When sitting in a first grader s desk as an adult,

CLINICAL ARTICLE The Case for a Universal, Valid, Reliable 5-Tier Triage Acuity Scale For US Emergency Departments Author: Polly Gerber Zimmermann, RN, MS, MBA, CEN, Chicago, Ill Polly Gerber Zimmermann, Illinois ENA, is Instructor, Department of Nursing, Harry S Truman College, and Associate Nurse, American Airlines, Chicago O Hare International Airport, Chicago, Ill. For reprints, write: Polly Gerber Zimmermann, RN, MS, MBA, CEN, 4200 N. Francisco, Chicago, IL 60618; E-mail: pzimmermann@ccc.edu. J Emerg Nurs 2001;27:246-54. Copyright 2001 by the Emergency Nurses Association. 0099-1767/2001 $35.00 + 0 18/1/115284 doi:10.1067/men.2001.115284 When sitting in a first grader s desk as an adult, it does not take long to realize that the desk no longer fits. Similarly, the US emergency departments traditional triage acuity rating scales no longer fulfill their intended purpose. ED visits have reached 100 million a year, whereas the number of emergency departments has decreased by 8%. 1,2 This situation is attributed to (1) ED volume increasing beyond the resources for immediate care; (2) emergency departments becoming the principle provider of primary medical care during off-duty hours; (3) emergency departments being used for primary care; and (4) emerging thirdparty payers. 3,4 These developments also emphasize the need to re-examine the US triage systems. The purpose of ED triage is to have the right person at the right place at the right time for the right reason. However, expectations for triage are now much greater than the original intent of battlefield triage, which was merely finding persons who were salvageable with the limited available resources. Now triage standards also seek to minimize morbidity, disfigurement, pain, emotional distress, and client dissatisfaction with their emergent care experience. Triage systems A triage system is the basic framework with which patients are sorted, using an acuity rating scale. In the most simplistic form, a traffic director greets and directs the patient to the correct treatment area based on an initial impression. This system was used by 4.8% of responding emergency departments in the ENA 1998 database 5 ; however, most emergency departments use 1 of 3 other triage systems: spot check, comprehensive, or 2-tier. 246 JOURNAL OF EMERGENCY NURSING 27:3 June 2001

TABLE 1 3-Tier acuity rating triage system Level Term Time frame for treatment/reassessment Examples Level 1 Emergent (red) Immediate/constant Cardiac chest pain Level 2 Urgent (yellow) <2 h/30 min Abdominal pain; open fracture Level 3 Nonurgent (green) >2 h/1-2 h Rash; vaginal discharge TABLE 2 4-Tier acuity rating triage system Level Term Time frame for treatment Examples Level 1 Emergent Immediate Chest pain Level 2 Acute/significant alteration As soon as possible; Pneumonia with shortness of breath 15-30 min Level 3 Urgent/in a timely fashion 30-60 min Hip fracture Level 4 When possible 60 min Cold/cough; urinary tract infection TABLE 3 Canada Triage and Acuity Scale Level Time framework Examples Level 1 Immediate Unconscious Level 2 Immediate to RN; Overdose; domestic abuse 15 min to physician Level 3 30 min Vomiting/diarrhea <2 y; acute psychosis Level 4 1 h Minor trauma; moderate pain (4-7 on a scale of 1-10); earache Level 5 2 h Vaginal bleeding, mild pain (less than 4 on a scale of 1-10); vomiting/diarrhea alone, no dehydration; >2 y Reprinted with permission from Canadian Association of Emergency Physicians. Canadian Emergency Department Triage and Acuity Scale implementation guidelines. J Can Assoc Emerg Phys 1999;1(Suppl). SPOT CHECK Spot check is a more detailed quick look during which a patient, based on his or her presenting complaint, is assigned a category within 2 minutes. This system, used by 28% of the responding emergency departments, 5 works best in a low-volume situation where treatment area space is readily available. COMPREHENSIVE The ENA Practice Standard is comprehensive triage, in which an initial across-the-room look (general appearance, work of breathing, circulation) is performed and vital signs, a complete history, and designated additional information (such as a domestic abuse screening or advanced directives) are obtained. According to the standard, triage should be completed by an experienced, specially trained RN who considers the patient s physical and psychosocial needs as well as the department s patient flow. 6 Comprehensive triage is to be completed in 2 to 5 minutes; however, according to one study, this goal is being met only 22% of the time. More time is needed for elderly and pediatric patients in particular. 7,8 TWO-TIER Two-tier triage, which is usually used in larger emergency departments, builds on a combination of the 2 previous systems. A registered nurse (RN) does an initial quick June 2001 27:3 JOURNAL OF EMERGENCY NURSING 247

TABLE 4 Canada Triage and Acuity Scale respiratory distress distinguishers Level Time frame Descriptors Level 1 Immediate Unable to speak; cyanosis; lethargic; tachycardia/bradycardia; PO 2 <90% Level 2 15 min Severe asthma; PEFR <40% predicted or previous best; dyspnea Level 3 30 min Mild/moderate asthma; frequent cough; night awakening; PEFR 40%- 60% previous best & PO 2 92%-94%; moderate dyspnea; SOB with exertion Level 4 1 h URI s/s; PO 2 95% Level 5 2 h URI; sore throat with normal VS PEFR, Peak expiratory flow rate; SOB, shortness of breath; URI, urinary tract infection; VS, vital sign. Reprinted with permission from Canadian Association of Emergency Physicians. Canadian Emergency Department Triage and Acuity Scale implementation guidelines. J Can Assoc Emerg Phys 1999;1(Suppl). look to sort all incoming patients. The emergent patients are readily identified and sent to the treatment area with minimal triage involvement. This nurse also often handles any family and visitor concerns. A second nurse performs in-depth, less interrupted assessments and histories for urgent and nonurgent patients. Variations include (1) sending a second staff member to assist when 5 or more patients are awaiting triage, or (2) having the triage nurse initially sort out nonurgent patients and sending them directly to a fast-track area where complete assessments and vital signs are then performed. 3,9 Acuity scales The ideal every patient being immediately seen in the treatment area is just not feasible. Therefore, the focus usually becomes the ordering of patients by acuity ratings within the system. One study found that, using a 3-tiered acuity rating, emergent determinations made by nurses of the same patient varied from 11% to 63%. NEED FOR RELIABILITY/REPLICABILITY An ideal triage system must be easily applied and statistically valid and must capture a novel presentation of a serious condition. Determining triage categories that are (1) reliable and reproducible and (2) independent of the nurse performing the role is important. Then comparisons can be made within and between emergency departments. When these criteria can be met, triage categories can have a significant role in legal defense, federal funding, case mix data, staffing levels, predictable resource consumption and budgeting, managed care decision making, and outcome measures for admission rates, ED length of stay, and complexity of care. 3,10 THREE- AND FOUR-TIER ACUITY RATINGS Currently, the 3-tier acuity system is the most popular system in the United States. Of the responding emergency departments, 60.8% use this system (Table 1). 5,11 A 4-level system, used by 6% of responding emergency departments in 1996, is promoted by the 1997 ENA triage manual. 11,12 Most versions of this scale make a more precise distinction in the middle category of the 3- level system (Table 2). 4 However, these 2 acuity rating scales lack reliability and replicability in comparing determinations made both by different persons (interrater agreement) or the same persons on different occasions (test-retest reliability). 13-15 One study found that, using a 3-tiered acuity rating, emergent determinations made by nurses of the same patient varied from 11% to 63% (κ = 0.38). 16 In another study, 87 nurses who used a 3-tiered triage scale to triage 5 standardized patient scenarios showed poor agreement (κ = 0.347) with poor test-retest agreement. Only 24% of the nurses rated all 5 cases the same on both occasions, and 46% changed the ratings by more than one severity level. 15 248 JOURNAL OF EMERGENCY NURSING 27:3 June 2001

FIGURE 1 Manchester Triage Group algorithm. PEFR, Peak expiratory flow rate; SaO 2, saturation of 95% on room air. (Reprinted with permission from Mackway-Jones K, editor. Emergency triage: Manchester Triage Group. London: BMJ Publishing Group; 1997. p. 46.) The National Triage Scale, the first standardized 5-level acuity system, originated in 1993 in Ipswich, Australia. It has shown good reliability. EXCEPTION DETERMINANTS To compensate for the large spans of time within the 3- or 4-tier systems, individuals and/or hospital protocols often develop official or unofficial prioritization factors. Some examples include patients who are: presenting within 2 hours of a respiratory treatment. experiencing very severe pain. TABLE 5 Manchester Triage Scale Level Level 1 Level 2 Level 3 Level 4 Level 5 Time frame Immediate 10 min 60 min 2 h 4 h Reprinted with permission from Mackway-Jones K, editor. Emergency triage: Manchester Triage Group. London: BMJ Publishing Group; 1997. p. 46. June 2001 27:3 JOURNAL OF EMERGENCY NURSING 249

TABLE 6 ESI level definitions ESI-1 ESI-2 Stability of vital functions (ABCs) Unstable Threatened Life threat or organ threat Obvious Likely but not always obvious How soon the patient should be seen by physician Immediately Minutes Expected resource intensity High resource intensity; staff at High resource intensity; multiple, bedside continuously; often often complex diagnostic studies; mobilization of team response frequent consultation; continuous (remote) monitoring Examples Cardiac arrest, intubated trauma Chest pain probably resulting from patient, severe overdose, SIDS ischemia, multiple trauma unless responsive, child with fever and lethargy, disruptive psychiatric patient ABCs, Airway, breathing, circulation; SIDS, sudden infant death syndrome. Reprinted with permission. Copyright 1999, Richard C. Wuerz, MD, and David R. Eitel, MD. arriving directly from a physician s office. returning to the emergency department within 24 hours after ED discharge. young children, especially late at night. behaving in a disruptive, violent, out-of-control, or incompetent manner. Although these or other discriminators may contribute to better patient prioritization, consequential individual practitioner subjectivity and variability are still an issue. Nursing experience is also a notable factor. Five-tier acuity ratings An awareness of the added precision and consistency of a 5- tier acuity rating system is growing. A 1996 ENA study found that 10% of the responding emergency departments already used some form of a 5-level acuity rating, 11 although considerable individual and regional variations exist. Internationally, national standardized 5-level systems with proven reliability and reproducibility are being instituted. To aid comparison, descriptions are made by 5 numerical levels, with 1 being the highest acuity, although the triage scale itself may use different terms, colors, or both. NATIONAL TRIAGE SCALE OF AUSTRALIA The National Triage Scale (NTS), the first standardized 5- level acuity system, originated in 1993 in Ipswich, Australia. It has shown good reliability with similar degrees of agreement when analyzed by site and years of nurse experience. 10,16-19 One study found that with the NTS, more than 50% of 110 Australian nurses had exact agreement on 100 written patient profiles. 20 CANADA TRIAGE AND ACUITY SCALE In 1995, a group of Canadian ED physicians and nurses developed the Canada Triage and Acuity Scale (CTAS)* (Table 3). 21 Its use became official policy in Canada in 1997, and the system is endorsed by the Canadian Association of Emergency Physicians and the National Emergency Nurses Affiliation of Canada. Description: The CTAS, similar to most US triage systems, is heavily based on a presumptive disease diagnosis, often made from an articulate, detailed history or description. One helpful attribute is the use of key objective data when tiering common complaints, such as respiratory symptoms or vaginal bleeding, into an appropriate triage category. With these discriminators, obtaining a complete history in the triage area is not necessary. For instance, with *CTAS implementation guidelines can be ordered from the Canadian Association of Emergency Physicians, 1785 Alta Vista Dr, Suite 104, Ottawa, Canada ON K1G 3Y6, phone (613)523-3343 or (800)463-1158, fax (613)523-0190, E-mail: guidelines@caep.ca. The price is $12 for a guidelines booklet and $5 for a pocket card. 250 JOURNAL OF EMERGENCY NURSING 27:3 June 2001

ESI-3 ESI-4 ESI-5 Stable Stable Stable Unlikely but possible No No Up to 1 h Could be delayed Could be delayed Medium/high resource intensity; multiple Low resource intensity; one Low resource intensity; examination only diagnostic studies or brief observation; simple diagnostic study; or complex procedure or a simple procedure Abdominal pain or gynecologic Closed extremity trauma, Cold symptoms, minor burn, recheck disorders unless in severe distress, simple laceration, cystitis, hip fracture in elderly patient typical migraine respiratory distress (regardless of the cause), the triage nurse focuses on the patient s ability to speak, PO 2, and peak expiratory flow resistance (Table 4). Age-specific parameters and educational implementation material with tips are also included. For instance, a true sudden onset headache (which is more likely to be a cerebral bleed) is when pain intensity peaks within a few seconds ( like being hit by a 2 4 ). Advantages/disadvantages: The CTAS has reliability and validity. One study found inter-rater agreement (κ = 0.66) 22 ; another study found an overall κ of 0.84. 13 Studies show that the CTAS yields results similar to the NTS in terms of seeing patients within their time frame. 21 However, up to 20% of level 4 and up to 10% of level 5 patients were admitted to the hospital. 21 These data raise the question of whether the scale is sensitive enough to discern the more obscure, serious problems. Obtaining necessary historical details for an accurate presumptive diagnosis can be time-consuming. In addition, the process (and triage decision) is hindered if the patient is a poor historian, has multiple complaints, or has a language problem. MANCHESTER TRIAGE GROUP In England, the Royal College of Nursing Accident and Emergency Association and the British Association for Accident and Emergency Medicine initially developed a 5- point scale (Table 5). In 1997, a group of British Accident and Emergency physicians and RNs met to develop a flowbased, reproducible training manual and reference. This system, endorsed by the Accident and Emergency Nurses Association, has become known as the Manchester Triage Group because they met in Manchester, England. 23,24 Description presentation over medical diagnosis: The Manchester Triage System notes that other systems with a basic premise of attempting to make a tentative medical diagnosis at triage are doomed to fail. This group believed that diagnosis-based triage scales could actually be dangerous because triage, by definition, has limited time, history, and objective data. The only appropriate focus of triage is to identify key signs/symptoms so as to place patients in an appropriate level for their generic acuity or risk. For instance, abdominal pain with shoulder pain, regardless of its cause, could signal free air and automatically moves the patient to a higher level of acuity. It is not up to the triage nurse to determine whether the cause is a ruptured ovarian cyst, a perforated gastric ulcer, or merely a muscle strain. Manchester Triage Group process: The system consists of 4 steps during triage, which are as follows 23 : 1. Identify the presenting complaint and pick an appropriate flow chart from the 52 choices. Flow charts overlap. An adult with a stiff neck and headache who does not feel well will be given the same priority whether the flow chart for an unwell adult, neck pain, or headache is chosen. June 2001 27:3 JOURNAL OF EMERGENCY NURSING 251

FIGURE 2 Emergency Severity Index triage algorithm. HR, Heart rate; RR, respiratory rate; SaO 2, saturation of 95% on room air. (Reprinted with permission from Wuerz RC, Travers D, Gilboy N, Eitel DR, Rosenau A, Yazhari R. Implementation and refinement of the Emergency Severity Index. Acad Emerg Med 2001;8:170-6. Published by Hanley & Belfus, Philadelphia, PA, www.hanleyandbelfus.com.) 2. Gather and analyze information using 6 general key discriminators to determine a level of priority. The general discriminators that apply to all patients include the following: Life threat (no airway, breathing, and circulation) Pain Hemorrhage Conscious level Temperature Acuteness (eg, started or acutely worse within the previous 7 days) The pain discriminator is determined by the patient s description rather than rating on a numeric scale. Discriminators specific to the complaints, such as pleuritic pain, are applied within the complaint-specific flow chart. Temperature, for example, hot child, is determined by feeling with the practitioner s hand, not a thermometer reading. 3. Evaluate and select alternatives, using general and specific discriminators within the flow chart to identify the patient s general acuity. The patient s triage level is determined by discriminators that consider both the disease history and current physical presentation. For example, the patient is a level 3 (60-minute wait before being seen) if a significant history of asthma exists (defined as a brittle asthmatic or has a history of a life-threatening incident) or no improvement is noted with bronchodilator therapy (given by either the patient or physician). The patient also currently exhibits the objective signs of physical compromise of low peak expiratory flow rate (PEFR) (PEFR of 50% or less of best or predicted PEFR) or low saturation of 95% on room air (SaO 2 ). However, the patient is elevated to a level 2 (10-minute wait before being seen) when more ominous physical symptoms are present. These symptoms include marked tachycardia (defined as greater than 120 bpm); very low SaO 2 (saturation 90% on room air or 95% with oxygen therapy); very low PEFR (33% or less of best or predicted PEFR); or inability to talk in sentences (Figure 1). 4. Simplified documentation. The routine triage documentation is only the presentation chart used, significant discriminators, and level assigned (eg, asthma, audible wheeze, PO 2 97% on room air, level 4). The triage nurse is not expected to obtain other standard historic information, such as the trigger for this episode (history of a recent upper respiratory infection) or whether immunizations are up-to-date. Patient education is an integral part of the triage process. Patients tend to like the system because they can readily observe how decisions are made and can determine that the system is equitable. 252 JOURNAL OF EMERGENCY NURSING 27:3 June 2001

Advantages/disadvantages: With this approach, less dependence on the patient s history or communication skills is required. A novel occurrence with key symptoms is more likely to receive a proper acuity rating. This presentation-based method is particularly advantageous for novice ED nurses. Practitioners quickly integrate the most commonly used presentation charts as part of their nursing knowledge. More important, this standardization system has shown reproducible inter-rater and test/retest reliability, 25 which has allowed intra- and inter-departmental comparison for resource issues. One major criticism of the system is the prominence given to the patient s subjective complaint about pain. The patient who states that he or she has unbearable pain theoretically places himself or herself in level 2, even if the pain is only from a strep throat. The developers of the system admit that the consideration of pain results in more patients being triaged into level 2. However, the intent was increased recognition of pain as a legitimate criterion and the use of immediate pain-relieving strategies in the triage area. Patients can be retriaged to a lower level after relief is obtained. Some experienced nurses complain that the system constrains an expert s gut instincts. However, developers counter that expertise is still important in other aspects of triage, such as interpersonal skills and management decisions. EMERGENCY SEVERITY INDEX Resource consumption and acuity are the basis of the Emergency Severity Index (ESI), a triage algorithm developed by the late Dr Richard Wuerz (at the Brigham and Women s Hospital in Boston) and Dr David Eitel (at York Hospital in York, Pa) (Figure 2, Table 6). After the initial decision about a life-threatening presentation, the distinguishers for level 2 are high risk, mental status alteration, severe pain, or abnormal vital signs/po 2. Severe pain/distress is determined by nurse clinical observation and/or a patient rating of a 7 on a 0 to 10 pain scale. The system then incorporates anticipated different types of resource interventions (eg, radiograph[s], laboratory test[s], injection, procedure, consult). It expands the concept of triage from beyond when should the patient be seen to also what does the patient need. A resource determination is made according to prudent standards of practice, not a particular physician s preference for more or less testing than is typically performed for this type of patient complaint. Advantages/disadvantages: Studies of the ESI indicate reproducibility. On written case studies, inter-rater reliability was excellent, with chance-corrected agreement of 80% between physician and nurse triage assignments and pairs weighted κ scores ranging from 0.83 to 0.96. One study showed improved reliability of this scale compared with the traditional 3-level triage scale (0.52 κ for 3-level scale; 0.59 κ for 5-level ESI). 26 The patient s ESI triage level accurately predicts the need for hospitalization and the patient s 6-month survival (Kaplan-Meier χ 2 = 25.9; P <.0001). 27-31 The ESI has been implemented and validated at 9 hospitals. Already this system has helped the Brigham and Women s emergency department objectively identify high patient acuity and the need for possible staffing adjustments. One criticism is that a minor variation in vital signs can result in a nonurgent patient being unnecessarily placed in a higher level. This can occur because the algorithm is sensitive in identifying patients who might be seriously ill, especially in an uncommon presentation. As a result, the initial vital sign parameters were modified to help with this issue. Conclusion The music has changed, so must the dance is an African proverb. So it is with the US ED triage acuity systems. Studies show that the widely used 3-tier system is ineffective in meeting the needs of today s burgeoning ED patient population. The solution seems to lie not in attempting to improve individuals performance within the same system but in embracing an improved process. The call for change is clear, even though the system that is best suited for use in the United States is not as evident. All standardized 5-tier systems have reliability and validity. Any one of the described systems could work well in the United States. The difficulty may be for traditionally individualistic ED nurses to adjust to a consistent adherence according to absolute criteria. All of the 5-tier systems except the ESI are being used in a socialized medicine environment, which could be an influence on their acceptance by the general population, as June 2001 27:3 JOURNAL OF EMERGENCY NURSING 253

well as by the emergency nursing community. The late Richard Wuerz, co-developer of the ESI, believed that each of the current international 5-tier systems uniquely fits the culture of its country of origin. The Canadian system will feel the most familiar to United States ED nurses because, like most US ED triage systems, it makes some type of diagnosis to help determine the triage level decision. Dr Wuerz believed that the ESI, developed in the United States, uniquely fits the litigious nature of the American society. It does include an element of practitioner expertise and experience application, such as what resources are needed to obtain a diagnosis for a female patient with abdominal pain. Converting to any of the 5-tier triage systems will require a commitment to provide an extensive training program for ED nurses and doctors to ensure they are comfortable with and skilled in consistent application of the system. It may be time for US ED professionals to begin the exciting journey of change by learning about the standardized 5-tier systems and selecting a standardized triage system for our nation. Acknowledgment I acknowledge with gratitude the input of Debbie Travers, RN, MSN; the late Richard C. Wuerz, MD; David Eitel, MD, MBA; Nicki Gilboy, RN, MS; Janet Marsden, MSc, BSC, RGN, OND, MiMgt; Lisa Hadfield-Law, RNG, A/E Cert.; and Brian Dolan, BSC (HONS), RMN RGN, CHSM. REFERENCES 1. Eisenberg D. Critical condition. Time 2000 Jan 31;32-4. 2. Nourjah P. National ambulatory medical care survey: 1997 emergency department summary (Vital and Health Statistic). Atlanta (GA): Centers for Disease Control and Prevention; 1999. 3. Gilboy N, Travers D, Wuerz R. Re-evaluating triage in the new millennium: a comprehensive look at the need for standardization and quality. J Emerg Nurs 1999;25:468-73. 4. Grossman VGA. Quick reference to triage 1999. Philadelphia: Lippincott. 5. Emergency Nurses Association. National emergency department benchmark and staffing database guide. Des Plaines (IL): The Association; 1999. 6. Emergency Nurses Association. Standards of emergency nursing practice. St. Louis: Mosby; 1998. 7. Travers D. Triage: how long does it take? How long should it take? J Emerg Nurs 1999;25:238-40. 8. Keddington RK. A triage vital sign policy for a children s hospital emergency department. J Emerg Nurs 1998;24:189-92. 9. Zimmermann PG. Manager s forum. J Emerg Nurs 2001;27: 286-92. 10. Cameron PA, Bradt DA, Ashby R. Emergency medicine in Australia. Ann Emerg Med 1997;28:342-6. 11. Emergency Nurses Association. Triage: meeting the challenge. 2nd edition. Des Plaines (IL): The Association; 1997. 12. Emergency Nurses Association. National emergency department benchmark and staffing database guide. Des Plaines (IL): The Association; 1996. 13. Brillman JC, Doezema D, Tandberg D, Sklar DP, Davis KD, Simms S, et al. Triage: limitations in predicting need for emergent care and hospital admission. Ann Emerg Med 1996;24:498-500. 14. Beveridge RC, Ducharme J, Janes L, Beaulieu S, Walter S. Reliability of the Canadian ED Triage & Acuity Scale: inter-rater agreement. Ann Emerg Med 1999;34:155-9. 15. Wuerz RC, Fernades CMB, Alarcon J. Inconsistency of emergency department triage. Ann Emerg Med 1998;32:431-5. 16. Gill JM, Reese CL, Diamond JL. Disagreement among healthcare professionals about the urgent needs of emergency department patients. Ann Emerg Med 1996;28:474-8. 17. Jelinek GA, Little M. Inter-rater reliability of the National Triage Scale of 11,500 simulated occasions of triage. Emerg Med 1996;8:226-30. 18. Hollis G, Sprivulis P. Reliability of the National Triage Scale with changes in emergency department acuity level. Emerg Med 1996;8:231-4. 19. Dilley S, Standen P. Victorian nurses demonstrate concordance in the application of the National Triage Scale. Emerg Med 1998;10:12-8. 20. FitzGerald GJ. Emergency department triage [dissertation]. Queensland, Australia: University of Queensland; 1989. 21. Canadian Association of Emergency Physicians. Canadian Emergency Department Triage and Acuity Scale implementation guidelines. J Can Assoc Emerg Phys 1999;1(Suppl):S1-S16. 22. Fernandez CMB, Wuerz RC, Clark S, Djurdjev O. How reliable is emergency department triage? Ann Emerg Med 1999;34:41-7. 23. Mackway-Jones K, editor. Emergency triage: Manchester Triage Group. London: BMJ Publishing Group; 1997. 24. Woolwich C. Nurse triage. In: Dolan B, Holt L, editors. Accident and emergency: theory into practice. New York: Baillière Tindall; 2000. p. 505-12. 25. Cook MW, Jinks S. Does the Manchester triage system detect the critically ill? J Accid Emerg Med 1999;16:179-81. 26. Travers DA, Waller AE, Bowling JM, Flowers DF. Comparison of 3-level and 5-level triage acuity systems [abstract]. Acad Emerg Med 2000;7:233. 27. Wuerz RC, Milne LW, Eitel DR, Travers D, Gilboy N. Reliability and validity of a new five-level triage instrument. Acad Emerg Med 2000;7:236-42. 28. Wuerz R, Travers D, Gilboy N, Yazhari R, Eitel D. Implementation of five-level triage at two university hospitals [poster presentation]. Acad Emerg Med 2000;7:233. 29. Wuerz R, Milne L, Eitel D, Wiencek J, Simonds W. Outcomes are predicted by a new 5-level triage algorithm. Acad Emerg Med 1999;6:398. 30. Wuerz R, Milne L, Eitel D, Wiencek J, Simonds W. Pilot phase reliability of a new five-level triage algorithm. Acad Emerg Med 1999;6:398-9. 31. Wuerz RC. Emergency Severity Index triage status and shortterm survival [abstract]. Presentation by D.R. Eitel at American College of Emergency Physicians Research Forum, 2000 Oct. 254 JOURNAL OF EMERGENCY NURSING 27:3 June 2001