• 
• 

Three Part Question

In [patients with suspected cardiac chest pain, who calls an ambulance], can [paramedics use troponin point of care test] to [safely ‘rule out’ Acute Coronary Syndromes in the pre-hospital emergency settings]?

Clinical Scenario

At 22:40 the ambulance service operation centre receives a call for a 56-year-old patient complaining of chest pain and an ambulance unit is dispatched to the patient. On paramedic arrival, the woman is alert and orientated and shows no evidence of diaphoresis. She has no previous medical history. The chest pain had self-resolved 5 minutes ago. On examination, she has a clear airway, respiration is shallow, talking in complete sentences; lung sounds clear, skin warm and not clammy. The electrocardiogram (ECG) recorded on the scene shows benign early repolarization with no other abnormalities. There are no other pertinent findings. Her vital signs are: respiratory rate 20/minute, heart rate 65 beats per minute, oxygen saturation 96% in air, blood pressure 124/62. She stated that she suddenly could not catch her breath while stood up, then the chest pain started. She got dizzy, sat down, and called 999 but is now feeling back to normal. You are aware that the history, physical examination and ECG cannot be used to ‘rule out’ an acute coronary syndrome (ACS) alone. You wonder whether a point of care troponin test could help you to ‘rule out’ ACS without requiring transfer to hospital.

Search Strategy

MEDLINE (Ovid) 1946 - March Week 3 2018
EMBASE (Ovid) 1974 - 2017 September 23

[Emergency Medical Services/ or paramedic*.mp. or Emergency Medical Technicians/ OR Ambulance*.mp. OR (pre-hospital.mp. OR prehospital.mp.) OR (out of hospital.mp. OR out-of-hospital.mp.)] AND [Acute Coronary Syndrome/ OR Myocardial Infarction/ OR Chest pain/ ] OR [Troponin I/ OR Troponin T/]

CINAHL Plus (EBSCOhost)

(troponin I OR troponin T OR ctni OR cTnI) AND (Emergency Medical Services OR Emergency Medical Technicians OR paramedic* OR pre-hospital OR prehospital OR ambulance* OR out of hospital OR out-of-hospital) AND (acute coronary syndrome OR Myocardial Infarction OR Chest Pain )

Search Outcome

The initial search and cross-referencing total of 187 papers were identified 11 papers (9 studies) of were relevant to the three-part question and the rest 176 were considered to be irrelevant or insufficient quality of evidence. We removed articles not written in English, which were published before the year 2000 (the first year when cardiac troponin was cited as the reference standard biomarker for diagnosing AMI) or where only the abstract was published.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Rasmussen et al, 2017, Denmark	16,449 individual patients who presented with symptoms suggestive of an AMI in the prehospital setting, and who underwent prehospital POC-cTnT testing before hospital admission. (19,615 cases with 18,712 POC-cTnT)	Observational population-based follow-up study This is a retrospective study (POC testing evaluated in routine practice)	Diagnostic accuracy for AMI	In total: Sensitivity 44.2% (42.1–46.3), specificity 92.8% (92.4–93.2), PPV 44.9% (42.8–47.1), NPV 92.6 %(92.2–93.0). Excluding patients with STEMI: Sensitivity 43.6% (41.0–46.2), specificity 92.8 (92.4–93.2), PPV 34.2% (32.0–36.5), NPV 95.1 (94.7–95.4).	Unclear reference standard. Final clinical diagnoses retrieved but the proportion of patients undergoing laboratory troponin testing and its timing are not stated Clinicians were not blinded to POC cTnT results, meaning that the study may be subject to important verification and incorporation bias
Ezekowitz, 2015, Canada	546 patients aged 30 years who called for an ambulance with symptoms of acute chest discomfort where paramedics suspected cardiac chest pain. All patients were transported to hospital, but were randomized to receive pre-hospital point of care troponin testing or standard care, followed by routine care in the ED Excluded: patients with ST elevation myocardial infarction (STEMI) on 12-lead ECG, cardiac arrest syncope or central nervous system symptoms, or a non-cardiovascular diagnosis (e.g. chest trauma, severe asthma), ventricular tachycardia, or atrial fibrillation with a heart rate >110 beats per minute.	Prospective, open-label, blinded-end point design randomised controlled trial	The time from first medical contact to ED discharge. (time from the arrival time of the paramedics on scene at the patient location to time of patient discharge from the ED or admission to hospital was documented.)	Despite the delay caused by doing the test at the scene, pre-hospital POC cTnI testing shortened the time from first medical contact to final discharge from the ED (median: 8.8 hours) compare to patients and usual care arm (median: 9.1 hours) for a patient with chest pain, however, no statistically significant difference in the time between the two groups.	Broad inclusion criteria, which could lead to the inclusion of patients without cardiac disease. Unclear reference standard to diagnose ACS and troponin.
			POC cTnI >0.03 ng/mL to diagnose ACS	Sensitivity 44%, specificity 96%, PPV 73.3%, NPV 87.2% for ACS.
			30-day events (all-cause mortality, rehospitalisation, or ED re-visits)	All-cause death (Control: 1.4%, POC:1.2%). Rehospitalisation (Control: 6.2% , POC: 6.8%), ED visit, (Control: 11.6%, POC:13.6%), ED visit or rehospitalisation (Control: 16.1%, POC:18.4%).
Ezekowitz et al, 2014, Canada	491 patients (480 patients were analysable) aged 18 years and over who called the ambulance for symptoms of acute chest discomfort where paramedics suspected cardiac chest pain. Those patients were randomized to usual care or point of care troponin testing. All patients were transported to hospital Excluded: patients with STEMI the initial 12-lead ECG, or other a previous non-cardiovascular diagnosis associated with recurrent dyspnoea or chest discomfort (eg, chest trauma, asthma).	Prospective, open-label, blinded-end point design randomised controlled trial	Time from first medical contact to ED discharge (time from the arrival time of the paramedics on scene to time of patient discharge from the ED or admission to hospital)	9.2 hours (IQR 7.3-11.1h) in the POC group vs 8.8 hours (IQR 6.3-12.1h) for controls, p=0.6	The trial was stopped early because the enrolment rate was less than expected without any apparent difference in the primary outcome.
			30-day events: all-cause mortality, rehospitalisation, or ED re-visits	No significant differences. Per protocol analysis (240 in each arm): All-cause death (Control: 5, POC: 4, p=1.000). Rehospitalisation (Control: 25, POC: 14, p=0.066), ED visit, (Control: 38, POC: 41, p=0.712), ED visit or rehospitalisation (Control: 57, POC:48, p=0.320).
			Diagnostic accuracy of POC cTnI >0.03ng/ml for ACS	Sensitivity 31%, Specificity 89.7%, PPV 34.6%, NPV 88.1%
Stengaard et al, 2013 Denmark	985 patients with ongoing or prolonged periods of chest pain or discomfort within the past 12 hours, acute dyspnoea. Excluded: known pulmonary disease, or a clinical suspicion of AMI All patients underwent testing for POC cTnT in the ambulance (Roche Cobas h232). Cut-off for ‘rule out’: 50ng/L Reference standard: adjudicated diagnosis based on all clinical information. (Requirements for subsequent troponin testing not clearly stated)	Prospective cohort study.	The feasibility and ability of quantitative POC troponin T assay to identify patients with AMI.	Blood sample obtained in 1,075/1,099 cases (98%) with a successful POC cTnT analysis in 990 (90%)	Some of the baseline data and duration of symptom were missing or incomplete. Unclear protocol for reference standard troponin testing No sample size calculation.
			Diagnostic accuracy of prehospital POC cTnT analysis for AMI in the full cohort (985)	Sensitivity 39% (95% CI, 32% to 46%) Specificity 95% (95% CI, 94% to 97%) PPV 68% (95% CI, 59% to 77%) NPV 86% (95% CI, 84% to 88%)
			Diagnostic accuracy of prehospital POC-cTnT analysis when testing 0-60 minutes from symptom onset for diagnosis AMI (923 cases with data on symptom duration)	Sensitivity 27% (95% CI, 18% to 38%) Specificity 95% (95% CI, 93% to 97%) PPV 58% (95% CI, 41% to 75%) NPV 84% (95% CI, 80% to 88%)
			Diagnostic accuracy of prehospital POC-cTnT analysis when testing 60-120 minutes from symptom onset for diagnosis AMI (923 cases with data on symptom duration)	Sensitivity 38% (95% CI, 25% to 52%) Specificity 97% (95% CI, 92% to 99%) PPV 80% (95% CI, 59% to 95%) NPV 81% (95% CI, 75% to 87%)
			Diagnostic accuracy of prehospital POC-cTnT analysis when testing >120 minutes from symptom onset for diagnosis AMI (923 cases with data on symptom duration)	Sensitivity 52% (95% CI, 40% to 65%) Specificity 94% (95% CI, 90% to 97%) PPV 69% (95% CI, 55% to 82%) NPV 89% (95% CI, 84% to 92%)
Sorensen et al, 2011, Denmark	958 patients (928 patient data was analysable) who were transported by the emergency medical services due to suspected acute coronary syndrome chest pain. All patients underwent pre-hospital POC cTnT testing (Roche Trop T Sensitive, qualitative assay, cut-off for ‘rule out’ 100ng/L) Reference standard: Adjudicated diagnosis, based on both pre-hospital POC cTnT concentrations and in-hospital cTnT concentrations (unclear timing)	Prospective cohort study.	Feasibility of pre-hospital cardiac troponin testing	Pre-hospital POC cTnT testing attempted in 958 patients and was successful in 928 (97%)	Unclear timing of reference standard cTnT testing Incorporation bias: pre-hospital POC cTnT concentrations considered when adjudicating diagnoses No sample size calculation.
			Diagnostic accuracy for AMI	Sensitivity 31% Specificity 99% PPV 91% NPV 84%
Ecollan et al, 2007, France	108 patients who called an emergency phone number for chest pain Patients underwent cTnI testing (Biosite Triage) in the pre-hospital phase (cut-off set at 400ng/L) Patients underwent serial laboratory cTnI testing every 6h for 24h after arrival at hospital. AMI defined as any positive cTn test	Prospective cohort study	Diagnostic accuracy for AMI	Sensitivity 21.8% (11.8%–35%), Specificity 100%	While patients underwent delayed cTn testing in this study, the definition of AMI was suboptimal (any raised cTn). Unconventionally high cTnI cut-off evaluated
Di Serio et al, 2006, Italy	A total of 53 patients with suspected cardiac chest pain who were transported to a hospital by an ambulance. Patients underwent cTnI testing (Abbott i-Stat, using the assay available in 2006; cut-off 90ng/L) Reference standard: In-hospital laboratory cTnI measured over 24 hours. AMI defined by European Society of Cardiology guidelines but unclear adjudication process	Prospective cohort study	Diagnostic accuracy for AMI	Sensitivity 91% (95% CI, 71% to 99%), Specificity 87% (95% CI, 70% to 96%), PPV 83%(95% CI, 66% to 92%), NPV 93 % (95% CI, 87% to 98%)	Small sample size. Cut-off bias. Unclear adjudication process for AMI, unclear cut-off for diagnosing AMI in hospital
Svensson et al, 2003, Sweden	511 patients were recruited on 538 occasions for patients who called for an ambulance with suspected cardiac chest pain for more than 15 minutes within the last 6 h, without known lung disease. Patients underwent POC cTnI testing using a qualitative assay with a cut-off of 100ng/L (Cardiac Status Spectral Diagnostics) AMI defined based on symptoms, ECG Q waves & CK-MB >10ng/L	Prospective cohort study	Diagnostic accuracy for AMI	Sensitivity 12.1% Specificity not stated	Outdated reference standard for AMI No sample size calculation Specificity not stated and cannot be calculated Unconventionally high cTnI cut-off used
Roth et al, 2001, Israel	777 patients with “non-specific symptoms that may have been due to a coronary event” for 6-48h. Non-diagnostic ECG in ambulance. Patients underwent cTnI testing (Cardiac Status; manufacturer and cut-off not stated) Follow-up for 7 days. Unclear reference standard for AMI. Discharge diagnoses retrieved	Prospective observational study	Diagnostic accuracy for AMI.	PPV 36.7% NPV100%	Unclear reference standard for AMI Manufacturer of cTnI assay not stated Study took place in 1999-2000. Limited validity in the high sensitivity troponin era Sensitivity and specificity not reported

Comment(s)

Diagnosing ACS in patients with chest pain in the pre-hospital settings is one of most challenging tasks. As cardiac biomarker testing is central to the diagnosis of AMI, there are incredible potential advantages to using cardiac biomarker POC assays but current evidence is limited in the pre-hospital setting. The evidence that is available suggests that POC troponin assays alone are insufficiently sensitive to ‘rule out’ ACS in the pre-hospital settings. Future research should focus on the use of POC troponin assays in conjunction with validated decision aids. The development of POC troponin assays with greater analytical sensitivity and precision is likely to improve diagnostic performance.

Editor Comment

JB

Clinical Bottom Line

Based on the currently available evidence, POC troponin assays are insufficiently sensitive to ‘rule out’ ACS in the pre-hospital environment.

References

1. Rasmussen MB, Stengaard C, Sørensen JT, et al. Predictive value of routine point-of-care cardiac troponin T measurement for prehospital diagnosis and risk-stratification in patients with suspected acute myocardial infarction. European Heart Journal: Acute Cardiovascular Care 2017;2048872617745893.
2. Ezekowitz JA, Welsh RC, Weiss D, et al. Providing Rapid Out of Hospital Acute Cardiovascular Treatment 4 (PROACT-4). Journal of the American Heart Association 2015;4(12):e002859.
3. Ezekowitz JA, Welsh RC, Gubbels C, et al. Providing Rapid Out of Hospital Acute Cardiovascular Treatment 3 (PROACT-3). Canadian Journal of Cardiology 2014;30(10):1208-1215.
4. Stengaard C, Sørensen JT, Ladefoged SA, et al. Quantitative point-of-care troponin T measurement for diagnosis and prognosis in patients with a suspected acute myocardial infarction. American journal of cardiology 2013;112(9):1361-1366.
5. Sørensen JT, Terkelsen CJ, Steengaard C, et al Prehospital troponin T testing in the diagnosis and triage of patients with suspected acute myocardial infarction. American journal of cardiology 2011;107(10), 1436-1440.
6. Ecollan P, Collet JP, Boon G, et al. Pre-hospital detection of acute myocardial infarction with ultra-rapid human fatty acid-binding protein (H-FABP) immunoassay. International journal of cardiology. 2007;119(3):349-354.
7. Di Serio F, Lovero R, Leone M, et al. Integration between the tele-cardiology unit and the central laboratory: methodological and clinical evaluation of point-of-care testing cardiac marker in the ambulance. Clinical Chemistry and Laboratory Medicine 2006;44(6):768-773.
8. Svensson L, Axelsson C, Nordlander R, et al. Elevation of biochemical markers for myocardial damage prior to hospital admission in patients with acute chest pain or other symptoms raising suspicion of acute coronary syndrome. Journal of internal medicine 2003;253(3):311-319.
9. Roth A, Malov N, Golovner M, et al. The “SHAHAL” experience in Israel for improving diagnosis of acute coronary syndromes in the prehospital setting. American Journal of Cardiology 2001;88(6):608-10.