Journal of Affective Disorders 190 (2016) 370375

Contents lists available at ScienceDirect

Journal of Affective Disorders

journal homepage: www.elsevier.com/locate/jad

Research report

Suicide and the Internet: Changes in the accessibility of suicide-related

information between 2007 and 2014
Lucy Biddle a, Jane Derges a,n, Becky Mars a, Jon Heron a, Jenny L. Donovan a, John Potokar a,
Martyn Piper b, Clare Wyllie c, David Gunnell a
a
School of Social and Community Medicine, University of Bristol, United Kingdom
b
Papyrus Prevention of Young Suicide, United Kingdom
c
Samaritans, United Kingdom

art ic l e i nf o a b s t r a c t

Article history: Background: Following the ongoing concerns about cyber-suicide, we investigate changes between 2007
Received 7 May 2015 and 2014 in material likely to be accessed by suicidal individuals searching for methods of suicide.
Received in revised form Methods: 12 search terms relating to suicide methods were applied to four search engines and the top
2 October 2015
ten hits from each were categorised and analysed for content. The frequency of each category of site
Accepted 15 October 2015
Available online 29 October 2015
across all searches, using particular search terms and engines, was counted.
Results: Key changes: growth of blogs and discussion forums (from 3% of hits, 2007 to 18.5% of hits,
Keywords: 2014); increase in hits linking to general information sites especially factual sites that detail and
Suicide evaluate suicide methods (from 9%, 2007 to 21.7%, 2014). Hits for dedicated suicide sites increased (from
Suicide methods
19% to 23%), while formal help sites were less visible (from 13% to 6.5%). Overall, 54% of hits contained
Internet
information about new high-lethality methods.
Limitations: We did not search for help sites so cannot assess the balance of suicide promoting versus
preventing sites available online. Social media was beyond the scope of this study.
Conclusions: Working with ISPs and search engines would help optimise support sites. Better site
moderation and implementation of suicide reporting guidelines should be encouraged.
& 2015 Elsevier B.V. All rights reserved.

1. Introduction
Against a backdrop of rapid developments in the online en-
vironment over more than a decade, there have been a substantial
number of cases of suicide in which the Internet has been im-
plicated. These have included instances of cyber-bullying and
cluster suicides (Birbal et al., 2009; Boyce, 2010; Becker and
Schmidt, 2004; Cheng et al., 2012; Daine et al., 2013). Scrutiny of
the Internet and social media has also emerged within academia,
where a small number of empirically-based studies have been
conducted looking at use of the Internet and its relationship to
self-harm and suicide (Baker and Fortune, 2008; Barak, 2007;
Biddle et al., 2012; Cash et al., 2013; Collings and Niederkro-
tenthaler, 2012; Dunlop et al., 2011; Gilat and Tobin, 2009; Harris
et al., 2014; McCarthy, 2010; Mitchell and Ybarra, 2007; Sisask and
Varnik, 2012). These studies have identied the potentially bene-
cial as well as harmful effects of the Internet on suicidal
n
Correspondence to: Canynge Hall, 39 Whatley Road, Bristol BS8 2PS, United
Kingdom. Fax: 44 117 928 7325.
E-mail address: jane.derges@bristol.ac.uk (J. Derges).
behaviour whilst emphasising its mixed content (Gunn and Lester,
2013; Hagihara et al., 2012; Westerlund, 2011; Yang et al., 2011),
which allows individuals who are suicidal to access help but also
the information with which to make a suicide attempt (Alao et al.,
2006; Daine et al., 2013; Durkee et al., 2011; Sueki et al., 2014).
In recent years there have been a series of Internet search
studies investigating the ease with which a suicidal individual can
access information about suicide on the web and what they might
nd when searching for information about methods (Biddle et al.,
2008; Till and Niederkrotenthaler, 2014; Pirkis et al., 2006; Re-
cupero et al., 2008; Sakarya et al., 2013; Wong et al., 2013). These
studies show that information on suicide is readily available;
through news sites, factual information-based sites and within
dedicated or pro-suicide sites. However, the online environment
is evolving and becoming more complex (Daine et al., 2013; Hi-
tosugi et al., 2007). In particular, there has been an expansion in
user-generated content, both social and information-based, such
as personal websites and participation in interactive discussion
forums and chat rooms, thus expanding opportunities to access
and exchange information on a global scale. This is perhaps
especially pertinent where site moderation is limited, or even
absent.

http://dx.doi.org/10.1016/j.jad.2015.10.028
0165-0327/& 2015 Elsevier B.V. All rights reserved.
 L. Biddle et al. / Journal of Affective Disorders 190 (2016) 370375 371

To begin assessing the impact of such changes on online con- Table 1

tent, we replicated our Internet search study of 2007 (Biddle et al., Coding frame.
2008) examining what a suicidal individual might nd on
Code Description
searching the web for information on suicide methods. As in 2007,
the study reported here focused specically on searches for A Dedicated suicide sites
methods of suicide rather than help or support, in order to es- A1 Suicide sites encouraging/promoting suicide
tablish specically, what someone who is suicidal and looking for A2 Sites that describe methods, but do not encourage
A3 Sites that portray suicide/self-harm in fashionable terms
methods would nd on the Internet. This will assist in addressing B Other sites/pages containing information about suicide methods
some of the concerns currently being expressed in the media and B1 Source of factual information e.g. Encyclopaedia/news article
the public health sector (Gov. Report, DOH, 2014; Gunnell et al., B2 Discussed in a joking fashion that is suggestive of real methods
2012; Boyce, 2010; Barak, 2007). As well as directly comparing our B3 Completely joking and all/most methods described are not real
C Against suicide sites
ndings with those obtained in 2007, we expanded our analysis to
C1 Explicitly forbids or opposes suicide
explore in more detail the content of the sites retrieved by C2 Does not forbid but discourages suicide
focussing on additional themes such as: peer support, images, site D Support/prevention sites offering information and access to support
moderation and help offered within discussion forums and blog D1 Formal organisations
sites. The study also took into consideration new and emerging D2 Individuals
E Academic/policy site focused on suicide
methods of suicide that were being discussed online and which F Search term not relevant
were reported to be increasing as methods of suicide, both in the G Newspaper article describing a suicide
UK and other parts of the world (Grassberger and Krauskopf, G1 General news report about a suicide
2007; Howard et al., 2010; WHO, 2014). G2 News report mentioning role of internet in a suicide
Ha General (i.e. not suicide) Interactive live sites: (discussion boards/
pages; chat rooms/ forums; blogs)
Ha.1 Discussion where focus is on suicide methods (individuals ask-
2. Methods ing for advice; discussing suicide methods; suggesting methods)
Ha1.1 Mostly pro-suicide
2.1. Search strategy Ha1.2 Mostly against suicide
Ha1.3 Mixed content
Ha.2 Discussion about various general issues relating to suicide
We conducted Internet searches between February and March where methods is not the focus (e.g. personal accounts of suicidal
2014 using the same 12 search terms as those in our previous feelings, coping, self-help, nding help, attempting suicide, causes of
study (Biddle et al., 2008): (a) suicide, (b) suicide methods, (c) suicide)
Ha2.1 Mostly pro-suicide
suicide sure methods, (d) most effective methods of suicide, (e)
Ha2.2 Mostly against suicide
methods of suicide, (f) ways to commit suicide, (g) how to commit Ha2.3 Mixed content
suicide, (h) how to kill yourself, (i) easy suicide methods, (k) best Hb Suicide specic Interactive live sites: (discussion boards/pages; chat
suicide methods, (l) pain-free suicide and, (m) quick suicide. These rooms/forums; blogs
were selected as terms a suicidal user might employ based on our Hb.1 Discussion where focus is on suicide methods (individuals
asking for advice; discussing suicide methods; suggesting methods)
previous research with participants who had made near fatal Hb1.1 Mostly pro-suicide
suicide attempts (Biddle et al., 2012), and search engine sugges- Hb1.2 Mostly against suicide
tions taken from entering the term suicide. These 12 broad search Hb1.3 Mixed content
terms were again applied to the four most popular search engines: Hb.2 Discussion about various general issues relating to suicide
where methods is not the focus (e.g. personal accounts of suicidal
Google, Bing (previously MSN), Yahoo and Ask. Google continues
feelings, coping, self-help, nding help, attempting suicide, causes of
to have the largest global market share of all the search engines suicide)
in February 2014 this was 88.6%. The other popular search engines Hb2.1 Mostly pro-suicide
are Yahoo (3.4%), Bing (6.4%) and others (1.6%). (http://theeword. Hb2.2 Mostly against suicide
co.uk/info/search_engine_market.html). Hb2.3 Mixed content
I Misc. sites reporting on or covering topical issues relating to suicide
Internet search results can vary despite the same search terms
J Page not available
being entered on the same search engine, according to the device K Collections of images (e.g. from Google/Yahoo/Bing/Ask)
used, its location, the users IP address, and previous search history
(Heerwegh, 2003; Kim and Kuljis, 2010). We explored this varia-
tion on each of the 4 search engines using a University desktop, coded using a coding frame derived from our previous study
and a home desktop, laptop and hand-held ipad in various loca- (Biddle et al., 2008) but updated to include newly emerging ca-
tions. We did this for two of our search terms suicide as a tegories (see Table 1).
general term likely to yield a vast range of hits, and how to kill
yourself as a more specic term. We recorded the top ten hits 2.2. Analysis
from each search, noting the extent and pattern of difference
across locations/devices. We found for example, that entering Two methods of analysis were used: coding the Internet sites,
suicide on ASK using rst a hand-held ipad and then a laptop, and a rudimentary content analysis of all the sites visited. The
elicited different results. Typically, the same sites were retrieved content analysis provided additional information on the nature of
across searches but the order in which they appeared differed, sites, including: references to celebrity suicide; links to help sites
including whether or not help sites were located within the rst or services; the presence of images including video clips, pictures
page. As variation did not seem to alter overall content in any and photographs of suicides; and information about novel high-
signicant way, for the main analysis we decided to run all sear- lethality methods identied from studies looking specically at
ches on a home desktop without any institutional controls or l- new and emerging methods, both within and outside the UK
ters, as in 2007, and started a new private browsing session for (Chan et al., 2005; Corkery et al., 2010; Grassberger and Krauskopf,
each search to remove the effects of previous searches. 2007; Gunnell et al., 2014).
As it is rare for searchers to look beyond the rst page (Ey- LB and JD independently categorised each hit using the coding
senbach and Kohler, 2002), the rst ten sites were retrieved from frame and undertook the content analysis of each site. For each hit,
each search, totalling 480 hits in all. All of the 480 hits were the whole site was taken as the unit of analysis rather than an
372 L. Biddle et al. / Journal of Affective Disorders 190 (2016) 370375

individual page, therefore we navigated around the site from our
initial entry point in order to reach a coding decision and complete
the content analysis. Any unresolved coding disagreements be-
tween LB and JD were referred to a third researcher (DG), who
analysed these blinded to the codes assigned to them by LB and JD.
In most cases this resolved the disagreement but where a decision
still could not be made LB, JD, and DG jointly visited the site and
discussed the rationale for their coding until agreement was
reached. Content analysis was checked in a similar way.
When all the sites had been coded, the frequency of each ca-
tegory of site across all searches and using particular search terms
and engines was counted. In each search, a rank was assigned to
the hits: 1 the rst site listed, 2 the second site listed, and on up
to 10. This indicated the status of each hit and therefore the
probability of it being accessed by a user. The median rank for each
type of site was calculated. We assessed the number of times a
particular site was identied across all searches. The 10 sites most
frequently retrieved were then identied and their median rank
calculated. We explored ndings overall and by search engine to
take account of Googles higher market share. Analysis was per-
formed using Stata version 10.
3. Results
We visited and analysed 480 web addresses (hits) and 135
different sites (taking account of duplicate hits). This compared to
240 different sites from the 480 hits in our 2007 search.
The frequency and accessibility of hits in the two time periods
is shown in Table 2. In 2014, 111 (23.1%) hits were for dedicated
suicide sites, somewhat higher than the number in 2007 (90
(19%)), the main growth occurring in relation to hits describing
methods without specically encouraging suicidal behaviour
(from 43 (9%) hits to 69 (14.4%)). The number of hits actively
promoting suicide was similar at the two time periods: 42 (8.8%)
in 2014 versus 45 (9%) in 2007. The biggest change was seen in the
number of hits leading to sites providing information about sui-
cide methods. The majority of these were factual information sites,
which had tripled from 24 (5%) hits in 2007 to 73 (15.2%) in 2014.
Such sites list or detail and sometime evaluate methods of suicide
and how to implement these. For example, Wikipedia, which
comprised 73.9% of the hits in this category, gives detailed ac-
counts of a variety of methods (traditional and novel) reporting on
issues such as effectiveness, dosage, side effects and often in-
cluding case examples from academic texts and the media.
Other notable changes were: a decrease in prevention/support
site hits, which halved from 62 (13%) to 31 (6.5%) hits; and a more
than six-fold increase in the number of hits leading to discussions
focusing on suicide methods from 12 (3%) in 2007 to 89 (18.5%)
in 2014 though the content of these was of a mixed nature
(below). There were decreases in academic sites. A new category
emerging in 2014 was hits for sites containing explicit images of
self-harm, suicide, and suicide methods (1.7% of all hits). These
included visual media images placed alongside text; postings of
photos and pictures of self-harm and suicide attempts (e.g. Bing
and Yahoo images), and videos of the same (e.g. YouTube).
There has been little change to median rank scores since 2007
(see Table 2). In 2014, both dedicated suicide sites and information
sites had the lowest median rank indicating that they tended to
appear earlier in each search output. This suggests these may have
had a higher probability of being accessed in each search, as
people come across these rst. Notably, the factual information
sites median score was 2. This means that at least half of the time
these sites were being retrieved, they were hit 1 or 2 in the search,
but such sites comprised 15.2% of all hits.
In terms of individual search engines (see Table 3), in 2014
Google searches accessed a higher proportion (30%) of dedicated
suicide sites than Yahoo, BING and Ask combined (20.8%), while
the three non-Google search engines yielded a higher proportion
of against suicide sites (11.9% versus 4.2%).
A comparison of the top 10 sites across the two time periods
(Table 4) further illustrates these changes. The top 10 sites in 2014
contained more information sites and Wikipedia had risen from
fourth to the top hit. Well-known dedicated suicide sites (Alter-
native Suicide Holiday (ASH), Satan Service) no longer appeared
within the top ten hits, as they had in 2007. However, their suicide
methods les did, information from this having been copied into
another top 10 site. The rise in social media is reected by the
occurrence of blogging/discussion sites in the 2014 top 10. Three
support/ prevention sites identied in the 2007 top ten had
dropped out of view in 2014.

Table 2
Comparison of frequency and accessibility of sites: 2007, 2014.

2007 2014

Type of site No (%) of hits (n 480) Median rank score No (%) of hits (n 480) Median rank score

All dedicated suicide sites: 90 (19) 5 111 (23.1) 4

Pro-suicide; encourage, promote, or facilitate 45 (9) 4 42 (8.8) 6
Describe methods but do not encourage 43 (9) 5 69 (14.4) 4
Portray suicide in fashionable terms 2 (0.4) 5
All information sites: 44 (9) 4.5 104 (21.7) 4
Factual information about suicide methods 24 (5) 3.5 73 (15.2) 2
Partly joking/tongue in cheek information but suggestive of real methods 12 (3) 7.5 1 (0.2) 5
Completely joking information: most or all methods described not real 8 (2) 5.5 30 (6.3) 7
Against suicide sites 59 (12) 5 48 (10.0) 5
Prevention or support sites 62 (13) 5 31 (6.5) 6
Academic or policy sites 70 (15) 7 38 (7.9) 6.5
Hit not relevant to search term 88 (18) 5.5 15 (3.1) 5
News reports of individual suicides 12 (3) 7.5 12 (2.5) 5
Chat rooms/blog sites focused on suicide methods 12 (3) 6 89 (18.5) 7
Chat rooms/blog sites discussing various general issues relating to 8 (2) 4.5 6 (1.3) 7
suicide
Miscellaneous sites reporting on topical issues relating to suicide 35 (7) 7 15 (3.1) 6
Page not available* 4 (0.8) 7.5
Collections of images* 8 (1.7) 6

*
Denotes new categories, 2014.
 L. Biddle et al. / Journal of Affective Disorders 190 (2016) 370375 373

Table 3
Frequency and accessibility of sites overall and by search engine, 2014.

Overall (n 480) Google (n 120) Yahoo, Bing, Ask (n360)

Type of site No (%) of Median rank No (%) of Median rank No (%) of hits Median rank score
hits score hits score (n 360)

All dedicated suicide sites: 111 (23.1) 4 36 (30.0) 4.5 75 (20.8) 4

Pro-suicide; encourage, promote, or facilitate 42 (8.8) 6 12 (10.0) 6 30 (8.3) 6
Describe methods but do not encourage 69 (14.4) 4 24 (20.0) 4 45 (12.5) 3
Portray suicide in fashionable terms
All information sites: 104 (21.7) 4 27 (22.5) 6 77 (21.4) 3
Factual information about suicide methods 73 (15.2) 2 18 (15.0) 3.5 55 (15.3) 2
Partly joking/tongue in cheek information but suggestive 1 (0.2) 5 1 (0.8) 5
of real methods
Completely joking information: most or all methods 30 (6.3) 7 8 (6.7) 8.5 22 (6.1) 6
described not real
Against suicide sites 48 (10.0) 5 5 (4.2) 3 43 (11.9) 5
Prevention or support sites 31 (6.5) 6 6 (5.0) 8 25 (6.9) 6
Academic or policy sites 38 (7.9) 6.5 9 (7.5) 3 29 (8.1) 7
Hit not relevant to search term 15 (3.1) 5 5 (4.2) 6 10 (2.8) 5
News reports of individual suicides 12 (2.5) 5 4 (3.3) 6 8 (2.2) 4.5
Chat rooms/blog sites focused on suicide methods 88 (18.3) 7 27 (22.5) 6 61 (16.9) 7
Chat rooms/blog sites discussing various general is- 6 (1.3) 7 6 (1.7) 7
sues relating to suicide
Miscellaneous sites reporting on topical issues relat- 15 (3.1) 6 1 (0.8) 6 14 (3.9) 6
ing to suicide
Page not available 4 (0.8) 7.5 4 (1.1) 7.5
Collections of images 8 (1.7) 6 8 (2.2) 6

3.1. Content analysis
Content analysis showed that over half of the hits (259 (54.0%))
mentioned new high-lethality methods. These hits linked to 41
individual sites, many of which were discussion forums and blogs.
Only 1 of these sites (a personal blog) linked to a formal help site.
A total of 6 sites (68 (14.2%) hits) contained graphic imagery of
methods, although one joking in nature. Blog sites/ discussion
boards were of mixed content and of those discussing suicide
methods, 48 hits were deemed to be primarily against suicide,
supporting individuals not to act, as opposed to 34 that were en-
couraging or pro-suicide, and 7 that took a neutral position. Help
services information such as links to formal webpages, or tele-
phone hotline numbers, was provided by 32 sites. A total of 179
(37.3%) hits linked to these sites.
Additional observations were that the overall intent behind
some sites is often blurred; for example, users could be drawn to a
site that promotes itself as offering information on suicide meth-
ods, but instead encourages users to seek help. Examples include:
http://ashishpaliwalunstoppable.blogspot and www.Cracked.com.
Conversely, sites such as Lostallhope.com, which claims to support
individuals who are suicidal, provides explicit and detailed in-
formation on suicide methods, thus adopting a paradigm in which
suicide is regarded as a legitimate option and where providing
information about methods is deemed one form of support. Dis-
claimers attached to sites were also more present in the 2014
search ndings.
4. Discussion
Our updated search of the Internet for suicide methods showed
that dedicated, pro-suicide sites were slightly more prominent
amongst the hits returned in our 2014 search compared to 2007.
Hits for prevention and support sites had reduced by nearly half,
whereas factual information sites many of which provide highly
explicit how to information on methods had tripled. Informa-
tion sites also typically ranked the highest in searches showing
they are easily accessed. Google searches returned more dedicated
suicide sites than other search engines, which is notable since
Google has the largest global market share of all search engines.
While the well-known Satan Service and ASH sites were no
longer amongst the top 10 sites retrieved, we found that the ASH
methods le was copied into other websites and personal blogs,
including one within the top 10 sites. Thus, the content remains
present and accessible though in less overt form. Changes between

Table 4
Top ten most frequently occurring sites: 2007 compared to 2014.

2007 2014

No. Site Category No. hits Median rank Site Category No. hits Median rank

1 Alt Suicide Holiday A1 24 5 Wikipedia B 44 1

2 Satan Service A1 19 2 Lostallhope A2 32 3
3 Suicidemethods.net A2 19 6 Martin Frost B 20 6
4 Wikipedia B 18 1.5 Suicidemethods.net B1 20 8
5 Extreme River C 17 5 Frater A2 19 3
6 Satan Service J 8 2.5 Uncyclopedia.wikia.com B3 19 6
7 New Hope Health Clinic C 7 4 Wantdeath.blogspot A1 19 6
8 Amazon I 7 4 Ashishpaliwalunstoppable.blogspot Ha 12 8.5
9 Soon D 7 5 Cracked.com C2 12 5
10 23nl people F 6 1 Experience Project Ha 11 5
374 L. Biddle et al. / Journal of Affective Disorders 190 (2016) 370375
2007 and 2014 indicate that information about suicide is now
widely disseminated through discussion forums and blogs, while
organisations supporting those who are suicidal continue to op-
erate alongside these. Discussion threads are of mixed content
some discouraging suicide and promoting help-seeking while
others providing knowledge, and in turn, cognitive availability of
suicide methods. Novel high-lethality methods are frequently
communicated. Thus while some are harmful, these also demon-
strate a level of informal help-seeking and peer support. Another
development since 2007 was that hits of graphic images of suicide
and suicide methods became present within our search results.
4.1. Strengths and limitations
Our ndings are consistent with other studies exploring sui-
cide-related internet content. These have also shown the accessi-
bility of online suicide information and noted its mixed content;
both information on methods and opportunities for support being
available within discussion forums and blog sites (Daine et al.,
2013; Hagihara et al., 2012; Ikunaga et al., 2013; Kemp and Coll-
ings, 2011; Marhan et al., 2012; Recupero et al., 2008). There are
currently no other studies providing a comparative and long-
itudinal approach to looking at changes in online content retrieved
when entering a set of search terms surrounding suicide and
suicide methods across two time points. The Internet rapidly
evolves and by replicating the methods of our previous study, we
have been able to explore some of these changes.
The vastness and evolving nature of the web which sees sites
change with considerable frequency, means that our searches
provide snapshots of web-based material that were easily accessed
via search engines when searching for information about suicide
and suicide methods on two particular occasions. The study is
based on the samples of material retrieved and does not inform
about absolute web content. Our focus on suicide information and
methods responds to ongoing concerns expressed about the In-
ternet in relation to self-harm and suicide. Our study was thus
designed to mimic some but not all possible types of Internet
usage by a suicidal individual, such as a non-technical individual
entering common sense search terms in a time of crisis, or an
individual deliberately researching methods of suicide. The results
return indicative information on some of the material that might
be accessed. We have not explored online help available and how
one accesses this for instance by entering specic help search
terms such as suicide help. Therefore our results do not indicate
the balance of suicide promoting versus suicide preventing sites.
This would be an important but separate study. On the other hand,
neither did we explicitly search for dedicated suicide sites. Nor
can we use our data to draw denitive conclusions about actual
online behaviour, how users interpret the material they view or to
understand the nature of online communities. We are currently
pursuing such issues through on-going qualitative study of people
who have used the Internet in relation to episodes of self-harm
(research in progress: Department of Health study ID no. 023/
0163). While demonstrating the complexity of the subject, our
preliminary qualitative ndings reafrm the validity of the data
reported here. Participants frequently make reference to highly
ranking sites, describe similar search strategies to those that we
undertook, and have acted upon information gained from sites
identied in our internet search. Some participants state that peer
support within discussion forums has increased value as it comes
through the direct, personal experience of other users.
Systematic searching and content analysis of specic social
media sites such as Tumblr, Twitter, and Pinterest was beyond the
scope of our study. Such sites have been implicated in recent high
prole cases of cybersuicide. Similarly, we could not provide clear
information on how searches differ according to the devices used,
nor the effect of personal cookies. Search engines can function to
personalise search results according to search history such that
previous searches for suicide-related material may increase the
volume obtained in subsequent searches. Our study may therefore
underestimate the accessibility of such material to returning users.
Our pilot work in this area indicated that the ordering of hits may
vary, thus rank scores should be viewed with some caution. Fur-
ther, our search did not access materials available on the deep
web. Suicidal thoughts and suicidal behaviour occur in a full
cross-section of people meaning that there is a varied skill set in
terms of accessing and using web-based technology, from the
computer literate to relative beginners or less sophisticated users.
It cannot therefore be assumed that all those who access the In-
ternet for suicide related purposes do so uniformly, and this
therefore affects how online searches would be conducted.
4.2. Policy implications
Our ndings show a variety of suicide information is easily acces-
sible online. Multiple strategies may be required to tackle these dif-
fering types of content and spaces. Findings also suggest that attempts
to block or remove content promoting suicide may be problematic. For
instance, there was evidence of information such as the ASH methods
le moving between sites, including personal blogs. The availability of
information on suicide methods has thus become more diffuse, per-
sonalised, and transient and may be more difcult to moderate. Si-
milarly, we detected a change in tone towards a more neutral pre-
sentation of information and an increased use of disclaimers, perhaps
as cases of cybersuicide and the potential legal ramications of these
have been publicised. Neutral presentation also occurs where a para-
digm is adopted in which both help-seeking and suicide are perceived
as legitimate options (e.g. Lostallhope.com). From a public health
perspective this presents a considerable danger to vulnerable in-
dividuals amenable to prevention who may instead become assisted in
arriving at a suicide plan. Our ndings also highlight the presence of
numerous online peer support blogs and discussion-based forums.
Removal of these would also remove the potential they offer for sui-
cide prevention. However, as with other suicide-related web-based
material, we found that many of these also contained exchanges that
could encourage or facilitate suicidal behaviour. Deriving policy actions
is therefore highly complex since strategies must manage a balance
between protecting positive elements contained within the web, re-
ducing the more harmful content that is also easily available (Niezen,
2013; Tam et al., 2007; Durkee et al., 2011), and also protecting the
rights to freedom of expression. Potential approaches could include:
greater site moderation, including by site users themselves to restrict
harmful material and ensure greater support for those most at risk;
working with ISPs and search engines to optimise supportive sites;
and expanding the application of media reporting guidelines on sui-
cide to also encompass web material.
Working with ISPs and search engines is likely to be a key
strategy for tipping the balance of the information that is accessed.
ISPs have the capacity to identify the frequency and types of
search terms being used which would give a clearer picture of how
people use the Internet in relation to suicide and allow support
organisations such as the Samaritans to target sectors of the po-
pulation more effectively. The way that search engines structure
data, rather than absolute web content, can determine site pro-
minence. Intervening in this process may make it possible to
promote some sites rather than others. Google in particular could
play a key public health role in managing sites and site optimi-
sation more effectively and safely. It continues to maintain its
place as the most popular search engine but also returned a higher
percentage of dedicated suicide sites in our searches (30% in
contrast to 20.8% among Yahoo, Bing and Ask combined). Detailed
investigation into how search engines function and structure data
 L. Biddle et al. / Journal of Affective Disorders 190 (2016) 370375
is required to pursue this approach.
Information sites, including Wikipedia which featured promi-
nently in our searches, present a signicant challenge in terms of
improving Internet safety. While they provide factual information
and may be thought of as innocuous compared to pro-suicide sites, in
fact they contain and disseminate widely a high volume of well-re-
searched and explicit how to information that can be used as a
resource by people researching and seeking to perfect suicide
methods (Biddle et al., 2012; Gunnell et al., in press). Our qualitative
interviews with internet users indicate that such sites are regularly
accessed by suicidal users and are perceived to be more trustworthy
by those seeking information on methods (work in progress). Mod-
erators and editors of such sites, including news sites, need to be
made more accountable for their content and urged to adopt media
reporting guidelines around suicide also online.
The Internets complexity lies in the degree and speed with
which the online environment changes and how online interac-
tions are now an integral part of most peoples lives. Calls for
greater regulation pose a challenge that is neither easily achiev-
able, nor desired by some if it risks eliminating some of the more
positive elements of the Internet. Striking a balance between
harnessing opportunities to reach out to vulnerable users, reduce
stigma and provide support, whilst protecting some of the more
vulnerable users is therefore necessary.
Acknowledgments
We are extremely grateful to Mike Matthews for running the 2014 searches for
this study. We also thank Rachel Holley, Jen Russell and Chris OSullivan for their
comments during the planning of this study, and Rachel Holley for her comments
on a draft of the paper.
References
http://theeword.co.uk/info/search_engine_market.html.
Alao, A., Soderberg, M., Pohl, E., Lolaalao, A., 2006. Cybersuicide: review of the role
of the Internet on suicide. CyberPsychol. Behav. 9, 489493.
Baker, D., Fortune, S., 2008. Understanding self-harm and suicide websites: a
qualitative interview study of young adult web users. Crisis 29 (3), 118122.
Barak, A., 2007. Emotional support and suicide prevention through the Internet: a
eld project report. Comput. Hum. Behav. 23, 971984.
Becker, K., Schmidt, K.H., 2004. Internet chat rooms and suicide, letters to the
editor. J. Am. Acad. Child Adolesc. Psychiatry 43, 3.
Biddle, L., Donovan, J., Hawton, K., Kapur, N., Gunnell, D., 2008. Suicide and Internet.
Br. Med. J. 336, 800.
Biddle, L., Gunnell, D., Owen-Smith, A., Potokar, J., Longson, D., Hawton, K., Kapur,
N., Donovan, J., 2012. Information sources used by the suicidal to inform choice
of method. J. Affect. Disord. 136, 702709.
Birbal, R., Maharajh, H.D., Clapperton, M., Jarvis, J., Ragoonath, A., Uppalapati, K.,
2009. Cybersuicide and the adolescent population: challenges of the future?
Int. J. Adolesc. Med. Health 21 (2), 151159.
Boyce, N., 2010. Pilots of the future: suicide prevention and the Internet. Lancet 376,
18891890. http://dx.doi.org/10.1016/S0140-6736(10)62199-X, DOI.
Cash, S., Thelwall, M., Peck, S., Ferrell, J., Bridge, J., 2013. Adolescent suicide state-
ments on MySpace. Cyberpsychol., Behav. Soc. Netw. 16, 166174.
Chan, K., Yip, P.S.-F., Au, J., Lee, D.T.S., 2005. Charcoal-burning suicide in post-
transition Hong Kong. Br. J. Psychiatry 186, 6773.
Cheng, Q., Chang, S.-S., Yip, P.S.-F., 2012. Correspondence: opportunities and chal-
lenges of online data collection for suicide prevention. Lancet 379 (www.the-
lancet.com) (May 26).
Collings, S., Niederkrotenthaler, T., 2012. Suicide prevention and emergent media:
Surng the opportunity. Crisis 33, 14.
Corkery, J.M., Button, J., Vento, A.E., Schifano, F., 2010. Two UK suicides using ni-
cotine extracted from tobacco instructions available on the Internet. Forensic
Sci. Int. 199, 913.
Daine, K., Hawton, K., Singaravelu, V., Stewart, A., Simkin, S., Montgomery, P., 2013.
The power of the web: a systematic review of studies of inuence of the In-
ternet on self-harm and suicide in young people. PLoS One 8 (10), e77555. http:
//dx.doi.org/10.1371/journal.pone.0077555, doi.
Dunlop, S.M., More, E., Romer, D., 2011. Where do youth learn about suicides on the
Internet, and what inuence does this have on suicidal ideation? J. Child Psy-
chol. Psychiatry 52 (10), 10731080.
Durkee, T., Hadlaczky, G., Westerlund, M., Carli, V., 2011. Internet pathways in
suicidality: a review of the evidence. Int. J. Environ. Res. Public Health 8,
39383952. http://dx.doi.org/10.3390/ijerph8103938, DOI.
375
Eysenbach, G., Kohler, C., 2002. How do consumers search for and appraise health
information on the world-wide web? Qualitative study using focus groups,
usability tests, and in-depth interviews. Br. Med. J. 324, 573577.
Gilat, I., Tobin, Y., 2009. I am so sick of this life: a semiotic analysis of suicidal
messages on the Internet. In: Sher, L., Vilens, A. (Eds.), Internet and Suicide.
Nova Science Publishers, New York, pp. 169184.
Gov. Report, DOH, 2014. Preventing suicide one year on, rst annual report on the
cross-government outcomes strategy to save lives, January http://www.gov.uk/
dh.
Grassberger, M., Krauskopf, A., 2007. Suicidal asphyxiation with helium: report of
three cases. Wien. Klin. Wochenschr. 119 (910), 323325.
Gunn, J.F., Lester, D., 2013. Using Google searches on the Internet to monitor sui-
cidal behavior. J. Affect. Disord. 148, 411412.
Gunnell, D., Derges, J., Chang, S-S., Biddle, L., 2015. Searching for suicide methods:
accessibility of information about helium as a method of suicide on the Inter-
net. Crisis 36 (5), 325331. http://dx.doi.org/10.1027/0227-5910/a000326.
Gunnell, D., Bennewith, O., Kapur, N., Simkin, S., Cooper, J., Hawton, K., 2012. The
use of the Internet by people who die by suicide in England: a cross-sectional
study. J. Affect. Disord. 14, 480483.
Gunnell, D., Coope, C., Fear, V., Wells, C., Chang, S.-S., Hawton, K., Kapur, N., 2014.
Suicide by gasses in England and Wales 20012011: evidence of the emergence
of new methods of suicide. J. Affect. Disord. 170, 190195. http://dx.doi.org/
10.1016/j.jad.2014.08.055, DOI.
Hagihara, A., Miyazaki, S., Abe, T., 2012. Internet suicide searches and the incidence
of suicide in young people in Japan. Eur. Arch. Psychiatry Clin. Neurosci. 262,
3946.
Harris, K.M., McLean, J.P., Shefeld, J., 2014. Suicidal and online: how do online
behaviors inform us of this high-risk population? Death Stud. 38, 387394.
http://dx.doi.org/10.1080/07481187.2013.768313, DOI.
Heerwegh, D., 2003. Internet Survey Paradata. In: Das, M., Ester, P., Kaczmirek, L.
(Eds.), Social Research and the internet. Taylor and Francis, New York,
pp. 325348.
Hitosugi, M., Nagai, T., Tokudome, S., 2007. A voluntary effort to save the youth
suicide via the Internet in Japan. Int. J. Nurs. Stud. 44, 157.
Howard, M.O., Hall, M.T., Edwards, J.D., Vaughn, M.G., Perron, B.E., Winecker, R.E.,
2010. Suicide by asphyxiation due to helium inhalation. Am. J. Forensic Med.
Pathol. 32 (1), 6170.
Ikunaga, A., Nath, S.R., Skinner, K.A., 2013. Internet suicide in Japan: a qualitative
content analysis of a suicide bulletin board. Transcult. Psychiatry 50 (2),
280302.
Kemp, C.G., Collings, S.C., 2011. Hyperlinked suicide assessing the prominence and
accessibility of suicide promoting websites. Crisis 32, 143151.
Kim, I., Kuljis, J., 2010. Applying content analysis to web-based content. J. Comput.
Inf. Technol. 18 (4), 369375. http://dx.doi.org/10.2498/cit.1001924, DOI.
Marhan, A.-M., Saucan, D., Popa, C., Danciu, B., 2012. Searching the Internet: a re-
port on accessibility, nature, and quality of suicide related information. Proce-
dia Soc. Behav. Sci. 33, 373377.
McCarthy, M.J., 2010. Internet monitoring of suicide risk in the population. J. Affect.
Disord. 122, 277279.
Mitchell, K.J., Ybarra, M.L., 2007. Online behavior of youth who engage in self-harm
provides clues for preventive intervention. Prev. Med. 45, 392396.
Niezen, R., 2013. Internet suicide: communities of afrmation and the lethality of
communication. Transcult. Psychiatry. 50, 303. http://dx.doi.org/10.1177/
1363461512473733, DOI.
Pirkis, J., Burgess, P.M., Francis, C., Blood, W.R., Jolley, D.J., 2006. The relationship
between media reporting of suicide and actual suicide in Australia. Soc. Sci.
Med. 62, 28742886.
Recupero, P.R., Harms, S.E., Noble, J.M., 2008. Googling suicide: surng for suicide
information on the Internet. J. Clin. Psychiatry 69, 878888.
Sakarya, D., Gunes, C., Sakarya, A., 2013. Googling suicide: evaluation of websites
according to the content associated with suicide. Turk. J. Psychiatry, 3740.
Sisask, M., Varnik, A., 2012. Media roles in suicide prevention: a systematic review.
Int. J. Env. Res. Public Health 9, 123138. http://dx.doi.org/10.3390/
ijerph9010123 DOI.
Sueki, H., Yonemoto, N., Takeshima, T., Inagaki, M., 2014. The impact of suicidality-
related internet use: a prospective large cohort study with young and middle-
aged internet users. PLoS One 9 (4), e94841. http://dx.doi.org/10.1371/journal.
pone.0094841, doi.
Tam, J., Tang, W.S., Fernando, D.J.S., 2007. The Internet and suicide: a double-edged
tool. Eur. J. Intern. Med. 18, 453455.
Till, B., Niederkrotenthaler, T., 2014. Surng for suicide methods and help: content
analysis of websites retrieved with search engines in Austria and the United
States. J. Clin. Psychiatry . http://dx.doi.org/10.4088/JCP.13m08861, DOI.
Westerlund, M., 2011. The production of pro-suicide content on the Internet: a
counter-discourse activity. New Media Soc. 14, 764780.
Wong, P.W., Fu, K.-W., Sai-Pong Yau, R., Hei-Man Ma, H., Law, Y.-W., Chang, S.-S.,
Yip, P.S.-F., 2013. Accessing suicide-related information on the Internet: a ret-
rospective observational study of search behavior. J. Med. Internet Res. 15 (1),
e3 (retrieved 20.02.14).
WHO, 2014, Preventing Suicide: A Global Imperative, ISBN 978 92 4 156477 9.
Yang, A.C., Tsai, S.-J., Huang, N.E., Peng, C.-K., 2011. Association of Internet search
trends with suicide death in Taipei City, Taiwan, 20042009. J. Affect. Disord.
132, 179184.