ORIGINAL ARTICLE

Care zoning in a psychiatric intensive care unit: piloting a model of

care in clinical risk assessment

Antony Mullen, RN, BN(Syd), MN(UTS), FACMHN,

Clinical Nurse Consultant,
Lake Macquarie Mental Health Service,
Hunter New England Local Health District.
Conjoint Lecturer, School of Nursing & Midwifery,
University of Newcastle.
PO Box 833
Newcastle NSW 2300
Antony.Mullen@hnehealth.nsw.gov.au
Phone: 02 4033 5031
Fax: 02 4033 5341

Vincent Drinkwater, RN, BHA(UNSW), GC Aged Care(UoN), MACMHN,

Nurse Unit Manager,
Psychiatric Emergencies Services,
Hunter New England Local Health District.

Terry J. Lewin, BCom(Psych)Hons.,

Research Manager, Mental Health Service,
Hunter New England Local Health District.
Conjoint Associate Professor, School of Medicine and Public Health,
University of Newcastle.

ABSTRACT:
Care Zoning is a model of care that guides nurses in assessing clinical risk and planning
care. Concerns about the lack of routine clinical assessment, and the varying quality of nursing
documentation, prompted a pilot of the care zoning model in a Psychiatric Intensive Care Unit
(PICU) within a regional mental health facility. The care zoning model assigns patients to a Zone
according to their clinical risk, encouraging nurses to document and implement targeted
interventions required to manage those risks.
The model was piloted for three months utilising the evaluation frameworks of previous
authors. This included a pre- and post-implementation questionnaire, a pre-pilot file audit, and a
weekly file audit during the pilot. Informal staff feedback was also sort via the surveys and regular
staff meetings.
Results of this pilot demonstrated improvements in the quality of mental state
documentation, and clinical risk information was identified more accurately. There was limited
improvement in the quality of care planning and the documentation of clinical interventions. Initial
concerns from staff over the introduction of the model shifted into overall acceptance and
recognition of the models benefits.

Key words: risk assessment; mental state examinations; psychiatric intensive care

INTRODUCTION
Psychiatric Intensive Care Units (PICU) are highly specialised units catering for acutely
disturbed patients, associated risk behaviours and vulnerabilities (Bowers et al., 2008) Therefore,
risk assessment is a fundamental part of nursing practice, and there needs to be a model of care that
allows for a systematic and continuous process to identify risks and implement targeted strategies.
Ongoing risk assessment, with monitoring and observation of risk, are seen as the main strategies to
manage risk behaviours within inpatient units (Bowles, 2000). However, clearly articulating
strategies to manage identified risks does not necessarily occur (Mullen, 2009).
Hospitalisation, is often seen as the intervention in itself, although incidents associated with
risk factors such as self harm or aggression still occur regularly within these facilities (Duxbury et
al., 2008). Admissions to a PICU facility can be seen as the solution in itself for managing such risk
factors (Bowers et. al, 2008). This is despite the questioning of the evidence to support PICU as an
intervention in itself (Crowhurst & Bowers, 2002). Recovery based psychosocial interventions that
attempt to facilitate greater awareness and understanding of self help strategies are not always used
(Mullen, 2009). Crowhurst and Bowers (2002) do describe psychosocial methods within PICU
facilities, but these seem to be generalised to cover any non-pharmacological approaches or those
processes which minimise patients restrictions. Taylor et al. (2011) blame the dominance of a crisis
approach within acute inpatient mental health facilities for a lack of knowledge and skill in
providing proactive and targeted strategies including psychosocial strategies.

Current risk assessment protocols follow the NSW Mental Health Outcome and Assessment
Tools (MHOAT). These documents do provide a framework for assessing risk, but the applicability
of these tools in certain situations is not always clear. Added to this, the information within these
tools is often poorly integrated within clinical care. Mental health nursing practice requires a model
that can guide all facets of practice from assessment and planning of care, through to the
implementation and evaluation of strategies.

Within the PICU investigated in the current study, risk assessment at the time of admission
was clearly identifiable. However, there was a lack of evidence of ongoing clinical risk assessment,
due, in part, to a lack of suitable standardised tools to guide and support clinicians in this process.
As a result, there was considerable variation in the amount of detail recorded in the nursing progress
notes, and, therefore, varying quality of information about risk assessment.
The authors had become aware of a model being used in similar PICU facilities, called Care
Zoning. This was originally developed by Ryrie et al. (1997) in a community setting, with the aim
of guiding nurses in assessing clinical risk and planning care (Ryrie et al., 1997). Since this original
work, others have applied this to various acute mental health facilities in Australia and the UK (Guy
& Henderson, 2004; Gamble, 2006; Gamble et al., 2010; Taylor et al., 2011).
The Care Zoning model utilises a traffic light system which identifies clinical risk and levels
of care required, according to three zones: the red zone identifies a high level of risk and high levels
of care; amber is for medium levels of risk and care; and green is for low levels of risk and care
(Ryrie et al., 1997). This model also emphasises the time nurses spend with patients, referred to as
protected time. It makes sense that the nurse needs to spend individual time with all patients, for
to be provided satisfactory care to be provided. Nurse patient interaction is intrinsically linked with
positive treatment outcomes (Rydon, 2005). Despite this, the literature reveals a decline in the
amount of time nurses spend with patients (Bowers et al., 2005; Bowles, 2000; Higgins et al., 1999;
Hurst et al., 1998; Mullen, 2009). Therefore the Care Zoning Model would have legitimate
applicability in this setting, as it proposes that nurses articulate specific strategies and interventions
in response to identified risk factors and behaviours.

Aims
The aims of this project were three-fold: to pilot the Care Zoning model of care as a
standardised tool within an 8 bed PICU; to evaluate the models ability to facilitate ongoing clinical
risk assessment and management, through improvements to the quality of nursing documentation

and the handover of clinical risk information; and to assess how Care Zoning can facilitate a greater
emphasis on nurse patient engagement.

METHODS
Study design
This study is essentially a quality improvement project, which seeks to develop nursing
practice in risk assessment and care planning.

Setting
This study took place in an 8 bed Psychiatric Intensive Care Unit (PICU) within a specialist
100 bed Mental Health Facility in a regional centre of New South Wales, Australia. This PICU
services the entire Hunter New England Local Health District that has a population of
approximately 840,000.

Procedure
As previously mentioned, the authors became aware of Care Zoning through the work of
colleagues in other Local Health Districts, from whom assistance was sought to implement the
model locally. Site visits and meetings with staff at the units where Care Zoning was the model of
care were carried out. This enabled the authors to experience and observe how Care Zoning
operated. Benefits as well as potential problems were discussed and worked through. A working
party to guide the project was formed from staff of the PICU, in addition to the authors, including a
Clinical Nurse Specialist, Registered Nurse, Occupational Therapist and Seclusion and Restraint
Project Officer. Major tasks for the working party were to finalise the tools, procedures and
guidelines for the models implementation, as well as establishing the risk items and planning for the
pilot roll-out.
The Care Zoning documents and tools were modified to suit the PICU. One
recommendation from our colleagues with experience in Care Zoning was to identify the relevant
risk items for ourselves, rather than adopting their existing items. This provided a critical
opportunity to engage staff in the project and ensure ownership. The working party decided upon
five risk items, considerably less than the nine items our colleagues used. Several items considered
to be similar were grouped together, as follows: mental emotional state;
violence/aggression/arousal; suicidality/self harm; impulse control/disinhibition; and treatment noncompliance.
The central document within the Care Zoning model was the Care Zoning Shift Review
Form, a one page document that incorporated the traffic light colours representing the clinical risk
across the three agreed zones: red (high risk); amber (medium risk); and green (low risk). In
addition, Mental State Examination headings were listed with space to document information, along
with proposed interventions and space to evaluate the outcome of these interventions. Information
sheets, including a step-by-step clinicians guide with examples, and a Mental State Examination
guide were also produced. A meeting with staff from the Clinical Information unit was essential to

ensure that the document design complied with local protocols and form design principles. General
discussions also took place within the working party, and with other staff, to identify ways to
support the protected time component of care zoning. This was a challenging exercise as it
represented a departure from the existing approach which prioritised managing the staff office and
being available for any unforseen circumstance. As a result of this, brief nurse: patient encounters
were encouraged in order for staff to ; respond to phone calls, provide information to medical staff
or relatives, and to support other colleagues.
The evaluation plan for this project essentially sought to replicate previous studies (e.g.,
Ryrie et al., 1997; Taylor et al., 2011). A 3 month pilot was implemented, between June and August,
2010. The similar staff survey tool to Ryrie et al. (1997) was administered both pre- and post-pilot,
primarily to compare care zoning with practices prior to its implementation. The 23 items asked
respondents to nominate whether current risk management strategies made no difference, some
difference, or a noticeable difference to their practice. A pre-pilot file audit was also carried out
on 7 patients, by auditing documentation from the morning and afternoon shifts over 7 days,
utilising the same audit tool as used by Taylor et al. (2011). Weekly file audits were conducted
during pilot period using the same audit tool. Regular meetings with staff took place throughout the
pilot to gauge progress and gather feedback about implementation.

Ethical issues
The Hunter New England Research Ethics Committee was consulted and concluded that the
proposed study represented an audit/quality improvement exercise and, therefore, did not require
formal approval.

Data coding and analysis

Data aggregation and analysis was conducted using the Statistical Package for Social
Sciences (SPSS version 17.0; Chicago, IL, USA). The pre-pilot file audit included 48 records from
day shifts and the corresponding records from afternoon shifts (and covered a one-week period),
while the audits conducted during the actual pilot period (i.e., post-implementation) included 283
records from day shifts and 289 from afternoon shifts (and covered a 15 week period). Following
Taylor et al. (2011), five of the eight audit indices simply noted the presence or absence of the
desired characteristic within the record for that shift (i.e., management measures, patient
observation, risk assessment, evidence of 1:1 time with the patient, and a relevant care plan), while
the remaining three audit indices (i.e., mental state, interventions, and evaluation outcomes)
characterised the quality of the information provided (not present, minimal, satisfactory or good).
Overall chi-square tests were used for each of these indices to compare the percentage of audited
records with the desired characteristic pre- versus post-implementation. An aggregate quality score
(potentially ranging from 0 to 14) was also calculated for each record using the weights listed on the
left-hand side of Table 1. This score was used to examine shift differences, overall improvement
from pre- to post-implementation, and changes across the post-implementation period; for these
analyses, Analysis of Variance (ANOVA) based techniques were used. The aggregate quality index
displayed good internal consistency, with a Chronbachs alpha of 0.86.
For convenience, the some difference and noticeable difference responses to the
individual items in the PICU staff surveys were collapsed together in calculating the percentages
reporting a difference to their practice; comparisons between these percentages pre- versus post-

implementation were undertaken using chi-square tests. These response profiles for the 23 survey
items were also compared with findings from the Taylor et al. (2011) study using correlational
analyses.

RESULTS
File audit
Table 1 summarises the findings from the pre- and post-implementation audit. As displayed
in the right-hand column, there were statistically significant improvements across all eight audit
indices. Some of these improvements simply reflected the fact that particular characteristics were
largely absent from the pre-pilot audit (e.g., information about interventions, evaluations of
outcomes, and shift level care plans). The most notable improvements related to risk assessments,
which rose from 13.5% to 81.5% of audited records, and the quality of mental state examinations,
for which initially two-thirds (65.6%) were coded as minimal quality, compared with two-thirds
(64.4%) displaying satisfactory or good quality during the post-implementation audit. Thus, the
introduction of Care Zoning improved the assessment of risk against standardised criteria, but also
led to improved accuracy of information regarding mental state. However, despite being an integral
part of the Care Zoning model, the quality of information recorded about interventions provided and
associated outcomes was still less than optimal, with only 31.5% and 42.5% respectively displaying
satisfactory or good quality post-implementation.

Insert Table 1 near here

Aggregate quality scores were analysed using a two-way ANOVA shift (day vs. afternoon)
by phase (pre- vs. post-implementation), which revealed a significant main effect for phase (F(1, 664) =
218.51, p < 0.001), but no shift or interaction effects. The mean quality score rose from 2.39 preimplementation (SD = 1.00, median =2.00) to 8.31 post-implementation (SD = 3.90, median =

9.00). The timing of post-implementation improvements was subsequently examined using a twoway ANOVA shift (day vs. afternoon) by audit period block (divided into three 5-week blocks),
which revealed a significant audit period effect (F(2, 566) = 6.21, p < 0.01), but no shift or interaction
effects. The post-implementation mean aggregate quality score was higher in the third 5-week block
(9.31, SD = 4.08), compared with the first block (7.87, SD = 4.00) or second block (8.07, SD =
3.62), suggesting that improvements in documentation took some time to be fully implemented; to
contextualise these post-implementation profiles, it should be noted that a score of 8 on this
aggregate audit index is the equivalent of all elements being at least minimally present.

Survey of PICU staff

The PICU staff surveys were completed by 15 staff pre-implementation and 13 staff postimplementation, representing response rates of approximately 74% and 72% respectively; several
staff changes occurred between these surveys, so there is only likely to be a modest overlap in
responders. For each of the 23 items surveyed, there was no statistically significant difference
(based on chi-square analyses) between pre- and post-implementation ratings of the extent to which
current risk management strategies made a difference (defined as some or a noticeable
difference). Consequently, to aid presentation and discussion, responses to the PICU staff surveys
have been grouped on the basis of postimplementation endorsement rates Table 2A: eight highly
endorsed questions (> 80% of respondents); Table 2B: nine moderately endorsed questions (60% to
80% of respondents); and Table 2C: six questions with relatively low endorsement (< 60% of
respondents). Applying similar criteria to the pre-implementation survey would have seen nine
highly endorsed, 13 moderately endorsed, and only one question (Q13 meetings commitment)
with a low endorsement rate. Thus, for many questions, there was little scope to demonstrate postimplementation improvement, particularly given the small sample size. On the other hand,
responses to four questions may raise some concerns, in that they displayed a tendency for lower

10

endorsement rates post-implementation (Q18 awareness of other patients needs, Q12 promoting
liaison, Q1 policy adherence, and Q23 resource reallocation).

Insert Tables 2A to 2C near here

As shown in Table 2A, the major perceived benefits of current risk management strategies
included: raising intuitive concerns (Q7); systematic, targeted and prioritised crisis interventions
(Q8, Q5, Q10); identification and sharing of clinical and risk information (Q3, Q6, Q4); and relapse
prevention (Q9). The questions with moderate endorsement rates (see Table 2B) focused largely on
relationships with (Q16, Q14) and communication (Q20, Q21, Q12) and support from colleagues
(Q2, Q19, Q22, Q18). As shown in Table 2C, the questions receiving the lowest postimplementation endorsement were less likely to be about direct patient care or immediate
relationships with colleagues, and included: staff absences (Q11); team structure (Q15); patient
loads (Q17); policy adherence (Q1); meetings commitment (Q13); and resource reallocation (Q23).
The right-hand columns of Tables 2A to 2C display the corresponding endorsement rates
from the study by Taylor et al. (2011). The overall profile of endorsement rates from the two studies
(across the 23 questions) was moderately correlated, r (21) = 0.70 (p < 0.001), suggesting that there
is some level of consensus between staff in PICU and those in other New South Wales mental
health units about the likely benefits of workable, high quality, risk management strategies.

Other observations
There was initial concern from staff about implementing a new form and the
perceived resulting increase in paperwork. This issue was worked through and the fact that this
form replaced current documentation allayed some of this concern. There was a perception among
some staff that the information within the Care Zoning model was already being documented,
despite the pre-pilot file audits revealing otherwise. Staff were also more prepared to commit to a

11

time limited 3-month pilot, with subsequent review and evaluation. Once implemented, it was
observed that staff were taking much more care in documenting information, particularly the mental
state examination. This process generated a lot of discussion and debate about psychiatric
terminology and the use of jargon and clinical descriptors. As the pilot progressed, staff verbalised
their appreciation of the structure and direction within the risk items and mental state sections of the
Care Zoning form provided. There was an obvious improvement in knowledge and understanding
of the mental state examination and presenting symptoms. However, the Care Zoning form was not
really utilised as a handover tool, with previous approaches to handover being preferred. Protected
time, as previously mentioned, was a challenging concept to implement. Attempts were made to
roster nurses to timeslots within their shift but this was not successful. However, the results did
demonstrate an improvement in 1:1 time spent with patients (see Table 1). This was linked to the
successful increase in the completion of Mental State Exams on the Care Zoning form. Nurses
spent more time with patients in order to more effectively assess and complete their mental state
examination.

DISCUSSION
This pilot project sought to implement a Care Zoning model within a PICU and returned
varying results. Overall, this pilot demonstrated a successful implementation, with high levels of
completion of the Care Zoning forms throughout the pilot. This is a comparable result to the Taylor
et al. (2011) study. Likewise, the audit and survey evaluation measures were similar to earlier
studies (e.g., Ryrie et al., 1997; Taylor et al., 2011) and performed well. The inclusion of an
aggregate quality score also facilitated an examination of other influences, such as potential shift
differences and the timing of changes during the post-implementation period. The consistency of
the audit findings across day and afternoon shifts probably reflects the similarity in nursing
characteristics and staffing demands across these shifts within the PICU, which may not be the case
in other acute inpatient units. Improvements in the overall quality of nursing documentation were

12

also more marked after the first ten weeks, suggesting that change takes time, and that longer-term
monitoring and evaluation is probably warranted.
In terms of specific sections within the Care Zoning form, mental state examination and risk
assessment information and documentation improved dramatically. Staff were able to embrace this
part of Care Zoning. Articulating targeted strategies and evaluating these strategies showed less
improvement. This is consistent with the literature that expresses concern over the lack of targeted
strategies to manage risk (Bowles, 2000; Mullen, 2009; Taylor et al., 2011).
The use of the Care Zoning form as a clinical handover tool did not get established. This is
at odds with results found in other studies within inpatient settings (Gamble, 2006; Taylor et al.,
2011). In some ways this is a puzzling finding, as the Care Zoning form provides an ideal
framework with which to deliver a verbal clinical handover. It summarises risk information, the
patients mental state, and subsequent management strategies, as well as an evaluation of outcomes.
Despite this, nurses did not opt to use the Care Zoning form in this way.
Protected time that quarantines 1:1 nurse patient interaction did not get formally introduced.
This is despite evidence from the audits that 1:1 time had improved as a result of the pilot. Existing
processes and conflicting responsibilities, where staff were torn between patient interaction and
administrative roles, was one of the major obstacles for this. This requires further examination, to
try to ascertain how protected time can be more formally established for the benefit of staff, patient
engagement and care provision. An interesting point to note on this the view among many nurses
that patients within a PICU environment, particularly psychotic or acutely disturbed patients, only
tolerate brief frequent interactions, or even no interactions at all. This is a valid point and is based
on clear assessment careful clinical decision making . The concept of protected time does not
prescribe intimate and continual face to face contact, but merely a quarantined time where the nurse
is solely available for that one patient (Taylor et al, 2010).
In addition, the staff survey did not show that Care Zoning made a difference to the clinical
practice of nurses in the PICU, perhaps because they were already reasonably positively disposed to

13

the existing strategies for managing risk and, therefore, it was difficult to demonstrate
improvement, especially with a small number of survey respondents. Again, this is a finding at odds
with previous studies (Gamble et al., 2009; Gamble, 2006; Taylor et al., 2011). This result may also
be explained in part because Care Zoning was not part of the clinical handover. Therefore, the
handover meeting was not seen as necessarily a place to gather support and communicate clinical
needs. If in fact handover meetings are seen in this way, it was not attributed to Care Zoning.
Despite these somewhat mixed results, there is anecdotal evidence that staff welcomed the
structure that Care Zoning provided and this also served as an educational tool as well as a clinical
assessment and documentation tool. The standard of mental state examination documentation
improved, as did the consistency of risk assessment information.

LIMITATIONS
Despite the encouraging results there are some limitations that need to be highlighted. This pilot did
not measure the impact of Care Zoning in terms of patient satisfaction or clinical outcomes. Other
variables that would add to an evaluation of the Care Zoning model, would be to compare patient
length of stays, seclusion rates, critical incidents rates, or to measure ward atmosphere. Furthermore
this pilot only involved one unit with a small sample of data. This needs to be considered, despite
the efforts to provide comparisons with the Taylor (et al. 2010) study.
CONCLUSIONS AND WAY FORWARD
This pilot has shown that formalising the mental state examination brings improvements in the
consistency and quality of assessment documentation. It is also a demonstration in the process of
implementing a change in clinical practice. In this pilot there was initial staff reluctance that
translated to overall acceptance, once the benefits to their practice were realised. Despite this,
establishing the concept of Protected Time was a more challenging exercise.
Since the pilot completion, discussions have led to the development of a Care Zoning sticker as
demonstrated by Taylor and colleagues (2010). This is largely to address the real problem of poor
file chronology due to the insertion of the Care Zoning form into the medical record. The
14

implementation of the sticker has a similar evaluation plan to that of this pilot discussed. It are also
discussion about rolling this sticker out more widely across other acute inpatient units.
One of the next challenges to be addressed by the authors is to improve the quality and consistency
of care planning and their evaluations. Another ongoing process, is continuing to facilitate a greater
emphasis on nurse patient engagement.

15

ACKNOWLEDGEMENTS
The authors would like to thank all the staff of the PICU where this pilot took place, Kelly
Hobden for entering the data, Rose Leahey for assisting with file audits and Jon Chesterson for
supporting the implementation of this pilot. Also, the authors express much gratitude to Justin Steel
and Belinda Weston for valuable contributions to the working party and the pilots implementation,
and to Stuart Guy for his generous provision of time and resources.

16

REFERENCES:

Bowles, A. (2000) Therapeutic nursing care in acute psychiatric wards: engagement over control.
Journal of Psychiatric and Mental Health Nursing (Commentary) 7, 179-184.

Bowers, L. (2005) Reason for admission and their implications for the nature of acute inpatient
psychiatric nursing. Journal of Psychiatric and Mental Health Nursing. 12, 231-236.

Bowers, L., Jeffery, D., Bilgin, H., Jarrett, M., Simpson, A., Jones, J. (2008) Psychiatric Intensive
Care Units: a literature review. International Journal of Social Psychiatry. 54(1):56-68.

Crowhurst, N. & Bowers, L. (2002) Philosophy, care and treatment on the psychiatric intensive care
unit: themes, trends and future practice Journal of Psychiatric and Mental Health Nursing. 9,
689695

Duxbury, J., Hahn, S., Needham, I. & Pulsford, D. (2008) The Management of Aggression and
Violence Attitude Scale (MAVAS): a cross-national comparative study. Journal of Advanced
Nursing, 62(5), 596606.

Gamble, C., Dodd, G., Grellier, J., Hever, M., O'Conner, C., Clarke, T., Chipere, R., Mellor, M.,
Ness, M. (2010) Zoning: focused support: a trust wide implementation. Journal of Psychiatric
& Mental Health Nursing. 17(1), 79-86.

Gamble, C. (2006) The zoning revolution. Mental Health Practice. 10 (4) 14-17.

17

Guy, S. & Henderson, B. (2004) The Stockport interim pilot report of zoning. Private
correspondence unpublished internal report.

Hurst, K., Wistow, G. & Higgins, R. (1998) Mental health nursing in acute settings. Mental Health
Practice, 2, 8-11.

Higgins, R., Hurst, K. & Wistow, G. (1999) Nursing acute psychiatric patients: a quantitative and
qualitative study, Journal of Advanced Nursing, 29 (1), 52-63.

Rydon, S. (2005) The attitudes, knowledge and skills needed in mental health nurses: The
perspective of users of mental health services. International Journal of Mental Health Nursing. 14,
78-87.

Ryrie, I. Hellard, L., Kearns, C., Robinson, D., Pathmanathan, I. & OSullivan, D. (1997) Zoning: A
system for managing case work and targeting resources in community mental health teams.
Journal of Mental Health, 6 (5), 515-523.

Taylor, K., Guy, S., Stewart, L., Ayling, M., Miller, G., Anthony, A., Bajuk, A., Le Brun, J., Shearer,
D., Gregory, R., & Thomas, M. (2011) Care Zoning: A Pragmatic Approach to Enhance the
Understanding of Clinical Needs As it Relates to Clinical Risks in Acute In-Patient Unit
Settings. Issues in Mental Health Nursing, 32 (5), 318326.

18

Table 1. Summary of pre- and post-implementation (Care Zoning) audit data 8 bed PICU
Audit Index:
% of records with identified
characteristic
(value in brackets = weight in
Aggregate Quality Score)
Number of Records
Management measures (1)
Patient observation (1)
Risk assessment (1)
Mental state:
Not present (0)
Minimal (1)
Satisfactory (2)
Good (3)
Interventions:
Not present (0)
Minimal (1)
Satisfactory (2)
Good (3)
Evaluation-Outcomes:
Not present (0)
Minimal (1)
Satisfactory (2)
Good (3)
Evidence of 1:1 (1)
Care Plan (1)
Aggregate Quality Score (0-14)
Mean (SD)

Pre-Care Zoning (Pilot)

Post-Implementation
Overall
Chi-square
(Pre vs. Post)

Day
Shift

Afternoon
Shift

Combined

Day
Shift

Afternoon
Shift

Combined

(N = 48)
66.7
43.8
4.2

(N = 48)
43.8
39.6
22.9

(N = 96)
55.2
41.7
13.5

(N = 283)
83.4
83.0
82.3

(N = 289)
80.3
79.9
80.6

(N = 572)
81.8
81.5
81.5

39.6
60.4
0.0
0.0

29.2
70.8
0.0
0.0

34.4
65.6
0.0
0.0

13.8
24.0
30.4
31.8

14.5
19.0
31.1
35.3

14.2
21.5
30.8
33.6

138.92***

75.0
25.0
0.0
0.0

100.0
0.0
0.0
0.0

87.5
12.5
0.0
0.0

15.2
59.0
12.0
13.8

16.6
46.4
21.1
15.9

15.9
52.6
16.6
14.9

219.27***

100.0
0.0
0.0
0.0
41.7
0.0

100.0
0.0
0.0
0.0
58.3
0.0

100.0
0.0
0.0
0.0
50.0
0.0

20.1
40.3
20.1
19.4
60.4
61.1

18.3
36.3
17.6
27.7
68.2
61.6

19.2
38.3
18.9
23.6
64.3
61.4

251.44***

2.42 (0.94)

2.35 (1.06)

2.39 (1.00)

8.14 (3.71)

8.49 (4.08)

8.31 (3.90)

33.91***
70.89***
186.95***

7.19**
124.14***

*** p <0.001; ** p < 0.01.

19

Table 2A. PICU staff Care Zoning survey Questions highly endorsed (> 80%) post-implementation

Survey Questions about the potential

benefits of current risk management strategies
Number of Respondents
Q7 helping clinicians raise intuitive concerns
regarding patients?
Q8 helping clinicians to focus on those patients
who require crisis intervention in a more
systematic manner?
Q5 helping to promote the delivery of targeted
proactive mental health care?
Q3 (helping clinicians) share clinical information
with colleagues?
Q6 helping clinicians identify clinical/risk issues
with patients?
Q4 (helping clinicians) share risk information with
colleagues?
Q10 helping focus the teams work on those
patients whose needs are considered serious?
Q9 preventing catastrophic relapse in patients?

% Reporting a difference
(Some or Noticeable)
Pre-Care
PostOverall
Zoning
Implementation
N=15
N=13
N=28

Corresponding
%
Improvement
from Taylor et
al., 2011

86.7

100.0

92.9

82

86.7

92.3

89.3

71

86.7

92.3

89.3

72

86.7

92.3

89.3

90

93.3

92.3

92.8

84

73.3

84.6

78.6

87

86.7

84.6

85.7

69

86.7

84.6

85.7

60

See Note to Table 2C.

20

Table 2B. PICU staff Care Zoning survey Questions moderately endorsed (60% to 80%) post-implementation

Survey Questions about the potential

benefits of current risk management strategies
Number of Respondents
Q20 providing the opportunity for clinicians to
discuss patients and gain ideas from
colleagues?
Q2 (helping clinicians) receive support from
colleagues?
Q19 providing the opportunity for clinicians to
request help from colleagues with patients?
Q16 complimenting existing clinical supervisory
frameworks?
Q21 providing the opportunity for clinicians to be
told they have done well with a patient?
Q22 providing the opportunity for clinicians to
receive informal peer support during the
remainder of the day?
Q14 promoting a culture of openness enabling
practitioners to discuss and safely voice
uncertainty surrounding patients in the care
they feel able or competent to offer?
Q18 raising clinicians awareness of other
clinicians patients whose needs are currently
considered serious?
Q12 promoting liaison and co-ordination with
others involved in patients care?

% Reporting a difference
(Some or Noticeable)
Pre-Care
PostOverall
Zoning
Implementation
N=15
N=13
N=28

Corresponding
%
Improvement
from Taylor et
al., 2011

66.7

76.9

71.4

72

66.7

76.9

71.4

74

80.0

76.9

78.6

62

73.3

69.2

71.4

47

60.0

61.5

60.7

28

60.0

61.5

60.7

43

73.3

61.5

67.8

68

86.7

61.5

75.0

76

86.7

61.5

75.0

60

See Note to Table 2C.

21

Table 2C. PICU staff Care Zoning survey Questions with lower endorsement (< 60%) post-implementation

Survey Questions about the potential

benefits of current risk management strategies
Number of Respondents
Q11 promoting the continuity of care for patients
whose allocated nurse/care coordinators are
on leave/off sick?
Q15 promoting structure within the team?
Q17 how clinicians manage patient loads?
Q1 (helping clinicians) adhere to policy?
Q13 commitment practitioners have to attend
meetings?
Q23 providing the opportunity for the reallocation
of resources?

% Reporting a difference
(Some or Noticeable)
Pre-Care
PostOverall
Zoning
Implementation
N=15
N=13
N=28

Corresponding
%
Improvement
from Taylor et
al., 2011

80.0

53.8

67.8

54

60.0
73.3
80.0

46.2
46.2
46.2

53.6
60.7
64.3

62
57
75

46.7

38.5

42.9

37

73.3

38.5

57.1

51

Note: There were no statistically significant differences between pre- and post-implementation rates. # Possible concerns lower postimplementation rates. Respective correlations with entries in last column: r (21) = 0.64 (Pre-), 0.62 (Post-) and 0.70 (Overall), p < 0.001.

22