Long-Term Readmission
Thirty-day readmission was 13.1%. At one and five years, 28.3% and 38.7% of patients were readmitted to the hospital for any reason (Figure 2 ). Overall rate of rate of cardiac-related readmission in this study period was 34.2%.
Competing risk regression was used to model risk factors for all-cause readmission. In a univariable analysis, postoperative ECMO (HR 0.37, 95% CI 0.19 to 0.74, P=0.01) was associated with a decreased hazards for readmission, which may reflect the high in-hospital mortality associated with its usage. Implantation of a ventricular assist device was associated with a four-fold risk of readmission (HR 4.31, 95% CI 2.80 to 6.63, P<0.001).
In a multivariable model, none of the bridging strategies were significantly associated with readmission after risk adjustment (Table 6 ). In this model, increasing baseline creatinine level (per 1 mg/dL, HR 1.86, 95% CI 1.03 to 3.36, P=0.04) was associated with increased risk of readmission. However, preoperative dialysis dependency was associated with drastically reduced hazards for readmission (HR 0.07, 95% CI 0.01 to 0.75, P=0.03), likely representing a high operative mortality in this subpopulation.