Statistical analysis
Categorical variables are described as numbers and percentages and
quantitative variables as means ± standard deviations. Groups were
compared using the chi square test, Fisher’s exact test, Student’s
t-test or the Wilcoxon test as appropriate. To determine independent
predictive risk factors for failure of mid-cavity VAD with the dependent
variable being binary, we first performed a bivariate analysis and then
a multivariate analysis using logistic regression23.
Variables with a level of significance less than or equal to 0.20 in the
bivariate analysis were included in the multivariate model, which was
analysed with a stepwise logistic regression. Interaction effects were
sought for all variables included in the model. Model discrimination was
assessed by the C-index, which is identical to the area under the
receiver operating characteristic (ROC) curve. Calibration was assessed
by the Hosmer-Lemeshow goodness-of-fit statistic23.
Internal validation was performed by resampling our study population
(5000 bootstrap samples). To predict the failure of mid-cavity VAD, a
ROC curve was used to determine a threshold of the probability issued
from the logistic multivariate regression.
All statistical analyses were performed with a significance level set at
0.05, with SAS software, version 8.02 (SAS Institute, Inc, Cary,
NC).