Evidence
Alert

Decision aids quickly and accurately rule out heart attack for almost half of all patients tested

Decision aids provided immediate reassurance to almost half of all patients arriving at emergency departments with a suspected heart attack. Use of the aids, which are based on a single blood test, accurately ruled out heart attacks in patients with relevant symptoms. Aids could provide early reassurance for worried patients and families and avoid many unnecessary hospital admissions, researchers found. 

This study was the first direct comparison of four decision aids used to decide which patients with symptoms of a possible heart attack need further tests. It found that the Troponin-only Manchester Acute Coronary Syndromes (T-MACS) and the Emergency Department Assessment of Chest Pain (EDACS) decision aids were the most accurate. 

What’s the issue?

Diagnosing heart attacks in patients arriving in the emergency department with chest pain and other related symptoms used to be time-consuming. Patients had a series of blood tests over several hours to measure the levels of cardiac troponins, which are proteins involved in regulating heartbeat.

More recently, highly accurate tests for cardiac troponin have helped doctors diagnose heart attacks more rapidly.  

Decision aids can be used to combine the outcome of a single blood test with information on the patient’s symptoms, previous medical history and findings from physical examinations, using a mathematical algorithm. Use of decision aids may improve diagnostic accuracy, but several decision aids exist, and it is not known which is most accurate. This study compared the diagnostic accuracy of four: 

  • Troponin-only Manchester Acute Coronary Syndromes (T-MACS), 
  • History, ECG, Age, Risk factors and Troponin (HEART), 
  • Thrombolysis in Myocardial Infarction (TIMI), and 
  • Emergency Department Assessment of Chest Pain (EDACS).

What’s new?

This study analysed 999 patients arriving at 14 emergency departments in England. The researchers took a blood test when patients first arrived in hospital. Each of the four decision aids was used to diagnose a suspected heart attack. It found:

  • T-MACS ruled out heart attacks in 46.5% patients and missed only one heart attack 
  • EDACS ruled out heart attacks in 48.3% patients but missed three heart attacks, a result only slightly less accurate, and not significantly different, from T-MACS 
  • HEART and TIMI had lower diagnostic accuracy than T-MACS and EDACS and therefore are unlikely to be used in clinical practice. 

As well as ruling out heart attacks, T-MACS identified a further one in ten patients with a suspected heart attack who were at very high risk and in need of rapid treatment. The remaining patients had a low or moderate risk of a heart attack and needed a second blood test for cardiac markers. 

Why is this important?

T-MACS and EDACS identified the likelihood of a heart attack in almost half of all patients tested, removing the need for multiple blood tests. This could save time and NHS resources, and either provide reassurance or identify the need to act urgently. 

The remaining patients needed to have a second blood test for cardiac markers. T-MACS identified low risk patients within this group who could be treated as outpatients, without taking up hospital beds. 

What’s next?

Some NHS hospitals already use the four decision aids in clinical practice. The paper’s lead author, Richard Body, is working with Health Innovation Manchester to implement T-MACS at hospitals across Greater Manchester. It is likely that almost all hospitals in the region will start using T-MACS in 2020. Others will first collect real-world data. 

The researchers are also working with the company Cambio CDS to develop an electronic interface to link T-MACS to electronic patient records automatically. This would simplify the process of putting it to use in new sites. The National Institute for Health and Care Excellence (NICE) has produced an adoption toolkit to support sites in designing diagnostic pathways tailored to their individual hospitals, and the Academic Health Science Network (AHSN) is producing a series of webinars, which will facilitate implementation of decision aids in hospitals.

Further NIHR-funded research (the PRESTO study) will determine whether the T-MACS and HEART decision aids can be used in an ambulance with a portable troponin blood test. If the decision aids are accurate in that environment, they could reduce the numbers of patients transferred to busy Emergency Departments.

The authors are investigating whether machine learning could continually and automatically refine the T-MACS algorithm during clinical use. This would help maintain its accuracy over time. 

You may be interested to read

The full paper: Body R, and others. Comparison of four decision aids for the early diagnosis of acute coronary syndromes in the emergency department. Emergency Medicine Journal. 2020;37:8-13

 T-MACS project page, Health Innovation Manchester

Support for the rapid uptake for diagnostic tests by the Accelerated Access Collaborative (AAC), Academic Health Science Network (AHSN) for the North East and North Cumbria

NICE adoption support resource for myocardial infarction (acute): Early rule out using high-sensitivity troponin tests (Elecsys Troponin T high-sensitive, ARCHITECT STAT High Sensitive Troponin-I and AccuTnI+3 assays) (2014)

Conflict of interest

RB has received speaker fees, grants, and reagents from various pharmaceutical companies.

Funding

The study was sponsored by Manchester University NHS Foundation Trust and supported by the NIHR Clinical Research Network (CRN). Siemens provided reagents without charge for the purpose of this study.

Commentaries

Study author

The results show that at least some of the decision aids are accurate enough to be used in clinical practice, and they could lead to tangible benefits for patients and the NHS.

But first, we need data on the effects of real-world implementation of the decision aids. Is their accuracy maintained over time and are they actually helping to reduce unnecessary hospital admissions?

Second, we need to build on the functionality of the decision aids. T-MACS calculates the probability of a heart attack for each individual patient. We now need to harness the power of that information. We could, for example, use this information to make personalised treatment recommendations to share with patients. By presenting alternative treatment options, we could engage patients in a shared decision-making process.

Richard Body, Professor of Emergency Medicine, University of Manchester

Researcher

The paper provides clinically relevant information in a contemporary UK population. The data should be useful to clinicians working in unselected acute medical care in the UK, and potentially further afield.

Colin Berry, Professor of Cardiology and Imaging, University of Glasgow