Access provided by
London Metropolitan University
• To be aware of recent studies on factors that influence emergency department (ED) nurses’ triage decisions
• To explore why there may be significant differences between triage algorithms’ results and nurses’ decisions
• To better understand what may negatively affect nurses’ ability to accurately triage patients in ED
Accurate triaging of patients in emergency departments (EDs) is crucial, since triage determines how quickly patients are assessed and treated. Understanding the factors that influence ED nurses’ triage decisions is important to ensure that patients are prioritised appropriately and cared for in a timely manner. This article reports and discusses the findings of a literature review on the factors that affect nurses’ triage decisions in the ED. Triage decisions by nurses in EDs are influenced by several factors relating to the patient, the nurse, the triage algorithm and the environment where triage takes place. Nurses’ ability to triage patients accurately is negatively affected by high patient numbers, inadequate staffing levels, lack of privacy and lack of training.
Emergency Nurse. doi: 10.7748/en.2022.e2123
Peer reviewThis article has been subject to external double-blind peer review and checked for plagiarism using automated software
Correspondence Conflict of interestNone declared
Gorick H (2022) Factors that affect nurses’ triage decisions in the emergency department: a literature review. Emergency Nurse. doi: 10.7748/en.2022.e2123
Published online: 28 February 2022
In 2018-19 there were almost 25 million emergency department (ED) attendances in England, which equates to nearly 68,000 per day (NHS Digital 2019). Patients attending EDs are triaged by nurses into acuity categories so that they can be seen in the most appropriate time frame. Triage uses algorithms, with patients being assigned to a category that represents their acuity level usually based on their symptoms and vital signs (Christ et al 2010, Mackway-Jones et al 2014). These categories indicate how quickly the patient needs to be seen – usually immediately for patients in the highest acuity category and within four hours for patients in the lowest acuity category (Mackway-Jones et al 2014). The triage algorithms most commonly used globally are the Manchester Triage System, the Emergency Severity Index and the Canadian Triage and Acuity Scale (Zachariasse et al 2019).
Correct triage is crucial, since triage affects how rapidly patients receive care and incorrect triage can result in potentially fatal delays (Christ et al 2010). In addition, incorrect triage has ramifications for the entire ED, affecting patient flow and potentially creating obstructions that could have significant negative effects on all patients’ outcomes (Christ et al 2010, Lentz et al 2017). Therefore, understanding the factors that affect the accuracy of nurses’ triage decisions enhances not only the identification of high-acuity patients, but also the performance of the ED as a whole.
This literature review aimed to examine the factors that affect nurses’ triage decisions in the ED.
• Accurate triage of patients presenting to the emergency department (ED) is crucial, since triage affects how rapidly they receive care
• When assigning an acuity category, ED nurses consider patient factors such as presenting complaint, vital signs, visual presentation and verbal history
• Experienced nurses tend to use their intuition in triage decisions, but caution is needed because of the risk of complacency
• Formal triage training is often lacking, resulting in a suboptimal understanding of triage and inaccurate decisions
• High patient numbers, inadequate staffing levels and lack of privacy negatively affect ED nurses’ ability to triage accurately
The databases MEDLINE, the Cumulative Index to Nursing and Allied Health Literature (CINAHL), PubMed, Wiley, Springer, SAGE Journals and Taylor & Francis were searched.
To be included, articles had to be peer reviewed, be published in English between 2015 and 2020, and relate to face-to-face triage decisions made by nurses in EDs. Articles from outside the UK were included if their findings were focused on general triage practices rather than on specific local practices. These articles were examined to ensure that their findings were relevant to triage in the UK and that their transferability and generalisability were sufficient.
Articles were excluded if they related to telephone triage, since telephone triage uses different protocols than face-to-face triage (McKinstry et al 2010) and, while there is overlap between the two methods, telephone triage was not the topic of the review. Articles on triage by doctors were excluded, since the focus of the review was on nurses. One article on doctor-led triage was included after closer examination because its conclusions focused on general triage practices rather than doctor-specific triage practices.
A total of 32 articles were included in the review. Four of them were literature reviews and the remaining ones were primary research articles. The 32 articles are summarised in an online-only appendix available at http://rcni.com/ED-triage-decision
Themes were identified using thematic analysis (Braun and Clarke 2022), reviewing articles and noting themes as they emerged, with recurrent examples of themes grouped together to establish frequency.
Four themes and 12 sub-themes were identified (Table 1), which are described and discussed below.
Triage algorithms had strong validation, with decisions that disregarded algorithms’ recommendations presented as inaccurate (Chang et al 2016, Tam et al 2018, Moon et al 2019, Lee et al 2020). However, some of the studies found that algorithm validation was often performed using patient scenarios instead of analysing actual patient data. That is, the algorithms did not always reflect real life, where patients’ circumstances such as past medical history, presentation and other factors may be significantly different from the ideal scenarios used by algorithms (Hinson et al 2018, Mistry et al 2018, Iversen et al 2019, Park et al 2019). This was reinforced by Zachariasse et al (2016), who reported that triage algorithms often only used connections between triage level and outcomes, rather than a full examination of all available factors, to determine validity. These findings suggest that while studies may describe triage decisions as accurate or inaccurate, these decisions do not always reflect the patient’s actual acuity but merely how well the patient’s presentation matches the algorithm used.
Varying opinions regarding the efficacy of triage algorithms were found, some nurses feeling that algorithms are reliable and others finding inaccuracies (Chang et al 2016, Hinson et al 2018, Tam et al 2018). Experienced nurses often used triage algorithms as a guide to aid decision-making and disregarded them at times in favour of their experience and intuition (Johannessen 2016, Mistry et al 2018). Less experienced nurses followed algorithms rigidly (Chang et al 2016, Johannessen 2016).
Some studies found that the algorithm itself was sometimes manipulated to change triage categories, nurses knowing that selecting a certain pathway would produce the category they wanted (Clarke et al 2015, Johannessen 2017, Park et al 2019). Other studies found that nurses went along with the category produced by the algorithm, despite their clinical knowledge and experience producing a different result, if they perceived the algorithm’s result to be more beneficial to the patient (Johannessen 2016, Adams et al 2017).
When nurses were uncertain, they often placed patients in an intermediate acuity category, which resulted in overpopulated intermediate acuity categories (Mistry et al 2018, Saban et al 2019). Some nurses felt that if they decided to disregard the algorithm’s result, they should always place the patient in a higher category and never ‘downgrade’ them because of the inherent risk (Ekins and Morphet 2015, Goldstein et al 2017, Johannessen 2017).
Several studies found that the patient’s presenting complaint was the main factor for assigning them to a triage category, especially when triage was undertaken by inexperienced nurses who tended to follow algorithms more strictly (Stanfield 2015, Moon et al 2019, Park et al 2019). Other studies found that the more severe the presenting complaint, the more attention nurses paid to secondary factors such as co-morbidities and underlying symptoms (Johannessen 2016, Adams et al 2017, Lee et al 2020). This resulted in already high acuity scores being further increased to take account of underlying symptoms that may or may not actually affect patient acuity. Furthermore, this suggests that nurses may tend to pay more attention, during triage, to patients who clearly present with higher acuity, therefore resulting in more accurate triage decisions for these patients. Symptom duration had an influence on decision-making, with longer duration seen as indicating lower urgency (Johannessen 2016).
Certain cues resulted in an instant change of triage category, these cues usually being red flags indicating a risk of rapid deterioration (Stanfield 2015, Roscoe et al 2016, Moon et al 2019). Red flags tended to be identified and used to change triage category more often by nurses with greater triage experience (Saban et al 2019), possibly because of their greater awareness of these red flags and greater ability to recognise them. When the patient’s presentation did not match the expected symptoms for the presenting complaint, nurses tended to increase the acuity score, possibly because they felt the patient may require further urgent investigation (Adams et al 2017).
In some studies, vital signs appeared to be an important component in triage decisions because they show whether and to what extent the patient’s vital signs differ from normal values (Chang et al 2016, Roscoe et al 2016, Bowen et al 2017). Nurses found it challenging to triage patients satisfactorily when they were unable to review their vital signs, and often changed their decision when they received vital signs (Hinson et al 2018, Wolf et al 2018). This was further explored by Adams et al (2017), in whose study junior emergency medicine doctors considered abnormal vital signs as directing them towards higher acuity categories.
However, Stanfield (2015) found that, in most cases, vital signs did not significantly affect triage decisions. Other studies found that vital signs were examined as complementary information to the patient’s presenting complaint, with the presenting complaint itself being what the triage decision was based on (Lukin et al 2015, Petruniak et al 2018, Lee et al 2020). Over-reliance on vital signs alone was associated with under-triage (under-triage in the sense of ‘being placed in an acuity category that is too low’) because other significant signs and symptoms were not given appropriate consideration (Cannavacciuolo et al 2018, Hinson et al 2018).
Nurses often initially used the common ‘quick-look triage’ technique, whereby they use their intuition to establish whether they think the patient is critically ill (Chang et al 2016, Adams et al 2017, Wolf et al 2018). Iversen et al (2019) found that the results of quick-look triage differed significantly from the results of triage algorithms and that they were more accurate in predicting short-term and long-term mortality. However, no other study explored this and further research in this area would be beneficial.
Patients’ behaviours were a significant factor in decision-making in Stanfield (2015), Johannessen (2016), Roscoe et al (2016), Adams et al (2017), Bowen et al (2017) and Yuliandari (2019). In these studies, how patients behaved was considered more important than their vital signs and how they said they felt. Patients who spoke calmly or played with their mobile phones while waiting for triage were more likely to receive lower acuity ratings than those in visible distress, whether or not these lower ratings accurately reflected actual acuity. Other studies reported that nurses felt visual cues could be confounding and could lead them to misjudge acuity, especially in patients presenting with trauma or mental health conditions (Clarke et al 2015, Lukin et al 2015, Goldstein et al 2017).
In a literature review on nurses’ use of discretion when triaging, Johannessen (2016) found that the mode of arrival was given significant weight, walk-in patients receiving lower acuity scores than those who arrived by ambulance, even when the results were adjusted to consider actual acuity.
Experienced nurses often felt that patients’ verbal histories increased their ability to triage accurately because patients provided details that could be missed by algorithms (Roscoe et al 2016, Johannessen 2017, Wolf et al 2018). However, in some cases, verbal histories were perceived as having a negative influence – for example if nurses felt that the patients exaggerated their symptoms, if patients emphasised elements of their condition that nurses considered secondary, or if patients downplayed embarrassing aspects of their symptoms that could be important cues (Johannessen 2016, 2017, Roscoe et al 2016).
Smyth and McCabe (2017) argued that communication issues reduced the effectiveness of verbal histories, for example if patients misunderstood questions or did not communicate important information. Nurses cited communication issues as a reason for placing patients in higher acuity categories to ensure their safety (Mistry et al 2018, Petruniak et al 2018).
Children and older people often had their acuity level increased because nurses were aware of potential challenges in recognising acute illness, and because the risk of deterioration is generally higher, in these patient groups (Ekins and Morphet 2015, Lukin et al 2015, Zachariasse et al 2016, Mistry et al 2018). However, in some of the other studies, patient age had no significant bearing on triage decisions (Johnson and Alhaj-Ali 2017, Petruniak et al 2018, Saban et al 2019).
These conflicting findings could be attributable to differences in definitions of age. Some studies used the age categories of under 50 years and over 50 years, finding that patients aged ≥50 years were more likely to be under-triaged compared with patients aged <50 years. Other studies used the age categories of under 70 years and over 70 years, finding that the likelihood of over-triage (over-triage in the sense of ‘being placed in an acuity category that is too high’) was reduced in patients aged ≥70 years compared with patients aged <70 years (Lukin et al 2015, Hinson et al 2018).
Nurses felt that experience was important to triage patients accurately, with less experienced nurses lacking the required familiarity with patient presentations and triage systems (Stanfield 2015, Chang et al 2016, Akta and Alemdar 2017, Bowen et al 2017, Wolf et al 2018, Reay et al 2020). Experienced nurses were more willing to make acuity category changes, potentially due to their greater knowledge, wider consideration of other factors and refined intuition (Bowen et al 2017, Johannessen 2017, Johnson and Alhaj-Ali 2017, Yuliandari 2019). Some studies found that junior nurses were more likely to over-triage patients since they feared missing a sick patient but also under-triaged patients because they lacked the knowledge to identify acute presentations unaccounted for in algorithms (Cannavacciuolo et al 2018, Mistry et al 2018).
Experience had a positive effect on the ability to assess acuity accurately (Stanfield 2015, Yuliandari 2019). However, experience did not necessarily result in an improved ability to triage patients, some nurses’ ability to make clinical decisions remaining at novice level despite significant time working in triage (Reisi et al 2018, Wolf et al 2018). Experienced nurses tended to take longer to make triage decisions because of a more in-depth exploration of patients’ histories and symptoms (van der Linden et al 2016, Johnson and Alhaj-Ali 2017). However, this was reversed when workload increased (Akta and Alemdar 2017, Yuliandari 2019, Reay et al 2020), maybe because experienced nurses have a better understanding of which parts of the assessment are crucial and which can be omitted.
Several studies found that nurses relied on intuition, while noting that this could lead to inaccurate triage decisions (Stanfield 2015, Adams et al 2017, Akta and Alemdar 2017, Johannessen 2017, Wolf et al 2018, Yuliandari 2019). Nurses tended to trust their intuition over algorithms, since they felt their experience enabled them to recognise a patient’s acuity more accurately (Stanfield 2015, Chang et al 2016, Johannessen 2016, Roscoe et al 2016) and since triage algorithms cannot account for all variables nor for contextual information, especially about patients with mental health conditions (Clarke et al 2015, Stanfield 2015, Johannessen 2016, 2017). Some studies found that using intuition too often led to complacency, nurses not examining their decisions closely and missing important details (Johannessen 2017, Smyth and McCabe 2017, Yuliandari 2019).
Training resulted in nurses relying less on intuition and more on deductive reasoning (Smyth and McCabe 2017), although some studies found that nurses’ ability to intuit correctly improved with effective triage training (Stanfield 2015, Akta and Alemdar 2017). Triage education was often found to be insufficient, however, and there were no significant differences observed between nurses who had regular refreshers and nurses who did not (Soontorn et al 2018, Tam et al 2018). This may be because training cannot replicate certain real-life factors that significantly affect decision-making, such as workload and environment (Soontorn et al 2018, Tam et al 2018).
Akta and Alemdar (2017) and Wolf et al (2018) found that many nurses lacked formal triage training, instead learning on the job, which resulted in a suboptimal understanding of triage and more triage inaccuracies. How this applies to the NHS and the quality and quantity of training provided to ED nurses working in the NHS would benefit from further exploration, coupled with assessing how best to improve nurses’ triage training.
Some nurses changed acuity categories if they felt patients would otherwise be sent to an inappropriate clinical area, whether because of pathways which nurses considered unsuitable or because patients needed specific equipment that was not available in the prescribed area (Clarke et al 2015, Chang et al 2016, Wolf et al 2018). Some nurses triaged patients to inappropriate areas such as urgent treatment centres despite high acuity, reasoning that patients would be seen and treated more quickly and that risk was lower than if they did not send them there (Reay et al 2016, van der Linden et al 2016, Mistry et al 2018).
Calm and private environments were considered crucial to correctly establish acuity, since they encourage the discussion of information that patients may feel uncomfortable sharing in an open environment (Stanfield 2015, Chang et al 2016). Johannessen (2016) found that open environments increased the likelihood of interruptions, resulting in disrupted thought processes and challenges returning to the assessment. However, Johnson and Alhaj-Ali (2017) found that interrupted triage often resulted in more accurate decisions, potentially because of reinforced concentration and nurses rethinking decisions.
Studies linked nurses who changed triage categories with high patient flow through the ED and how many patients had been triaged to each category (van der Linden et al 2016, Adams et al 2017, Johannessen 2017, Wolf et al 2018, Reay et al 2020). History taking was impaired by high patient numbers (Roscoe et al 2016, van der Linden et al 2016, Smyth and McCabe 2017, Reay et al 2020), notably because nurses attempted to gain time by reducing complex cases to simple sentences. The risk in such cases was that nurses would miss important details, which suggests that triage nurses must be given sufficient time to make their assessments.
Inadequate numbers of staff or inadequate skill levels strongly affected triage (Johannessen 2016, Wolf et al 2018, Yuliandari 2019), increased workload resulting in nurses being unable to carry out full assessments. Increased workload was also noted as increasing the risk of burnout, because staff would feel overwhelmed by the number of patients they had to see and unable to make decisions safely (Reay et al 2016, 2020, Wolf et al 2018). This led to an unwillingness to explore patients’ presentations in depth and consequently to incorrect acuity scores. This suggests that there is a need for adequate staffing levels in triage to ensure patient and staff safety.
Further research exploring the accuracy of quick-look triage and the factors that affect it would assist in streamlining triage, although care needs to be taken to ensure that assessments are made safely. Ensuring that there are suitable triage environments and adequate staffing levels in triage would benefit nurses and patients. Exploring how to improve teaching and training in triage, especially from an educational perspective, may benefit practice.
Triage decision-making by nurses in EDs is influenced by several factors, those intrinsic to the patient and those relating to the triage algorithm, the nurse and the physical environment. ED nurses use a range of information to assign patients to acuity categories, including algorithm results, visual presentation, vital signs and verbal history. There can be significant differences between the acuity category produced by the algorithm and that selected by the nurse, which may occur because nurses disagree with the algorithm based on their experience or intuition, because they lack adequate training or because they manipulate the algorithm to achieve the outcome they think is best. The triage environment has a strong influence on nurses’ ability to triage accurately, with high numbers of patients, inadequate staffing levels and a lack of privacy all negatively affecting the triage process.
03 September 2024 / Vol 32 issue 5
Emergency assessment bay for cancer patients
The annual Quality in Care (QiC) Excellence in Oncology...
Auditing a chemotherapy triage telephone service
The increased need for a system of assessment, intervention...
Appropriate categorisation of mild pain at triage: a diagnostic study
The Manchester Triage System (MTS) offers emergency...
Early warning scores: are they needed in emergency care?
This article reviews an audit to assess how adequately...
Developing a triage tool for paediatric care
Most paediatric triage systems used in UK emergency...
