Emergency Medicine (2003) 15, 6 10

Blackwell Science, Ltd

Systems Management Series

Canadian perspective on ED triage

The Canadian Triage and Acuity Scale: A Canadian perspective on emergency department triage
Michael J Murray Emergency Services, Royal Victoria Hospital, Barrie and Associate Clinical Professor Department of Family Medicine, McMaster University, Ontario, Canada

Abstract
The Canadian Triage and Acuity Scale has received widespread acceptance in Canada as a reliable and valid tool for emergency department triage. The importance of accurate triage becomes more apparent as emergency department volumes increase, and resources shrink. The need to ensure that those patients requiring more urgent care receive care rst is the basis for all triage scales. Through the Canadian Triage and Acuity Scale National Working Group, the scale became the recommended triage tool for Canadian emergency departments. Work has been done on the interrater reliability of Canadian Triage and Acuity Scale among health care providers. There is a need to further assess the validity of the scale. This scale has now been applied in the out of hospital setting by paramedics and is being used in measurements of emergency physician workload. The future may see an electronic triage tool develop for emergency department use to reduce variability in its application. The Canadian Triage and Acuity Scale has become an integral component of Canadian emergency departments.
Key words: Canadian Triage and Acuity Scale, emergency department, triage, workload.

Introduction
Triage is accepted as an integral part of the emergency department (ED) patient assessment. It has become important to the safe and efcient operation of the ED by allowing for the prioritization of patients based on acuity. The key role of triage is to assign a treatment priority to patients to ensure that those needing urgent care, access care rst. If triage is faulty or incorrect, seriously ill patients may wait too long for care and

this may result in worse outcomes. Those who are triaged as less urgent or non-urgent experience longer delays to care and some will leave without being seen. Therefore accuracy of the triage scale in predicting urgency is essential. With signicant overcrowding in ED the need for triage is even more apparent. There has been some suggestion that triage could be used to effectively identify those patients that may not require emergency care. In order for triage to be effective the triage scale must be reliable, valid and easy to use. It

Correspondence:

Dr Michael J Murray, Emergency Services, Royal Victoria Hospital, 201 Georgian Drive, Barrie, Ontario, Canada L4M 6M2. Email: michaelj.murray@sympatico.ca

Michael J Murray, MD CCFP(EM), Medical Director, Assistant Clinical Professor and Chair, CAEP CTAS National Working Group.

Canadian perspective on ED triage

needs to allow for rapid assessment and determination of priority based upon often times limited information. There has been a signicant interest in ED triage in Canada and in other regions (Australasia, UK) in recent years. The recognition of the importance of triage in patient assignment, the ow in the ED, and effective use of resources have resulted in renewed interest in triage scales that serve multiple functions. Triage scales are of primary importance clinically, but they may also be linked to performance criteria for ED efciency, as well as to ED workload, human resource requirements and funding. All of these have sparked intense interest in triage and triage scales.

Goal of triage development in Canada

There has been no universal agreement on the most reliable and valid triage scale. Scales varied from three to ve levels with differing denitions for each level of acuity. These scales historically have been either insufciently reliable or unproven to be valid as predictors of patient acuity.17 In Brillmans study of a four-level triage scale there was a high degree of inconsistency between physician and nurse assignment of triage levels.8 All of this was disappointing, considering the greater importance that triage was assuming in emergency patient care and ED function. Furthermore, triage scales were also being considered for use to refer patients out of the ED.9,10 The goal in the Canadian setting has been to develop a triage scale that has good interrater reliability and is valid, and relatively easy to learn and apply.

Recent history of triage development in Canada

Prior to the 1995 recommendation of the Canadian Association of Emergency Physicians (CAEP) to use a single ve-level triage scale,11 Canadian ED that were performing triage were using the three-level scale of emergent, urgent and deferrable. There was no literature that evaluated either the reliability or the validity of this scale despite its widespread application. Many larger volume ED had some form of formal triage; however, smaller ED often did not have triage and where triage existed this was done by clerks at the time of registration. Concurrently, Jelinek and Fitzgerald published their work on the Australasian National Triage Scale

(NTS)12 and Jelinek had shown good interrater reliability of the NTS.13 They had also shown that there was good correlation between triage level and several markers of severity of illness such as admission rate to hospital and ICU, mortality, injury severity scale, revised trauma score, Apache 2 scores, and average length of stay in the ED. This was in contrast to other triage scales that had been evaluated. Furthermore, the philosophy of care delivery in Canadian ED was changing in the 1990s. The focus was not only on those requiring the most immediate care but also on customer service. This customer service was focused on providing those with less urgent problems service within a reasonable time frame that met public expectations. The Australasian NTS has time frames for the patient to be seen by the nurse and the physician, that embodied this same philosophy. Furthermore the term deferrable was felt to be inappropriate, since the ED was often the only source for care for some patients with even minor problems. It is for these reasons that CAEP began to look at the Australasian work of Fitzgerald and Jelinek14 with the Australasian NTS. Beveridge rst introduced the CAEP ve level triage and acuity scale in 1995. This was based on the Australasian NTS and modied for the Canadian context. Following its publication by CAEP there was a renewed national interest in a standardized triage scale for all Canadian ED. This led to the formation of the CAEP National Working Group (NWG) on triage. This had representation from CAEP, the National Emergency Nurses Afliation, AMUQ (emergency physicians association in the province of Quebec), and the Society of Rural Physicians of Canada. The NWG reviewed the work of Beveridge and through a series of consensus meetings agreed on supporting one Canadian ED Triage and Acuity Scale (CTAS). This was endorsed by CAEP, AMUQ and NENA in May 1998. This followed with the development and publication of the CTAS Implementation Guidelines in October 1999.15 The CTAS were felt to contain the important elements of a triage scale. The fact that it was ve levels versus three would enable a more comprehensive splitting of patients based on acuity that could better reect the need for care. This was most noticeable in the move from a single urgent category to both urgent (CTAS 3) and less urgent (CTAS 4) categories. It also allowed for identication of those requiring immediate resuscitation (CTAS 1) in which both the physician and RN needed to be in attendance immediately, from 7

MJ Murray

Table 1. The Canadian Triage and Acuity Scale and fractile response rates for time to nurse and time to physician

Triage level

Time to nurse (min) Immediate Immediate 30 60 120 Time to physician (min) Immediate 15 30 60 120 Fractile response rate 98% 95% 90% 85% 80%
The time to nurse and to physician is from triage time and is the same for both nurse and physician for levels 3, 4 and 5.

those that were emergency (CTAS 2) in which there was time for the RN to perform an initial assessment prior to the attendance of the physician. Lastly CTAS contained a non-urgent (CTAS 5) level that eliminated the deferrable category and recognized that those with non-urgent needs may be appropriately and cost effectively seen in the ED. The CTAS also set time frames in which patients would ideally be seen in the ED (Table 1). These time frames had fractile response rates that measured system performance. They were arbitrarily set, based on both knowledge of time to care for some problems (for example, immediate for cardiac arrest, within 15 min for cardiac chest pain to capture potential thrombolysis within 30 min) and reasonable expectations for symptom relief (time to pain intervention) and consumer factors (waiting time for less urgent and non-urgent problems). The CTAS contains examples of typical presentations for each category (usual presentation), and potential differential diagnoses (sentinel diagnosis) with their respective ICD-9 codes. On the basis of the work of Jelinek with the NTS and admission rates by triage level, CTAS contained expected admission rates for each level. The CTAS gained popularity quickly once endorsed by the national organizations and centres began to implement it across Canada. The Minister of Health in Ontario introduced CTAS as mandatory in all ED in the province in July 1999. By 2001 the majority of provinces in Canada had sites using CTAS and in some provinces, it had been introduced province wide (report to the CTAS NWG).

Interrater reliability
Although most agreed that CTAS had face validity, it came under some criticism by academics for its widespread application before formal research evaluation 8

of interrater reliability and validity. The research on the NTS was used as an argument for the reliability of the CTAS, since no studies had been done with CTAS. This was until Beveridge et al. published the rst study that showed excellent rates of agreement between physicians and nurses using CTAS for written cases (quadratic weighted kappa = 0.80).16 The kappa statistic was used to assess interrater agreement beyond that expected by chance.17 The quadratically weighted kappas are described in the study and are based on the work described by Fleiss.18 The probability of choosing the same level for nurses and physicians combined was: CTAS 1 (0.69), CTAS 2 (0.50), CTAS 3 (0.46), CTAS 4 (0.50) and CTAS 5 (0.58). A further study reproduced the high rate of agreement between RN, physicians and paramedics (quadratic weighted = 0.77). The probability of choosing the same level for rst and second observer was: CTAS 1 (0.78), CTAS 2 (0.49), CTAS 3 (0.37), CTAS 4 (0.41) and CTAS 5 (0.49). This latter study attempted to meet the criticism that weighted kappas may not be a fair statistical method for evaluation. This work has increased the credibility of the CTAS as a reliable tool for triage. Unpublished reports have suggested that CTAS is a valid marker for acuity on the basis of admission rates to hospital and ICU, death rates, rates of transfer out to tertiary care centres, and ED length of stay. The widespread acceptance and application of CTAS in ED nationally has led to paramedics in some jurisdictions (province of Ontario) being trained to apply CTAS in the out-of-hospital setting. The goal of this initiative was to improve communication with ED through the use of the same language and to apply the triage determination in decisions on hospital bypass.19 This formed the backbone of the Patient Priority System implemented in Ontario to assist with identication of the most appropriate facility for patients. There is some discussion of adopting this in other provinces in the prehospital setting. A recent study found moderate to substantial agreement between paramedics and nurses on CTAS level assigned to 1636 patients presenting to the ED by ambulance (Tables 2 and 3).20 Both paramedics and nurses independently assigned the CTAS levels on ED arrival. The probability of agreement between the two observers on a given patient was 0.599 and the overall agreement within one level was 96%. The overall chance corrected agreement using quadratic weights was 0.61 (95% Cl 0.560.66). The highest agreement occurred at the most urgent triage levels. This study suggested that paramedics can reliably use

Canadian perspective on ED triage

Table 2. Probability of nurses and paramedics choosing the same triage level on presentation to the emergency department (ED)

Table 4. The Canadian Triage and Acuity Scale (CTAS) level and emergency department physician time for workload formula

Level chosen by ED triage nurse 1 2 3 4 5

n = 1437.

1 0.80 0.04 0.002 0.003 0.00

Level chosen by paramedic (PCP) 2 3 4 5 0.17 0.55 0.14 0.03 0.00 0.02 0.36 0.60 0.34 0.29 0.00 0.05 0.23 0.48 0.47 0.00 n = 46 0.00 n = 322 0.03 n = 719 0.14 n = 312 0.24 n = 38 Level 1 Level 2 Level 3 Level 4 Level 5

Average time per patient (minutes) (hours) 75.8 40.2 25.1 12.9 7.4 1.26 0.69 0.42 0.21 0.123

Number of patients per hour (pt/h) 0.8 1.5 2.4 4.6 8.1

CTAS. Furthermore, decisions around destination of the most urgent patients could be made on the basis of CTAS assessment.

The CTAS and its application to physician workload and funding

There has been interest in the recent application of triage data in determining ED physician workload and levels of physician stafng and funding using the workload formula developed by Murray (Table 4), (Equation 1). A = (% in each triage category) (average time per patient per triage category) =

and the workload and volumes of a large community hospital ED. There were several assumptions made in the development of the workload formula that were accepted by consensus. The times have not been validated in any time studies to date. The provincial health ministry in Ontario is funding a time study in 2003.

Current and future directions

Following CTAS implementation and more widespread use, it became apparent that there needed to be an enhancement of the triage scale for paediatrics. As a result of this, the CTAS NWG set out to develop the Canadian Paediatric Triage and Acuity Scale (PedsCTAS).24 The NWG expanded to include members from the Canadian Paediatric Society Section on Paediatric Emergency Medicine. Jarvis, Warren, and Leblanc produced the implementation guidelines for PedsCTAS for the NWG who then endorsed them nationally. An educational program has recently been developed by a subcommittee of the NWG for training beginning in the fall of 2002. The CTAS is currently recommended as a standard data element in the data set for Canadian ED Information Systems (CEDIS),25 developed by the CAEP CEDIS committee. The Canadian Institute of Health Information has included CTAS as a mandatory element in their National Ambulatory Care

( Vn / Vt )( Tn )
5

(1)

A = V1 / Vt T1 + V2 /Vt T2 + V3 /Vt T3 + V4 / Vt T4 + V5 /Vt T5 = V1 / Vt (1.26) + V2 /Vt (0.69) + V3 /Vt (0.42) + V4 / Vt (0.21) + V5 /Vt (0.123)

Workload = 1/A PPH (patients per hour) At least two provinces (Ontario and Nova Scotia) have funding models for physician services based upon this formula using CTAS levels. The times for each triage level were based upon several studies2123
Table 3. Interrater reliability between paramedics and nurses

Treatment of CTAS scale and analysis approach Five categories: unweighted kappa Five categories: weighted kappa Dichotomized (level 1 vs 25) kappa Dichotomized (level 1 or 2 vs 35) kappa
CTAS, The Canadian Triage and Acuity Scale.

Statistic value 0.34 0.61 0.74 0.55

95% condence limits 0.310.37 0.560.66 0.690.79 0.490.60

MJ Murray

Reporting System (NACRS). Work is ongoing in Canada by Bullard to develop an electronic triage tool (e triage) based on the CTAS. This has the potential to reduce interrater variability in the application of CTAS. The CTAS has become an integral part of ED triage in Canada. It is a dynamic document evidenced by the recent enhancements to develop PedsCTAS. The 1999 CTAS guidelines will be reviewed in 2002/3 by the NWG to identify areas that can be modied and improved.

7. 8.

Slay LE, Riskin WG. Algorithm directed triage in an emergency department. JACEP 1976; 5: 869 76. Brillman JC, Doezema D, Tandeberg D et al. Triage: Limitations in predicting need for emergent care and hospital admission. Ann. Emerg. Med. 1996; 27: 493 500. Derlet RW, Nishio DA. Refusing care to patients who present to an emergency department. Ann. Emerg. Med. 1990; 19: 26 262 7.

9.

10. Derlet RW, Kinser D, Ray L et al. Prospective identication and triage of non emergency patients out of an emergency department: a 5 year study. Ann. Emerg. Med. 1995; 25: 215 23. 11. Beveridge RC. The Presidents Letter. Ottawa, Ontario, Canada: Canadian Association of Emergency Physicians Communiqu, April 1 3, 1995. 12. Policy Document Australasian College for Emergency Medicine. National Triage Scale. Emerg. Med. 1994; 6: 145 6. 13. Jelinek G, Little M. Interrater reliability of the National Triage Scale. Emerg. Med. 1996; 8: 226 30. 14. Fitzgerald GJ. Emergency Department Triage [ Thesis]. Brisbane, Queensland, Australia: University of Queensland, 1989. 15. Beveridge R, Clarke B, Janes L et al. Canadian Emergency Department Triage and Acuity Scale implementation guidelines. CJEM 1999; 1 (Suppl.) S1 A24. 16. Beveridge R, Ducharme J, Janes L, Beaulieu S, Walter S. Reliability of the Canadian triage and acuity scale: interrater agreement. Ann. Emerg. Med. 1999; 34: 155 9. 17. Landis J, Koch G. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 159 74. 18. Fleiss J. Statistical Methods for Rates and Proportions, 2nd edn. New York, NY: Wiley 1981. 19. Murray M, Leblanc L, Beveridge R et al. The Canadian Triage and Acuity Scale for Ontario Paramedics. EHS Branch: Ontario Ministry of Health and Long Term Care, June 2001.

Conclusions
There is sufcient evidence that CTAS has good interrater reliability, face validity, and validity as a marker of acuity. It will continue to receive widespread acceptance and use across Canada. Other countries, including the United States, have expressed an interest in CTAS. There are some USA sites that have already adopted CTAS and are currently using it in their ED. There are many similarities between the Canadian, Australasian and UK triage scales which will allow for international comparison and benchmarking. Certainly the adoption of a similar ve-level scale in the USA would well serve the interests of emergency medicine there and worldwide. Canada will continue to use CTAS as the single nationally endorsed triage and acuity scale.

References
1. 2. 3. Haerden RD. Critical appraisal of papers describing triage systems. Acad. Emerg. Med. 1999; 6: 116671. Williams RM. Triage and emergency department services. Ann. Emerg. Med. 1996; 27: 506 8. Magnusson G. The hospital emergency department as the primary source of medical care. Scand. J. Soc. Med. 1980; 8: 14956. Weill TP, Spivey WH, Brown CG et al. Editorial: Clintons health reform and emergency department volumes: a return visit. Ann. Emerg. Med. 1993; 22: 117 19. Milner PC, Nocholl JP, Williams BT. Variation in demand for A & E departments in England 1974 84. J. Epidemiol. Commun. Health 1988; 43: 274 8. Pane GA, Farner MC, Salness KA. Cost implications of delayed access to care in patients admitted through the emergency department. Ann. Emerg. Med. 1991; 20: 730 3.

20. Murray M, Bondy S. The reliability of the Canadian triage and acuity scale in the prehospital setting: Interrater reliability between paramedics and nurses (Abstract). CJEM 2002; 4: 128. 21. Graff LG, Wolf S, Dinwoodie R, Buono D, Mucci D. Emergency physician workload: a time study. Ann. Emerg, Med. 1993; 22: 1156 63. 22. Bond M, Erwich-Nijhout M, Phillips D, Baggoley C. Urgency, disposition, and age groups: A casemix model for emergency medicine. Emerg. Med. 1998; 10: 103 10. 23. Bond M, Erwich-Nijhout M, Phillips D, Baggoley C. The identication of costs associated with emergency department attendances. Emerg. Med. 1997; 9: 181 8. 24. Warren D, Jarvis A, Leblanc L. Canadian Pediatric Triage and Acuity Scale: Implementation guidelines for emergency departments. CJEM 2001; 3 (Suppl.) S1 S27. 25. Innes G, Murray M, Grafstein E. A consensus based process to dene standard national data elements for a Canadian emergency department information system. CJEM 2001; 3: 277 84.

4.

5.

6.

10