|
Activity Number:
|
25
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 2, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #303706 |
|
Title:
|
Multiple-Outcome Meta-Analysis Methods: Application to Orthopedics Research
|
|
Author(s):
|
Yan Ma*+ and Madhu Mazumdar
|
|
Companies:
|
Weill Medical College of Cornell University and Weill Medical College of Cornell University
|
|
Address:
|
Hospital for Special Surgey, 400 E 71st Street, New York, NY, 10021,
|
|
Keywords:
|
Meta-analysis ; orthopedics ; multiple correlated outcomes
|
|
Abstract:
|
Meta-analysis (M-A) is broadly accepted as a means for combining many quasi-experiments for the purpose of establishing the weight of scientific evidence for a certain research question.In the field of orthopedics, the number of M-A is quickly growing but advanced methods are underutilized. Typically the clinical questions have multiple outcomes but the analytic methods used are often based on single outcomes. Methods for multiple outcomes available in the statistical literature need to be promoted in this field. We reanalyze one M-A examining the use of Antifibrinolytic agents (AA) in spine surgery Gill et al,2008)using the correlated nature of the two outcomes, estimated total blood loss and blood transfusions. We find that the methods for dealing with more than two outcomes and outcomes of mixed nature are not available. Some new approaches are suggested and applied.
|
