White Papers
Using SAS to Implement Proportional Odds Ratio Model for Meta-Analysis of Diagnostic Tests
Overview The work discussed in this paper shows how one can use SAS system to implement the Proportional Odds Ratio (POR) model. POR is used to pool results of several studies together in a meta-analysis of diagnostic tests. The POR uses Odds Ratio (OR) as performance measure, and makes no assumptions about the shape of OR0, a baseline function capturing the way OR changes across papers. The POR merely specifies a relationship between the ORs of two tests instead of assuming homogeneity of ORs. This paper briefly introduces the POR model. The focus will be on how to use SAS/STAT on Windows to fit the model for a real example, and estimate the discrimination accuracy of competing tests.
| Publisher | SAS Institute | File Format | |
|---|---|---|---|
| Date Published | March 2005 | ||
| Format | White Papers | ||
| Topics |
|
||


