Subsequently, variables significant in univariate models (p < 0.10) were inserted in a multivariate model, applying a backwards elimination technique retaining variables with a p-value < 0.10.
In a Cox regression multivariate model applied to ICU survivors, only age was an independent predictor for mortality both at 1 year (RR = 1.04, 95 % CI: 1.01-1.07) and 2 years (RR = 1.02, 95 % CI: 1.002-1.045) (Table 3).
Multivariate models were applied to adjust for the effect of potential confounders.
Univariate logistic regression models, supported by multivariate models, were applied to identify groups of people who had adverse interviewing experiences.
Multivariate models were applied in only two studies, one of them showing overall inconsistent findings with respect to our review question [ 41].
Two multivariate models (ANCOVA), applied separately to each SF-36 scale allow tiered analysis of the effect of diabetes status on quality-of-life measures.
Furthermore, 2 distinct multivariate models were applied to consider colinearity between feed type and feed factories in the foodborne risk study.
Multivariate models were applied to investigate the association of sex, age, education and modality of Euroclot recruitment (twins enrolled in the ITR and volunteers) with the targeted dimensions.
Three multivariate models were applied to evaluate the contribution from standard clinical parameters alone (Model 1), from standard parameters plus genomic proliferation (Model 2), and from standard parameters plus proliferation and the intrinsic subtype (Model 3).
Two multivariate models were applied to evaluate the association of CD44 and CD24 expression with gastric cancer recurrence: Model 1 considered CD44 and CD 24 as two variables, while Model 2 considered the combined CD44/CD24 expression as one variable.
