The acute evaluation of chest pain of potentially ischaemic origin remains a persistent dilemma, motivating an increasingly resource-intensive undertaking driven by cultural ‘zero-miss’ expectations. An evolving management strategy proposes the utilisation of non-invasive CT coronary angiography (CTCA) for ‘low-risk’ Emergency Department (ED) patients, espoused by its proponents to efficiently allow safe and cost-effective disposition. Three high-profile trials, known by their acronyms of ACRIN-PA, CT-STAT and ROMICAT II, have recently presented their findings.1–3 Together, these trials randomised nearly 3500 low-risk chest pain patients evaluated in an ED setting to a CTCA-based strategy or a conventional strategy, typically consisting of inpatient stress myocardial perfusion imaging. Based on the evidence presented, the various authors endorse CTCA as a promising option.

The premise of its use in the acute setting is rapid triage via radiographic demonstration of the absence of coronary artery disease (CAD), thereby conferring an acceptable prognosis for short-term event-free survival. The authors in ACRIN-PA, a prospective, multi-centre, safety evaluation, emphasise findings that none of the patients with a CTCA negative for significant CAD suffered death or myocardial infarction (MI) within 30 days. However, only two patients out of 1357 who were not diagnosed with MI at the index visit subsequently sustained an MI within 30 days regardless of CTCA findings. At issue is whether the CTCA confers any incremental prognostic value over a clinical and biomarker-only risk-stratification strategy. Even with prior generation cardiac troponin assays, two negative troponins within 6 h of chest pain onset conferred a ∼99% event-free short-term prognosis.4 Modern, ‘high-sensitivity’ troponin assays now envisage 2 h or single-troponin strategies for early discharge from the ED.5–7

Efficiency and cost endpoints were evaluated by CT-STAT and ROMICAT II, demonstrating that patients randomised to the CTCA cohort had reductions in ED length-of-stay and decreased costs associated with their ED visit. In CT-STAT, median expenditures are reported as reduced by 38% in the CTCA cohort, but their analysis is limited only to ED costs. Following their ED stay, the CTCA cohort underwent a sixfold greater number of non-invasive imaging studies, but the total cost data were not subsequently tracked. Similarly, in ROMICAT II, ED costs were reported as reduced by 19%, but eventual hospital costs were increased by 50% over the standard care arm. This short-term observation of additional downstream testing following CTCA normalises any cost benefit seen in the ED and is consistent with other studies raising similar resource utilisation concerns. An elderly cohort free of known CAD undergoing outpatient CTCA as their initial non-invasive test incurred total healthcare costs up to three times greater than other non-invasive strategies over a 6-month follow-up period.8 Twice as many CTCA patients in this study underwent revascularisation procedures, yet all-cause mortality was not observed to significantly differ. Therefore, unfortunately, it does not appear routine utilisation of CTCA results in the useful cost savings needed to justify its lack of additional protective prognostic effect.

If the incremental prognostic value of CTCA in the acute setting is uncertain, downstream resource utilisation is not reduced and patients receiving CTCA undergo additional revascularisation procedures of questionable benefit, the role for CTCA remains open for contention.

A further issue entails pragmatic consideration of its use—and potential for abuse—outside the controlled academic trial setting featured in each of these recent trials. A potential cautionary parallel may be drawn against the sharp increase in testing for pulmonary embolism (PE) following widespread availability of CT pulmonary angiogram. In the 8 years following the widespread availability of CT pulmonary angiogram, diagnoses of PE have risen 81% in the USA.9 However, these additional diagnoses had no impact on the overall mortality trend for PE, while complications from anticoagulation increased by 71%. In a quixotic quest for perfection, it is not unreasonable to envision that elimination of barriers to a CTCA-based strategy may result in the scope of testing ballooning to include patients having only an infinitesimal pretest probability for significant CAD. If the routine use of CTCA for low-risk chest pain becomes common, it is likely significant patient harms will be realised secondary to false-positives and detection of ‘pseudodisease.’

The alternative conclusion suggested by this evidence, contrary to the authors’, is that CTCA confounds the acute evaluation of low-risk chest pain. Current National Institute of Health and Clinical Excellence guidelines in the UK include CTCA as an element of the diagnostic strategy, but emphasise that it should not be used to screen for CAD and should be applicable to only a very small number of people.10 It is likely superior in effectiveness to rely on clinical and biomarker risk-stratification instruments for early discharge, potentially improving ED flow while avoiding the deleterious risks of false positives, peri-procedural events, contrast-induced injuries and medical radiation exposure associated with CTCA.11–13 Additional caution should be exercised in interpreting the presented results considering multiple conflict-of-interest disclosures acknowledged by several authors in each of these trials, the most salient of which are associated with the imaging technology manufacturers. While safe evaluation and treatment of patients remains the goal of every physician, further clarification and narrowing of the appropriate practice role for CTCA are needed to prevent unintended patient harms.