Article summary

Accident Analysis and Prevention 37 (2005) 625–632

Suicide prevention through means restriction: Assessing the risk of substitution

A critical review and synthesis

Marc S. Daigle∗

Centre for Research and Intervention on Suicide and Euthanasia and Philippe Pinel Institute, University of Qu´ebec at Trois-Rivi`eres, P.O. Box 500, Trois-Rivi`eres, Que., Canada G9A 5H7

Received 30 December 2004; accepted 13 March 2005

Abstract

The effectiveness of restricting access to certain means of committing suicide has been demonstrated, at least as regards toxic domestic gas, firearms, drugs and bridges. At the individual level, studies tend to indicate that many persons have a preference for a given means, which would limit the possibility of substitution or displacement towards another method. Similarly, the fact that suicidal crisis are very often s iven means w opulation s ver a longer t ©

K

1

s K H f a t v t s s c C s

tegy s, s ev- 2003 e- t the o cer-

en- - l., tion

vid- tting eth- e of nd

0 d

hort-lived (and, what is more, influenced by ambivalence or impulsiveness) suggests that an individual with restricted access to a g ould not put off his plans to later or turn to alternative methods. This has been more difficult to demonstrate scientifically in p tudies. Nevertheless, it appears that, should such a shift occur towards other means, it would be put into effect only in part and o erm.

2005 Elsevier Ltd. All rights reserved.

eywords: Suicide; Means; Restriction; Substitution; Displacement

. Introduction

The field of suicide prevention was marked early on by uch characters as the Reverend Chad Varah in the United ingdom and Dr. Edwin Shneidman in the United States. owever, these often cited men, as well as all the volunteers

rom Samaritans and Suicide Prevention Centres after them nd all the professionals of the mental health field, contributed

o give an essentially clinical or individual character to inter- entions in this domain. As early as in 1968,Bagley (1968) urned his focus to measuring the effect of hotline services on uicide rates even though, contemporaneously, other factors uch as the detoxification (reduction of the carbon monoxide ontent) of domestic gas could be at play (Kreitman, 1976). linicians have thus tended to overlook broad sections of uicide prevention, namely those centered on environmental

∗ Tel.: +1 819 376 5011; fax: +1 819 376 5195. E-mail address:marc [email protected].

approaches even if it could be the most promising stra for the future (Johansson et al., 2005). Among these, mean restriction has been drawing more and more attention, a idenced by the fact that nine of the presentations at the congress of theInternational Association for Suicide Pr ventionreferred to this measure, compared with none a 1988 congress. The effectiveness of restricting access t tain means has been demonstrated (Leenaars et al., 2000), at least as regards toxic domestic gas (Kreitman, 1976), firearms (Bridges and Kunselman, 2004; Conwell et al., 2002; Hem way and Miller, 2002), drugs (Hawton et al., 2001; Wilkin son et al., 2002) and bridges (Beautrais, 2001; Berman et a 1994). However, there remains the question of substitu or displacement, that is, the possibility that suicidal indi uals, once deprived of their preferred means of commi suicide, shift their consideration towards alternative m ods. The purpose of this article is to examine the issu means substitution from the point of view of individuals a populations.

001-4575/$ – see front matter © 2005 Elsevier Ltd. All rights reserved.

oi:10.1016/j.aap.2005.03.004

626 M.S. Daigle / Accident Analysis and Prevention 37 (2005) 625–632

2. Methods

Printed documentation, in English or French, was searched in the following electronic databases: MedLine and PsychInfo. Two specialized documentation centers, both lo- cated in Canada, were also consulted: the Centre for Research and Intervention on Suicide and Euthanasia (Montreal) and the Suicide Information and Education Center (Calgary). The keywords “displacement” or “substitution” not being very productive, the following search strategy was also used: sui- cide AND (epidemiology OR rate OR rates OR trends OR “statistics & numerical data” OR incidence) AND (“cause of death” OR “prevention & control” OR mortality OR method OR methods).

3. At the individual level

In this regard, two types of studies can be distinguished, namely those that have focused on a preference for a par- ticular method of committing suicide and those that have examined the fact that suicide crises tend to be short-lived.

3.1. Preference for a particular means of committing suicide

. For i on, a m rob- a easy t ing. P acter i f b ess ( c e the d d be p s put t ly no w , a t t ng- l and h using o

ts in- d pain- l oth w con- c ger o to s ard, i that t ecif- i ople c e

than from the nearbySan Francisco-Oakland Bay Bridge (Seiden and Spence, 1983), the factors mentioned included, in particular, the more or less morbid and contagious public- ity that surrounded cases at theGolden Gate. The symbolism (“suicide shrine”) and romanticism associated with the site would also explain why many suicidal persons chose it, even though the other bridge was sometimes definitely closer to their homes.

A certain copycat and contagion effect has been observed in particular when the media put out information in more or less adequate fashion regarding suicides already perpetrated in a given place. It is believed that this type of coverage pushes certain individuals, especially the most vulnerable, to adopt a given method of suicide rather than another. This contagion phenomenon has been particularly well documented over the years, beginning with the example of Werther, the charac- ter created by 18th century German writer Goethe, up until more recent cases, such as those in the Vienna subway and at Mount Mihara in Japan (Clarke and Lester, 1989; Ellis, 1996; Gunnell and Nowers, 1997).

Certain personal and cultural factors are also at play. De- pending on the individuals concerned, certain forms of death seem more inculpating or exculpating for the bereaved or more terrifying, more painful or more shameful. Persons who are more ambivalent or who could eventually manip- ulate the people around them might also prefer a less lethal m le c hoice s lity; f ng- i ills); l igh b ty); d dan- g essi- n ved o me); s ance ( (per- c ons, w ore, s ans- i t.

tries a stric- t ural a ther ( 92 T ada, A can s one c eath c s, N e by fi uch l mon,

Suicide is not a homogeneous category of behaviour nstance, regarding suicide from domestic gas intoxicati

ethod once widely used in England, suicidal persons p bly considered this a highly lethal, painless, clean and

o use method that required little “courage” and plann robably, no other means of suicide shares these char

stics (Clarke and Lester, 1989). Individuals who jump of ridges, for their part, often suffer from severe mental illn see many studies in the summary list ofTable 1). In such ases, auditory hallucinations, for example, can influenc ecision to commit suicide in a specific manner. It shoul ointed out that among the suicides by all other mean

ogether, the percentage of psychotic individuals is usual ay as high as in this group (Cantor et al., 1989). However t least one study byNowers and Gunnell (1996)found tha

hose jumping from the Clifton Suspension Bridge in E and were no more likely to have psychiatric histories ave been psychiatric inpatients than matched controls ther suicide methods.

In response to a questionnaire, female college studen icated their preference for methods deemed quick and

ess and entailing no risk of disfigurement. However, b omen and men would use a firearm as they were more erned with getting it over with quickly than with the dan f disfigurement (Lester, 1988). This type of study tends how that, notwithstanding gender differences in this reg ndividuals have a preference for a specific means and hey would not easily resolve to change method. More sp cally, in a study that sought to understand why more pe ommitted suicide by jumping off theGolden Gate Bridg

-

ethod (Clarke and Lester, 1989). To understand how peop hoose a particular method, these authors suggest 20 “c tructuring properties of methods of suicide”: availabi amiliarity with the method; technical skills needed (ha ng, gasing); planning necessary (buy a gun, save up p ikely pain (cutting wrists); so-called “courage” needed (h uilding, train); consequence of failure (disability, publici isfigurement after death (hanging versus overdose); er/inconvenience to others (car crash, subway leap); m ess, bloodiness (wrist cutting); discovery of body (by lo nes or strangers); contamination of nest (i.e., avoid ho cope for concealing or publicizing death—shame, insur car crash, drowning, subway leap); certainty of death eived/actual); time taken to die while conscious (pois rist cutting); scope for second thoughts (swim back to sh witch off gas); chances of intervention; symbolism (cle ng by fire); masculine/feminine (guns); dramatic impac

Also, the methods used vary greatly across the coun nd even within the countries; consequently, a means re

ion approach that might work in under one set of cult nd environmental conditions might not work under ano Cantor, 2000; Cheng and Lee, 2000; Marzuk et al., 19). able 2 shows some comparisons between USA, Can ustralia and England but also gender differences. One ee that data are not always reported the same way from ountry to the other despite the now growing use in the d ertificates of theInternational Classification of Disease inth Revision. Nevertheless, that table shows that suicid rearm is very common in USA for both the sexes but m ess in the other countries where hanging is most com

M.S. Daigle / Accident Analysis and Prevention 37 (2005) 625–632 627

Table 1 Summary of included studies on risk of substitution

Focus Authors Location Population/samplea Key findings

Individual level Preferences (vs. mental illness)Prasad and Lloyd (1983) Scotland 18 survivors jumping from height All have a mental illness

Kontaxakis et al. (1988) Greece 46 survivors jumping from height and 214 from drug overdose

More major psychopathologies in the first group

Cantor et al. (1989) Australia 40 survivors and 47 suicides from bridge Most suffer from severe mental illness Pommereau et al. (1989) France 110 survivors jumping from height and

908 from drug overdose More psychiatric history in first group

Nowers and Gunnell (1996) England 127 suicides from bridge and 127 other suicides

Same number of individuals with mental illness

De Moore and Robertson (1999)

Australia 31 survivors jumping from height and 51 from using firearms

More psychiatric history in first group

Coman et al. (2000) Australia 7 survivors and 55 suicides from bridge Most suffer from severe mental illness Beautrais (2001) Australia 15 suicides from bridges Most suffer from severe mental illness Lindqvist et al. (2004) Sweden 50 suicides from bridge Most had psychiatric problems

Preferences (other) Rosen (1975) USA 7 survivors from 2 bridges Preference for one out of two available bridges Seiden and Spence (1983) USA 555 suicides from one bridge and 112

from the nearby other Publicity, symbolism, romanticism influence the choice of the bridge

Lester (1988) USA 429 college students Men and women choose different methods Cantor and Hill (1990) Australia 73 suicides from 2 bridges 2 different profiles of suicides Kaplan and Geling (1999) USA Total population Regional cultural factors influence choice Cheng and Lee (2000) Asia and the Far

East Total populations Methods vary across and within countries and

across genders Hawton et al. (2004a,b) UK 428 survivors from self-cutting and

11 065 from self-poisoning More men and previous deliberate self-harm in self-cutters

seeTable 2 USA, Canada, Australia, England

Total populations Methods vary across countries and across genders

Short-lived crises Seiden (1978) USA 515 survivors bridges and 184 survivors from other types of attempts

Long term follow-up shows 90% do not die later on from violent death

Glatt (1987) USA 30 survivors from bridge Ambivalence about death Hawton et al. (2001) UK Total population Impulsive or ambivalent people do not accu-

mulate lethal drugs Lindqvist et al. (2004) Sweden 50 suicides from bridges Peak of alcohol-related suicides during

weekends Suominen et al. (2004) Finland 100 survivors from self-poisoning 87% will not commit suicide after 37 years Sakinofsky (2000)(review) Many Survivors from suicide attempts 89% will not commit suicide on the long term Owens et al. (2002)(review) Many Survivors from suicide attempts 93% will not commit suicide on the long term

Population level Toxic gas Kreitman (1976) UK Total population Almost no substitution

Burvill (1980) Australia Total population Substitution for men only Lester and Abe (1989) Japan Total population Almost no substitution Lester (1990) USA Total population Almost no substitution

Firearms Carrington and Moyer (1994) Canada Total population No substitution Lester (1995) USA Total population No substitution Leenaars and Lester (1996) Canada Total population Substitution for men only Kaplan and Geling (1998) USA Total population Substitution for men only Conner and Zhong (2003) USA Total population No substitution De Leo et al. (2003) Australia Male population No substitution Bridges (2004) Canada Total population Substitution Caron (2004) Canada Total population Substitution Clarke and Lester (1989)(re- view)

All Total populations Mixed results

Lambert and Silva (1998)(re- view)

All Total populations No substitution

Drugs Oliver and Hetzel (1972) Australia Total population Almost no substitution Whitlock (1975) Australia Total population Almost no substitution Yamasawa et al. (1980) Japan Total population Almost no substitution Wilkinson et al. (2002) England Total population Substitution Hawton et al. (2004a,b) UK Total population Some substitution Nordentoft et al. (2004) Denmark Total population No substitution

Bridges Berman et al. (1994) USA Total population No substitution Beautrais (2001) Australia Suicides from height No substitution

a The survivors samples include attempted suicides and failed suicides.

628 M.S. Daigle / Accident Analysis and Prevention 37 (2005) 625–632

Table 2 Percentage of utilization of different methods of suicide in four countries

Method Males Females

USA (2000)a

Canada (1998)b

Australia (1998)c

England (2003)d

USA (2000)a

Canada (1998)b

Australia (1998)c

England (2003)d

Firearms 61.2 26.2 10.1 N/Re 37.2 6.6 3.2 N/Re

Hanging, strangulation, suffocation 20 40 47.1 46.1 16.7 33.9 32.3 25.6 Solid and liquid poisoning 7.1 9.6 8.3 18.5 30.6 32.4 29.6 43.9 Gas poisoning (motor vehicle exhaust and other) 4.8 12.4 21 8.4 5.5 9 14.4 3.1 Jumping from height 1.8 3.9 N/Rf 3.2 3.1 5.8 N/Rf 4.6 Cutting and piercing N/Rg 1.6 1.8 N/Rg N/Rg 1.4 1.9 N/Rg

All other methods 5.1 6.2 11.8 23.8 7 7 18.6 22.8 a McIntosh (2003). b Langlois and Morrison (2002)(Population aged 10 or older). c Steenkamp and Harrison (2000). d National Institute for Mental Health in England (2005). e Not reported for England. Included in “all other methods”. f Not reported for Australia. Included in “all other methods”. g Not reported for USA and England. Included in “all other methods”.

at least for men. As for applying means restriction to this last method, this represents a real challenge except in hospital- ized or incarcerated populations.Dahlberg et al. (2004)also report that, in their American sample, 76.3% of all suicides occurred in the home, which is also a challenge for suicide prevention, except if intervention takes place before the suici- dal crisis, in controlling the purchase or ownership of means of suicide like firearms.

Regarding more particularly the symbolism associated with suicide from bridges, there is no doubt that the spec- tacular aspect of the deed is a major factor. Survivors have indicated that not only had they planned to commit suicide by jumping off a bridge but also off a very specific bridge. They generally referred to an association that they had made between death and the particular “beauty/grace” of the given bridge (Rosen, 1975, 1976). Moreover, all this symbolism seems to rest at least in part on a popular myth to the effect that committing suicide by jumping off a bridge is easy to do and hassle free and that the person in question gently dis- appears under the surface (Seiden, 1978). Be this as it may, studies as a whole indicate that many individuals prefer a particular means of suicide, if not a particular place for it (De Moore and Robertson, 1999). In Australia, researchers found major differences between the persons who committed suicide at two Australian bridges (Cantor and Hill, 1990). It ensues from this that these individuals would not readily go t ther p

3

ort d they w the m eir p an- o ents

of transitory impulsiveness have been found to be signifi- cant in suicidal behaviour, especially among young people and even more specifically in suicides committed by jump- ing from high places (Pommereau et al., 1989). At the same time, it is important to acknowledge that ambivalence is typ- ical of the suicidal process (Centers for Disease Control and Prevention, 1992), which is precisely what has been observed in suicidal persons who contact crisis lines using telephones installed on bridges (Glatt, 1987). The study byHawton et al. (2001), showing that restricting pack sizes of drugs lower suicide rates, could also be explained by the fact that impul- sive or ambivalent suicidal people do not accumulate their pills and that they change their mind in the mean time.

Immediate help is required just after the attempt. In fact, the risk of repeating a suicide attempt is highest in the pe- riod immediately following the first suicidal act. Longitudinal studies as a whole have demonstrated that, after a year, from 1 to 3% only of these individuals will in the end die from suicide. After 5 years, this percentage reaches 9% and in longer studies, 11% (Sakinofsky, 2000). This means, there- fore, that of the individuals prevented from using their method of choice, 89% will not commit suicide by some other means on the long term. The percentage of persons saved can even be higher if effective help is provided at the same time. Over- all, although it is true that the suicide risk is about 100 times higher for individuals with a history of suicidal behaviour, t all (

4

cific m hort- l ould n y the s other

hrough with their intentions using other means or in o laces.

.2. Short-lived crises

It is recognized that suicidal crisis are usually of sh uration and that, if their fatal outcome is prevented, ill not be repeated or help will be made available in eantime. At worst, suicidal individuals deprived of th referred means of committing suicide will fall back on ther, possibly less lethal, method. In this regard, elem

he majority of these, then, will not die from suicide after Seiden, 1978; Owens et al., 2002).

. Population studies

If individuals seem to have a preference for a spe eans of suicide and if they essentially experience s

ived crises, restricting access to a specific method sh ot bring about an increase in the use of other means. B ame token, overall suicide rates should decrease, all

M.S. Daigle / Accident Analysis and Prevention 37 (2005) 625–632 629

things remaining the same. This is what population studies should confirm. However, it is evident that the direct effects of restricting access to means of suicide are hard to observe, as is an eventual shift towards other means, when the method tar- geted is already seldom used in the population (seeTable 2). At the statistical level, proving a phenomenon whatsoever is rendered much more difficult. This is true for suicides, com- pleted or not, committed from bridges and for an eventual displacement towards other methods (Gunnell and Nowers, 1997). This is why it would be interesting also to look at what has been reported about means that are much more com- monly used and relatively easier to restrict like firearms. Con- sequently, we will examine the issue of displacement when access is restricted to the following four means of suicide: domestic toxic gas, firearms, drugs and bridges.

4.1. Domestic toxic gas

It has been demonstrated that the detoxification of domes- tic gas in Great Britain brought about a considerable reduc- tion in both the number of suicides by this method and the overall suicide rate (down by one-third) (Clarke and Lester, 1989). Kreitman (1976)was among the first to make the con- nection between the decline in overall suicide rates and the detoxification of domestic gas in the United Kingdom. As domestic gas poisoning had always been the preferred means o in t t on t ent t ved in J , 1 not b thods o imply r ce- m ards i be- c L f d en.

4

ss to fi the U is m n s ntrol ( in- v re- s pted t ations b rates a eans u a ce t ct to

men (Leenaars and Lester, 1996). However, the methodology of this study has been contested rather convincingly by the first team (Carrington and Moyer, 1994; Carrington, 1999). In fact,Leenaars (2001, p. 260)himself concluded that “people did not switch to other methods of suicide”. Two more recent Canadian studies (Bridges, 2004; Caron, 2004) analyzed the effect of another Canadian law (Bill C-17) enforced in 1992 on the safe storage of firearms. In a specific region of northern Quebec, Caron compared suicide methods 6 years prior to the enactment of the law and 5 years after. Suicide by firearms decreased for both sexes but not among men over 45 who may be the owners of the key giving access to the firearms. In his study, hanging increased among youths and suicide by poisoning doubled in the female population. Caron concludes that his findings support the substitution hypothesis despite the unexplained high rise in hanging method.Bridges (2004) reaches similar conclusion for the total Canadian population. De Leo et al. (2003)were particularly interested in such an apparent displacement from firearms towards hanging in Aus- tralia. Rates of suicide by hanging were found to have begun increasing prior to the decline in firearm suicide in Australia. They concluded that individual suicide method choice may be related to independent changes in the social acceptability of each method as well as to an increasing prevalence of sui- cide in younger males who are more likely to use the hanging method. It seems that such a method is no more associated to s

4

effect o L - i lated t eth- o of B mon- s s by p this m such a ., 1 t a eta- m n the m en o ths. N the s in a rease i

4

E .67 s such

f suicide in this country (49.8% in 1958), a reduction he gas’s lethality is believed to have had a direct effec he overall suicide rates, without giving rise to displacem owards other means. The same effect has been obser apan (Lester and Abe, 1989) and the United States (Lester 990). The gradual detoxification of domestic gas did ring about a genuine displacement towards other me f suicide. Thousands of lives have thus been saved by s educing the toxicity level of domestic gas. If any displa ent occurred in these studies, it was very weak and tow

ntoxication by motor vehicle exhaust fumes which had ome much more easily available in the meantime (Clarke and ester, 1989). In Australia,Burvill (1980) found this type o isplacement towards car exhaust fumes but only for m

.2. Firearms

Results have been mixed regarding restricted acce rearms, at least as far as the very specific situation in nited States (Clarke and Lester, 1989) but a recent review ore optimistic (Lambert and Silva, 1998). A large Canadia

tudy reached such conclusions in favour of firearms co Carrington and Moyer, 1994). In this case, researchers estigated the effect of Bill C-51 which imposed certain trictions on firearms possession as of 1978. Their interru ime series statistical analyses revealed positive associ etween restrictions on firearms and reduction of suicide nd found no evidence of displacement regarding the m sed in Canada. Other researchers, instead, have noted

ain displacement in Canada, although only with respe

r-

hame in new generations.

.3. Drugs and other toxic substances

Numerous studies have shown that no displacement ccurs when access to lethal drugs is restricted (Clarke and ester, 1989). Oliver and Hetzel (1972)showed that, follow

ng control on sales of sedatives in Australia, suicides re o such a means decreased with no substitution of other m ds.Whitlock (1975)did the same in the specific region risbane. In this regard, a study conducted in Japan de trated that since 1956, the availability of certain drug rescription only had diminished the rate of suicide by ethod without displacement towards other intoxicants s chemical farming products or cyanide (Yamasawa et al 980). More recently,Hawton et al. (2004a,b)showed tha legislation restricting pack sizes of drugs like parac ol and salicylates had substantial beneficial effects o ortality and morbidity in UK; in the mean time, ibuprof

verdoses increased, but with little or no effect on dea evertheless, looking only at hospital admissions for ame period,Wilkinson et al. (2002)found also a decrease dmissions following paracetamol overdoses but an inc

n other types of admissions.

.4. Bridges

The installation of an anti-suicide fence on theDuke llington Bridgein Washington, DC, where an average 3 uicides a year were committed, resulted in only one

630 M.S. Daigle / Accident Analysis and Prevention 37 (2005) 625–632

death in the following 5 years. Meanwhile, the average num- ber of suicides remained essentially the same at a less popular bridge nearby with no such fence (1.69 before and 2 after). Over the same period of time, the overall number of suicides in Washington, DC, declined from 76.4 to 71.6, thereby indi- cating that no displacement had occurred towards other meth- ods (Lester, 1993; Berman et al., 1994). This last paper being a “case consultation” between three specialists, one should otherwise note that the second author, O’Carroll, is less con- vinced, pointing to the fact that this was not a real controlled study. Actually, most studies reported on means restriction are ecological ones. He argues that only interrupted time se- ries analyses, if really specific data could be made available, would give us the right answer about displacement in this case. These analyses would then take into account the over- all and specific trends in suicide rates and show if real changes appear when a specific intervention is implemented.

In an Australian study examining the effects of the re- moval of an anti-suicide fence from an overpass often used by suicidal psychiatric patients in large part suffering from schizophrenia, the number of suicides was shown to have in- creased from 3 in 1992–1995 to 15 in 1997–2000, while the total number of suicides in the region from jumping off high places remained stable. At first glance, this seemed to indi- cate that displacement occurred towards the overpass that was once again accessible, all the more so that, over this period, f tes. H files o rom o ped f enia c igh p ly no d rsons w e, a d (for o as a r in- t , like s ven- t t

t own t t this s all h ced c h no s

5

dis- p L ,

1996). At the individual level, studies tend to demonstrate that many suicidal persons have a preference for a specific method. Similarly, the fact that suicidal crises are very often short-lived (and, what is more, influenced by ambivalence or impulsiveness) suggests that a given individual who has lim- ited access to a given means will not put off his intentions to later or shift to another method. However, where population studies are concerned, scientific proof to this effect is some- times harder to establish and most studies are ecological. This notwithstanding, means restriction can in the meantime save lives and buy time for interventions of a more clinical nature.

Acknowledgements

Marc Daigle acted as a consulting expert for the engineer- ing firm Tecsult Inc. of Montreal as part of Contract 3-00-02- 17—Jacques Cartier Bridge, Consultant Services, Feasibility Study to Erect an Anti-Jumping Barrier (2002) awarded by The Federal Bridge Corporation Limited of Ottawa (Canada).

References

Bagley, C.R., 1968. The evaluation of a suicide prevention scheme by an

B sites:

B tion. riers.

B cide

B guns s 98,

B lia,

C van and

C Z.

C ted iatry

C adian es. 8,

C de in

C ada: v. 3,

C reven- uman

C t. In: book , pp.

ewer suicides were committed by jumping off other si owever, a more in-depth analysis revealed that the pro f the persons who died from suicide were very different f ne site to another. Indeed, 78% of the persons who jum

rom the overpass in question suffered from schizophr ompared with just 21% of those who jumped off other h laces. This would indicate, then, that there was actual isplacement, but rather an increase in suicides (by pe ith schizophrenia) off the overpass and at the same tim ecrease in suicides from jumping off other high places ther reasons) (Beautrais, 2001). As seen above, this w gain the case of a very specific site holding a particula

erest for certain persons. Note also that bridge barriers ome control on firearms, may have another injury pre ive effect on homicides (Berman et al., 1994) but also on raumatization of surrounding people.

On a similar topic,Reisch and Michel (2004)reported hat, after the installation of a protective net at a well kn errace in Berne (Switzerland), no suicides occurred a ite. Furthermore, the number of people jumping from igh places (including from bridges) in this city was redu ompared to the years before the installation, althoug afety means were installed at the other sites.

. Conclusion

Generally speaking, then, the risk of substitution or lacement towards other methods seems small (Clarke and ester, 1989; Gunnell and Nowers, 1997; Prévost et al.

ecological method. Soc. Sci. Med. 2, 1–14. eautrais, A.L., 2001. Effectiveness of barriers at suicide jumping

a case study. Aust. N. Z. J. Psychiatry 35, 557–562. erman, A.L., O’Carroll, P.W., Silverman, M.M., 1994. Case consulta

Community suicide prevention: the effectiveness of bridge bar Suicide Life Threat. Behav. 24, 89–99.

ridges, F.S., 2004. Gun control law (Bill C-17), suicide, and homi in Canada. Psychol. Rep. 94, 819–826.

ridges, F.S., Kunselman, J.C., 2004. Gun availability and use of for suicide, homicide, and murder in Canada. Percept. Mot. Skill 594–598.

urvill, P.-W., 1980. Changing patterns of suicide in Austra 1910–1977. Acta Psychiatr. Scand. 62 (3), 258–268.

antor, C.H., 2000. Suicide in the western world. In: Hawton, K., Heeringen, K.P. (Eds.), The International Handbook of Suicide Attempted Suicide. Wiley, Chichester, England, pp. 9–28.

antor, C.H., Hill, M.A., 1990. Suicide from river bridges. Aust. N. J. Psychiatry 24, 377–380.

antor, C.H., Hill, M.A., McLachlan, E.K., 1989. Suicide and rela behaviour from river bridges—a clinical perspective. Br. J. Psych 155, 829–835.

aron, J., 2004. Gun control and suicide: possible impact of Can legislation to ensure safe storage of firearms. Arch. Suicide R 361–374.

arrington, P.J., 1999. Gender, gun control, suicide and homici Canada. Arch. Suicide Res. 5, 71–75.

arrington, P.J., Moyer, S., 1994. Gun availability and suicide in Can testing the displacement hypothesis. Stud. Crime Crime Pre 168–178.

enters for Disease Control and Prevention, 1992. Youth Suicide P tion Programs: A Resource Guide. Department of Health and H Service, Centers for Disease Control and Prevention, Atlanta.

heng, A.T.A., Lee, C.-S., 2000. Suicide in Asia and the Far Eas Hawton, K., van Heeringen, K.P. (Eds.), The International Hand of Suicide and Attempted Suicide. Wiley, Chichester, England 29–48.

M.S. Daigle / Accident Analysis and Prevention 37 (2005) 625–632 631

Clarke, R.V.G., Lester, D., 1989. Suicide: Closing the Exits. Springer Verlag, New York.

Coman, M., Meyer, A.D., Cameron, P.A., 2000. Jumping from the West- gate Bridge, Melbourne. Med. J. Aust. 172, 67–69.

Conner, K.R., Zhong, Y., 2003. State firearm laws and rates of suicide in men and women. Am. J. Prev. Med. 25, 320–324.

Conwell, Y., Duberstein, P.R., Connor, K., Eberly, S., Cox, C., Caine, E.D., 2002. Access to firearms and risk for suicide in middle-aged and older adults. Am. J. Geriatr. Psychiatry 10, 407–416.

Dahlberg, L.L., Ikeda, R.M., Kresnow, M.J., 2004. Guns in the home and risk of a violent death in the home: findings from a national study. Am. J. Epidemiol. 160 (10), 929–936.

De Leo, D., Dwyer, J., Firman, D., Neulinger, K., 2003. Trends in hanging and firearm suicide rates in Australia: substitution of method? Suicide Life Threat. Behav. 33 (2), 151–164.

De Moore, G.M., Robertson, A.R., 1999. Suicide attempts by firearms and by leaping from heights: a comparative study of survivors. Am. J. Psychiatry 9, 1425–1431.

Ellis, A.M., 1996. Suicide from the Clifton Suspension Bridge in England. J. Epidemiol. Community Health 50, 474.

Glatt, K.M., 1987. Helpline: suicide prevention at a suicide site. Suicide Life Threat. Behav. 17, 299–309.

Gunnell, D., Nowers, M., 1997. Suicide by jumping. Acta Psychiatr. Scand. 96, 1–6.

Hawton, K., Harriss, L., Simkin, S., Bale, E., Bond, A., 2004a. Self- cutting: patient characteristics compared with self-poisoners. Suicide Life Threat. Behav. 34 (3), 199–208.

Hawton, K., Simkin, S., Deeks, J., Cooper, J., Johnston, A., Waters, K., Arundel, M., Bernal, W., Gunson, B., Hudson, M., Suri, D., Simpson, K., 2004b. UK legislation on analgesic packs: before and after study of long term effect on poisonings. Br. Med. J. 329 (7474), 1076.

H en- ack ited-

H hold icidal ev. 8,

J rvey Accid.

K the . Soc.

K tterns ce 5,

K anis, res.

K ates,

L ntrol

L mpts.

L . In: nitz,

L iong le-

Con- tives.

L ntrol

Lester, D., 1988. Why do people choose particular methods for suicide? Activitas Nervosa Superior (Praha) 30 (4), 312–314.

Lester, D., 1990. The effects of detoxification of domestic gas on suicide in the United States. Am. J. Public Health 80, 80–81.

Lester, D., 1993. Suicide from bridges in Washington, DC. Percept. Mot. Skills 77, 534.

Lester, D., 1995. The availability of guns and the rates of personal vio- lence (homicide and suicide). Ital. J. Suicidol. 5 (2), 73–76.

Lester, D., Abe, K., 1989. The effect of restricting access to lethal meth- ods for suicide: a study of suicide by domestic gas in Japan. Acta Psychiatr. Scand. 80, 180–182.

Lindqvist, P., Jonsson, A., Erikson, A., Hedelin, A., Björntig, U., 2004. Are suicides by jumping off bridges preventable? An analysis of 50 cases from Sweden. Accid. Anal. Prev. 36, 691–694.

Marzuk, P.M., Leon, A.G., Tardiff, K., Morgan, E.B., Stajic, M., Mann, J.J., 1992. The effect of access to lethal methods of injury on suicide rates. Arch. Gen. Psychiatry 49 (6), 451–458.

McIntosh, J.L., 2003. 2000 Official Final Statistics U.S.A. Sui- cide. American Association of Suicidology, Washington, DC, retrieved February 11, 2005, from the World Wide Web: http://mypage.iusb.edu/∼jmcintos/SuicideStats.html.

National Institute for Mental Health in England, 2005. National Suicide Prevention Strategy for England: Annual Report 2004. Department of Health, London, England.

Nordentoft, M., Qin, P., Helweg-Larsen, K., Juel, K., 2004. Development in suicide rates in Denmark and availability of means for suicide. In: Proceedings of the 10th European Symposium on Suicide and Suicidal Behaviour, Copenhagen, p. 54.

Nowers, M., Gunnell, D., 1996. Suicide from the Clifton Suspension Bridge in England. J. Epidemiol. Community Health 50, 30–32.

Oliver, R.G., Hetzel, B.S., 1972. Rise and fall of suicide rates in Australia:

O tition

P ping an,

ersity

P Psy-

P with riptive 377–

R talla- influ- uro-

agen,

R who kland

R ns of

S K., Sui- 385–

S icide ehav.

S sui- d Bay

S Harm titute

awton, K., Townsend, E., Deeks, J., Appleby, L., Gunnell, D., B newith, O., Cooper, J., 2001. Effects of legislation restricting p sizes of paracetamol and salicytate on self-poisoning in the Un Kingdom, before and after study. Br. Med. J. 19, 1203–1207.

emenway, D., Miller, M., 2002. Association of rates of house handgun ownership, lifetime major depression, and serious su thoughts with rates of suicide across US census regions. Inj. Pr 313–316.

ohansson, L., Stenlund, H., Lindqvist, P., Eriksson, A., 2005. A su of teenager unnatural deaths in northern Sweden 1981–2000. Anal. Prev. 37 (2), 253–258.

aplan, M.S., Geling, O., 1998. Firearm suicides and homicides in United States: regional variations and patterns of gun ownership Sci. Med. 46 (9), 1223–1227.

aplan, M., Geling, O., 1999. Sociodemographic and geographic pa of firearm suicide in the United States, 1989–1993. Health Pla 179–185.

ontaxakis, V., Markadis, M., Vaslamatzis, G., Ionnidis, H., Steph C., 1988. Attempted suicide by jumping: clinical and social featu Acta Psychiatr. Scand. 77, 435–437.

reitman, N., 1976. The coal gas story—United Kingdom suicide r 1960–1971. Br. J. Prev. Soc. Med. 30, 86–93.

ambert, M.T., Silva, P.S., 1998. An update on the impact of gun co legislation on suicide. Psychiatr. Q. 69 (2), 127–134.

anglois, S., Morrison, P., 2002. Suicide deaths and suicide atte Health Rep. 13 (2), 9–22.

eenaars, A.A., 2001. Controlling the environment to prevent suicide Wasserman, D. (Ed.), Suicide: An Unnecessary Death. Martin Du London, pp. 259–263.

eenaars, A., Cantor, C., Connolly, J., EchoHawk, M., Gailiene, D., X He, Z., Kokorina, N., Lester, D., Lopatin, A.A., Rodriquez, M., Sch busch, L., Takahashi, Y., Vijayakumar, L., Wenckstern, S., 2000. trolling the environment to prevent suicide. International perspec Can. J. Psychiatry 45, 639–644.

eenaars, A.A., Lester, D., 1996. Gender and the impact of gun co on suicide and homicide. Arch. Suicide Res. 2, 223–234.

relation to sedative availability. Med. J. Aust. 2, 919–923. wens, D., Horrocks, J., House, A., 2002. Fatal and non-fatal repe

of self-harm: systematic review. Br. J. Psychiatry 181, 193–199. ommereau, X., Tedo, P., Penouil, F., 1989. Attempted suicide by jum

from a height: a five-year retrospective study. In: Platt, S.D., Kreitm N.P. (Eds.), Current Research on Suicide and Parasuicide. Univ of Edinburgh, Edinburgh, pp. 153–162.

rasad, A., Lloyd, G.G., 1983. Attempted suicide by jumping. Acta chiatr. Scand. 68, 394–396.

ŕevost, C., Julien, M., Brown, B.P., 1996. Suicides associated the Jacques-Cartier Bridge, Montreal, Quebec 1988–1993: desc analysis and intervention proposal. Can. J. Public Health 87, 380.

eisch, T., Michel, K., 2004. Suicide data before and after ins tion of the safety net at the Berne Muenster Terrace and its ence on nearby jumping sites. In: Proceedings of the 10th E pean Symposium on Suicide and Suicidal Behaviour, Copenh p. 54.

osen, D.H., 1975. Suicide survivors, a follow-up study of persons survived jumping from the Golden gate and San Francisco-Oa Bay Bridges. West. J. Med. 122, 289–294.

osen, D.H., 1976. Suicide survivors: psychotherapeutic implicatio egocide. Suicide Life Threat. Behav. 6, 209–215.

akinofsky, I., 2000. Repetition of suicidal behaviour. In: Hawton, van Heeringen, K.P. (Eds.), The International Handbook of cide and Attempted Suicide. Wiley, Chichester, England, pp. 404.

eiden, R.H., 1978. Where are they now? A follow-up study of su attempters from the Golden Gate bridge. Suicide Life Threat. B 8, 203–216.

eiden, R.H., Spence, M., 1983. A tale of two bridges: comparative cide incidence on the Golden Gate and San Francisco Oaklan Bridges. Omega 14, 201–209.

teenkamp, M., Harrison, J.E., 2000. Suicide and Hospitalised Self- in Australia. Injury Research and Statistics Series. Australian Ins of Health and Welfare, Adelaide, Australia.

632 M.S. Daigle / Accident Analysis and Prevention 37 (2005) 625–632

Suominen, K., Isometsa, E., Suokas, J., Haukka, J., Achte, K., Lonnqvist, J., 2004. Completed suicide after a suicide attempt: a 37-year follow- up study. Am. J. Psychiatry 161 (3), 562–563.

Whitlock, F.A., 1975. Suicide in Brisbane, 1956 to 1973: the drug-death epidemic. Med. J. Aust. 1, 737–743.

Wilkinson, S., Taylor, G., Templeton, L., Mistral, W., Salter, E., Bennett, P., 2002. Admissions to hospital for deliberate self-harm in England

1995–2000: an analysis of hospital episode statistics. J. Public Health Med. 24, 179–183.

Yamasawa, K., Nishimukai, H., Ohbora, Y., Inoue, S., 1980. Statistical study of suicides through intoxication. Acta Med. Leg. et Soc. 30, 187–192.