International Emergency Nursing (2013) 21, 89–96

Avai lab le a t www.sc ienced i rec t .com

journal homepage: www.elsevierheal th .com/ journals /aaen

Triage assessment of registered nurses inthe emergency department

Torunn Kitty Vatnøy RN, MSc a,*, Mariann Fossum RN, PhD a,b,Nina Smith RN, MSc c, Ashild Slettebø RN, PhD a

a Department of Health and Nursing Science, Faculty of Health and Sports Sciences, University of Agder, Grimstad, Norwayb School of Health and Medical Sciences, Orebro University, Orebro, Swedenc Sørlandet Hospital, Arendal, Norway

Received 22 February 2012; received in revised form 15 June 2012; accepted 18 June 2012

17ht

*

P.

KEYWORDSEmergency department;Nursing assessment;Triage;Vital signs

55-599X/$ - see front matttp://dx.doi.org/10.1016/j.i

Corresponding author. AddO. Box 509, 4898 Grimstad,

E-mail address: torunn.vat

er ª 201enj.2012

ress: DeNorway.

noy@uia

Abstract

Standardised triage systems have been implemented in emergency departments (EDs) toimprove the efficacy of assessment strategies as performed by registered nurses (RNs). How-ever, the exact effect the standardised triage systems have on the decision-making processremains unclear.Aim: To evaluate decision making in the triage setting before and after implementation of theMedical Emergency Triage and Treatment System Adult in one hospital’s ED.Methods: A descriptive intervention design with a quantitative approach. A total of 655patients before and 413 patients after the intervention were included. A questionnaire wasused to evaluate how the RNs assessed the patients before intervention while the emergencypatient records were used for data collection after intervention.Results: Before the intervention, a majority of the assessments were founded on signs andsymptoms and medical diagnoses, whereas vital parameters were rarely used. After the inter-vention, nearly two thirds of the patients were assessed according to a triage system with vitalparameters and standardised algorithm for symptoms and signs included in the assessment pro-cedure.Conclusion: Implementing a standardised triage system, including vital parameters and stand-ardised algorithms for signs and symptoms, increased the use of vital parameters and signs andsymptoms for decision making and acuity assignment.ª 2012 Elsevier Ltd. All rights reserved.

2 Elsevier Ltd. All rights reserved..06.004

partment of Health and Nursing Science, Faculty of Health and Sports Sciences, University of Agder,Tel.: +47 48129644/37233775.

.no (T.K. Vatnøy).

90 T.K. Vatnøy et al.

Introduction

The aim of triage assessment in the initial encounter be-tween the nurse and the patient in emergency departments(EDs) is to quickly determine and classify the patients in theorder of urgency based on the need for treatment (Brabrandet al., 2010; Gilboy et al., 1999; Huryk, 2006). Becauseaccuracy in triage is critical, acuity assignment affects theprioritisation of limited medical resources among patientsin acute need of medical care. Triage assessment consti-tutes a challenge and responsibility for nurses in EDs. Be-cause patients have unknown and potentially very seriousillnesses, there is a high degree of uncertainty and acuitythat complicates the assessment process (Goranssonet al., 2008; Hale and Tippett, 2009; Wolf, 2010a,b). Indeveloped countries several triage protocols and scoringsystems are available to support patient quality and safety(Brabrand et al., 2010; Forsgren et al., 2009; Goranssonet al., 2008; Odell et al., 2009). Despite different findingson the validity and reliability of triage protocols and scoringsystems, the literature supports the use of standardisedmethods to identify patients at risk for developing criticalillnesses (Considine and McGillivray, 2010; Odell et al.,2009; Twomey et al., 2007).

A national review of Norwegian EDs conducted by theNorwegian Board of Health Supervision (NBHS) in 2007 re-ported a lack of established guidelines in several EDs to se-cure the reception and priority of the patients as theyarrive. Based on these findings, the NBHS concluded that astandardised method for triage to support patient safetyand quality care was recommended in all EDs (NorwegianBoard of Health Supervision, 2008). This intervention studywas based on the implementation of a triage protocol andscoring system called the Medical Emergency Triage andTreatment System Adult (METTS-A) in an ED in a regionalhospital in Southern Norway.

Literature

Validating triage protocols and scoring systems has been animportant goal for research (Brabrand et al., 2010; Twomeyet al., 2007; van Veen and Moll, 2009), even though uncer-tainty exists regarding the assessment strategies that regis-tered nurses (RNs) use to assign acuity in EDs (Wolf,2010a,b). RNs use a wide range of thinking strategies indetecting deterioration (Odell et al., 2009) or preformingtriage (Goransson et al., 2008). Studies have shown thatnursing strategies and decision-making in triage settingsare complex and not only a result of an assessment basedon professional, well-considered foundations (Edwards andSines, 2008; Wolf, 2010a,b) or data relevant for the acuityassignment (Jessica, 2011). Brannon and Carson (2003) sug-gest that contextual information can incur biases in a waythat might lead nurses not to properly explore the physicalconditions of their patients. In diagnostic decision-makingprocesses nurses tend to dismiss the physical symptoms ofthe patient in favour of less serious conditions that maybe present. Some studies report that vital parameters areignored or disregarded as the basis for decision making innursing practice (Cooper et al., 2002; Day and Oldroyd,

2010; Odell et al., 2009). According to other studies con-ducted in EDs, the inclusion of vital parameters in the triageassessment allows outcomes to be more predictive (Cooperet al., 2002; Domagala, 2009; Sklar et al., 2007). Measuringvital signs contributes to low inter-rater variability amongRNs, supporting validity by predicting mortality and identi-fying patients that are either at high risk or low risk of dete-rioration during their stay in the EDs (Thompson et al., 2009;Widgren and Jourak, 2011).

Other studies have underscored the pitfalls that affectthe assessment process in a way that can lead to distortionsin the RN decision-making process (Edwards and Sines, 2008;Wolf, 2010a,b). Wolf (2010a,b), for instance, found that theprocess of acuity assignment in the ED is influenced by irrel-evant factors, such as the interplay of elements among theindividual nurses, the immediate unit environment, and thegeneral care environment. From their findings, Edwards andSines (2008) conclude that nurses tend to perform triageprocesses that include judging clinical data based on theway that the patient behaves. They found that nurses in tri-age settings determined credibility by the way patients de-picted the problem: outward clinical signs were not viewedas a neutral manifestation on pathology but as a representa-tion of the physical discomfort of the patient. Some studieshave noted the potential impact health care providers’emotions may have on decision making in clinical practice(Croskerry et al., 2010). Croskerry et al. (2010) examinedthe literature across multiple disciplines to review the inter-relationships between emotion, decision making, andbehaviour to assess their potential impact on patient safety.The authors found that the emotional state of the healthcare provider leads to an affective bias in decision making.This may be influenced by many factors, including the char-acteristics of the patient, ambient conditions in the healthcare setting, (diurnal, circadian, infradian, and seasonalvariables), and the endogenous disorders of the individualprovider (Croskerry et al., 2010). Hence, emotions may alsoinfluence the assessment process in a way that can impacttriage assignment. Croskerry et al. (2010) emphasise theneed for strategies to prevent emotional influences thatmay impact care.

Some literature and studies have focused on the workingconditions of RNs in EDs as factors that may have a negativeimpact on validity and reliability in RNs’ decision making intriage settings (Adriaenssens et al., 2011; Forsgren et al.,2009; Gail and Nora, 2007; Goransson et al., 2008). Adria-enssens et al. (2011) found that in most EDs nurses work un-der adverse conditions compared with nurses working inother settings (Adriaenssens et al., 2011). The perceptionof RNs regarding overcrowded EDs and the constant flowof patients (Gail and Nora, 2007), time pressure, the lackof adequate work procedures, and the complexity of the tri-age role (Goransson et al., 2008) may cause psychosomaticdistress and a lack of job satisfaction among RNs (Forsgrenet al., 2009). Adriaenssens et al. (2011) found that usingguidelines and adequate work procedures have a positive ef-fect on job satisfaction, work engagement, and turnover forRNs, in addition to contributing to clarity and reducedstress. Working within protocols and clinical guidelines areviewed as important for providing trust and assuranceamong nurses working in EDs (Melby et al., 2011).

Triage assessment of registered nurses in the emergency department 91

Aim

The study aimed to evaluate decision making in the triagesetting before and after implementation of the METTS-A inan ED at a regional hospital.

Methods

Study design

This study had a descriptive intervention design with a quan-titative approach. Data were collected before and after theimplementation of METTS-A in an ED. See Fig. 1 for designand timeline.

Setting and sample

This study was conducted in an ED of a regional hospital inSouthern Norway between April 2008 and November 2009.This hospital receives approximately 20,000 patients peryear with acute need of medical care. Patients with mentalhealth problems are in addition to these numbers and areassessed in a separate ED. The half of these patients arrivesdirectly to the injury policlinic or the medical policlinic. Therest 9700 are assessed in the ED and admitted to hospitali-sation and those are the population of this study. The pa-tients who arrive at the ED are referred from the hospitalspoliclinics, a general practitioner, by emergency medicalservice, or the emergency medical service communicationsystem. The ED is covered by six full-time employees, fiveRNs and one licensed practical nurse (LPN) and 48 part-timeemployees, 44 RNs and four LPNs. Eight of the RNs werespecialised as critical care RNs or RNs specialised in anaes-thesia for work in the operating theatre. None of the nurseswas specialised as emergency RNs, but all new RNs under-went a training programme before working in the ED. Therewere no turnovers in the nursing staff during the time of thestudy. The participants in this study were RNs who receivedpatients as they arrive in the ED. During each shift, one RNon duty was responsible for receiving and assessing patientswho arrive in the ED. The exception was incoming patientswith a known critical condition and where an accepting phy-sician receives the patient upon arrival to the ED. Before theintervention, there was no prioritisation system used to sup-port the RNs’ decision-making process.

Ethical considerations

The aim of the study and the data collection methods werepresented to the head of the Regional Committee for

2008 2008 2

Before intervention 2008

Participants: n=655

Information and educational sessions

Implementationemergency pati

Timeline April - May May - December July - D

Fig. 1 Overview of the study design

Medical Research Ethics (RCMRE) in Southern Norway in2007. However, the study was deemed by the head of theRCMRE, not necessary to be presented to the RCMRE. TheRNs were given written and oral information about the nat-ure of the study and were encouraged to participate. Afterthe intervention, the basis for decision making was ab-stracted from the ED’s patient records. The patients wereinformed verbally and in writing about the study, includingthat participation was voluntary and that they could with-draw without consequences on the care they received.The patients gave their written informed consent to usethe ED patient record in the study.

Intervention

The METTS-A triage protocol involves assigning a colour-coded category of priority based on vital parameters andsymptoms and signs. METTS-A, a protocol for intra-hospitaltriage of adult patients (Widgren et al., 2008), was devel-oped in 2004 at Sahlgrenska University Hospital. This instru-ment provides the basis for sorting, prioritising andidentifying the risk of poor outcome of all adult patientsin EDs (Widgren and Jourak, 2011). The METTS-A triage pro-tocol has demonstrated high inter-rater reliability and is asensitive tool for identifying those in need of immediatemedical attention and for the early detection of thosewho deteriorate during their ED stay. The tool is also sensi-tive in predicting mortality in the ED as well as throughoutthe hospitalisation period (Widgren and Jourak, 2011; Wid-gren et al., 2008).

Before implementation of the intervention, some adjust-ments were made to the original METTS-A. The originalMETTS-A protocol was a five-level scale that included thelevels red, orange, yellow, green, and blue. The lowest le-vel (blue) was given to patients who did not need emergencycare or hospital facilitation (Widgren and Jourak, 2011). Theblue level was excluded as a priority level and the Glasgowcoma scale (GCS) (Teasdale and Jennett, 1974) was used forthe assessment of consciousness, replacing the Reaction Le-vel Scale (RLS85). These adjustments should not affect thevalidity because the blue level implied no constraint obser-vations or measurements. The rationale for these adjust-ments was that the blue priority included patients who didnot need emergency care or hospital facilitation. This cate-gory of patients would not normally come to the ED butwould instead be treated by the policlinic at the hospital re-ferred from a general practitioner or treated in the primaryhealthcare system in Norway. The GCS replaced the RLS85because the GCS was the standard used for assessing con-sciousness in the study hospital. The authors of METTS-Awere informed about the adjustments.

008 2009 2009

Implementation of the triage system METTS-A, establish use and follow up

After intenvention 2009

Participants: n= 413

of an ent record form

ecember January - April October

and timeline for the intervention.

Method Red Orange Yellow Green

A: Airway Inspection Airway compromised

Stridor

Not used Not used Not used

B: Respiration RR/min

POX %

>30 or <8

SpO2 < 90 with

with O2

inhalation therapy

>25

SpO2 < 90

without O2

inhalation therapy

< 25

SpO2 90-95

without O2

inhalation therapy

9- 25

SpO2 > 95

without O2

inhalation therapy

C: Circulation HR

BP

HR > 130, HR ir

> 150

SBP < 90 mmHg

HR> 120 or < 40 HR > 110 or < 50 HR 51-109

D: Disability GCS Unconscious

(GCS 3-9) or on-

going seizure

Somnolent

(GCS: 10-12)

Acute confused

(GCS: 13-14)

Alert (GCS 15)

E: Exposure Body temp Not used > 41° or < 35° > 38.5° 35° - 38.5°

RR=respiratory rate, POX=pulse oximetry, O 2= medical oxygen, HR= heart rate, HR ir=irregular heart rate, BP=blood pressure, SBP=systolic blood pressure, GCS= Glasgow coma scale

Fig. 2 The METTS-A vital parameters limits referring to each priority (Widgren and Jourak, 2011), including the adjustments forimplementation in the ED.

ESS algorithm no. 5 METTS-A

Chest pain

ST elevation in ECG in ambulance or in ED (Red ESS)

Pathological ECG and chest pain -or history of chest pain during the last 24 h combined with vegetative (autonomous) symptoms -or chest pain + dyspnea -or symptoms of unstable angina (Orange ESS)

Moderate chest pain with no signs of unstable angina but with one or more risk factors of cardiovascular disease (Yellow ESS)

None of the above (Green ESS)

Fig. 3 Example of an emergency symptoms and signs (ESS)algorithm (Widgren and Jourak, 2011).

92 T.K. Vatnøy et al.

The triage method of METTS-A includes two steps as-sessed simultaneously: step one by algorithm for vitalparameters shown in Fig. 2 and step two by one of 96 algo-rithms (emergency symptoms and signs, ESS). An example ofone of the ESS algorithms is shown in Fig. 3.

ESS outcomes are determined on the basis of the pa-tients’ chief complaints, symptoms, and signs, as assessedaccording to the METTS-A protocol. The combined outcomeof both these algorithms is given the final priority level. TheESS gives a higher priority in cases of deterioration when thephysiological vital parameters are within normal limits. Pri-ority, according to the METTS-A, determines the time thatpatients could wait for a medical examination by a physicianand the level of monitoring received. The priority levels andinterventions for triage priority are shown in Fig. 4 (Widgrenand Jourak, 2011).

During the nurse-patient encounter, the RNs measuredphysiological vital signs and decided the categories regard-ing the METTS-A vital parameter limits. While in dialoguewith the patient the RNs made observations to determinethe ESS algorithm that best described the problem. Fig. 5depicts the assessment process according to the METTS-Aprotocol.

The ED patient records and a triage protocol adapted tothe formal triage system were used to support the triageprocess and for documentation. The physician was informedabout the patients’ triage priority and prioritised examina-tion according to the patients’ priority level (Fig. 4). The

first medical examination and treatment of the patient inthe ED were administered by a junior physician who had

Patient arrives to ED

Admittance to hospital by a general practitioner, the emergency medical service, or the acute medical emergency communication central.

Triage area

The RN follows the METTS-A protocol by step 1, and step 2, assessed simultaneously.

Step 1

The RNs measuring all vital parameter according to A, B, C, D and E (figure 2).The RNs assesses acuity level. i.e.: all parameters within normal limits hens’ acuity level: GREEN.

Step 2

The RN assesses which ESS algorithms to follow on basis of the patient’s chief complain, symptoms and signs; i.e. patient complaint are chest pain (figure 3). I addition the patient has a history of chest pain during the last 24 h combined with vegetative (autonomous) symptoms like pail an unwell. This gives the acuity level: ORANGE.

RNs assess priority level

Highest level of step 1 and step 2 gives the finale priority level. i.e: This patient’s final acuity level and priority is ORANGE.

Fig. 5 Triage decision-making process in accordance to the METTS-A protocol including an example.

Red = Life-threatening condition that requires full monitoring (telemetry if necessary, ischemia supervision, BP,

SpO2, respiratory rate, consciousness), and the nurses remain bedside until the doctor responsible for treating the

patient has down-graded the patient to orange or lower. Time to Doctor (TTD) = 0 minutes

Orange = Condition that require full monitoring (same as red), but with intervals of 20 minutes, and supervision by

the patient charge nurse until the TTD. Time to Doctor (TTD) = <20 minutes

Yellow = Standard process controls and further controls when necessary, selective monitoring and control of the

parameters with a marked impact on triage, upgrading TTD. Time to Doctor (TTD) = <120 minutes

Green = No monitoring, but regular supervision by the RN. Time to Doctor (TTD) = <240 minutes

Fig. 4 Interventions regarding triage priority.

Triage assessment of registered nurses in the emergency department 93

the opportunity to consult with a senior physician if unsureof the situation.

The METTS-A was implemented after informing andeducating the employees about the ED. The information

included a review of the METTS-A, how the implementationwould change the workflow, and impact tasks and responsi-bilities. A 4-h seminar consisting of lectures and casework insmaller groups followed by a plenary discussion was offered

94 T.K. Vatnøy et al.

to the RNs who would be performing the triage according tothe METTS-A. These RNs were also offered 2 days in theclinic to perform triages while being closely followed andguided by an experienced user of the METTS-A. The RNs fre-quently received feedback regarding their performance ofthe triage process based on documentation in the ED patientrecords to improve the assessments preformed with theMETTS-A. Follow-up and feedback were provided by one ofthe authors (TKV).

Data collection and procedures

A questionnaire was used to collect data before the inter-vention. The RNs select whether they assessed the medicalcondition of the patient as ‘urgent’ or ‘not urgent’ based onthe patient’s need for a medical check and treatment. Fur-thermore, the RNs should provide an explanation of theirreasons for these assessments.

Data collection after the intervention was performed9 months after the implementation of the METTS-A. Datawere collected from ED patient records to document suchvital parameters as respiratory rate, oxygen saturation,heart rate, blood pressure, and disability score accordingto the Glasgow coma scale (Teasdale and Jennett, 1974)(Fig. 2). The ESS algorithm number and urgency are givenin Fig. 3.

Results

The data collected before the intervention were analysedusing descriptive and comparative statistics. Frequency ta-bles and cross tables were used to present the distribution

Table 3 Registered nurse basis for decision-making in acuity assigimplementation of the METTS-A, N = 655.

Basis for decision-making in acuity assignment Used emergesymptoms an

Vital parameters assessed 6 (1%)Vital parameters not assessed 261 (40%)

Table 1 The registered nurses basis for decision making ininitial patient encounter before intervention N = 655.

Vital parameters Emergency symptomsand signs

Used 24 (4%) 267 (41%)Not used 631 (96%) 198 (30%)

Table 2 The registered nurses basis for decision-making ininitial patient encounter 2009, N = 413.

Vital parameters Emergency symptomsand signs

Used 361 (87%) 318 (77%)Not used 52 (13%) 95 (23%)

of the data. A vertical bar chart presents the variance.The software Statistical Package for the Social Sciencesv.15.0 for Windows was used for data management andanalyses.

Data collection was conducted on 655 adult patients whoarrived at the ED over a period of 6 weeks before the inter-vention (table 1) and 413 adult patients over a period of4 weeks after the intervention (Table 2). Before the inter-vention, 4% of the patients were assessed on the basis of vi-tal parameters. In 96% of the triage assessments, vitalparameters were not mentioned (Table 1).

Before the intervention, 41% of the patients were as-sessed based on symptoms and signs, whereas vital parame-ters were rarely used when no standardised system fortriage was available. Medical diagnoses according to theInternational Classification of Diseases (ICD 10) were usedas a basis of assessment in 29% of the patients (Table 3).After the intervention, the use of vital parameters for deci-sion making increased from 4% to 87% (Tables 1 and 2). Theuse of vital parameters and symptoms and signs as a basisfor decision making increased from 1% to 69% (Tables 3and 4). Medical diagnoses were not used as basis for decisionmaking and acuity assignment after the implementation ofthe METTS-A (Tables 3 and 4). Table 2 shows that afterthe intervention 87% of the patients were assessed basedon vital parameters according to the METTS-A and 77% wereassessed based on emergency symptoms and signs (ESS)according to the METTS-A protocol. Vital parameters werenot completely documented in 13% of the assessments(Table 4).

A difference in acuity assignation before and after inter-vention is illustrated in Fig. 6. After the intervention, therewas a decrease in the number of patients assessed as ur-gent. The mean time of initial assessment by a doctor wasless than 20 min.

Discussion

Implementing a standardised triage protocol and scoringsystem, including validated ESS algorithms and vital param-eters, increased the use of vital parameters and symptomsand signs for decision making and acuity assignment. Our re-sults from the intervention show that, to a small extent, theRNs’ decision making and acuity assignment were based onthe patients’ vital parameters. This result corresponds tothose from other studies that describe vital parameters tobe ignored or disregarded as the basis for decision makingin nursing practice (Cooper et al., 2002; Day and Oldroyd,2010; Odell et al., 2009). Not using vital parameters as basisfor decision making may occur in low inter-rater variabilityamong RNs in the triage decision-making process (Widgrenand Jourak, 2011) and may have a negative effect on patient

nment in the initial patient nurse encounter before (2008) the

ncyd signs

Did not use the emergencysymptoms and signs

Used medicaldiagnoses

13 (2%) 5 (1%)185 (28%) 185 (28%)

Table 4 Registered nurse basis for decision-making in acuity assignment in the initial patient nurse encounter after (2009) theimplementation of the METTS-A, N = 413.

Basis for decision-making in acuity assignment Used emergencysymptoms and signs

Did not use emergencysymptoms and signs

Vital parameters assessed 286 (69%) 75 (18%)Vital parameters not assessed 32 (8%) 20 (5%)

Fig. 6 Patients assessed to be urgent or not urgent, as statedby the RNs, before and after the implementation.

Triage assessment of registered nurses in the emergency department 95

safety in the ED (Cooper et al., 2002). An unexpected findingwas that the RNs used medical diagnoses in their decision-making process before the intervention. When using medi-cal diagnosis as a foundation in triage decisions, the diagno-sis is most likely tentative because it is not necessarilyverified by a physician but might be determined instead bycontextual information and conditions. Contextual informa-tion can be irrelevant in decision making regarding theintention of the triage process in which the aim is to deter-mine and classify the priority of patients according to theirneed of urgent treatment in a resource-limited setting(Brannon and Carson, 2003).

Without a standardised protocol, the nurses’ assessmentincluded symptoms and signs that were observed and inter-preted based on the individual RN’s competence (or lackthereof). Studies concluding that the process of acuityassignment were influenced by the nurses knowledge base,critical cue recognition, and social context (Wolf,2010a,b) where depended on intuition (Odell et al., 2009)and interpreting patient behaviour (Edwards and Sines,2008). The studies suggest that decision making under theseconditions might be performed on a misleading basis. Part-time working nurses and nurses without specialisation, as inour study, are targets for competence improvement. Fur-ther assessments during the initial nurse-patient encounterfor the purpose of acuity assignment in the ED imply thecapacity to assess complicated medical conditions thatmay exceed ordinary nursing competence. Such compli-cated assessments may require special qualifications, edu-cation, and skills (Gilboy et al., 1999). The METTS-A ESSprotocol represents support in RN assessment and may im-prove the RNs’ competence in acuity assignment.

In our study, the RNs adopted the METTS-A such thatchanged the basis of their decision-making process.Whether the implementation of a standardised triage sys-tem improves patient safety in the ED requires additionalresearch. Further studies should focus on whether motiva-tion influences changes in practice.

Limitations

This sample included all patients >15 years of age who vis-ited the ED. Different data collection tools were used be-fore and after the intervention. When analysing the datacollected before the intervention, any mention by a nurseof a vital parameter was considered ample information toregister ‘use of vital parameters’ (e.g., an explanation oftheir reasons for these assessments in the free text spacewas: ’blood pressure high’; the registration was ‘use of vitalparameters’). Thus, the analysis of data before the inter-vention can only interpret whether RNs regarded the vitalparameters as important parameters in their decision-mak-ing process. The data source for the assessment after theintervention was the ED’s patient records. All of the vitalparameters recommended by the METTS-A had to be docu-mented if the registration reported ‘use of vital parame-ters’ in the data analyses (e.g., if the respiration rate wasnot documented, the registration reported ‘vital parame-ters not used’). Adjustments were performed to the originalMETTS-A. There is a risk that the adjustments preformed tothe original METTS-A has affected the validity of the proto-col. This risk is considered to be low because the excludedblue level did not imply any constraint observation ormeasurements.

Conclusions

Because patients have unknown and potentially very seriousillnesses, there is a high degree of uncertainty and acuitythat complicate the triage decision process. Performing tri-age under these difficult conditions is a demanding processand a challenge for the RNs working in an emergency med-ical service unit. Implementing a standardised triage sys-tem, including vital parameters and standardisedalgorithms for signs and symptoms, was found to increasethe use of vital parameters and symptoms and signs for deci-sion making and acuity assignment.

Acknowledgments

The authors wish to thank all of the RNs and patients whoparticipated in this study. We also express our gratitude

96 T.K. Vatnøy et al.

to all of the ED staff members who contributed to data col-lection and who photocopied records. We are grateful forthe implementation project members and the managementfor their support. Finally, we are thankful for the financialsupport of the hospital.

References

Adriaenssens, J., De Gucht, V., Van Der Doef, M., Maes, S., 2011.Exploring the burden of emergency care: predictors of stress-health outcomes in emergency nurses. J. Adv. Nurs. 67, 1317–1328.

Brabrand, M., Folkestad, L., Clausen, N., Knudsen, T., Hallas, J.,2010. Risk scoring systems for adults admitted to the emergencydepartment: a systematic review. Scand. J. Trauma Resusc.Emerg. Med. 18, 8.

Brannon, L.A., Carson, K.L., 2003. The representativeness heuris-tic: influence on nurses’ decision making. Appl. Nurs. Res. 16,201–204.

Considine, J., McGillivray, B., 2010. An evidence-based practiceapproach to improving nursing care of acute stroke in anAustralian Emergency Department. J. Clin. Nurs. 19, 138–144.

Cooper, R.J., Schriger, D.L., Flaherty, H.L., Lin, E.J., Hubbell,K.A., 2002. Effect of vital signs on triage decisions. Ann. Emerg.Med. 39, 223–232.

Croskerry, P., Abbass, A., Wu, A.W., 2010. Emotional influences inpatient safety. J. Patient Saf. 6, 199–205.

Day, A., Oldroyd, C., 2010. The use of early warning scores in theemergency department. J. Emerg. Nurs. 36, 154–155.

Domagala, S.E., 2009. Discharge vital signs: an enhancement to EDquality and patient outcomes. J. Emerg. Nurs. 35, 138–140.

Edwards, B., Sines, D., 2008. Passing the audition – the appraisal ofclient credibility and assessment by nurses at triage. J. Clin.Nurs. 17, 2444–2451.

Forsgren, S., Forsman, B., Carlstrom, E.D., 2009. Working withManchester triage – job satisfaction in nursing. Int. Emerg. Nurs.17, 226–232.

Gail, C., Nora, C., 2007. Safety in the emergency department: it’sabout time. Kansas Nurse 82, 3.

Gilboy, N., Travers, D., Wuerz, R., 1999. Re-evaluating triage in thenew millennium: a comprehensive look at the need for stan-dardization and quality. J. Emerg. Nurs. 25, 468–473.

Goransson, K.E., Ehnfors, M., Fonteyn, M.E., Ehrenberg, A., 2008.Thinking strategies used by registered nurses during emergencydepartment triage. J. Adv. Nurs. 61, 163–172.

Hale, N., Tippett, J., 2009. Safeguarding vulnerable adults in theemergency department. Int. Emerg. Nurs. 17, 141–142.

Huryk, L.B., 2006. Is it really an emergency? When nonurgent casesoverwhelm the ED. Am. J. Nurs. 106 (1), 69–72.

Jessica, C., 2011. Emergency department triage: what data arenurses collecting? J. Emerg. Nurs. 37, 417–422.

Melby, V., Gillespie, M., Martin, S., 2011. Emergency nursepractitioners: the views of patients and hospital staff at a majoracute trust in the UK. J. Clin. Nurs. 20, 236–246.

Norwegian Board of Health Supervision, 2008. ‘‘While we arewaiting. . .’’. – do patients receive adequate treatment inaccident and emergency units? Report from the NorwegianBoard of Health, Supervision 2/2008.

Odell, M., Victor, C., Oliver, D., 2009. Nurses’ role in detectingdeterioration in ward patients: systematic literature review. J.Adv. Nurs. 65, 1992–2006.

Sklar, D.P., Crandall, C.S., Loeliger, E., Edmunds, K., Paul, I.,Helitzer, D.L., 2007. Unanticipated death after discharge homefrom the emergency department. Ann. Emerg. Med. 49, 735–745.

Teasdale, G., Jennett, B., 1974. Assessment of coma and impairedconsciosness: a practical scale. Lancet 304, 81–84.

Thompson, M., Coad, N., Harnden, A., Mayon-White, R., Perera, R.,Mant, D., 2009. How well do vital signs identify children withserious infections in paediatric emergency care? Arch. Dis. Child.94, 888–893.

Twomey, M., Wallis, L.A., Myers, J.E., 2007. Limitations invalidating emergency department triage scales. Emerg. Med.J. 24, 477–479.

van Veen, M., Moll, H., 2009. Reliability and validity of triagesystems in paediatric emergency care. Scand. J. Trauma Resusc.Emerg. Med. 17, 38.

Widgren, B.R., Jourak, M., 2011. Medical emergency triage andtreatment system (METTS): a new protocol in primary triage andsecondary priority decision in emergency medicine. J. Emerg.Med. 40, 623–628.

Widgren, B.R., Jourak, M., Ann, M., 2008. Ny traffsaker triageme-tod. METTS-A ger underlag for prioritering till ratt vardniva.Lakartidningen 105, 201–204 (in Swedish).

Wolf, L., 2010a. Acuity assignation: an ethnographic exploration ofclinical decision making by emergency nurses at initial patientpresentation. Adv. Emerg. Nurs. J. 32, 234–246.

Wolf, L., 2010b. Does your staff really ‘‘get’’ initial patientassessment? Assessing competency in triage using simulatedpatient encounters. J. Emerg. Nurs. 36, 370–374.