Accepted Manuscript Title: The impact of a flow strategy for patients who presented to an australian emergency department with a mental health illness Author: Nerolie Bost, Julia Crilly, Karen Wallen PII: DOI: Reference:
S1755-599X(15)00009-9 http://dx.doi.org/doi:10.1016/j.ienj.2015.01.005 IENJ 405
To appear in:
International Emergency Nursing
Received date: Revised date: Accepted date:
5-11-2014 29-1-2015 31-1-2015
Please cite this article as: Nerolie Bost, Julia Crilly, Karen Wallen, The impact of a flow strategy for patients who presented to an australian emergency department with a mental health illness, International Emergency Nursing (2015), http://dx.doi.org/doi:10.1016/j.ienj.2015.01.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Title:
2
The impact of a flow strategy for patients who presented to an Australian emergency
3
department with a mental health illness
4
Running Title:
5
The ED mental health patient flow strategy
6
Authors:
7
Nerolie Bost MN, RN (Corresponding)
8
Nurse Researcher
9
Gold Coast Hospital & Health Service & Centre for Health Practice Innovation, Griffith
10
University
11
Emergency Department
12
1 Hospital Boulevard
13
Southport 4215 Qld Australia
14
Telephone: +61 5687 5273
15
Email: [email protected]
16 17
Associate Professor Julia Crilly PhD, RN
18
Centre for Health Practice Innovation, Griffith University &
19
Queensland Emergency Research Collaborative
20
Gold Coast Hospital & Health Service
1
Page 1 of 37
21
Emergency Department
22
1 Hospital Boulevard
23
Southport 4215 Qld Australia
24
Email: [email protected]
25 26 27
Karen Wallen MN (Hons), RN
28
Nurse Educator
29
Gold Coast Hospital & Health Service
30
Mental Health & Integrated Care
31
1 Hospital Boulevard
32
Southport 4215 Qld Australia
33
Email: [email protected]
34 35
Author Contributions:
36
NB contributed to study concept and design, acquisition of data, data analysis,
37
manuscript preparation and revision; JC contributed to study concept and design, data
38
analysis, manuscript preparation and revision; KW contributed to study concept and
39
critical revision of manuscript.
40 41
Competing Interests:
2
Page 2 of 37
42
No potential or actual conflicts of interest have been identified by any of the authors
43
of this manuscript.
44 45 46
TITLE
47
The impact of a flow strategy for patients who presented to an Australian emergency
48
department with a mental health illness
49
HIGHLIGHTS
50 51
Almost 40% of patients diagnosed with mental health illness were admitted from ED
52
A new model of care to improve patient flow from ED was implemented
53
The proportion of patients admitted or discharged from ED within 4 hours increased
54
Time spent waiting to see a medical officer did not change significantly
55
Improvements in service delivery can be realised without additional staffing
56 57 58
ABSTRACT
59
Objectives: To describe and compare characteristics, care delivered, and outcomes of
60
patients who presented to an emergency department (ED) with a mental health illness
61
before and after the implementation of a patient flow strategy.
62
Methods: This was a retrospective, descriptive study. Health care data of patients who
63
presented to a public teaching hospital ED in Queensland, Australia diagnosed with a mental
3
Page 3 of 37
64
health illness before (5th September 2011 – 4th March 2012) and after (5th March 2012 - 4th
65
September 2012) the implementation of a patient flow strategy were analysed.
66
Results: A total of 3,037 (before: n=1511; after: n=1526) mental health presentations (4.5%
67
of all ED presentations) were made to the ED. Following the implementation of a patient
68
flow strategy, improvements in ED length of stay, tests performed and nursing observations
69
were seen. These varied by mental health diagnosis.
70
Conclusion: Our results indicate that a targeted approach to improving service delivery for a
71
specific cohort of ED patients can make a difference without additional staffing. Further
72
focused refinement of the strategy (such as time waiting for treatment) may be required.
73 74
Key words: mental health, emergency department, patient flow
75 76 77 78
Introduction Emergency departments (EDs) are the primary entry point to public hospitals in
79
Australia with patient presentations increasing by an average of 3% each year from 5.7
80
million in 2008 to 6.7 million in 2012 (Australian Institute of Health and Welfare [AIHW],
81
2013). Internationally, patient presentations and acuity are reported to be increasing and,
82
coupled with a finite supply of hospital beds, has contributed to issues of access block (also
83
known as boarding) and over crowding within many EDs (Bond et al., 2007; Forero et al,
84
2011; Pines et al., 2011). In Australia, access block is defined as a situation where “patients
85
are unable to gain access to appropriate hospital beds within a reasonable time, no greater
86
than 8 hrs” (Forero and Hillman, 2008; p4) and leads to ED crowding. ED crowding has been
4
Page 4 of 37
87
linked to inadequate patient care due to prolonged wait times, delays to treatment,
88
communication and medical errors, adverse events and increased risk of in-hospital
89
mortality (Forero and Hillman, 2008; Pines and Hollander, 2008; Sun et al., 2013).
90
Similar to the general ED population in Australia, the numbers of patients
91
presenting to public hospital EDs for mental health (MH) illness have increased by an
92
average of 3% per year, from 236,654 in 2009/10 to 243,444 in 2010/11 (AIHW MH
93
Services, 2012; p2). There has also been a reported increase in MH ED presentations
94
over the last decade in other countries such as USA and Canada (Chang et al., 2012;
95
Leon et al., 2013; Hefflefinger, 2014). Previous studies have suggested that patients
96
who present to the ED for MH illness may be susceptible to longer ED lengths of
97
stays (Atzema et al., 2012; Bost et al., 2014).
98 99
In 2012, a targeted patient flow strategy was introduced at one hospital ED located in Queensland, Australia. The aim of the strategy was to assist with improving time to
100
assessment and treatment for MH patients within the ED, improve flow for patients
101
admitted to MH wards and streamline the discharge process. Expected benefits were a
102
reduced ED LoS (Length of Stay) and a reduction of access block for patients who presented
103
to the ED with a MH illness.
104 105
Mental Health Patient Flow Strategy
106
The new patient flow strategy involved several elements that included: education
107
regarding the use of a mental health triage tool by the ED triage nurse (Aust. Gov. Dept. of
108
Health and Ageing, 2014; [see Table 1]); a streamlined assessment referral process from ED
109
to MH clinicians and the use of the ED short stay ward for patients likely to be admitted.
110
Other elements included the use of a “Mental Health Rapid Emergency Admission
5
Page 5 of 37
111
Destination Initiative” checklist (see Appendix 1) and the introduction of the “pull from the
112
ED service delivery model.”
113
The pull model is based on the concept of hospital wards actively pursuing the
114
transfer of patients from ED (Queensland Government [Qld Gov.], 2014). Via the hospital
115
bed manager, a MH hospital bed is located for the patient to be admitted from ED. A nurse
116
from the MH ward attends the ED, gives a verbal handover from the ED nurse, completes
117
the checklist with the ED nurse and escorts the patient to the ward (see Figure 1).
118 119
INSERT FIGURE 1 ABOUT HERE
120 121
Prior to the new patient flow strategy, the push model had been used where the ED
122
nurse escorted the patient to the ward and transferred the patient and information
123
(handover) to the nurse on the ward. The final element of the strategy was for the admitted
124
patient to be given a brochure that provided names and contact information regarding the
125
MH ward, nurse in charge and treating doctor, patients’ expected date of discharge,
126
allocated community case worker and information for themselves, family and/or carer to
127
use after the patient was discharged (Qld Gov., 2014).
128
The project manager responsible for implementation of the patient flow strategy
129
oversaw the education of the nurses. Education was delivered through informal information
130
sessions at around 14:00 hrs during the change of shift times between the day shift (07:00 to
131
15:30) and the afternoon shift (14:30 to 23:00). ED staff resources were unchanged during
132
the implementation of the strategy.
6
Page 6 of 37
133
Additionally, prior to the implementation of the patient flow strategy, the National
134
Emergency Access Target (NEAT) was introduced as a national performance benchmark. The
135
goal of NEAT is that by 31 December 2015, 90% of patients are discharged from ED,
136
admitted to a hospital ward or transferred to another hospital within 4 hrs (AIHW, 2012).
137
The implementation of the patient flow strategy provided an opportunity to
138
undertake research to answer the following question: What is the impact of a targeted
139
patient flow strategy for patients presenting to the ED with a MH illness in terms of
140
characteristics, care delivered and outcomes (e.g. ED LoS, access block, admission rate)?
141 142
Methods
143
Design and Setting
144
This was a retrospective descriptive study, using a before and after design, of all
145
patients who presented to a Queensland hospital ED with a MH illness between 5th
146
September 2011 and 4th September 2012; six months before and six months after the
147
implementation of the patient flow strategy. The study site was an urban public teaching
148
hospital with over 350 beds that serviced a population of around 280,000 (Australian Bureau
149
of Statistics, 2011).
150 151 152
Data collection Two main data sources were used: i) electronic ED data and ii) hospital health care
153
records. Electronic data were extracted from the ED database by a member of the hospital’s
154
Decision Support Services and provided to the researchers in a Microsoft Excel spreadsheet.
155
Data pertained to MH patient presentations made over the 12 month study period, as
7
Page 7 of 37
156
defined by one of 26 ED International Classification of Diseases -10 diagnostic codes (ICD-10,
157
2010).
158
Along with ICD-10 codes, data also included demographic information (age, gender),
159
mode of arrival, Australasian Triage Scale (ATS) category, date and time of triage, presenting
160
problem, date and time seen and treated, date and time of ED departure. From this data,
161
waiting time to medical assessment and total time in ED was calculated. The ATS is
162
underpinned by guidelines that indicate urgency and acuity where a number corresponds to
163
the recommended timeframe in which a patient should be seen by a doctor in ED
164
(Australasian College of Emergency Medicine, 2013). The original guidelines were adapted
165
to assist nurses to triage patients presenting with MH illness (Australian Government,
166
Department of Health and Ageing, 2014) (see Table 1).
167 168
INSERT TABLE 1 ABOUT HERE
169
To provide further, in-depth, understanding of care given and comparisons
170
between diagnostic groups, additional data were collected from the hospital health
171
care records for patient presentations that were in five of the most frequent ED ICD-
172
10 codes for MH illness: 1. X84: Suicidal ideation/Self harm; 2. F43.9: Emotional
173
crisis; 3. F41.9: Anxiety; 4. F32.9: Depression 5. F20.9: Schizophrenia. Two hundred
174
health care record reviews (100 pre; 100 post) were conducted for each of the five
175
diagnostic groups (1,000 in total) with equal representation between pre and post
176
groups. This sample was selected sequentially from a generated list of hospital health
177
care record numbers from each of the five diagnostic groups within the time frames of
178
pre and post patient flow strategy.
8
Page 8 of 37
179
Specific data collected and entered into an Excel spreadsheet included
180
documentation of the following: A MH diagnosis by the reviewing ED medical
181
officer, co-morbidities (medical or psychiatric), previous case management in the
182
community, drug or alcohol intoxication on presentation, drug or alcohol frequency of
183
use, observations performed while in ED, medical investigations performed while in
184
ED, status of presentation (voluntary or involuntary), use of restraint (if required),
185
discharge disposition (e.g. admitted, discharged or transferred to other facility) and
186
where the patient was referred to for ongoing treatment following discharge. Two
187
trained research nurse assistants collected data from the health care records. Inter-rater
188
reliability testing was undertaken using 20 records and yielded 95% agreement. This study was approved by the Health Service District’s Human Research
189 190
Ethics Committee.
191 192
Data analysis
193
Collected de-identified data were transferred to and analysed using SPSS 21.0.
194
Descriptive statistics (e.g., median, interquartile range, frequencies, percentages) were used
195
to analyse patient demographics, patient information (e.g. age, ATS, diagnosis code, mode of
196
transport, drug and alcohol use), time related outcomes such as ED LoS and care delivery
197
aspects (such as use of restraint and investigations performed). Inferential statistics (t-tests
198
and chi- square tests) were used to compare differences pre and post strategy
199
implementation. A p-value of
S1755-599X(15)00009-9 http://dx.doi.org/doi:10.1016/j.ienj.2015.01.005 IENJ 405
To appear in:
International Emergency Nursing
Received date: Revised date: Accepted date:
5-11-2014 29-1-2015 31-1-2015
Please cite this article as: Nerolie Bost, Julia Crilly, Karen Wallen, The impact of a flow strategy for patients who presented to an australian emergency department with a mental health illness, International Emergency Nursing (2015), http://dx.doi.org/doi:10.1016/j.ienj.2015.01.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Title:
2
The impact of a flow strategy for patients who presented to an Australian emergency
3
department with a mental health illness
4
Running Title:
5
The ED mental health patient flow strategy
6
Authors:
7
Nerolie Bost MN, RN (Corresponding)
8
Nurse Researcher
9
Gold Coast Hospital & Health Service & Centre for Health Practice Innovation, Griffith
10
University
11
Emergency Department
12
1 Hospital Boulevard
13
Southport 4215 Qld Australia
14
Telephone: +61 5687 5273
15
Email: [email protected]
16 17
Associate Professor Julia Crilly PhD, RN
18
Centre for Health Practice Innovation, Griffith University &
19
Queensland Emergency Research Collaborative
20
Gold Coast Hospital & Health Service
1
Page 1 of 37
21
Emergency Department
22
1 Hospital Boulevard
23
Southport 4215 Qld Australia
24
Email: [email protected]
25 26 27
Karen Wallen MN (Hons), RN
28
Nurse Educator
29
Gold Coast Hospital & Health Service
30
Mental Health & Integrated Care
31
1 Hospital Boulevard
32
Southport 4215 Qld Australia
33
Email: [email protected]
34 35
Author Contributions:
36
NB contributed to study concept and design, acquisition of data, data analysis,
37
manuscript preparation and revision; JC contributed to study concept and design, data
38
analysis, manuscript preparation and revision; KW contributed to study concept and
39
critical revision of manuscript.
40 41
Competing Interests:
2
Page 2 of 37
42
No potential or actual conflicts of interest have been identified by any of the authors
43
of this manuscript.
44 45 46
TITLE
47
The impact of a flow strategy for patients who presented to an Australian emergency
48
department with a mental health illness
49
HIGHLIGHTS
50 51
Almost 40% of patients diagnosed with mental health illness were admitted from ED
52
A new model of care to improve patient flow from ED was implemented
53
The proportion of patients admitted or discharged from ED within 4 hours increased
54
Time spent waiting to see a medical officer did not change significantly
55
Improvements in service delivery can be realised without additional staffing
56 57 58
ABSTRACT
59
Objectives: To describe and compare characteristics, care delivered, and outcomes of
60
patients who presented to an emergency department (ED) with a mental health illness
61
before and after the implementation of a patient flow strategy.
62
Methods: This was a retrospective, descriptive study. Health care data of patients who
63
presented to a public teaching hospital ED in Queensland, Australia diagnosed with a mental
3
Page 3 of 37
64
health illness before (5th September 2011 – 4th March 2012) and after (5th March 2012 - 4th
65
September 2012) the implementation of a patient flow strategy were analysed.
66
Results: A total of 3,037 (before: n=1511; after: n=1526) mental health presentations (4.5%
67
of all ED presentations) were made to the ED. Following the implementation of a patient
68
flow strategy, improvements in ED length of stay, tests performed and nursing observations
69
were seen. These varied by mental health diagnosis.
70
Conclusion: Our results indicate that a targeted approach to improving service delivery for a
71
specific cohort of ED patients can make a difference without additional staffing. Further
72
focused refinement of the strategy (such as time waiting for treatment) may be required.
73 74
Key words: mental health, emergency department, patient flow
75 76 77 78
Introduction Emergency departments (EDs) are the primary entry point to public hospitals in
79
Australia with patient presentations increasing by an average of 3% each year from 5.7
80
million in 2008 to 6.7 million in 2012 (Australian Institute of Health and Welfare [AIHW],
81
2013). Internationally, patient presentations and acuity are reported to be increasing and,
82
coupled with a finite supply of hospital beds, has contributed to issues of access block (also
83
known as boarding) and over crowding within many EDs (Bond et al., 2007; Forero et al,
84
2011; Pines et al., 2011). In Australia, access block is defined as a situation where “patients
85
are unable to gain access to appropriate hospital beds within a reasonable time, no greater
86
than 8 hrs” (Forero and Hillman, 2008; p4) and leads to ED crowding. ED crowding has been
4
Page 4 of 37
87
linked to inadequate patient care due to prolonged wait times, delays to treatment,
88
communication and medical errors, adverse events and increased risk of in-hospital
89
mortality (Forero and Hillman, 2008; Pines and Hollander, 2008; Sun et al., 2013).
90
Similar to the general ED population in Australia, the numbers of patients
91
presenting to public hospital EDs for mental health (MH) illness have increased by an
92
average of 3% per year, from 236,654 in 2009/10 to 243,444 in 2010/11 (AIHW MH
93
Services, 2012; p2). There has also been a reported increase in MH ED presentations
94
over the last decade in other countries such as USA and Canada (Chang et al., 2012;
95
Leon et al., 2013; Hefflefinger, 2014). Previous studies have suggested that patients
96
who present to the ED for MH illness may be susceptible to longer ED lengths of
97
stays (Atzema et al., 2012; Bost et al., 2014).
98 99
In 2012, a targeted patient flow strategy was introduced at one hospital ED located in Queensland, Australia. The aim of the strategy was to assist with improving time to
100
assessment and treatment for MH patients within the ED, improve flow for patients
101
admitted to MH wards and streamline the discharge process. Expected benefits were a
102
reduced ED LoS (Length of Stay) and a reduction of access block for patients who presented
103
to the ED with a MH illness.
104 105
Mental Health Patient Flow Strategy
106
The new patient flow strategy involved several elements that included: education
107
regarding the use of a mental health triage tool by the ED triage nurse (Aust. Gov. Dept. of
108
Health and Ageing, 2014; [see Table 1]); a streamlined assessment referral process from ED
109
to MH clinicians and the use of the ED short stay ward for patients likely to be admitted.
110
Other elements included the use of a “Mental Health Rapid Emergency Admission
5
Page 5 of 37
111
Destination Initiative” checklist (see Appendix 1) and the introduction of the “pull from the
112
ED service delivery model.”
113
The pull model is based on the concept of hospital wards actively pursuing the
114
transfer of patients from ED (Queensland Government [Qld Gov.], 2014). Via the hospital
115
bed manager, a MH hospital bed is located for the patient to be admitted from ED. A nurse
116
from the MH ward attends the ED, gives a verbal handover from the ED nurse, completes
117
the checklist with the ED nurse and escorts the patient to the ward (see Figure 1).
118 119
INSERT FIGURE 1 ABOUT HERE
120 121
Prior to the new patient flow strategy, the push model had been used where the ED
122
nurse escorted the patient to the ward and transferred the patient and information
123
(handover) to the nurse on the ward. The final element of the strategy was for the admitted
124
patient to be given a brochure that provided names and contact information regarding the
125
MH ward, nurse in charge and treating doctor, patients’ expected date of discharge,
126
allocated community case worker and information for themselves, family and/or carer to
127
use after the patient was discharged (Qld Gov., 2014).
128
The project manager responsible for implementation of the patient flow strategy
129
oversaw the education of the nurses. Education was delivered through informal information
130
sessions at around 14:00 hrs during the change of shift times between the day shift (07:00 to
131
15:30) and the afternoon shift (14:30 to 23:00). ED staff resources were unchanged during
132
the implementation of the strategy.
6
Page 6 of 37
133
Additionally, prior to the implementation of the patient flow strategy, the National
134
Emergency Access Target (NEAT) was introduced as a national performance benchmark. The
135
goal of NEAT is that by 31 December 2015, 90% of patients are discharged from ED,
136
admitted to a hospital ward or transferred to another hospital within 4 hrs (AIHW, 2012).
137
The implementation of the patient flow strategy provided an opportunity to
138
undertake research to answer the following question: What is the impact of a targeted
139
patient flow strategy for patients presenting to the ED with a MH illness in terms of
140
characteristics, care delivered and outcomes (e.g. ED LoS, access block, admission rate)?
141 142
Methods
143
Design and Setting
144
This was a retrospective descriptive study, using a before and after design, of all
145
patients who presented to a Queensland hospital ED with a MH illness between 5th
146
September 2011 and 4th September 2012; six months before and six months after the
147
implementation of the patient flow strategy. The study site was an urban public teaching
148
hospital with over 350 beds that serviced a population of around 280,000 (Australian Bureau
149
of Statistics, 2011).
150 151 152
Data collection Two main data sources were used: i) electronic ED data and ii) hospital health care
153
records. Electronic data were extracted from the ED database by a member of the hospital’s
154
Decision Support Services and provided to the researchers in a Microsoft Excel spreadsheet.
155
Data pertained to MH patient presentations made over the 12 month study period, as
7
Page 7 of 37
156
defined by one of 26 ED International Classification of Diseases -10 diagnostic codes (ICD-10,
157
2010).
158
Along with ICD-10 codes, data also included demographic information (age, gender),
159
mode of arrival, Australasian Triage Scale (ATS) category, date and time of triage, presenting
160
problem, date and time seen and treated, date and time of ED departure. From this data,
161
waiting time to medical assessment and total time in ED was calculated. The ATS is
162
underpinned by guidelines that indicate urgency and acuity where a number corresponds to
163
the recommended timeframe in which a patient should be seen by a doctor in ED
164
(Australasian College of Emergency Medicine, 2013). The original guidelines were adapted
165
to assist nurses to triage patients presenting with MH illness (Australian Government,
166
Department of Health and Ageing, 2014) (see Table 1).
167 168
INSERT TABLE 1 ABOUT HERE
169
To provide further, in-depth, understanding of care given and comparisons
170
between diagnostic groups, additional data were collected from the hospital health
171
care records for patient presentations that were in five of the most frequent ED ICD-
172
10 codes for MH illness: 1. X84: Suicidal ideation/Self harm; 2. F43.9: Emotional
173
crisis; 3. F41.9: Anxiety; 4. F32.9: Depression 5. F20.9: Schizophrenia. Two hundred
174
health care record reviews (100 pre; 100 post) were conducted for each of the five
175
diagnostic groups (1,000 in total) with equal representation between pre and post
176
groups. This sample was selected sequentially from a generated list of hospital health
177
care record numbers from each of the five diagnostic groups within the time frames of
178
pre and post patient flow strategy.
8
Page 8 of 37
179
Specific data collected and entered into an Excel spreadsheet included
180
documentation of the following: A MH diagnosis by the reviewing ED medical
181
officer, co-morbidities (medical or psychiatric), previous case management in the
182
community, drug or alcohol intoxication on presentation, drug or alcohol frequency of
183
use, observations performed while in ED, medical investigations performed while in
184
ED, status of presentation (voluntary or involuntary), use of restraint (if required),
185
discharge disposition (e.g. admitted, discharged or transferred to other facility) and
186
where the patient was referred to for ongoing treatment following discharge. Two
187
trained research nurse assistants collected data from the health care records. Inter-rater
188
reliability testing was undertaken using 20 records and yielded 95% agreement. This study was approved by the Health Service District’s Human Research
189 190
Ethics Committee.
191 192
Data analysis
193
Collected de-identified data were transferred to and analysed using SPSS 21.0.
194
Descriptive statistics (e.g., median, interquartile range, frequencies, percentages) were used
195
to analyse patient demographics, patient information (e.g. age, ATS, diagnosis code, mode of
196
transport, drug and alcohol use), time related outcomes such as ED LoS and care delivery
197
aspects (such as use of restraint and investigations performed). Inferential statistics (t-tests
198
and chi- square tests) were used to compare differences pre and post strategy
199
implementation. A p-value of
