Point-of-care testing in ED t”>American Journal of Emergency Medicine (2013) 31, 291-296

Original Contribution

Utility of Point-of-care testing in ED triage^?,??

Olanrewaju A. Soremekun MD, MBA ^a,?, Elizabeth M. Datner MD ^a,

Simon Banh RN, BSN ^a, Lance B. Becker MD ^a, Jesse M. Pines MD, MBA, MSCE ^b

^aDepartment of Emergency Medicine, University of Pennsylvania, Philadelphia PA 19104

^bDepartments of Emergency Medicine and Health Policy, George Washington University, Washington, DC

Received 1 June 2012; revised 28 June 2012; accepted 20 July 2012

Abstract

Background: Triage systems are commonly used in emergency departments (ED) to prioritize patients. Laboratory testing is not typically used to help risk-stratify patients at triage.

Objectives: We studied the utility of point-of-care (POC) testing at triage in ED patients with high-risk complaints.

Methods: We conducted a prospective observational study on a convenience sample of ED patients at an urban academic hospital with 60,000 annual visits. Patients who were triaged to the waiting area with any of the following criteria were approached for enrollment: (1) chest pain or shortness of breath in patients older than 40 years, (2) possible infection in the presence of two or more systemic inflammatory response system criteria in patients older than 18 years, and (3) patients N 65 years with non-traumatic complaints. A total of 300 subjects were enrolled. All enrolled patients received POC testing that included a combination of Chem8+, hemoglobin, troponin, B-type natriuretic peptide, and lactate. The triage nurse completed a survey after receiving the results.

Results: POC results was reported to be helpful in 56% of patients, changed the triage level in 15% of patients and led to 6% of patients being brought back for rapid physician evaluation. Overall, 50% of patients had one or more abnormal POC laboratory tests. There was no relationship between ED census and the likelihood of being helpful, changing the triage level, changing management, or bringing patients back any faster.

Conclusion: POC testing at triage is a helpful adjunct in triage of patients with high-risk ED complaints.

(C) 2013

Introduction

Triage systems are used commonly in hospital-based emergency departments (ED) to prioritize patients with

^? Prior Presentations: No.

^?? Financial Support: Abbott Point-of-Care, Princeton NJ.

* Corresponding author. Tel.: +1 215 615 3477.

E-mail address: [email protected] (O.A. Soremekun).

regard to severity of illness and resource needs. The need for triage is borne out of periods where resources (ie, staff or space) become constrained, requiring newly arriving patients to wait. Effective triage systems are especially important in settings with prolongED wait times. Commonly, an experi- enced nurse considering the presenting complaint, overall appearance, comorbidities, and vital signs assigns a triage level. More recently, some EDs are incorporating physician triage and advance Triage protocols into their traditional triage systems to improve triage accuracy and effectiveness

0735-6757/$ - see front matter (C) 2013 http://dx.doi.org/10.1016/j.ajem.2012.07.025

[1]. The most commonly used triage system in the United States is the Emergency Severity Index , which has 5 levels, with level 1 corresponding to the highest acuity, most resource-intensive patients and level 5 corresponding to the lowest acuity, least resource-intensive patients [2,3].

Although the incorporation of physician or nursing- initiated orders in triage is becoming common place, the main focus of these interventions in the published literature has been the role of early orders in reducing the total length of stay and ED bed time with no data reported on the impact on patient safety or the quality of the triage process [4,5]. Given typical laboratory turnaround times, incorporating laboratory results into the triage of patients with high-risk complaints has been very limited. To date, none of the current triage systems integrate laboratory results into assigning triage levels. Integrating laboratory testing into triage presents an opportunity to increase the accuracy of triage by helping enhance risk-stratification. Historically, laboratory results have been logistically difficult to obtain and incorporate into triage systems because results are not immediately available. Recently, newly available point-of- care (POC) testing with handheld devices presents an opportunity to add important clinical information to the ED triage process to help identify patients who need immediate care or could potentially wait for longer periods of time. Several studies have explored the feasibility of using POC testing in pre-hospital settings or at ED triage [6-9]; however, to our knowledge, no studies have tested the impact on patient management or value of this information to Triage nurses.

We assessed the utility of POC testing in patients who

were triaged to the waiting area who had potentially high-risk complaints: chest pain, shortness of breath, possible infection with two or more systemic inftammatory response system (SIRS) criteria, and older adults with non-traumatic com- plaints. We determined the rates of abnormal POC laboratory results and how this information impacted waiting area management decisions.

Methods

We conducted a prospective observational study on a convenience sample of ED patients who were triaged to the waiting room area at a single inner-city academic hospital. The hospital sees approximately 60,000 annual patient visits per year and has a 4-year emergency medicine residency training program. At the study site, there are periods during peak times, in which patients wait for prolonged periods of time in the waiting room area prior to physician evaluation. Triage is performed by an experienced nurse using a 5-level ESI triage, where 1 = immediate care, 2 = high-risk, multiple- resources needed, 3 = multiple resources needed, 4 = a single resource needed, and 5 = no resources needed [10].

Patients were eligible for enrollment that were triaged as ESI levels 2, 3, or 4 met one or more of the inclusion criteria

and were expected to wait in the waiting room. Patients were included if they gave informed consent to be enrolled in this study and received triage POC laboratory testing. Patients were included who met any of the following criteria: (1) chest pain or shortness of breath in patients older than 40 years, (2) possible infection in the presence of two or more SIRS criteria (by vital signs only) in patients older than 18 years, and (3) older patients (N 65 years of age) with non- traumatic complaints [11]. Exclusion criteria were known pregnancy, in police custody, or an inability to provide informed consent. Because of overlapping criteria, patients could be members of one or more study groups. These specific groups of patients were chosen because of their ease of identification, high potential for unexpected laboratory results, and for whom POC laboratory results could lead to a change in triage level or management.

Patients who were triaged to the waiting room were

enrolled as a convenience sample when the study nurse was present. The study nurse was not involved at all in the triage or management of patients in the waiting room. Enrollment was performed during the hours of 1 PM to 11 PM to correspond to episodes of high patient volume to increase the likelihood that patients would be triaged to the waiting area. After completing study forms and obtaining patient consent, the study nurse drew blood by either a single stick or placed an intravenous line and collected a series of samples that could be used for care in the main ED to avoid additional needle sticks. POC laboratory testing was done using the i-STAT system (Abbott Point of Care, Princeton, NJ). The following labs were performed:

(1) chest pain or shortness of breath in patients older than 40 years: CHEM8+, hemoglobin (Hgb), troponin, B-type Natriuretic Peptide , lactate; (2) possible infection with two or more SIRS criteria: CHEM8+, Hgb, lactate; and 3) older adults (N 65 years of age) with non-traumatic complaints: CHEM8+, Hgb, troponin, lactate. The range in time it took to run all the tests and obtain results varied by test from 2 to 10 minutes, according to the manufacturer’s specifications. The Pathology department at the study site validated the i-STAT system and set the normal ranges for each of the POC tests based on already established test ranges. POC results were given to the triage nurse who performed the initial triage evaluation by the study nurse. The triage nurse acted upon the laboratory results as clinically warranted. For patients who met criteria for multiple inclusion groups, the more comprehensive set of POC laboratory tests was obtained. Several data points were captured on each patient, including demographics, comorbid conditions, triage levels, and ED census at the time of enrollment, which was used as a proxy for ED crowding [12]. In addition, a survey was conducted with the consent of the triage nurse who received the POC results. The survey queried the helpfulness of the results, and any changes in levels of Clinical concern, triage levels, or clinical management. The triage nurse survey is provided as an online Appendix.

In this pilot study, our goal was to enroll a total of 300 patients, with a minimum enrollment of 50 patients per study group. The primary outcomes of the study were the presence of any abnormal laboratory values which included: (1) Troponin-I level N .05 ng/mL, (2) Hgb b 9 g/dL, (3) glucose N 300 mmol/L or b 50 mmol/L, (4) sodium level b 130 mmol/L or N 150 mmol/L, (5) lactate level N 3.0 mmol/L, (6) potassium level N 5.5 mmoL/L or b 3.0 mmol/L, (7) bicarbonate level b 20 mmol/L, (8) anion gap N 15 mmol/L, and (9) BNP N 100 pg/mL. The normal ranges for each POC test was determined by the Pathology department at the study site. Data were entered from paper study forms into a Microsoft Excel spreadsheet (Redmond, WA) and were imported into Stata v. 11 (College Station, TX). The Institutional Review Board at the University of Pennsylvania approved this study.

Simple descriptive statistics were used to tabulate data. Data were stratified by ESI level to assess if triage nurses were more likely to change management for a given ESI level. Fisher’s exact test was used to compare groups. In addition, we tested if patients were more likely to have Abnormal test results at higher levels of ED crowding. We similarly used Fisher’s exact test to assess if management was more likely to be changed or any abnormal POC test result was more likely in four quartiles of ED census, which was used as a proxy for ED crowding. P b .05 was considered significant. No hypothesis testing was conducted across the study groups because patients could meet inclusion criteria for more than one group. This research was supported by an unrestricted grant from Abbott Point of Care, Inc.

Results

A total of 300 patients were enrolled over a 12-month period. The average age was 57 years, 41% were men, and 72% were black. (Table 1) Overall, POC lab results were reported as helpful to the triage nurse in 56% of the patients with POC results increasing the level of concern in 16%, changing the management in 14%, and led to 6% of patients being brought back for rapid physician evaluation. After the POC results, 11 (4%) of 286 had their triage levels increased to higher Acuity levels (10 were triaged from ESI 3 to 2 and one from an ESI 4 to 2) and 29 (10%) of 286 were triaged from ESI 2 to an ESI 3. Similar results were seen across study groups (see Table 2).

We compared ESI 2 and ESI 3 patients to assess if there were any significant differences in helpfulness of the POC results (58% v. 55%, P = .28), change of the triage level (19% v. 21%, P = .70), or Change in management (8% v. 17%, P = .06) based on POC results: none were significantly different. We also assessed whether quartiles of ED census were associated with any differences in helpfulness, changing the triage level, or management. None of these comparisons were significant. A total of 152 patients had one

Mean (+-SD)

Age 57 years (16) Male 122 (40.7%) Race White 68 (22.7%) Black 215 (71.7%) Other 17 (5.3%) Pre-POC triage level 2 111 (37.0%) 3 184 (61.3%) 4 5 (1.7%) Comorbid conditions Hypertension 157 (52.3%) Coronary artery disease 19 (6.3%) Diabetes 73 (24.3%) Active smoker 24 (8.0%) Hypercholesterolemia 42 (14.0%) Family history of coronary disease 7 (2.3%) ED census (median, IQR) 66 (59-73)

IQR, interquartile range.

or more abnormal POC lab results (see Table 2). A total of 69 (62%) of 111 of patients triaged as ESI 2 had at least one or more abnormal test results as compared to 81/184 (44%) of patients triaged as ESI 3, P = .006.

Table 1 Demographics of the study population (n = 300)

Limitations

This study has several limitations. The first is that it was conducted as a convenience sample of patients at a single study site with high levels of ED crowding: an average census of 66 patients in an ED with 41 licensed ED beds. This may have caused us to over-estimate the prevalence of positive findings in the waiting room because in other less crowded settings, such a high-risk patient population may not be triaged to wait. One of the reasons we may not have found any differences by ED crowding is that we sampled patients during the most crowded times, so there may not have been sufficient instances of “low-crowding” to show any differences across groups. Another factor that limits the generalizability of the findings is that triage-levels them- selves do not have particularly high inter-rater reliability, as might triage decision-making with or without POC labora- tory tests. However, ESI, which is used at the study site, has been shown to be more reliable than other systems [13,14]. We also relied on triage nurses’ clinical judgments about the helpfulness of triage POC results, which may not be an ideal standard for the clinical value of the result. We only enrolled a small number of patients in this pilot study and our group of patients with possible infection and two or more SIRS criteria was the smallest (n = 50). We may have had different results had we enrolled a larger number of patients; however, it is not possible to estimate how precisely this would have changed our results. Finally, while we report the role of

Chest pain or shortness Suspected infection Older adults (AgeN65) Total (n = 300) of breath in patients and 2 + SIRS with non-traumatic N 40 y (n = 184) (n = 50) complaints (n = 91)
management changes/clinical perception
Increased level of concern	32	(17.4%)	9 (18.0%)	13	(14.3%)	48 (16.0%)
Decreased level of concern	7	(3.8%)	1 (2.0%)	7	(7.7%)	15 (5.0%)
Increased the triage level ?	6	(3.4%)	1 (2.3%)	5	(5.7%)	11 (4.2%)
Decreased the triage level ?	22	(12.4%)	4 (9.1%)	7	(8.0%)	29 (10.1%)
Changed management	26	(14.1%)	7 (14.0%)	12	(13.2%)	41 (13.7%)
in any way
Brought patient back faster	12	(6.5%)	4 (8.0%)	3	(3.3%)	18 (6.0%)
POC lab results reported as	105	(57.1%)	26 (52.0%)	51	(56.0%)	167 (55.6%)
“helpful” to triage nurse
Positive result rates
Troponin (N 0.05 mmol/L)	14	(7.7%)	1 (2.0%)	5	(5.5%)	19 (6.3%)
B-Type Natriuretic Peptide	51	(27.7%)	3 (6.0%)	11	(12.1%)	52 (17.3%)
(N 100 pg/dL)
Glucose (N 300 mmol/L	12	(6.5%)	1 (2.0%)	1	(1.1%)	13 (4.3%)
or b 50 mmol/L)
Lactate level N 3.0 mmol/L	9	(4.9%)	1 (2.0%)	1	(4.4%)	14 (4.7%)
Potassium level	11	(6.0%)	5 (5.0%)	4	(4.4%)	19 (6.3%)
N 5.5 mmol/L or b 3.0 mmol/L
Sodium level	1	(0.5%)	3 (6.0%)	2	(2.2%)	5 (1.7%)
(b 130 mmol/L or N 150 mmol/L)
Bicarbonate level	3	(1.6%)	-	3	(3.3%)	5 (1.7%)
(b 20 mmol/L)
Anion Gap N 15	48	(26.1%)	10 (20.0%)	21	(23.1%)	73 (24.3%)
Hemoglobin (b 9.0 g/dL)	3	(1.6%)	1 (2.0%)	6	(6.6%)	9 (3.0%)
Any positive result	107	(58.2%)	21 (42.0%)	42	(46.2%)	152 (50.7%)
* n = 286.

triage POC results in changing triage levels and waiting room management, the study was not designed to measure the impact of these changes on clinical outcomes.

Table 2 How POC tests were used, helpful or changed clinical management

Discussion

We found that POC testing was reported to be helpful to the triage nurse in more than half of patients in this pilot study; and a change in patient management or triage level in a significant number of patients being sent to the waiting room (14% and 15% respectively). POC results expedited physician evaluation in 6% of patients because those patients were placed in main ED rooms sooner. This suggests our current ESI triage procedure may miss clinically significant findings in about 1 in 20 patients in this population.

In our experience, the key advantage to POC over our standard lab is the shorter turnaround time. Given the limited steps from Blood draw to results with POC testing, the turnaround time at our institution was much shorter when compared to standard lab testing. This short turnaround time allows for the triage nurse to easily incorporate the results into their triage decisions and waiting room management. For departments with short standard lab turnaround times,

incorporating laboratory testing into the triage process using standard labs may be an alternative to POC lab testing.

On the surface, it may appear that even in high-risk ED populations in a crowded ED setting, POC Laboratory test results infrequently change triage levels or management decisions, casting doubt on whether POC tests are a useful adjunct at triage. However, an alternative viewpoint is that management changes in 15% of patients who were triaged as safe to wait is clinically important, particularly for the 6% of patients whose results prompted expedited care or the 4% whose acuity level was upgraded from ESI 3 to ESI 2 because of abnormal test results.

The value of individual elements of the history or vital signs in assessing triage is not well-studied. Age, vital signs, and a need for Timely interventions have been associated with nurses assigning triage ESI 2, while expected numbers of resources are the major reason for assigning ESI 3 [15]. When comparing actual resource utilization, ESI 2 patients are more likely to require cardiac monitoring, electrocardio- gram, medications, and specialty consultation than ESI 3 patients [16]. These studies suggest that the risk of decline and intervention differentiate ESI 2 from 3. Our study was meaningful different from these because we specifically addressed the incremental value of POC labs at triage. This was in the context of an experienced clinician already having

evaluated a patient as safe for the ED waiting room using an ESI triage system that is based on historical and some physiological information, specifically vital signs [10]. In pediatric patients, vital signs alone have been shown to be helpful in distinguishing serious infections [17].

Overall, we found that more than half of patients had at least one abnormal test result, which was defined by being outside of the specified normal ranges specified by the Department of Pathology at the study site. Certain high-risk results, such as an elevated Troponin that could represent ongoing myocardial ischemia, a low Hgb — in the absence of knowing a patient’s baseline value — could represent acute, Life-threatening bleeding, or an abnormal potassium level would result in immediately raising concern. In addition, several studies have shown that an elevated lactate is an indicator commonly used to detect tissue hypoperfu- sion, particularly in the case of sepsis and requires early resuscitation to improve outcomes [18,19]. In the broad cohort, 6% of the Troponin results were elevated, 3% of Hgb were low, 6% of potassiums were abnormal, and 5% of lactates were elevated indicating that these high-risk laboratory values are not uncommon in high-risk ED patients that are triaged to the waiting area. The most common abnormalities were Elevated BNP levels and elevated anion gaps, which individually were noted in about 1 in 4 patients. Out of context, these test results may not be immediately alarming; however, BNP alone has independently been associated with higher risk of mortality [20,21]. Our results suggest that a relatively high-risk group of patients is being triaged to wait.

In crowded EDs, the waiting room is often seen as the

highest risk area. After triage, waiting before physician evaluation can be potentially a patient safety issue in several situations. First, if a time-sensitive disease is not apparent on initial evaluation. Specifically, the patient’s immediate physiology does not suggest a time-sensitive illness. As a group, older adults are at higher risk for under-triage than younger patients, which is supported by our relatively high rate of abnormal test results and management changes in this broad population [13]. Adding more objective physiological information may be helpful in risk-stratification. There are two other related situations where POC laboratory results may be particularly helpful: when the ED is very crowded — and high-risk patients are required to wait — or the expected waiting time will be prolonged. Therefore, an ED may choose to deploy POC laboratory testing to triage in certain high-risk populations when the expected wait will be long or there are high-risk patients who are required to wait. In these situations, it could actually be more helpful in identifying the patients who can wait rather than those who can’t, which is supported by our results where more than 3 times as many patients had their triage levels downgraded as upgraded. In addition, several studies on physician or nursing initiated lab and radiology orders in triage have demonstrated a reduction in the LOS and time in ED bed [5,4]. While our intervention focused on improving the sensitivity of triage, there may be

some additional benefit on improvement in LOS from early initiation of lab testing.

In conclusion, we found that POC test results were helpful to the triage nurse and changed the management of a clinically significant proportion of high-risk patients triaged to the ED waiting area in our hospital. A strategy to deploy POC-testing to the waiting area when high-risk patients are required to wait may be useful in improving patient safety. This strategy must be designed to minimize the impact of the time to triage or other key waiting room metrics such as time to electrocardiogram. Future studies in this area should explore the impact of POC testing in triage on patient outcomes as well as the overall cost effective- ness of this intervention.

Acknowledgments

The authors are grateful for the assistance from Dr Nachamkin, Amy Reiprish, and the rest of the Department of Pathology for the pre-study validation as well as support during the study. We are also very grateful to the entire nursing staff for their support and assistance. This research was supported by an unrestricted grant from Abbott Point of Care Inc.

References

1. Wiler JL, Gentle C, Halfpenny JM, Heins A, Mehrotra A, Mikhail MG, et al. Optimizing emergency department front-end operations. Ann Emerg Med 2010;55:142-60.e1.
2. Singer RF, Infante AA, Oppenheimer CC, West CA, Siegel B. The use of and satisfaction with the emergency severity index. J Emerg Nurs 2012;38:120-6.
3. McHugh M, Tanabe P, McClelland M, Khare RK. More patients are triaged using the Emergency Severity Index than any other triage acuity system in the United States. Acad Emerg Med 2012;19:106-9.
4. Retezar R, Bessman E, Ding R, Zeger SL, McCarthy ML. The effect of triage diagnostic standing orders on emergency department treatment time. Ann Emerg Med 2011;57:89-99.e2.
5. Russ S, Jones I, Aronsky D, Dittus RS, Slovis CM. Placing physician orders at triage: the effect on length of stay. Ann Emerg Med 2010;56: 27-33.
6. Sorensen JT, Terkelsen CJ, Steengaard C, Lassen JF, Trautner S, Christensen EF, et al. Prehospital troponin T testing in the diagnosis and triage of patients with suspected acute myocardial infarction. Am J Cardiol 2011;107:1436-40.
7. Calzavacca P, Licari E, Tee A, Bellomo R. Point-of-care testing during medical emergency team activations: a pilot study. Resuscitation 2012.
8. Arora S, Henderson SO, Long T, Menchine M. Diagnostic accuracy of point-of-care testing for diabetic ketoacidosis at emergency-depart- ment triage: {beta}-hydroxybutyrate versus the Urine dipstick. Diabetes Care 2011;34:852-4.
9. Goyal M, Pines JM, Drumheller BC, Gaieski DF. Point-of-care testing at triage decreases time to lactate level in septic patients. J Emerg Med 2010;38:578-81.
10. Gilboy N, Tanabe P, Travers DA, Rosenau AM, Eitel DR. Emergency Severity Index, Version 4: Implementation Handbook. 2012, 04/18.
11. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus

Conference Committee. American College of Chest Physicians/Soci- ety of Critical Care Medicine. Chest 1992;101:1644-55.

1. Hwang U, McCarthy ML, Aronsky D, Asplin B, Crane PW, Craven CK, et al. Measures of crowding in the emergency department: a systematic review. Acad Emerg Med 2011;18:527-38.
2. Platts-Mills TF, Travers D, Biese K, McCall B, Kizer S, LaMantia M, 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;17:238-43.
3. Storm-Versloot MN, Ubbink DT, Kappelhof J, Luitse JS. Comparison of an informally structured triage system, the emergency severity index, and the Manchester triage system to distinguish patient priority in the emergency department. Acad Emerg Med 2011;18:822-9.
4. Garbez R, Carrieri-Kohlman V, Stotts N, Chan G, Neighbor M. Factors influencing patient assignment to level 2 and level 3 within the 5-level ESI triage system. J Emerg Nurs 2011;37:526-32.
5. Garbez RO, Carrieri-Kohlman V, Stotts N, Chan G. Level 2 and level 3 patients in emergency severity index triage system: comparison of characteristics and resource utilization. Adv Emerg Nurs J 2011;33: 322-35.
6. Thompson M, Coad N, Harnden A, Mayon-White R, Perera R, Mant D. How well do vital signs identify children with serious infections in paediatric emergency care? Arch Dis Child 2009;94: 888-93.
7. Jones AE, Puskarich MA. Sepsis-induced tissue hypoperfusion. Crit Care Nurs Clin North Am 2011;23:115-25.
8. Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2008. Crit Care Med 2008;36:296-327.
9. Fonarow GC, Peacock WF, Phillips CO, Givertz MM, Lopatin M, ADHERE Scientific Advisory Committee and Investigators. Admis- sion B-type natriuretic peptide levels and in-hospital mortality in acute decompensated heart failure. J Am Coll Cardiol 2007;49: 1943-50.
10. Grewal J, McKelvie RS, Persson H, Tait P, Carlsson J, Swedberg K, et al. Usefulness of N-terminal pro-brain natriuretic peptide and brain natriuretic peptide to predict cardiovascular outcomes in patients with heart failure and preserved left ventricular ejection fraction. Am J Cardiol 2008;102:733-7.