Background: Although the use of clopidogrel in patients "unlikely" to require coronary artery bypass grafting (CABG) is recommended in current guidelines of acute coronary syndrome (ACS) management, an important minority of patients require CABG. We assessed the ability to predict need for CABG from demographics known at the time of ACS presentation, using data from SYNERGY.
Methods: Patients undergoing CABG at any time after the index angiogram were included. Early CABG was defined as surgery <72 hours after angiography. The relationship between cessation of enoxaparin and glycoprotein IIb/IIIa inhibition, CABG timing, and 30-day death or MI and bleeding events was assessed. Demographic and clinical factors and geographic location were assessed as predictors of early CABG or CABG at any time. The discriminatory utility is reported with the c-index.
Results: Of the 9053 patients undergoing angiography, 1793 (18.1%) received CABG. Early CABG (n = 972) was associated with more bleeding events (39.2% vs 29.4%, P < .001) but not death or MI. The risk of bleeding events diminished when surgery was delayed >18 hours after cessation of enoxaparin and glycoprotein IIb/IIIa inhibition. Clinical factors associated with early CABG included diabetes and lack of prior CABG or clopidogrel. However, overall the logistic regression model had poor discriminatory ability to predict patients likely to require CABG in the setting of an ACS presentation (c-index 0.671).
Conclusions: It is difficult to predict those high-risk patients with ACS who will undergo surgical revascularization based on baseline clinical characteristics.