p.52 抽出テキスト

Circulation. 2021;144:e368–e454. DOI: 10.1161/CIR.0000000000001029 November 30, 2021 e419 CLINICAL STATEMENTS AND GUIDELINES

other diagnostic procedures.1-4 In these models, costeffectiveness was influenced by an improved diagnostic accuracy for stress echocardiography, which led to longer life expectancy.1 In a recent systematic review, the evidence supports that stress echocardiography or stress MPI are cost-effective for those patients at intermediate pretest risk.5 The improved CAD detection for exercise echocardiography resulted in fewer office and ED visits and hospital days, yielding a 20% cost savings when compared with the exercise ECG.6 There were 2204 patients that underwent stress echocardiography in the PROMISE trial, and 3-year mean costs were similar to that of CCTA (CCTA – stress echocardiography mean cost difference: –$363; 95% CI: –$1562 to $818).7 5.3.4. Stress Nuclear MPI Cost-Value Considerations Among intermediate-risk patients, evidence synthesis supports that stress MPI is a cost-effective test option.1 From the SPARC (Study of Myocardial Perfusion and Coronary Anatomy Imaging Roles in Coronary Artery Disease) registry, observed 2-year mortality rate was highest for PET MPI (5.5%) compared with CCTA (0.7%) or SPECT MPI (1.6%), with 2-year cost highest for patients undergoing PET.2 In the PROMISE trial, nearly two-thirds of patients underwent stress MPI, and the mean cost was similar when compared with CCTA.3 Mean costs were also similar in a randomized trial of 457 patients comparing stress MPI with exercise ECG (P>0.05).4 Among higher-likelihood patients in the UK enrolled in the CECaT (Cost Effectiveness of Noninvasive Cardiac Testing) trial, MPI SPECT was the most cost-effective approach.5 5.3.5. Stress CMR Cost-Value Considerations A synthesis of this cost evidence reveals a pattern whereby CMR perfusion and scar imaging is associated with a favorable incremental cost-effectiveness ratio of <$50 000 per quality-adjusted life years saved.1,2 From a single report, a CMR strategy informed by the CE-MARC trial was more cost-effective than stress MPI, largely because of the higher diagnostic accuracy for CMR.2 However, the most cost-effective strategy was that of initial exercise ECG followed by selective stress CMR and invasive angiography; for this tiered testing approach, additional testing was deemed appropriate in the setting of abnormal or inconclusive findings. In a decision model for intermediate pretest risk patients, a strategy of CMR followed by selective ICA had projected reduced costs by ∼25% when compared with direct referral to ICA.2,3 From the Stress CMR Perfusion Imaging in the United States registry,4 patients with negative findings for ischemia and scar had low downstream costs.5 6 EVIDENCE GAPS AND FUTURE

RESEARCH

Chest pain is one of the most common symptoms for which a person seeks medical care, and it should therefore be the target of substantial research investigation.

1.	 For patients with ACS, considerable success has been achieved in reducing door-to-balloon times for STEMI, but little progress has been made in reducing the important delays from symptom onset to presentation. Further research is needed to develop approaches to shorten this interval including studies of other methods of evaluating patients with chest pain using technologies that permit acquisition and transmission of ECGs from home and remote evaluations (eg, telehealth) for those with acute symptoms.1,2 2.	 An important, increasing patient population includes women and men with angina and ACS associated with angiographically normal or nonobstructive coronary arteries.3,4 Prognosis is not benign, pathophysiology has not been clarified, and optimal therapy is unclear in these heterogeneous groups, which are now considered in terms of INOCA5 and MINOCA.6 Adequately identifying patients with INOCA, and completing an evaluation to make such a diagnosis, is necessary but often not done, regardless of whether chest pain is assessed in the ED, inpatient, or outpatient setting. Further investigation to clarify disease mechanisms in these challenging syndromes is needed to provide the basis for therapeutic advances. 3.	 One of the initial challenges in the evaluation of patients with chest pain, either in the emergency or office setting, is symptom classification. Methods to elicit symptoms and clusters of symptoms that provide improved pretest probabilities of symptomatic CAD may be aided with machinelearning algorithms. It is already clear that some common dogma about chest pain descriptions, such as differences between men and women, may not be as prevalent as has been reported7 and may impede care of both sexes if they do not fit preconceived notions of the clinical significance of their symptoms. However, reducing the differences in both sex and racial differences in treatment and outcomes are important future goals of research and clinical care. 4.	 Clinical risk stratification and decision tools will likely continue to grow in popularity because they are incorporated into electronic health records, but it would be useful to test them in large randomized trials to rigorously determine their

Downloaded from http://ahajournals.org by on October 7, 2026