Risk modeling for ventricular assist device support in post-cardiotomy shock

Background: post-cardiotomy shock (pcs) has a complex etiology. although treatment with inotrops and intra-aortic balloon pump (iabp) support improves cardiac performance, end-organ injuries are common and lead to prolonged icu stay, extended hospitalization and increased mortality. early consideration of mechanical circulatory support may prevent such complications and improve outcome. methods: between january 1997 and january 2002, 321 patients required iabp and inotropic support for pcs following coronary artery bypass grafting (cabg) at our institution. perioperative variables including age, mixed venous saturation (mvo2), inotropic requirements and lv function were analyzed using multivariate statistical methods. all explanatory variables with a univariate p value <0.10 were entered into a stepwise logistic regression model to predict hospital mortality. odds ratios from significant variables (p < 0.05) in the regression model were used to compose a risk score. results: overall hospital mortality was 16%. the independent risk factors for mortality in this population were: mvo2 < 60% (or = 3.2), milrinone > 0.5 lg/kg/min (or = 3.2), age > 75 (or = 2.7), adrenaline > 0.1 lg/kg/min (or = 1.5). a 15-point risk score was developed based on the regression model. hospital mortality in patients with a score >6 was 46% (n = 13/28), 3–6 was 31% (n = 9/29) and <3 was 11% (n = 29/264). conclusions: a significant proportion of patients with pcs continue to face high mortality despite iabp and inotropic support. advanced age, heavy inotropic dependency and poor oxygen delivery all predicted increased risk for death. further investigation is needed to assess whether early institution of vad support could improve outcome in this high-risk group of patients.