An elevation in the level of cardiac troponin (cTn) is the biochemical criterion to define myocardial infarction (MI). High-sensitivity (hs) assays, introduced in clinical routine since 2010, can quantify minute amounts of circulating cTn. Compared to conventional methods, these assays allow for a more precise assessment of patients presenting to the emergency department (ED) with symptoms suggestive of MI. Not surprisingly, this has generated a tremendous interest in new assessment strategies applying cTn results obtained from hs-assays.

Research on hs-cTn has pushed forwards into different directions: the development of algorithms applying very low thresholds, the use of serial samples at early time points including specific delta values, or of accelerated diagnostic pathways (ADP) that integrate cTn results with other indicator of cardiac risk, e.g. patient history, symptom characteristics, vital signs and/or ECG findings. Studies are published in increasing numbers: using ‘chest pain’ and ’assessment’ as entry terms on PubMed yields 350 publications in 2017, corresponding to almost one per day, weekends included. Recognizing that this overwhelms clinicians with information, we have reviewed the literature focusing on the performance metrics of different assessment strategies but also on studies investigating the consequences of their clinical implementation [1].

The use of very low cTn thresholds, e.g. at the level of detection or slightly above, increases sensitivity and is useful to identify non-diseased subjects suitable for early ED discharge (‘rule-out’) by a single blood test only. However, the utility of this strategy depends on the prevalence of MI and even more, on the analytical performance of the applied cTn assay. True high-sensitivity is a must.

Algorithms based on serial sampling and data-driven thresholds for cTn levels and their changes represent an alternative. Such algorithms, recommended by the European Society of Cardiology, group patients into cohorts for rule-out and rule-in, the latter being targeted for hospital admission. While these algorithms have shown clinical promise, sole reliance on cTn results still implies some risk of misclassification of MI. In addition, the proportion of patients left without a management decision (‘observational zone’) may be as high as 32%, and there is no data from clinical implementation studies.

A different strategy is pursued by ADPs that integrate cTn results and clinical risk indicators. ADPs indicate the probability of MI (or of cardiac risk) and provide medical guidance, similar as cTn algorithms. Overall, they tend to identify smaller proportions of patients suitable for rule-out. This depends on the incorporation of other risk estimates apart from cTn results which on the other hand will provide stronger reassurance to clinicians. Example of such tools used in clinical practice are the ADAPT-ADP, EDACS-ADP and HEART pathway. Prospective implementation studies demonstrate that their use tends to lower admission rates at maintained high patient safety.

So, what next? Optimally, a new strategy employing hs-cTn assays for the assessment of patients with suspected MI should provide:

  1. High efficacy, i.e. a low proportion of subjects left without a management decision,

  2. High safety, i.e. a low proportion of diseased subjects in the rule-out cohort,

  3. Evidence from successful real-world clinical implementation studies,

  4. User-friendliness.

Currently, there is no strategy that meets all these criteria but ADPs get close. A proposal for structured patient assessment is presented in our review [1]. Nonetheless, clinicians need to identify the key requirements that best fit with the main purpose of their intended strategy. When used appropriately, this has the potential to improve patient care and hospital resource utilization, thereby leading to considerable cost savings.


1. Eggers KM, Jernberg T, Ljung L, Lindahl B. High-Sensitivity Cardiac Troponin-Based Strategies for the Assessment of Chest Pain Patients-A Review of Validation and Clinical Implementation Studies. Clin Chem 2018. 64:1572-85.