Methods:

Patients ≥45 years of age undergoing major inpatient NCS between 2016–2020 at one of four academic medical centers were retrospectively identified, and all who had undergone non-ECG-gated chest CT within a 12-month period prior to surgery were included. Exclusion criteria included previous coronary revascularization or valve prosthesis. From these non-ECG-gated CT scans, severity of calcification in each of three major epicardial coronary arteries (left main-left anterior descending, left circumflex, and right coronary artery) was determined based on the length of calcium visualized, and in each vessel an ordinal score was assigned (0 = no CAC seen; 1 = mild, 1-24% vessel involved; 2 = moderate, 25-29% of vessel involved; and 4 = severe, with ≥50% of vessel involved). Estimated coronary calcification burden (ECCB) was calculated by summation of the ordinal score from all three vessels, and clinically relevant thresholds were chosen for further analysis. Revised Cardiac Risk Index (RCRI) score was calculated for each subject. The primary outcome was all-cause mortality or myocardial infarction within 30 days of surgery. The ability of RCRI versus ECCB to discriminate risk was determined by receiver operating characteristic (ROC) curve model comparison. To assess pragmatic feasibility, a sample of 100 studies (50 contrast and 50 noncontrast) was reviewed by two nonradiologist physicians to estimate inter-reader reliability.