Is reoperation still a risk factor in coronary artery bypass surgery?

Background. Hospital mortality for reoperative coronary artery bypass grafting (CABG) is approaching that of primary CABG. This raises two questions: (1) has experience neutralized the risk of reoperation attributable to its greater difficulty, or (2) has experience neutralized the risk attributable to the higher-risk profile of reoperative patients? Methods. From 1990 to 2003, 21,568 CABG procedures were performed, of which 4,518 (21%) were reoperations: 3,919 first, 552 second, 43 third, 3 fourth, and 1 fifth. Reoperative patients had a higher-risk profile than primary patients, with more vascular disease, left ventricular dysfunction, and coronary artery disease (all p < 0.0001). Logistic regression was used to identify factors associated with hospital death and to develop a propensity score for reoperation, which was used to (1) adjust multivariable analyses of death and (2) compare outcomes in matched patients. Results. Hospital mortality was 4.3% (168 of 3,919) for first reoperation, 5.1% (28 of 552) for second, and 6.4% (3 of 47) for third or more, compared with 1.5% (263 of 17,050) for primary operations. Risk of both primary and reoperative CABG decreased with experience (p > 0.0002); however, reoperative risk fell markedly in the mid-1990s. In both the overall and matched-pairs analyses, reoperation was a risk factor before 1997 (p <= 0.008), but not after (p = 0.2). Reoperation within 1 year of previous CABG increased risk (p < 0.0001). Risk attributable to left ventricular dysfunction decreased with experience (p = 0.05). Conclusions. Hospital mortality for reoperative CABG has been consistently higher than for primary operation, but this difference has narrowed considerably. Patient characteristics, not reoperation itself, now have greater influence.