You are on page 1of 18

CNE

Earn Contact Hours

Safewards Training in Victoria, Australia


A Descriptive Analysis of Two Training Methods and
Subsequent Implementation
Justine Fletcher, MPsych(Clin), BPsych(Hons); John Reece, PhD, MAPS;
Stuart A. Kinner, PhD, BA(Hons); Lisa Brophy, PhD, MPolLaw, BBSc, BSW; and
Bridget Hamilton, PhD, RN, BPsychN, BN(Hons)

I
ABSTRACT npatient mental health care has
Safewards is a psychosocial intervention designed to improve practice and staff– been evolving during the past 2 de-
cades in English-speaking countries,
patient interactions in mental health wards. However, evidence regarding the
with substantial efforts being made to
impact of training on implementing change initiatives in this setting is mixed. reduce restriction (Baker et al., 2014;
Pre- and post-training surveys were completed by staff from 18 inpatient wards Duxbury et al., 2019; Ramluggun &
across seven health services in Victoria, Australia. Fidelity audits were undertak- Chalmers, 2018; Wale et al., 2011).
Initiatives to reduce restrictive in-
en to assess implementation of Safewards into routine practice. Staff knowledge,
terventions, such as forced medica-
confidence, and motivation increased significantly from pre- to post-training, tion, seclusion, and restraint (Bowers,
with no difference between two different methods of training. Most wards were 2009; Cowman et al., 2017; Happell
implementing six or more of the interventions at the end of the trial. A struc- & Harrow, 2010), and shifts toward a
recovery-oriented approach have been
tured approach to training, with flexibility of delivery options, produced posi-
two significant policy and practice
tive changes in staff and translation to practice. Substantial investment in train- change initiatives in Australia and
ing from government and organizations appears to strengthen the uptake and internationally (Australian Human
impact of training, and the current study provides evidence that the interven- Rights Commission, 1993; Boumans
et al., 2015; LeBel et al., 2014; Oster
tions were implemented as intended. [Journal of Psychosocial Nursing and Mental
et al., 2016). Evidence-based interven-
Health Services, 58(12), 32-42.] tions to facilitate improvements in in-

32 Copyright © SLACK Incorporated


patient mental health care have been mental health services to achieve the these positive findings. Hence, the cur-
used to support and implement policy goal of reducing the use of restrictive rent study has two aims: (1) to evaluate
and practice changes. Development of interventions, in line with the 2014 whether Safewards training increased
psychosocial interventions and subse- Mental Health Act. In the Victorian staff knowledge, confidence, and mo-
quently training inpatient staff in psy- Safewards Trial, wards in seven health tivation to implement the Safewards
chosocial interventions, is a common services were given funding and access model and 10 interventions; and (2) to
method to enable the transfer of policies to Safewards training via a 3-day train- assess the degree to which training was
into practice (Edmunds et al., 2013). the-trainer workshop. Wards then par- translated into practice. The current
Training mental health staff in psy- ticipated on an opt-in basis to imple- article reports on the evaluation of lo-
chosocial interventions can result in ment Safewards over a 12-week trial cal health service training efforts. The
more positive attitudes toward patients, period, with follow up for 1 year. process of training adopted by local
increased knowledge and confidence of Safewards was developed in the health services in the Victorian Safe-
staff to work in new ways (Bradshaw et United Kingdom after extensive re- wards Trial provides a unique oppor-
al., 2007; Doyle et al., 2007; Meadows views of the literature and empirical tunity to add to the literature about
et al., 2019), implementation of the research regarding conflict and con- Safewards training and the subsequent
target interventions (Redhead et al., tainment events in acute inpatient impact on implementation.
2011), and positive outcomes for con- mental health wards (Bowers, 2014;
sumers (Azeem et al., 2011). However, Bowers et al., 2014). Safewards pro- METHOD
training in psychosocial interventions vides a conceptual framework and 10 Study Setting and Participants
alone is not always sufficient to bring psychosocial interventions designed The current study took place in
about changes in practice (Brennan & to reduce conflict and use of contain- seven health services that opted into
Gamble, 1997; Peters et al., 2013). An ment strategies in acute inpatient a trial of Safewards, involving 18 of
Australian study of transfer of a Col- mental health wards (Bowers, 2014). 68 wards in Victoria. Inpatient men-
laborative Recovery Model training (For a full description of the model, tal health wards providing services to
into mental health services to increase see Fletcher et al. [2017].) Fidelity of adolescent, adult, and aged acute and
recovery orientation found low rates implementation refers to how faith- secure extended care units were in-
of translation of training into routine fully Safewards is reproduced during volved. The anticipated pool of nurs-
practice (37% over 5 months) (Uppal implementation in each context. To ing and allied staff employed across
et al., 2010). Key barriers identified in date, fidelity results are mixed from as the 18 wards at the time of the train-
the study were institutional constraints, few as two or three interventions be- ing was approximately 400 individuals.
lack of staff knowledge and confidence ing implemented (Price et al., 2016), Training took place over a 4-month pe-
after collaborative recovery training, to as many as seven or eight interven- riod, from November 2014 to February
and lack of motivation to implement tions being implemented (Kipping et 2015. It consisted of two parts: a central
the new protocol. This finding illus- al., 2018; Maguire et al., 2018). The train-the-trainer program, followed by
trates that research about training in uptake and quality of training has local health service training.
psychosocial interventions is mixed. been one variable suggested as impact-
One effective model of facilitating dis- ing the level of fidelity of Safewards Train-The-Trainer
semination of evidence-based interven- implementation (Higgins et al., 2018; Three train-the-trainer days were
tions to health care professionals, increas- Kipping et al., 2018; Maguire et al., commissioned by the Victorian Gov-
ing staff knowledge, improving clinical 2018; Price et al., 2016). ernment. Clinical nurse educators
behavior, and improving outcomes for The researchers were commissioned (CNEs) and select ward staff from each
consumers is the train-the-trainer model to conduct an evaluation in the con- health service attended the training,
(Pearce et al., 2012). In this model, an text of “real-world” conditions of the which covered the Safewards model
expert leads a training event involv- inpatient unit, thus the evaluators had and 10 interventions (Table 1).
ing practitioners who then train their limited control of study conditions In this jurisdiction, the CNE is a dis-
colleagues and implement the new in- and were required to navigate com- tinct role, differentiated in the employ-
tervention locally. Pearce et al. (2012) plexities that occur with unexpected ment award and professional structure.
also found that a multifaceted approach changes. This real-world evaluation of Typically, CNEs are nurses recruited
involving interactive components and all components of the Victorian Safe- from clinical specialist roles; they have
course materials was most effective. wards Trial found that seclusion rates mental health specialist postgradu-
were reduced by 36% in the Safewards ate qualifications at either Graduate
THE VICTORIAN SAFEWARDS TRIAL project wards by 12-month follow up Diploma or Master’s level. CNEs are
In 2014 in Australia, the Victo- (Fletcher et al., 2017). Further research members of their service level (local)
rian Government initiated the Vic- was required to illuminate the condi- nurse education team, with access to a
torian Safewards Trial to support tions that were relevant in producing statewide community of practice. The

JOURNAL OF PSYCHOSOCIAL NURSING • VOL. 58, NO. 12, 2020 33


12-Week Implementation Phase
TABLE 1 After the training phase, the trial
SAFEWARDS INTERVENTIONSa phase involved monitoring all wards
Intervention Description Purpose implementing the Safewards interven-
tions during a 12-week period, from
Mutual Help Patients offer and receive Strengthens patient
March to May 2015. The goal stated
Meeting mutual help and support community; opportunity to
by the Victorian Government was that
through a daily, shared give and receive help.
all wards would implement all 10 Safe-
meeting.
wards interventions from the start of
Know Each Other Patients and staff share some Builds rapport, connection, the 12 weeks and would have embed-
personal interests and ideas and sense of common ded Safewards into routine practice af-
with each other, displayed in humanity. ter the 12-week trial.
unit common areas.
Clear Mutual Patients and staff work Counters some power Study Design and Ethical Considerations
Expectations together to create mutually imbalances; creates a The current study used a pragmatic
agreed aspirations that apply stronger sense of shared real-world evaluation design to inves-
to both groups equally. community. tigate both training methods, using a
Calm Down Staff support patients to draw Strengthen patient non-matched pre-post survey design;
Methods on their strengths and use/ confidence and skills to the surveys were online and sent out by
learn coping skills before the cope with distress. the local trainers. There was concern
use of PRN medication or that staff would not complete the sur-
containment. vey if they believed they could be iden-
tified, so the evaluators did not require
Discharge Before discharge, patients Strengthens patient
staff to identify themselves. There was
Messages leave messages of hope for community; generates
an optional question allowing staff to
other patients on a display in hope.
make a unique code using their date of
the unit.
birth and suburb, but most staff chose
not to enter this code. Therefore, we
were not able to match pre- and post-
centralized policy branch developed a staff attending the train-the-trainer training survey data. Health services
train-the-trainer package and deliv- workshops to plan and implement lo- self-selected into the training method
ered this to CNEs and other nurses in cal training, in agreement with man- that was most suitable to the organi-
local Safewards implementation roles. agers of the participating wards. The zation’s needs. Figure 1 presents the
In addition, training materials primary goal was to train all staff work- summary of the Safewards project com-
were provided for use in local health ing on the wards. Hence, two methods ponents and data collection timeline.
service training. To ensure fidelity to were used to train local staff in Safe- Ethics approval was obtained from the
Safewards, the training package used wards: health services provided a full University of Melbourne Human Eth-
the freely available UK Safewards day of training to ward staff (top-up in- ics Sub-Committee (ID 1443604), as
material (Safewards, 2015), which in- service sessions were offered by some well as Victorian Human Research Eth-
cludes an array of practical exercises services to ensure coverage of staff ics Multi-site (ID 15225L) approval for
and opportunities for brainstorming who could not attend a whole day ses- each of the involved health services.
implementation ideas. The materials sion), from this point forward referred
were organized by each of the 10 inter- to as “1-day plus in-service,” or they Data Sources and Collection
ventions to enable each health service provided multiple short in-service ses- Three sources of data were collected:
to tailor the process of training to best sions, referred to as “in-service.” Most pre-training surveys, post-training sur-
suit local circumstances. commonly, one in-service session cov- veys, and a fidelity checklist. The pre-
ered two Safewards interventions, and and post-training survey questions were
Local Health Service Training five sessions would be considered the developed by the research team, based
Training was provided within each optimal number to cover the 10 inter- on the components of Safewards and
health service to staff who would be ventions. Both methods of training of- previous research findings about impor-
implementing Safewards. The Victo- fered the same content multiple times, tant constructs related to training staff
rian Government did not provide a set to ensure coverage of all shift-working in health care.
process for the delivery of Safewards staff. Notwithstanding differences in Pre-Training Staff Survey. The pre-
training in the health services; rather, training delivery, staff in all services training survey included questions
this was left to the discretion of the received similar training content. about staff knowledge, confidence, and

34 Copyright © SLACK Incorporated


motivation to implement the Safewards
model and each of the 10 interven- TABLE 1 (CONTINUED)
tions (response scale: 1 = none, 2 = fair, SAFEWARDS INTERVENTIONSa
3 = good, 4 = very good, 5 = excellent), Intervention Description Purpose
and demographic and professional role
Soft Words Staff take great care with their Reduces a common
questions. This survey was completed
tone and use of collaborative flashpoint (i.e., a time or
prior to undertaking the training. This
language. Staff reduce the situation when something
survey was administered between De-
limits faced by patients, could go wrong, often seen
cember 2014 and February 2015.
create flexible options, and as a trigger that signifies
Post-Training Staff Survey. The post-
use respect if limit setting is potential conflict); builds
training survey included the questions
unavoidable. respect, choice, and dignity.
from the pre-training survey, as well as
additional quantitative questions relat- Talk Down De-escalation process focuses Increases respect,
ed to the frequency with which staff in- on clarifying issues and collaboration, and mutually
tended to implement each intervention finding solutions together. positive outcomes.
(1 = never, 2 = rarely, 3 = sometimes, Staff maintain self-control,
4 = usually, 5 = always), and staff sat- respect, and empathy.
isfaction with their training experience Positive Words Staff say something Increases positive
(e.g., interactive group activities, clar- positive in handover about appreciation and helpful
ity of trainers, group discussions, imple- each patient. Staff use information for colleagues
mentation planning activities, respons- psychological explanations to to work with patients.
es to questions) and materials (e.g., describe challenging actions.
written handouts, PowerPoint® slides, Bad News Staff understand, proactively Reduces impact of
UK videos) provided (1 = none, 2 = fair, Mitigation plan for, and mitigate the common flashpoints; offers
3 = good, 4 = very good, 5 = excellent). effects of bad news received extra support.
Staff also had the opportunity to pro- by patients.
vide written responses about their ex-
Reassurance Staff touch base with every Reduces a common
perience with each component of the
patient after every conflict flashpoint; increases
training. Staff completed the survey at
on the unit and debrief as patients’ sense of safety and
the completion of training. The survey
required. security.
was open until April 2015.
Fidelity Checklist. The fidelity check-
a
list is a 14-item, standardized audit tool Adapted from Victoria Department of Health and Human Services Safewards training
used by the UK Safewards Trial team resources (access https://www2.health.vic.gov.au/mental-health/practice-and-service-quality/
(Bowers et al., 2015). This observational safety/safewards/training-resources).
tool is completed during a walkthrough
of the ward, designed to assess if each of
the interventions is being implemented. cross checking their observations and re- 1 to 5, and together these 11 compo-
The fidelity checklist was slightly altered solved any differences in interpretation as nents result in a potential total score
for the Victorian trial. In the UK trial, an a team to enable consistency of ratings. of 55. Weighted averages were used to
additional end of shift measure, the Pa- The timing of data collection is depicted evaluate the impact of training on staff
tient Staff Conflict Checklist (PSCC), in Figure 1. knowledge, confidence, and motiva-
was gathered, and the records were used tion; and intention to implement for
to inform some components of the fidel- Data Analysis the entire sample. The same approach
ity checklist. However, in the Victorian The authors calculated composite was used to understand the impact of
trial it was not feasible to use the PSCC, scores, for each construct, of a total 55 the two methods of training on staff
so all reference to it was removed from points using data from the survey ques- knowledge, confidence, motivation,
the fidelity checklist. Independent ob- tions. There is one composite score and satisfaction; and intention to im-
servational walkthroughs of each ward for each of the following: knowledge, plement the Safewards model.
involved in the trial were conducted by confidence, motivation, and satisfac- The fidelity checklist scores each in-
members of the research team (J.F., B.H.) tion (post-training survey only); and tervention 1 to 10, with 100 being the
at four points during the project (T1, T3, intention to implement the Safewards highest possible total score. Data from
T4, T6), and local educators completed model and 10 interventions (post- the fidelity checklist were analyzed by
T2 and T5. The research team using the training survey only). The model and training method to assess if the method
fidelity checklist maintained a process of interventions each received a score of of training had an impact on the fidel-

JOURNAL OF PSYCHOSOCIAL NURSING • VOL. 58, NO. 12, 2020 35


Figure 1. Safewards project and data collection timeline.

to use the in-service method, and four


TABLE 2 services the 1-day plus in-service train-
ing approach. Thus, more wards were
NUMBER OF HEALTH SERVICES, WARDS, AND STAFF FOR EACH exposed to the 1-day plus in-service
TRAINING METHOD training method.
Staff Surveys (%) Demographic characteristics of
Training Method Health Services Wards Pre Post survey participants are presented in
Table 3. Most participants were female
In-service 3 7 101 (37) 82 (54)
and spoke English as their first language.
1-day plus in-service 4 11 174 (63) 71 (46) The type of professional role of staff was
similar among pre- and post-training
survey participants; however, the 1-day
ity of implementation. Fidelity data et al., 2017), these rely on knowledge plus in-service training group had no
were also analyzed to assess the transfer of which participants have provided peer or consumer workers and had a
of training into routine practice, with matched data and which have not, slightly higher rate of associate nurse
higher fidelity scores over time indicat- which was not the case in the current unit managers. A substantial number of
ing higher transfer of training. study. Although we are aware of the staff identified as other; these staff were
A range of univariate and multivari- potential problem with correlated error most frequently occupational therapists,
ate parametric statistical tests were per- in analyzing the data in this way, it was social workers, or psychologists.
formed. All assumptions (e.g., normal- believed that the reported approach Post-training surveys were received
ity and homogeneity of variance) were provided the best alternative for an in- from two health services that deliv-
adequately met for these analyses. An ferential analysis of the results. ered in-service training and four health
effect size measure and its associated services that delivered 1-day plus
95% confidence interval (CI) are re- RESULTS in-service training. Table 4 shows the
ported for each significance test. ␣ was The current study reports on 275 number of training events that post-
set at 0.01 for each analysis to compen- staff who attended local health service training survey participants reported
sate for multiple testing. training and completed the pre-training attending at their local health service.
Because of the partially overlapping survey, and 153 staff who completed Fifty-nine percent of staff working in
nature of the design—that is, some the post-training survey, reflecting a re- health services with the 1-day plus in-
participants provided data at pre- and sponse rate of approximately 60% and service training method reported at-
post-training, whereas others provided 34%, respectively, of those who poten- tending one training session. Similarly,
data at only one time point—and the tially received training across the seven 60% of staff who worked in health
inability to match results at the two health services. services delivering multiple in-service
time points, the data were analyzed us- Table 2 indicates the distribution training sessions attended between two
ing a between-subjects framework. Al- of health services, wards, and pre- and and five sessions. A further 20% attend-
though there are methods for analyzing post-training surveys to each of the two ed between six and 11 sessions; 19% of
partially overlapping designs (Derrick training methods. Three services chose the in-service training staff reported at-

36 Copyright © SLACK Incorporated


TABLE 3
PARTICIPANT CHARACTERISTICS
n (%)
In-Service 1-Day Plus In-Service
Pre-Training Post-Training Total Pre-Training Post-Training Total
Characteristic (n = 101) (n = 82) (n = 183) (n = 174) (n = 71) (n = 245)
Gender
Male 38 (37.6) 29 (35.4) 67 (36.6) 45 (25.9) 20 (28.2) 65 (26.5)
Female 63 (62.4) 48 (58.5) 111 (60.7) 129 (74.1) 50 (70.4) 179 (73.1)
Other 0 (0) 3 (3.7) 3 (1.6) 0 (0) 0 (0) 0 (0)
Missing 0 (0) 2 (2.4) 2 (1.1) 0 (0) 1 (1.4) 1 (0.4)
Language
English 82 (81.2) 66 (80.5) 148 (80.9) 156 (89.7) 64 (90.1) 220 (89.8)
Other 19 (18.8) 13 (15.9) 32 (17.5) 18 (10.3) 6 (8.5) 24 (9.8)
Missing 0 (0) 3 (3.7) 3 (1.6) 0 (0) 1 (1.4) 1 (0.4)
Professional role
Nurse unit 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)
manager
Associate nurse 9 (8.9) 6 (7.3) 15 (8.2) 17 (9.8) 11 (15.5) 28 (11.4)
unit manager
Clinical nurse 3 (3) 3 (3.7) 6 (3.3) 3 (1.7) 2 (2.8) 5 (2)
educator
Clinical nurse 0 (0) 2 (2.4) 2 (1.1) 1 (0.6) 0 (0) 1 (0.4)
specialist
RN 47 (46.5) 42 (51.2) 89 (48.6) 81 (46.6) 33 (46.5) 114 (46.5)
Enrolled nurse 12 (11.9) 12 (14.6) 24 (13.1) 31 (17.8) 8 (11.3) 39 (15.9)
Consumer 2 (2) 0 (0) 2 (1.1) 0 (0) 0 (0) 0 (0)
consultant
Peer worker 0 (0) 1 (1.2) 1 (0.5) 0 (0) 0 (0) 0 (0)
Other 25 (24.8) 13 (15.9) 38 (20.8) 37 (21.3) 13 (18.3) 50 (20.4)
Missing 3 (3) 3 (3.7) 6 (3.3) 4 (2.3) 4 (5.6) 8 (3.3)

tending only one session. This finding Safewards increased from very good be- Single-factor between-subjects
indicates that staff from the in-service fore training to excellent after training analysis of variance (ANOVA) con-
training group received substantially (Figure 2). ducted separately on the pre- and
less training than would be considered Figure 3 displays staff ratings of post-training scores for knowledge,
optimal for coverage of the Safewards knowledge, confidence, and motiva- confidence, and motivation found no
model and 10 interventions. tion before and after training, stratified significant differences across the seven
by training method. Prior to training, service sites. Site location was, there-
Impact of Training on Knowledge, both groups rated themselves as average fore, excluded from all subsequent
Confidence, and Motivation to good on measures of knowledge and models for either nesting or covariance
On average, staff reported that their confidence; this increased to very good purposes.
knowledge and confidence were good after training. Motivation increased Impact of training was analyzed us-
in the pre-training surveys and very from good to excellent after training. ing a 2 ⫻ 2 between-subjects multivari-
good in the post-training surveys. Mean This pattern of change was consistent ate ANOVA (MANOVA). The scores
self-reported motivation to implement across both groups. on knowledge, confidence, and motiva-

JOURNAL OF PSYCHOSOCIAL NURSING • VOL. 58, NO. 12, 2020 37


significant multivariate phase main
TABLE 4 effect (⌳ = 0.68, F[3, 571] = 88.08,
TRAINING RECEIVED BY POST-SURVEY RESPONDENTS p < 0.001, ␩p2 = 0.32, 95% CI [0.25,
n (%) 0.37]) with significant univariate re-
Number of Training
sults found for all three outcomes:
Events Attended In-Service 1-Day Plus In-Service knowledge (F[1, 573] = 263.46,
1 16 (19) 42 (59) p < 0.001, ␩2 = 0.32, 95% CI [0.26,
2 to 5 49 (60) 25 (35) 0.37]); confidence (F[1, 573] = 193.97,
6 to 11 17 (21) 4 (6) p < 0.001, ␩p2 = 0.25, 95% CI [0.20,
0.31]); and motivation (F[1, 573] =
Total 82 (100) 71 (100)
45.09, p < 0.001, ␩p2 = 0.07, 95% CI
[0.04, 0.12]). These results were all in
the direction of pre- to post-training
improvement, with the strongest effect
evident in knowledge improvement.
Using a Pearson’s correlation, we found
no significant relationship between
number of sessions attended and post-
intervention scores.

Satisfaction With Training


Both groups expressed satisfaction
(on average good to very good) with
most aspects of the training, particu-
larly the clarity of the trainers, the way
their questions were answered, and the
group discussions (Figure A, available
in the online version of this article).
With regard to training method,
Figure 2. Pre- and post-training mean knowledge, confidence, and motivation. participants who engaged in whole-
day training (mean = 3.41, SD = 0.87,
range = 1 to 5) were found to be sig-
nificantly more satisfied with the train-
ing videos than in-service participants
(mean = 2.91, SD = 1.24, range = 1 to 5)
(t[166] = 3.03, p = 0.003, d = 0.47,
95% CI [0.16, 0.77]). This was the only
significant difference related to train-
ing method. No significant correlations
were found between number of training
sessions and satisfaction.
Significant differences in two as-
pects of satisfaction were found across
the seven sites: written materials
(F[6, 161] = 2.91, p = 0.01, ␩p2 = 0.25,
95% CI [0.01, 0.16]) and training
videos (F[6, 161] = 3.94, p = 0.001,
␩p2 = 0.25, 95% CI [0.02, 0.20]). For
Figure 3. Mean knowledge, confidence, and motivation pre- and post- training, per written materials, Health Service
training model. 5 reported the highest level of sat-
isfaction (mean = 4.67, SD = 0.58,
tion formed the multivariate outcome. No significant multivariate or univari- range = 1 to 5), with Health Service 1
The two factors were phase (pre- ver- ate phase by training method interac- (mean = 3.40, SD = 0.76, range = 1 to 5)
sus post-intervention) and training tions or training method main effects and Health Service 3 (mean = 3.27,
mode (whole day versus in-service). were evident. There was, however, a SD = 0.59, range = 1 to 5) reporting

38 Copyright © SLACK Incorporated


the lowest (Figure B, available in the 95% CI [0.10, 0.32]). Despite this sig- confidence, and motivation, as well as
online version of this article). The pat- nificant difference, the actual range intention to implement and satisfac-
tern was different for training videos, between the highest and lowest fre- tion with training; and (2) implemen-
with Health Service 6 (mean = 4.06, quency items was not large, with tation of Safewards in practice. Staff
SD = 1.06, range = 1 to 5) and Health use of bad news mitigation record- who participated in local health service
Service 7 (mean = 4.00, SD = 0.00, ing the lowest mean (mean = 4.04, training reported significant pre- to
range = 1 to 5) reporting the high- SD = 0.88, range = 1 to 5) and use of post-training improvements in their
est level of satisfaction, and Health reassurance the highest (mean = 4.42, knowledge of the Safewards model
Service 2 the lowest (mean = 2.86, SD = 0.69, range = 1 to 5). Figure D and 10 interventions, as well as their
SD = 1.23, range = 1 to 5) (Figure (available in the online version of this confidence and motivation to imple-
C, available in the online version of article) illustrates the mean score for ment Safewards into their practice. We
this article). the extent to which participants in observed no difference between staff
Results of a within-subjects ANOVA each training group could envisage in- who received a 1-day plus in-service
showed an overall significant difference corporating each Safewards interven- training method compared with an in-
in mean satisfaction levels across the tion into their practice. service training method on any of these
eight items (⌳ = 0.55, F[7, 161] = 18.80, Similar to the satisfaction items, an outcomes. However, satisfaction with
p < 0.001, ␩p2 = 0.45, 95% CI [0.32, exploratory factor analysis found that the training videos was significantly higher
0.52]), with the highest overall level of 11 frequency items loaded onto a single for the 1-day plus in-service training
satisfaction being reported for clarity of factor accounting for 62% of observed group. Other observed differences in
the trainers (mean = 3.96, SD = 0.91, variance with all communalities >0.50. satisfaction were between health ser-
range = 1 to 5), and the lowest for the Results from the Fidelity Checklist at vices and not related to the training
training videos (mean = 3.15, SD = 1.10, the end of the 12-week trial period are method. An exploratory factor analy-
range = 1 to 5). presented by ward in Figure E (avail- sis indicated that all satisfaction items
Finally, an exploratory factor analy- able in the online version of this arti- were related to the one construct of
sis on the seven satisfaction items re- cle). The 1-day plus in-service training satisfaction.
vealed that all items were associated wards (n = 11) showed an average fidel- Staff believed they would usually in-
with a single factor that accounted for ity of 82 (SD = 18.54, range = 0 to 100) corporate Safewards interventions into
65% of variance with all communali- and implemented between four and their practice regardless of the method
ties >0.50, indicating that responses 10 interventions. In-service wards of training they received or the health
to the satisfaction items were different (n = 7) showed an average fidelity of service they worked in. Small but sig-
dimensions of a single underlying satis- 77 (SD = 15.49, range = 0 to 100) and nificant differences were found between
faction construct. implemented between five and 10 in- the degree to which each intervention
terventions. may be delivered, with staff reporting
Implementing Safewards into Routine No significant difference in fidel- bad news mitigation to be the least fre-
Practice ity was found between the two types of quently incorporated intervention and
Regardless of the training method, training methods or across the training reassurance being the most incorporated
staff on average reported that they could sites. In addition, there was no signifi- intervention. There was some vari-
envisage including Safewards into their cant association between the number of ability regarding the fidelity checklist.
practice usually (in-service: mean = 47, visits and either training method or site. Three wards from each group were im-
SD = 7.12, range = 10 to 50; 1-day A trend analysis was used to model plementing between four and six of the
plus in-service: mean = 46, SD = 6.86, the change in fidelity scores across the 10 Safewards interventions by the end
range = 10 to 50). There were no sig- six visits. Of the five trends tested, only of the 12-week trial. This low fidelity oc-
nificant differences between the two the first order linear trend was found curred despite some of these wards hav-
training methods on either the total to be significant (F[1, 84] = 29.18, ing access to the same training sessions
frequency score or on any of the 11 p < 0.001, ␩p2 = 0.25, 95% CI [0.11, as other wards in the same health ser-
individual frequency items. Similarly, 0.40]), indicating that fidelity increased vice that were implementing between
there were no significant differences in a consistent linear pattern with in- eight and 10 interventions. As such,
across the seven training sites on any creasing visits. This pattern is evident variability in fidelity was not shown to
frequency items, and no significant cor- in Figure F (available in the online be a function of the training method.
relations between the number of training version of this article). Rather, these findings suggest that other
sessions taken and any frequency item. (unmeasured) factors influenced transla-
A within-subjects ANOVA found DISCUSSION tion into practice. The degree to which
a significant difference across the 11 The current study evaluated Safewards was being implemented by
mean frequency scores (⌳ = 0.75, F[10, Safewards training in two parts: (1) the the end of the 12-month follow-up pe-
159] = 5.41, p < 0.001, ␩p2 = 0.25, impact of training on staff knowledge, riod showed a significant linear trend

JOURNAL OF PSYCHOSOCIAL NURSING • VOL. 58, NO. 12, 2020 39


overtime. There was no significant dif- 38% with associated reductions in con- trend was not evident in the Victorian
ference found between the methods of flict and containment events (Bowers trial. There are some possible explana-
training or health service and the degree et al., 2015). The UK trial relied on re- tions for the success of this trial. First,
to which the interventions were being search assistants conducting in-service the investment from the Victorian
implemented in the 9 months after the training and staff accessing materials Government sent a clear message to
implementation phase. that were freely available online (James services and staff that Safewards was
These findings are encouraging from et al., 2017). In contrast, the Victorian valued, and that they had “top down”
a health service perspective as the re- Government contracted trainers to support for implementation. Second,
sults reveal that, regardless of which collate all UK training materials, cre- gathering staff in groups for interac-
method of training staff received, it had ate learning activities, and deliver a tive training may create shared mo-
a positive impact on staff knowledge, train-the-trainer workshop to a small tivation and vision to make practice
confidence, and motivation. Notwith- group of nurse educators from each changes. In this case, group training
standing that a small number of wards health service participating in the trial. may have provided opportunities for
had low implementation fidelity at the Educators were provided with the ma- staff with reservations about the mod-
end of the implementation phase, at terials they would need to deliver train- el and interventions to raise questions
the end of the follow-up phase most ing to staff at their health service. This and problem solve how implementa-
wards had also successfully translated streamlined approach resulted in staff tion may work in their own workplace
their training into routine practice. at each ward receiving similar materi- (Hamilton & Manias, 2007). This no-
The findings of the current study are als and training presentations, regard- tion is consistent with findings from
consistent with previous studies that less of the method of training delivery. Grundy et al. (2017), who evaluated
have shown evidence of training in The flexibility in training delivery en- a staff training package and conclud-
psychosocial interventions leading to sured that training could be delivered ed that staff were grateful for having
practice change (Bradshaw et al., 2007; in a way that accommodated the needs team-oriented time enabling reflec-
Doyle et al., 2007; Kipping et al., 2018; and goals of each health service and tion of current practice. The format of
Maguire et al., 2018; Redhead et al., ward—a key factor when attempting to training that involved work teams in
2011). train all shift-working staff. It appears interactive exercises is “an acceptable
Implementation science explores that this flexibility in training method and engaging format for learning for
many factors beyond training that may and taking a team-based approach im- mental health professionals” (Grundy
have an impact on practice change, proved the uptake of this intervention. et al., 2017, p. 365). Kipping et al.’s
such as organizational buy-in at a senior Furthermore, the current study in- (2018) findings concur with this,
level, staff turnover, actual and per- volved a ward team being trained in where a cocreation approach to imple-
ceived time constraints, and other com- the implementation of Safewards. This mentation that involved substantial
peting change agendas (Damschroder team training approach is in contrast investment in staff training also re-
et al., 2009; Sadeghi-Bazargani et al., to the collaborative recovery model sulted in staff buy-in and high fidelity.
2014; Weatherson et al., 2017). Previ- reported by Uppal et al. (2010), in
ous evaluation reporting has identified which individual staff accessed train- STRENGTHS AND LIMITATIONS
a number of factors that may have im- ing and volunteered to be part of the A key methodological strength of
pacted implementation of Safewards in evaluation and subsequently, low rates this real-world study is the independent
Victoria, including lack of support from of translation of training into routine observation measuring transfer of train-
nurse unit managers to release staff practice were observed. Hence, the cur- ing into routine practice using a fidelity
for training, organizational barriers to rent study provides an indication that checklist. The current study also had
spending allocated funds to support transfer of training into practice may some limitations. First, we were unable
purchase of materials for implementa- be more successful when using a team- to match data across pre- and post-
tion, and staff ambivalence about some based approach. Further evidence from training surveys, thus precluding use
interventions (Hamilton et al., 2016). the current study highlights that “suc- of more powerful statistical methods.
Implementation of Safewards has prov- cessful” training (as indicated by im- Second, as the surveys were voluntary
en challenging in some settings, with provements in knowledge, confidence, there was varying representation across
prior studies noting that a larger invest- and motivation) is a vital foundation wards, which may have impacted find-
ment in training would likely have im- to the successful implementation of ings; for example, staff who participat-
proved staff understanding of the mod- Safewards and other psychosocial prac- ed may have been more or less positive
el and buy-in and thus implementation tice change initiatives. than those who did not. Given these
success (Higgins et al., 2018; Price et Practice change initiatives often methodological limitations, the degree
al., 2016). falter because of staff resistance to en- to which these findings can be general-
Fidelity to implementation in the gaging with new methods of practice ized to other mental health inpatient
UK Safewards Trial was reported to be (Corrigan et al., 2001); however, this settings is unclear.

40 Copyright © SLACK Incorporated


Dack, C., James, K., Jarrett, M., Jeffery, D.,
PRACTICE IMPLICATIONS settings, there is a noteworthy lesson Nijman, H., Owiti, J. A., Papadopoulos, C.,
The method of training adopted in from this study regarding the impact of Ross, J., Wright, S., & Stewart, D. (2014).
the Victorian Safewards Trial adds to substantial investment in training. Safewards: The empirical basis of the model
the body of knowledge about Safewards Future research in this area would be and a critical appraisal. Journal of Psychi-
atric and Mental Health Nursing, 21(4),
training and its subsequent positive strengthened by ensuring that all pre-
354–364. https://doi.org/10.1111/jpm.12085
impact on implementation. The cur- and post-training data are in matched PMID:24460906
rent study also adds to the mental pairs to allow for more detailed statis- Bowers, L., James, K., Quirk, A., Simpson, A.,
health training literature more broadly. tical analysis to be conducted into the Stewart, D., Hodsoll, J., & the SUGAR.
Training with structured content that impact of the dose of training and lo- (2015). Reducing conflict and contain-
ment rates on acute psychiatric wards: The
allowed for flexibility of delivery pro- cal cultural differences. In addition,
Safewards cluster randomised controlled
vided optimal conditions for reaching a future research should link training ef- trial. International Journal of Nursing Studies,
large number of inpatient shift-working fort with detailed implementation data, 52, 1412–1422. https://doi.org/10.1016/j.
staff. Importantly, the approach of including the service environment and ijnurstu.2015.05.001 PMID:26166187
training staff teams as opposed to in- key stakeholder involvement. Such a Bradshaw, T., Butterworth, A., & Mairs, H.
(2007). Does structured clinical supervision
dividuals appears to strengthen trans- link could inform a range of explana-
during psychosocial intervention education
fer of training into practice, and thus tions as to why the training effort in enhance outcome for mental health nurses
has a positive impact on the fidelity of the Victorian Safewards Trial has led and the service users they work with? Jour-
implementation to the Safewards mod- to strong and sustained implementa- nal of Psychiatric and Mental Health Nursing,
el and its 10 interventions. Safewards tion over the short- and long-term with 14(1), 4–12. https://doi.org/10.1111/j.1365-
2850.2007.01021.x PMID:17244000
training enables nurses to directly focus associated significant reductions in the
Brennan, G., & Gamble, C. (1997). Schizophre-
on prevention of conflict, a previously use of seclusion (Fletcher et al., 2017). nia family work and clinical practice. Mental
under-theorized but key area of rela- Health Nursing, 17(4), 12–15.
tional work in mental health nursing. REFERENCES Corrigan, P. W., Steiner, L., McCracken, S. G.,
Using fidelity measures to assess the Australian Human Rights Commission. (1993). Blaser, B., & Barr, M. (2001). Strategies for
Report of the national inquiry into the human disseminating evidence-based practices to
impact of training and the extent to
rights of people with mental illness. https:// staff who treat people with serious mental ill-
which the intervention has been ad- humanrights.gov.au/our-work/publications/ ness. Psychiatric Services (Washington, D.C.),
opted in practice can support staff to report-national-inquiry-human-rights- 52, 1598–1606. https://doi.org/10.1176/appi.
use reflective practices to monitor the people-mental-illness ps.52.12.1598 PMID:11726749
impact of Safewards and promote posi- Azeem, M. W., Aujla, A., Rammerth, M., Cowman, S., Björkdahl, A., Clarke, E., Gethin,
Binsfeld, G., & Jones, R. B. (2011). Effective- G., Maguire, J., & the European Violence
tive change.
ness of six core strategies based on trauma in Psychiatry Research Group. (2017). A
informed care in reducing seclusions and descriptive survey study of violence manage-
CONCLUSION restraints at a child and adolescent psychi- ment and priorities among psychiatric staff
Despite the current study’s limita- atric hospital. Journal of Child and Adolescent in mental health services, across seventeen
tions, it is clear that this training was Psychiatric Nursing, 24(1), 11–15. https://doi. European countries. BMC Health Services
org/10.1111/j.1744-6171.2010.00262.x Research, 17, 59. https://doi.org/10.1186/
well received, and that participants
Baker, J., Sanderson, A., Challen, K., & Price, s12913-017-1988-7 PMID:28103871
were ready and motivated to implement O. (2014). Acute inpatient care in the UK. Damschroder, L. J., Aron, D. C., Keith, R. E.,
Safewards, regardless of the method of Part 1: Recovery-oriented wards. Mental Kirsh, S. R., Alexander, J. A., & Lowery, J. C.
training. Substantial investment by the Health Practice, 17(10), 18–24. https://doi. (2009). Fostering implementation of health
Victorian Government to develop and org/10.7748/mhp.17.10.18.e883 services research findings into practice: A
Boumans, C. E., Walvoort, S. J. W., Egger, J. I. consolidated framework for advancing imple-
implement a train-the-trainer package
M., & Hutschemaekers, G. J. M. (2015). mentation science. Implementation Science, 4,
has provided a sound foundation from The methodical work approach and the re- 50. https://doi.org/10.1186/1748-5908-4-50
which staff across seven health services duction in the use of seclusion: How did it PMID:19664226
in metropolitan and regional Victoria work? The Psychiatric Quarterly, 86(1), 1–17. Derrick, B., Toher, D., & White, P. (2017). How
have been trained and subsequently https://doi.org/10.1007/s11126-014-9321-7 to compare the means of two samples that
PMID:25270895 include paired observations and independent
successfully implemented Safewards.
Bowers, L. (2009). Association between staff fac- observations: A companion to Derrick, Russ,
Interactive group training for Safewards tors and levels of conflict and containment Toher and White (2017). Tutorials in Quanti-
successfully improved the knowledge, on acute psychiatric wards in England. Psychi- tative Methods for Psychology, 13(2), 120–126.
confidence, and motivation of staff atric Services (Washington, D.C.), 60(2), 231– https://doi.org/10.20982/tqmp.13.2.p120
to implement Safewards into routine 239. https://doi.org/10.1176/ps.2009.60.2.231 Doyle, M., Kelly, D., Clarke, S., & Braynion, P.
PMID:19176418 (2007). Burnout: The impact of psychosocial
practice, resulting in the staff teams im-
Bowers, L. (2014). Safewards: A new model interventions training. Mental Health Prac-
plementing a high number of Safewards of conflict and containment on psychiat- tice, 10(7), 16–19. https://doi.org/10.7748/
interventions successfully and quickly. ric wards. Journal of Psychiatric and Mental mhp2007.04.10.7.16.c4298
Where services are embarking on im- Health Nursing, 21(6), 499–508. https://doi. Duxbury, J., Thomson, G., Scholes, A., Jones,
plementation of Safewards or other org/10.1111/jpm.12129 PMID:24548312 F., Baker, J., Downe, S., Greenwood, P.,
Bowers, L., Alexander, J., Bilgin, H., Botha, M., Price, O., Whittington, R., & McKeown, M.
psychosocial interventions in inpatient

JOURNAL OF PSYCHOSOCIAL NURSING • VOL. 58, NO. 12, 2020 41


(2019). Staff experiences and understandings LeBel, J. L., Duxbury, J. A., Putkonen, A., Uppal, S., Oades, L. G., Crowe, T. P., & Deane,
of the REsTRAIN Yourself initiative to mini- Sprague, T., Rae, C., & Sharpe, J. (2014). F. P. (2010). Barriers to transfer of collab-
mize the use of physical restraint on mental Multinational experiences in reducing and orative recovery training into Australian
health wards. International Journal of Mental preventing the use of restraint and seclu- mental health services: Implications for the
Health Nursing, 28(4), 845–856. https://doi. sion. Journal of Psychosocial Nursing and Men- development of evidence-based services.
org/10.1111/inm.12577 PMID:30887624 tal Health Services, 52(11), 22–29. https:// Journal of Evaluation in Clinical Practice,
Edmunds, J. M., Beidas, R. S., & Kendall, P. C. doi.org/10.3928/02793695-20140915-01 16(3), 451–455. https://doi.org/10.1111/
(2013). Dissemination and implementa- PMID:25310674 j.1365-2753.2009.01141.x PMID:20337837
tion of evidence-based practices: Training Maguire, T., Ryan, J., Fullam, R., & McKenna, Wale, J. B., Belkin, G. S., & Moon, R. (2011).
and consultation as implementation strate- B. (2018). Evaluating the introduction of Reducing the use of seclusion and restraint
gies. Clinical Psychology: Science and Practice, the Safewards model to a medium- to long- in psychiatric emergency and adult inpatient
20(2), 152–165. https://doi.org/10.1111/ term forensic mental health ward. Journal of services : Improving patient-centered care.
cpsp.12031 PMID:24072959 Forensic Nursing, 14(4), 214–222. https:// The Permanente Journal, 15(2), 57–62. https://
Fletcher, J., Spittal, M., Brophy, L., Tibble, H., doi.org/10.1097/JFN.0000000000000215 doi.org/10.7812/TPP/10-159 PMID:21841927
Kinner, S., Elsom, S., & Hamilton, B. (2017). PMID:30433910 Weatherson, K. A., Gainforth, H. L., & Jung, M. E.
Outcomes of the Victorian Safewards trial Meadows, G., Brophy, L., Shawyer, F., Enticott, (2017). A theoretical analysis of the barriers and
in 13 wards: Impact on seclusion rates and J. C., Fossey, E., Thornton, C. D., Weller, facilitators to the implementation of school-
fidelity measurement. International Journal of P. J., Wilson-Evered, E., Edan, V., & Slade, based physical activity policies in Canada: A
Mental Health Nursing, 26, 461–471. https:// M. (2019). REFOCUS-PULSAR recovery- mixed methods scoping review. Implementa-
doi.org/10.1111/inm.12380 PMID:28960739 oriented practice training in specialist mental tion Science, 12(1), 41. https://doi.org/10.1186/
Grundy, A. C., Walker, L., Meade, O., Fraser, C., health care: A stepped-wedge cluster ran- s13012-017-0570-3 PMID:28347322
Cree, L., Bee, P., Lovell, K., & Callaghan, P. domised controlled trial. The Lancet. Psy-
(2017). Evaluation of a co-delivered training chiatry, 6, 103–114. https://doi.org/10.1016/ Ms. Fletcher is Research Fellow, Dr. Brophy is
package for community mental health profes- S2215-0366(18)30429-2 PMID:30635177 Honorary Principal Research Fellow, Centre for
sionals on service user- and carer-involved Oster, C., Gerace, A., Thomson, D., & Muir- Mental Health, Dr. Kinner is Professor, Melbourne
care planning. Journal of Psychiatric and Men- Cochrane, E. (2016). Seclusion and restraint School of Population and Global Health, and Dr.
tal Health Nursing, 24(6), 358–366. https:// use in adult inpatient mental health care: An Hamilton is Associate Professor, Centre for Psy-
doi.org/10.1111/jpm.12378 PMID:28218977 Australian perspective. Collegian (Royal Col- chiatric Nursing, School of Health Sciences, The
Hamilton, B., Fletcher, J., Sands, N., Roper, lege of Nursing, Australia), 23(2), 183–190. University of Melbourne, Parkville, Victoria; and
C., & Elsom, S. (2016). Safewards Vic- https://doi.org/10.1016/j.colegn.2015.03.006 Dr. Reece is Professor of Psychological Science,
torian trial final evaluation report. https:// Pearce, J., Mann, M. K., Jones, C., van Buschbach, Australian College of Applied Psychology, Mel-
healthsciences.unimelb.edu.au/__data/assets/ S., Olff, M., & Bisson, J. I. (2012). The most bourne, Victoria. Dr. Kinner is also Professor and
pdf_file/0004/2472718/Safewards-Victoria- effective way of delivering a train-the-trainers Head, Justice Health Group, Centre for Adoles-
Evaluation-Final-Report-July-2016-3.pdf program: A systematic review. The Journal of cent Health, Murdoch Children’s Research Insti-
Hamilton, B., & Manias, E. (2007). Rethink- Continuing Education in the Health Professions, tute, Parkville, Victoria, Adjunct Professor, Griffith
ing nurses’ observations: Psychiatric nurs- 32(3), 215–226. https://doi.org/10.1002/ Criminology Institute, Griffith University, Honor-
ing skills and invisibility in an acute in- chp.21148 PMID:23173243 ary Professor, Mater Research Institute-UQ, Uni-
patient setting. Social Science & Medicine, Peters, D. H., Adam, T., Alonge, O., Agyepong, versity of Queensland, Queensland, and Associate
65(2), 331–343. https://doi.org/10.1016/j. I. A., & Tran, N. (2013). Implementation Professor, School of Public Health and Preventive
socscimed.2007.03.025 PMID:17459545 research: What it is and how to do it. BMJ Medicine, Monash University, Melbourne, Victoria;
Happell, B., & Harrow, A. (2010). Nurses’ at- (Clinical Research Ed.), 347, f6753. https:// and Dr. Brophy is also Professor, School of Allied
titudes to the use of seclusion: A review of doi.org/10.1136/bmj.f6753 PMID:24259324 Health, Human Services and Sport, La Trobe Uni-
the literature. International Journal of Men- Price, O., Burbery, P., Leonard, S.-J., & Doyle, M. versity, Melbourne, Victoria.
tal Health Nursing, 19(3), 162–168. https:// (2016). Evaluation of Safewards in forensic The authors have disclosed no potential
doi.org/10.1111/j.1447-0349.2010.00669.x mental health. Mental Health Practice, 19(8), conflicts of interest, financial or otherwise. This
PMID:20550639 14–21. https://doi.org/10.7748/mhp.19.8.14.s17 independent evaluation was financially supported
Higgins, N., Meehan, T., Dart, N., Kilshaw, M., Ramluggun, P., Chalmers, C., & Anjoyeb, M. by the Office of the Chief Mental Health Nurse, in
& Fawcett, L. (2018). Implementation of the (2018). The practice of seclusion: A review the Department of Health and Human Services,
Safewards model in public mental health fa- of the discourse on its use. Mental Health Prac- Government of Victoria. This article forms part
cilities: A qualitative evaluation of staff per- tice, 21(7), 17–23. https://doi.org/10.7748/ of the work toward a PhD, which is supported
ceptions. International Journal of Nursing Stud- mhp.2018.e1258 through an Australian Government Research
ies, 88, 114–120. https://doi.org/10.1016/j. Redhead, K., Bradshaw, T., Braynion, P., & Training Program Scholarship. Ms. Fletcher is
ijnurstu.2018.08.008 PMID:30236863 Doyle, M. (2011). An evaluation of the out- supported by NHMRC PhD Research Scholarship
James, K., Quirk, A., Patterson, S., Brennan, comes of psychosocial intervention training 1133627. Dr. Kinner is supported by NHMRC
G., & Stewart, D. (2017). Quality of inter- for qualified and unqualified nursing staff Research Fellowship APP1078168.
vention delivery in a cluster randomised working in a low-secure mental health unit. The researchers are indebted to staff and
controlled trial: A qualitative observational Journal of Psychiatric and Mental Health Nurs- consumers at the trial sites, for cooperation with
study with lessons for fidelity. Trials, 18, 548. ing, 18(1), 59–66. https://doi.org/10.1111/ fidelity measurement.
https://doi.org/10.1186/s13063-017-2189-8 j.1365-2850.2010.01629.x PMID:21214685 Address correspondence to Justine Fletcher,
PMID:29149915 Sadeghi-Bazargani, H., Tabrizi, J. S., & Azami- MPsych(Clin), BPsych(Hons), Research Fellow,
Kipping, S. M., De Souza, J. L., & Marshall, L. Aghdash, S. (2014). Barriers to evidence-based Centre for Mental Health, Level 4, 207 Bouverie
A. (2018). Co-creation of the Safewards medicine: A systematic review. Journal of Evalu- Street, Carlton, Victoria, 3053, Australia; email:
model in a forensic mental health care facil- ation in Clinical Practice, 20(6), 793–802. https:// justine.fletcher@unimelb.edu.au.
ity. Issues in Mental Health Nursing, 40, 2–7. doi.org/10.1111/jep.12222 PMID:25130323 Received: March 24, 2020
https://doi.org/10.1080/01612840.2018.1481 Safewards. (2015). Interventions. http://www. Accepted: June 29, 2020
472 PMID:30067417 safewards.net/table/english/interventions doi:10.3928/02793695-20201013-08

42 Copyright © SLACK Incorporated


3.90
Questions answered to your satisfaction
3.72

3.27
Adequate time allowed for questions
3.32

3.65
Group discussions
3.68

4.01
Clarity of the trainers
3.79

3.54
Power point slides
3.51

3.56
Interactive group activities
3.61

2.83
Videos
3.34

3.70
Written materials
3.54

0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
Participants rating on 5-point likert scale

Inservice (n=82) One day (n=71)

Figure A. Mean satisfaction of training components, per training model.


5.00

4.50

4.00

3.50
Mean satisfaction rating

3.00

2.50

2.00

1.50

1.00

0.50

0.00
1 2 3 4 5 6 7
Ward

Figure B. Mean satisfaction with written materials by ward.


5.00

4.50

4.00

3.50
Mean satisfaction rating

3.00

2.50

2.00

1.50

1.00

0.50

0.00
1 2 3 4 5 6 7
Ward

Figure C. Mean satisfaction with videos by ward.


Safewards Model 4.16
4.17

Reassurance 4.45
4.39

Soft Words 4.40


4.20

Discharge Messages 4.23


4.01

Calm Down Methods 4.45


4.25

Mutual Help Meeting 4.26


3.97

Know Each Other 4.18


4.24

Bad News Mitigation 4.00


4.03

Positive Words 4.38


4.30

Talk Down 4.29


4.18

Clear Mutual Expectations 4.26


4.04

1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00


Participants rating on 5-point likert scale

Inservice (n=82) One day (n=71)

Figure D. Frequency with which staff could envisage implementing Safewards into their practice
after participating in training, by training type.
100 97 100
95 95
90 90 89 89 90 90
90 86 86.5

80

70 65 65
Total fidelity scores

60
60 57.5
55

50 45

40

30

20

10

0
one day plus in-service (n = 11 wards) in-service (n = 7 wards)
Individual wards

Figure E. Fidelity score of each ward at the end of the 12-week trial, per training model.
100

90

80

70
Mean Fidelity

60

50

40

30

20

10

0
Visit 1 Visit 2 Visit 3 Visit 4 Visit 5 Visit 6
Visit Number
Figure F. Mean fidelity across visits.
Reproduced with permission of copyright owner. Further reproduction
prohibited without permission.

You might also like