Acute kidney injury following on-pump or off-pump coronary artery bypass grafting in elderly patients: a retrospective propensity score matching analysis

OBJECTIVES: This single-centre, retrospective propensity score matching (PSM) study designed to study the impact of cardiopulmonary bypass (CPB) on postoperative acute kidney injury (AKI) and the relationship between AKI and long-term outcomes in elderly patients undergoing coronary artery bypass grafting (CABG). METHODS: After PSM, 466 pairs of patients (A group, on-pump; B group, off-pump) who were aged≥70 years undergoing first isolated CABG surgery from January 2012 to December 2016 entered the study. AKI was defined and classified according to the Acute Kidney Injury Network (AKIN) criteria. The incidence and severity of in-hospital AKI were compared. The impacts of AKI on the long-term outcomes including new onset of dialysis and mortality were analyzed. RESULTS: The two PSM groups had similar baseline and procedure except whether the CPB was used or not. In hospital and 30-day mortality was of no difference(χ2 = 0.051, p = 0.821). AKI of any severity occurred in 40.3% of all patients, with stage 1 accounting for most cases. No difference regarding the incidence and severity of AKI could be found: AKIN stage 1: 139 (29.8%) vs 131 (28.1%); AKIN stage 2: 40 (8.6%) vs 35 (7.5%); AKIN stage 3: 18 (3.9%) vs 13 (2.8%), (u = 0.543, p = 0.532). No difference was observed in the in-hospital new onset of dialysis (χ2 = 0.312, P = 0.576). The use of CPB was not found to influence long-term new onset of dialysis (χ2 = 0.14, p = 0.708) and mortality (χ2 = 0.099, p = 0.753). Comparing with non-AKI patients, AKI patients were associated with an increased rates of new onset of dialysis (χ2 = 8.153, p = 0.004) and mortality (χ2 = 6.277, p = 0.012) during the follow-up. Multivariable logistic regression manifested that the HR for long-term new onset of dialysis and mortality in AKI patients vs. non-AKI patients was 1.83 and 1.31 respectively (95%CI 1.12–2.86, p = 0.007; 95%CI 1.17–2.58, p = 0.015). The HR for long-term new onset of dialysis and mortality in on-pump group vs. off-pump group was 1.07 and 1.11 respectively (95%CI 1.03–1.23, p = 0.661; 95%CI 1.09–1.64, p = 0.702). CONCLUSIONS: For elderly CABG patients, AKI was common, but deterioration of dialysis was a seldom incidence. Comparing with on-pump, off-pump did not decrease the rates or severity of AKI, long-term new onset of dialysis or mortality. AKI was associated with an increased long-term new onset of dialysis and decreased long-term survival.