Abstract
Background. The Manchester Triage Scale is used in Irish emergency departments. This fails to provide guidance on triaging psychiatric presentations. A Mental Health Triage scale is recommended by the National Institute of Clinical Excellence. Aim. To examine the effectiveness of a Mental Health Triage scale in assessing patients presenting with self-harm. Method. Ten vignettes were created, detailing cases of deliberate self-harm. Nurses () were given five vignettes and asked to assign each vignette to a triage category, using The Manchester Triage Scale. Each nurse was subsequently asked to reevaluate the same vignettes using the Mental Health Triage Scale. Triage with each method was deemed safe or unsafe, using the benchmark triage categories assigned by a consultant in psychiatry and a consultant in emergency medicine departments. Results. 245 cases were triaged. There was a significant change in the categories assigned when the Mental Health Triage scale was in use, . The triage categories assigned using the Mental Health Triage scale were significantly safer than under the Manchester Triage Scale (79% versus 60% safe, respectively, ).
1. Introduction
1.1. Background
The National Institute of Excellence in Britain defines self-harm as “intentional self-poisoning or injury, irrespective of the apparent purpose of the act.” The incidence in Ireland of deliberate self-harm has increased dramatically. In the four-year period since 2007, there has been a 27% increase in the incidence of deliberate self-harm cases in men and a 17% increase in women presenting to Irish emergency departments [1]. Furthermore, in 2011 alone, there were 12,216 cases of deliberate self-harm presentations to Irish emergency departments [1].
Deliberate self-harm is the single most important risk factor for suicide [2, 3]. A mortality follow-up study of 11,586 patients in the United Kingdom found that the risk of suicide in the first year following an act of deliberate self-harm was 0.7%, which was 66 times the annual risk of suicide in the general population [3]. Furthermore, a recent study found that physical health and life expectancy are severely compromised in individuals who self-harm compared with the general population [4].
Triage in Irish emergency departments is conducted using the Manchester Triage Scale. This focuses on medical and surgical presentations but fails to provide guidance on the triage of patients with psychiatric presentations. This is despite the fact that up to 5% of people attending the emergency departments present with primary psychiatric problems and another 20–30% have psychiatric symptoms in addition to physical symptoms [5]. The National Institute of Clinical Excellence (NICE) recommends that consideration should be given to introducing the Australian Mental Health Triage scale [6]. It noted that this is a comprehensive assessment scale for rating clinical urgency [6]. In Ireland, the National Emergency Medicine Programme Report (June 2012) noted the need for development of guidelines on mental health triage.
Smart et al. introduced a Mental Health Triage scale in 1994 at the Royal Hobart Hospital, Tasmania. Prior to implementation of the Mental Health Triage scale, patients with mental health presentations had longer waiting times compared to those with medical and surgical presentations, and they were more likely to receive a lower triage score [7]. The introduction of the Mental Health Triage scale led to a statistically significant () reduction in waiting time and posttrial reduction in the number of people who “did not wait” () [7]. In 2001, Broadbent introduced a temporary Mental Health Triage scale which reinforced the findings of Smart et al. With regard to waiting time, after implementation, 88.8% of patients were seen in an appropriate time frame compared with only 26% prior to implementation [8]. Broadbent noted that 70.9% () of patients were initially given triage codes that misrepresented the acuity of their presentation, most () received a triage code that assigned less priority than that may have been allocated with the Mental Health Triage scale [8].
The primary objective was to evaluate if a Mental Health Triage scale would facilitate the safe triage of patients presenting with deliberate self-harm. Secondary objectives included an assessment of familiarity with the concept of mental health triage, the impact of years’ work experience on accuracy of a new triage scale, and the exposure of nurses to psychiatric training.
2. Methods
2.1. Setting and Participants
This was an experimental interventional study. This study was carried out at Cork University Hospital Emergency Department. This hospital provides acute care for an urban and rural catchment area of Cork City and County. It also acts as a referral hospital for hospitals located in the Munster area. Cork University Hospital handles over 58,000 emergency presentations annually. Liaison psychiatry services are located on site.
2.2. Data Collection
Emergency department staff nurses participated in this study. Inclusion criteria for participants included emergency department staff nurses only, participants must have prior experience in triage, and participants must give informed consent before participating. Agency staff nurses were not included. Participants were approached during their shift and invited to participate in the study.
Ten vignettes detailing cases of deliberate self-harm were created (Appendix A). The pertinent details from the case files of 10 patients who presented over a 3-week period in March 2012 to Cork University Hospital Emergency Department provided the information for these original vignettes. Each vignette contained information regarding the background provided, information on arrival, and the behaviour witnessed. Vignettes were laid out on an A4 page and laminated with one vignette per page. The cases chosen reflected a variety of presentations and severity of deliberate self-harm. Original data sheets were created to record responses during data collection. This Mental Health Triage scale used is derived from the Australian Mental Health Triage scale and is recommended by the National Institute of Excellence guidelines on self-harm (Appendix B). A copy of the Manchester Triage Scale was not provided to nurses. Both scales categorise patients based on their presentation findings. Each category has as a recommended time frame in which the patient should be seen.
Two expert opinions, namely, a consultant Psychiatrist and a consultant in Emergency Medicine, assigned what they considered to be the appropriate triage categories to the vignettes. The triage categories assigned acted as the gold standard to compare the triage categories assigned by nurses. The consultant Psychiatrist is taken as having assigned the correct MHTS category and the consultant in Emergency Medicine is taken as having assigned the correct Manchester Triage Scale category.
The data was collected over the space of an 11-day period from the 21st September to the 1st October, 2012, inclusively. All data was obtained at Emergency Department, Cork University Hospital.
Each nurse was then asked to assign a triage category to five of the ten vignettes using the existing Manchester Triage Scale. Each nurse triaged vignettes detailing two cases considered to be a mental health triage category one and the remaining three detailed cases represented triage categories two, three, and four. Nurses were then asked if they were familiar with the concept of mental health triage.
Each nurse was subsequently shown a copy of the Mental Health Triage scale. Following this, they were asked to reevaluate the same five vignettes and to assign a triage category again; keeping in mind the Mental Health Triage scale, they had been shown. Participants were allowed to refer to the Mental Health Triage scale when reevaluating the vignettes. The comparison of triage categories assigned by emergency department nurses to the benchmark set by the Psychiatric consultant and Emergency Medicine consultant allowed an evaluation of how cases were triaged. If a nurse assigned a triage category of the same priority or greater, they were considered to have triaged the case safely. Finally, nurses were asked to indicate how long it was since their last formal psychiatric training. Nurses were asked not to discuss the study with other staff members until data collection was complete to maintain the validity of the research.
2.3. Ethical Approval
Ethical approval was granted by the Cork Research Ethics Committee (CREC) on the 3 April, 2012. Permission was obtained from the Accident and Emergency Department in Cork University Hospital.
2.4. Statistical Analysis
Data was entered into Microsoft Excel and subsequently imported to IBM SPSS Statistics 20 and coded. Frequency analysis and chi-square analysis were carried out. One sample Wilcoxon test was used to compare the triage categories assigned with the gold standard set by the Psychiatrist.
3. Results
3.1. Descriptive Statistics
There were 85 emergency department staff nurses on the monthly working roster at Cork University Hospital during the study period. 70 nurses made up the study population after exclusion criteria were applied. Of these, 49 (70%) participated in the study.
The number of years’ experience working in the emergency department was recorded for each nurse as was familiarity with the concept of mental health triage; see Table 1 and Figure 1, respectively.
3.2. Change in Triage Categories Assigned after Using Mental Health Triage Scale
Each nurse () assigned a triage category to five vignettes, twice. Thus, 245 cases were triaged under the existing Manchester Triage Scale and the same 245 cases were retriaged under the Mental Health Triage scale. There was a statistically significant difference between triage scores assigned under the existing Manchester Triage Scale compared to when cases were retriaged using the Mental Health Triage scale, . Table 2 shows the change in distribution of triage categories assigned using the Manchester Triage Scale and the Mental Health Triage scale.
Vignette four and eight showed the most notable change in triage score assigned. Tables 3 and 4 show the change in distribution of triage categories assigned by nurses who triaged cases four and eight, respectively. Using chi-square analysis, the change in triage categories assigned for vignette four is statistically significant, , and likewise the change for vignette eight is statistically significant.
A nonparametric one sample Wilcoxon test was employed to compare how vignettes were triaged compared to the gold standard set by the Psychiatrist when the Mental Health Triage scale was in use (Table 5).
For five of the ten vignettes (3, 6, 7, 9, and 10), there is not sufficient evidence to say that nurses disagree with the gold standard ().
3.3. Safety of Triage Categories Assigned
A greater number of patients were triaged safely using the Mental Health Triage scale, when the triage category assigned by the Psychiatrist was used as a benchmark. When vignettes were first evaluated, 146 cases (60%) were triaged safely. Subsequently, 194 cases (79%) were triaged safely when the Mental Health Triage scale was in use. The change was found to be statistically significant, .
The two vignettes that were assigned a triage category one by the Psychiatrist were triaged notably safer when nurses had knowledge of the Mental Health Triage scale. For vignette four, 14.6% () of nurses were considered to have triaged the case safely initially compared with 55% () of nurses when the Mental Health Triage scale was in use, . An even greater change was seen for vignette eight, 6% () of nurses triaged safely initially compared to 75.5% () of nurses when reevaluated, .
An equal number of nurses triaged vignettes one, two, three, and five as safe before and after viewing the Mental Health Triage scale. For vignettes 6, 7, 9, and 10, there was a reduction in the number of cases triaged as safe when vignettes were reevaluated using the Mental Health Triage scale. However, the reduction was negligible and none were statistically significant.
The triage categories assigned by nurses were also compared to benchmark triage categories assigned by the Emergency Medicine consultant using the Manchester Triage Scale. Initially, 164 cases (67%) were triaged safely. On reevaluating the vignettes, 168 cases (68.5%) were considered to have been triaged as safe. This change was not statistically significant, .
The effect of years’ experience working in the emergency department was examined. 70% of cases triaged by those with 11–15 years’ working experience were considered to have received a safe triage category in the initial assessment. Only 53% of cases were initially triaged safely by those in the one- to five-year group and 21+ year’s group. When vignettes were reevaluated using the Mental Health Triage scale, 76% of cases triaged by those in the 11–15 years’ experience group were safe. The number of cases triaged safely increased for all groups with the greatest increase seen in the one to five (23% increase).
4. Discussion
This study has two principle findings. Firstly, there was a significant change in the triage categories assigned to cases when nurses had knowledge of the Mental Health Triage scale. Secondly, there was an increase in the number of cases that were triaged safely when the Mental Health Triage scale was used. Additional findings included the lack of familiarity with mental health triage and the apparent lack of recent formal psychiatric training for emergency department nurses.
4.1. Assigned Triage Categories
The change in the triage categories assigned when nurses reevaluated the vignettes was statistically significant. There was a significant uptriaging of cases. The number of cases which assigned a triage category one increased by 22.5% and the number which assigned a triage category two or three decreased. Notably, there was a 10% increase in the number of cases which assigned a triage category four. These findings correlate largely with findings published by Happell et al. [9]. The uptriaging of patients represents a challenge to the emergency department. More patients will need immediate attention from a service already stretched and lacking resources for adequate patient care. However, it would appear logical that all patients, regardless of their presenting complaint, should be treated equally and evidence-based triage guidelines should be followed.
4.2. Safety of Assigned Triage Categories
The number of cases considered to be triaged safely increased by 19% when the Mental Health Triage scale was used. Vignettes four and eight made a substantial contribution to this figure. These two vignettes detailed presentations that warranted a triage category one from the Psychiatric consultant. Each nurse was intentionally given both these vignettes to triage as we wanted in particular to assess how nurses triaged serious psychiatric presentations. The significant initial undertriaging of patients with the existing Manchester Triage Scale for these vignettes is concerning. Patients that receive an inappropriately low priority triage category are at a greater risk of harming themselves and others.
The results following the introduction of the Mental Health Triage scale were by no means perfect; 21% of cases were still triaged unsafely. This may be explained by the fact that this is the first time the nurses got to see the Mental Health Triage scale. Alternatively, it may indicate that education is needed to help nurses interpret and apply these new guidelines correctly. Happell et al. noted that experienced psychiatric nurses were better at using the Mental Health Triage scale than their emergency department nonpsychiatric nursing colleagues [9]. A more in-depth analysis may uncover the reason and provide guidance on how to reduce the number triaged unsafely even further.
4.3. Familiarity and Education in Mental Health Triage
The relative lack of familiarity with the concept on mental health triage is evident in this study. Only 41% indicated familiarity, which is concerning given the reported prevalence of psychiatric admissions to the emergency department in the literature [1, 5]. This problem is confounded by the lack of formal psychiatric training for emergency department staff nurses. The majority (87.7%) of nurses in this study reported that they had received no formal psychiatric training in the last five years. The lack of education has also been noted in the literature; in the University of Melbourne, it was found that only 38% of the nurses had been given education or training in mental health triage [10]. This should be a target area in order to provide nurses with the optimal skills to manage those with psychiatric problems.
4.4. Strengths and Weaknesses
There are some limitations to this study. Nurses could have been subjected to the Hawthorne effect. Nurses were instructed to triage the cases as if the case had presented to Emergency Department on a normal day to minimise this limitation. The fact that nurses may be more familiar with the details of the case on reevaluation may be a limitation. Reevaluation can often lead to a change in a person’s decision without intervention and encourage more conservative answering. There is also the potential that nurses may have discussed the content of vignettes and the triage categories they assigned to individual cases. Nurses were asked to avoid discussing the content of the study with other staff and the duration of the study was limited to 11 days to help mitigate this limitation.
The strengths of this project rest in three main areas. Firstly, this was a novel study. No previous study has directly intervened and compared how nurses would triage the same case under existing guidelines and a Mental Health Triage scale. Secondly, real life cases were used to create the vignettes. These cases were exactly the type of cases that emergency triage nurses deal with on a day to day basis. Thirdly, nurses were intentionally interviewed during their shift. Thus, nurses were evaluated in the same environment, in which they would actually carry out triage. Furthermore, it eliminated the formal atmosphere of a lecture room and the potential for nurses to influence each other’s answers.
Future similar research in multicentre sites with larger sample sizes may reproduce our findings with greater validity. A more in-depth analysis of education provided to nurses is also needed. We need to see why nurses are not receiving continued education to keep their skills up to date. Education is likely to form the cornerstone to any advances in the area of psychiatric triage.
5. Conclusion
We have established that the introduction of the Mental Health Triage scale would influence how nurses triage patients with psychiatric presentations. The change was largely positive with a greater proportion of patients receiving a safe triage category. This study has identified a paucity of familiarity with mental health triage and education in the area of psychiatric triage. These findings support the need for the modification of existing triage guidelines and further education in the area of mental health triage.
Appendices
A. Vignettes
A.1. Vignette 1
A.1.1. Background Provided
(i)23-year-old female BIBA at 03.30 having taken an overdose of friend’s medication at around 01.30 Zopiclone (Zimovane) 7.5 mg × 14, Sertraline (Lustral) 100 mg × 15.(ii)She had also consumed alcohol.(iii)She contacted her sister who called the ambulance.(iv)Sister said she took another overdose 6/12 ago.
A.1.2. On Arrival
(i)Airway clear.(ii)RR 16.(iii)HR 100, regular.(iv)Sats 100% r/a.(v)BP 108/78.(vi)T-35.5.(vii)GCS 15/15.(viii)PEARL.(ix)Glucose 5.6.(x)No LOC before arrival of ambulance.(xi)No vomiting.
A.1.3. Behaviour
(i)Mildly intoxicated but coherent.(ii)Mildly Anxious.(iii)Cooperative.
A.2. Vignette 2
A.2.1. Background Provided
(i)30-year-old male self-referral to the emergency department at 22.30.(ii)He took an overdose of his prescribed medications (Mirtazapine (Zispin) 270 mg and Escitalopram (Lexapro) 30 mg) at 21.15.(iii)He had also consumed alcohol.(iv)After taking the medications, he informed his mother and she brought him to the emergency department.(v)His mother found a note and said he has been attending psychiatric services.
A.2.2. On Arrival
(i)HR 110.(ii)RR 18.(iii)T-36.6.(iv)Sats 98% r/a.(v)BP 117/85.(vi)GCS 15/15.(vii)No vomiting.
A.2.3. Behaviour
(i)Intoxicated.(ii)Withdrawn.(iii)Expressing a wish to die.(iv)Not sure he wants to stay for treatment.
A.3. Vignette 3
A.3.1. Background Provided
(i)41-year-old male BIBA at 02.20. Found by gardai lying on the ground in a car park for an unknown amount of time.(ii)Patient was confused and drowsy with cool extremities when found.(iii)Patient consumed alcohol and took an unknown quantity of his parents Zopiclone (Zimovane) and Zolpidem (Stilnoct).(iv)Exwife received texts earlier that night from patient threatening to kill himself.(v)Patient also took an overdose of anti-inflammatories 1/12 ago.
A.3.2. On Arrival
(i)Airway clear.(ii)HR 82, regular.(iii)RR 16.(iv)Sats 100% r/a.(v)BP 119/65.(vi)T-34.4.(vii)GCS 15/15.(viii)Laceration above his left eye.
A.3.3. Behaviour
(i)Appears confused.(ii)Intermittent agitation and restlessness.(iii)Verbally aggressive threatening staff when approached.
A.4. Vignette 4
A.4.1. Background Provided
(i)49-year-old female BIBA at 10.00.(ii)Found by her son unresponsive in bed. Left a suicide note.(iii)Empty packets of Olanzapine 10 mg × 20, Venlafaxine (Effexor) 3 g total, and Paracetamol 500 mg × 36.(iv)Recent depression and treatment in a Psychiatric Unit.
A.4.2. On Arrival
(i)Airway clear.(ii)HR 62, regular.(iii)RR 24.(iv)T-36.1.(v)Sats 95% r/a.(vi)BP 101/65.(vii)GCS 12/15 on scene, 14/15 on arrival in emergency department.(viii)Slurred speech.(ix)Glucose 4.8.
A.4.3. Behaviour
(i)Withdrawn, poor communication.(ii)Said she is a bad person who deserves to go to hell.(iii)Witnessed wrapping a drip line around her neck in ED.
A.5. Vignette 5
A.5.1. Background Provided
(i)28-year-old female BIBA at 17.00 following an overdose at 13.30 of her own medication (Xanax 0.25 mg × 30; Paracetamol 500 mg × 12; Inderal 10 mg × 100; Spasmonal 60 mg × 20).(ii)She was found by her partner at 14.00 who called ambulance.(iii)She had sent texts to relatives.(iv)Past history of anxiety. On Lexapro and Xanax regularly.
A.5.2. On Arrival
(i)Airway clear.(ii)HR 60, regular, sinus rhythm with PVCs.(iii)RR 16.(iv)Sats 98% in r/a.(v)BP 113/74.(vi)GCS 15/15.(vii)Glucose 5.3.(viii)Vomit × 1.
A.5.3. Behaviour
(i)Mildly drowsy but coherent.(ii)Depressed but appears remorseful.
A.6. Vignette 6
A.6.1. Background Provided
(i)17-year-old female self-referral to the emergency department at 22.45 following an overdose of Piriton (10–12 × 4 mg) taken at 22.15.(ii)She has vomited and is nauseous.(iii)She reports being stressed at school. No known psychiatric history.
A.6.2. On Arrival
(i)Airway clear.(ii)BP 105/65.(iii)HR 87, regular.(iv)Sats 96% r/a.(v)RR 18.(vi)T-36.1.(vii)GCS 15/15.
A.6.3. Behaviour
(i)Anxious and restless.(ii)Wants to go home.
A.7. Vignette 7
A.7.1. Background Provided
(i)30-year-old male BIBA at 02.50.(ii)Found hanging from a door frame by bed sheets at around 01.45.(iii)CPR performed by the family as he was initially unresponsive but responsive when the ambulance arrived.(iv)Had consumed alcohol and illicit drugs earlier.(v)Got recently out of prison. Known history of alcohol and benzodiazepine abuse.
A.7.2. On Arrival
(i)Airway clear.(ii)HR 60, regular.(iii)RR 12, shallow breaths.(iv)Sats 99% r/a.(v)BP 118/72.(vi)Glucose 4.8.(vii)GCS 10/15 (E2, V3, M5).
A.7.3. Behaviour
(i)Agitated in response to painful stimuli.(ii)Currently otherwise uncommunicative.
A.8. Vignette 8
A.8.1. Background Provided
(i)35-year-old male BIBA at 23.00(ii)Taken from Rive Lee by a Garda after seen jumping from bridge 5 minutes earlier.(iii)Restrained by Gardai as physically aggressive at the scene.(iv)Known to attend mental health services.
A.8.2. On Arrival
(i)Airway clear.(ii)HR 60, regular, sinus rhythm.(iii)RR 16.(iv)Sats 98% in r/a.(v)T 35.5 C.(vi)BP 110/70.(vii)GCS 15/15.(viii)Glucose 5.3.
A.8.3. Behaviour
(i)Is very agitated and appears afraid.(ii)Is shouting incessantly about the devil.(iii)Has verbal and physical aggression when approached.(iv)Security present.
A.9. Vignette 9
A.9.1. Background Provided
(i)44-year-old female BIBA at 10 a.m.(ii)Jumped out from second floor window of home, found by husband at 9 am.(iii)History of epilepsy, cluster of seizures previous day, behaving strangely.(iv)No known psychiatric history.
A.9.2. On Arrival
(i)Airway clear.(ii)RR 16.(iii)BP 108/67.(iv)HR 80, regular.(v)Sats 100% r/a.(vi)T-35.5.(vii)GCS 15/15.(viii)Obvious pain on movement of right leg.
A.9.3. Behaviour
(i)Very agitated and moaning.(ii)Appears confused and frightened when approached.(iii)Reciting Prayer “Hail Mary” repeatedly.
A.10. Vignette 10
A.10.1. Background Provided
(i)50-year-old male self-referral at 13.00 with self-inflicted cuts on both legs and having drunk 100 mls of bleach.(ii)No prior history.
A.10.2. On Arrival
(i)Airway clear.(ii)HR 84, regular.(iii)RR 18.(iv)Sats 97% r/a.(v)BP 110/75.(vi)T-36.4.(vii)Glucose 5.2.(viii)GCS 15/15.
A.10.3. Behaviour
(i)Anxious man, wants help.(ii)Insisting he has an infestation under his skin.(iii)Think it will kill him if he does not get rid of it.
B. Mental Health Triage Scale
See Figure 2.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.