Third, multivariate regression with a forward selection procedure [ 15] was used to identify independent composite predictors of presence of injuries on chest CT.
> -wrap-foot> Note: a ED, emergency department; b PE, physical examination; c CR, conventional radiography Second, univariate logistic regression analysis was used to study the ability of each composite predictor to distinguish patients with injuries from patients without injuries on chest CT.
The adjusted odds ratios modeled by composite predictor variables of preference, measurement and training after adjustment for Place of living, Age and/or Sex are presented in Table 5.
This method searches all possible combinations of factors joined by and/or operators and identifies the best composite predictor L that is a single dichotomous factor.
The composite predictor VSL1 was developed to improve the prediction performance on short disordered regions (<30 residues); the VSL1 implementation showed the best prediction performance in an independent assessment of 20 order/disorder predictors [58].
Definitions of positive composite predictors are displayed in Table 2 Figure 2 shows the ROC curve of the predictive model containing these nine predictors with an R-square of 0.478 and an AUC of 0.85 (95% CI: 0.83 0.87).
Crude odds ratios (OR) of all positive composite predictors were derived for the dependent variable "presence of injuries on CT (yes, no)".
Table 2 shows the definitions of positive composite predictors.
> -wrap-foot>> -wrap-foot> Notes: OR, crude odds ratio; 95% CI, 95% confidence interval Definitions of positive composite predictors are displayed in Table 2 Note: OR, odds ratio adjusted to all other predictors in the model; CI, confidence interval.
