Prehospital HEART scoring with point-of-care troponin ruled out 30-day death with 99.8% negative predictive value

Design

systematic review and individual patient data meta-analysis of 6 prospective studies

Population

5,239 patients with suspected NSTE-ACS assessed by emergency medical services

Primary outcome

sensitivity and negative predictive value of low-risk classification for 30-day outcomes

Effect

mortality sensitivity 93.2% (83.5-98.1), NPV 99.8% (99.5-99.9); MACE sensitivity 92.8% (88.7-95.5), NPV 97.2% (92.1-99.0)

An individual patient data meta-analysis pooled six prospective studies of prehospital risk stratification in suspected non-ST-elevation acute coronary syndrome, 5,239 patients in total. All six used a clinical risk score derived from HEART - history, ECG, age, risk factors and troponin - with troponin measured at the point of care by the ambulance crew.

For patients classified low risk, sensitivity for 30-day all-cause mortality was 93.2% (83.5 to 98.1) with a negative predictive value of 99.8% (99.5 to 99.9). For death or acute myocardial infarction, sensitivity was 91.8% and NPV 97.3%. For major adverse cardiac events, sensitivity 92.8% and NPV 97.2%. Lowering the cut-off raised both sensitivity and NPV, at the cost of classifying fewer patients as low risk - the trade-off every rule-out pathway makes.

The authors are careful about where this holds, and the caution is the useful part. These figures come from well-trained emergency medical services with prompt and reliable follow-up for the patients who are not conveyed. In a system where a low-risk patient left at home may not be seen again, the negative predictive value of the score is not the negative predictive value of the pathway. For Indian practice, where prehospital troponin is rare outside a few metropolitan services and follow-up after non-conveyance is not systematised, this reads as a description of what a mature system can achieve rather than a protocol to adopt tomorrow.

• Ask what happens to the low-risk patient after the ambulance leaves before adopting any prehospital rule-out
• Check which troponin assay the point-of-care device uses and how it compares with your laboratory's
• Keep the ECG central - the score does not work without a properly acquired and interpreted one
• Audit 30-day outcomes locally if you implement this; the pooled figures are from six selected systems
• Remember the trade-off: a safer cut-off rules out fewer patients, so the workload saving shrinks as safety rises

Why it matters

It moves the rule-out decision from the emergency department to the ambulance - which is only safe in a system that can catch the patients it sends home.

Don't overread it

These are six selected, well-resourced EMS systems; the pooled performance is not transportable to a service without organised follow-up.

The statistics, in plain English

A negative predictive value of 99.8% for death sounds definitive, but it depends heavily on how uncommon death is in this population - with a low event rate, almost any rule looks good on NPV. Sensitivity is the more honest number, and at 93.2% with a lower confidence bound of 83.5%, the pooled data are compatible with the score missing as many as one in six deaths. The MACE figures are similar. That is why the follow-up arrangement, not the score, decides whether the pathway is safe.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.