a b s t r a c t

Background: In evaluating patients with chest pain, emergency department Observation Units (EDOUs) may use a staffing model in which emergency physicians determine patient testing (EP model) or a model similar to a chest pain unit (CPU) in which cardiologists determine provocative testing (CPU model).

Methods: We performed a prospective study with 30-day telephone follow-up for all chest pain patients placed in our EDOU. Halfway through the study period, our EDOU transitioned from an EP model to a CPU model. We com- pared provocative testing rates and outcomes between the 2 models.

Results: Over the 34-month study period, our EDOU evaluated 1190 patients for chest pain. Patients placed in the EDOU during the 17-month CPU model were more likely to be moderate risk (Thrombolysis in Myocardial Infarc- tion score 3-5) than those during the 17-month EP model: 24.9% vs 18.8%, P = .011. Despite this difference, rates of provocative testing (Stress testing or coronary computed tomography) were lower during the CPU model: 47.1% vs 56.5%, P = .001. This reduction was particularly evident among low-risk patients (Thrombolysis in Myo- cardial Infarction score 0-2): 49.8% vs 58.1%, P = .011. Rates of myocardial infarction, percutaneous coronary in- tervention, or coronary artery bypass graft were similar between the 2 groups (2.8% vs 3.2%, P = .140). We noted no significant events or missed diagnoses in either group during the 30-day follow-up.

Conclusion: An EDOU model that used mandatory cardiology consultation resulted in decreased provocative testing, particularly among Low-risk chest pain patients. Future research should explore the cost-effectiveness of this model.

(C) 2016

Introduction

One of the more challenging patient care issues in the emergency department (ED) is determining the type and extent of testing to be per- formed in patients presenting with chest pain. Emergency department observation units (EDOUs) may fill an important role in the evaluation of these patients, but appropriate staffing models for the observation units have not been determined. Some centers use dedicated Chest pain units (CPUs), which are staffed by cardiologists, whereas others may use an EDOU staffed primarily by emergency physicians (EPs) or advanced practice clinicians (APCs).

Our EDOU used a model in which EPs placed patients with chest pain in the unit on a chest pain protocol. Cardiology consultation was avail- able and was typically obtained in the observation unit only for patients who had positive testing results. Our model changed, however, to a model in which cardiologists consulted on all patients with chest pain who were in the observation unit to determine the necessity and

? Grant support: University of Utah medical group.

?? Conflicts of interest: None.

* Corresponding author at: Division of Emergency Medicine, University of Utah, 30 N. 1900 E. 1C26, Salt Lake City, UT 84132. Tel.: +1 801 581 2417; fax: +1 801 585 6699.

E-mail address: [email protected] (T. Madsen).

appropriate provocative testing, regardless of whether these patients had positive results in their testing to that point.

The purpose of this study was to determine how this change in our EDOU staffing model affected testing and outcomes among patients with chest pain. Specifically, we wanted to note whether the presence of man- datory cardiology consultation led to changes in the rates of provocative testing and how this change may have affected patient outcomes.

Methods

We performed a prospective, observational study with 30-day follow-up for all patients admitted to our EDOU with chest pain over a 34-month period. The University of Utah ED is an urban ED with ap- proximately 39 000 visits annually. The ED is staffed by attending, board-certified EPs who supervise emergency medicine residents, ro- tating residents, and medical students in the care of ED patients. The EDOU is a 10-bed unit that opened in April 2006. The observation unit is under the direction of EPs and APCs (physician assistants and nurse practitioners) and cares for more than 2500 patients per year. The unit is designed to care for a variety of patients who are admitted under var- ious Treatment protocols. Each protocol has specific inclusion and exclu- sion criteria for the observation unit and protocolized order sets for

http://dx.doi.org/10.1016/j.ajem.2016.09.032

0735-6757/(C) 2016

26 T. Madsen et al. / American Journal of Emergency Medicine 35 (2017) 25-28

these patients. The University of Utah EDOU uses 21 different treatment protocols, including chest pain, abdominal pain, transient ischemic at- tack, and pulmonary embolism, with the chest pain protocol being the most frequently used.

The chest pain protocol allows physicians to admit patients with chest pain to the observation unit if they have a nondiagnostic electro- cardiogram (ECG), have normal or indeterminate troponin I on initial ED testing, and are hemodynamically stable. Patients requiring escala- tion of care for continued chest pain are admitted to the inpatient set- ting. Additional exclusion criteria include significant comorbidities (cancer, hepatic disease, congestive heart failure, or renal disease). A history of coronary artery disease is not an exclusion criterion, and pa- tients may be admitted to the observation unit at the EP’s discretion.

All patients with chest pain admitted to the EDOU receive serial tro- ponin I measurements every 6 hours for a total of 3 (including the initial troponin I in the ED) and ECG as needed for recurrent chest pain or as deemed necessary by the EDOU provider. This serial troponin rule-out protocol was considered standard practice at the time and was devel- oped for the EDOU in cooperation with the Division of cardiovascular medicine. Before November 1, 2010, EPs made the decision to place pa- tients in the EDOU and ordered all testing to be performed during the EDOU stay, including provocative testing. Patients placed in the EDOU typically stayed overnight for a period of up to 23 hours for additional testing and evaluation. The EDOU was staffed by APCs who carried out the orders which had been placed by the EP. If the EP ordered provoca- tive testing, this was typically performed during the morning of the EDOU stay, and cardiology was consulted at that point for performance of the provocative testing and interpretation. admission decisions were based on results of the patient’s testing in the EDOU, determination for further inPatient evaluation based on the patient’s history and presenta- tion, or additional diagnoses or conditions which may have been identi- fied during the EDOU stay.

On November 1, 2010, the EDOU model changed to one in which car-

diologists evaluated all EDOU chest pain patients. Cardiologists consulted on all patients during the morning after their placement in the EDOU to review testing results and to determine if additional testing was neces- sary, and, if so, which testing modality would be appropriate (coronary stress testing, coronary computed tomography angiography [CTA], or car- diac catheterization). The decision to implement mandatory cardiology consultation was a joint decision between the ED and Division of Cardio- vascular Medicine. The intent of this change was to increase the level of care provided to chest pain patients in the EDOU and to bring this staffing level to something more consistent with a traditional CPU.

We performed the study as part of an ongoing quality assessment of patient selection and outcomes in the EDOU. Emergency physicians and cardiologists were not aware that we would be comparing provocative testing rates and patient outcomes before and after the change in the staffing model. Research associates were blinded to the study hypothesis, and data collection methods, questions, and follow-up procedures were identical in the periods before and after the change in staffing model.

We performed a prospective evaluation of all patients admitted to

the observation unit under the chest pain protocols for the 34-month period from June 1, 2009, through March 31, 2012. This study period in- cluded the 17 months before the change on November 1, 2010, and 17 months after the change. The study received approval from the hospital’s institutional review board, and all participants provided in- formed consent for participation and contact via telephone after 30 days. Trained research associates staffed the observation unit 7 days a week to obtain baseline data on all patients placed in the observation unit under the chest pain protocols. Baseline data were recorded at the time of the patient’s observation unit placement. For patients placed in the observation unit between midnight and 0700 hours, research as- sociates recorded baseline data during the morning of the patient’s stay. We reviewed Electronic medical records for patient outcomes related to the EDOU stay, including positive troponin I, provocative test- ing, cardiac catheterization, percutaneous coronary intervention (PCI),

coronary artery bypass graft (CABG) surgery, and inpatient admission from the EDOU. We noted adverse events during the observation unit ad- mission, including bleeding, clinically significant arrhythmia requiring medical intervention, unstable vital signs, and death. Our laboratory tro- ponin I assay defined a negative troponin I as <=0.05 ng/mL, an indetermi- nate troponin I as 0.06-0.49 ng/mL (recommended retesting in 4-6 hours), and a positive troponin I as >=0.50 ng/mL, consistent with myocar- dial Ischemic injury (troponin I chemiluminescent immunoassay). For our study purposes, we defined a positive troponin I as >=0.50 ng/mL.

We determined patient outcomes during the 30-day follow-up

through telephone contact with patients and review of the EMR at least 30 days after the EDOU admission. Trained research associates attempted to contact patients by telephone at least 30 days after their EDOU stay. They used a prepared script with questions regarding events since the EDOU stay, hospitalizations, follow-up visits, or procedures performed. In our analysis, we included only events that occurred dur- ing the EDOU stay or during the 30 days after the ED visit. For patients presenting to the ED more than once during the study period, we in- cluded only their initial ED visit in the analysis, with subsequent visits used for outcomes related to the index visit if they occurred during the 30-day follow-up period. We contacted 68.2% of patients by phone, with follow-up by EMR review for the remaining 31.8%. For pa- tients not contacted by phone, we searched the Social Security Death Index to ensure no deaths were missed.

We included all patients in the final database. The primary study outcome was the rate of provocative testing when comparing the 2 groups. For classification purposes, we refer to the period before manda- tory cardiology consultation as the EP model, given that the EP primarily determined provocative testing in the EDOU. We refer to the period after the change as the CPU model, given the implementation of manda- tory cardiology consultation similar to the staffing model in a CPU.

We additionally noted rates of myocardial infarction (MI), PCI, or CABG as well as missed events. A “missed event” included any patient who experienced MI or death during the 30-day follow-up period. Addi- tional missed events included PCI or CABG during the 30-day follow-up period which was not part of the arranged outpatient follow-up from the EDOU. We calculated the Thrombolysis in Myocardial Infarction score using baseline variables. We classified patients as low risk if they had a TIMI score of 0 to 2 and moderate risk with a TIMI score 3 to 5. Because of EDOU exclusion of individuals with significant ECG changes or a positive troponin I, patients placed in the EDOU did not have TIMI scores higher than 5.

We also evaluated rates of cardiac catheterization, focusing specifically on patients who went to cardiac catheterization from the EDOU without a positive troponin I during the EDOU stay and without having undergone provocative testing before catheterization. We evaluated whether differ- ences in provocative testing rates may have been offset by differences in rates of cardiac catheterization, particularly among patients who may not have had objective findings during the EDOU stay which precipitated cardiac catheterization. We refer to this as “primary cardiac catheteriza- tion,” given that these were cases in which cardiac catheterization was used at the primary means of evaluation for coronary disease, rather than using a less invasive modality before this procedure.

The principal investigator (TM) reviewed 20% of the medical re- cords, selected randomly, as quality assessment of data obtained from the EMR. In addition, the principal investigator reviewed the records of all patients who had MI, inpatient admission, cardiac catheterization, positive stress test result, PCI, or CABG to assure the accuracy of the medical record review and to confirm details as reported by patients in telephone follow-up (? = 0.96). Research associates entered data into Research Electronic Data Capture database and imported data to a Microsoft Excel database for statistical analysis (Microsoft Corporation, Redmond, WA). We performed statistical analysis using Stata (v. 12.0; Statacorp, College Station, TX), with P b .05 considered statistically sig- nificant. We used Pearson ?² test for comparison of categorical variables and Student t test for continuous variables.

T. Madsen et al. / American Journal of Emergency Medicine 35 (2017) 25-28 27

Percent female 53.6% (317) 52.8% (316) .760

Table				20
Baseline characteristics				18
	EP model (n or SD)	CPU model (n or SD)	P	16
Total patients	591	599		14

Average age (y) 54 (+-12.7) 54 (+-13.4) .980

Average TIMI score 1.5 (+-1.1) 1.9 (+-1.1) b.001

History of CAD 17.6% (104) 13.1% (78) .035

Diabetes 19.6% (116) 17.7% (106) .392

Hypercholesterolemia 36.7% (217) 43.1% (258) .025

Hypertension 16.9% (100) 47.6% (285) b.001

Family history of CAD 38.2% (226) 57.1% (342) b.001

Current smoker 19.5% (115) 19.5% (117) .974

>=3 cardiac risk factors 23.5% (139) 30.4% (182) .007 Abbreviation: CAD, coronary artery disease.

12

10

11.7%

10.5%

3.2% 2.8%

Percent

8

6

4

2

0

MI/PCI/CABG Inpatient admission

EP model CPU model

Results

During the 34-week study period, 1190 EDOU chest pain patients agreed to participate in the study. The “EP model” group included 591 patients who presented during the first 17 months of the study and be- fore the implementation of mandatory cardiology consultation. The “CPU model” group included 599 patients presenting during the second 17 months of the study and after mandatory cardiology consultation was in place. More than half (53.6%) of patients in the EP model and 52.8% in the CPU model were female (P = .760). Average age was 54 years in the EP model and 54 years in the CPU model (P = .980). Aver- age TIMI score was higher in the CPU model: 1.5 vs 1.9 (P b .001). Medi- an TIMI score was 1 for the EP model and 2 for the CPU model (25th and 75th percentile TIMI scores were equal for both groups: 1 and 2, respec- tively). Patients in the CPU model were less likely to report a history of coronary disease (previous MI, PCI, or CABG) but were more likely to re- port multiple cardiac risk factors (Table).

We compared diagnostic testing rates between groups. In the EP model, 56.5% of patients underwent diagnostic testing (stress testing or coronary CTA) compared to 47.1% of patients in the CPU model (P = .001). When selecting a testing modality, cardiologists were signif- icantly more likely to perform myocardial perfusion imaging over stress echocardiography or coronary CTA: 24.5% of tests were myocardial per- fusion tests in the CPU model vs 5.7% in the EP model (P b .001). We stratified patients by TIMI score into low risk (score of 0-2) and moder- ate risk (score of 3-5). Overall, 21.8% of patients were moderate risk, and patients in the CPU model were more likely to be moderate risk: 18.8% vs 24.9% (P = .011). Reductions in testing rates were significant among low-risk patients: 58.1% of low-risk patients in the EP model underwent provocative testing vs 49.8% in the CPU model (P = .011). Differences in testing rates among the moderate-risk patients were not statistically significant: 49.5% EP model vs 38.9% CPU model (P =

.087; Fig. 1). Among all study patients, 3.9% underwent primary cardiac

70

58.1%

49.8%

49.5%

38.9%

60

50

40

Percent

30 EP model

Fig. 2. Comparison of rates of MI, PCI, and CABG, as well as inpatient admission. Differences were nonsignificant for both (P = .140, P = .430).

catheterization (cardiac catheterization without a preceding diagnostic test or a positive troponin I). In the EP model, 3.9% of patients had pri- mary cardiac catheterization, whereas 4% of patients in the CPU model had primary cardiac catheterization performed (P = .919).

Finally, we evaluated outcomes related to the EDOU stay. The com- posite rate of MI, PCI, or CABG was similar between the EP model and CPU model: 3.2% in the EP model vs 2.8% in the CPU model (P = .140). Similarly, inpatient admission rates were similar between the 2 groups, with 10.5% of patients admitted in the EP model and 11.7% of patients admitted in the CPU model (P = .430; Fig. 2). There were no deaths in the study, and no patient in either group experienced MI or underwent unplanned revascularization during the 30-day follow-up.

Discussion

In evaluating patients who present to the ED with chest pain, it is critical to appropriately assess and diagnose acute coronary syndrome. Even as improved testing modalities arise, a substantial number of pa- tients who present to the ED with an MI are sent home [1-6]. The EDOU has become a common feature of the ED in recent years to help address this population. Although used for many complaints, chest pain is the most commonly reported reason for placement within the observation unit [7]. These units provide an area to monitor low- and intermediate-risk patients who present to the ED with chest pain while awaiting diagnostic results to further stratify risk and determine Optimal treatment. Multiple studies have shown observation units to be cost-effective when compared with hospital admission and equally as effective in diagnosis and treatment [8-11].

Although hospitals typically use rigorous protocols for placement and discharge from these units, these protocols often vary between in- stitutions. Chest pain protocols typically include provocative testing for patients who do not demonstrate initial ischemia or infarction. As ob- servation unit models differ in terms of management, ranging from car- diac nurses to physician assistants, EPs, and cardiologists, so does the decision making regarding which patients will receive provocative test- ing. Historically, the liberal use of provocative testing has been used for patients in the EDOU with chest pain [12-15]. However, recent studies have shown that physician discretion by ED physicians in conjunction with cardiology consultation may decrease the numbers of provocative

20

10

0

Low-risk patients Moderate-risk

patients

CPU model

tests without an increase in adverse events [16,17]. Additional studies have suggested that unnecessary provocative testing leads to higher rates of unnecessary intervention and adverse outcomes [18-21].

We found a significant reduction in the rates of provocative testing in the EDOU after the implementation of mandatory cardiology consulta- tion. It is worth noting that this reduction was primarily among low- risk patients, which would seem to be the patient group in which

Fig. 1. Comparison of provocative testing rates. Difference between low-risk patients was significant (P = .011) but was not significant between moderate-risk patients (P = .087).

selective use of provocative testing would be most appropriate. This re- duction in provocative testing was not offset by a higher rate of primary

28 T. Madsen et al. / American Journal of Emergency Medicine 35 (2017) 25-28

cardiac catheterization in the EDOU, in which cardiologists may have performed this testing modality rather than opting for less invasive pro- vocative testing measures.

We did find it notable that patients placed in the EDOU during the pe- riod of mandatory cardiology consultation were more likely to be moder- ate risk by TIMI score. Emergency physicians placed patients in the EDOU at their discretion. It is possible that physicians may have been more com- fortable placing moderate-risk patients in the EDOU knowing that they would be seen by a cardiologist, where they may have been more likely to admit these patients to an Inpatient unit in the period in which cardiol- ogists did not consult on all EDOU patients. Despite this increase in moderate-risk patients, the overall rate of provocative testing decreased in the period of mandatory cardiology consultation, and the rates among moderate-risk patients were reduced (but not statistically significant).

Although this study was not designed to assess cost-effectiveness, it suggests the potential for additional analysis related to various EDOU staffing models and the cost of care for low- and intermediate-risk pa- tients with chest pain. Consultation by specialist services often incurs additional cost, but, as we found, having cardiology consult on all pa- tients in the EDOU could decrease unnecessary testing, thus potentially decreasing cost. This would be particularly true in low-risk populations. It remains to be seen whether reduced costs in testing and intervention offset the cost associated with the higher level of care with regular spe- cialty consultation in the EDOU. In addition, we did find that cardiolo- gists were more likely to order myocardial perfusion imaging over stress echocardiography or coronary CTA, with this more expensive test possibly offsetting cost savings from decreased testing rates.

Our study contains several limitations. The results of our study rep- resent the population of a single center, and the demographics and risk factors of this population cannot be directly translated to those of other areas. Institutional practices, particularly as these relate to patient selection for provocative testing, may vary between institutions, and it is unclear whether our results would translate to other centers.

We recorded study data from direct questioning of patients, reviewing the EMR, and a 30-day telephone follow-up, which convey inherent limitations. Accuracy of patient understanding as well as mem- ory of events presents the possibility for recall bias. In addition, only 68.2% of patients were contacted by telephone. For those unavailable by telephone, review of the electronic medical record as well as the So- cial Security Death Index was used to assess 30-day outcome. However, in the event of an adverse outcome treated at an outside facility, the re- sults would not be available for assessment within this study.

Patterns of patient participation also introduce the potential for bias. Inconsistencies in records related to refusal of participation did not allow us to report an overall participation rate of eligible subjects; anec- dotally, nearly all eligible chest pain patients in the EDOU agreed to be part of the study. Without knowing exact numbers of those who refused to participate, however, we acknowledge that this may affect the inter- pretation of study findings and leave in question the potential impact of outcomes missed among these individuals.

The staffing model of the observation unit at this facility, including the utilization of APCs in association with emergency medicine attend- ing physicians, might differ greatly from other centers. In addition, the characteristics of a teaching hospital may affect decision making. The ready availability of a cardiology service within this academic setting may have affected both the willingness of the service to regularly con- sult on observation patients as well as the ease with which provocative testing was obtained. This may have affected our stress testing rates, particularly when compared with a community facility which may not have comparable access to regular in-house cardiology consultation.

Additional limitations may relate to our study design. We collected data before and after a staffing model change, rather than performing a randomized study in which cardiology consultation may have been man- datory on certain days while the EP primarily made provocative testing de- cisions on other days. Given the increasing attention to overtesting among patients with chest pain, it is possible that the reduction in provocative

testing rates may have related more to the temporal nature of the CPU model rather than to the mandatory cardiology consultation. Our study de- sign did not allow for determination whether the EP model, if continued, would have led to reduced provocative testing rates due to increased awareness of the adverse effects of unnecessary provocative testing.

In summary, we found that implementation of mandatory cardiology consultation resulted in a reduction in rates of provocative testing among patients evaluated for chest pain in the EDOU. This reduction was partic- ularly notable among low-risk chest pain patients. We did not find that the reduced rate of provocative testing was due to a higher rate of patients going directly to cardiac catheterization. The reduction in provocative testing did not lead to adverse outcomes or missed diagnoses among pa- tients. The cost-effectiveness of such a model remains to be seen.

References

1. Lloyd-Jones D, Adams R, Carnethon M, De Simone G, Ferguson TB, Flegal K, et al. Heart disease and stroke statistics-2009 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circula- tion 2009;119(3):480-6.
2. Lee TH, Rouan GW, Weisberg MC, Brand DA, Acampora D, Stasiulewicz C, et al. Clin- ical characteristics and natural history of patients with acute myocardial infarction sent home from the emergency room. Am J Cardiol 1987;60(4):219-24.
3. McCarthy BD, Beshansky JR, D’Agostino RB, Selker HP. Missed diagnoses of acute myocardial infarction in the emergency department: results from a multicenter study. Ann Emerg Med 1993;22(3):579-82.
4. Pope JH, Aufderheide TP, Ruthazer R, Woolard RH, Feldman JA, Beshansky JR, et al. Missed diagnoses of Acute cardiac ischemia in the emergency department. N Engl J Med 2000;342(16):1163-70.
5. Schull MJ, Vermeulen MJ, Stukel TA. The risk of missed diagnosis of acute myocardial infarction associated with emergency department volume. Ann Emerg Med 2006; 48(6):647-55.
6. Vukmir RB. Medical malpractice: managing the risk. Med Law 2004;23(3):495-513.
7. Osborne A, Weston J, Wheatley M, O’Malley R, Leach G, Pitts S, et al. Characteristics of hospital observation services: a society of cardiovascular patient care survey. Crit Pathw Cardiol 2013;12(2):45-8.
8. Goodacre S, Nicholl J, Dixon S, Cross E, Angelini K, Arnold J, et al. Randomised controlled trial and economic evaluation of a chest pain observation unit compared with routine care. BMJ 2004;328(7434):254.
9. Farkouh ME, Smars PA, Reeder GS, Zinsmeister AR, Evans RW, Meloy TD, et al. A clin- ical trial of a chest-pain observation unit for patients with unstable angina. Chest Pain Evaluation in the Emergency Room (CHEER) Investigators. N Engl J Med 1998;339(26):1882-8.
10. Storrow AB, Gibler WB. Chest pain centers: diagnosis of acute coronary syndromes. Ann Emerg Med 2000;35(5):449-61.
11. Roberts RR, Zalenski RJ, Mensah EK, Rydman RJ, Ciavarella G, Gussow L, et al. Costs of an emergency department-based accelerated Diagnostic protocol vs hospitalization in patients with chest pain: a randomized controlled trial. JAMA 1997;278(20):1670-6.
12. Anderson JL, Adams CD, Antman EM, Bridges CR, Califf RM, Casey Jr DE, et al. ACC/ AHA 2007 guidelines for the management of patients with unstable angina/non- ST-elevation myocardial infarction. J Am Coll Cardiol 2007;50(7):e1-157.
13. Gomez MA, Anderson JL, Karagounis LA, Muhlestein JB, Mooers FB. An emergency department-based protocol for rapidly ruling out myocardial ischemia reduces hos- pital time and expense: results of a randomized study (ROMIO). J Am Coll Cardiol 1996;28(1):25-33.
14. Amsterdam EA, Kirk JD, Diercks DB, Lewis WR, Turnipseed SD. Immediate exercise testing to evaluate low-risk patients presenting to the emergency department with chest pain. J Am Coll Cardiol 2002;40(2):251-6.
15. Amsterdam EA, Kirk JD, Diercks DB, Lewis WR, Turnipseed SD. Exercise testing in chest pain units: rationale, implementation, and results. Cardiol Clin 2005;23(4):503-16 [vii].
16. Napoli AM, Arrighi JA, Siket MS, Gibbs FJ. Physician discretion is safe and may lower stress test utilization in emergency department chest pain unit patients. Crit Pathw Cardiol 2012;11(1):26-31.
17. Napoli AM. The association between pretest probability of coronary artery disease and stress test utilization and outcomes in a chest pain observation unit. Acad Emerg Med 2014;21(4):401-7.
18. Foy AJ, Liu G, Davidson Jr WR, Sciamanna C, Leslie DL. Comparative effectiveness of diagnostic testing strategies in emergency department patients with chest pain: an analysis of downstream testing, interventions, and outcomes. JAMA Intern Med 2015;175(3):428-36.
19. Aldous S, Richards AM, Cullen L, Pickering JW, Than M. The incremental value of stress test- ing in patients with acute chest pain beyond serial cardiac Troponin testing. Emerg Med J 2016;33(5):319-24.
20. Diercks DB, Mumma BE, Frank Peacock W, Hollander JE, Safdar B, Mahler SA, et al. Incremental value of objective cardiac testing in addition to physician impression and serial contemporary troponin measurements in women. Acad Emerg Med 2013;20(3):265-70.
21. Poldervaart JM, Six AJ, Backus BE, de Beaufort HW, Cramer MJ, Veldkamp RF, et al. The predictive value of the exercise ECG for Major adverse cardiac events in patients who presented with chest pain in the emergency department. Clin Res Cardiol 2013; 102(4):305-12.