What do you think is an acceptable miss rate for acute coronary syndrome in patients presenting with chest pain to the ED? If you discharge one hundred chest pain patients, how many are allowed return to ED in the next month with an MI and yet you still be considered a safe clinician? Is it possible to get to zero? Before reading on, write down your personally acceptable miss rate. What would it be for your department? How do you go about achieving that target?
Our department has recently undertaken a quality improvement project for the rapid rule out of acute coronary syndrome (ACS) in low risk chest pain patients. This QI project was borne out of the ADAPT trial* (2-hr Accelerated Diagnostic Protocol to Assess Patients with chest pain symptoms using contemporary Troponin). This project was heralded with fanfare and enthusiasm, ostensibly so that we could risk stratify low risk chest pain patients (LRCPP) even faster than before with the help of ultra-sensitive troponin assays.
The ADAPT trial provides a data set which enables us to look at the efficacy of how two Australasian EDs risk stratify LRCPP patients, with serial troponins forming the cornerstone of the assessment. The paper highlights the safety of using a 0 + 2hr troponins for ruling out ACS compared with 0 + 6hr troponins. The way the sensitivity of troponin assays have been evolving lately, that’s not an unexpected finding. Buried within the data set, however, is something far more interesting (although in my opinion, no less unsurprising): LRCPP with a non-ischamic ECG and a TIMI score of zero have a 30-day MACE rate of 1.26%.
With the use of troponins in this study, the event rate dropped to 0.26%, which again is pretty phenomenal. But a MACE rate of 1.26% without troponins is pretty low anyway. If we are happy to use the PE rule-out criteria (PERC) with its 98% sensitivity, as an emergency medicine community we should have a serious discussion about what we feel is a reasonable ACS miss rate. If we’re happy with a similar miss rate to PE, what does that mean for troponins in low risk chest pain? There are undoubtly patients who, if given the option, would always choose for further risk stratification with diagnostic testing. For me, if I had a normal ECG and resolved chest pain that an emergency physician thought was low risk (i.e. 1.26% 30-day MACE rate), and then that doctor told me I could be further risk stratified an additional one percent if I waited around in the ED for another 3hrs…. I’d head for the door.
Troponins, at least in my practice, have become a bit of a security blanket as part of practising defensive medicine. A substantial proportion of blood tests I order are so I can justify to my colleagues that I have been a thorough/safe/thoughtful doctor, even if I didn’t think they were really warranted. It’s far easier when trying to explain to a senior my rationale for discharging a patient that they are “troponin negative”, instead of give a nuanced account of their history of symptoms and ECG interpretation. The corollary of this paradigm is that the assay has become a surrogate for perceived safety i.e. “the troponin is negative so the patient must not have the disease”. But ignoring a concerning history or ECG in the face of a negative troponin is done so at the patient’s peril.
– History + ECG + TIMI 0 = 1.26% 30-day MACE (ADAPT data set)
– History + ECG + TIMI 0 + 2hr Troponin = 0.26% 30-day MACE (ADAPT data set)
– Negative troponin ≠ negative for ACS or ischaemic heart disease
Points to ponder
– What does your department consider to be a reasonable “miss rate” for ACS?
– What is the baseline rate of MACE in the general population who don’t have a sentinel chest pain episode?
Than M, Cullen L, Aldous S, et al. 2-Hour Accelerated Diagnostic Protocol to Assess Patients With Chest Pain Symptoms Using Contemporary Troponins as the Only Biomarker: The ADAPT Trial. J Am Coll Cardiol. 2012;59(23):2091-2098.