We need to create EDs where the first priority is making the patient comfortable, not filling out onerous triage questionnaires
As the director of a single-doctor covered community ED for 25 years, I’ve been able to observe the evolution of ED triage. What I have concluded is that triage has grown into an intrusive, largely irrelevant part of the ED process that I want my patients to avoid if possible.
Sure, in a very busy ED that has no room for any new patients, you have no choice but to do triage. And in EDs that have multiple physicians or advanced practice clinicians there is the option for the “provider in triage” model. But in a single-covered ED this really isn’t an option.
This is not to say that we never did triage. But we only did it when there were no beds or chairs available. Yes, chairs. I would rather have new patients in chairs in the hallway than have them sit in the waiting room. At least when they are in the hallway you can go over and introduce yourself and get some stuff started and they can see that the ED is humming and that you are moving quickly and nobody is on coffee break. In the waiting room they get no sense of this urgency.
But back to why I don’t like triage.
There are lots of bad things that result from triage. First off, the initial registration and triage process makes a very bad first impression with our patients. I love the quote, “You never get a second chance to make a first impression.” I heard it on a “Head and Shoulder’s” shampoo commercial, but it is true nonetheless.
Generally when you go to an ED you pass a guard at the door (why don’t they have guards at the hospital’s front door?), then you sit in front of someone who is going to talk to you through a one-inch bullet-proof window (the patients may hurt us) in a narrow half cubicle (a real welcoming greeting). Then, when it’s your turn, you go to see the triage nurse. Now remember, in most EDs about 80% of the patients go home, so most have no serious conditions. Let’s say you hobble into the triage room with a sprained ankle.
Then it begins: What’s wrong? How did it happen? Any other injuries? Did you pass out? Any numbness in your foot? On a scale of 1-10, with 10 being worst, how would you rate your pain? What chronic diseases do you have? Diabetes? Hypertension? High lipids? Emphysema? Allergies? Have you had any recent hospitalizations? What surgery have you had in the past? What prescription medications do you take on a regular basis? Medication reconciliation – name, dose, frequency? Do you take any over-the-counter medications? Have you taken any medication prior to arrival today? Last oral intake? Last menstrual period? Any birth control? Pregnancy history? Are your immunizations current? Any recent travel to West Africa? Do you feel safe at home? Do you have limited English proficiency? What is your preferred language? Would you like a translator? How do you learn most effectively? Visual? Auditory? Do you have an advanced directive? Do you have a history of falling? Suicidal ideation? Are you interested in being screened for HIV or hepatitis C?
Now for the elements of the physical exam: Vital signs, weight, ankle / foot assessment – swelling, tenderness, deformity, pulse, sensation; nutritional assessment.
Now perhaps, depending on the department, the initiation of some tests or the like may occur and, maybe even an ice bag or pain pill will be given–
but most likely not.
What would you expect a person of normal intelligence to think when asked this battery of intrusive, largely irrelevant questions? Does it set the tone for a personalized, compassionate response to what is, in fact, a straightforward injury? Does it make it clear that the injury comes first and all of this other information, if mandated, could be asked later?
It seems that every regulatory agency wants a piece of the triage pie. In California we are “required” to ask about domestic violence and, if positive, report it – without asking the patient’s consent. Talk about being outrageously paternalistic! When this became the law in California I asked my wife what she would have thought about being screened for domestic violence at a routine ED visit. Her response— “It is none of their @#$%^& business.” She has a way with words. But then again the California legislature knows best. What do the nurses do? Check the box and move on.
And what about all of these other questions that are totally unrelated to the reason for the visit. Some “do gooders” feel the ED is the perfect place to do health screening. But I never heard one of our nurses upon completion of a nutritional assessment tell a person they needed to go on a diet. What do the nurses do? Check the box and move on.
There are two papers in the EMA database that reflect the frustration with the ever-expanding triage process. The first is a survey of ENA members in New York. The response rate was marginal, but the conclusions were clear – many of the questions had nothing to do with assigning a triage level and, oddly enough, the authors concluded that nurses believed that capturing lots of data was somehow important. My guess—most nurses actually agree that collecting the vast majority of this data is a waste of time, irrelevant and/or unnecessarily intrusive. Based on estimates made by the respondents, on average, triage took nine minutes. This is likely the least amount of time it takes when considering all of the work affiliated with the triage process (remember all of this key information has to be entered into an EMR).
EMERGENCY DEPARTMENT TRIAGE: WHAT DATA ARE NURSES COLLECTING?
Castner, J., J Emerg Nurs 37(4):417, July 2011
BACKGROUND: In the ED setting, triage is generally perceived to be a process of assessing a patient’s level of acuity. In actuality, triage in the ED requires an overall evaluation of the status of patients in the waiting room, provider workflow, treatment room congestion, and “boarding” of inpatients. Furthermore, the role of triage is expanding to include collection of increasing amounts of data, ordering preliminary diagnostic tests, and screening for various conditions.
METHODS: The author, from Buffalo General Hospital, surveyed a convenience sample of 1600 members of the Emergency Nurses Association to ascertain their perceptions of selected elements of the triage process.
RESULTS: Responses were received from 27% of those surveyed (430), 67% of whom were staff nurses. When asked to list the importance of 18 potential triage items, the highest mean scores were assigned to (4.07-4.92 on a scale of 1-5) were assigned to vital signs, allergies, pain score, medical history, pain variables (onset, location, intensity), surgical history, last menstrual period, weight, and maltreatment screening. Mean scores of 3.16-3.96 were assigned to the importance of limited English proficiency, immunization status, advance directives, medication reconciliation, pregnancy history, existing vascular access, last oral intake and height. The lowest mean score (2.98) was assigned to skin/wound screening. The respondents noted that documentation of most of these items is required in their EDs “always” or “most of the time.” The respondents estimated that the mean time spent for triage was 9 minutes per patient, although the range was wide (from 1 to 80 minutes).
CONCLUSIONS: Data collected by emergency nurses at triage appear to extend well beyond assessment of patient acuity and to exceed the 3-5 minutes estimated to be required for triage. 14 references (firstname.lastname@example.org – no reprints)
© 2011 by EMA – All Rights Reserved 12/11 – #12
The next paper expresses appropriate frustration regarding the growing number of conditions being screened in triage and the fact that most do not focus on the specific goals of triage – establishing the degree of urgency which is required to appropriately care for a patient.
LET’S ASK THAT OUT FRONT: HEALTH AND SAFETY SCREENINGS IN TRIAGE
Foley, A.L., et al, J Emerg Nurs 37(5):515, September 2011
The authors, nurses from the University of Washington Medical Center in Seattle, note that three-fourths of emergency department (ED) patients are ambulatory and present through triage. Comprehensive triage—a complete history, vital signs, and department-specific screening questions —takes 2-5 minutes and, the authors feel, can lead to a backlog of patients waiting to see the triage nurse. They argue that triage could be streamlined by eliminating department-specific screenings, which, while important for patient care, may not be necessary to determine triage acuity. The authors do not quarrel with the Joint Commission requirement of triage screening for suicidality in at-risk patients, but merely suggest that every patient need not be screened for suicidality. Screening for tuberculosis in at-risk patients with respiratory complaints, as recommended by the CDC-P, is not required for triage assessment, but the authors believe that such screening is required to protect the ED and the community. They stress that triage vital signs are an adjunct to nursing judgment in assigning acuity level, and are less important for the acutely emergent or completely nonemergent patient. The authors believe that only medications related to the patient’s chief complaint should be recorded in triage; a more complete list should be taken later. They feel that knowledge of immunization status is irrelevant to triage, “does not affect the chief complaint and is not going to change the course of care.” The authors also note that screening for falls and domestic violence does not affect triage acuity. The authors conclude that ED triage nurses face increasing pressure to identify those patients requiring immediate treatment. They suggest that careful thought be given regarding the essential components of screening in the ED in order to optimize the triage process without compromising patient safety. 7 references (email@example.com – no reprints)
© 2012 by EMA – All Rights Reserved 2/12 – #37
Although it may be “motherhood and apple pie” to use the triage process to screen for potentially serious chronic conditions in the ED, just what is the yield? The next paper makes it clear that universal screening for HIV is an extraordinary low-yield process (the authors say the impact is “modest”—a most charitable description). Want to do something serious for public health?—identify smokers and high blood pressure patients and see if even a small percentage can be interdicted. A far greater return-on-investment will be generated.
MODEST PUBLIC HEALTH IMPACT OF NONTARGETED HUMAN IMMUNODEFICIENCY VIRUS SCREENING IN 29 EMERGENCY DEPARTMENTS
d’Almeida, K.W., et al, Arch Intern Med 172(1):12, January 9, 2012
METHODS: In this French study, clinically stable patients aged 18-64 without self-reported HIV infection who presented to one of 29 EDs in metropolitan Paris and who were able to provide informed consent for rapid HIV testing were offered such testing. Rapid HIV antibody testing (OraQuick Advance Rapid HIV-1/2 Antibody Test), which provides results (negative, reactive or invalid) in 20-40 minutes, was performed by nurses at triage. Positive results were confirmed with additional testing, and HIV-positive patients were referred for follow-up care with infectious disease specialists.
RESULTS: Of 78,411 eligible patients (56.5% of all patients seen in the 29 EDs during the study period), rapid testing was offered to 20,962 (26.7% of eligible patients and 15.1% of all patients seen in the study EDs), and was actually performed in 12,754 (16.3% of eligible patients and 9.1% of all patients presenting to the EDs). HIV infection was confirmed in 37 of 38 patients with reactive rapid HIV tests (0.29% of patients tested and 0.05% of eligible patients), but most admitted that they had been aware of their HIV positivity and infection was newly diagnosed in only 18 patients (0.02% of eligible patients and 0.14% of those actually tested), all but one of whom belonged to traditional high-risk groups. Eight of the 18 patients presented to the ED with HIV-related symptoms, and would have been tested for diagnostic purposes. Only ten patients were asymptomatic. Of the 18 patients with confirmed HIV infection, one-third (six) did not return for follow-up care despite repeated attempts at contact.
CONCLUSIONS: These findings do not support nontargeted screening for HIV in the general ED population. 38 references (firstname.lastname@example.org – no reprints)
© 2012 by EMA – All Rights Reserved 8/12 – #37
The next paper looked at screening for self-harm during triage. It is clear that some EDs did it selectively (as would seem appropriate) while others had the screening on autopilot (the range was 3.5% of patients in one ED to 31% in another with one outlier screening 95% of its patients for self-harm risk).
EVALUATING CURRENT PATTERNS OF ASSESSMENT FOR SELF-HARM IN EMERGENCY DEPARTMENTS: A MULTICENTER STUDY
Caterino, J., et al, Acad Emerg Med 20(8):807, August 2013
BACKGROUND: Evidence suggests that assessment for self-harm (including suicide attempts, suicidal ideation and nonsuicidal self-injury) is suboptimal in patients presenting to the ED with psychiatric conditions.
METHODS: The authors, coordinated at Ohio State University, conducted a prospective observational cohort study of self-harm assessments in eight urban Level 1 trauma centers and 94,354 triaged adults.
RESULTS: Self-harm was identified in 2.7% of charts. Excluding one institution, 11% of ED patients were assessed for self-harm, with a range of 3.5-31%. There was one outlier, an institution in which 95% of patients were assessed. Rates of assessment were highest for institutions with specific self-harm policies. With the exception of the outlier ED, triage nurses assessed the risk of self-harm based on the patient’s complaint, behavior or history. Being male was a predictor for being assessed, with or without inclusion of the outlier ED (adjusted risk ratio (aRR) 1.17, 95% CI 1.10-1.26). Among all EDs, patients aged 65 or older were less likely to be assessed (aRR 0.56, 95% CI 0.35-0.92), and assessments were more likely on weekends (aRR 1.3, 95% CI 1.04-1.62) and after 3 PM (aRR 1.36, 95% CI 1.01-1.85). Although likely due to the outlier ED’s population, assessment rates were higher for American Indians or Alaskan natives (aRR 2.3, 95% CI 1.53-3.47) and Hispanics (aRR 2.41, 95% CI 1.67-3.48) among all EDs.
CONCLUSIONS: This study demonstrated substantial variability in assessment of patients for self-harm in the ED, and the importance of development of standardized self-harm assessment policies. 28 references (email@example.com for reprints)
© 2014 by EMA – All Rights Reserved 1/14 – #34
And what about “medication reconciliation?” I remember when the JC started mandating it and it was viewed as a nightmare. Did the patients know what they were taking? What it was for? And did the emergency physicians have any idea what was appropriate or not (heart meds, eye drops, statins)? What were we supposed to do about it – change patient medications? Call the family doctor and suggest a change? What a nightmare.
The next paper didn’t even need to be published. As predictable as the sun rising in the morning, everyone knows to expect ED medication reconciliation to be garbage.
MEDICATION HISTORY TAKING IN EMERGENCY DEPARTMENT TRIAGE IS INACCURATE AND INCOMPLETE
Mazer, M., et al, Acad Emerg Med 18(1):102, January 2011
BACKGROUND: It is estimated that 1.5 million preventable adverse drug events (ADEs) occur each year and that nearly 10% of patients who suffer an ADE are left with a disability. The prevention of such events begins with an accurate medication history.
METHODS: The authors, from the University of Pennsylvania, prospectively evaluated discrepancies between the medication history obtained by nurses at the time of triage in the ED and survey-based history subsequently recorded by research associates (considered the criterion standard). The study included 1,657 stable adult ED patients.
RESULTS: Mean patient age was 39 years. The patients were taking up to 21 medications each (mean, 2.6 per patient); 31% stated that they took no medications. Discrepancies between the survey and the medication list obtained by the nurses at triage were noted for 626 patients (nearly 38%). Discontinued mediations were included in the triage list for 10% of patients and medications were omitted for 28%. Thirty-eight percent of patients (632) reported taking over-the-counter medications not listed in the EMR.
CONCLUSIONS: Medication histories recorded by ED triage nurses were often inaccurate. The authors suggest that in view of the relatively low mean patient age and number of medications taken in this sample, and failure to include an allergy history, the study likely underestimated the scope of the problem. 10 references (firstname.lastname@example.org for reprints)
© 2011 by EMA – All Rights Reserved 7/11 – #34
Given all of the distractions associated with current day triage, it is a wonder that nurses can get the core process of triage down pat. The following paper notes that there are relatively low levels of interrater reliability when it comes to nurse triage, irrespective of the nurse’s level of experience. But remember, physicians don’t do so well either when it comes to practicing with limited variability despite, in many cases, clear-cut, evidence-based guidelines.
EMERGENCY DEPARTMENT TRIAGE: DO EXPERIENCED NURSES AGREE ON TRIAGE SCORES?
Dallaire, C., et al, J Emerg Med 42(6):736, June 2012
BACKGROUND: The accuracy of triage in the ED setting impacts departmental efficiency and patient safety.
METHODS: In this prospective study, from Universite Laval in Canada, written scenarios detailing the cases of 100 adult patients transported to the ED by ambulance were presented to five senior ED nurses experienced in the use of the Canadian Triage and Acuity Scale (CTAS), which is used in most of the country’s EDs. CTAS categories assigned for each patient by the five participating nurses were compared with the triagecategories assigned at the time of actual patient presentation to the ED. The nurses participating in this study were not provided with recent training in use of the CTAS, but reported experience with utilization of this instrument that ranged from two to six years.
RESULTS: Interrater agreement across pairs of nurses ranged between fair and good. The overall level of agreement was judged to be moderate (Kappa=0.44, 95% CI 0.40-0.48).
CONCLUSIONS: These results are consistent with relatively low levels of interrater agreement between experienced nurses in the application of the CTAS instrument, in the absence of recent retraining. 31 references
© 2012 by EMA – All Rights Reserved 11/12 – #33
The bottom line – create an ED that has open beds as much as possible and don’t triage when beds (or chairs) are available. Have the registration clerk call into the ED and have a nurse go out, greet the patient and escort the patient to an appropriate treatment area and have the nurse who is going to care for the patient during their stay ask all of the questions your ED feels are required to keep you from being cited by the JC or other regulators. And ask these questions well after the patient is made comfortable and their primary concerns are addressed.