Fast assessment and management of chest pain patients without ST-elevation in the pre-hospital gateway (FamouS Triage): ruling out a myocardial infarction at home with the modified HEART score

Abstract

Background:

The first study of the FamouS Triage project investigates the feasibility of ruling out a myocardial infarction in pre-hospital chest pain patients without electrocardiographic ST-segment elevation by using the modified HEART score at the patient’s home, incorporating only a single highly sensitive troponin T measurement.

Methods:

A venous blood sample was drawn in the ambulance from 1127 consecutive chest pain patients for measurement of the pre-hospital highly sensitive troponin T levels, in order to establish a pre-hospital HEART score (i.e. the modified HEART score) and evaluate the possibility of triage at the patient’s home. The primary endpoint was the occurrence of a major adverse cardiac event (MACE) i.e. acute myocardial infarction, percutaneous coronary intervention, coronary artery bypass grafting or death within 30 days after initial presentation.

Results:

Two hundred and six patients (18%) developed a MACE during 30 days of follow-up. Thirty-six per cent of the patients (n=403) had a low modified HEART score (0–3 points) and none of them developed a MACE during follow-up. Forty-four per cent of the patients (n=494) had an intermediate modified HEART score (4–6 points) and 18% of them developed a MACE. Twenty per cent of the patients (n=230) had a high modified HEART score (7–10 points) of which 52% developed a MACE during follow-up.

Conclusion:

It seems feasible to rule out a myocardial infarction at home in chest pain patients without ST-segment elevation by using the modified HEART score.

TRIAL ID: NTR4205. Dutch Trial Register [http://www.trialregister.nl]: trial number 4205.

Keywords

Pre-hospital triage chest pain non-ST-segment elevation acute coronary syndrome acute myocardial infarction modified HEART score highly sensitive troponin

Introduction

Acute myocardial infarction (AMI) is globally known to be a major cause of disability and death.¹ In Europe and the United States alone 15–20 million patients per year present to the emergency department (ED) with sudden onset of chest pain and symptoms suggestive of AMI. Approximately 10% of all ED consultations are of patients presenting with symptoms indicative of AMI.² The 12-lead electrocardiogram (ECG) and cardiac troponins (cTn) along with clinical evaluation form the diagnostic cornerstone for AMI.³ While ST-segment elevation myocardial infarction (STEMI) provides a straightforward diagnosis in most patients based on ECG changes and clinical evaluation, STEMI patients only represent 5% of all patients presenting with acute chest pain.¹ Therefore, based on the ECG alone, an initial diagnostic uncertainty remains in the majority of patients presenting with acute chest pain and these patients usually undergo investigations at the ED for about 4–8 hours before an AMI can safely be excluded. Therefore, risk stratification in chest pain patients prior to transportation to the hospital would be of great use in this large group of patients.

While there are several risk stratification tools available for patients with a confirmed diagnosis of acute coronary syndrome (ACS), for example the TIMI and the GRACE risk scores,^4,5 the so-called HEART score (see Appendix 1 for HEART score algorithm) is specifically aimed at undifferentiated chest pain patients and has proved to be an easy, quick and effective risk stratification tool in this patient population.^6-8 Based on the patients’ history, ECG, age, cardiovascular risk factors and cTn measurements, a score between 0 and 10 points is calculated, representing the patients’ risk of developing a major adverse cardiac event (MACE) within 6 weeks after initial presentation.⁸ However, the HEART score has only been validated for use at the ED and not in a pre-hospital setting prior to hospital transportation. Regarding cTn, the HEART score also incorporates conventional (4th generation) cTn measurements, while novel assays, the so-called 5th generation or ‘highly sensitive’ cardiac troponins (hs-cTn), offer greater diagnostic accuracy concerning the diagnosis of AMI and even at an earlier stage compared to conventional cTn.⁹ It would be interesting to investigate whether it is feasible to rule out an AMI at the earliest moment possible, i.e. at first medical contact (FMC) at the patients’ home before a decision for a hospital transfer is made by ambulance paramedics. Ruling out an AMI in a pre-hospital setting might be feasible by combining a pre-hospital hs-cTnT measurement with an easy to use clinical risk score like the HEART score. We have modified the original HEART score by replacing the in-hospital 4th generation cTn measurements by a single pre-hospital hs-cTnT measurement, thus creating the ‘modified HEART score’, which is thought to have a better diagnostic accuracy in a pre-hospital setting. In this first study of the FamouS Triage project we sought to investigate whether it is feasible to rule out an AMI at FMC in a pre-hospital setting using the modified HEART score.

Methods

Study design and participants

The design and sample size calculation of the fast assessment and management of chest pain without ST-elevation in the pre-hospital gateway (FamouS Triage) project has been described previously.¹⁰ In brief, the FamouS Triage project is a prospective, multicentre and multiphase study, performed in The Netherlands to assess a new pre-hospital triage strategy in chest pain patients suggestive of AMI, by using the HEART score in combination with a single hs-cTnT measurement. This first prospective, observational study of the Famous Triage project was conducted in the ambulance region of Zwolle in The Netherlands, covering an area of 300,000 inhabitants. The study involved 28 emergency medical services (EMS) vehicles of the regional ambulance service (RAV IJsselland) with approximately 90 ambulance paramedics (registered nurses, specialised in pre-hospital care). Two hospitals participated in this study: Isala Zwolle and Deventer Ziekenhuis, both in the province of Overijssel in The Netherlands. The first FamouS Triage study has been completed and its results are presented in this paper.

Pre-hospital patient selection and data acquisition

The inclusion and exclusion criteria of the FamouS Triage study are presented in Appendix 2 of this paper.¹⁰ Briefly, all men or non-pregnant women aged 18 years or older, who called the EMS between June 2012 and December 2014 due to chest pain, were prospectively evaluated at FMC by ambulance paramedics according to a standard pre-hospital chest pain protocol. All patients underwent routine assessment including a brief history, monitoring of vital signs (blood pressure, heart rate, pulse oximetry), 12-lead ECG and intravenous access. Patients with chest pain suggestive of AMI without ST-segment elevation on the ECG, but who were to be transported to the ED, were eligible for study participation. In participating patients, trained ambulance nurses acquired intravenous access, as is required for usual care, and before any intravenous administration of substances (saline or medication) a venous blood sample was drawn from the intravenous cannula into a lithium heparin blood collection tube (BD Vacutainer) and kept separate until hospital arrival (all 28 EMS vehicles were equipped with the required blood collection tubes). Assessment of the HEART score at the patients’ home or in the ambulance was not mandatory for study participation. Due to the observational character of this study, all participating patients were transported to the ED of the nearest hospital and managed according to standard care as any other non-participating chest pain patient with a suspicion of non-ST-segment elevation acute coronary syndrome (NSTE-ACS).

In-hospital patient inclusion and calculation of the modified HEART score at the ED

Upon arrival at the ED the pre-hospital venous blood sample was sent to the clinical chemistry laboratory for measuring the pre-hospital hs-cTnT value. The pre-hospital hs-cTnT test result was not disclosed to the physicians in the ED in order to adhere to standard care with clinical decision-making based on a full history, physical examination, pulse oximetry, 12-lead ECG, telemetry and the results of in-hospital (serial) representative measurement(s) of hs-cTnT. Therefore, both ambulance paramedics and emergency physicians involved in this study were blinded to the pre-hospital hs-cTnT result. This study complies with the Declaration of Helsinki and did not require an informed consent procedure for study participation as the medical ethics committee approved the study as observational and considered all study activities to fit within the boundaries of usual care.

Calculating the modified HEART score

Based on the medical records of participating patients, the clinical components of the HEART score (i.e. history, ECG, age and risk factors) were assessed by two cardiology residents assigned to the study. For assessment of the biochemical component of the HEART score (troponin), the pre-hospital hs-cTnT value was assessed, yielding the modified (or pre-hospital) HEART score. The venous blood samples taken in the ambulance were analysed at the clinical chemistry laboratory using the hs-cTnT Modular and Cobas 8000 system of Roche Diagnostics, Manheim Germany, with a limit of detection of 0.003 ng/mL and inaccuracy corresponding to a 10% coefficient of variation at the 99th percentile upper reference limit (URL) of the reference population. A hs-cTnT value exceeding the 99th percentile URL (0.014 ng/mL) was considered elevated. The hs-cTnT results were not corrected for haemolysis. The five components of the modified HEART score were assessed according to the same criteria of the original HEART score where each component was awarded 0, 1 or 2 points, resulting in a total score ranging from 0 to 10 points. Risk categorisation into low, intermediate or high risk for MACE was also based on the same cut-off points as in the original HEART score (0–3 points = low risk; 4–6 points = intermediate risk and 7–10 points = high risk).

Outcome measures

Based on medical records of participating patients, follow-up was performed in terms of the occurrence of a MACE within 30 days after initial presentation, which was the primary outcome of the study. A MACE was defined as an AMI, a percutaneous coronary intervention (PCI), coronary artery bypass grafting (CABG) or death. The diagnosis of AMI was defined according to the third universal definition of myocardial infarction.¹¹ A PCI was defined as a percutaneous strategy of therapeutic intracoronary balloon dilatation and/or stent placement. A CABG was defined as an open surgical procedure of coronary artery grafting as to bypass a coronary stenosis.

In case an AMI was ruled out in the ED in patients of the low-risk category (modified HEART score 0–3), the discharge diagnosis was evaluated based on the hospital medical records. This was done to investigate whether patients with a modified HEART score of 0–3 points could have been left at home after FMC without worrying about a life-threatening (non-cardiac) cause of chest pain.

Statistical analysis

We refer to our design paper for the sample size calculation of this study cohort.¹⁰

Statistical analysis was performed with IBM SPSS Statistics for Windows, version 22.0. Descriptive statistics are given as number (percentage), as average (±SD) or median (interquartile range). For the assessment of differences between subgroups the one-way analysis of variance test was used when continuous data had a normal distribution, otherwise the Kruskall–Wallis test was used to assess subgroup differences. For assessment of differences between subgroups concerning categorical data the Pearson’s chi-square test was used. When contingency tables did not meet the required criteria of the chi-square test (‘no more than 20% of the expected counts are less than 5 and all individual expected counts are 1 or greater’), the Fisher’s exact test was used. Endpoint probabilities within risk categories of the modified HEART score were calculated as the percentage of cases reaching an endpoint within the same risk category. Statistical significance was defined as P<0.05 two-sided.

Results

Population

Between June 2012 and December 2014 a total of 1426 consecutive chest pain patients were enrolled in the ambulance by paramedics. At the ED two cardiology residents who were assigned to the study performed further screening of these patients based on medical records. Patients who met the inclusion criteria (see Appendix 2) were enrolled chronologically in order to prevent selection bias, until the powered number of 1127 patients was reached by December 2014. A 30-day follow-up after initial presentation on the occurrence of the primary endpoint (MACE) was performed in all 1127 patients and all study patients were eligible for analysis.

Patient characteristics

Patient characteristics are presented in Table 1. The average age of the entire study cohort was 63.8±14.4 years (mean ±SD) and 57.7% of the patients were men. Four hundred and six patients (36%) had a history of atherosclerotic disease at baseline (i.e. prior coronary artery disease, myocardial infarction, revascularisation, peripheral artery disease or stroke). Three hundred and thirteen patients (27.8%) had an elevated pre-hospital hs-cTnT value (hs-cTnT >0.014 ng/mL). The median time interval between symptom onset and pre-hospital blood sampling of the entire study population was 2.5 hours and there were no significant differences in the median time intervals between the three risk groups (see Table 1).

Table 1.

Patient characteristics at baseline (overall and per risk category).

	All patients	Modified HEART score0–3	Modified HEART score4–6	Modified HEART score7–10	P value
Population, n (% of study population)	1127 (100%)	403 (35.8%)	494 (43.8%)	230 (20.4%)
Age, years (SD)	63.8 (14.4)	53.5 (12.2)	67.0 (11.9)	75.1 (10.8)	<0.001
Male, n (%)	650 (57.7%)	209 (51.9%)	295 (59.7%)	146 (63.5%)	0.008
Prior atherosclerotic disease,^a n (%)	406 (36.0%)	32 (7.9%)	226 (45.7%)	148 (64.3%)	<0.001
Pre-hospital hs-cTnT >0.014 ng/mL, n (%)	313 (27.8%)	5 (1.2%)	118 (23.9%)	190 (82.6%)	<0.001
Median time from symptom onset to blood sampling, hh:mm:ss (Q1–Q3)	2:30:00(01:03:00–06:39:00)	02:18:00(01:03:00–06:39:00)	02:45:00(01:10:00–07:06:15)	02:33:30(01:23:00–07:30:00)	0.232

History of coronary artery disease, previous myocardial infarction, prior revascularisation, peripheral artery disease or previous stroke.

hs-cTnT: highly sensitive cardiac troponin T; SD: standard deviation; hh:mm:ss: hours:minutes:seconds; Q1–Q3: first quartile and third quartile.

Four hundred and three patients (35.8%) were calculated as being in the low-risk category of the modified HEART score (0–3 points), 494 patients (43.8%) were in the intermediate-risk category (4–6 points) and 230 patients (20.4%) were in the high-risk category (>6 points). Patients stratified to the intermediate-risk category were significantly older than patients in the low-risk group (67.0 vs. 53.5 years; P<0.001) and patients in the high-risk category were older than patients in the intermediate-risk group (75.1 vs. 67.0 years; P<0.001). A higher risk category of the modified HEART score also had significantly more male patients, more patients with a prior atherosclerotic disease and a higher proportion of patients with an elevated pre-hospital hs-cTnT value (see Table 1).

Primary endpoint

A total of 206 patients developed a MACE during follow-up, which means an 18% pre-test probability of developing a MACE in this specific population (chest pain without ST-segment elevation in the ambulance).

In the low-risk category (n=403) five patients (1.2%) had an elevated pre-hospital hs-cTnT value; however, none of the patients in the low-risk category developed a MACE during follow-up. In the intermediate-risk group (n=494) 118 patients (23.9%) had an elevated pre-hospital hs-cTnT value, of which 87 patients (17.6%) developed a MACE during follow-up. In the high-risk group (n=230) 190 patients (82.6%) had an elevated pre-hospital hs-cTnT value and 119 (51.7%) of these high-risk patients developed a MACE. The occurrence of the primary endpoint (MACE) within each risk category of the modified HEART score, as well as the occurrence of the separate MACE components (AMI, PCI, CABG, death) all differed significantly between the three risk categories. An overview of these results is presented in Table 2.

Table 2.

Endpoints within 30 days after evaluation at the ED.

Modified HEART score				P value
	0–3 points(n=403)	4–6 points(n=494)	7–10 points(n=230)
Pre-hospital hs-cTnT>0.014 ng/mL, n (%)	5 (1.2%)	118 (23.9%)	190 (82.6%)	<0.001
MACE (AMI, PCI, CABG or death), n (%)	0 (0.0%)	87 (17.6%)	119 (51.7%)	<0.001
ACS, n (%)	0 (0.0%)	76 (15.4%)	116 (50.4%)	<0.001
AMI, n (%)	0 (0.0%)	54 (10.9%)	101 (43.9%)	<0.001
PCI, n (%)	0 (0.0%)	35 (7.1%)	46 (20.0%)	<0.001
CABG, n (%)	0 (0.0%)	18 (3.6%)	33 (14.3%)	<0.001
Death, n (%)	0 (0.0%)	4 (0.8%)	5 (2.2%)	0.013

ACS: acute coronary syndrome; AMI: acute myocardial infarction; CABG: coronary artery bypass grafting; hs-cTnT: highly sensitive cardiac troponin T; MACE: major adverse cardiac event; PCI: percutaneous coronary intervention.

Proportion of patients with MACE in each separate modified HEART score

In Table 3 all study patients are clustered in subgroups of corresponding outcomes of the modified HEART score. MACE only occurred in patients with a modified HEART score of 4 or higher and the proportion of patients developing MACE rises significantly (P<0.001) with each higher category of the modified HEART score. Figures 1 and 2 display the proportion of patients developing MACE within each outcome of the modified HEART score.

Table 3.

Proportion of patients developing MACE per separate modified HEART score.

Modified HEART score	Number of patients	Patients developing MACE (% within subgroup)
0	13	0 (0.0%)
1	57	0 (0.0%)
2	149	0 (0.0%)
3	184	0 (0.0%)
4	178	10 (5.6%)
5	170	34 (20.0%)
6	146	43 (29.5%)
7	102	39 (38.2%)
8	89	53 (59.6%)
9	34	23 (67.6%)
10	5	4 (80.0%)
Total	1127	206 (18.3%)

MACE: major adverse cardiac event.

Cause of chest pain within the low-risk category of the modified HEART score

Three hundred and seventy patients of the low-risk category (92%) were discharged directly from the ED and most of them had a benign non-cardiac chest pain. The causes of chest pain of the patients who were discharged from the ED are presented in Figure 3. Thirty-three patients of the low-risk category were hospitalised for further clinical observation, extra diagnostics or treatment, of which only two patients (0.4%) required urgent clinical treatment, one for pulmonary embolism and one for acute pancreatitis. The causes of chest pain of the hospitalised patients are presented in Figure 4.

Discussion

Based on the results of our first observational study of the FamouS Triage project it seems feasible to rule out safely an AMI in a pre-hospital setting using the modified HEART score at the home of patients with chest pain without ST-segment elevation on their ECG. Our study is the first study to demonstrate the feasibility of ruling out an AMI in patients presenting with symptoms suspected of NSTE-ACS using a pre-hospital risk stratification tool and we report several novel findings.

First, more than a third (36%) of our study patients calling the EMS for chest pain suggestive of AMI had a modified HEART score of 0–3 points (low-risk group) and did not develop any MACE during a follow-up period of 30 days. This number of low-risk chest pain patients (36%) is similar to the number of low-risk chest pain patients with a HEART score of 0–3 points in the validation study of the original HEART score by Backus et al.⁸ However, by using our modified HEART score it seems feasible to identify these low-risk chest pain patients even before transporting the patient to the ED using a single hs-cTnT test.

Second, a substantial number of patients (20%) at high risk for developing a MACE, as reflected by a 52% MACE and 2% mortality rate in the group with a modified HEART score of 7–10 points, can be identified earlier by assessing the modified HEART score at the patients’ home. Pre-hospital identification of high-risk chest pain patients may change their pre-hospital pathway, directing their transfer to a hospital with PCI capability to get the treatment earlier, which would also reduce the amount of expensive interhospital transfers of high-risk chest pain patients from non-interventional hospitals to interventional hospitals. In the high-risk group a PCI was performed in only 20% and a CABG in 14% of the patients (Table 2), which means the modified HEART score is more suitable for ruling out an AMI, but also capable of narrowing down high-risk patients into a subgroup of 20% of the chest pain population. The ability of the modified HEART score to select chest pain patients in a pre-hospital setting who are at intermediate or high risk for MACE could enable cost-effective use of non-invasive diagnostic testing, for example the use of coronary computed tomography angiography (CCTA). CCTA could increase the in-hospital diagnostic accuracy of the modified HEART score in patients stratified to an intermediate or high risk for MACE.¹² The use of CCTA in patients with a modified HEART score of 0–3 points (low-risk group) may not be cost-effective based on the 0% event rate of this group.

Third, only two patients of the low-risk category (0.4%) required hospital admission for treatment of a life-threatening cause of chest pain (Figure 4). Due to the observational character of this study, all patients were transferred to the ED after FMC in the ambulance and therefore these two patients with pulmonary embolism and acute pancreatitis were diagnosed appropriately at the ED, after which they received successful treatment. If the modified HEART score was already implemented in the ambulance to identify low-risk chest pain patients, based on their clinical status these two serious patients would have been transported to the hospital anyway, despite scoring a low modified HEART score. Using a pre-hospital risk score for chest pain patients does not mean that ambulance paramedics will stop using their clinical judgement.

This first prospective study of the FamouS Triage project shows that the modified HEART score might provide an excellent tool in risk stratifying patients with undifferentiated chest pain in a pre-hospital setting. Having a simple, effective and validated pre-hospital risk stratification tool in this large group of patients prior to presentation to the ED could potentially lead to several benefits. Identifying low-risk chest pain patients in a pre-hospital setting and avoiding unnecessary hospital transfers would reduce ED overcrowding and stimulate a more cost-effective pathway by managing these low-risk patients in a primary care setting instead of at the ED, which may also reduce patient anxiety and avoid the use of empirical treatment with all its potential harmful side-effects. Further, using pre-hospital risk stratification could enable healthcare professionals to consider alternative diagnoses in low-risk chest pain patients sooner, which may lead to improvement of patient care and hospital logistics.

Limitations prior to implementation

In this first FamouS Triage study the modified HEART score was established by cardiology residents and not by ambulance paramedics at FMC in the ambulance. Assessment of the HEART score components in the ambulance by paramedics was not mandatory in this study, as the main responsibility of ambulance paramedics was to obtain pre-hospital venous blood samples for the measurement of pre-hospital hs-cTnT values in all patients. In the next studies of the FamouS Triage project the assessment of the HEART score in the ambulance by paramedics is mandatory, which will make it possible to study the interobserver variability between HEART score assessment by ambulance paramedics and assessment by cardiology residents in the ED.

Another limitation prior to implementation is the use of pre-hospital venous blood sampling instead of using a point-of-care (POC) hs-cTnT test. In this study ambulance paramedics collected a venous blood sample from every patient and took the sample along with the patient to the hospital, where the venous blood sample was analysed for the pre-hospital hs-cTnT value. This was the only way to have a pre-hospital hs-cTnT value as a component of the pre-hospital HEART score, because unfortunately to this date, there is no POC device available for hs-cTnT testing. In the next study of the FamouS Triage project ambulance paramedics will get to practise with POC testing of cTn, for it will be mandatory to assess the entire HEART score in the ambulance in order to include the patient. As stated, in the absence of hs-cTnT POC devices, the available 4th generation cTn POC devices will be used.

A general limitation related to troponin testing concerns the time interval between symptom onset and blood sampling. As troponin has a delayed release after myocardial injury, there is the risk of false negative test results when troponin is tested shortly after symptom onset. As demonstrated in Table 1 the median time interval between symptom onset and pre-hospital blood sampling was 02:30 hours, resulting in 29% (45 of 155 AMI patients) of false negative pre-hospital hs-cTnT test results (see Appendix 3). Appendix 3 shows the poor post-test probabilities of isolated pre-hospital hs-cTnT testing regarding the diagnosis of AMI. However, in the modified HEART score hs-cTnT testing is only one of the five components and therefore risk stratification is not relying on troponin testing alone. The main goal of the modified HEART score is to rule out an AMI at an early stage rather than ruling in of AMI, and based on our results this seems successful, as reflected by the absence of any MACEs in the low-risk group.

Finally, the inclusion of patients was solely left at the discretion of ambulance paramedics, hence the element of selection bias cannot be fully excluded. However, inclusion was further selected at the ED by cardiology residents according to pre-defined inclusion and exclusion criteria as published in our design paper.¹⁰ Therefore, our cohort does represent consecutive eligible undifferentiated chest pain patients in whom the modified HEART score can be applied appropriately.

Conclusion

In patients who call the EMS with chest pain, it is feasible to rule out an AMI at FMC in a pre-hospital setting by assessment of the modified HEART score incorporating a single pre-hospital hs-cTnT measurement. An effective pre-hospital triage strategy that could reliably identify chest pain patients with a low probability of having an AMI could lead to a reduction of unnecessary empirical treatments, overcrowding at the ED, healthcare costs and unnecessary patient anxiety.

Footnotes

Appendix

Appendix 3.

Diagnostic characteristics of a single pre-hospital hs-cTnT measurement regarding AMI.Contingency table.

	AMI	No AMI	Total
Positive hs-cTnT test	110	203	313
Negative hs-cTnT test	45	769	814
Total	155	972	1127

Sensitivity: 71.0%; specificity: 79.1%; positive predictive value: 35.1%; negative predictive value: 94.5%; positive likelihood ratio: 3.40 (95% CI, 2.90–3.98); negative likelihood ratio: 0.37 (95% CI 0.29–0.47).

Acknowledgements

The authors would like to thank Mireille Edens (clinical epidemiologist at Isala Zwolle, The Netherlands) and David de Jong (data manager at Isala Zwolle, The Netherlands) for their assisting contributions to the FamouS Triage study.

Conflict of interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This first study of the FamouS Triage project has been completed without extramural funding.

References

Mueller

New ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Swiss Med Wkly 2012; 142: w13514.

Nawar

Niska

National Hospital Ambulatory Medical Care Survey: 2005 emergency department summary. Adv Data 2007; 386: 1–32.

Roffi

Patrono

Collet

J-P

. 2015 ESC Guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 2016; 37: 267–315.

Antman

Cohen

Bernink

. The TIMI risk score for unstable angina/non-ST elevation MI: a method for prognostication and therapeutic decision making. JAMA 2000; 284: 835–842.

Granger

Goldberg

Dabbous

.; Global Registry of Acute Coronary Events Investigators. Predictors of hospital mortality in the global registry of acute coronary events. Arch Intern Med 2003; 163: 2345–2353.

Six

Backus

Kelder

JC.

Chest pain in the emergency room: value of the HEART score. Neth Heart J 2008; 16: 191–196.

Backus

Six

Kelder

. Chest pain in the emergency room: a multicenter validation of the HEART Score. Crit Pathw Cardiol 2010; 9: 164–169.

Backus

Six

Kelder

. A prospective validation of the HEART-score for chest pain patients at the emergency department. Int J Cardiol 2013; 168: 2153–2158.

Reichlin

Hochholzer

Bassetti

. Early diagnosis of myocardial infarction with sensitive cardiac troponin assays. N Engl J Med 2009; 361: 858–867.

10.

Ishak

Ali

Fokkert

Slingerland

. Fast assessment and management of chest pain without ST-elevation in the pre-hospital gateway: rationale and design. Eur Heart J Acute Cardiovasc Care 2015; 4: 129–136.

11.

Thygesen

Alpert

Jaffe

. Third universal definition of myocardial infarction. Eur Heart J 2012; 33: 2551–2567.

12.

D’Ascenzo

Cerrato

Biondi-Zoccai

. Coronary computed tomographic angiography for detection of coronary artery disease in patients presenting to the emergency department with chest pain: a meta-analysis of randomized clinical trials. Eur Heart J Cardiovasc Imaging 2013; 14: 782–789.

13.

HEART-score algorithm [File Name: Astellasback – table.pdf]. http://www.heartscore.nl/resources/table.pdf (accessed 13 March 2014).