Accepted: 11 October 2017

DOI: 10.1111/jocn.14107

ORIGINAL ARTICLE

Risk assessment and subsequent nursing interventions in a

forensic mental health inpatient setting: Associations and
impact on aggressive behaviour

Tessa Maguire BN, RN MMentHlthSc FBS, Clinical Nurse Consultant1,2 | Michael Daffern
MPsych Clin, PhD Professor of Clinical Forensic Psychology1,2 | Steven J Bowe PhD, Bed
se Maths, MMed. Stats, Senior Research Fellow3 | Brian McKenna RN, PhD, Professor in
Forensic Mental Health1,4

1
Centre for Forensic Behavioural Science,
Swinburne University of Technology,
Aim and objectives: To examine associations between risk of aggression and nurs-
Melbourne, Vic., Australia ing interventions designed to prevent aggression.
2
Victorian Institute of Forensic Mental Background: There is scarce empirical research exploring the nature and effective-
Health, Forensicare, Melbourne, Vic.,
Australia ness of interventions designed to prevent inpatient aggression. Some strategies may
3
Faculty of Health, Biostatistics Unit, be effective when patients are escalating, whereas others may be effective when
Deakin University, Melbourne, Vic.,
aggression is imminent. Research examining level of risk for aggression and selection
Australia
4
School of Clinical Sciences, Auckland and effectiveness of interventions and impact on aggression is necessary.
University of Technology, Auckland, New Design: Archival case file.
Zealand
Methods: Data from clinical files of 30 male and 30 female patients across three
Correspondence forensic acute units for the first 60 days of hospitalisation were collected. Risk for
Tessa Maguire, Centre for Forensic
Behavioural Science, Swinburne University imminent aggression as measured by the Dynamic Appraisal of Situational Aggres-
of Technology, Clifton Hill, Vic, Australia. sion, documented nursing interventions following each assessment, and acts of
Email: tessa.maguire@forensicare.vic.gov.au
aggression within the 24-hours following assessment were collected. Generalised
estimating equations were used to investigate whether intervention strategies were
associated with reduction in aggression.
Results: When a Dynamic Appraisal of Situational Aggression assessment was com-
pleted, nurses intervened more frequently compared to days when no Dynamic
Appraisal of Situational Aggression assessment was completed. Higher Dynamic
Appraisal of Situational Aggression assessments were associated with a greater
number of interventions. The percentage of interventions selected for males dif-
fered from females; males received more pro re nata medication and observation,
and females received more limit setting, one-to-one nursing and reassurance. Pro re
nata medication was the most commonly documented intervention (35.9%) in this
study. Pro re nata medication, limit setting and reassurance were associated with an
increased likelihood of aggression in some risk bands.
Conclusions: Structured risk assessment prompts intervention, and higher risk rat-
ings result in more interventions. Patient gender influences the type of interven-
tions. Some interventions are associated with increased aggression, although this
depends upon gender and risk level.

J Clin Nurs. 2018;27:e971–e983. wileyonlinelibrary.com/journal/jocn © 2017 John Wiley & Sons Ltd | e971
e972 | MAGUIRE ET AL.

Relevance to clinical practice: When structured risk assessments are used, there is
greater likelihood of intervention. Intervention should occur early using least restric-
tive interventions.

KEYWORDS
aggression management, forensic mental health, intervention, mental health nursing, nursing,
risk assessment

1 | INTRODUCTION
What does this paper contribute to the wider
Aggression occurs frequently in inpatient mental health settings,
global clinical community?
resulting in multiple adverse consequences for patients, staff, as well
as for the milieu and operation of the unit (Bowers et al., 2011; Cut- • Few studies have examined the impact of aggression risk
cliffe & Riahi, 2013; Daffern, Howells, & Ogloff, 2007; Daffern, assessment on nursing initiated preventative interven-
Maguire, Carroll, & McKenna, 2015; Davidson, 2005; Duxbury, tions within mental health settings.
Hahn, Needham, & Pulsford, 2008; Needham et al., 2004). There • We examined the use of the Dynamic Appraisal of Situa-
may also be undesirable outcomes for patients who engage in tional Aggression, and interventions nurses documented
aggression, including being subject to restrictive practices such as to prevent aggression, and the impact of these interven-
sedation, seclusion or restraint (Daffern et al., 2015). tions on aggressive behaviour in male and female foren-
sic mental health inpatients.
• Results suggest that structured assessment may prompt
2 | BACKGROUND nurses to intervene, different interventions are used for
males and females, and more interventions are applied
2.1 | Nursing interventions designed to reduce with higher risk patients.
aggression
• Greater consideration is required when selecting inter-
There has been limited investigation into the impact of nursing inter- ventions to prevent aggression. Attention should be
ventions that are used to prevent and manage aggression (Irwin, given to risk level and patient gender, with early inter-
2006), and little guidance as to the most effective methods, resulting vention and least restrictive strategies used before
in uncertainty for nurses (Stevenson, Jack, O’Mara, & Le Gris, 2015). patient’s risk escalates.
Therefore, the selection and implementation of preventative strategies
often occur in the absence of a framework that would assist in provid-
ing pre-emptive and targeted interventions before the aggression nurses may be applying the same intervention but in different ways.
becomes imminent (Taylor et al., 2011). Without a systematic Furthermore, although some of these interventions have been sub-
approach, the interventions used may lack consistency, or be unsuit- ject to empirical investigation, each intervention is rarely investigated
able for the patient or the situation (Daffern et al., 2007), and may lead independently, leading to uncertainty in determining the impact of
to the use of reactive and restrictive practices (Taylor et al., 2011). any one intervention. For example, Bobier et al. (2015) examined the
Some commonly used nursing interventions include increased introduction of sensory modulation; however, the introduction of
observations, one-to-one engagement with a nurse, reassurance, dis- sensory modulation was one of a range of interventions (including
traction techniques (including sensory modulation), limit setting, ver- personal safety assessment tools) that were introduced as part of an
bal de-escalation and the use of pro re nata (PRN) medication organisation-wide restraint and seclusion reduction initiative.
(Department of Health, 2013; MacNeela et al., 2010; NICE, 2015). Evaluation of nursing interventions is typically achieved through
Some interventions such as limit setting have been associated with scrutinising medical records. However, there are limits to this
an increased risk of aggression (Bjørkly, 1999; Sheridan, Henrion, approach. Documentation has been noted to be vague in relation to
Robinson, & Baxter, 1990) and the manner in which limits are set the recording of the specific details of interventions provided (Hale,
may influence how patients respond to this intervention (Lancee, Thomas, Bond, & Todd, 1997; Martin & Street, 2003; Mullen,
Gallop, McCay, & Toner, 1995; Maguire, Daffern, & Martin, 2014). Drinkwater, & Lewin, 2013). Furthermore, the rationale for choosing
Often these interventions are poorly defined (e.g., de-escalation and and enacting the intervention, and outcomes arising from the inter-
limit setting; Irwin, 2006; Johnson & Hauser, 2001; Maguire et al., vention may be unclear or absent (Curtis, Baker, & Reid, 2007;
2014; Roberton, Daffern, Thomas, & Martin, 2012), leading to the O’Brien & Cole, 2004). For example, a nursing file note stating “pa-
possibility of inconsistent application of the intervention, where tient was de-escalated” may not adequately describe the actual
MAGUIRE ET AL. | e973

practice, which may make it difficult for others to replicate or to

2.2 | Risk assessment instruments
know which elements of the intervention were effective.
Another limitation has been the absence of regard for gender dif- The assessment of an individual’s risk of aggression is considered
ferences when considering the type and effectiveness of nursing inter- an important skill for forensic mental health nurses, along with the
ventions used to prevent aggression (Nicholls, Brink, Greaves, Lussier, implementation of evidence-based interventions following assess-
& Verdun-Jones, 2009). Certain interventions may be more frequently ment (Bowring-Lossock, 2006; Martin et al., 2013; McKenna,
offered to male or female patients or may be more readily accepted by Maguire, & Martin, 2016). There are several risk assessment instru-
either males or females. For example, some studies have reported that ments that have demonstrated validity with regard to predicting
sensory modulation was used more by female patients (Bobier et al., imminent aggression in the inpatient setting. The Brøset Violence
2015; Novak, Scanlan, McCaul, MacDonald, & Clarke, 2012) whilst Checklist (BVC) and the Dynamic Appraisal of Situational Aggres-
another study found that PRN medication was administered nearly sion (DASA) are valid instruments that have been recommended by
twice as often to male patients (Nicholls et al., 2009). There is little NICE (2015); these instruments are conceptually and structurally
advice offered in the literature differentiating the effectiveness, appli- similar and often used by nurses. Both the BVC and DASA are
cation or selection of interventions for male and female patients. quick and easy to use. Each incorporates dynamic items (including
Proactive approaches should be used to ameliorate risk (Taylor some overlapping items) allowing the possibility of regular and
et al., 2011) and these strategies should be used in a manner that is timely assessment of fluctuating risk levels, which can then prompt
suitable for the patient and their level of risk. The principles of preven- the intervention to prevent aggressive behaviours (Chu, Daffern, &
tion outlined by the World Health Organization (WHO) (Krug, Dahl- Ogloff, 2013).
berg, Mercy, Zwi, & Lozano, 2002) classify interventions according to Several studies have investigated the psychometric properties of
three levels: primary, secondary and tertiary. Primary prevention the BVC (Almvik, Woods, & Rasmussen, 2000; Bjorkdahl, Olsson, &
approaches include interventions aimed at preventing aggression Palmstierna, 2005; Hvidhjelm, Sestoft, Skovgaard, & Bue Bjorner,
before it develops. Secondary prevention approaches are recom- 2014; Yao et al., 2014) and the DASA (Barry-Walsh, Daffern, Duncan,
mended when aggression is considered to be imminent, and tertiary & Ogloff, 2009; Chu et al., 2013; Dumais, Larue, Michaud, & Goulet,
interventions are used when aggression is occurring and needs to be 2012; Griffith, Daffern, & Godber, 2013; Vojt, Marshall, & Thomson,
controlled to reduce its harmful effects. Priority should be given to 2010). Two cluster randomised controlled trials have demonstrated
interventions that prevent aggression from occurring in the first that the use of the BVC resulted in less aggression and a decreased
instance, using primary intervention strategies (Krug et al., 2002). use of restrictive practices (e.g., seclusion, mechanical restraint and the
Due to the risks (e.g., physical injury, positional asphyxia, the forced injection of medication) (Abderhalden et al., 2008; van de
experience of trauma) that are associated with some of the tertiary Sande et al., 2011). In the van de Sande et al. (2011) study, experi-
interventions commonly used to manage aggression (e.g., rapid tran- mental wards had a significantly lower number of aggressive incidents
quilisation, restraint and seclusion), the National Institute of Clinical and a lower number of patients engaging in aggression following intro-
Excellence’s (NICE) guidelines for the short-term management of vio- duction of the BVC. Some possible reasons for the results were noted;
lence and aggression suggest that these practices should only be risk assessment led to earlier identification of imminent aggression,
considered once other interventions (such as de-escalation) have leading to “more focused discussions on how to deal with observed
failed to effectively calm the person (NICE, 2015). changes in risks, such as timely verbal de-escalation, behavioural limit
Internationally, there is increasing pressure to reduce the use of setting, close observation and reintegration to the ward after seclu-
restrictive practices. It is now widely acknowledged that these prac- sion” (p. 474).
tices are not therapeutic (Department of Health, 2013; Finke, 2001; Although the predictive validity of the BVC and DASA are estab-
LeBel & Goldstein, 2005), have the potential to cause physical and lished and the impact of the BVC on aggression and the use of
psychological harm (Bonner, Lowe, Rawcliff, & Wellman, 2002; Lam, restrictive practices have been demonstrated, there has been no
2002; Ryan & Happell, 2009; Sullivan et al., 2005) and may inadver- study focusing specifically on the DASA and nursing interventions,
tently encourage rather than prevent aggression (Daffern et al., 2007). including (i) whether the use of the DASA encourages nurses to
Tertiary interventions (that may involve the use of restrictive prac- modify their response and initiate interventions to prevent aggres-
tices) should therefore only be used as a last resort after other inter- sion; (ii) whether different levels of assessed risk are associated with
ventions have been offered, tried and excluded (McKenna, Furness, & different interventions; and (iii) the effectiveness of these nursing
Maguire, 2014). However, what constitutes last resort is often unclear interventions for patients at different levels of risk. There also
(Riahi, Thomson, & Duxbury, 2016). The Royal Australian and New remains uncertainty about the impact of both measures on nursing
Zealand College of Psychiatrists (2010) have commented on the lack initiated interventions and the effectiveness of specific interventions
of staff knowledge and skills in the use of nonrestrictive interventions. at different risk levels for males and females. One of the main chal-
Other researchers have highlighted that there is currently limited lenges facing nurses in providing care following risk assessment
research and evaluation of aggression prevention interventions (Fos- involves the planning and implementation of interventions that are
ter, Bowers, & Nijman, 2007; Kynoch, Wu, & Chang, 2010). As a result, designed to reduce the risk, while also trying to avoid the use of
aggression is often managed in a reactive and uncoordinated manner. restrictive practices.
e974 | MAGUIRE ET AL.

likelihood of imminent aggression, and a final risk rating is translated

2.3 | Aim of the study
into one of three risk bands, low (0), moderate (1–3) and high (4 or
The aim of this study was to investigate what nursing interventions greater) (Maguire et al. 2017). The bands serve as a guide for inter-
were used to prevent aggression in male and female inpatients in a preting the scores and assist in indicating the level of risk (Ogloff &
forensic inpatient setting. It also sought to determine whether com- Daffern, 2004). The DASA, while identifying patients who are at
pletion of a structured aggression risk assessment instrument (DASA) imminent risk of aggression, currently does not provide any proce-
and the risk bands it yields is associated with initiation of more or dures or guidance about interventions following the DASA assess-
different aggression prevention interventions. This study was also ment. At TEH, the DASA assessment is generally scored each day at
designed to examine the effectiveness of interventions at different around midday to handover the DASA assessment and for the nurs-
levels of DASA risk (low, moderate and high), and establish whether ing team to discuss possible intervention strategies.
there are interventions that are more effective at different risk levels
for males and females.
3.3 | Ethical issues
Access to conduct the study was granted by the Forensicare Opera-
3 | METHOD
tional Research Committee. Ethical approval to conduct the study
was granted by the Swinburne University of Technology Human
3.1 | Setting and participants
Research Ethics Committee (SUHREC) (SHR Project 2014/199). The
Data were collected in a statewide inpatient forensic mental health study was conducted in compliance with the ethical guidelines of
service in Victoria, Australia, the Thomas Embling Hospital (TEH). The the Declaration of Helsinki (World Medical Association, 2001). No
hospital has 116 beds and provides assessment and treatment for ethical issues were encountered during the study.
patients who have been found not guilty by reason of mental impair-
ment, patients who may be sentenced or on remand from prison and
3.4 | Data collection
who require mental health inpatient assessment and treatment and
patients who may be considered high risk of offending who are Patient demographic information was gathered from patient files, as
referred from civil mental health services. Thirty male and 30 female was the data for the first 60 days of admission. The following informa-
patients admitted to the acute mental health units (there are two 15- tion was collected: (i) DASA scores on each day of each participant’s
bed male acute units with a total of 30 beds, and one 10-bed female hospital stay, (ii) the nursing interventions that were documented in
unit) between 2009–2013 and who remained in the hospital for at the following 24 hr after the DASA assessment prior to any act of
least 60 days were randomly selected for inclusion. The sample size aggression (if one occurred) and (iii) if aggression occurred in the sub-
was selected to provide a large enough number of patients and days sequent 24 hr. Aggression was documented and categorised according
of observation (3,600 patient days) and included an equal number of to the following groupings: (i) raised voices/verbal aggression/verbal
males and female patients so that comparisons between genders were altercation/threats, (ii) hitting/pushed another person, (iii) use of
possible. The data set used for this study has been described in an ear- weapon or physical threats to attack another person, or an attack on
lier paper (Maguire et al. 2017), which focused on the predictive valid- another person, and (iv) throwing, striking, kicking/hitting or damaging
ity of DASA; however, the focus of this study is associated with objects/furnishings/fittings (Royal College of Psychiatrists, 2007). Any
nursing interventions and their impact on aggression, which has not questions about the classifications for aggression were discussed by
previously been reported. TM and MD until a consensus was reached.

3.2 | Instruments 3.5 | Data analysis

Statistical analyses were conducted using the Statistical Package for
3.2.1 | The dynamic appraisal of situational
Social Sciences version 23 (SPSS, Chicago, IL, USA). Cross-tabulations
aggression
that report proportions and frequencies were used to explore associa-
Dynamic Appraisal of Situational Aggression contains seven items: tions between patients who received an intervention and whether
negative attitudes, impulsivity, irritability, verbal threats, sensitivity they went on to engage in aggression; however, due to the correlated
to perceived provocation, easily angered when requests are denied nature of the data, any further statistical inference was made using
and unwillingness to follow directions. These items are all indepen- binary generalised estimating equation (GEE) models with the binomial
dently moderately related to aggression within 24 hr following the family and logit link with an exchangeable working correlation matrix.
DASA assessment (AUC > 0.70) (Ogloff & Daffern, 2006). The DASA The GEE models were undertaken to evaluate the relationship
takes approximately five minutes to complete and the assessments between the use of intervention strategies and aggression. The seven
involve a nurse trained in the use of the DASA scoring each item for intervention strategies were binary variables (Yes/No) and GEE mod-
its presence or absence in the 24 hr prior to assessment (Daffern els are reported as odds ratios (OR). Odds ratios were generated from
et al., 2009). DASA items are then totalled to determine the the exponentiated values of the beta estimates of the binary GEE
MAGUIRE ET AL. | e975

models. Due to the exploratory nature and the use of multiple GEE that they had provided one type of intervention in the 24 hr follow-
models, the level of significance (a) was set to a value of .05. ing DASA assessment and prior to an act of aggression (n = 754;
36.9%), followed by two interventions 29.6% (n = 604), three inter-
ventions 20.9% (n = 426), four interventions 8% (n = 164), five inter-
4 | RESULTS
ventions 3.7% (n = 75) and six interventions <1% (n = 18). The most
commonly documented intervention was PRN medication 35.9%
4.1 | Demographic and clinical characteristics of
(n = 733), followed by reassurance 18.1% (n = 369), distraction
participants
10.9% (n = 223), limit setting 10.5% (n = 214), one-to-one nursing
The mean age of the sample was 36.8 years (SD = 10.4, ranging 9.7% (n = 198), increased observations 9.1% (n = 185) and de-esca-
from 22–68); for males, the mean age was 34.5 (SD = 10.3, ranging lation 5.8% (n = 119).
from 22–68), and the mean age for females was 39.2 (SD = 10.2, Figure 1 reports the percentage of interventions that were pro-
ranging from 26–68). Diagnoses were as follows: schizophrenia 70% vided to males and females. Patients are reported multiple times in
(n = 42), schizoaffective disorder 13.3% (n = 8), first-episode psy- any or all of the three DASA bands (low, moderate, high). Pearson’s
chosis 5% (n = 3), major depressive disorder 3.3% (n = 2), personality chi-squared tests were considered to explore associations in both
disorder 3.3% (n = 2), schizophreniform 1.7% (n = 1), organic psy- Figure 1 and Table 1 but are not reported because these tests do
chosis 1.7% (n = 1) and bipolar disorder 1.7% (n = 1). not account for the correlated nature of these data. In Figure 1,
there does not appear to be an association between gender and the
use of distraction, or for gender and de-escalation. A possible associ-
4.2 | Aggression
ation appears to exist between gender and limit setting, with more
There were 546 days from a total possible 3,600 patient days when females receiving limit setting as an intervention than males. There
some form of aggression occurred meaning patients did not engage also appears to be an association between gender and one-to-one
in aggressive behaviour on 84.8% of days. When aggression was nursing, with more females receiving one-to-one nursing than males.
documented, 79.9% incidents (n = 436) were categorised as raised Likewise, there appears to be an association between gender and
voices/verbal aggression/verbal altercation/threats; 11.7% (n = 64) reassurance, with more females receiving reassurance than males.
were categorised as throwing, striking, kicking/hitting or damaging More males received PRN medication than females. However, there
objects/furnishings/fittings; 6.4% (n = 35) were classified as hitting/ also appears to be a possible association between gender and obser-
pushing another person, and 2% (n = 11) were incidents involving vations, with more males receiving observations than females.
the use of a weapon or threatening, or attacking another person.

4.4 | DASA assessments

4.3 | Nursing interventions
There were 2,175 DASA risk assessments completed (60.4%, from a
In total, there were 1,257 patient days when interventions were possible 3,600 patient days). The DASA ratings in order from highest
documented and 2,041 interventions were documented (1 to a maxi- number to lowest number recorded were DASA 0 (n = 1,300), DASA
mum of six interventions daily). Nurses most commonly documented 1 (n = 225), DASA 2 (n = 201), DASA 3 (n = 123), DASA 7

80
Males Females
68.9
70 65.1

60 54.8
51.8 53.3 51.5
% of interventions

48.2 48.5 50 50
50 45.2 46.7

40 34.9
31.1
30

20

10

0
Medication Reassurance Observations Distraction Limit setting 1-1 nursing De-escalation
n = 733 n = 369 n = 183 n = 199 n = 214 time n = 198 n = 108
FIGURE 1 Percentage of interventions
by gender Interventions
e976 | MAGUIRE ET AL.

T A B L E 1 Proportions of reported acts of patient aggression following an intervention

Males Females

Intervention Aggression n Proportion, %* Aggression n Proportion, %*

Limit setting
Low Band No 519 6.4 No 742 5.1
Yes 17 11.8 Yes 22 27.3
Moderate Band No 241 17.8 No 255 22.7
Yes 34 29.4 Yes 21 42.9
High Band No 151 49.0 No 110 52.7
Yes 26 65.4 Yes 37 54.1
Medication
Low Band No 434 5.3 No 708 4.8
Yes 102 11.8 Yes 56 17.9
Moderate Band No 171 21.1 No 230 23.0
Yes 104 16.3 Yes 46 30.4
High Band No 87 52.9 No 90 51.1
Yes 90 50.0 Yes 42 56.1
Reassurance
Low Band No 494 6.5 No 694 5.3
Yes 42 7.1 Yes 70 10.0
Moderate Band No 236 16.9 No 234 21.8
Yes 39 33.3 Yes 42 38.1
High Band No 147 49.0 No 119 50.4
Yes 30 63.3 Yes 28 64.3
Distraction
Low Band No 516 6.4 No 735 5.6
Yes 20 10.0 Yes 29 10.3
Moderate Band No 246 18.7 No 259 23.2
Yes 29 24.1 Yes 17 41.2
High Band No 161 52.8 No 127 52.0
Yes 16 37.5 Yes 20 60.0
De-escalation
Low Band No 523 6.5 No 742 5.7
Yes 13 7.7 Yes 22 9.1
Moderate Band No 262 18.7 No 265 24.2
Yes 13 30.8 Yes 11 27.3
High Band No 170 51.2 No 141 51.8
Yes 7 57.1 Yes 6 83.3
Observations
Low Band No 492 6.5 No 752 5.7
Yes 44 6.8 Yes 12 8.3
Moderate Band No 250 19.2 No 263 24.0
Yes 25 20.0 Yes 13 30.8
High Band No 170 51.8 No 137 51.1
Yes 7 42.9 Yes 10 80.0
One-one
Low Band No 508 6.3 No 724 5.7
Yes 28 10.7 Yes 40 7.5

(Continues)
MAGUIRE ET AL. | e977

TABLE 1 (Continued)
Males Females

Intervention Aggression n Proportion, %* Aggression n Proportion, %*

Moderate Band No 247 18.6 No 259 23.6
Yes 28 25.0 Yes 17 35.3
High Band No 164 50.6 No 131 51.1
Yes 13 61.5 Yes 16 68.8

*Patients are reported multiple times in any or all of the three Dynamic Appraisal of Situational Aggression bands (Low, Moderate, High).

(n = 109), DASA 4 (n = 85), DASA 5 (n = 77) and a DASA score of 6 67.6% (n = 219). Results from a GEE model (n = 2,175) suggest there
(n = 53). According to the specified DASA risk bands, 59.8% was a significant association between the DASA bands and the inter-
(n = 1,300) were low DASA ratings (DASA score of 0), 25.2% ventions provided (v2(2) = 110.64, p < .0001). There were more inter-
(n = 549) were moderate DASA ratings (DASA score of 1-3) and ventions provided as the level of risk increased. Figure 2 shows the
14.9% (n = 324) were high DASA ratings (DASA scores 4–7). When number of interventions that were provided per DASA risk band.
a DASA was completed, there was a total of 1,347 patient days
when there were no documented interventions, either documented
4.6 | Examining the effectiveness of interventions
as planned following the DASA assessment or in the subsequent
at different DASA bands
24 hr. Of these, 71.6% (n = 964) of these ratings were in the low
DASA band, 20.6% (n = 278) were in the moderate DASA band, and Using cross-tabulation, the difference between those patients identi-
7.8% (n = 105) were in the high-risk band (including 29 ratings of a fied as having the intervention and then engaging in aggression were
DASA score of 7, where no intervention was documented). compared for each DASA band (results are displayed in Table 1).
Out of the 1,257 times when interventions were documented, a Exploratory descriptive results suggested that higher proportions of
DASA risk assessment had been completed and nursing staff then females who were offered limit setting in the low and moderate
documented intervention strategies to prevent aggression a total of bands subsequently engaged in aggression. Higher proportions of
828 times. As the data reported are correlated, such that a patient is males and females in the low DASA band engaged in aggression
counted on more than one occasion, chi-squared tests from GEE after being offered medication, and higher proportions of both males
models were used. GEE model results suggest there was a significant and females in the moderate-risk band engaged in aggression follow-
association between DASA completions and interventions provided, ing reassurance. To draw further inference on these possible associa-
(v2 (1) = 21.35, p < .0001), whereby the patients who had a DASA tions, usual chi-squared tests were deemed inappropriate due to the
assessment (38.0%) had more interventions provided as compared to correlated nature of these data. Hence, binary GEE models were
those without a DASA assessment (30.2%). Males had more docu- performed and are displayed as odds ratios (OR) in Tables 2 and 3.
mented interventions (40.33%, n = 726) than the females (29.50%, Each of the seven intervention strategies displayed were analysed as
n = 531), v2(1) = 3.05, p = .081). separate models. In the low-risk band, limit setting was significant,
meaning there was an increased likelihood of aggression for females
when limit setting was used. Medication was also significant in the
4.5 | Interventions provided at different risk bands
low-risk band; therefore, there was also an increased likelihood of
There were 2,175 occasions when a DASA was recorded. As the aggression for males and females when PRN was administered in the
DASA band level increased so did the percentage of interventions pro- low-risk band. In the moderate-risk band, there was an increased
vided within the DASA bands, with the low band at 25.8% (n = 336), likelihood of aggression for females when limit setting was used.
the moderate band at 49.5% (n = 273) and the high DASA band at Reassurance was also significant for both males and females in the

250
Number of intervenons

201

200
158

150
112
96

87

82

100
68

56
52

52
49
46
39
36

35
34
30
28

50
23
19

14

0
PRN Reassurance Limit seng One-One Distracon ObservaonsDe-escalaon
F I G U R E 2 Interventions per Dynamic Intervenons
Appraisal of Situational Aggression band DASA low band DASA moderate band DASA high band
e978 | MAGUIRE ET AL.

T A B L E 2 Separate generalised
Males Females
estimating equation models for the seven
95% CI 95% CI interventions documented in the low
OR p OR p
Dynamic Appraisal of Situational
Limit setting 1.585 0.567 4.283 .363 4.178 1.156 15.104 .029*
Aggression band by males (n = 30) and
No limit setting 1 1 females (n = 30)
Medication 2.025 1.122 3.654 .019* 4.275 1.711 10.679 .002*
No medication 1 1
One-one 1.568 0.473 5.197 .462 1.226 0.476 3.157 .674
No one-one 1 1
Reassurance 1.009 0.272 3.747 .989 1.838 0.969 3.485 .062
No reassurance 1 1
Distraction 0.868 0.331 2.278 .774 1.767 0.691 4.521 .235
No distraction 1 1
De-escalation 0.877 0.157 4.852 .877 1.378 0.429 4.425 .590
No de-escalation 1 1
Observation 1.023 0.336 3.111 .968 1.607 0.385 6.707 .515
No observations 1 1

*<.05.

T A B L E 3 Separate generalised
Males Females
estimating equation models for the seven
95% CI 95% CI interventions documented in the moderate
OR p OR p
Dynamic Appraisal of Situational
Limit setting 1.863 0.576 6.025 .299 2.411 1.045 5.562 .039*
Aggression band by males (n = 27) and
No limit setting 1 1 females (n = 27)
Medication 0.786 0.416 1.488 .460 1.115 0.541 2.299 .769
No medication 1 1
One-one 1.379 0.688 2.675 .379 1.548 0.771 3.109 .219
No one-one 1 1
Reassurance 2.423 1.144 5.135 .021* 2.092 1.291 3.388 .003*
No reassurance 1 1
Distraction 1.302 0.434 3.910 .638 2.018 0.899 4.533 .089
No distraction 1 1
De-escalation 1.677 0.690 4.073 .254 1.209 0.403 3.629 .735
No de-escalation 1 1
Observations 1.011 0.350 2.914 .984 1.646 0.577 4.695 .352
No observations 1 1

*<.05.

moderate band, with an increased likelihood of aggression. There moderate and high) for males and females to determine whether
were no significant results for the high-risk band. there may be interventions that are more effective at preventing
aggression in different risk bands.

5 | DISCUSSION
5.1 | Nursing interventions
The aim of this study was to elucidate the nursing interventions that This study offers an insight into the types of interventions used to
were used to prevent aggression in male and female patients in an prevent aggression in a forensic inpatient setting. The use of PRN
acute forensic setting, whether different and more interventions medication was the most commonly documented intervention, con-
were associated with higher risk levels, and whether these interven- sistent with studies by Haw and Wolstencroft (2014) and Richardson
tions prevented aggression. This study also aimed to examine the et al. (2015). PRN medication may have been the most common
effectiveness of interventions at different levels of risk (low, type of intervention documented as there are more stringent
MAGUIRE ET AL. | e979

requirements in regard to documentation of medication, whereas patients are perceived as less dangerous than their male counter-
psycho-social interventions such as those that involve engagement parts, or perhaps as being more receptive to interpersonal
may be documented less reliably. While the use of PRN medication approaches.
can be an effective intervention, it is generally considered that medi-
cation should be used after other nonpharmacological interventions
5.4 | Impact of nursing interventions
have failed, due to potential side effects (Usher & Luck, 2004). Possi-
ble reasons for the use of PRN over other interventions may include Some interventions were associated with an increased likelihood of
the pressures on nurses due to the nature of busy inpatient wards aggression, but none that prevented aggression. For males in the
where PRN medication may be favoured over more time-consuming low-risk band, the use of PRN medication was associated with an
interventions (Barlow, 2014). increased likelihood of subsequently engaging in aggression. In the
moderate-risk band for males, the use of reassurance as an interven-
tion was also associated with being more likely to engage in aggres-
5.2 | Interventions per risk band
sion. For females in the low-risk band, PRN medication was also
As the DASA score increased so did the number of interventions associated with being more likely to engage in aggressive behaviour,
that were documented, which corresponds with the intention of the along with limit setting. In the moderate band, the use of the inter-
DASA, which is to assist in prompting clinicians to start planning and ventions reassurance and limit setting was all associated with a
providing suitable interventions to reduce the risk of aggression female being more likely to engage in aggression. Inpatient, aggres-
(Ogloff & Daffern, 2004). While the literature describing interven- sion may in part be associated with the quality of the interaction
tions for preventing aggression and/or restrictive interventions often between staff and patients (Lancee et al., 1995), and in particular,
mentions the need to use several interventions (e.g., Bowers et al., limit setting has been linked with aggressive responses from patients,
2015; Gaskin, Elsom, & Happell, 2007; NICE, 2015; Stewart, Van der which may in part be due to the manner in which limits have been
Merwe, Bowers, Simpson, & Jones, 2010), little attention has been set. For example, a more authoritarian style of limit setting may
paid to which interventions should be used, in which combination, engender a hostile response from patients, whereas limit setting
whether males and females should be cared for differently with using an authoritative approach may enhance positive outcomes
regard to aggression prevention and when interventions should be (Maguire et al., 2014). Patients may also view limit setting as an
initiated. As seen in this study, the majority of interventions are pro- intervention that is more restrictive in nature; as such, this interven-
vided once behaviour reaches a threshold within the high-risk band. tion should be reserved for when the level of risk has escalated and
While it seems reasonable that there would be more interventions is imminent rather than when the person is presenting in the low-
applied in the high-risk band, it may be more effective to intervene and moderate-risk bands.
earlier to prevent the aggression risk from escalating (Krug et al., While assessment can alert nurses to imminent risk, preventing
2002). Intervening early may also increase the efficacy of interven- patients from engaging in aggressive behaviour will often require
tions as the patient may be more receptive (Fluttert et al., 2008). nursing intervention along with certain resources and/or procedures
When a person is assessed as being in the high-risk DASA band, (Kling, Yassi, Smailes, Lovato, & Koehoorn, 2010), which might
they are already in a state of irritability and disagreeableness (Barry- include additional staff to facilitate interventions such as close obser-
Walsh et al., 2009), which may impact on their willingness to engage vations, equipment for distraction or space on the unit suitable for
in the intervention. engaging in one-to-one nursing, distraction techniques and de-esca-
lation. However, the engagement and selection of appropriate inter-
ventions (from primary to tertiary) for patients who present as a risk
5.3 | Comparing the interventions provided for
of engaging in aggression remains a significant challenge for nurses
males and females
working in the acute inpatient setting.
Our findings suggest there were differences between males and This study found that the impact of interventions differs for male
females in relation to the type of interventions provided. Similar and female patients. The differences could be due to how the inter-
to the findings by Nicholls et al. (2009), there was a tendency to vention is performed, and when the intervention is applied (is an
provide males in this study with more restrictive interventions; intervention such as limit setting being applied too early, e.g., when
more PRN medication was administered to males as compared to someone is in the low-risk band when primary interventions should
females, and they were also more likely to be subjected to be instigated, or perhaps they are initiated too late, such as the use
increased observations. While the reasons for this discrepancy are of reassurance in the moderate DASA band when this intervention
unclear, others have noted that these differences are explained by may have been more successful when a person was at low risk). The
males being perceived to be more dangerous (Wynn, 2002). lack of empirical research along with a lack of clear definitions of
Females in this study were more likely than the males to receive interventions, and clear procedures that articulate how to success-
one-to-one nursing, reassurance and limit setting, which by their fully apply these nursing interventions for males and females at dif-
very nature involve a more interpersonal approach and have an ferent risk levels, is hampering preventative action (Barlow, 2014;
emphasis on communication. This may be because female forensic Johnson & Hauser, 2001; Roberton et al., 2012).
e980 | MAGUIRE ET AL.

Following the DASA assessment, a plan needs to be made about interventions and acts of aggression. Using this methodology, we
how to prevent aggression, taking into account knowledge of what may not have provided a comprehensive representation of the care
works for each patient, with regard to his or her gender and risk that was provided, and may have failed to capture some of the other
level. Patient preferences may also be usefully considered although factors that influence aggression and its prevention (e.g., staffing,
this was not a focus of this study. When the DASA was first devel- ward milieu). The reporting of nursing interventions and aggression
oped, guidance suggested the intensity of the interventions should is reliant on the quality of the documentation, and it is likely that
correspond with the level of risk (Ogloff & Daffern, 2006). The both nursing interventions and aggression were underreported.
DASA instrument in its current form does not provide any sugges- Another limitation is that patient risk levels may have changed from
tions or framework in regard to what might be suitable nursing inter- the time of the DASA assessment, to the time that the intervention
ventions at each DASA risk band. However, a framework including was applied, which may mean some of the interventions were actu-
primary, secondary and tertiary prevention linked to DASA risk ally appropriate if the person’s risk had increased. The modest num-
bands may enhance the use of the DASA, providing structure for ber of patients in this study and the forensic mental health setting
nursing intervention. may also limit the generalisability of these results to other hospital
In the light of our findings, perhaps the focus may need to be on settings.
engaging patients and applying primary interventions (such as one-to-
one nursing, distraction techniques and reassurance) in the low-risk
band. In the moderate DASA risk band, interventions might also 6 | CONCLUSION
include the use of de-escalation techniques as indicated (and possibly
distraction techniques and one-to-one nursing). Finally, for patients Although numerous interventions for managing aggression are men-
who are assessed as being in the high DASA band, tertiary interven- tioned in the literature and documented in practice, some interven-
tions (along with primary and secondary interventions) might include tions appear contraindicated for males and females assessed as low
the use of PRN medication, increased observation and limit setting. to moderate risk of aggression. Further, these results suggest that
These are more restrictive interventions and should be used only there is no particular intervention that is particularly effective for
after other less restrictive interventions have been tried and when preventing aggression for patients at high risk of imminent aggres-
risk of aggression is imminent. The more restrictive interventions such sion; it is likely that a combination of various interventions is
as the use of PRN medication, increased observations and limit set- required to prevent aggression for high-risk patients (e.g., offering
ting appear unsuitable when a patient is presenting as low or moder- reassurance, providing PRN medication and observing and engaging
ate risk. These interventions are restrictive and may be viewed as with the patient). This is reflected by the higher number of interven-
unnecessary and provocative by patients at low risk. They should be tions being initiated for patients who are assessed in the moderate-
used after other less restrictive (primary and secondary) practices to high-risk bands. Gender may have an influence on the type of
have been attempted. Attention should also be given to the choice interventions provided. To improve patient care and prevent inpa-
and application of interventions in the moderate-risk band. In this tient aggression, attention needs to be directed towards improving
study, limit setting (for females) and reassurance for males and the interventions designed to reduce aggression and the use of
females were associated with an increased risk of aggression. restrictive interventions. Early intervention should be a priority.

5.5 | Completion of risk assessment

7 | RELEVANCE TO CLINICAL PRACTICE
Despite risk assessment being considered an important aspect of
forensic mental health nursing (McKenna et al., 2016), and a require- Findings from this study can be used to enhance clinical practice in
ment for all patients admitted to the acute units in this service, on several ways. Given the importance of assessing for risk of imminent
40% of the days in this study a DASA was not completed. Compet- aggression, patients in the acute inpatient setting should have a daily
ing demands for nurses, documentation not being completed, and risk assessment completed by nurses. Once a risk assessment has
staff not valuing structured risk assessment may explain this result. been completed, a plan should be developed about how to manage
Identifying when patients might be at risk of engaging in aggressive the risk; this should take into account the patient’s gender and risk
behaviour is an important component of forensic mental health nurs- level. Consideration needs to be given to the type of intervention
ing practice and should be done using an evidence-based approach and how it is applied (e.g., how reassurance is provided, and how
(Bowring-Lossock, 2006). Further research is required to understand limits are set—this requires elucidation of best practice and consis-
why risk assessments are not completed when required. tency in implementation). As noted, many of the commonly sug-
gested interventions are not defined clearly and practice can vary
considerably (Roberton et al., 2012). Staff training in relation to the
5.6 | Limitations
details of the interventions and when to apply interventions could
The most significant limitation of this study was its reliance on docu- possibly enhance the effectiveness, consistency and documentation
mented notes as the means of collecting the data for nursing of these interventions. Consideration also needs to be given to
MAGUIRE ET AL. | e981

intervening as early as possible and in the least restrictive manner, Bjørkly, S. (1999). A ten-year prospective study of aggression in a special
using primary, secondary and tertiary interventions according to risk secure unit for dangerous patients. Scandinavian Journal of Psychology,
40, 57–63. https://doi.org/10.1111/sjop.1999.40.issue-1
level. Early intervention may be more successful at averting aggres-
Bobier, C., Boon, T., Downward, M., Loomes, B., Mountford, H., & Swadi,
sion as patients may be more receptive, although care should be H. (2015). Pilot investigation of the use and usefulness of a sensory
given not to intervene with low-risk patients using restrictive/ter- modulation room in a child and adolescent psychiatric inpatient unit.
tiary strategies. Occupational Therapy in Mental Health, 31(4), 385–401. https://doi.
org/10.1080/0164212X.2015.1076367
Bonner, G., Lowe, T., Rawcliff, D., & Wellman, N. (2002). Trauma for all:
ACKNOWLEDGMENTS A pilot study of the subjective experience of physical restraint for all
mental health inpatients and staff in the UK. Journal of Psychiatric
The authors would like to acknowledge Forensicare’s support for TM and Mental Health Nursing, 9, 465–473. https://doi.org/10.1046/j.
1365-2850.2002.00504.x
through the further study incentive scheme. We are also very grate-
Bowers, L., James, K., Quirk, A., Sugar, , Stewart, D., & Hodsoll, J. (2015).
ful for the ongoing support and assistance from Ms. Jo Ryan, Reducing conflict and containment rates on acute psychiatric wards:
Director of Nursing at Forensicare and Mr. Murray Bruce for his The Safewards cluster randomised controlled trial. International Jour-
insightful comments on an early draft. nal of Nursing Studies, 52, 1412–1422. https://doi.org/10.1016/j.ijnur
stu.2015.05.001
Bowers, L., Stewart, D., Papadopoulos, C., Dack, C., Ross, J., Khanom, H.,
CONTRIBUTIONS & Jeffery, D. (2011). Inpatient violence and aggression: A literature
review. London: Kings College. Retrieved from http://www.kcl.ac.uk/
Substantial data analysis, manuscript preparation: TM; statistical iop/depts/hspr/research/ciemh/ mhn/projects/litreview/LitRe-
analysis, manuscript preparation and revision: MD, SB, BM. vAgg.pdf
Bowring-Lossock, E. (2006). The forensic mental health nurse – A litera-
ture review. Journal of Psychiatric and Mental Health Nursing, 13(6),
780–785. https://doi.org/10.1111/jpm.2006.13.issue-6
AUTHORSHIP
Chu, C. M., Daffern, M., & Ogloff, J. R. P. (2013). Predicting aggression in
All authors listed met the criteria according to the guidelines of the acute inpatient psychiatric setting using BVC, DASA, and HCR-20
Clinical scale. The Journal of Forensic Psychiatry & Psychology, 24(2),
International Committee of Medical Journal Editors and are in
269–285. https://doi.org/10.1080/14789949.2013.773456
agreement with the manuscript. Curtis, J., Baker, J. A., & Reid, A. R. (2007). Exploration of therapeutic inter-
ventions that accompany the administration of p.r.n. (‘as required’) psy-
chotropic medication within acute mental health settings: A
CONFLICT OF INTEREST retrospective study. International Journal of Mental Health Nursing, 16,
318–326. https://doi.org/10.1111/j.14470349.2007.00487.x
The authors report no financial support or conflict of interest. Cutcliffe, J. R., & Riahi, S. (2013). Systemic perspective of violence and
aggression in mental health care: Towards a more comprehensive
understanding and conceptualisation: Part 1. International Journal of
ORCID Mental Health Nursing, 22, 558–567. https://doi.org/10.1111/inm.
12029
Tessa Maguire http://orcid.org/0000-0002-1050-6094 Daffern, M., Howells, K., Hamilton, L., Mannion, A., Howard, R., & Lilly,
M. (2009). The impact of structured risk assessments followed by
management recommendations on aggression in patients with per-
REFERENCES sonality disorder. The Journal of Forensic Psychiatry & Psychology, 20
(5), 661–679. https://doi.org/10.1080/14789940903173990
Abderhalden, C., Needham, l., Dassen, T., Halfens, R., Haug, H.-J., & Fis- Daffern, M., Howells, K., & Ogloff, J. P. (2007). The Interaction between
cher, J. E. (2008). Structured risk assessment and violence in acute individual characteristics and the function of aggression in forensic
psychiatric wards: Randomised controlled trial. The British Journal of psychiatric Inpatients. Psychiatry Psychology and Law, 14(1), 17–25.
Psychiatry, 193(1), 44–50. https://doi.org/10.1192/bjp.bp.107.045534 https://doi.org/10.1375/pplt.14.1.17
Almvik, R., Woods, P., & Rasmussen, K. (2000). The Brøset violence Daffern, M., Maguire, T., Carroll, A., & McKenna, B. (2015). The problem
checklist – Sensitivity, specificity, and interrater reliability. Journal of of workplace violence: A focus in the mental health sector. In A. Day,
Interpersonal Violence, 15(12), 1284–1296. https://doi.org/10.1177/ & E. Fernandez (Eds.), Violence in Australia: Perspectives, policy and
088626000015012003 solutions (pp. 104–116). Annandale: Federation Press.
Barlow, E. (2014). Acute mental health nursing and PRN medication Davidson, S. (2005). The management of violence in general psychiatry.
administration: A review of the literature. Mental Health Nursing, 34 Advances in Psychiatric Treatment, 11, 362–370. https://doi.org/10.
(6), 13–15. 1192/apt.11.5.362
Barry-Walsh, J., Daffern, M., Duncan, S., & Ogloff, J. (2009). The predic- Department of Health. (2013). Providing a safe environment for all. Frame-
tion of imminent aggression in patients with mental illness and/or work for reducing restrictive interventions. Melbourne, Vic.: Mental
intellectual disability using the dynamic appraisal of situational Health, Drugs and Regions Division, Department of Health, Victorian
aggression instrument. Australasian Psychiatry, 17(6), 493–496. Government.
https://doi.org/10.1080/10398560903289975 Dumais, A., Larue, C., Michaud, C., & Goulet, M.-H. (2012). Predictive
Bjorkdahl, A., Olsson, T., & Palmstierna, D. (2005). Nurses short-term pre- validity and psychiatric nursing staff’s perception of the clinical use-
diction of violence in acute psychiatric intensive care. Acta Psychi- fulness of the French version of the dynamic appraisal of situational
atrica Scandinavica, 113, 224–229. https://doi.org/10.1111/j.1600- aggression. Issues in Mental Health Nursing, 33(10), 670–675.
0447.2005.00679.x https://doi.org/10.3109/01612840.2012.697254
e982 | MAGUIRE ET AL.

Duxbury, J., Hahn, S., Needham, I., & Pulsford, D. (2008). The Manage- aloud study. Journal of Advanced Nursing, 66(6), 1297–1307. https://d
ment of Aggression and Violence Attitude Scale (MAVAS): A cross- oi.org/10.1111/(ISSN)1365-2648
national comparative study. Journal of Advanced Nursing, 62(5), 596– Maguire, T., Daffern, M., Bowe, S., & McKenna, B. (2017). Predicting
606. https://doi.org/10.1111/j.1365-2648.2008.04629.x aggressive behaviour in acute forensic mental health units: A re-
Finke, L. M. (2001). The use of seclusion is not evidence-based practice. examination of the dynamic appraisal of situational aggression’s pre-
Journal of Child and Adolescent Psychiatric Nursing, 14(4), 186–190. dictive validity. International Journal of Mental Health Nursing, 26,
https://doi.org/10.1111/jcap.2001.14.issue-4 472–481. https://doi.org/10.1111/inm.2017.26.issue-5
Fluttert, F., Van Meijel, B., Webster, C., Nijman, H., Bartels, A., & Gryp- Maguire, T., Daffern, M., & Martin, T. (2014). Exploring nurses’ and
donck, M. (2008). Risk management by early recognition of warning patients’ perspectives of limit setting in a forensic mental health set-
signs in patients in forensic psychiatric care. Archives of Psychiatric ting. International Journal of Mental Health Nursing, 23(2), 153–160.
Nursing, 22(4), 208–216. https://doi.org/org/10.1016/j.apnu.2007.06. https://doi.org/10.1111/inm.12034
012 Martin, T., Maguire, T., Quinn, C., Ryan, J., Bawden, L., & Summers, M.
Foster, C., Bowers, L., & Nijman, H. (2007). Aggressive behaviour on (2013). Standards of practice for forensic mental health nurses –
acute psychiatric wards: Prevalence, severity and management. Jour- Identifying contemporary practice. Journal of Forensic Nursing, 9(3),
nal of Advanced Nursing, 58(2), 140–149. https://doi.org/10.1111/jan. 171–178. https://doi.org/10.1097/JFN.0b013e31827a593a
2007.58.issue-2 Martin, T., & Street, A. F. (2003). Exploring evidence of the therapeutic
Gaskin, C. J., Elsom, S. J., & Happell, B. (2007). Interventions for relationship in forensic psychiatric nursing. Journal of Psychiatric and
reducing the use of seclusion in psychiatric facilities. The British Mental Health Nursing, 10, 543–551. https://doi.org/10.1046/j.1365-
Journal of Psychiatry, 191(4), 298–303. https://doi.org/10.1192/ 2850.2003.00656.x
bjp.bp.106.034538 McKenna, B., Furness, T., & Maguire, T. (2014). A literature review and
Griffith, J. J., Daffern, M., & Godber, T. (2013). Examination of the pre- policy analysis on the practice of restrictive interventions. Melbourne,
dictive validity of the Dynamic Appraisal of Situational Aggression in Vic.: State of Victoria.
two mental health units. International Journal of Mental Health Nurs- McKenna, B., Maguire, T., & Martin, T. (2016). Forensic mental health
ing, 22, 485–492. https://doi.org/10.1111/inm.12011 nursing. In K. Evans, D. Nizette, & A. O’Brien (Eds.), Psychiatric and
Hale, C. A., Thomas, L. H., Bond, S., & Todd, C. (1997). The nursing mental health nursing (4th ed., pp. 315–338). Chatswood, NSW: Else-
record as a research tool to identify nursing interventions. Journal of vier Australia.
Clinical Nursing, 6, 207–214. https://doi.org/10.1111/jcn.1997.6.is Mullen, A., Drinkwater, V., & Lewin, T. J. (2013). Care zoning in a psychi-
sue-3 atric intensive care unit: Strengthening ongoing clinical risk assess-
Haw, C., & Wolstencroft, L. (2014). A study of the use of sedative PRN ment. Journal of Clinical Nursing, 23, 731–743. https://doi.org/10.
medication in patients at a secure hospital. The Journal of Forensic 1111/jocn.12493
Psychiatry and Psychology, 25(3), 307–320. https://doi.org/10.1080/ National Institute for Health and Clinical Excellence (NICE) (2015). Vio-
14789949.2014.911948 lence and aggression: Short-Term management in mental health, health
Hvidhjelm, J., Sestoft, D., Skovgaard, L. T., & Bue Bjorner, J. (2014). Sen- and community settings. London, UK: National Institute for Health
sitivity and specificity of the Brøset Violence Checklist as predictor and Clinical Excellence.
of violence in forensic psychiatry. Nordic Journal of Psychiatry, 68(8), Needham, I., Abderhalden, C., Meer, R., Dassen, T., Haug, H., Halfens, R., &
536–542. https://doi.org/10.3109/08039488.2014.880942 Fischer, J. E. (2004). The effectiveness of two interventions in the man-
Irwin, A. (2006). The nurse’s role in the management of aggression. Jour- agement of patient violence in acute mental health inpatient settings.
nal of Psychiatric and Mental Health Nursing, 13, 309–318. https://doi. Report on a pilot study. Journal of Psychiatric and Mental Health Nursing,
org/10.1111/jpm.2006.13.issue-3 11, 595–601. https://doi.org/10.1111/j.1365-2850.2004.00767.x
Johnson, M. E., & Hauser, P. M. (2001). The practices of expert psychi- Nicholls, T. L., Brink, J., Greaves, C., Lussier, P., & Verdun-Jones, S.
atric nurses: Accompanying the patient to a calmer personal space. (2009). Forensic psychiatric inpatients and aggression. An exploration
Issues in Mental Health Nursing, 22, 651–668. https://doi.org/10. of incidence, prevalence, severity, and interventions by gender. Inter-
1080/016128401750434464 national Journal of Law and Psychiatry, 32, 23–30. https://doi.org/10.
Kling, R. N., Yassi, A., Smailes, E., Lovato, C. Y., & Koehoorn, M. (2010). 1016/j.ijlp.2008.11.007
Evaluation of a violence risk assessment system (the Alert System) Novak, T., Scanlan, J., McCaul, D., MacDonald, N., & Clarke, T. (2012).
for reducing violence in an acute hospital: A before and after study. Pilot study of a sensory room in an acute inpatient psychiatric unit.
International Journal of Nursing Studies, 48, 534–539. Australasian Psychiatry, 20(5), 401–406. https://doi.org/10.1177/
Krug, E. G., Dahlberg, L. L., Mercy, J. A., Zwi, A. B., & Lozano, A. R., & 1039856212459585
World Health Organization. (2002). World report on violence and O’Brien, L., & Cole, R. (2004). Mental health nursing practice in acute
health. Geneva, Switzerland: World Health Organization. psychiatric close-observation areas. International Journal of Mental
Kynoch, K., Wu, C.-J., & Chang, A. M. (2010). Interventions for prevent- Health Nursing, 13, 89–99. https://doi.org/10.1111/inm.2004.13.is
ing and managing aggressive patients admitted to an acute hospital sue-2
setting: A systematic review. Worldviews on Evidence-Based Nursing, Ogloff, J., & Daffern, M. (2004). The dynamic appraisal of situational
8, 76–86. https://doi.org/10.1111/j.1741-6787.2010.00206.x aggression (DASA). Melbourne, Vic.: Centre for Forensic Behavioural
Lam, L. T. (2002). Aggression exposure and mental health among Science, Monash University.
nurses. Australian e-Journal for the Advancement of Mental Health, 1 Ogloff, J. R. P., & Daffern, M. (2006). The dynamic appraisal of situational
(2), 1–12. aggression: An instrument to assess risk for imminent aggression in
Lancee, W. J., Gallop, R., McCay, E., & Toner, B. (1995). The relationship psychiatric inpatients. Behavioural Sciences and the Law, 24, 799–813.
between nurses’ limit setting styles and anger in psychiatric inpa- https://doi.org/10.1002/(ISSN)1099-0798
tients. Psychiatric Services, 6, 609–613. Riahi, S., Thomson, G., & Duxbury, J. (2016). An integrative review
LeBel, J., & Goldstein, R. (2005). The economic cost of using restraint and exploring decision-making factors influencing mental health nurses in
the value added by restraint reduction or elimination. Psychiatric Ser- the use of restraint. Journal of Psychiatric and Mental Health Nursing,
vices, 56(9), 1109–1114. https://doi.org/10.1176/appi.ps.56.9.1109 23(2), 116–128. https://doi.org/10.1111/jpm.12285
MacNeela, P., Clinton, G., Place, C., Scott, A., Treacy, P., Hyde, A., & Richardson, M., Brennan, G., James, K., Lavelle, M., Renwick, L., Stewart,
Dowd, H. (2010). Psychosocial care in mental health nursing: A think D., & Bowers, L. (2015). Describing the precursors to and
MAGUIRE ET AL. | e983

management of medication non adherence on acute psychiatric Taylor, K., Guy, S., Stewart, L., Ayling, M., Miller, G., Anthony, A., . . . Tho-
wards. General Hospital Psychiatry, 37, 606–612. https://doi.org/10. mas, M. (2011). Care zoning a pragmatic approach to enhance the
1016/j.genhosppsych.2015.06.017 understanding of clinical needs as it relates to clinical risks in acute
Roberton, T., Daffern, M., Thomas, S., & Martin, T. (2012). De-escalation and in-patient unit settings. Issues in Mental Health Nursing, 32, 318–326.
limit setting in forensic mental health units. Journal of Forensic Nursing, 8, https://doi.org/10.3109/01612840.2011.559570
94–101. https://doi.org/10.1111/j.1939-3938.2011.01125.x Usher, K., & Luck, L. (2004). Psychotropic PRN: A model for best practice
Royal Australian and New Zealand College of Psychiatrists. (2010). Posi- management of acute psychotic behavioural disturbance in inpatient
tion statement 61, Minimising the use of seclusion and restraint in peo- psychiatric settings. International Journal of Mental Health Nursing, 13,
ple with mental illness. Melbourne, Vic.: Royal Australian and New 18–21. https://doi.org/10.1111/inm.2004.13.issue-1
Zealand College of Psychiatrists. Vojt, G., Marshall, L. A., & Thomson, L. D. G. (2010). The assessment of
Royal College of Psychiatrists (2007). Healthcare commission national imminent inpatient aggression: A validation study of the DASA-IV in
audit of violence 2006–7. London, UK: Royal College of Psychiatrists. Scotland. The Journal of Forensic Psychiatry & Psychology, 21, 789–
Ryan, B., & Happell, B. (2009). Learning from experience: Using action 800. https://doi.org/10.1080/14789949.2010.489952
research to discover consumer needs in post-seclusion debriefing. Wynn, R. (2002). Medicate, restrain or seclude? Strategies for dealing
International Journal of Mental Health Nursing, 18, 100–107. https://d with violent and threatening behaviour in a Norwegian university
oi.org/10.1111/inm.2009.18.issue-2 psychiatric hospital. Scandinavian Journal of Caring Science, 16, 287–
van de Sande, R., Nijman, H. L. I., Noorthoorn, E., Wierdsma, A., Hellen- 291. https://doi.org/10.1046/j.1471-6712.2002.00082.x
doorn, E., van der Staak, C., & Mulder, C. L. (2011). Aggression and World Medical Association (2001). World medical association declara-
seclusion on acute psychiatric wards: Effect of short-term risk assess- tion of Helsinki: Ethical principles for medical research involving
ment. The British Journal of Psychiatry, 199, 473–478. https://doi.org/ human subjects. Bulletin of the World Health Organization 79(4),
10.1192/bjp.bp.111.095141 373–374.
Sheridan, M., Henrion, R., Robinson, L., & Baxter, V. (1990). Precipitant Yao, X., Li, Z., Arthur, D., Hu, L., An, F.-R., & Cheng, G. (2014). Accept-
of violence in a psychiatric inpatient setting. Hospital and Community ability and psychometric properties of Brøset Violence Checklist in
Psychiatry, 41(7), 776–780. psychiatric care settings in China. Journal of Psychiatric and Mental
Stevenson, K. N., Jack, S. M., O’Mara, L., & Le Gris, J. (2015). Registered Health Nursing, 21(9), 848–855.
nurses’ experiences of patient violence on acute care psychiatric
inpatient units: An interpretive descriptive study. Bio Med Central
Nursing, 14, 35. https://doi.org/10.1186/s12912-015-0079-5
Stewart, D., Van der Merwe, M., Bowers, L., Simpson, A., & Jones, J. (2010). How to cite this article: Maguire T, Daffern M, Bowe SJ,
A review of interventions to reduce mechanical restraint and seclusion McKenna B. Risk assessment and subsequent nursing
among adult psychiatric inpatients. Issues in Mental Health Nursing, 31
interventions in a forensic mental health inpatient setting:
(6), 413–424. https://doi.org/10.3109/01612840903484113
Sullivan, A. M., Bezmen, J., Barron, C. T., Rivera, J., Curley-Casey, L., & Associations and impact on aggressive behaviour. J Clin Nurs.
Marino, D. (2005). Reducing restraints: Alternatives to restraints on 2018;27:e971–e983. https://doi.org/10.1111/jocn.14107
an inpatient psychiatric service-utilising safe and effective methods
to evaluate and treat the violent patient. Psychiatric Quarterly, 76(1),
51–65. https://doi.org/10.1007/s11089-005-5581-3