• 
• 

Three Part Question

In [adults with suspected cardiac chest pain], can [the HEAR score] be used to [identify very low-risk patients without the need for troponin testing]?

Clinical Scenario

A fit and well 30-year-old male presents to the emergency department (ED) with central lower chest pain that came on at rest today and lasted 4 hours. He describes a “heavy ache” that radiated to his neck and made him feel sick, but denies vomiting or diaphoresis. He has had indigestion in the past but this felt higher and more severe than previous episodes. His observations, examination, and ECG are normal. He has no risk factors for coronary artery disease, but is worried about the possibility of a heart attack.

You clinically suspect indigestion but feel a cardiac cause might be possible. He hasn’t had bloods sent and has been in the ED for 3 hours already. His HEAR score is 1 (due to a moderately suspicious history) and you wonder whether you really need to send a troponin in order to complete the HEART score and exclude an acute coronary syndrome (ACS).

Search Strategy

Ovid Medline (1946-present) and Embase (1974-present) databases searched on 17th November 2021 (Sarah Rudd, BRISTOL, UK: North Bristol Library and Information Service):

1) (“History, Electrocardiogram, Age, and Risk” adj3 scor*).mp
2) “HEAR scor*”.mp
3) 1 or 2

Search Outcome

The Medline database returned 17 results, Embase returned 33 results. Initial screening found six original research papers that evaluated the HEAR score in the ED setting,1-6 and two further papers from the pre-hospital setting.7,8 Manual search of reference and citation lists found two additional papers, one from the primary care setting,9 and one based in ED.10 The seven articles from the ED setting are summarised in Table 1.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Smith L et al, 2020 USA	Patients aged > 21 with chest pain due to possible ACS, with a completed HEAR score and at least one troponin requested. Exclusions: STEMI; acute ECG ischaemia (> 1 mm ST depression or contiguous T-wave inversion); known Ischaemic Heart Disease (IHD)	Secondary analysis of prospective cohort study of 3,809 patients from three EDs	30d MACE:	182/3809 = 4.8%	543/3809 (14.3%) patients incomplete data presumed MACE free 2 of 4 MACE events were deaths due to malignancy complications rather than cardiac events Analysis excludes patients with known IHD, under-estimate sensitivity, exaggerate % of patients classified HEAR<2
			% HEAR<2:	447/3809 = 11.7%
			HEAR<2 sensitivity:	178/182 = 97.8% (95%CI 94.5% to 99.4%)
			HEAR<2 NPV:	443/447 = 99.1% (95%CI 97.7% to 99.8%)

Comment(s)

The HEAR score (Table 1) is a clinical decision rule which, like the PERC score,11 attempts to identify low-risk patients with chest pain that don’t stand to benefit from further testing. Although one small study defined ‘low-risk’ as a HEAR score <4, the other six studies all used a HEAR score <2 which will therefore be the focus of this review. In these six studies a HEAR score <2 identified between 11% and 31% of patients for discharge without the need for troponin testing. The sensitivity for MACE at 30-45 days was between 97.6% and 100% with lower 95%CI bounds between 87.7% and 95.6%. At least 5 of the 14 ‘missed MACE’ events from the six studies however were from non-IHD causes (sepsis, malignancy, aortic aneurysm, pericarditis).1,3 The negative predictive value (NPV) is inversely proportional to the prevalence of MACE, meaning we should interpret Moumneh et al.’s (2021) results with caution as the 30d MACE was only 1.5%.3 In the other five studies however (30d or 45d MACE 4.8-10.7%), the NPV was between 99.1% and 100% (i.e. 0-0.9% risk of MACE) with lower 95%CI bounds between 96.6% and 98.1%, giving a maximum 3.4% risk of MACE at 30-45 days. The findings come with several caveats. All studies include patients that had troponins sent as part of routine care which may have influenced the ‘History’ aspect of the HEAR score if known to the clinician at the time. Only one study (by Todd et al) excluded patients in whom the troponin result was already known.6 Incomplete data classified as ‘MACE-free’ and incomplete troponin testing may have also missed some MACE. Furthermore, all studies were reanalyses of observational data which means they don’t tell us how the HEAR score performs in real practice, or if its use will actually reduce testing rates and ED length of stay.

Clinical Bottom Line

These observational studies suggest a HEAR score <2 can identify patients with a less than 1% risk of MACE at 30-45 days without the need for troponin testing. However, the 95% CIs imply this risk could be as high as 3.4%, which would be unacceptable to most clinicians. Prospective implementation studies are needed to assess real world diagnostic performance, impact on blood test utilisation, and ED length of stay.

References

1. Smith L, Ashburn N, Snavely A et al. Identification of very low-risk acute chest pain patients without troponin testing. Emerg Med J 2020; 37:690-695.