confidential: for review only - bmj...2015/02/17 · manuscript id bmj.2017.037996 article type:...

Confidential: For Review O

nly

Non-selective high-sensitivity cardiac troponin testing in

the Emergency Department

Journal: BMJ

Manuscript ID BMJ.2017.037996

Article Type: Research

BMJ Journal: BMJ

Date Submitted by the Author: 15-Feb-2017

Complete List of Authors: Shah, Anoop; Univerity Of Edinburgh, Centre of Cardiovascular Sciences Noaman, Ala; University of Edinburgh College of Medicine and Veterinary Medicine Vaswani, Amar; University of Edinburgh, Gibbins, Matthew; University of Edinburgh College of Medicine and

Veterinary Medicine Griffiths, Megan; University of Edinburgh College of Medicine and Veterinary Medicine Chapman, Andrew; University of Edinburgh College of Medicine and Veterinary Medicine Strachan, Fiona; University Of Edinburgh, Anand, Atul; University of Edinburgh College of Medicine and Veterinary Medicine Adamson, Philip; University of Edinburgh College of Medicine and Veterinary Medicine D'Souza, Michelle; University of Edinburgh College of Medicine and Veterinary Medicine

Gray, Alasdair; Royal Infirmary of Edinburgh, Emergency Medicine Research Group, Department of Emergency Medicine McAllister, David; University of Glasgow School of Medicine Newby, David; University of Edinburgh, BHF/University Centre for Cardiovascular Sciences Mills, Nicholas; University of Edinburgh, BHF/University Centre for Cardiovascular Sciences

Keywords: troponin, high-sensitivity troponin, positive predictive value, prevalence, emergency department, acute coronary syndrome, myocardial infarction

https://mc.manuscriptcentral.com/bmj

BMJ


nly

1

ORIGINAL ARTICLE

Non-selective high-sensitivity cardiac troponin testing

in the Emergency Department

Anoop SV Shah MD MPH PhD,1 Ala Noaman MD,

1 Amar Vaswani MD,

1 Mathew Gibbins

MSc,1 Megan Griffiths MD,

1 Andrew R Chapman MD,

1 Fiona S Strachan PhD,

1 Atul Anand

MD,1

Philip D Adamson MD1, Michelle D Souza,

1 Alasdair J Gray MD,

3,4 David A

McAllister MD,2 David E Newby MD PhD

1 and Nicholas L Mills MD PhD

1

1BHF Centre for Cardiovascular Science, University of Edinburgh, United Kingdom 2Centre for Population Health Sciences, University of Edinburgh, United Kingdom 3Department of Emergency Medicine, Royal Infirmary of Edinburgh, United Kingdom 4Emergency Medicine Research Group Edinburgh (EMeRGE), United Kingdom

Correspondence and requests for reprints:

Dr Anoop S V Shah

BHF/University Centre for Cardiovascular Science

Chancellor’s Building

University of Edinburgh

Edinburgh EH16 4SB

United Kingdom

Tel: +44 7766 544 156

Fax: +44 131 242 6379

E-mail: [email protected]

Abstract: 295

Word Count: 2,952

Table and Figures: 5

References: 46

Support: British Heart Foundation Special Project Grant (SP/12/10/29922),

Project Grant (PG/15/51/31596) and Senior Clinical Research

Fellowship (FS/16/14/32023)

Page 1 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

2

Abstract

Objective: High-sensitivity cardiac troponin assays may improve the diagnosis of myocardial

infarction, but increase the detection of myocardial injury in patients without acute coronary

syndrome. We evaluate how cardiac troponin testing in unselected patients would impact on

the diagnosis of myocardial infarction.

Design: Prospective study of two independent consecutive patient cohorts presenting to the

Emergency Department.

Setting: Secondary and tertiary care hospitals in Scotland

Participants: High-sensitivity cardiac troponin I concentrations were measured in two

independent consecutive patient cohorts presenting to the Emergency Department: 1,054

unselected patients, and 5,815 patients in whom the clinician suspected acute coronary

syndrome following initial assessment. The final diagnosis of type 1 or type 2 myocardial

infarction, or myocardial injury was independently adjudicated.

Main outcome measures: The positive predictive value (PPV) of an elevated cardiac

troponin concentration for a diagnosis of type 1 myocardial infarction.

Results: Cardiac troponin concentrations were elevated in 13.7% (144/1,054) of unselected

patients with 17, 13 and 114 patients classified as type 1 myocardial infarction, type 2

myocardial infarction and myocardial injury respectively. The prevalence of type 1

myocardial infarction was just 1.6% (17/1,054) with a PPV of 13.6% (95% confidence

interval [CI] 8.0-19.7%). When testing was performed only in patients with suspected acute

coronary syndrome, the prevalence of type 1 myocardial infarction was higher at 14.5%

(843/5,815) and the PPV increased to 59.7% (95%CI 57.1-62.2%). The PPV was highest in

patients with chest pain (67.5% [95%CI 64.7-70.3]) compared to those without (34.1%

[95%CI 29.0-39.2]).

Page 2 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

3

Conclusions: When high-sensitivity cardiac troponin testing is performed without prior

clinical assessment, elevated troponin concentrations are common, and predominantly reflect

myocardial injury rather than type 1 myocardial infarction. Our observations highlight the

importance of prior clinical assessment to guide patient selection for testing to optimize the

diagnostic utility of high-sensitivity cardiac troponin.

Page 3 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

4

Introduction

Cardiac troponin is integral to the diagnosis of myocardial infarction, but troponin

concentrations are often elevated in patients who do not have acute coronary syndrome. The

universal definition now classifies myocardial infarction as spontaneous or type 1, due to

plaque rupture and coronary thrombosis, and secondary or type 2 due to myocardial oxygen

supply-demand imbalance.1-5

Patients with elevated cardiac troponin concentrations in the

absence of myocardial ischemia are classified as having myocardial injury. Whilst patients

with type 2 myocardial infarction or myocardial injury are increasingly recognized in clinical

practice,1 they represent a heterogeneous group with overt or covert major illness for whom

there is currently no consensus on the optimal cardiac investigation, management or treatment

strategy.

We have shown that lowering the diagnostic threshold with a more sensitive cardiac troponin

assay reduced recurrent myocardial infarction or death in patients redefined as having type 1

myocardial infarction.6 However, use of these lower diagnostic thresholds more than doubled

the number of patients with type 2 myocardial infarction or myocardial injury with no

improvement in their outcome despite undergoing additional cardiac investigation.1 The

introduction of high-sensitivity cardiac troponin assays will further increase the frequency of

type 2 myocardial infarction or myocardial injury6,7

and may lead to diagnostic uncertainty

and unnecessary investigation of patients without acute coronary syndrome.7,8

Patients attending the Emergency Department often undergo simultaneous testing for both

cardiac and non-cardiac conditions,9 to facilitate early diagnosis or discharge. In this context,

a non-selective approach to high-sensitivity cardiac troponin testing may lead to inappropriate

Page 4 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

5

diagnosis of type 1 myocardial infarction.10

Our aim was to evaluate the impact of non-

selective high-sensitivity cardiac troponin testing on the diagnosis of myocardial infarction.

Page 5 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

6

Methods

Study populations

This prospective observational cohort study was performed in two pre-specified patient

populations with the approval of the national research ethics committee, and in accordance

with the Declaration of Helsinki. In an unselected cohort, we identified all consecutive

patients (n=1,054) presenting to the Emergency Department at the Royal Infirmary of

Edinburgh, Scotland in whom the attending clinician performed blood sampling irrespective

of their clinical presentation (Figure 1). In a separate independent cohort, we identified all

consecutive patients with suspected acute coronary syndrome (n=5,815) presenting to

secondary and tertiary care hospitals in the southeast of Scotland.11,12

Across both cohorts the

attending clinician reviewed all patients and identified those with suspected acute coronary

syndrome through a dedicated proforma embedded into the electronic patient record prior to

troponin testing. Patients across both cohorts were excluded if they had ST-segment elevation

myocardial infarction, or a previous presentation during the study period, or were not resident

in Scotland.

Cardiac troponin I assay

Cardiac troponin testing was performed at the discretion of the attending physician using a

contemporary cardiac troponin I assay (Abbott Laboratories, Abbott Park, IL). Serum surplus

to clinical requirements was used to measure cardiac troponin I concentration using the

ARCHITECTSTAT high-sensitive troponin I assay (Abbott Laboratories, Abbott Park, IL). This

assay has a limit of detection of 1.2 ng/L and the inter-assay CV <10% at 4.7 ng/L. The 99th

centile upper reference limit is 34 ng/L in men and 16 ng/L in women.12

The results from the

high-sensitivity assay were not made available to clinicians and across both cohorts only

Page 6 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

7

results from the contemporary assay, where requested by the attending clinician, were used to

guide patient care.

Clinical characteristics

Baseline clinical characteristics and investigations were obtained from a standardized

electronic patient record as previously described.1,6,12

The National Early Warning Score was

calculated for patients incorporating six simple physiological parameters: pulse rate, blood

pressure, temperature, oxygen saturation, level of consciousness and respiratory rate

measured at admission (Appendix Table 1).13

Classification of myocardial injury and infarction

In all patients with elevated high-sensitivity cardiac troponin I concentrations, the diagnosis

was adjudicated by two cardiologists independently as previously described.1,11,12

Type 1

myocardial infarction was defined as myocardial necrosis in the context of a presentation with

suspected acute coronary syndrome with chest pain or evidence of myocardial ischemia on

the electrocardiogram. Patients with symptoms or signs of myocardial ischemia due to

increased oxygen demand or decreased supply (e.g. tachyarrhythmia, hypotension or anemia)

secondary to an alternative pathology and myocardial necrosis were classified as type 2

myocardial infarction. Myocardial injury was defined as evidence of myocardial necrosis in

the absence of any clinical features of myocardial ischemia. A third adjudicator resolved any

discrepancies in adjudication through in-depth review of source documents. All patients were

followed up for death from any cause using regional and national registries.14

This method

allowed capture of all in-hospital and community deaths ensuring complete follow-up.

Page 7 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

8

Patient and Public Involvement

Patient and public involvement has been incorporated throughout the trial development and

conduct. At study design stage, we assessed feasibility and received feedback on the potential

acceptability of patient participation in the trial. Lay representatives contribute to the

membership of the Trial Steering Committee for the High-STEACS clinical trial and all

related studies (NCT01852123).

Statistical analysis

Unselected patients were categorized into five groups a priori. Those patients with elevated

cardiac troponin concentrations >99th

centile upper reference limit (men >34 ng/L and women

>16 ng/L)11

using the high-sensitivity assay were grouped together (group 5). The remaining

patients separated into four quartiles (group 1-4) (see Appendix for full statistical analysis

plan). The distribution of cardiac troponin differed in men and women and were therefore

grouped separately. Prevalence, sensitivity, specificity and positive likelihood ratios were

estimated by sampling from Beta distributions with Jeffrey’s priors (see Appendix for full

statistical analysis plan). All analyses were performed in R Version 3.2.3. Statistical

significance was taken as a two-sided P<0.05.

Page 8 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

9

Results

There were 3,619 visits to the Emergency Department during which 1,130 patients underwent

blood sampling for their presenting complaint (Figure 1). There were 76 patients who met

our exclusion criteria giving a final study population of 1,054 patients with a mean age of

55±23 years (52.1% women) (Table 1). Cardiac troponin was requested by the attending

physician in 3.8% (n=136/3,619) of all visits and in 12.9% (n=136/1,054) of the study

population. Patients who had cardiac troponin requested by the attending physician were

older, and were more likely to be male, to have cardiovascular risk factors or to present with

chest pain and have intermediate or high GRACE scores (Appendix Table 2).15

There were a total of 5,815 consecutive patients with suspected acute coronary syndrome

mean age of 64±16 years (43.9% females) (Figure 1 and Table 1). Patients with suspected

acute coronary syndrome were more likely to present with chest pain and had a higher

prevalence of cardiovascular risk factors compared to unselected patients in the Emergency

Department.

Prevalence and positive predictive value of cardiac troponin

In unselected patients attending the Emergency Department, 13.7% (n=144/1,054) had

cardiac troponin concentrations above the 99th

centile with 17, 13 and 114 patients classified

as type 1 myocardial infarction, type 2 myocardial infarction and myocardial injury

respectively. Of all patients with cardiac troponin concentrations above the 99th

centile, chest

pain (n=36/144, 25%), falls or collapse (n=40/144, 28%), and dyspnea (n=13/144, 9%) were

the most common presenting complaints (Appendix Table 3). The most common diagnoses

were cardiac (n=35/144, 24%), respiratory (n=23/144, 16%) and infectious disease

(n=21/144, 15%) (Appendix Table 4). Overall, the prevalence of type 1 myocardial infarction

Page 9 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

10

was 1.6% (n=17/1,054) with a positive predictive value for type 1 myocardial infarction of

13.6% (95% CI 8.0 to 19.7%) (Figure 2a, 2b and Appendix Figure 1).

In patients with suspected acute coronary syndrome, cardiac troponin was elevated in 24.3%

(95% CI 23.2 to 25.4%) (n=1,403/5,815) of all patients. Type 1 myocardial infarction was

adjudicated in 843 (14.5%) patients, with 229 (3.9%) and 341 (5.9%) of patients classified as

type 2 myocardial infarction and myocardial injury respectively. The prevalence of type 1

myocardial infarction and positive predictive value was 14.5% (n=843/5,825) and 59.7%

(95% CI 57.1 to 62.2%) respectively (Figure 2a, 2b and Appendix Figure 1). The positive

predictive value was highest in patients with chest pain (67.5 [64.7-70.3]%), myocardial

ischemia (69.8 [65.9-73.7]%), or known ischemic heart disease (68.1 [64.2-71.9]%) compared

to those without (34.1 [29.0-39.2]%, 57.8 [54.0-61.0]% and 55.1 [51.6-58.6]% respectively)

(Figure 3). The positive predictive value was 83.0% (95% CI 77.4 to 88.5%) in patients with

all three of these clinical features.

Agreement in the adjudication of type 1 myocardial infarction across both cohorts was

excellent at 0.86 (95% CI 0.83-0.89) and remained robust whether patients underwent serial

sampling or not (0.83 [95%CI 0.79 - 0.87] and 0.86 [95%CI 0.81 – 0.89] respectively).

In a sensitivity analysis, we evaluated the combined prevalence of type 1 or type 2 myocardial

infarction in unselected patients and those with suspected acute coronary syndrome. In

unselected patients, the prevalence of type 1 or type 2 myocardial infarction was 2.9% (95%

CI 1.9 to 3.9%, n=30/1,054) with a positive predictive value of 21.0% (95% CI 14.6 to

27.7%). In patients with suspected acute coronary syndrome, the prevalence was 18.4% (95%

Page 10 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

11

CI 17.4 to 19.4%, n=1,072/5,815) with a positive predictive value of 75.8% (95% CI 73.6 to

78.1%). The presence of chest pain, myocardial ischemia on the electrocardiogram or history

of ischemic heart disease was associated with a higher positive predictive value (Appendix

Table 5).

Determinants and outcomes of patients with elevated cardiac troponin

Unselected patients with elevated cardiac troponin concentrations (group 5), were older

(77±16 versus 35±13 years, P<0.001) and more likely to have comorbid conditions than those

with the lowest cardiac troponin concentrations (group 1) (Table 2). Patients with elevated

cardiac troponin concentrations were also more likely to have abnormal physiology (medium

or high risk National Early Warning Score, 16.8% versus 2.3%, P<0.001), lower hemoglobin

(122±23 versus 139±15 mg/dL, P<0.001) and higher creatinine concentrations (1.3±1.0

versus 0.8±0.2 mg/dL). The prevalence of elevated cardiac troponin concentrations increased

with age, the presence of ischemic heart disease and the number of co-morbid conditions

(Appendix Figure 2). The majority of patients found to have an elevated cardiac troponin

concentration (group 5) were admitted to hospital (124/144, 86.1%).

On logistic regression both age (model 1: odds ratio 2.06, 95% CI 1.75 to 2.47 per 10 year

increment) and comorbidity (model 2: odds ratio 6.88, 95% CI 1.83 to 23.21 for four or more

comorbidities compared to none) were strongly associated with elevated cardiac troponin

concentrations after adjusting for sex, renal function and the National Early Warning Score

(Appendix Table 6). When adjusting for both age and comorbidity, age remained a powerful

determinant of an elevated cardiac troponin concentration (odds ratio 1.99, 95% CI 1.66 to

2.42 per 10 years’ increment, P<0.001).

Page 11 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

12

Discussion

We have evaluated the impact of a non-selective approach to high-sensitivity cardiac troponin

testing on the diagnosis of myocardial infarction in consecutive patients attending the

Emergency Department. We make a number of important observations. First, if testing is

performed in all patients without prior clinical assessment, elevated cardiac troponin

concentrations are frequent, occurring in one in every eight patients. The majority of these

patients were admitted to hospital with an alternative primary diagnosis and were adjudicated

as having type 2 myocardial infarction or myocardial injury. Second, cardiac troponin testing

without prior clinical assessment results in a very low prevalence of type 1 myocardial

infarction (2%). As such, the positive predictive value of an elevated cardiac troponin

concentration for type 1 myocardial infarction is low at 14% in unselected patients. Third, if

prior clinical assessment is used to guide cardiac troponin testing, the prevalence and positive

predictive value increases more than four-fold. These findings highlight the importance of

clinical assessment and patient selection to optimize the diagnostic utility of high-sensitivity

cardiac troponin.

This is the first prospective evaluation of high-sensitivity cardiac troponin in all patients

undergoing blood sampling in the Emergency Department. This is an important distinction

from previous consecutive1,11,12

or selected16,17

patient cohorts of suspected acute coronary

syndrome. Our study has a number of strengths. First, we minimized selection bias by

identifying all consecutive patients presenting to the Emergency Department. Second, we did

not rely on the contemporary cardiac troponin assays, but instead two cardiologists

independently adjudicated the diagnosis using the high-sensitivity cardiac troponin I assay

with sex-specific thresholds as the reference standard.18,19

Together these approaches reduced

Page 12 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

13

case ascertainment bias and the potential to underestimate the prevalence of elevated cardiac

troponin concentrations.

Our study has implications for the adoption of high-sensitivity cardiac troponin assays,

particularly in those regions where the frequency of testing is high. The positive predictive

value is dependent on the prevalence of type 1 myocardial infarction, which in turn is

dependent on the selection of patients for testing which differs widely across health care

systems. In a representative sample of over 44,000 patients attending over 500 Emergency

Departments in the United States, 17% of all patients and 47% of those hospitalized, had

cardiac biomarkers tested,5 compared to just 3% and 16% respectively in our cohort.

Furthermore, over a third of patients tested in this US population did not present with cardiac

symptoms.5 The only previous study of high-sensitivity cardiac troponin I testing in a US

Emergency Department reported a prevalence of type 1 myocardial infarction of 3.2%,20

which we estimate would give a positive predictive value of 13.4% (Figure 2). High-

sensitivity cardiac troponin assays are now being introduced worldwide, with the exception of

the United States where they have only recently been approved by the Food and Drug

Administration.19

Whilst cardiac troponin testing in the undifferentiated patient may be

justified where myocardial infarction is a possibility,21,22

it is important that clinicians are

aware that elevated cardiac troponin concentrations are not exclusive to type 1 myocardial

infarction. If implemented without adopting a considered approach to testing and

interpretation, high-sensitivity cardiac troponin assays may increase diagnostic uncertainty

and increase the need for further invasive and non-invasive cardiac investigations with cost

implications for the health care system.

Page 13 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

14

Where clinical gestalt was used to select patients for cardiac troponin testing, we observed a

much improved positive predictive value. Perhaps unsurprisingly the positive predictive value

was highest in patients with chest pain, myocardial ischemia on the electrocardiogram or

known ischemic heart disease. A more selective approach to cardiac troponin testing is clearly

preferable, but could it potentially lead to clinicians missing a diagnosis of type 1 myocardial

infarction? Our analysis suggests that this is unlikely. In 1,054 unselected patients there were

only two patients with type 1 and three patients with type 2 myocardial infarction in whom

cardiac troponin was not requested by their attending clinician. These patients were already

admitted to hospital for further assessment with the exception of one patient who was

discharged with what was considered to be atypical chest pain (Appendix Table 7).

Whilst elevated cardiac troponin concentrations outwith acute coronary syndrome may be

challenging to interpret, they do convey potentially important clinical information. In these

patients cardiac troponin concentrations reflect physiological and hemodynamic stress. Nearly

all of these patients were already recognized by their attending physician as being acutely

unwell and were admitted to hospital. Although cardiac troponin is a powerful prognostic

marker in patients with type 2 myocardial infarction or myocardial injury, there is currently

no guidance on how to investigate these patients, and as yet no evidence to suggest that

cardiovascular treatments will improve outcomes.1 Until we have a better understanding of

the mechanisms and underlying pathophysiology, or effective treatment strategies for these

patients routine screening for these conditions is not recommended.

We acknowledge that whilst the majority of troponin determinations in the Emergency

Department are intended to evaluate patients with suspected acute coronary syndrome,

guidelines do recommend cardiac biomarker testing in other acute presentations including

Page 14 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

15

pulmonary embolus and acute heart failure.23,24

As such, whilst we have demonstrated how

cardiac troponin testing without consideration of pre-test probability impacts on the positive

predictive value of type 1 myocardial infarction, we accept that cardiac troponin testing is not

used exclusively to evaluate patients with suspected acute coronary syndrome.

There are some limitations to our study that merit discussion. First, we did not perform serial

cardiac troponin testing or systematically undertake coronary investigations in our cohort of

unselected patients. As such, in this cohort our adjudication was based on a single

measurement at presentation and those investigations undertaken as part of routine clinical

care. Patients with an elevated cardiac troponin concentration were not classified as having a

type 1 myocardial infarction where the clinical presentation and subsequent investigations

strongly supported an alternative primary diagnosis. Despite our careful attempt to classify

patients we accept a number of patients may have been misclassified. Whilst serial testing

may help to differentiate between acute and chronic myocardial injury, it does not distinguish

between type 1 and type 2 myocardial infarction.25,26

However, critically our study design

eliminated selection bias by systematically measuring cardiac troponin in wastage samples

from all patients without influencing clinical care. This precluded serial sampling, but ensured

our findings are generalisable and relevant to clinical practice. This approach was essential to

evaluate how the routine clinical assessment of pre-test probability influences the positive

predictive value of cardiac troponin testing. We accept this approach may have

underestimated acute myocardial injury in our unselected patients, as a small proportion of

patients with cardiac troponin concentrations in the normal reference range at presentation

may have had elevated concentrations on serial testing. Second, the cohort of unselected

patients was derived from the Emergency Department at a single site. Differing health

systems, including reimbursement systems, and the availability of primary care physicians

Page 15 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

16

may result in regional differences in the patient populations presenting to the Emergency

Department. This may have an impact on the overall prevalence and subsequent positive

predictive value of an elevated cardiac troponin concentration for the diagnosis of type 1

myocardial infarction. However, we modelled the impact of differences in prevalence on the

positive predictive value of cardiac troponin and the findings from our cohort are consistent

with multiple previous studies from centres across the world (Appendix Figure 1).16,27-46

Conclusions

When high-sensitivity cardiac troponin testing is performed in unselected patients, elevated

troponin concentrations are common and predominantly reflect myocardial injury rather than

type 1 myocardial infarction. Our observations highlight the importance of prior clinical

assessment to guide patient selection for testing to optimize the diagnostic utility of high-

sensitivity cardiac troponin.

Page 16 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

17

Acknowledgements

This research was funded by the British Heart Foundation (SP/12/10/29922 and

PG/15/51/31596). NLM and DEN are supported by the Butler Senior Clinical Research

Fellowship (FS/16/14/32023) and John Wheatley Chair (CH/09/002) awards respectively

from the British Heart Foundation. Abbott Laboratories provided high-sensitivity cardiac

troponin I assay reagents, calibrators, and controls.

Data sharing

Patient level data and statistical code will be made available following publication of the

primary study (HighSTEACS: NCT01852123) from the corresponding author at

[email protected]. Participant consent to share data was not obtained but the presented

data are anonymized and risk of identification is low.

Transparency Declaration

Dr Anoop Shah affirms that the manuscript is an honest, accurate, and transparent account of

the study being reported; that no important aspects of the study have been omitted; and that

any discrepancies from the study as planned (and, if relevant, registered) have been explained.

Copyright Statement

The Corresponding Author has the right to grant on behalf of all authors and does grant on

behalf of all authors, an exclusive licence on a worldwide basis to the BMJ Publishing Group

Ltd to permit this article (if accepted) to be published in BMJ editions and any other

BMJPGL products and sublicences such use and exploit all subsidiary rights, as set out in our

licence.

Page 17 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

18

Conflicts of interest

NLM and ASVS has acted as a consultant for Abbott Laboratories and Beckman-Coulter. All

other authors have no conflicts of interest.

Contribution statement

ASVS and NLM initially conceived the design and carried out the initial acquisition, analysis,

or interpretation of data. All authors were involved in drafting the manuscript and revising it,

and have given final approval of the version to be published.

Page 18 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

19

References

1. Shah AS, McAllister DA, Mills R, et al. Sensitive troponin assay and the classification

of myocardial infarction. Am J Med. 2015;128:493-501 e493.

2. Newby LK, Jesse RL, Babb JD, et al. ACCF 2012 expert consensus document on

practical clinical considerations in the interpretation of troponin elevations: a report of

the American College of Cardiology Foundation task force on Clinical Expert

Consensus Documents. J Am Coll Cardiol. 2012; 60: 2427–63.

3. Saaby L, Poulsen TS, Hosbond S, et al. Classification of myocardial infarction:

frequency and features of type 2 myocardial infarction. Am J Med. 2013;126:789-797.

4. Saaby L, Poulsen TS, Diederichsen AC, et al. Mortality rate in type 2 myocardial

infarction: observations from an unselected hospital cohort. Am J Med. 2014;127:295-

302.

5. Sandoval Y, Smith SW, Thordsen SE, Apple FS. Supply/demand type 2 myocardial

infarction: should we be paying more attention? J Am Coll Cardiol. 2014;63:2079-

2087.

6. Mills NL, Churchhouse AM, Lee KK, et al. Implementation of a sensitive troponin I

assay and risk of recurrent myocardial infarction and death in patients with suspected

acute coronary syndrome. JAMA. 2011;305:1210-1216.

7. Melanson SE, Conrad MJ, Mosammaparast N, Jarolim P. Implementation of a highly

sensitive cardiac troponin I assay: test volumes, positivity rates and interpretation of

results. Clin Chim Acta. 2008;395:57-61.

8. Makam AN, Nguyen OK. Use of cardiac biomarker testing in the emergency

department. JAMA Intern Med. 2015;175:67-75.

Page 19 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

20

9. Yiadom MY, Jarolim P, Jenkins C, Melanson SE, Conrad M, Kosowsky JM.

Diagnostic implications of an elevated troponin in the emergency department. Disease

markers. 2015;2015:157812.

10. Baxi S, Lakin J, Stapleton S, Redberg R. High-sensitivity troponin: elevated without

infarction, is the horse out of the barn? Emerg Med J. 2014;31:354-355.

11. Shah AS, Anand A, Sandoval Y, et al. High-sensitivity cardiac troponin I at

presentation in patients with suspected acute coronary syndrome: a cohort study.

Lancet. 2015;386:2481-2488.

12. Shah AS, Griffiths M, Lee KK, et al. High sensitivity cardiac troponin and the under-

diagnosis of myocardial infarction in women: prospective cohort study. BMJ.

2015;350:g7873.

13. National early warning score (NEWS): standardising the assessment of acute-illness

severity in the NHS. Royal College of Physicians 2012. Available at

https://www.rcplondon.ac.uk/projects/outputs/national-early-warning-score-news

14. Ford I, Murray H, Packard CJ, Shepherd J, Macfarlane PW, Cobbe SM. Long-term

follow-up of the West of Scotland Coronary Prevention Study. N Engl J Med. 2007;

357: 1477–86.

15. Fox KA, Dabbous OH, Goldberg RJ, et al. Prediction of risk of death and myocardial

infarction in the six months after presentation with acute coronary syndrome:

prospective multinational observational study (GRACE). BMJ. 2006;333:1091.

16. Keller T, Zeller T, Ojeda F, et al. Serial changes in highly sensitive troponin I assay

and early diagnosis of myocardial infarction. JAMA. 2011;306:2684-2693.

17. Reichlin T, Hochholzer W, Bassetti S, et al. Early diagnosis of myocardial infarction

with sensitive cardiac troponin assays. N Engl J Med. 2009; 361: 858–67

Page 20 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

21

18. Apple FS. A new season for cardiac troponin assays: it's time to keep a scorecard. Clin

Chem. 2009;55:1303-1306.

19. Korley FK, Jaffe AS. Preparing the United States for high-sensitivity cardiac troponin

assays. J Am Coll Cardiol. 2013;61:1753-1758.

20. Korley FK, Schulman SP, Sokoll LJ, et al. Troponin elevations only detected with a

high-sensitivity assay: clinical correlations and prognostic significance. Acad Emerg

Med. 2014;21:727-735.

21. Jairam S, Jones P, Samaraie L, Chataline A, Davidson J, Stewart R. Clinical diagnosis

and outcomes for Troponin T 'positive' patients assessed by a high sensitivity

compared with a 4th generation assay. Emerg Med Australas. 2011;23:490-501.

22. Sherwood MW, Kristin Newby L. High-sensitivity troponin assays: evidence,

indications, and reasonable use. J Am Heart Assoc. 2014;3(1):e000403.

23. Konstantinides SV, Torbicki A, Agnelli G, et al. 2014 ESC guidelines on the

diagnosis and management of acute pulmonary embolism. Eur Heart J. 2014;35:3033-

3069, 3069a-3069k.

24. Ponikowski P, Voors AA, Anker SD, et al. 2016 ESC Guidelines for the diagnosis and

treatment of acute and chronic heart failure: The Task Force for the diagnosis and

treatment of acute and chronic heart failure of the European Society of Cardiology

(ESC). Developed with the special contribution of the Heart Failure Association

(HFA) of the ESC. Eur J Heart Fail. 2016;18:891-975.

25. Sandoval Y, Thordsen SE, Smith SW, et al. Cardiac troponin changes to distinguish

type 1 and type 2 myocardial infarction and 180-day mortality risk. Eur Heart J Acute

Cardiovasc Care. 2014;3:317-325.

26. Collinson P, Lindahl B. Type 2 myocardial infarction: the chimaera of cardiology?

Heart. 2015;101:1697-1703.

Page 21 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

22

27. Invernizzi L, Doka M, Cappellini F, et al. Effectiveness of highly sensitive troponin T

assay for early diagnosis of acute myocardial infarction (AMI). Biochimica Clinica.

2013; 37: 36–9.

28. Lotze U, Lemm H, Heyer A, Muller K. Combined determination of highly sensitive

troponin T and copeptin for early exclusion of acute myocardial infarction: first

experience in an emergency department of a general hospital. Vasc Health Risk

Manag. 2011;7:509-515.

29. Bahrmann P, Bahrmann A, Breithardt OA, et al. Additional diagnostic and prognostic

value of copeptin ultra-sensitive for diagnosis of non-ST-elevation myocardial

infarction in older patients presenting to the emergency department. Clin Chem Lab

Med. 2013;51:1307-1319.

30. Bahrmann P, Christ M, Bahrmann A, et al. A 3-hour diagnostic algorithm for non-ST-

elevation myocardial infarction using high-sensitivity cardiac troponin T in unselected

older patients presenting to the emergency department. J Am Med Dir Assoc.

2013;14:409-416.

31. Hammerer-Lercher A, Ploner T, Neururer S, et al. High-sensitivity cardiac troponin T

compared with standard troponin T testing on emergency department admission: how

much does it add in everyday clinical practice? J Am Heart Assoc. 2013;2(3):e000204.

32. Thelin J, Borna C, Erlinge D, Ohlin B. The combination of high sensitivity troponin T

and copeptin facilitates early rule-out of ACS: a prospective observational study. BMC

Cardiovasc Disord. 2013;13:42.

33. Freund Y, Chenevier-Gobeaux C, Bonnet P, et al. High-sensitivity versus

conventional troponin in the emergency department for the diagnosis of acute

myocardial infarction. Crit Care. 2011;15:R147.

Page 22 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

23

34. Santalo M, Martin A, Velilla J, et al. Using high-sensitivity troponin T: the importance

of the proper gold standard. Am J Med. 2013;126:709-717.

35. Hoeller R, Rubini Gimenez M, Reichlin T, et al. Normal presenting levels of high-

sensitivity troponin and myocardial infarction. Heart. 2013;99:1567-1572.

36. Body R, Carley S, McDowell G, et al. Rapid exclusion of acute myocardial infarction

in patients with undetectable troponin using a high-sensitivity assay. J Am Coll

Cardiol. 2011;58:1332-1339.

37. Aldous SJ, Florkowski CM, Crozier IG, Than MP. The performance of high

sensitivity troponin for the diagnosis of acute myocardial infarction is underestimated.

Clin Chem Lab Med. 2012;50:727-729.

38. Aldous SJ, Richards AM, Cullen L, Than MP. Early dynamic change in high-

sensitivity cardiac troponin T in the investigation of acute myocardial infarction. Clin

Chem. 2011;57:1154-1160.

39. Christ M, Popp S, Pohlmann H, et al. Implementation of high sensitivity cardiac

troponin T measurement in the emergency department. Am J Med. 2010;123:1134-

1142.

40. Giannitsis E, Kehayova T, Vafaie M, Katus HA. Combined testing of high-sensitivity

troponin T and copeptin on presentation at prespecified cutoffs improves rapid rule-

out of non-ST-segment elevation myocardial infarction. Clin Chem. 2011;57:1452-

1455.

41. Collinson P, Goodacre S, Gaze D, Gray A. Very early diagnosis of chest pain by

point-of-care testing: comparison of the diagnostic efficiency of a panel of cardiac

biomarkers compared with troponin measurement alone in the RATPAC trial. Heart.

2012;98:312-318.

Page 23 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

24

42. Sebbane M, Lefebvre S, Kuster N, et al. Early rule out of acute myocardial infarction

in ED patients: value of combined high-sensitivity cardiac troponin T and

ultrasensitive copeptin assays at admission. Am J Emerg Med. 2013;31:1302-1308.

43. Inoue K, Suwa S, Ohta H, et al. Heart fatty acid-binding protein offers similar

diagnostic performance to high-sensitivity troponin T in emergency room patients

presenting with chest pain. Circ J. 2011;75:2813-2820.

44. Eggers KM, Venge P, Lindahl B. High-sensitive cardiac troponin T outperforms novel

diagnostic biomarkers in patients with acute chest pain. Clin Chim Acta.

2012;413:1135-1140.

45. Normann J, Mueller M, Biener M, Vafaie M, Katus HA, Giannitsis E. Effect of older

age on diagnostic and prognostic performance of high-sensitivity troponin T in

patients presenting to an emergency department. Am Heart J. 2012;164:698-705 e694.

46. Khan DA, Sharif MS, Khan FA. Diagnostic performance of high-sensitivity troponin

T, myeloperoxidase, and pregnancy-associated plasma protein A assays for triage of

patients with acute myocardial infarction. Korean J Lab Med. 2011;31:172-178.

Page 24 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

25

Figure legends

Figure 1. Flow diagram summarizing the patients recruited in the unselected Emergency

Department cohort (Red) and the in the suspected acute coronary syndrome cohort (Green).

Figure 2.

A) Infographic summarizing the prevalence of elevated cardiac troponin and type 1

myocardial infarction in the unselected Emergency Department patient cohort (Red)

and in patients with suspected acute coronary syndrome (Green)

B) Influence of prevalence on the positive predictive value of an elevated cardiac

troponin concentration for the diagnosis of type 1 myocardial infarction.

Red dot represents cohort of unselected patients in the Emergency Department

(n=1,054) and the green dot represents cohort of selected patients with suspected

acute coronary syndrome (n=5,815). Red line represents the central estimate of the

positive predictive value with the 95% confidence interval (dashed blue) derived from

the unselected Emergency Department cohort.

Figure 3. Influence of clinical characteristics on the positive predictive value of an elevated

cardiac troponin for a diagnosis of type 1 myocardial infarction in patients with suspected

acute coronary syndrome.

Dashed red line represents the positive predictive value across the whole cohort with dashed

grey lines representing the 95% confidence intervals.

Page 25 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

26

Table 1. Baseline characteristics of unselected patients in the Emergency Department

and selected patients with suspected acute coronary syndrome

Unselected patients

n=1,054

Suspected acute

coronary syndrome

n=5,815

Sex, male (%) 549 (52.0) 2,552 (43.9)

Age (years) 54 (23) 64 (16)

Chest pain, n (%) 183 (17.6) 4,825 (83.0)

Risk factors

Smoker, n (%) 183 (17.6) 1,105 (30.6)

Hypertension, n (%) 337 (32.4) 1,969 (37.6)

Hyperlipidaemia, n (%) 299 (28.7) 1,611 (30.8)

Past medical history

Ischemic heart disease, n (%) 193 (18.5) 1,846 (35.2)

Myocardial infarction, n (%) 109 (10.5) 1,082 (20.7)

Cerebrovascular disease, n (%) 99 (9.5) 475 (9.1)

Diabetes mellitus, n (%) 106 (10.1) 842 (16.1)

PCI, n (%) 52 (5.0) 611 (11.7)

CABG, n (%) 32 (3.1) 330 (6.3)

Medications at presentation

Aspirin, n (%) 180 (17.5) 1,344 (33.7)

Clopidogrel, n (%) 81 (7.9) 468 (11.8)

Beta blockers, n (%) 149 (14.5) 1,082 (27.2)

ACE-I/ARB, n (%) 189 (18.3) 1,311 (32.9)

Statin, n (%) 249 (24.2) 1,578 (39.6)

Warfarin, n (%) 44 (4.3) 278 (7.0)

Haemodynamics

Systolic blood pressure (mmHg) 130.5 (22.2) 137.5 (26.0)

Pulse rate (bpm) 86.9 (22.3) 81.2 (22.9)

Killip class, n (%)

I 930 (89.7) 4,847 (90.8)

II 85 (8.2) 408 (7.6)

III 15 (1.4) 75 (1.4)

IV 1 (0.1) 6 (0.1)

Baseline electrocardiography

ST elevation, n (%) 13 (2.0) 218 (4.2)

ST depression, n (%) 21 (3.2) 397 (7.7)

T-wave inversion, n (%) 58 (8.9) 726 (14.1)

Diagnosis

Type 1 myocardial infarction 17 (1.6) 843 (14.5)


Myocardial injury 114 (10.8) 341 (5.9)

Abbreviations: bpm – breaths/beats per minute; PCI – percutaneous coronary intervention, CABG – coronary artery bypass grafting,

ACE-I/ARB: angiotensin converting enzyme inhibitor/ Angiotensin receptor blocker

Page 26 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only 27

Table 2. Baseline characteristics in unselected patients attending the Emergency Department stratified by cardiac troponin concentration

Group 1

Group 2

Group 3

Group 4

Group 5

(TnI > 99th centile)

n 341 150 210 209 144

Troponin concentration, ng/L

Male ≤2 3 4 - 8 9 - 34 >34

Female ≤1 2 3 - 4 5 - 16 >16

Sex, male (%) 174 (51.0) 66 (44.0) 108 (51.4) 102 (48.8) 55 (38.2)

Age (years) 35.3 (14.0) 47.3 (16.0) 59.9 (17.8) 70.2 (17.8) 77.0 (15.7)

Chest pain, n (%) 42 (12.5) 26 (17.6) 41 (19.6) 38 (18.4) 36 (25.4)

Risk factors

Smoker, n (%) 109 (36.9) 50 (38.8) 60 (33.5) 46 (25.0) 34 (29.6)

Hypertension, n (%) 23 (6.9) 33 (22.1) 87 (41.8) 101 (48.8) 93 (65.5)

Hyperlipidaemia, n (%) 25 (7.5) 31 (20.8) 71 (34.1) 96 (46.6) 76 (53.1)


Ischemic heart disease, n (%) 16 (4.8) 14 (9.4) 44 (21.2) 60 (29.0) 59 (41.3)

Myocardial infarction, n (%) 11 (3.3) 7 (4.7) 26 (12.5) 33 (15.9) 32 (22.4)

Cerebrovascular disease, n (%) 4 (1.2) 10 (6.7) 27 (13.0) 27 (13.0) 31 (21.7)

Diabetes mellitus, n (%) 14 (4.2) 11 (7.4) 25 (11.9) 35 (16.8) 21 (14.7)

Previous revascularisation

PCI, n (%) 9 (2.7) 6 (4.1) 12 (5.7) 14 (6.7) 11 (7.7)

CABG, n (%) 1 (0.3) 1 (0.7) 6 (2.9) 11 (5.3) 13 (9.1)

Physiological parameters

Respiratory rate (bpm) 17.6 (3.6) 17.3 (2.9) 17.6 (3.4) 18.6 (4.0) 20.7 (6.2)

Temperature (oC) 36.8 (0.8) 36.9 (0.8) 36.8 (0.8) 37.0 (0.9) 37.0 (1.1)

Pulse rate (bpm) 85.7 (19.3) 84.6 (20.2) 86.0 (22.6) 86.1 (22.1) 94.1 (28.3)

Systolic blood pressure (mmHg) 126.9 (16.7) 129.0 (21.2) 132.0 (23.7) 135.4 (24.8) 131.0 (26.2)

Oxygen saturations (%) 97.8 (1.9) 97.3 (2.3) 96.7 (2.4) 95.9 (3.3) 95.3 (4.3)

Page 27 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


On oxygen, n (%) 5 (1.7) 1 (0.7) 12 (6.4) 17 (8.9) 27 (20.5)

Consciousness, n (%)

Alert 331 (98.2) 144 (96.6) 205 (98.1) 202 (97.1) 132 (91.7)

Verbal 3 (0.9) 3 (2.0) 2 (1.0) 1 (0.5) 5 (3.5)

Pain 2 (0.6) 0 (0.0) 1 (0.5) 3 (1.4) 2 (1.4)

Unresponsive 1 (0.3) 2 (1.3) 1 (0.5) 2 (1.0) 5 (3.5)

NEWS risk (medium/high), n (%) 5 (2.3) 1 (1.1) 6 (4.4) 7 (5.1) 17 (16.8)

Killip class, n (%)

I 1 (0.3) 0 (0.0) 0 (0.0) 2 (1.0) 3 (2.1)

II 327 (98.5) 145 (98.0) 184 (88.9) 172 (82.7) 102 (71.8)

III 4 (1.2) 3 (2.0) 19 (9.2) 28 (13.5) 31 (21.8)

IV 0 (0.0) 0 (0.0) 4 (1.9) 6 (2.9) 5 (3.5)

Haematology and biochemistry

Haemoglobin (g/dL) 13.9 (1.5) 13.6 (1.7) 13.4 (1.8) 12.7 (2.1) 12.2 (2.3)

Creatinine (mg/dL), median (IQR) 0.8 (0.7-0.9) 0.79 (0.7-0.9) 0.8 (0.7-0.9) 0.9 (0.8-1.2) 1.0 (0.8-1.4)

Urea (mg/dL) 4.4 (1.6) 5.0 (2.2) 5.6 (2.7) 7.5 (4.4) 9.5 (5.4)

Medication at presentation

Aspirin, n (%) 13 (3.9) 16 (10.8) 35 (17.3) 65 (31.7) 51 (36.2)

Clopidogrel, n (%) 8 (2.4) 9 (6.1) 17 (8.4) 29 (14.1) 18 (12.8)

Beta blockers, n (%) 12 (3.6) 13 (8.8) 34 (16.8) 48 (23.4) 42 (29.8)

ACE-I/ARB, n (%) 14 (4.2) 21 (14.2) 53 (26.2) 51 (24.9) 50 (35.5)

Statin, n (%) 19 (5.7) 29 (19.6) 58 (28.7) 83 (40.5) 60 (42.6)

Oral anticoagulant, n (%) 2 (0.6) 4 (2.7) 9 (4.5) 16 (7.8) 13 (9.3)

Admission ECG

ST-segment elevation, n (%) 1 (0.6) 2 (2.4) 1 (0.8) 3 (2.0) 6 (5.1)

ST-segment depression, n (%) 0 (0.0) 4 (4.8) 1 (0.8) 2 (1.4) 14 (12.0)

T-wave inversion, n (%) 4 (2.3) 7 (8.3) 9 (7.1) 22 (14.9) 16 (13.7)

Management

Page 28 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Chest pain nurse referral, n (%) 3 (0.9) 5 (3.4) 6 (2.9) 7 (3.4) 15 (10.4)

Cardiology referral, n (%) 1 (0.3) 3 (2.0) 7 (3.3) 15 (7.2) 24 (16.7)

Admitted to hospital, n (%) 140 (41.5) 74 (49.7) 121 (57.9) 150 (72.1) 124 (86.1)

Echocardiogram, n (%) 0 (0.0) 1 (0.7) 0 (0.0) 6 (2.9) 10 (6.9)

Coronary angiogram, n (%) 1 (0.3) 1 (0.7) 1 (0.5) 4 (1.9) 10 (6.9)

PCI, n (%) 1 (0.3) 0 (0.0) 1 (0.5) 2 (1.0) 6 (4.2)

CABG, n (%) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (0.7)

Abbreviations: bpm – breaths/beats per minute, PCI – percutaneous coronary intervention, CABG – coronary artery bypass grafting,

ACE-I/ARB: angiotensin converting enzyme inhibitor/ Angiotensin receptor blocker, NEWS – National early warning score

SI conversion factors: To convert creatinine to µmol/L, multiply values by 88.4; To convert urea to mmol/L, multiply values by 0.357

Page 29 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only

3,619 ED attendances

Patients undergoing blood sampling

n = 1,130

76 exclusions• inadequate volume• repeat attenders• non-residents

Final study populationn=1,054

Unselected testing

232,000 ED attendances

Suspected acute coronary syndrome

1,155 exclusions• repeat attenders• unable to link records• non-residents

Final study populationn=5,815


Patients undergoing blood sampling

n = 54,000

* Troponin only used to guide clinical care in patients with suspected acute

coronary syndrome

Yes*n=136

Non=918


Yes*n=6,970

Non=47,030

Figure1 Page 30 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Unselectedpatients

TnI > 99th centile(13.7%)

Type 1 myocardialinfarction (1.6%)


TnI > 99th centile(24.2%)

Type 1 myocardialinfarction (14.5%)

Figure2aPage 31 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review OnlyUnselected

Suspected ACS

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50

Prevalence (%)

Pred

icte

d po

sitiv

e pr

edic

tive

valu

e (%

)

Figure 2b Page 32 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


32

Figure 2.

0 10 20 30 40 50 60 70 80 90 100Positive predictive value (95% Confidence Intervals)

Age

Chest pain

Ischaemic ECG

Hypertension

Hyperlipidaemia

Diabetes

Ischaemic heart disease

<40

40−49

50−59

60−69

70−79

>80

Yes

No

Yes

No

Yes

No

Yes

No

Yes

No

Yes

No

3.7 (2.0−5.5)

8.8 (6.8−10.9)

10.4 (8.6−12.2)

14.2 (12.2−16.2)

18.2 (16.1−20.4)

22.5 (20.1−25.0)

15.1 (14.1−16.2)

11.5 (9.5−13.5)

31.0 (28.4−33.6)

11.0 (10.0−12.0)

19.2 (17.4−20.9)

13.1 (11.9−14.3)

20.3 (18.3−22.2)

13.2 (12.1−14.3)

18.0 (15.4−20.6)

14.8 (13.8−15.9)

20.5 (18.7−22.3)

12.6 (11.5−13.7)

1 2 3

Variable StrataPrevalence.of.type.1.

myocardial.infarction,. %.(95%.CI)

≥"80

Positive(predictive( value(for(type(1(myocardial(infarction( (95%(CI)

Figure 3

Page 33 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

1

APPENDIX


in the Emergency Department Anoop SV Shah MD MPH PhD,1 Ala Noaman MD,1 Amar Vaswani MD,1 Mathew Gibbins

MSc,1 Megan Griffiths MD,1 Andrew R Chapman MD,1 Fiona Strachan PhD,1 Atul Anand

MD,1 Philip Adamson MD1, Michelle D Souza,1 Alasdair J Gray MD,3,4 David A McAllister

MD,2 David E Newby MD PhD1 and Nicholas L Mills MD PhD1

1BHF Centre for Cardiovascular Science, University of Edinburgh, United Kingdom 2Centre for Population Health Sciences, University of Edinburgh, United Kingdom 3Department of Emergency Medicine, Royal Infirmary of Edinburgh, United Kingdom 4Emergency Medicine Research Group Edinburgh (EMeRGE), United Kingdom Correspondence and requests for reprints: Dr Anoop S V Shah BHF/University Centre for Cardiovascular Science Chancellor’s Building University of Edinburgh Edinburgh EH16 4SB United Kingdom Fax: +44 131 242 6379 E-mail: [email protected]

Supplementary Tables and Figures: 10

Page 34 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

Appendix Table 1. Parameters and scoring system of the National Early Warning Score (NEWS)

Points 3 2 1 0 1 2 3

Physiological parameter

Respiratory rate, breaths per minute

≤8 - 9 to 11 12 to 20 - 21 to 24 ≥25

Oxygen saturations, % ≤91 92 to 93 94 to 95 ≥96 - -

Any supplemental oxygen - Yes - No - - -

Temperature, oC ≤35.0 35.1 to 36.0 36.1 to 38.0 38.1 to 39 ≥39.1 -

Systolic blood pressure, mmHg

≤90 91 to 100 101 to 110 111 to 219 - - -

Heart rate, beats per minute ≤40 - 41 to 50 51 to 90 91 to 110 111 to 130 ≥131

Consciousness - - - Alert - Voice, Pain or Unresponsive -

Page 35 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

3

Appendix Table 2. Baseline characteristics of unselected patients in the Emergency Department with and without cardiac troponin testing requested by the attending clinician Cardiac troponin testing following

clinical assessment No (n=918) Yes (n=136)

Sex, male (%) 426 (46.4) 79 (58.1)

Age (years), mean (SD) 53.0 (23.0) 64.9 (16.4)

Chest pain, n (%) 75 (8.3) 108 (79.4)

Risk factors

Smoker, n (%) 258 (33.2) 41 (32.5)

Hypertension, n (%) 257 (28.4) 80 (59.3)

Hyperlipidaemia, n (%) 230 (25.4) 69 (51.5)


Ischemic heart disease, n (%) 127 (14.0) 66 (48.9)

Myocardial infarction, n (%) 61 (6.7) 48 (35.6)

Cerebrovascular disease, n (%) 83 (9.2) 16 (11.9)

Diabetes mellitus, n (%) 81 (8.9) 25 (18.5)

Percutaneous coronary intervention, n (%) 28 (3.1) 24 (17.8)

CABG, n (%) 19 (2.1) 13 (9.6)

Admission physiological parameters

Respiratory rate (bpm), mean (SD) 18.3 (4.3) 17.8 (3.4)

Temperature (oC), mean (SD) 36.9 (0.9) 36.6 (0.6)

Pulse rate (bpm), mean (SD) 87.7 (22.1) 81.8 (23.0)

Systolic blood pressure (mmHg), mean (SD) 130.5 (22.3) 130.8 (21.9)

Oxygen saturations (%), mean (SD) 96.8 (3.0) 96.7 (2.2)

On oxygen, n (%) 52 (6.3) 10 (8.1)

Consciousness, n (%)

Alert 880 (96.6) 134 (98.5)

Verbal 14 (1.5) 0 (0.0)

Pain 7 (0.8) 1 (0.7)

Unresponsive 10 (1.1) 1 (0.7)

Killip class, n (%)

I 819 (90.8) 111 (82.2)

II 65 (7.2) 20 (14.8)

III 11 (1.2) 4 (3.0)

IV 1 (0.1) 0 (0.0)

Haematology and biochemistry

Haemoglobin (mg/dL), mean (SD) 132.8 (19.5) 131.9 (16.8)

Creatinine (mg/dL), mean (SD) 1.0 (0.7) 1.0 (0.6)

Urea (mg/dL), mean (SD) 5.9 (3.71) 6.8 (3.8)

Troponin I (ng/L), median [IQR] 3.0 [2.0, 9.0] 5.0 [3.0, 19.5]

Medications on presentation

Page 36 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

4

Aspirin, n (%) 130 (14.5) 50 (37.6)

Clopidogrel, n (%) 62 (6.9) 19 (14.3)

Beta blockers, n (%) 109 (12.1) 40 (30.1)

ACE-I/ARB, n (%) 143 (15.9) 46 (34.6)

Statin, n (%) 192 (21.4) 57 (42.9)

Warfarin, n (%) 35 (3.9) 9 (6.8)

Baseline electrocardiography

ST-segment elevation, n (%) 5 (1.0) 8 (6.0)

ST-segment depression, n (%) 9 (1.7) 12 (9.0)

T-wave inversion, n (%) 34 (6.5) 24 (18.0)

GRACE Risk category

Low 260 (59.0) 45 (40.2)

Medium/ High 179 (41.0) 67 (59.8)

Management

Chest pain nurse referral, n (%) 5 (0.5) 31 (22.8)

Cardiology referral, n (%) 14 (1.5) 36 (26.5)

Admitted to hospital, n (%) 512 (56.2) 97 (71.3)

Echocardiogram, n (%) 5 (0.5) 12 (8.8)

Coronary angiogram, n (%) 2 (0.2) 15 (11.0)

PCI, n (%) 0 (0.0) 10 (7.4)

CABG, n (%) 0 (0.0) 1 (0.7)

Diagnosis

No myocardial injury 804 (87.6) 106 (77.9)



Myocardial injury 109 (11.9) 5 (3.7)

Page 37 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

5

Appendix Table 3. Presenting complaint in patients with cardiac troponin concentrations > 99th centile in the unselected cohort (n=144).

Presenting complaint Number Percentage

Fall/Collapse 40 28%

Chest pain 36 25%

Dyspnea 13 9%

Abdominal pain 8 6%

Dizziness 4 3%

Confusion 3 3%

Other 42 29%

Page 38 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

6

Appendix Table 4. Primary clinical diagnostic in patients with cardiac troponin concentrations > 99th centile in the unselected cohort (n=144).

Diagnostic category Number Percentage

Cardiac* 35 24%

Respiratory* 23 16%

Infective / sepsis 21 15%

Other 16 11%

Mechanical fall 15 10%

Trauma / Orthopaedic 10 7%

Gastrenterology / hepatology 9 6%

Renal 6 4%

Surgical 5 3%

Neurological 4 3% * In patients with a cardiac diagnosis (n=35/144), acute coronary syndrome (n=17) and atrial fibrillation (n=10) accounted for the majority of patients. In those patients with a respiratory diagnosis (n=23/144), pneumonia (n=11) and chronic obstructive pulmonary disease (n=7) accounted for the majority of patients.

Page 39 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

7

Appendix Table 5. Influence of clinical characteristics on the positive predictive value of an elevated cardiac troponin for a diagnosis of type 1 or type 2 myocardial infarction in patients with suspected acute coronary syndrome.

Variable Strata PPV (95%CI) Age, years <40 76.8% (61.3-91.2)

40-49 88.7% (81.9-95.0)

50-59 83.5% (77.7-89.2)

60-69 81.3% (76.3-86.1)

70-79 75.7% (71.4-79.9)

≥80 68.8% (64.8-72.7)

Chest pain Yes 83.5% (81.3-85.7)

No 50.9% (45.5-56.3)

Ischemic ECG Yes 87.9% (85.1-90.6)

No 72.9% (69.7-76.0)

Hypertension Yes 78.0% (74.7-81.2)

No 76.2% (73.0-79.3)

Hyperlipidaemia Yes 83.7% (80.4-87.0)

No 73.4% (70.4-76.3)

Diabetes Yes 75.0% (69.6-80.2)

No 77.6% (75.1-80.0)

Ischemic heart disease Yes 82.1% (78.9-85.2)

No 73.7% (70.6-76.8)

Page 40 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8

Appendix Table 6. Logistic regression models for determinants of elevated cardiac troponin in unselected patients attending the Emergency Department Variable Model 1, OR (95% CI) Model 2, OR (95% CI) Model 3, OR (95% CI)

Age, per 10 years 2.06 ( 1.75 - 2.47 )*** - 1.99 ( 1.66 - 2.42 )***

Creatinine, per 0.1 mg/dL 1.03 ( 1.00 - 1.07 )* 1.03 ( 1.00 - 1.06 ) 1.03 ( 1.00 - 1.06 )*

Sex (Male) 0.56 ( 0.33 - 0.93 )* 0.40 ( 0.24 - 0.66 )*** 0.51 ( 0.30 - 0.86 )*

NEWS, per 1 unit 1.46 ( 1.27 - 1.70 )*** 1.51 ( 1.32 - 1.73 )*** 1.48 ( 1.28 - 1.72 )***

Comorbidity

None (referent) - 1 1

One - 3.78 ( 2.05 – 7.00 )*** 1.1 ( 0.55 - 2.19 )

Two - 5.69 ( 2.97 - 10.98 )*** 1.33 ( 0.64 - 2.76 )

Three - 8.13 ( 3.5 - 18.57 )*** 1.73 ( 0.69 - 4.25 )

Four - 6.88 ( 1.83 - 23.21 )** 2.58 ( 0.65 - 9.28 )

AIC (model fit) 411 478 413

***P value <0.001 **P value <0.05

Page 41 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


9

Appendix Table 7. Case series of patients classified as type 1 or type 2 myocardial infarction in whom the attending clinician did not request cardiac troponin Age Sex Classification Presenting

complaint Known IHD

TnI, ng/L

Ischemia on ECG

Admitted Clinical details

94 M Type 1 Chest pain No 1,377 Yes Yes Late presentation myocardial infarction, not for escalation of care

44 M Type 1 Chest pain No 132 No No Atypical history of chest pain 65 M Type 2 Breathlessness No 46 Yes Yes Patient with multiple comorbidities and

admitted with shortness of breath secondary to exacerbation of COPD

94 F Type 2 Chest pain No 24 No Yes Admitted with fast ventricular response to atrial fibrillation and associated chest tightness

72 M Type 2 Chest pain No 36 Yes Yes Admitted with fast ventricular response to atrial fibrillation and associated chest pain with rate related left bundle branch block

Page 42 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10

Appendix Figure 1. Influence of prevalence on the positive predictive value of an elevated cardiac troponin concentration for the diagnosis of type 1 myocardial infarction.

Red dots represent our cohorts of unselected patients in the Emergency Department (n=1,054) and selected patients with suspected acute coronary syndrome (n=5,815). Black dots represent the reported positive predictive values for cardiac troponin by prevalence of type 1 myocardial infarction in previously published cohorts using high-sensitivity cardiac troponin T (blue) and high-sensitivity cardiac troponin I (red) assays.1-21 Dot size reflects the number of patients in each cohort (small dot <500 patients, medium dot 500-1,000 patients, large dot >1,000 patients). Red line represents the central estimate of the positive predictive value with the 95% confidence interval (dashed blue) derived from the unselected Emergency Department cohort.

31

Figure 1.

Unselected

Aldous

Aldous

Bahrmann

Body

Christ

Collinson

Eggers

Freund

Hammerer−Lercher

HoellerInoue

Invernizi

Lotze

SantalolSebbaneGiannitis

Khan

Normann

Thelin

Suspected ACS

KellerHoeller

Korley

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50

Prevalence of type 1 myocardial infarction (%)

Posi

tive

pred

ictiv

e va

lue

(%)

Page 43 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11

Appendix Figure 2. Prevalence of myocardial injury in unselected patients attending the Emergency Department

All

Sex

Age (years)

Ischaemic heart disease

No. of comorbidities

GFR (ml/min)

Males

Females

<40

40 to 49

50 to 59

60 to 69

70 to 79

Over 80

No

Yes

Zero

One

Two

Over three

< 30

30 to 59

Over 60

144 /1054

55 /505

89 /549

5 /312

6 /157

8 /128

19 /128

31 /142

75 /187

84 /849

59 /193

32 /602

38 /200

42 /151

29 /85

98 /559

17 /81

29 /407

13.7 (11.6−15.8)

11 (8.3−13.7)

16.3 (13.2−19.4)

1.8 (0.5−3.2)

4.1 (1.3−7.2)

6.6 (2.7−10.9)

15.1 (9.2−21.4)

22 (15.4−28.9)

40.2 (33.2−47.2)

9.9 (8−12)

30.7 (24.3−37.2)

5.4 (3.6−7.2)

19.2 (13.8−24.6)

28 (21−35.1)

34.3 (24.5−44.3)

17.6 (14.5−20.8)

21.3 (12.8−30.3)

7.2 (4.8−9.8)

1 2 3 4 0 5 10 15 20 25 30 35 40 45

Prevalence of elevated cardiac torponin > 99th centile (95% Confidence Intervals)

Variable Strata Number Prevalence1(95%1CI)

≥80

≥3*

None*

1

2*

≥*60*

Page 44 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

12

References

1. Invernizzi L, Doka M, Cappellini F, et al. Effectiveness of highly sensitive troponin T

assay for early diagnosis of acute myocardial infarction (AMI). Biochimica Clinica

2013;37(1):36-39.

2. Lotze U, Lemm H, Heyer A, et al. Combined determination of highly sensitive troponin T

and copeptin for early exclusion of acute myocardial infarction: first experience in an

emergency department of a general hospital. Vasc Health Risk Manag 2011;7:509-15.

doi: 10.2147/VHRM.S21753

3. Bahrmann P, Bahrmann A, Breithardt OA, et al. Additional diagnostic and prognostic

value of copeptin ultra-sensitive for diagnosis of non-ST-elevation myocardial infarction

in older patients presenting to the emergency department. Clin Chem Lab Med

2013;51(6):1307-19. doi: 10.1515/cclm-2012-0401

4. Bahrmann P, Christ M, Bahrmann A, et al. A 3-hour diagnostic algorithm for non-ST-

elevation myocardial infarction using high-sensitivity cardiac troponin T in unselected

older patients presenting to the emergency department. J Am Med Dir Assoc

2013;14(6):409-16. doi: 10.1016/j.jamda.2012.12.005

5. Hammerer-Lercher A, Ploner T, Neururer S, et al. High-sensitivity cardiac troponin T

compared with standard troponin T testing on emergency department admission: how

much does it add in everyday clinical practice? J Am Heart Assoc 2013;2(3):e000204.

doi: 10.1161/JAHA.113.000204

6. Thelin J, Borna C, Erlinge D, et al. The combination of high sensitivity troponin T and

copeptin facilitates early rule-out of ACS: a prospective observational study. BMC

Cardiovasc Disord 2013;13:42. doi: 10.1186/1471-2261-13-42

7. Freund Y, Chenevier-Gobeaux C, Bonnet P, et al. High-sensitivity versus conventional

troponin in the emergency department for the diagnosis of acute myocardial infarction.

Crit Care 2011;15(3):R147. doi: 10.1186/cc10270

8. Santalo M, Martin A, Velilla J, et al. Using high-sensitivity troponin T: the importance of

the proper gold standard. Am J Med 2013;126(8):709-17. doi:

10.1016/j.amjmed.2013.03.003

9. Hoeller R, Rubini Gimenez M, Reichlin T, et al. Normal presenting levels of high-

sensitivity troponin and myocardial infarction. Heart 2013;99(21):1567-72. doi:

10.1136/heartjnl-2013-303643

Page 45 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

13

10. Body R, Carley S, McDowell G, et al. Rapid exclusion of acute myocardial infarction in

patients with undetectable troponin using a high-sensitivity assay. J Am Coll Cardiol

2011;58(13):1332-9. doi: 10.1016/j.jacc.2011.06.026

11. Aldous SJ, Florkowski CM, Crozier IG, et al. The performance of high sensitivity

troponin for the diagnosis of acute myocardial infarction is underestimated. Clin Chem

Lab Med 2012;50(4):727-9. doi: 10.1515/cclm.2011.830

12. Aldous SJ, Richards AM, Cullen L, et al. Early dynamic change in high-sensitivity

cardiac troponin T in the investigation of acute myocardial infarction. Clin Chem

2011;57(8):1154-60. doi: 10.1373/clinchem.2010.161166

13. Christ M, Popp S, Pohlmann H, et al. Implementation of high sensitivity cardiac

troponin T measurement in the emergency department. Am J Med 2010;123(12):1134-

42. doi: 10.1016/j.amjmed.2010.07.015

14. Giannitsis E, Kehayova T, Vafaie M, et al. Combined testing of high-sensitivity troponin

T and copeptin on presentation at prespecified cutoffs improves rapid rule-out of non-

ST-segment elevation myocardial infarction. Clin Chem 2011;57(10):1452-5. doi:

10.1373/clinchem.2010.161265

15. Collinson P, Goodacre S, Gaze D, et al. Very early diagnosis of chest pain by point-of-

care testing: comparison of the diagnostic efficiency of a panel of cardiac biomarkers

compared with troponin measurement alone in the RATPAC trial. Heart

2012;98(4):312-8. doi: 10.1136/heartjnl-2011-300723 [published Online First:

2011/11/15]

16. Sebbane M, Lefebvre S, Kuster N, et al. Early rule out of acute myocardial infarction in

ED patients: value of combined high-sensitivity cardiac troponin T and ultrasensitive

copeptin assays at admission. Am J Emerg Med 2013;31(9):1302-8. doi:

10.1016/j.ajem.2013.04.033

17. Inoue K, Suwa S, Ohta H, et al. Heart fatty acid-binding protein offers similar diagnostic

performance to high-sensitivity troponin T in emergency room patients presenting with

chest pain. Circ J 2011;75(12):2813-20.

18. Eggers KM, Venge P, Lindahl B. High-sensitive cardiac troponin T outperforms novel

diagnostic biomarkers in patients with acute chest pain. Clin Chim Acta 2012;413(13-

14):1135-40. doi: 10.1016/j.cca.2012.03.011 [published Online First: 2012/03/30]

19. Normann J, Mueller M, Biener M, et al. Effect of older age on diagnostic and prognostic

performance of high-sensitivity troponin T in patients presenting to an emergency

department. Am Heart J 2012;164(5):698-705 e4. doi: 10.1016/j.ahj.2012.08.003

Page 46 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

14

20. Khan DA, Sharif MS, Khan FA. Diagnostic performance of high-sensitivity troponin T,

myeloperoxidase, and pregnancy-associated plasma protein A assays for triage of

patients with acute myocardial infarction. Korean J Lab Med 2011;31(3):172-8. doi:

10.3343/kjlm.2011.31.3.172

21. Keller T, Zeller T, Ojeda F, et al. Serial changes in highly sensitive troponin I assay and

early diagnosis of myocardial infarction. Jama 2011;306(24):2684-93. doi:

10.1001/jama.2011.1896 [published Online First: 2011/12/29]

Page 47 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

1

APPENDIX: STATISTICAL ANALYSIS PLAN


in the Emergency Department

Anoop SV Shah MD MPH PhD,1 Ala Noaman MD,

1 Amar Vaswani MD,

1 Mathew Gibbins MSc,

1

Megan Griffiths MD,1 Andrew R Chapman MD,1 Fiona Strachan PhD,1 Atul Anand MD,1 Philip

Adamson MD1, Michelle D Souza,

1 Alasdair J Gray MD,

3,4 David A McAllister MD,

2 David E

Newby MD PhD1 and Nicholas L Mills MD PhD1

1BHF Centre for Cardiovascular Science, University of Edinburgh, United Kingdom

2Centre for Population Health Sciences, University of Edinburgh, United Kingdom

3Department of Emergency Medicine, Royal Infirmary of Edinburgh, United Kingdom

4Emergency Medicine Research Group Edinburgh (EMeRGE), United Kingdom

Correspondence and requests for reprints:

Dr Anoop S V Shah

BHF/University Centre for Cardiovascular Science

Chancellor’s Building

University of Edinburgh

Edinburgh EH16 4SB

United Kingdom

Fax: +44 131 242 6379

E-mail: [email protected]

Page 48 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

2

Contents

Contents ............................................................................................................................... 2

1 Introduction .................................................................................................................. 3

1.1 Background ............................................................................................................................. 3

1.2 Study objectives ....................................................................................................................... 4

1.2.1 Hypothesis .......................................................................................................................... 4

2 Statistical analysis plan ................................................................................................. 5

2.1.1 Objectives ........................................................................................................................... 5

2.1.2 General Statistical Principles .............................................................................................. 5

2.1.3 Analysis population ............................................................................................................ 5

2.1.4 Software ............................................................................................................................. 6

3 List of analyses .............................................................................................................. 7

3.1 Baseline Characteristics ......................................................................................................... 7

3.2 Primary outcome ..................................................................................................................... 9

3.2.1 Event summary ................................................................................................................... 9

3.2.2 Primary analysis ................................................................................................................. 9

3.2.3 Secondary analysis ........................................................................................................... 10

4 References ................................................................................................................... 12

Page 49 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

3

1 Introduction

1.1 Background

Patients attending the Emergency Department often undergo simultaneous testing for both cardiac and

non-cardiac conditions to facilitate early diagnosis or discharge (Yiadom et al., 2015). Cardiac

troponin is integral to the diagnosis of myocardial infarction (Mills et al., 2011), and testing is used

widely in the Emergency Department although approach to testing varies between region, country and

health care system (Melanson et al., 2008, Jairam et al., 2011, Yiadom et al., 2015). In a

representative sample of patients attending the Emergency Department in the USA, cardiac troponin

was measured in one in five and over half of those patients who were hospitalised. Moreover, a third

of patients did not have any cardiac symptoms (Makam and Nguyen, 2015). Myocardial injury,

defined as an elevation in cardiac troponin concentration in the absence of myocardial ischemia, is

common in many cardiac and non-cardiac conditions and increasingly recognised using high-

sensitivity cardiac troponin assays (Shah et al., 2015c, Newby et al., 2012). Here cardiac troponin

testing may contribute to diagnostic uncertainty, particularly in those regions where testing is

performed without prior clinical assessment (Melanson et al., 2008, Makam and Nguyen, 2015).

Across three prospective cohorts of patients attending the Emergency Department with and without

suspected acute coronary syndrome, we evaluate the impact of clinical assessment and patient

selection for cardiac troponin testing on the diagnosis of myocardial infarction.

This analysis will include patients with suspected acute coronary syndrome recruited to all sites in

NHS Lothian during the validation phase of the High-Sensitivity Troponin in the Evaluation of

patients with Acute Coronary Syndrome (HighSTEACS) randomised clinical trial (n=5,815)

(clinicaltrials.gov: NCT01852123). The High-STEACS study is a stepped-wedge cluster randomised

controlled trial divided into validation and implementation phases across ten centres in Scotland. The

ARCHITECT STAT high-sensitivity troponin I assay will be run in parallel with the existing

contemporary sensitive troponin I assay (standard care) throughout the validation phase. Only the

results from the contemporary assay are used for clinical decision-making during the validation phase

Page 50 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

4

and the diagnostic threshold for myocardial infarction will remain unchanged. This phase will provide

baseline information on patients with suspected acute coronary syndrome for each site. Centres will

then be randomised 1:1 to early or late implementation of the high-sensitivity assay. In the

implementation phase, all centres will implement the high-sensitivity assay for at least 6 months.

An independent unselected cohort, we identified all consecutive patients (n=1,054) presenting to the

Emergency Department at the Royal Infirmary of Edinburgh, Scotland in whom the attending

clinician performed blood sampling irrespective of their clinical presentation.

1.2 Study objectives

To evaluate the impact of clinical assessment and patient selection for high-sensitivity cardiac

troponin testing in the Emergency Department on the diagnosis of myocardial infarction.

1.2.1 Hypothesis

Clinical assessment to guide the selection of patients for high-sensitivity cardiac troponin testing will

improve the positive predictive value for a diagnosis of myocardial infarction compared to an

unselected approach to testing in the Emergency Department.

Page 51 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

5

2 Statistical analysis plan

2.1.1 Objectives

The objective of this SAP is to describe the statistical analyses contributing to the final report and

primary publication(s) of the analyses described below.

2.1.2 General Statistical Principles

Baseline data will be summarised as proportions for categorical data. Parametric and non-parametric

continuous data will be presented as mean and standard deviations or median and interquartile range

respectively.

2.1.3 Analysis population

2.1.3.1 Unselected emergency department cohort

The unselected Emergency Department cohort will consist of consecutive patients presenting to the

Emergency Department at the Royal Infirmary of Edinburgh, Scotland in whom the attending

clinician has performed blood sampling irrespective of their clinical presentation. Excess serum stored

will be used to measure high-sensitivity cardiac troponin I. The results will not be used to guide

clinical care and requests for cardiac troponin for clinical use will remain at the discretion of the

attending clinician. To ensure confidence in our estimates of the prevalence of type 1 myocardial

infarction in the unselected cohort, we determined that at least 1,000 patients would be required to

estimate a prevalence of 2.5% with an upper 95% confidence limit of less than 5% at greater than

90% power with an α of 0.05.

Page 52 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

6

2.1.3.2 Suspected acute coronary syndrome

Approximately 6,000 patients with suspected acute coronary syndrome will be included from two

cohorts across three centres as previously described (Shah et al., 2015b, Shah et al., 2015a). In all of

these patients troponin was measured using both the contemporary sensitive and high-sensitivity

troponin I assay. Only the contemporary assay results were used for clinical decision-making.

2.1.3.3 Adjudication of diagnosis

The diagnosis of each patient was adjudicated using the high-sensitivity troponin assay. Where serial

samples were available, we used peak and change in troponin concentration. Two independent

cardiologists reviewed all clinical information, including non-invasive and invasive investigations,

and outcomes from admission to 30-days. Patients were classified as having type 1 myocardial

infarction, type 2 myocardial infarction or myocardial injury.

2.1.4 Software

Analyses will be performed in R (Vienna, Austria).

Page 53 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

7

3 List of analyses

3.1 Baseline Characteristics

Summary statistics will be provided for the baseline characteristics for both the Emergency

Department cohort and suspected acute coronary syndrome cohorts. The following variables where

available will be reported:

• Age (years) – mean(standard deviation)

• Sex (male/female) n(%)

• Cardiovascular risk factors (smoking status, hypertension, hyperlipidaemia, diabetes mellitus)

n(%)

• Past medical history (ischaemic heart disease, stroke, previous coronary revascularisation)

n(%)

• Medical therapy on admission (dichotomous) n(%)

o Aspirin

o Clopidogrel

o Beta blocker

o Angiotensin converting enzyme (ACE) inhibitor / angiotensin receptor blocker

(ARB)

o Statin

• 12-lead electrocardiogram n(%)

o ST elevation (yes/no)

o ST depression (yes/no)

o T wave inversion (yes/no)

Patients prospectively recruited into the Emergency Department cohort will be divided into five

groups a priori; those patients with cardiac troponin concentrations >99th centile upper reference limit

(men >34 ng/L and women >16 ng/L) (group 5) and the remaining patients separated into four equal

Page 54 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

8

quartiles (group 1-4). Continuous variables will be compared using parametric and non-parametric

tests as appropriate. Categorical variables will be compared using the chi-square or fisher’s exact test

where appropriate.

Page 55 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

9

3.2 Primary outcome

3.2.1 Event summary

The following primary analyses will be presented:

• Predicted positive predictive value for a range of pre-test probabilities from 1% to 50%

• Impact of clinical features and bedside investigations on the positive predictive value in

patients with suspected acute coronary syndrome.

Positive predictive value refers to the post-test probability of a diagnosis of type 1 myocardial

infarction in patients with any cardiac troponin concentration above the 99th centile sex-specific upper

reference limit.

3.2.2 Primary analysis

3.2.2.1 Impact of pre-test probability on the diagnosis of myocardial infarction

Using the unselected Emergency Department cohort, we will calculate the sensitivity, specificities and

positive predictive values for a range of pre-test probabilities, including that observed in this cohort.

Using the unselected Emergency Department cohort we will calculate the observed prevalence of type

1 myocardial. We sampled from a Beta distribution in R (10,000 iterations) in order to estimate

confidence intervals for all proportions, using a Jeffrey’s prior (both shape parameters = 0.5), as this

approach allows us to straightforwardly estimate the positive predictive values for different a priori

values, and this approach also has excellent frequentist properties.

3.2.2.2 Impact of clinical features and baseline investigations in patients with suspected acute

coronary syndrome on the diagnosis of myocardial infarction

Using the same method as in the unselected Emergency Department cohort, in the two cohorts of

consecutive patients with suspected acute coronary syndrome we will calculate the observed positive

Page 56 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

10

predictive value and the 95% confidence interval with both cohorts combined. We will then calculate

the observed positive predictive value across pre-specified groups stratified by age, presenting

symptoms, risk factors, presence of ischemia on ECG and known ischemic heart disease.

3.2.3 Secondary analysis

3.2.3.1 Prevalence and determinants of myocardial injury in the Emergency Department

In patients recruited to the Emergency Department cohort we will calculate the prevalence of

myocardial injury (troponin concentrations >99th centile upper reference limit) across the whole

population and stratified by pre-specified groups including age, sex, known ischemic heart disease,

renal function and number of comorbidities (hypertension, ischemic heart disease, peripheral vascular

disease, diabetes mellitus or cerebrovascular disease).

We will develop logistic regression models to explore clinical factors associated with myocardial

injury (defined as cardiac troponin above the 99th centile upper reference limit). We will sequentially

adjust for age, sex, renal function, comorbidity and the National Early Warning Score (NEWS). The

NEWS score will be calculated for patients incorporating six simple physiological parameters (pulse

rate, blood pressure, temperature, oxygen saturations, level of consciousness and respiratory rate

measured at admission.

3.2.3.2 Association of cardiac troponin levels at presentation and mortality in the unselected

Emergency Department

All patients recruited in the Emergency Department cohort will be followed up for death from any

cause using regional and national registries including the General Register of Scotland (Ford et al.,

2007). This method allowed capture of all in-hospital and community deaths ensuring complete

follow-up. Occurrence of death will be modeled using a Cox proportional hazards regression model

Page 57 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

11

with troponin (natural log) as the explanatory variable both as a single parameter and after adjusting

for age, sex, renal function and NEWS score. We will examine the linearity of the association

between troponin concentration and risk of death using penalized splines .

Page 58 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

12

4 References

FORD, I., MURRAY, H., PACKARD, C. J., SHEPHERD, J., MACFARLANE, P. W. & COBBE, S.

M. 2007. Long-term follow-up of the West of Scotland Coronary Prevention Study. The New

England Journal Of Medicine, 357, 1477-86.

JAIRAM, S., JONES, P., SAMARAIE, L., CHATALINE, A., DAVIDSON, J. & STEWART, R.

2011. Clinical diagnosis and outcomes for Troponin T 'positive' patients assessed by a high

sensitivity compared with a 4th generation assay. Emerg Med Australas, 23, 490-501.

MAKAM, A. N. & NGUYEN, O. K. 2015. Use of cardiac biomarker testing in the emergency

department. JAMA Intern Med, 175, 67-75.

MELANSON, S. E., CONRAD, M. J., MOSAMMAPARAST, N. & JAROLIM, P. 2008.

Implementation of a highly sensitive cardiac troponin I assay: test volumes, positivity rates

and interpretation of results. Clin Chim Acta, 395, 57-61.

MILLS, N. L., CHURCHHOUSE, A. M., LEE, K. K., ANAND, A., GAMBLE, D., SHAH, A. S.,

PATERSON, E., MACLEOD, M., GRAHAM, C., WALKER, S., DENVIR, M. A., FOX, K.

A. & NEWBY, D. E. 2011. Implementation of a sensitive troponin I assay and risk of

recurrent myocardial infarction and death in patients with suspected acute coronary

syndrome. JAMA : the journal of the American Medical Association, 305, 1210-6.

NEWBY, L. K., JESSE, R. L., BABB, J. D., CHRISTENSON, R. H., DE FER, T. M., DIAMOND,

G. A., FESMIRE, F. M., GERACI, S. A., GERSH, B. J., LARSEN, G. C., KAUL, S.,

MCKAY, C. R., PHILIPPIDES, G. J. & WEINTRAUB, W. S. 2012. ACCF 2012 expert

consensus document on practical clinical considerations in the interpretation of troponin

elevations: a report of the American College of Cardiology Foundation task force on Clinical

Expert Consensus Documents. Journal Of The American College Of Cardiology, 60, 2427-63.

SHAH, A. S., ANAND, A., SANDOVAL, Y., LEE, K. K., SMITH, S. W., ADAMSON, P. D.,

CHAPMAN, A. R., LANGDON, T., SANDEMAN, D., VASWANI, A., STRACHAN, F. E.,

FERRY, A., STIRZAKER, A. G., REID, A., GRAY, A. J., COLLINSON, P. O.,

MCALLISTER, D. A., APPLE, F. S., NEWBY, D. E., MILLS, N. L. & HIGH, S. I. 2015a.

High-sensitivity cardiac troponin I at presentation in patients with suspected acute coronary

syndrome: a cohort study. Lancet, 386, 2481-8.

SHAH, A. S., GRIFFITHS, M., LEE, K. K., MCALLISTER, D. A., HUNTER, A. L., FERRY, A. V.,

CRUIKSHANK, A., REID, A., STODDART, M., STRACHAN, F., WALKER, S.,

COLLINSON, P. O., APPLE, F. S., GRAY, A. J., FOX, K. A., NEWBY, D. E. & MILLS, N.

L. 2015b. High sensitivity cardiac troponin and the under-diagnosis of myocardial infarction

in women: prospective cohort study. BMJ, 350, g7873.

SHAH, A. S., MCALLISTER, D. A., MILLS, R., LEE, K. K., CHURCHHOUSE, A. M., FLEMING,

K. M., LAYDEN, E., ANAND, A., FERSIA, O., JOSHI, N. V., WALKER, S., JAFFE, A. S.,

FOX, K. A., NEWBY, D. E. & MILLS, N. L. 2015c. Sensitive troponin assay and the

classification of myocardial infarction. Am J Med, 128, 493-501 e3.

YIADOM, M. Y., JAROLIM, P., JENKINS, C., MELANSON, S. E., CONRAD, M. &

KOSOWSKY, J. M. 2015. Diagnostic implications of an elevated troponin in the emergency

department. Dis Markers, 2015, 157812.

Page 59 of 59


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960