Influence of previous coronary artery bypass grafting in the difficulty of acute coronary syndrome diagnosis

OBJECTIVES To investigate whether the diagnosis of acute coronary syndrome (ACS) in coronary artery bypass grafting (CABG) patients with chest pain (CP) is more difficult based on the initial clinical and electrocardiogram (ECG) findings. METHODS We included consecutive CP patients attended at a single emergency department (ED) during a 10-year period. CABG status and the final diagnosis of ACS were considered as stated in the ED discharge report. We evaluated the frequency of 21 CP characteristics (CPC) and four ECG signatures, their individual and collective association with ACS, and ED length of stay (LOS) in CABG and non-CABG patients. RESULTS We included 34 429 patients [median age: 61 years; female: 41.8%; CABG: 2204 patients (6.4%)], and ACS was diagnosed in 6727 (19.5%; CABG/non-CABG 37.2%/18.3%; P < 0.001). CABG patients more frequently had CPC and ECG findings typically associated with ACS, but their final association with ACS was weaker than in non-CABG patients (only significant after adjustment for attendant diaphoresis, throat irradiation, ST-segment elevation and T-wave inversion). The collective discriminative capacity was significantly lower in CABG patients (area under the curve 0.710 vs. 0.793; P < 0.001), even after adjustment (0.708 vs. 0.790; P < 0.001). ED LOS was longer for CABG patients, overall (P < 0.001) and for patients diagnosed with ACS (P = 0.008) and non-ACS (P < 0.001), but these differences disappeared after adjustment. CONCLUSION CABG substantially reduces the diagnostic performance of CPC and ECG findings to suggest ACS. A longer LOS in the ED in CABG patients is more related to their baseline characteristics than to CABG itself.