Implications of the North American 2021 Chest Pain guidelines in the diagnostic approach to patients with stable chest pain and low pretest probability of obstructive coronary artery disease

Background The North American 2021 Chest Pain Guidelines recommend not testing stable patients with low pretest likelihood of obstructive coronary artery disease (CAD), defined as pretest probability <15% using contemporary models (Class I recommendation). In selected cases among this subset of patients, coronary artery calcium (CAC) score is considered a “reasonable first-line test” (Class IIa). Despite some supporting evidence, the clinical implications of a widespread adoption of these recommendations remain unclear. The purpose of this study was to assess the results of three different testing strategies for patients with pretest probability <15%: A) defer testing; B) perform CAC score and withhold further testing if = 0, and proceed to coronary CT angiography (CCTA) if >0; C) perform CCTA in all. Methods We conducted a two-center cross-sectional study assessing symptomatic patients with suspected CAD who underwent CAC score and CCTA. Patients with known CAD, suspected acute coronary syndrome, or symptoms other than chest pain or dyspnea were excluded. Pretest probability of obstructive CAD was calculated based on age, sex and symptom typicality. Obstructive CAD was defined as any luminal stenosis ≥50% on CCTA. Results A total of 2259 patients were screened, of which 1385 (61.3%) had pretest probability <15% and were included in the analysis (mean age 57±11 years, 79% women). Symptom characteristics were: 48% non-anginal chest pain, 26% atypical angina, 21% dyspnea, and 5% typical chest pain. Overall, the prevalence of obstructive CAD was 10.3% (n=142). In the 786 patients (56.6%) with a CAC score of 0, 8.5% (n=67) had some degree of CAD [1.9% (n=15) obstructive, and 6.6% (n=52) nonobstructive]. Among those with CAC >0 (n=599), 21.2% (n=127) had obstructive CAD. The results that would be reached with each of the 3 diagnostic strategies are presented in Figure 1. The number of patients needed to scan with strategy B (CAC as gatekeeper) vs. A (no testing) to identify one patient with obstructive CAD was 11, whereas the number needed to scan with strategy C (CCTA for all) vs. strategy B was 91. Conclusions Not testing patients with suspected CAD and pretest likelihood <15% would lead to missing obstructive CAD in 1 out of 10 patients. Using CAC as a gatekeeper in this subgroup would decrease the use of CCTA by more than 50%, at the cost of missing obstructive CAD in 1 out of 100 patients. These findings may be used to inform decisions on testing, which will ultimately depend on how much diagnostic uncertainty and missed diagnoses patients and their physicians are willing to accept. Funding Acknowledgement Type of funding sources: None.