Does prior contact with police reduce the likelihood of suicide? Examining the predictive ability of different incident types and the interaction effects
| Date | 01 September 2017 |
| Published date | 01 September 2017 |
| Author | Judy Li,Rhiannon Newcombe,Sonia Barnes,Darren Walton |
| DOI | 10.1177/1461355717717997 |
| Subject Matter | Articles |
Article
International Journal of
Police Science & Management
Does prior contact with police reduce
2017, Vol. 19(3) 148–158
ª The Author(s) 2017
the likelihood of suicide? Examining the
Reprints and permission:
sagepub.co.uk/journalsPermissions.nav
predictive ability of different incident
DOI: 10.1177/1461355717717997
journals.sagepub.com/home/psm
types and the interaction effects
Darren Walton
Research and Evaluation, New Zealand Police, New Zealand; Psychology Department,
University of Canterbury, New Zealand
Judy Li
Research and Evaluation, New Zealand Police, New Zealand
Sonia Barnes
Research and Evaluation, New Zealand Police, New Zealand
Rhiannon Newcombe
Research and Evaluation, New Zealand Police, New Zealand
Abstract
This study explores the relationship between prior contact with the police as a victim or person at risk and death by
suicide. Findings from this study will improve our understanding of the role the police can have in preventing suicide. The
sample comprised working age men who died suddenly in New Zealand during 2007–2011. Univariate and multivariate
logistic regression models were used to assess the ability of age, ethnicity and prior police contact to predict whether
the death was from suicide (n ¼ 1140) or other causes (n ¼ 7236). The analysis focused on prior police contact relating to:
(1) threatened/attempted suicide, (2) domestic disputes and (3) potential mental health issues. Age, ethnicity and the three
prior police contact types all independently predicted suicide. After controlling for demographic variables, prior police
contact remained a predictor, with those relating to threatened/attempted suicide having the strongest predictive ability
(Adjusted odds ratio [AOR] ¼ 6.05, 4.75–7.70). Importantly, interaction effects were found between prior police contact
types. The predictive ability of contact relating to threatened/attempted suicide reduced significantly if the person also had
prior contact relating to a domestic dispute or potential mental health issue. Consistent with previous research, the main
effects reported in this study suggest that the police have a role in suicide prevention. This study also makes a novel
contribution by exploring the interaction effects between types of prior contact. Future research may investigate the
interaction effects in more depth to reveal the contributing factors that reduce the odds of suicide.
Keywords
Victims, suicide, preventing, gatekeeper, mental health
Submitted 24 Jan 2017, Revise received 18 May 2017, accepted 01 Jun 2017
Corresponding author:
Judy Li, Research and Evaluation, New Zealand Police, PO Box 3107, Wellington 6014, New Zealand.
Email: Judy.Li@police.govt.nz
Walton et al.
149
Introduction
those who had been in conflict with the law (compared with
the general population) could be attributed in part to the social
Suicide and suicide attempts have huge financial and emo-
and economic disadvantages commonly experienced by this
tional costs to society (Brent et al., 1996; O’Dea and Tucker,
group (Sveticic and De Leo, 2011), addiction problems
2005; Palmer et al., 1995; Rudestam, 1977). Proximal and
(Haglund, 2015; Sveticic and De Leo, 2011), and previous
distal predictors of suicides that are well established in the
suicide attempts (Haglund, 2015; Sveticic and De Leo, 2011).
literature include genetic and biological factors, social and
The police are an agency that has frequent contact with
demographic factors, mental health and distress, drug and
people who are at risk of suicide. Understanding the rela-
substance abuse, and exposure to traumatic life events
tionship between prior contact with the police and suicide
(Beautrais et al., 2005; Hawton and van Heeringen, 2009;
provides an additional pathway to identify and triage people
Nock et al., 2008). Understanding the predictors of suicide
who require support. However, only one study has investi-
allows for the development of suicide prevention strategies,
gated the link between prior contact with the police and
including improved engagement between social and health
suicide. Linsley et al. (2007) combined suicide death and
agencies, and people who are at risk.
police records in County Durham and Darlington in the
UK, and noted that 20% of the people who died from suicide
The role of health providers in suicide prevention
had contact with the police in the 3 months prior to death.
There was an equal proportion of those who were a victim or
Clinical practitioners play a significant role in suicide pre-
an alleged perpetrator of crime. This study had three limita-
vention because of the opportunity they have to intervene
tions. First, the sample size was small: 133 confirmed sui-
with those who are at risk. A systematic review collated
cides and 72 probable suicides over 3 years. Second, merely
findings from 24 studies conducted in Europe, the USA,
knowing that there was an association between prior contact
Canada, and Australia on suicide and prior healthcare con-
with police and suicide is not sufficient for strategy devel-
tact, with a combined sample size of 17,149 (Pirkis and
opment. Instead, what is needed is a more focused analysis
Burgress, 1998). Although the review was compiled 20 years
that would help to identify at-risk groups and to understand
ago, it is still the most comprehensive information about this
the nature of the type of contact. Third, although the study
topic available to date. The review revealed that a notable
provided a contact rate for those who died from suicide,
proportion of those who suicided had prior contact with a
there was no control group against which to compare the
health provider, including mental health providers and gen-
contact rate. The analysis therefore did not provide the rela-
eral practitioners (GPs). For example, up to 83% had visited
tive probability of those who had a prior police contact dying
a GP in the previous 12 months, and up to 20% had visited a
by suicide, or another cause. Despite these limitations, the
health provider in the week prior to suicide. Higher contact
findings from Linsley and colleagues (2007) highlight the
rates were found among psychiatric in-patient care (41% had
need to understand the nature of prior police contact and
prior contact in the past year, 9% for the day prior to death)
incorporate police in wider suicide prevention strategies.
compared with community-based care (11% and 4%,
respectively).
Purpose of the study
Recognising the limited empirical evidence on prior contact
Prior police contact and suicide
with police and suicide, this study examines the relationship
Research conducted in Australia, the UK and other European
between prior contact with New Zealand Police as a victim
countries has shown that suicide is also more likely to occur
or person at risk and death by suicide. The scope of the
in other vulnerable populations such as those who commit
analysis was limited by the data available on confirmed
crime. Several international studies have investigated the
suicide and prior police contact, and is therefore restricted
relationship between suicide and imprisonment, showing
to working age men. ‘Working age men’ is defined as males
that the incidence of suicide among this population is higher
who were between 25 and 64 years old at the time of death
than for the general population (Fazel et al., 2005; Haglund,
(Suicide Mortality Review Committee, 2016). Working age
2015). There is also a growing body of research to suggest
men have a high rate of suicide, and are therefore a priority
increased risk of suicide among those serving a community
group for suicide prevention (Suicide Mortality Review
sentence (MacKenzie et al., 2013). Recent studies from
Committee, 2016).
Australia and the UK also suggest that the relationship
To enable a better understanding of the nature of the
between offending and suicide risk may extend beyond those
interaction with the police, prior police contact was included
with custodial or community sentences to those having con-
in the analysis when the individuals were a victim of the
tact earlier in the criminal justice system (King et al., 2015;
incident or a person at risk, and limited to three specific
Sveticic and Leo, 2011). The higher rate of suicide among
types of police contact (‘incidents’). The incidents were
150
International Journal of Police Science & Management 19(3)
coded in the police occurrence record as: (1) threatened/
Each incident that is entered into the police occurrence
attempted suicide, (2) domestic dispute or (3) mental health
database has an incident code, which serves as a quick
incidents. Definitions of these incidents are provided in the
reference to the nature of the incident. Each person
Methods section. Choice of these incidents was informed by
involved in the incident is linked to that occurrence. Where
the literature around risk factors for suicide (Beautrais et al.,
different code options exist to describe the incident, the most
2005; Devries et al., 2011; Hawton and van Heeringen,
specific option should be selected; and therefore, the code
2009; Kilpatrick et al., 1985).
given is believed to be the best description of the incident
To overcome a methodology limitation in the study of
based on information available to the police at the time the
Linsley et al. (2007), a comparison group was included in
data are extracted (New Zealand Police, 2016).
the study. The comparison group consisted of working age
The three types of incidents used for the...
To continue reading
Request your trial