• EBM
• Quality of Care and Outcomes

Terkelsen et al imply that the guidelines have distorted the evidence, particularly with respect to acceptable delays in delivering percutaneous coronary intervention (PCI), in order to favour prehospital thrombolysis over primary PCI as the recommended reperfusion strategy for STEMI, and that doing so will deprive patients from lifesaving treatment.

This is a complete misinterpretation and misrepresentation of the guidelines. First, the guidelines clearly favour primary PCI and state “Primary PCI (...) is the preferred reperfusion strategy in patients with STEMI”. Secondly, the guidelines do not suggest thrombolysis over PCI if the PCI delay is 60 min, but in fact state (in several locations) that a “PCI related delay of 120 min is useful in selecting primary PCI over immediate thrombolysis as the preferred mode of reperfusion”. This statement clearly reflects the choice of the committee to favour primary PCI for reperfusion and the fact that, contrary to the statement of Terkelsen et al, the guidelines do not “reduce the window of opportunity for primary PCI”. However, the guidelines also emphasise that the decision to choose primary PCI over thrombolysis should not lead to complacency with respect to delays; this is because the evidence is very consistent that shorter delays to reperfusion translate into improved outcomes and there is extensive evidence demonstrating the efficacy of the very early administration of fibrinolytic therapy on infarct size, left ventricular function, and mortality. Therefore, the document emphasises quality targets that are distinct from the PCI related acceptable delay and recommends a maximum delay of 90 min for transfer to primary PCI as a goal for quality improvement, and an even shorter delay (60 min) if the patient presents early and with a large infarct or if the patient presents directly to an interventional centre. The guideline document clearly states, both in tables and in the text, the distinction between acceptable PCI related delays (for choosing primary PCI over thrombolysis) and target delays for primary PCI in the various settings. We strongly disagree with the assertion that the two terms are used interchangeably.

Terkelsen et al are obviously concerned that, because of the fear of quality indicators focusing on the proportion of patients within the guideline recommended delays, physicians may elect to use thrombolysis as it may be easier to achieve reperfusion within the recommended period. There is no question that shorter delays translate into better outcomes and this is reflected in the guidelines which recommend monitoring and reporting of actual delays, in an effort to improve time to reperfusion, not as a tool to skew the choice of reperfusion strategies towards lysis.

We strongly disagree with the assertion that the guideline document erroneously equates PCI related delay and time from first medical contact to primary PCI; in reality the document says (page 13): “The ‘PCI related delay’ is the theoretical difference between the time of FMC [first medical contact] to balloon inflation, minus the time from FMC to start of fibrinolytic therapy (ie, ‘door-to-balloon’ minus ‘door-to-needle’). The extent to which the PCI related delay diminishes the advantages of PCI over fibrinolysis has been the subject of extensive debates. Because no specifically designed study has addressed this issue, caution is needed when interpreting the results of these post-hoc analyses.”

Terkelsen et al suggest that CAPTIM (Comparison of primary angioplasty and pre-hospital fibrinolysis in acute myocardial infarction) evidence, published by the first author of the guidelines,3 was weighted too heavily and used to make recommendations regarding maximal delays for primary PCI. First, the guidelines state that “In settings where primary PCI cannot be performed within 120 min of FMC by an experienced team, fibrinolysis should be considered, particularly if it can be given pre-hospital within the first 120 min of symptom onset”, and the CAPTIM trial was only referenced to support the latter part of this statement. The discussion on delays subsequently focuses at length on the two observational analyses by Pinto et al 4 ,5 which Terkelsen et al also discuss.

We also wish to point out that the guidelines were not written by a single author, but by a wide consensus from a large group of 23 authors and that the process included several rounds of thorough review by 19 experienced external reviewers. In regard to the delivery of reperfusion therapy, ‘one size does not fit all’ and there are parts of the world where distance, climate, and the availability of facilities will result in significant delays in the delivery of primary PCI—the guidelines recognise this by defining a role for the pharmacoinvasive strategy in specific settings. It is not clear to us why Terkelsen et al take exception to such an eminently reasonable approach.

We also feel that the authors’ interpretation of the results of Strategic Reperfusion Early after Myocardial Infarction (STREAM) (with tenecteplase and antithrombotic treatment) early after myocardial infarction) is simplistic and does not take into account the results obtained after a protocol amendment reduced the dose of tenecteplase in the elderly.

The guidelines represent an honest attempt by the task force to summarise the existing evidence into recommendations for clinical practice; they are always open to improvement, particularly as new evidence accrues. While we welcome open scientific debate, we cannot condone misrepresentation by Terkelsen et al for two reasons: first, the evidence they provide in regard to relating delays and outcomes in STEMI is observational and inherently weak; secondly, to state that the guidelines are ‘threatening’ the delivery of optimal care when the guidelines are pushing strongly for shorter delays in primary PCI is at least paradoxical if not offensive.