Determinants of outcomes for postcardiotomy VAD placement: an 11-year, two-institution study.

OBJECTIVE

Ventricular assist device (VAD) placement after postcardiotomy failure is associated with a high mortality. We sought to determine prognostic factors in order to help better select patients who may benefit from VAD placement.

METHODS

From January 1992 to January 2003, 63 patients out of approximately 30,000 cardiac surgery patients (0.21%) developed postcardiotomy failure requiring VADs. Planned VAD for bridge to transplant or bridge to recovery were excluded. Multivariate logistic regression analysis, based on pre-VAD placement parameters, was used to determine prognostic factors for in-hospital 30-day mortality.

RESULTS

Overall operative mortality was 73% (46 of 63). Multivariate logistic regression analysis identified median age </=50 (each additional year) (OR = 0.85, 95% CI = 0.77 to 0.95, p = 0.004) and median base deficit >/=0 mEq/L (each additional mEq/L) prior to VAD placement (OR = 0.60, 95% CI = 0.48 to 0.91, p = 0.012) to be independent predictors of improved 30-day survival.

CONCLUSIONS

Postcardiotomy failure is a rare event but is associated with a very poor prognosis despite salvage therapies utilizing VADs. Age </=50 years and base deficit >/=0 (mEq/L) prior to VAD placement are associated with improved 30-day survival.