Predictors of in-hospital mortality in children after long-term ventricular assist device insertion

Abstract

Objectives: This study aimed to determine the pre-implantation predictors for in-hospital mortality in children with ventricular assist device (VAD) support.

Background: Candidate selection is of critical importance for improved outcomes in patients supported with VAD. However, risk factors for post-VAD survival in children are still not clearly understood.

Methods: From June 1996 to December 2009, 92 children underwent implantation of a long-term VAD at Germany Heart Institute Berlin. Data on all these patients were retrospectively analyzed, and pre-operative risk factors for in-hospital survival after VAD implantation were identified by multivariate logistic regression.

Results: Of the 92 subjects, the median age at implantation was 7 years (range 12 days to 18 years), and the median support time was 35 days (range 1 to 591 days). The overall survival rate to transplantation or recovery of ventricular function was 63%. Independent predictors of in-hospital mortality in children included congenital etiology (odds ratio [OR]: 11.2; 95% confidence interval [CI]: 2.6 to 47.5), norepinephrine requirement (OR: 6.9; 95% CI: 1.4 to 31), C-reactive protein level >6.3 mg/dl (OR: 4.9; 95% CI: 1.1 to 22.1), and central venous pressure >17 mm Hg (OR: 4.6; 95% CI: 1.1 to 20).

Conclusions: Congenital etiology, pre-operative norepinephrine requirement, higher serum C-reactive protein, and central venous pressure were associated with increased in-hospital mortality in children with VAD support. Optimal candidate selection and timing of VAD insertion may be of great importance for improved outcomes in children with advanced heart failure.