The determination of cut-off score that represents a better trade-off between sensitivity and specificity of a measure is straightforward. However, for multivariate ROC curve analysis, I have noted that most of the researchers have focused on algorithms to determine the overall accuracy of a linear combination of several indicators (variables) in terms of AUC. For instance, one commonly used method for getting a linear combination of multiple variables and to determine the diagnostic accuracy of this combination of variables is to conduct a logistic regression first and then to save the predicted probabilities. And using this saved probability as an indicator one may conduct the unvariate ROC in the usual manner and test whether the resulting AUC for combined variables is significantly better than any of the variables taken alone.
However, these methods do not mention how to decide a combination of cut-off scores associated with the multiple indicators that gives the best diagnostic accuracy.
I would highly appreciate the help of esteemed members working with ROC to suggest the method and some software/ SPSS macro/ Excel spreadsheet or any other tools to accomplish such analysis.
The following is a link to a paper that probably deals with the issue that I have raised but I am not able to comprehend how to accomplish such analysis
http://www.clinchem.org/content/41/8/1248.full.pdf