- Email: [email protected]
Using Simulation to Assess the Impact of Triage Interruptions
RESEARCH
USING SIMULATION TO ASSESS THE IMPACT OF TRIAGE INTERRUPTIONS Authors: Kimberly D. Johnson, PhD, RN, CEN, and Abeer Alhaj-Ali, BSN, Cincinnati, OH
Earn Up to 5.5 CE Hours. See page 489.
Contribution to Emergency Nursing Practice
• This research is innovative because simulation has not previously been used to assess the impact of interruptions on triage. • Emergency nurse educators could use the methods and results discussed in this project to guide departmental training of triage nurses. • Practicing in a simulated setting can help new triage nurses hone their skills at managing interruptions while providing veteran triage nurses with ways to recognize factors that are affecting the accuracy of their triage. Abstract Introduction: Interruptions are common in the emergency
was used to collect information about the nurses’ education and experience. The Emergency Severity Index (ESI) was used for triage categorization. Each participant completed 2 scenarios (one interrupted and one uninterrupted). After completion of the scenarios, video-simulated recall interviews were used to assess the simulation experience and the impact that interruptions had on the triage decision-making process. Results: Triage time had a mean of 10 minutes and ranged between 4.34 minutes and 13.45 minutes. However, triage was significantly longer during the interrupted scenarios. Seventy-seven percent of the acuity assessments (ESI) were correct. Of the 18 scenarios, 3 uninterrupted scenarios had incorrect ESI scores, and one interrupted scenario had a missing acuity score.
department and contribute to catastrophic errors. Care priorities and acuity levels are assigned during triage, meaning that mistakes and omissions during the triage process could have detrimental effects on patients. The purpose of this project was to assess the feasibility of investigating the impact of interruptions on triage and the decision-making process in a simulated setting.
Discussion: This study provides the basis for future work that looks at how nurses successfully manage interruptions and tests interventions to assist triage nurses in managing or reducing interruptions during this important patient assessment process.
Methods: A 2-phase, sequential exploratory mixed method design was used. Nine nurses from 3 emergency departments in a Midwest area participated. A short demographic questionnaire
Key words: Triage; Interruptions; Distractions; Emergency department; Triage nurse role; Simulation; Video-simulated recall interviews
lmost 100,000 deaths per year are attributed to health care–related errors. 1,2 Interruptions have been identified as a cause of errors and can be
detrimental to patient safety. 1,2 Frequently the care provided during ED visits is interrupted. 3,4 One particular area that is fraught with interruptions is the triage process, and thus this part of the patient’s ED experience is particularly vulnerable to errors. 5–7 With annual ED visits increasing to more than 130 million per year, 8 triage nurses are pressured to quickly and accurately assess each patient. Triage assessment, defined as assigning acuity to patients to determine treatment priority, is the critical beginning of the treatment plan for ED patients. The initial triage assessment can affect both the ED visit quality and length. 5 Because of the nature of the emergency department, nurses are exposed to frequent interruptions, which may lead to inaccuracies in acuity assessment, thus resulting in errors that may affect the quality of care. 6 Although errors and delays adversely affect patient outcomes, 9
A
Kimberly D. Johnson, Member, Greater Cincinnati Chapter, is Assistant Professor, College of Nursing, University of Cincinnati, Cincinnati, OH. Abeer Alhaj-Ali is a PhD student, University of Cincinnati, Cincinnati, OH. This work was supported by a ENA Foundation/Sigma Theta Tau International Research Grant. For correspondence, write: Kimberly D. Johnson, PhD, RN, CEN, University of Cincinnati, 3110 Vine St, Cincinnati, OH 45221; E-mail: [email protected]. J Emerg Nurs 2017;43:435-43. 0099-1767 Copyright © 2017 Emergency Nurses Association. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jen.2017.04.008
September 2017
VOLUME 43 • ISSUE 5
WWW.JENONLINE.ORG
435
RESEARCH/Johnson and Alhaj-Ali
the true impact of triage interruptions on patients’ clinical outcomes is unknown because few studies have investigated this phenomenon. The purpose of this research was to determine the feasibility of using a simulated setting to investigate the impact of interruptions and how triage nurses make data-based decisions in the presence of interruptions.
sets priorities for the treatment team, and also begins the health care customer experience, it stands to reason that the triage process should be a primary target for intervention. To best design an improvement intervention, understanding the effect of triage interruptions on the quality of health care is vital. 5,17
Background and Significance
TRIAGE INTERRUPTIONS
The triage nurse’s assessment of a patient is an important first step in an episode of care and can be an indicator of how the patient’s ED experience will progress. 5 Interruptions may lead to an incorrect triage decision, missed symptom identification, incomplete assessment, or unasked questions. Any of these factors could potentially delay care, resulting in significant morbidity or mortality. 10,11 A patient seeking treatment may have a triage interview interrupted for myriad reasons that include addressing the needs of other patients, visitors, or staff. The interruptions then create delays in getting patients into a treatment area, keep nurses from collecting appropriate triage data, or cause nurses to make poor or erroneous triage decisions. 12–14 Errors during triage can decrease quality of care and have an adverse impact on patient outcomes. 1,9,10
Previous work by Johnson et al 5 identified the types and frequency of interruptions that occur during the triage process. These investigators reported that triage nurses were interrupted an average of 16.6 times per shift for a variety of reasons; top occurrences included being asked to grant ED access to staff and visitors, being asked to furnish conveniences to visitors (eg, a cup of water or a blanket) 13.2 times per shift, responding to patient care–related requirements (eg, new patient arrivals and leaving the triage area for missing supplies) 8.1 times per shift, and responding to patients or family who ask, “How much longer?” 7.1 times per shift. 5 Recurrent interruptions have been shown to interfere with triage tasks that may affect patient care. 5,17 Methods
ED INTERRUPTIONS
DESIGN
The release of the Institute of Medicine report “HospitalBased Emergency Care: At the Breaking Point” alerted the public that interruptions were one of the challenges contributing to the struggle to provide high-quality emergency care to patients. 15 Interruptions have been shown to occur more often in emergency care than in other health care settings. 6,7,13 Additionally, interruptions of ED providers have been linked to both errors and delays in patient care. 12,16 The results of a prospective time-and-motion study showed that emergency physicians were interrupted 6.6 times per hour and that the interruptions were associated with a significant increase in the time required to complete tasks such as writing orders, dictating notes, or assessing patients; providers failed to return to the original task 18.5% of the time. 10 Another study reported that physicians and nurses failed to return to task once interrupted 13% of the time. 3 Generalizing the results of these studies is difficult, primarily because the categorization of interruptions and the target of observation varied. It becomes clear that reducing medical errors, as well as improving the efficiency and quality of care, can be accomplished by tackling the underlying causes of interruptions. 13 Unfortunately, few prior studies on interruptions have considered the importance of the triage process. Given that triage begins the sequence of clinical care events,
We used a 2-phase, sequential exploratory mixed methods design 18 to investigate the impact of interruptions on the triage assessment process. Mixed methods designs allow researchers to gain dynamic insight into phenomena by combining the strengths of quantitative and qualitative methods. Our study included a quantitative arm in which we used an observational immersive simulation design (Phase I) followed by a qualitative arm in which we used a qualitative descriptive approach (Phase II). In phase I, study participants conducted simulated, video-recorded triage assessments on standardized patients using 2 scripted scenarios. One scenario included interruptions at predetermined time points, whereas the other scenario was uninterrupted. This design allowed us to observe how select interruptions specifically affect triage accuracy and determine how participants reacted to these interruptions as they occurred. Although participants were aware that the patients were actors, they were given no details about the scripts, including assigned diagnoses. In phase II, we conducted video-simulated recall interviews (VSRIs) in which the participants were shown their simulated triage videos and were interviewed to explore their experiences and decision making during the 2 scenarios. VSRI has been shown to produce useful and insightful data for examining the way people experience a specific event. 19 We received
436
JOURNAL OF EMERGENCY NURSING
VOLUME 43 • ISSUE 5
September 2017
Johnson and Alhaj-Ali/RESEARCH
TABLE 1
Description of interruptions per scenario Simulation scenario
Interrupted Male, chest pain (ESI 2) Male, abdominal pain (ESI 3) Uninterrupted Female, pulmonary embolism (ESI 2) Female, animal bite (ESI 4)
Times scenario used
No. of interruptions, mean (SD)
Type of interruption
Triage length, min, mean (SD)
% Accuracy of ESI
5
1.20 (0.447)
12.30 (1.599)
100
4
1.75 (0.500)
Family’s cell phone (2) Coworker (4) Family’s cell phone (3) Coworker (4)
9.96 (2.167)
75
3 6
0 0
8.27 (3.566) 8.7 (2.027)
100 50
n/a n/a
ESI, Emergency Severity Index; SD, standard deviation.
Institutional Review Board approval from the first author’s academic institution prior to initiation of study procedures.
triage room, and 2 ED treatment bays with a cot, chair, and medical diagnostic equipment.
SAMPLE AND SETTING
PROCEDURES
We recruited emergency nurses trained in triage assessment from hospitals within a Midwestern city, including: (1) a level I adult academic trauma center, (2) a suburban level III trauma center, and (3) an urban hospital without a trauma designation. Participants were eligible for the study if they were an emergency nurse with at least 2 years of clinical experience assessing patient acuity using the Emergency Severity Index (ESI). We planned to recruit 20 nurses using E-mail, flyers, presentations at ED staff meetings, and discussions at shift change meetings. Recruitment materials included an explanation of the scope and purpose of the project, anticipated time commitment, and participant payment information.
Standardized Patient Training
Standardized Patients We partnered with actors who served as associates in the study. Standardized patient (SP) simulation includes the use of trained individuals to act as patients, family members, or others, thus allowing students and practitioners to practice history taking, physical examinations, and communication and diagnostic skills. Standardized patient simulation has been used extensively in evaluating clinical applications of knowledge and improving skill development. 18–20 The simulation experiences occurred in a simulation center maintained in the College of Medicine at our university. The center replicated the ED structure, including the patient waiting area, an admission area,
September 2017
VOLUME 43 • ISSUE 5
SPs were hired from the College of Medicine’s SP pool to simulate ED patients in an ED waiting room. The SP profiles were created from the Agency for Healthcare Research and Quality’s sample case studies, 19 which were part of actual emergency care situations, and thus were used to have a positive impact on the realism. Each of the 4 created scenarios included details from one practice case, such as symptoms and vital signs. The SPs were instructed regarding personal history details that were acceptable to add to their story and statements to avoid that would modify their acuity scores. Half of the scenarios included scripted interruptions that occurred at specified times during the triage process (see Table 1 for a list of sample scenarios). Each SP was given 2 different scenarios to perform. During the training session, a research team member played the role of triage nurse. This 2-hour training prior to the arrival of the participants allowed the principal investigator (PI) to answer SP questions, provide instruction/guidance on roles and scripts, and make changes to the scripts for clarity and uniformity. Research team members, unaware of the ESI level of scenarios, participated in the training and assigned each scenario an ESI score with 100% agreement to the Agency for Healthcare Research and Quality ESI assignment. Training continued until the SPs could perform consistently for 2 of the researchers. Additionally, the training triage interviews were observed
WWW.JENONLINE.ORG
437
RESEARCH/Johnson and Alhaj-Ali
by the PI while they occurred, and the SPs were then given feedback and re-educated when deviations occurred. Simulated Triage Interviews When participants arrived at the simulation laboratory, a study team member reviewed the informed consent document with them, answered all their questions, and obtained their signatures prior to beginning the study. Next, each participant received a $100 gift card and completed a demographic survey. Participants were oriented to the ED environment and the chart they would be expected to complete during the simulated triage process. The participants were each given 30 minutes to review a blank ED chart that would be used to complete the triage assessments. Participants were instructed to complete as much of the chart as necessary to determine the patient’s acuity during the simulation. To simulate the need for quick, accurate triage, participants were informed that their triage interviews would need to be completed within 15 minutes. When they had familiarized themselves with the charting, the participants were escorted to a “Triage Room” and instructed to call the patient from the waiting room when ready. Simulations took place over 2 days, with 9 participants completing the scenarios. Each participant was randomly assigned to 2 simulated scenarios, one interrupted and one uninterrupted, with the order of the scenarios randomly assigned. Each scenario was video recorded with audio for use in phase II. Video-Stimulated Recall Interviews Immediately after the second scenario, each participant watched the video-recorded simulation of their interaction with the SP in its entirety. Then participants met with a study team member and watched the recording again. It was during this viewing that the researcher stopped the recording after each interruption episode to ask the participant about the decision-making process. All interviews were recorded and transcribed verbatim. Accuracy of the transcripts was assessed by 2 independent researchers. Transcripts were then reviewed by 3 researchers and coded. MATERIALS
Demographic Questionnaire A short demographic questionnaire was used to collect information about each participant’s education and experience in nursing, emergency nursing, and triage using the ESI for triage categorization.
438
JOURNAL OF EMERGENCY NURSING
ESI The ESI, a well-established psychometrically sound tool, 20 is used in the triage process to categorize ED patients. The ESI focuses on patient acuity and resources needed in treatment. 17 ESI levels range from 5—the least urgent level with minimal resources needed—to 1, the most urgent level, with patients requiring interventions that entail use of multiple resources. The ESI was used in this project to assign patient acuity during the simulation experiences. All study participants were familiar with using the ESI because of their experience in their normal work setting. Triage Chart After the simulations were completed, the charts were independently reviewed for completeness by the research team. Charts included (1) screening questions common to our local emergency departments (eg, Ebola, falls, and safety); (2) vital signs; (3) allergies; (4) acuity (ESI) assignment; (5) medications; (6) surgical and medical history; and (7) focused system assessments. Qualitative Interview Guide The research team, in consultation with on-site simulation experts and experienced qualitative researchers, developed open-ended interview questions to assess the impact of interruptions on the participants’ decision-making process and to evaluate their opinions of their simulation experience. Interview questions included: (1) What were your thoughts of doing these simulated activities with standardized patients? (2) What were your thoughts at the time when the interrupting person approached you? (3) Why did you feel that you responded this way? (4) How did you refocus on the patient after the interruption occurred? (5) How big of an impact did this interruption have on your triage effectiveness? (6) Was this simulation realistic? (7) How would you improve the simulation experience? DATA ANALYSIS
Phase 1 Data Analysis We used descriptive and univariate statistics to clean the data and test whether our data met assumptions. Frequency distributions were created to understand sample demographics, triage duration and length, and the number and length of interruptions experienced across the sample. We analyzed the associations between work experience on triage duration in both interrupted and uninterrupted scenarios using Pearson correlation matrices and Point biserial statistics. We ran χ 2 analyses to determine the impact of
VOLUME 43 • ISSUE 5
September 2017
Johnson and Alhaj-Ali/RESEARCH
TABLE 2
Correlations between triage duration and participant characteristics Years as nurse
Triage time Years as nurse Years in ED Highest nursing degree
0.253
Years in ED
Highest nursing degree
No. of interruptions
0.490 * 0.913 **
–0.598 ** 0.010 –0.294
0.352 0.030 –0.009 .085
ED, Emergency department. * P b .05. ** P b .01.
the number of interruptions, hospital, and highest nursing degree on triage accuracy. Finally, we conducted t-tests to determine whether triage durations between interrupted and uninterrupted scenarios were significantly different. Phase 2 Data Analysis Guided by Colaizzi’s 7 steps in phenomenological data analysis, 3 research team members independently analyzed each transcript. First, the team members read the transcripts in their entirety multiple times to gain an overall sense of the entire content of the interview. The researchers coded each transcript line by line, extracting significant statements and recording these statements and their corresponding locations within the interviews in a separate document. Each team member recorded his or her interpretation of the meanings of these statements and clustered the meanings into preliminary categories. The team then met to discuss their findings and determine agreement of the initial coding scheme between researchers. The protocol for disagreement was to thoroughly discuss inconsistencies between team members to reach consensus. In the event of an unresolved disagreement, the PI would make the final decision regarding the preliminary theme for further analysis (this problem did not occur in this analysis). The researchers reanalyzed the data using the developed preliminary coding scheme and elevated the preliminary categories into themes. The analysts met again to formulate a description and fundamental structure of the phenomenon “experience of decision making in the presence of triage interruptions.” Once consensus had been reached among the analysts, the researchers validated their findings by talking with participants to receive feedback on the accuracy of their description. The coding schema, description of the themes, and representative exemplars were provided to 3 of the study participants (triage nurses) to assess their validity prior to finalization of the study findings.
September 2017
VOLUME 43 • ISSUE 5
Results DESCRIPTIVE ANALYSIS
Although 12 participants were recruited, 3 did not attend. The 9 participants were all female and came from 3 local hospitals; 2 locations were represented by 4 participants each, and 1 participant came from the third hospital. Almost half (4/9) of the participants were certified emergency nurses. Six participants reported their highest degree as a bachelor of science in nursing, 2 held diploma or associate degrees, and 1 participant possessed a doctoral degree. The years of nursing experience in nursing varied from 4 to 40 years (mean, 17.8; median, 14.0; interquartile range [IQR], 4.75-30.5), while the years of ED experience ranged from 3 to 36 years (mean, 13.5; median, 9.5; IQR, 4.75-17.0). The participants completed 75.0% of the provided triage documentation. However, focused assessments were only completed in 45.8% of scenarios. The patient’s name, height, weight, pain, and vital signs were recorded for all the patients with 100% accuracy. Both a medication list and medical history were recorded in 16 of the 18 scenarios. Surgical history, smoking quit date, falls risk assessment, and alcohol intake were addressed in 94.4% of scenarios. The video-stimulated recall interviews, which included the time to watch the interruption occur, varied in length from 15 to 27 minutes, with the mean being 21.8 minutes. The average interruption lasted 21.77 seconds and ranged from 7 seconds to 49 seconds. Table 1 shows the details of each scenario. Phase I Triage duration: The triage duration for all scenarios was normally distributed with a mean of 9.93 minutes (standard deviation [SD], 2.59), with a range between 4.34 and 13.45 minutes. Triage duration was significantly longer in the
WWW.JENONLINE.ORG
439
RESEARCH/Johnson and Alhaj-Ali
interrupted scenarios (t = –2.475, P = .025), with uninterrupted scenarios having a mean of 8.60 minutes (SD, 2.410), whereas interrupted scenarios had a mean of 11.25 minutes (SD, 2.137). No significant difference was found in the triage duration (t = –.107, P = .211) between scenarios with a correct acuity score with a mean of 9.97 minutes (SD, 2.88) and an incorrect ESI with a mean of 9.80 minutes (SD, 1.515). Correlations between triage duration and both ED and general nursing experience can be found in Table 2. Although triage durations were longer for participants who possessed more ED experience, higher educational degrees were associated with shorter triage duration. Triage accuracy: The participants’ acuity assessments (ESI) were correct in 77% of cases. Of the 18 scenarios (2 scenarios per participant), 3 had incorrect ESI scores and 1 had a missing acuity score. The correct ESI was assigned for every interrupted scenario except for one instance where no ESI was documented. Conversely, of the 9 uninterrupted scenarios, there were 3 incidents of incorrect ESI assignment. The 3 cases of incorrect ESI scores were all for the same uninterrupted scenario and were all assigned a more acute score by nurses with more than 15 years of ED experience each. Because 3 uninterrupted scenarios were assigned incorrect acuity scores, further investigation was warranted. To ensure that the researchers set the correct ESI value for the scenario, 2 outside reviewers watched all of the scenario’s videos and then provided an ESI score based on the SP’s performance. In all cases, the scenario was scored as a 4, which was the expected ESI score. The χ 2 analysis revealed no relationship between triage accuracy and the participants’ place of employment χ 2 (2, N = 9) = 1.36, P = .505; the presence of interruptions χ 2 (1, N = 18) = 1.286, P = .257; the number of interruptions χ 2 (2, N = 18) = 2.089, P = .352; or the highest educational degree of the participant χ 2 (2, N = 9) = 4.018, P = .134. Phase II The qualitative analysis of the interviews revealed many themes concerning how the participants dealt with the interruptions and consequently how they made decisions during that time. These themes were: (1) control, (2) common, (3) nature of the interruption, (4) refocus, and (5) rationale for incorrect ESI assignment. Control: One third of the participants reported the “need to control” the triage environment. The participants reported that interruptions have a “big impact because they take the control away from the nurse but also that the interruptions need to be controlled by the nurse…. Interruptions impacted effectiveness because it decreases the nurse’s control of the situation” (Participant 8). One participant stated that
440
JOURNAL OF EMERGENCY NURSING
interruptions “could have a big impact on triage because it disturbs the flow,” thus leading to loss of concentration and the potential to miss important patient information. Common: All of the participants reported that interruptions are common in the triage environment. One participant reported that interruptions did not bother her because she had “a mindset of ER. You know how to deal with the situation” (Participant 3). A second participant said, “These interruptions are a common problem in the ER” (Participant 5). Another remarked that it is “just sort of part of the usual triage, like, usual experience” (Participant 6). Another echoed the sentiment by stating, “I’m really used to being interrupted” (Participant 4). Interruption nature: Patient care–related interruptions are less offensive than noncare- related interruptions. Although not all of the participants reported having trouble getting back on track, all of the participants who experienced non-patient care–related interruptions reported needing to take time to refocus on the patient they were triaging. Participants reported that they felt annoyed and wanted to say, “I have stuff to do!! Get out of my way!!” (Participant 8). Most of the persons who participated in the scenario in which the “registration clerk” enters to verify the right chart stated that this interruption was not disruptive because “she was relevant with the situation.” Refocus: Most participants reported using the chart as a prompt for refocusing. However, the participants who reported being annoyed with the interruption said that they had difficulty refocusing because they lost their place in the chart. Another participant reported that the ability to refocus “depends on person’s reaction, like, how they are acting. If they are cooperative, it’s easier” (Participant 6). Two other participants reported that they have a normal order of triage they follow. Rationale for incorrect ESI assignments: Some ESI assignment decisions were influenced by emotional reactions to the patient. The 3 participants who assigned inaccurate ESI scores explained that they “bumped up” the acuity score because of the patient’s anxiety level (Participant 4 and Participant 7). The one participant who forgot to assign an ESI during her interrupted scenario reported that she forgot because she was “thrown off track” (Participant 3). Simulation experience: The participants were asked to evaluate the usefulness of this simulation experience for assessing triage accuracy. All participants reported that the simulations were realistic and that the SPs were believable. Four participants provided recommendations for improvement. They stated that having the vital signs already available to them was disruptive to their assessment process. Two of those participants reported the collection of vital signs as an important time for patient assessment and observation. One
VOLUME 43 • ISSUE 5
September 2017
Johnson and Alhaj-Ali/RESEARCH
participant (#8) explained that she “uses vitals to get to know the patient.” Two participants reported that using the paper charting method was not ideal. One participant reported that only having one interruption during a triage interview was unrealistic and that we should “have more interruptions for the next time” (Participant 1). Discussion
This study is the first to assess the practicality of using SPs to test how participants respond to interruptions during triage. Previous work has shown that simulation, both computer-based and patient-based simulators (eg, SIM MAN), has been used successfully to train nurses. 21 Although using SPs is not a new trend in health research, this study is the first to use SPs to explore triage accuracy in relation to interruptions. 22–24 TRIAGE DURATION
Longer triage durations were associated with more ED experience. Gerdtz and Bucknall 25 reported that assigning acuity during triage averaged 3.36 minutes, with noninterrupted triage lasting 3.47 minutes and interrupted triage lasting 4.05 minutes, which is significantly lower than the 8.6 and 11.3 minutes, respectively, in this study. However, the difference may be due to the data collection methods. In our scenarios, the participant was required to retrieve the SP from the waiting room, and our participants were tasked with completing the triage assessment, which included more than assigning acuity scores. Higher degrees were associated with shorter triage durations, which may be indicative of the role of experience, coupled with an ability to focus on the assessment’s important aspects and skip the screenings that are often added into the triage assessment. However, we were unable to assess mediating, moderating, and multicollinear relationships for these variables because of the small sample size. One explanation for this finding is that a more experienced nurse becomes more confident in performing the necessary tasks and therefore required less time to make a triage decision compared with more novice nurses. 26 TRIAGE ACUITY
The correct ESI was assigned for every interruption scenario except one, in which no ESI was documented. Conversely, the uninterrupted scenario had 3 incidents of incorrect ESI assignment. The incorrect ESI scores were all for the same uninterrupted scenario, and in each instance, the participant gave a higher acuity score. The ESI for the uninterrupted scenario was a 4, but 2 participants assigned it an ESI of 3,
September 2017
VOLUME 43 • ISSUE 5
and another participant assigned it an ESI of 2. Upon discussion with the participants, each reported that she would have assigned an ESI of 4 except for the anxiety that the woman displayed. Because of the patient’s anxiety, they made the decision to assign a more acute ESI score. This decision may be explained by connecting a judgment confidence with perceived difficulty of that judgment. Previous studies have shown overconfidence with difficult judgment and underconfidence with easy judgment. 27–29 Interestingly, most of the incorrect ESI assignments occurred in the uninterrupted scenarios. More experienced participants were more likely to assign an incorrect ESI, but participants consistently reported that interruptions and distractions had less of an impact on more experienced nurses. Retraining in ESI algorithms may be useful in the more experienced nurses, and enhanced education added to quality improvement projects might improve overall accuracy rates. 30–32 Providing this training in a simulation laboratory using SPs may be an appropriate method to refresh ESI protocols for more experienced nurses. Previous research on overtriaging (assigning a higher acuity) is mixed. Although our findings of overtriage coincides with Kilner’s findings 25 of the nurse’s tendency to overtriage, it contradicts the work of Considine et al 33 and Platts-Mills et al, 11 who reported that ED nurses were more likely to undertriage (ie, assign a lower acuity to) older patients. Although in our study the triage acuity was incorrect in 22% of scenarios, no significant difference was found between the interrupted and uninterrupted scenarios. Again, this finding may be due to the small sample size. When considering all of the scenarios, interrupted and uninterrupted, the participants in this study correctly assigned an ESI 78% of the time, whereas previous research 34,35 places accurate acuity rates much lower, at 58% and 56%. It is unclear why our study had a higher accuracy. Perhaps the simulated environment affected the amount of time the participants spent assessing the patient. Participants did not have the pressure of a full waiting room, but we did place a 15-minute limit on the scenario, thus mimicking the time pressure of triage in reality; the participants were aware of this stipulation before beginning. Working under pressure (timed simulation scenario) tends to increase the level of urgency that led to increased time spent in completing the task. 27 DECISION MAKING
It is important for triage nurses to make accurate acuity assessments. To accomplish this task, they need to be able to control their environment. When controlling the environment is not possible because of circumstances such as
WWW.JENONLINE.ORG
441
RESEARCH/Johnson and Alhaj-Ali
interruptions, the nurse needs to be able to refocus on the patient they are triaging. Refocusing may be easier to do when an interruption is based on patient care rather than other issues. The participants who were interrupted for non-patient care–related issues reported feeling annoyed in those instances, and it was during one such occasion that an acuity score was not recorded for a patient in this study. Similarly, Berg et al 3 found that interruptions are handled differently based on whether they are viewed as disruptive or nondisruptive on the basis of the nature of interruption and how relative it is to patient care. As in a previous study, 5 nurses reported that interruptions are common in the emergency department. The participants in our study stated that they are accustomed to dealing with them. Because interruptions are ubiquitous to the emergency department, we must be able to identify which interruptions are problematic and how they affect the care of patients. Other factors that affect nurse triage decisions that may lead to errors should be explored as well. Some of these factors include ED crowding level, ED nurse shortage, patient anxiety level, family member presence, and nurses’ personal factors. Because triage can be affected by so many outside influences, we must determine which factors have the biggest impact on patient care. The ramifications of interruptions on the triage process require a comprehensive assessment and evaluation to design interventions to effectively assist the nurse in managing interruptions. Simulated environments have been used successfully to improve confidence in triage training 21 and to assist in designing interventions to decrease the effects of interruptions on patient care. 13
Limitations
This study enabled a comparison of interrupted and uninterrupted simulation scenarios and their effects on triage duration and accuracy. However, a larger sample is required to determine if the errors and missed acuity assignments are generalizable to a larger population. We did not simulate a realistic arrival time between new ED patients, so the accurate pressures of triage may not have been fully realized. Although we purposely had the participants use paper charts instead of computerized point-and-click charting, participants who were familiar with paper charting may have had an advantage when reverting back to this charting method. The use of paper charts also may have contributed to the lower rate of chart completion. 36 Further research designed to examine the effect of ED interruptions on triage duration, accuracy, and decision making in different situations using larger samples and more rigorous data collection strategies is also needed. However, this project lays the foundation for designing those studies.
442
JOURNAL OF EMERGENCY NURSING
Implications for Emergency Nursing
This study forms a foundation for understanding how ED interruptions affect triage nurses and how quality of triage care affects patient safety and quality of care. The results of this study support the importance of using simulation as a learning strategy in nursing. Practicing in a simulated setting might be a good way for new triage nurses to hone their skills at managing interruptions. Practicing in this manner also provides veteran triage nurses with ways to recognize what is affecting the accuracy of their triage. This information will aid researchers, educators, and clinicians in evaluating competencies using simulation, as well as collaboratively designing interventions that may decrease interruptions and their effects. Furthermore, regularly incorporating triage simulations into nurses’ training or retraining could improve ESI assignment accuracy and patient care. 31
Conclusion
Interruptions in the emergency department have been shown to lead to errors and delays. Errors and delays degrade the quality of care 1 while adversely affecting outcomes. The impact of interruptions is compounded in triage, where errors and delays during this time propagate through the patient’s entire care episode. The findings of this study will be the basis for future work that examines how nurses successfully manage interruptions. Additionally, it tests interventions to assist triage nurses in managing or reducing interruptions during the important patient assessment process. Triage is the important initial assessment in which acuity is assigned to a patient. Interruptions during this process can cause delays and interfere with providing safe and efficient patient care. Interruptions during this time increase the process time for triage and can cause system delays to the ED process. Additionally, interruptions can lead to errors in assessment and documentation. Therefore, the necessity to identify potentially harmful interruptions and develop interventions to mitigate the adverse impact of these interruptions becomes apparent, thus allowing us to decrease errors and delays, which will lead to improving patient outcomes and ED process time. REFERENCES 1. Raban MZ, Walter S, Douglas H, et al. Measuring the relationship between interruptions, multitasking and prescribing errors in an emergency department: a study protocol. BMJ Open. http://bmjopen. bmj.com/content/5/10/e009076. Published October 2015. Accessed May 1, 2017.
VOLUME 43 • ISSUE 5
September 2017
Johnson and Alhaj-Ali/RESEARCH
2. Westbrook JI, Woods A, Rob MI, et al. Association of interruptions with an increased risk and severity of medication administration errors. Arch Intern Med. 2010;170(8):683-690. 3. Berg LM, Kallberg AS, Goransson KE, et al. Interruptions in emergency department work: an observational and interview study. BMJ Qual Saf. 2013;22(8):656-663. 4. Kosits LM, Jones K. Interruptions experienced by registered nurses working in the emergency department. J Emerg Nurs. 2011;37(1):3-8. 5. Johnson KD, Motavalli M, Gray D, et al. Causes and occurrences of interruptions during ED triage. J Emerg Nurs. 2014;40(5):434-439. 6. Cole G, Stefanus D, Gardner H, et al. The impact of interruptions on the duration of nursing interventions: a direct observation study in an academic emergency department. BMJ Qual Saf. 2016;25(6):457-465. 7. Jeanmonod R, Boyd M, Loewenthal M, et al. The nature of emergency department interruptions and their impact on patient satisfaction. Emerg Med J. 2010;27(5):376-379. 8. Centers for Disease Control and Prevention, National Center for Health Statistics. Emergency department visits. 2013. https://www.cdc.gov/ nchs/fastats/emergency-department.htm. Updated March 17, 2017. Accessed May 1, 2017. 9. Chisholm CD, Dornfeld AM, Nelson DR, et al. Work interrupted: a comparison of workplace interruptions in emergency departments and primary care offices. Ann Emerg Med. 2001;38(2):146-151. 10. Grossmann FF, Zumbrunn T, Cirprian S, et al. At risk of undertriage? Testing the performance and accuracy of the emergency severity index in older emergency department patients. Ann Emerg Med. 2012;60(3):317-325. 11. Platts-Mills TF, Travers D, Biese K, et al. Accuracy of the Emergency Severity Index triage instrument for identifying elder emergency department patients receiving an immediate life-saving intervention. Acad Emerg Med. 2010;17(3):238-243. 12. Burley D. Better communication in the emergency department: Duncan Burley outlines results of a literature review on how practitioners can improve history taking by adopting a more patient-centered approach. Emerg Nurse. 2011;19(2):32-36. 13. Werner NE, Holden RJ. Interruptions in the wild: development of a sociotechnical systems model of interruptions in the emergency department through a systematic review. Appl Ergon. 2015;51:244-254. 14. Wears RL, Woloshynowych M, Brown R, et al. Reflective analysis of safety research in the hospital accident & emergency departments. Appl Ergon. 2010;41(5):695-700. 15. Warden G, Altman S, Asplin B, et al. Hospital-Based Emergency Care: At the Breaking Point. Washington, DC: National Academies Press; 2007. 16. Hughes R. Patient Safety and Quality: An Evidence-Based Handbook for Nurses. Rockville, MD: Agency for Healthcare Research and Quality; 2008. 17. Aacharya RP, Gastmans C, Denier Y. Emergency department triage: an ethical analysis. BMC Emerg Med. 2011;11:16. 18. Creswell JW. Qualitative Inquiry and Research Design: Choosing Among Five Approaches. Thousand Oaks, CA: Sage Publications; 2012. 19. Agency for Healthcare Research and Quality. Emergency Severity Index (ESI): A triage tool for emergency department. https://www.ahrq.gov/
September 2017
VOLUME 43 • ISSUE 5
professionals/systems/hospital/esi/esi9.html. Reviewed October 2014. Accessed May 1, 2017. 20. Tanabe P, Gimble R, Yarnold PR, et al. Reliability and validity of scores on the Emergency Severity Index version 3. Acad Emerg Med. 2004;11(1):59-65. 21. Wolf L. The use of human patient simulation in ED triage training can improve nursing confidence and patient outcomes. J Emerg Nurs. 2008;34(2):169-171. 22. Gates DM, Fitzwater E, Telintelo S. Using simulations and standardized patients in intervention research. Clin Nurs Res. 2001;10(4):387-400. 23. Geurtzen R, Hogeveen M, Rajani AK, et al. Using simulation to study difficult clinical issues: prenatal counseling at the threshold of viability across American and Dutch cultures. Simul Healthc. 2014;9(3):167-173. 24. Rogers ES, Gilespie C, Zabar S, et al. Using standardized patients to train telephone counselors for a clinical trial. BMC Res Notes. 2015;7:341. 25. Gerdtz MF, Bucknall TK. Triage nurses' clinical decision making. An observational study of urgency assessment. J Adv Nurs. 2001;35(4):550-561. 26. Dallaire C, Poitras J, Aubin K, et al. Emergency department triage: do experienced nurses agree on triage scores? J Emerg Med. 2012;42(6):736-740. 27. Yang H, Thompson C, Bland M. The effect of clinical experience, judgment task difficulty and time pressure on nurses' confidence calibration in a high fidelity clinical simulation. BMC Med Inform Decis Mak. 2012;12:113. 28. Yang H, Thompson C. Nurses' risk assessment judgements: a confidence calibration study. J Adv Nurs. 2012;66(12):2751-2760. 29. Friedman CP, Gatti GG, Franz TM, et al. Do physicians know when their diagnoses are correct? Implications for decision support and error reduction. J Gen Intern Med. 2005;20(4):334-339. 30. Travers DA, Waller AE, Katznelson J, et al. Reliability and validity of the emergency severity index for pediatric triage. Acad Emerg Med. 2009;16(9):843-849. 31. van der Wulp I, Schrijvers AJ, van Stel HF. Predicting admission and mortality with the Emergency Severity Index and the Manchester Triage System: a retrospective observational study. Emerg Med J. 2009;26(7):506-509. 32. Dalawari P, Sanning J, Pan D, et al. Emergency severity index version 4 during the first year of implementation at an academic institution. J Hosp Admin. 2015;5(2):35. 33. Considine J, Levasseur SA, Villanueva E. The Australasian Triage Scale: examining emergency department nurses' performance using computer and paper scenarios. Ann Emerg Med. 2004;44(5):516-523. 34. Göransson KE, Ehrenberg A, Ehnfors M. Triage in emergency departments: national survey. J Clin Nurs. 2005;14(9):1067-1074. 35. Chen SS, Chen JC, Ng CJ, et al. Factors that influence the accuracy of triage nurses' judgement in emergency departments. Emerg Med J. 2010;27(6):451-455. 36. Benda NC, Meadors ML, Hettinger AZ, et al. Electronic health records and interruptions: the need for new interruption management strategies. Proc Int Symp Hum Fact Ergon Health Care. 2015;4(1):70.
WWW.JENONLINE.ORG
443
USING SIMULATION TO ASSESS THE IMPACT OF TRIAGE INTERRUPTIONS Authors: Kimberly D. Johnson, PhD, RN, CEN, and Abeer Alhaj-Ali, BSN, Cincinnati, OH
Earn Up to 5.5 CE Hours. See page 489.
Contribution to Emergency Nursing Practice
• This research is innovative because simulation has not previously been used to assess the impact of interruptions on triage. • Emergency nurse educators could use the methods and results discussed in this project to guide departmental training of triage nurses. • Practicing in a simulated setting can help new triage nurses hone their skills at managing interruptions while providing veteran triage nurses with ways to recognize factors that are affecting the accuracy of their triage. Abstract Introduction: Interruptions are common in the emergency
was used to collect information about the nurses’ education and experience. The Emergency Severity Index (ESI) was used for triage categorization. Each participant completed 2 scenarios (one interrupted and one uninterrupted). After completion of the scenarios, video-simulated recall interviews were used to assess the simulation experience and the impact that interruptions had on the triage decision-making process. Results: Triage time had a mean of 10 minutes and ranged between 4.34 minutes and 13.45 minutes. However, triage was significantly longer during the interrupted scenarios. Seventy-seven percent of the acuity assessments (ESI) were correct. Of the 18 scenarios, 3 uninterrupted scenarios had incorrect ESI scores, and one interrupted scenario had a missing acuity score.
department and contribute to catastrophic errors. Care priorities and acuity levels are assigned during triage, meaning that mistakes and omissions during the triage process could have detrimental effects on patients. The purpose of this project was to assess the feasibility of investigating the impact of interruptions on triage and the decision-making process in a simulated setting.
Discussion: This study provides the basis for future work that looks at how nurses successfully manage interruptions and tests interventions to assist triage nurses in managing or reducing interruptions during this important patient assessment process.
Methods: A 2-phase, sequential exploratory mixed method design was used. Nine nurses from 3 emergency departments in a Midwest area participated. A short demographic questionnaire
Key words: Triage; Interruptions; Distractions; Emergency department; Triage nurse role; Simulation; Video-simulated recall interviews
lmost 100,000 deaths per year are attributed to health care–related errors. 1,2 Interruptions have been identified as a cause of errors and can be
detrimental to patient safety. 1,2 Frequently the care provided during ED visits is interrupted. 3,4 One particular area that is fraught with interruptions is the triage process, and thus this part of the patient’s ED experience is particularly vulnerable to errors. 5–7 With annual ED visits increasing to more than 130 million per year, 8 triage nurses are pressured to quickly and accurately assess each patient. Triage assessment, defined as assigning acuity to patients to determine treatment priority, is the critical beginning of the treatment plan for ED patients. The initial triage assessment can affect both the ED visit quality and length. 5 Because of the nature of the emergency department, nurses are exposed to frequent interruptions, which may lead to inaccuracies in acuity assessment, thus resulting in errors that may affect the quality of care. 6 Although errors and delays adversely affect patient outcomes, 9
A
Kimberly D. Johnson, Member, Greater Cincinnati Chapter, is Assistant Professor, College of Nursing, University of Cincinnati, Cincinnati, OH. Abeer Alhaj-Ali is a PhD student, University of Cincinnati, Cincinnati, OH. This work was supported by a ENA Foundation/Sigma Theta Tau International Research Grant. For correspondence, write: Kimberly D. Johnson, PhD, RN, CEN, University of Cincinnati, 3110 Vine St, Cincinnati, OH 45221; E-mail: [email protected]. J Emerg Nurs 2017;43:435-43. 0099-1767 Copyright © 2017 Emergency Nurses Association. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jen.2017.04.008
September 2017
VOLUME 43 • ISSUE 5
WWW.JENONLINE.ORG
435
RESEARCH/Johnson and Alhaj-Ali
the true impact of triage interruptions on patients’ clinical outcomes is unknown because few studies have investigated this phenomenon. The purpose of this research was to determine the feasibility of using a simulated setting to investigate the impact of interruptions and how triage nurses make data-based decisions in the presence of interruptions.
sets priorities for the treatment team, and also begins the health care customer experience, it stands to reason that the triage process should be a primary target for intervention. To best design an improvement intervention, understanding the effect of triage interruptions on the quality of health care is vital. 5,17
Background and Significance
TRIAGE INTERRUPTIONS
The triage nurse’s assessment of a patient is an important first step in an episode of care and can be an indicator of how the patient’s ED experience will progress. 5 Interruptions may lead to an incorrect triage decision, missed symptom identification, incomplete assessment, or unasked questions. Any of these factors could potentially delay care, resulting in significant morbidity or mortality. 10,11 A patient seeking treatment may have a triage interview interrupted for myriad reasons that include addressing the needs of other patients, visitors, or staff. The interruptions then create delays in getting patients into a treatment area, keep nurses from collecting appropriate triage data, or cause nurses to make poor or erroneous triage decisions. 12–14 Errors during triage can decrease quality of care and have an adverse impact on patient outcomes. 1,9,10
Previous work by Johnson et al 5 identified the types and frequency of interruptions that occur during the triage process. These investigators reported that triage nurses were interrupted an average of 16.6 times per shift for a variety of reasons; top occurrences included being asked to grant ED access to staff and visitors, being asked to furnish conveniences to visitors (eg, a cup of water or a blanket) 13.2 times per shift, responding to patient care–related requirements (eg, new patient arrivals and leaving the triage area for missing supplies) 8.1 times per shift, and responding to patients or family who ask, “How much longer?” 7.1 times per shift. 5 Recurrent interruptions have been shown to interfere with triage tasks that may affect patient care. 5,17 Methods
ED INTERRUPTIONS
DESIGN
The release of the Institute of Medicine report “HospitalBased Emergency Care: At the Breaking Point” alerted the public that interruptions were one of the challenges contributing to the struggle to provide high-quality emergency care to patients. 15 Interruptions have been shown to occur more often in emergency care than in other health care settings. 6,7,13 Additionally, interruptions of ED providers have been linked to both errors and delays in patient care. 12,16 The results of a prospective time-and-motion study showed that emergency physicians were interrupted 6.6 times per hour and that the interruptions were associated with a significant increase in the time required to complete tasks such as writing orders, dictating notes, or assessing patients; providers failed to return to the original task 18.5% of the time. 10 Another study reported that physicians and nurses failed to return to task once interrupted 13% of the time. 3 Generalizing the results of these studies is difficult, primarily because the categorization of interruptions and the target of observation varied. It becomes clear that reducing medical errors, as well as improving the efficiency and quality of care, can be accomplished by tackling the underlying causes of interruptions. 13 Unfortunately, few prior studies on interruptions have considered the importance of the triage process. Given that triage begins the sequence of clinical care events,
We used a 2-phase, sequential exploratory mixed methods design 18 to investigate the impact of interruptions on the triage assessment process. Mixed methods designs allow researchers to gain dynamic insight into phenomena by combining the strengths of quantitative and qualitative methods. Our study included a quantitative arm in which we used an observational immersive simulation design (Phase I) followed by a qualitative arm in which we used a qualitative descriptive approach (Phase II). In phase I, study participants conducted simulated, video-recorded triage assessments on standardized patients using 2 scripted scenarios. One scenario included interruptions at predetermined time points, whereas the other scenario was uninterrupted. This design allowed us to observe how select interruptions specifically affect triage accuracy and determine how participants reacted to these interruptions as they occurred. Although participants were aware that the patients were actors, they were given no details about the scripts, including assigned diagnoses. In phase II, we conducted video-simulated recall interviews (VSRIs) in which the participants were shown their simulated triage videos and were interviewed to explore their experiences and decision making during the 2 scenarios. VSRI has been shown to produce useful and insightful data for examining the way people experience a specific event. 19 We received
436
JOURNAL OF EMERGENCY NURSING
VOLUME 43 • ISSUE 5
September 2017
Johnson and Alhaj-Ali/RESEARCH
TABLE 1
Description of interruptions per scenario Simulation scenario
Interrupted Male, chest pain (ESI 2) Male, abdominal pain (ESI 3) Uninterrupted Female, pulmonary embolism (ESI 2) Female, animal bite (ESI 4)
Times scenario used
No. of interruptions, mean (SD)
Type of interruption
Triage length, min, mean (SD)
% Accuracy of ESI
5
1.20 (0.447)
12.30 (1.599)
100
4
1.75 (0.500)
Family’s cell phone (2) Coworker (4) Family’s cell phone (3) Coworker (4)
9.96 (2.167)
75
3 6
0 0
8.27 (3.566) 8.7 (2.027)
100 50
n/a n/a
ESI, Emergency Severity Index; SD, standard deviation.
Institutional Review Board approval from the first author’s academic institution prior to initiation of study procedures.
triage room, and 2 ED treatment bays with a cot, chair, and medical diagnostic equipment.
SAMPLE AND SETTING
PROCEDURES
We recruited emergency nurses trained in triage assessment from hospitals within a Midwestern city, including: (1) a level I adult academic trauma center, (2) a suburban level III trauma center, and (3) an urban hospital without a trauma designation. Participants were eligible for the study if they were an emergency nurse with at least 2 years of clinical experience assessing patient acuity using the Emergency Severity Index (ESI). We planned to recruit 20 nurses using E-mail, flyers, presentations at ED staff meetings, and discussions at shift change meetings. Recruitment materials included an explanation of the scope and purpose of the project, anticipated time commitment, and participant payment information.
Standardized Patient Training
Standardized Patients We partnered with actors who served as associates in the study. Standardized patient (SP) simulation includes the use of trained individuals to act as patients, family members, or others, thus allowing students and practitioners to practice history taking, physical examinations, and communication and diagnostic skills. Standardized patient simulation has been used extensively in evaluating clinical applications of knowledge and improving skill development. 18–20 The simulation experiences occurred in a simulation center maintained in the College of Medicine at our university. The center replicated the ED structure, including the patient waiting area, an admission area,
September 2017
VOLUME 43 • ISSUE 5
SPs were hired from the College of Medicine’s SP pool to simulate ED patients in an ED waiting room. The SP profiles were created from the Agency for Healthcare Research and Quality’s sample case studies, 19 which were part of actual emergency care situations, and thus were used to have a positive impact on the realism. Each of the 4 created scenarios included details from one practice case, such as symptoms and vital signs. The SPs were instructed regarding personal history details that were acceptable to add to their story and statements to avoid that would modify their acuity scores. Half of the scenarios included scripted interruptions that occurred at specified times during the triage process (see Table 1 for a list of sample scenarios). Each SP was given 2 different scenarios to perform. During the training session, a research team member played the role of triage nurse. This 2-hour training prior to the arrival of the participants allowed the principal investigator (PI) to answer SP questions, provide instruction/guidance on roles and scripts, and make changes to the scripts for clarity and uniformity. Research team members, unaware of the ESI level of scenarios, participated in the training and assigned each scenario an ESI score with 100% agreement to the Agency for Healthcare Research and Quality ESI assignment. Training continued until the SPs could perform consistently for 2 of the researchers. Additionally, the training triage interviews were observed
WWW.JENONLINE.ORG
437
RESEARCH/Johnson and Alhaj-Ali
by the PI while they occurred, and the SPs were then given feedback and re-educated when deviations occurred. Simulated Triage Interviews When participants arrived at the simulation laboratory, a study team member reviewed the informed consent document with them, answered all their questions, and obtained their signatures prior to beginning the study. Next, each participant received a $100 gift card and completed a demographic survey. Participants were oriented to the ED environment and the chart they would be expected to complete during the simulated triage process. The participants were each given 30 minutes to review a blank ED chart that would be used to complete the triage assessments. Participants were instructed to complete as much of the chart as necessary to determine the patient’s acuity during the simulation. To simulate the need for quick, accurate triage, participants were informed that their triage interviews would need to be completed within 15 minutes. When they had familiarized themselves with the charting, the participants were escorted to a “Triage Room” and instructed to call the patient from the waiting room when ready. Simulations took place over 2 days, with 9 participants completing the scenarios. Each participant was randomly assigned to 2 simulated scenarios, one interrupted and one uninterrupted, with the order of the scenarios randomly assigned. Each scenario was video recorded with audio for use in phase II. Video-Stimulated Recall Interviews Immediately after the second scenario, each participant watched the video-recorded simulation of their interaction with the SP in its entirety. Then participants met with a study team member and watched the recording again. It was during this viewing that the researcher stopped the recording after each interruption episode to ask the participant about the decision-making process. All interviews were recorded and transcribed verbatim. Accuracy of the transcripts was assessed by 2 independent researchers. Transcripts were then reviewed by 3 researchers and coded. MATERIALS
Demographic Questionnaire A short demographic questionnaire was used to collect information about each participant’s education and experience in nursing, emergency nursing, and triage using the ESI for triage categorization.
438
JOURNAL OF EMERGENCY NURSING
ESI The ESI, a well-established psychometrically sound tool, 20 is used in the triage process to categorize ED patients. The ESI focuses on patient acuity and resources needed in treatment. 17 ESI levels range from 5—the least urgent level with minimal resources needed—to 1, the most urgent level, with patients requiring interventions that entail use of multiple resources. The ESI was used in this project to assign patient acuity during the simulation experiences. All study participants were familiar with using the ESI because of their experience in their normal work setting. Triage Chart After the simulations were completed, the charts were independently reviewed for completeness by the research team. Charts included (1) screening questions common to our local emergency departments (eg, Ebola, falls, and safety); (2) vital signs; (3) allergies; (4) acuity (ESI) assignment; (5) medications; (6) surgical and medical history; and (7) focused system assessments. Qualitative Interview Guide The research team, in consultation with on-site simulation experts and experienced qualitative researchers, developed open-ended interview questions to assess the impact of interruptions on the participants’ decision-making process and to evaluate their opinions of their simulation experience. Interview questions included: (1) What were your thoughts of doing these simulated activities with standardized patients? (2) What were your thoughts at the time when the interrupting person approached you? (3) Why did you feel that you responded this way? (4) How did you refocus on the patient after the interruption occurred? (5) How big of an impact did this interruption have on your triage effectiveness? (6) Was this simulation realistic? (7) How would you improve the simulation experience? DATA ANALYSIS
Phase 1 Data Analysis We used descriptive and univariate statistics to clean the data and test whether our data met assumptions. Frequency distributions were created to understand sample demographics, triage duration and length, and the number and length of interruptions experienced across the sample. We analyzed the associations between work experience on triage duration in both interrupted and uninterrupted scenarios using Pearson correlation matrices and Point biserial statistics. We ran χ 2 analyses to determine the impact of
VOLUME 43 • ISSUE 5
September 2017
Johnson and Alhaj-Ali/RESEARCH
TABLE 2
Correlations between triage duration and participant characteristics Years as nurse
Triage time Years as nurse Years in ED Highest nursing degree
0.253
Years in ED
Highest nursing degree
No. of interruptions
0.490 * 0.913 **
–0.598 ** 0.010 –0.294
0.352 0.030 –0.009 .085
ED, Emergency department. * P b .05. ** P b .01.
the number of interruptions, hospital, and highest nursing degree on triage accuracy. Finally, we conducted t-tests to determine whether triage durations between interrupted and uninterrupted scenarios were significantly different. Phase 2 Data Analysis Guided by Colaizzi’s 7 steps in phenomenological data analysis, 3 research team members independently analyzed each transcript. First, the team members read the transcripts in their entirety multiple times to gain an overall sense of the entire content of the interview. The researchers coded each transcript line by line, extracting significant statements and recording these statements and their corresponding locations within the interviews in a separate document. Each team member recorded his or her interpretation of the meanings of these statements and clustered the meanings into preliminary categories. The team then met to discuss their findings and determine agreement of the initial coding scheme between researchers. The protocol for disagreement was to thoroughly discuss inconsistencies between team members to reach consensus. In the event of an unresolved disagreement, the PI would make the final decision regarding the preliminary theme for further analysis (this problem did not occur in this analysis). The researchers reanalyzed the data using the developed preliminary coding scheme and elevated the preliminary categories into themes. The analysts met again to formulate a description and fundamental structure of the phenomenon “experience of decision making in the presence of triage interruptions.” Once consensus had been reached among the analysts, the researchers validated their findings by talking with participants to receive feedback on the accuracy of their description. The coding schema, description of the themes, and representative exemplars were provided to 3 of the study participants (triage nurses) to assess their validity prior to finalization of the study findings.
September 2017
VOLUME 43 • ISSUE 5
Results DESCRIPTIVE ANALYSIS
Although 12 participants were recruited, 3 did not attend. The 9 participants were all female and came from 3 local hospitals; 2 locations were represented by 4 participants each, and 1 participant came from the third hospital. Almost half (4/9) of the participants were certified emergency nurses. Six participants reported their highest degree as a bachelor of science in nursing, 2 held diploma or associate degrees, and 1 participant possessed a doctoral degree. The years of nursing experience in nursing varied from 4 to 40 years (mean, 17.8; median, 14.0; interquartile range [IQR], 4.75-30.5), while the years of ED experience ranged from 3 to 36 years (mean, 13.5; median, 9.5; IQR, 4.75-17.0). The participants completed 75.0% of the provided triage documentation. However, focused assessments were only completed in 45.8% of scenarios. The patient’s name, height, weight, pain, and vital signs were recorded for all the patients with 100% accuracy. Both a medication list and medical history were recorded in 16 of the 18 scenarios. Surgical history, smoking quit date, falls risk assessment, and alcohol intake were addressed in 94.4% of scenarios. The video-stimulated recall interviews, which included the time to watch the interruption occur, varied in length from 15 to 27 minutes, with the mean being 21.8 minutes. The average interruption lasted 21.77 seconds and ranged from 7 seconds to 49 seconds. Table 1 shows the details of each scenario. Phase I Triage duration: The triage duration for all scenarios was normally distributed with a mean of 9.93 minutes (standard deviation [SD], 2.59), with a range between 4.34 and 13.45 minutes. Triage duration was significantly longer in the
WWW.JENONLINE.ORG
439
RESEARCH/Johnson and Alhaj-Ali
interrupted scenarios (t = –2.475, P = .025), with uninterrupted scenarios having a mean of 8.60 minutes (SD, 2.410), whereas interrupted scenarios had a mean of 11.25 minutes (SD, 2.137). No significant difference was found in the triage duration (t = –.107, P = .211) between scenarios with a correct acuity score with a mean of 9.97 minutes (SD, 2.88) and an incorrect ESI with a mean of 9.80 minutes (SD, 1.515). Correlations between triage duration and both ED and general nursing experience can be found in Table 2. Although triage durations were longer for participants who possessed more ED experience, higher educational degrees were associated with shorter triage duration. Triage accuracy: The participants’ acuity assessments (ESI) were correct in 77% of cases. Of the 18 scenarios (2 scenarios per participant), 3 had incorrect ESI scores and 1 had a missing acuity score. The correct ESI was assigned for every interrupted scenario except for one instance where no ESI was documented. Conversely, of the 9 uninterrupted scenarios, there were 3 incidents of incorrect ESI assignment. The 3 cases of incorrect ESI scores were all for the same uninterrupted scenario and were all assigned a more acute score by nurses with more than 15 years of ED experience each. Because 3 uninterrupted scenarios were assigned incorrect acuity scores, further investigation was warranted. To ensure that the researchers set the correct ESI value for the scenario, 2 outside reviewers watched all of the scenario’s videos and then provided an ESI score based on the SP’s performance. In all cases, the scenario was scored as a 4, which was the expected ESI score. The χ 2 analysis revealed no relationship between triage accuracy and the participants’ place of employment χ 2 (2, N = 9) = 1.36, P = .505; the presence of interruptions χ 2 (1, N = 18) = 1.286, P = .257; the number of interruptions χ 2 (2, N = 18) = 2.089, P = .352; or the highest educational degree of the participant χ 2 (2, N = 9) = 4.018, P = .134. Phase II The qualitative analysis of the interviews revealed many themes concerning how the participants dealt with the interruptions and consequently how they made decisions during that time. These themes were: (1) control, (2) common, (3) nature of the interruption, (4) refocus, and (5) rationale for incorrect ESI assignment. Control: One third of the participants reported the “need to control” the triage environment. The participants reported that interruptions have a “big impact because they take the control away from the nurse but also that the interruptions need to be controlled by the nurse…. Interruptions impacted effectiveness because it decreases the nurse’s control of the situation” (Participant 8). One participant stated that
440
JOURNAL OF EMERGENCY NURSING
interruptions “could have a big impact on triage because it disturbs the flow,” thus leading to loss of concentration and the potential to miss important patient information. Common: All of the participants reported that interruptions are common in the triage environment. One participant reported that interruptions did not bother her because she had “a mindset of ER. You know how to deal with the situation” (Participant 3). A second participant said, “These interruptions are a common problem in the ER” (Participant 5). Another remarked that it is “just sort of part of the usual triage, like, usual experience” (Participant 6). Another echoed the sentiment by stating, “I’m really used to being interrupted” (Participant 4). Interruption nature: Patient care–related interruptions are less offensive than noncare- related interruptions. Although not all of the participants reported having trouble getting back on track, all of the participants who experienced non-patient care–related interruptions reported needing to take time to refocus on the patient they were triaging. Participants reported that they felt annoyed and wanted to say, “I have stuff to do!! Get out of my way!!” (Participant 8). Most of the persons who participated in the scenario in which the “registration clerk” enters to verify the right chart stated that this interruption was not disruptive because “she was relevant with the situation.” Refocus: Most participants reported using the chart as a prompt for refocusing. However, the participants who reported being annoyed with the interruption said that they had difficulty refocusing because they lost their place in the chart. Another participant reported that the ability to refocus “depends on person’s reaction, like, how they are acting. If they are cooperative, it’s easier” (Participant 6). Two other participants reported that they have a normal order of triage they follow. Rationale for incorrect ESI assignments: Some ESI assignment decisions were influenced by emotional reactions to the patient. The 3 participants who assigned inaccurate ESI scores explained that they “bumped up” the acuity score because of the patient’s anxiety level (Participant 4 and Participant 7). The one participant who forgot to assign an ESI during her interrupted scenario reported that she forgot because she was “thrown off track” (Participant 3). Simulation experience: The participants were asked to evaluate the usefulness of this simulation experience for assessing triage accuracy. All participants reported that the simulations were realistic and that the SPs were believable. Four participants provided recommendations for improvement. They stated that having the vital signs already available to them was disruptive to their assessment process. Two of those participants reported the collection of vital signs as an important time for patient assessment and observation. One
VOLUME 43 • ISSUE 5
September 2017
Johnson and Alhaj-Ali/RESEARCH
participant (#8) explained that she “uses vitals to get to know the patient.” Two participants reported that using the paper charting method was not ideal. One participant reported that only having one interruption during a triage interview was unrealistic and that we should “have more interruptions for the next time” (Participant 1). Discussion
This study is the first to assess the practicality of using SPs to test how participants respond to interruptions during triage. Previous work has shown that simulation, both computer-based and patient-based simulators (eg, SIM MAN), has been used successfully to train nurses. 21 Although using SPs is not a new trend in health research, this study is the first to use SPs to explore triage accuracy in relation to interruptions. 22–24 TRIAGE DURATION
Longer triage durations were associated with more ED experience. Gerdtz and Bucknall 25 reported that assigning acuity during triage averaged 3.36 minutes, with noninterrupted triage lasting 3.47 minutes and interrupted triage lasting 4.05 minutes, which is significantly lower than the 8.6 and 11.3 minutes, respectively, in this study. However, the difference may be due to the data collection methods. In our scenarios, the participant was required to retrieve the SP from the waiting room, and our participants were tasked with completing the triage assessment, which included more than assigning acuity scores. Higher degrees were associated with shorter triage durations, which may be indicative of the role of experience, coupled with an ability to focus on the assessment’s important aspects and skip the screenings that are often added into the triage assessment. However, we were unable to assess mediating, moderating, and multicollinear relationships for these variables because of the small sample size. One explanation for this finding is that a more experienced nurse becomes more confident in performing the necessary tasks and therefore required less time to make a triage decision compared with more novice nurses. 26 TRIAGE ACUITY
The correct ESI was assigned for every interruption scenario except one, in which no ESI was documented. Conversely, the uninterrupted scenario had 3 incidents of incorrect ESI assignment. The incorrect ESI scores were all for the same uninterrupted scenario, and in each instance, the participant gave a higher acuity score. The ESI for the uninterrupted scenario was a 4, but 2 participants assigned it an ESI of 3,
September 2017
VOLUME 43 • ISSUE 5
and another participant assigned it an ESI of 2. Upon discussion with the participants, each reported that she would have assigned an ESI of 4 except for the anxiety that the woman displayed. Because of the patient’s anxiety, they made the decision to assign a more acute ESI score. This decision may be explained by connecting a judgment confidence with perceived difficulty of that judgment. Previous studies have shown overconfidence with difficult judgment and underconfidence with easy judgment. 27–29 Interestingly, most of the incorrect ESI assignments occurred in the uninterrupted scenarios. More experienced participants were more likely to assign an incorrect ESI, but participants consistently reported that interruptions and distractions had less of an impact on more experienced nurses. Retraining in ESI algorithms may be useful in the more experienced nurses, and enhanced education added to quality improvement projects might improve overall accuracy rates. 30–32 Providing this training in a simulation laboratory using SPs may be an appropriate method to refresh ESI protocols for more experienced nurses. Previous research on overtriaging (assigning a higher acuity) is mixed. Although our findings of overtriage coincides with Kilner’s findings 25 of the nurse’s tendency to overtriage, it contradicts the work of Considine et al 33 and Platts-Mills et al, 11 who reported that ED nurses were more likely to undertriage (ie, assign a lower acuity to) older patients. Although in our study the triage acuity was incorrect in 22% of scenarios, no significant difference was found between the interrupted and uninterrupted scenarios. Again, this finding may be due to the small sample size. When considering all of the scenarios, interrupted and uninterrupted, the participants in this study correctly assigned an ESI 78% of the time, whereas previous research 34,35 places accurate acuity rates much lower, at 58% and 56%. It is unclear why our study had a higher accuracy. Perhaps the simulated environment affected the amount of time the participants spent assessing the patient. Participants did not have the pressure of a full waiting room, but we did place a 15-minute limit on the scenario, thus mimicking the time pressure of triage in reality; the participants were aware of this stipulation before beginning. Working under pressure (timed simulation scenario) tends to increase the level of urgency that led to increased time spent in completing the task. 27 DECISION MAKING
It is important for triage nurses to make accurate acuity assessments. To accomplish this task, they need to be able to control their environment. When controlling the environment is not possible because of circumstances such as
WWW.JENONLINE.ORG
441
RESEARCH/Johnson and Alhaj-Ali
interruptions, the nurse needs to be able to refocus on the patient they are triaging. Refocusing may be easier to do when an interruption is based on patient care rather than other issues. The participants who were interrupted for non-patient care–related issues reported feeling annoyed in those instances, and it was during one such occasion that an acuity score was not recorded for a patient in this study. Similarly, Berg et al 3 found that interruptions are handled differently based on whether they are viewed as disruptive or nondisruptive on the basis of the nature of interruption and how relative it is to patient care. As in a previous study, 5 nurses reported that interruptions are common in the emergency department. The participants in our study stated that they are accustomed to dealing with them. Because interruptions are ubiquitous to the emergency department, we must be able to identify which interruptions are problematic and how they affect the care of patients. Other factors that affect nurse triage decisions that may lead to errors should be explored as well. Some of these factors include ED crowding level, ED nurse shortage, patient anxiety level, family member presence, and nurses’ personal factors. Because triage can be affected by so many outside influences, we must determine which factors have the biggest impact on patient care. The ramifications of interruptions on the triage process require a comprehensive assessment and evaluation to design interventions to effectively assist the nurse in managing interruptions. Simulated environments have been used successfully to improve confidence in triage training 21 and to assist in designing interventions to decrease the effects of interruptions on patient care. 13
Limitations
This study enabled a comparison of interrupted and uninterrupted simulation scenarios and their effects on triage duration and accuracy. However, a larger sample is required to determine if the errors and missed acuity assignments are generalizable to a larger population. We did not simulate a realistic arrival time between new ED patients, so the accurate pressures of triage may not have been fully realized. Although we purposely had the participants use paper charts instead of computerized point-and-click charting, participants who were familiar with paper charting may have had an advantage when reverting back to this charting method. The use of paper charts also may have contributed to the lower rate of chart completion. 36 Further research designed to examine the effect of ED interruptions on triage duration, accuracy, and decision making in different situations using larger samples and more rigorous data collection strategies is also needed. However, this project lays the foundation for designing those studies.
442
JOURNAL OF EMERGENCY NURSING
Implications for Emergency Nursing
This study forms a foundation for understanding how ED interruptions affect triage nurses and how quality of triage care affects patient safety and quality of care. The results of this study support the importance of using simulation as a learning strategy in nursing. Practicing in a simulated setting might be a good way for new triage nurses to hone their skills at managing interruptions. Practicing in this manner also provides veteran triage nurses with ways to recognize what is affecting the accuracy of their triage. This information will aid researchers, educators, and clinicians in evaluating competencies using simulation, as well as collaboratively designing interventions that may decrease interruptions and their effects. Furthermore, regularly incorporating triage simulations into nurses’ training or retraining could improve ESI assignment accuracy and patient care. 31
Conclusion
Interruptions in the emergency department have been shown to lead to errors and delays. Errors and delays degrade the quality of care 1 while adversely affecting outcomes. The impact of interruptions is compounded in triage, where errors and delays during this time propagate through the patient’s entire care episode. The findings of this study will be the basis for future work that examines how nurses successfully manage interruptions. Additionally, it tests interventions to assist triage nurses in managing or reducing interruptions during the important patient assessment process. Triage is the important initial assessment in which acuity is assigned to a patient. Interruptions during this process can cause delays and interfere with providing safe and efficient patient care. Interruptions during this time increase the process time for triage and can cause system delays to the ED process. Additionally, interruptions can lead to errors in assessment and documentation. Therefore, the necessity to identify potentially harmful interruptions and develop interventions to mitigate the adverse impact of these interruptions becomes apparent, thus allowing us to decrease errors and delays, which will lead to improving patient outcomes and ED process time. REFERENCES 1. Raban MZ, Walter S, Douglas H, et al. Measuring the relationship between interruptions, multitasking and prescribing errors in an emergency department: a study protocol. BMJ Open. http://bmjopen. bmj.com/content/5/10/e009076. Published October 2015. Accessed May 1, 2017.
VOLUME 43 • ISSUE 5
September 2017
Johnson and Alhaj-Ali/RESEARCH
2. Westbrook JI, Woods A, Rob MI, et al. Association of interruptions with an increased risk and severity of medication administration errors. Arch Intern Med. 2010;170(8):683-690. 3. Berg LM, Kallberg AS, Goransson KE, et al. Interruptions in emergency department work: an observational and interview study. BMJ Qual Saf. 2013;22(8):656-663. 4. Kosits LM, Jones K. Interruptions experienced by registered nurses working in the emergency department. J Emerg Nurs. 2011;37(1):3-8. 5. Johnson KD, Motavalli M, Gray D, et al. Causes and occurrences of interruptions during ED triage. J Emerg Nurs. 2014;40(5):434-439. 6. Cole G, Stefanus D, Gardner H, et al. The impact of interruptions on the duration of nursing interventions: a direct observation study in an academic emergency department. BMJ Qual Saf. 2016;25(6):457-465. 7. Jeanmonod R, Boyd M, Loewenthal M, et al. The nature of emergency department interruptions and their impact on patient satisfaction. Emerg Med J. 2010;27(5):376-379. 8. Centers for Disease Control and Prevention, National Center for Health Statistics. Emergency department visits. 2013. https://www.cdc.gov/ nchs/fastats/emergency-department.htm. Updated March 17, 2017. Accessed May 1, 2017. 9. Chisholm CD, Dornfeld AM, Nelson DR, et al. Work interrupted: a comparison of workplace interruptions in emergency departments and primary care offices. Ann Emerg Med. 2001;38(2):146-151. 10. Grossmann FF, Zumbrunn T, Cirprian S, et al. At risk of undertriage? Testing the performance and accuracy of the emergency severity index in older emergency department patients. Ann Emerg Med. 2012;60(3):317-325. 11. Platts-Mills TF, Travers D, Biese K, et al. Accuracy of the Emergency Severity Index triage instrument for identifying elder emergency department patients receiving an immediate life-saving intervention. Acad Emerg Med. 2010;17(3):238-243. 12. Burley D. Better communication in the emergency department: Duncan Burley outlines results of a literature review on how practitioners can improve history taking by adopting a more patient-centered approach. Emerg Nurse. 2011;19(2):32-36. 13. Werner NE, Holden RJ. Interruptions in the wild: development of a sociotechnical systems model of interruptions in the emergency department through a systematic review. Appl Ergon. 2015;51:244-254. 14. Wears RL, Woloshynowych M, Brown R, et al. Reflective analysis of safety research in the hospital accident & emergency departments. Appl Ergon. 2010;41(5):695-700. 15. Warden G, Altman S, Asplin B, et al. Hospital-Based Emergency Care: At the Breaking Point. Washington, DC: National Academies Press; 2007. 16. Hughes R. Patient Safety and Quality: An Evidence-Based Handbook for Nurses. Rockville, MD: Agency for Healthcare Research and Quality; 2008. 17. Aacharya RP, Gastmans C, Denier Y. Emergency department triage: an ethical analysis. BMC Emerg Med. 2011;11:16. 18. Creswell JW. Qualitative Inquiry and Research Design: Choosing Among Five Approaches. Thousand Oaks, CA: Sage Publications; 2012. 19. Agency for Healthcare Research and Quality. Emergency Severity Index (ESI): A triage tool for emergency department. https://www.ahrq.gov/
September 2017
VOLUME 43 • ISSUE 5
professionals/systems/hospital/esi/esi9.html. Reviewed October 2014. Accessed May 1, 2017. 20. Tanabe P, Gimble R, Yarnold PR, et al. Reliability and validity of scores on the Emergency Severity Index version 3. Acad Emerg Med. 2004;11(1):59-65. 21. Wolf L. The use of human patient simulation in ED triage training can improve nursing confidence and patient outcomes. J Emerg Nurs. 2008;34(2):169-171. 22. Gates DM, Fitzwater E, Telintelo S. Using simulations and standardized patients in intervention research. Clin Nurs Res. 2001;10(4):387-400. 23. Geurtzen R, Hogeveen M, Rajani AK, et al. Using simulation to study difficult clinical issues: prenatal counseling at the threshold of viability across American and Dutch cultures. Simul Healthc. 2014;9(3):167-173. 24. Rogers ES, Gilespie C, Zabar S, et al. Using standardized patients to train telephone counselors for a clinical trial. BMC Res Notes. 2015;7:341. 25. Gerdtz MF, Bucknall TK. Triage nurses' clinical decision making. An observational study of urgency assessment. J Adv Nurs. 2001;35(4):550-561. 26. Dallaire C, Poitras J, Aubin K, et al. Emergency department triage: do experienced nurses agree on triage scores? J Emerg Med. 2012;42(6):736-740. 27. Yang H, Thompson C, Bland M. The effect of clinical experience, judgment task difficulty and time pressure on nurses' confidence calibration in a high fidelity clinical simulation. BMC Med Inform Decis Mak. 2012;12:113. 28. Yang H, Thompson C. Nurses' risk assessment judgements: a confidence calibration study. J Adv Nurs. 2012;66(12):2751-2760. 29. Friedman CP, Gatti GG, Franz TM, et al. Do physicians know when their diagnoses are correct? Implications for decision support and error reduction. J Gen Intern Med. 2005;20(4):334-339. 30. Travers DA, Waller AE, Katznelson J, et al. Reliability and validity of the emergency severity index for pediatric triage. Acad Emerg Med. 2009;16(9):843-849. 31. van der Wulp I, Schrijvers AJ, van Stel HF. Predicting admission and mortality with the Emergency Severity Index and the Manchester Triage System: a retrospective observational study. Emerg Med J. 2009;26(7):506-509. 32. Dalawari P, Sanning J, Pan D, et al. Emergency severity index version 4 during the first year of implementation at an academic institution. J Hosp Admin. 2015;5(2):35. 33. Considine J, Levasseur SA, Villanueva E. The Australasian Triage Scale: examining emergency department nurses' performance using computer and paper scenarios. Ann Emerg Med. 2004;44(5):516-523. 34. Göransson KE, Ehrenberg A, Ehnfors M. Triage in emergency departments: national survey. J Clin Nurs. 2005;14(9):1067-1074. 35. Chen SS, Chen JC, Ng CJ, et al. Factors that influence the accuracy of triage nurses' judgement in emergency departments. Emerg Med J. 2010;27(6):451-455. 36. Benda NC, Meadors ML, Hettinger AZ, et al. Electronic health records and interruptions: the need for new interruption management strategies. Proc Int Symp Hum Fact Ergon Health Care. 2015;4(1):70.
WWW.JENONLINE.ORG
443