals, many of whom werevisitors to the area, and were being used for car exhaust poisonings.A multi-agency alliance was formed, comprising representatives of theForestry Commission, the Samaritans, the local health authority, thedistrict council and the university. Preventative measures wereintroduced, including strategic placing of signs displaying theSamaritans’ telephone number. The scheme was rigorously evaluatedand was shown to result in a significant and sustained reduction inthe number of car park suicides occurring in the New Forest district(King & Frost, 2005).

For further details, contact: University Department of Psychiatry,Royal South Hants Hospital, Southampton. Tel: 023 8082 5537.

Page 27: Guidance on action to be taken at suicide hotspots - SPRC

25

Amos T, Appleby L, Kiernan K (2001). Changes in rates ofsuicide by car exhaustasphyxiation in England andWales. Psychol Med, 31(5): 935-9.

Beautrais AL (2001).Effectiveness of barriers atsuicide jumping sites: a casestudy. Australian and NewZealand Journal of Psychiatry,35:557-562.

Beautrais AL, Fergusson DM,Horwood LJ (2006). Firearmslegislation and reductions infirearm-related suicide deathsin New Zealand. Australian andNew Zealand Journal ofPsychiatry; 40:253-259.

Berman, AL et al (1990). Suicideprevention in public places. In:AL Berman (Ed.), SuicidePrevention: Case Consultations.New York: Springer.

Cantor CH, Baume PJ (1998).Access to methods of suicide. Australian and NewZealand Journal of Psychiatry,32(1): 8-14.

Cantor CH, Slater PJ (1995).The impact of firearm controllegislation on suicide inQueensland: preliminaryfindings. Med J Aust, 162(11): 583-5.

Cooper PN, Milroy CM (1995).The coroner’s system andunder-reporting of suicide. Med Sci Law, 35(4): 319-26.

Department of Health (1999). Saving Lives: OurHealthier Nation. London:Stationery Office.

Department of Health (2002).National Suicide PreventionStrategy for England. London:Department of Health.

Department of Health (2006a).Our Health, Our Care, Our Say.London: Department of Health.

Department of Health (2006b).Best practice guidance on therole of Director of Adult SocialServices. London: Departmentof Health.

Gallagher M, Hares T, SpencerJ, Bradshaw C, Webb I (1993).The Nominal Group Technique:a Research Tool for GeneralPractice? Family Practice, 10(1): 76-81.

Glatt K M (1987). H e l p l i n e :Suicide Prevention at a SuicideS i t e . Suicide and Life Thre a t e n i n gB e h a v i o r, 17(4): 299-309.

Gunnell D, Middleton N, FrankelS (2000). Method availabilityand the prevention of suicide: are-analysis of secular trends inEngland and Wales 1950-1975.Soc Psychiatry PsychiatrEpidemiol, 35: 437-443

Gunnell D, Nowers M (1997).Suicide by jumping. ActaPsychiatr Scand, 96:1-6.

Hawton K, Appleby L, Platt S,Foster T, Cooper J, Malmberg A,Simkin S (1998). Thepsychological autopsyapproach to studying suicide: a review of methodologicalissues. Journal of AffectiveDisorders 50, 269-276.

Hawton K, Townsend E, DeeksJ, Appleby L, Gunnell D,Bennewith O, Cooper J (2001).Effects of legislation restrictingpack sizes of paracetamol andsalicylate on self poisoning inthe United Kingdom: beforeand after study. BMJ, 322:1203-7.

Hawton K, Simkin S, Deeks J,Cooper J, Johnston A, Waters K et al (2004). UK legislation onanalgesic packs: before andafter study of long term effecton poisonings. BMJ, 329:1076-1079.

Kerkhof A (2003). RailwaySuicide: Who is Responsible?Crisis, 24(2): 47-48.

King E, Frost N (2005). The NewForest Suicide PreventionInitiative (NFSPI). Crisis, 26(1): 25-33.

REFERENCES

Page 28: Guidance on action to be taken at suicide hotspots - SPRC

26

Kreitman N (1976). The coal gasstory. United Kingdom suiciderates, 1960-71. Br J Prev SocMed, 30(2): 86-93.

Lindqvist P, Jonsson A, ErikssonA, Hedelin A, Björnstig U (2004).Are suicides by jumping offbridges preventable? Ananalysis of 50 cases fromSweden. Accident Analysis andPrevention, 36: 691-694.

Office for National Statistics,2005. National Statistics Online(http://www.statistics.gov.uk/downloads/theme_health/Dh2_31/Table2.19.xls)

Pearson VAH (1993).Suicide in North and WestDevon: a comparative studyusing Coroner's inquestrecords. Journal of Public Health Medicine, 15(4): 320-326.

Pirkis J, Blood RW (2001).Suicide and the Media. Part 1:Reportage in Non-fictionalMedia. Crisis, 22(4): 146-154.

Reisch T, Michel K (2005).Securing a suicide hotspot:effects of a safety net at theBern Muenster Terrace. SuicideLife Threat Behav, 35(4):460-7.

Rosen DH (1975). SuicideSurvivors: A Follow-up Study ofPersons Who SurvivedJumping from the Golden Gateand San Francisco-OaklandBay Bridges. Western Journal ofMedicine, 122: 289-294.

Sampson HH, Rutty GN (1999).Under-reporting of suicide inSouth Yorkshire (West). J ClinForensic Med, 6(2): 72-6.

Simkin S, Hawton K, Sutton L,Gunnell D, Bennewith O, KapurN (2005). Co-proxamol andsuicide: preventing thecontinuing toll of overdosedeaths. QJM, 98(3): 159-70.

Sonneck G, Etzersdorfer E,Nagel-Kuess S (1994). Imitativesuicide on the Viennesesubway. Soc Sci Med, 38(3): 453-7.

Spicer RS, Miller TR (2004).Suicide acts in 8 states:Incidence and case fatalityrates by demographics andmethod. Am J Public Health, 90:1885-1891.

Yip P (2005). A Public HealthApproach to SuicidePrevention. Hong Kong JPsychiatry, 15:29-31.

Williams C, Miller J, Watson G,Hunt N (1994). A strategy fortrauma debriefing after railwaysuicides. Soc Sci Med, 38(3):483-7.

Page 29: Guidance on action to be taken at suicide hotspots - SPRC

27

APPENDIX 1:Resourcesassociated withprogramme General administrative support:4 hours per week @ NHS Band 3-4

Initial data collection: 2-3weeks @ NHS Band 5

Cleaning & formatting data: 1-2weeks @ NHS Band 5

Initial data analysis: 2-3 weeks@ NHS Band 7-8

Software (GIS)

Regular updating andinterrogation of database: 5days per 6 months @ NHS Band7-8

Travel costs: visits to coronersand records offices

Stakeholder conference:• Additional administration: 1

week full-time either side ofevent, plus 80 hours overpreceding 3-4 months @ NHSBand 3-4

• Conference venue: £1,000• Conference catering: £720

(£12 per delegate x 60)

Programme management(overseeing, planning andaudit): 1 day per week @ NHSBand 7-8

Travel costs associated withsite visits

Evaluation (services ofuniversity or RDSU)

APPENDIX 2:List of potentialstakeholders andpartner agenciesThis list gives suggestions foragencies that may need to beinvolved in the local consultationp rocess. It is not exhaustive and thekey stakeholders for any local are awill be influenced by its geography.

HM Coroners

NHS Direct• Mental Health Site Lead

General Practitioners with Special Interest in Mental Health

Primary Care Trusts• Chief Executive• Director of Public Health• Standard Seven Lead• Commissioner for Mental Health

Mental Health and LearningDisability Trust• Chief Executive• Standard Seven Lead • Liaison Psychiatry or Self-

Harm Team • Crisis Resolution Team• Community Mental Health Te a m s• Drug and Alcohol Team

Local Implementation Groups(LIGs) and Teams (LITs)

Children's Trust: CAMHS Self-Harm Lead

Social Services: Director ofAdult Social Services

Drug Action Teams

Samaritans

Mind

Other local voluntary agencies

Mental health service user andcarer groups

CSIP/NIMHE RegionalDevelopment Consultants

Regional Government Office:Public Health Consultant

Police• Chief Inspector or above• Mental Health Liaison Officer• Ambulance Service

Fire and Rescue Service

Prison (if applicable)• Prison Health Care Manager• Suicide Prevention Lead

Probation Service

British Transport Police: AreaCoroners' Liaison Officer

County Council • County Surveyor• Highways Department

Town/City Council• Car Parking Department

Parish Councils

Network Rail Area GeneralManager

Highways Agency Area GeneralManager

Bridge authorities (ifapplicable): Bridge Manager

Coastguard Service

Forestry Commission

National Trust

National Parks Authority

English Nature/CountrysideAgency/Rural DevelopmentService

Churches and faithcommunities

Local Media• Television• Radio• newspapers

Page 30: Guidance on action to be taken at suicide hotspots - SPRC

28

APPENDIX 3:List of variables toinclude

Coroner

Full name

Date of Birth

Date of Death

Age

Age group: under 25; 25-34;35-44; 45-54; 55-64; 65-74;75+

Sex: male; female

Ethnicity/first language (willinform decisions as to whethersigns need to be provided inminority languages)

Home address (for purposes ofascertaining whether suicidetook place at home)

Home postcode

Resident in county: yes; no

Verdict: suicide; open

Method of suicide: • drug-related poisoning • other poisoning including

motor gas • hanging/strangulation/

suffocation • jumping from a high place • jumping/lying in front of a

moving object • drowning • cutting or stabbing • firearms • burning • other

Location of act (Place nameand as much narrative detail aspossible to enable preciselocation to be pinpointed onOrdnance Survey map)

Postcode of location (This willbe needed for mapping usingGIS software. It can looked upand entered later, but adatabase field will be required)

Status of location: private;public*

Location type: bridge; building;cliff; road; rail; rural car park orlayby; other

Date of act (if clearly differentfrom date of death)

Time of act (if ascertainable)

* Suggested definitions

Private location: any privateaddress, including a farmer'sown land, but excluding hotelsand guest houses unless thepermanent residence of thedeceased. Psychiatric in-patientunits, prisons, hostels and carehomes in which the individualwas living/being cared for attime of death.

Public location: all open landnot owned by the deceasedindividual, transport networks,public buildings, and hotels andguest houses in which thedeceased was a temporaryresident.

Page 31: Guidance on action to be taken at suicide hotspots - SPRC

APPENDIX 4:Example of localsuicide data mappedusing GIS software

N.B. The locations shown arebased on fictitious datagenerated for demonstrationpurposes only.

29

Legend20002001200220032004

Legend20002001200220032004

Page 32: Guidance on action to be taken at suicide hotspots - SPRC

APPENDIX 5:Stakeholderconference:pre-conferencequestionnaire andoutline of consensusmethod for use inpriority setting

30

Inter-agency Forum on Self-Harm & Suicide SuicideHotspots Conference

Pre-conference questionnaire

We are defining a suicide hotspot as a specific, public place that isf requently used as a location for suicide and which provides either meansor opportunity for suicide. An example would be a bridge from whichindividuals have jumped to their deaths on more than one occasion.

Are there any locations in that you regard as suicide hotspots? If so,please list them below, together with your reason for including eachone (i.e. relevant personal or professional experience). Please bringthe completed sheet with you to the conference.

(i)

(ii)

(iii)

(iv)

(v)

Page 33: Guidance on action to be taken at suicide hotspots - SPRC

31

Workshop timetable

2.00 - 2.30 Presentation of local data

Formation of 6-8 groups, with appointed facilitators (chairpersons)

2.30 - 2.45

2.45 - 3.15

3.15 - 3.45

3.45 - 3.50

3.50 - 4.10

4.10 - 4.30

Total: 15 mins

20 mins

10 mins

Total: 30 mins

15 mins

15 mins

5 mins

10 mins

10 mins

Total: 20 mins

20 mins

Personal stories

Please introduce yourselves to each other andtake a few minutes to share stories, from yourown professional experience, of suicides thathave occurred in public places in [county].

Does _____________ have any hotspots and, if so, where are they?

Using our data and your own local knowledge,and working as a group, please compile a list ofall locations in that you consider to be hotspots.

N.B. Our maps show completed suicides only.Please think about and include locations of anyserious suicide attempts or near misses.

Chairs to feed back results

Scribe to compile complete list

Which, if any, should we prioritise?

Stage 1 (chairs take teabreak)As a group, please select your top 5 hotspotsfrom the complete list

Stage 2 (group members take teabreak)Chairs get together to pool the groups' 'top 5s'and decide overall priorities

Overall prioritiesConference chair to present the decision onoverall priority sites

Risk factors and risk-management strategiesFor each of the final priority locations, pleaseconsider in your groups:a) what makes it attractive to suicidal individuals; b) how might site-specific risks be addressed?

Chairs to feed back

Formation of planning groups and round-up of day

Page 34: Guidance on action to be taken at suicide hotspots - SPRC

APPENDIX 6:Useful websites NIMHE Primary Care SuicideAudit Toolhttp://www.eastmidlands.csip.org.uk/suicide_db/index.html

Coroners' Societyh t t p : / / w w w. c o r o n e r. o r g . u k / p u b l i c /search.asp

Ordnance Survey NHShelpdesk, for help with GIS andmapping: tel: 0845 458 0650 or e: [email protected]

Office for National Statistics(ONS)http://www.statistics.gov.uk/

Samaritans http://www.samaritans.org/

Rail Safety and StandardsBoardhttp://www.rssb.co.uk/

Office of Rail Regulation.Fatalities database:http://www.rail-reg.gov.uk/server/show/ConWebDoc.8070

Highways agency http://www.highways.gov.uk/

Institution of Civil Engineershttp://www.ice.org.uk/homepage/index.asp

County Surveyors' Societyhttp://www.cssnet.org.uk/

Press Complaints CommissionCode of Practicehttp://www.pcc.org.uk/cop/practice.html

32

Page 35: Guidance on action to be taken at suicide hotspots - SPRC
Page 36: Guidance on action to be taken at suicide hotspots - SPRC