Objective:
To create a risk model to assess outcomes of aortic valve replacement after cardiovascular surgery (ReAVR) using a national Japanese database.
Methods:
The Japan Adult Cardiovascular Surgery Database is a web-based data system involving more than 500 hospitals. Between 2007 and 2012, 2,227 patients who underwent ReAVR for aortic stenosis were retrospectively analyzed. Patients with a previous history of AVR were also included.
Results:
The background of prior surgery (including overlapping cases) was: CABG, 30.9%; valve, 65.4%; and thoracic aorta, 14.7%. The mean age was 70.4. Types of prosthesis used were: bioprosthesis, 51.3%; and mechanical valve, 48.7%. The rate of isolated ReAVR was 59.5%. Concomitant procedures were: CABG, 14.6%; mitral valve surgery, 30.4%; and aortic surgery, 5.7%. The overall hospital mortality rate was 8.7%. Major complications (reoperation, prolonged ventilation, mediastinitis, stroke, and newly required dialysis) occurred in 26.0%. The incidence of stroke was 3.7%, and that of AV block was 3.5%. ORs for hospital mortality were as follows: age, 1.4 (reference ≦60, 5-year increments); male, 1.4; urgency, 1.7; EF ≦30%, 2.0; NYHA classification IV, 1.7; MR ≧2, 1.5; creatinine >2.0 mg/dL, 2.7; liver dysfunction (cirrhosis, AST or ALT >100 U/L, or bilirubin >1.5 mg/dL), 2.3; peripheral artery disease, 1.6; and recent stroke (<2 weeks), 4.8. The type of previous surgery was not a predictor for mortality. In addition to the above-mentioned risk factors, ORs for the composite outcome (mortality and major complications) were as follows: carotid stenosis, 2.5; respiratory dysfunction (FEV1.0 <75%, or bronchodilator use), 1.8; prior valve surgery, 1.3; preoperative inotropic agents, 2.0; endocarditis, 1.7; concomitant CABG, 1.4; and concomitant valve surgery, 1.5. As a performance metric model, C-indexes of hospital mortality and the composite outcome were 0.77 and 0.71, respectively.
Conclusions:
Based on a national Japanese database, early outcomes after ReAVR were satisfactory, despite these operations being associated with a higher risk than the primary operations. The type of previous surgery was not a risk for mortality after ReAVR.