CON-J NEWS


Grand Round Meta-Analysis: A Statistical Perspective (Feb. 23, 2017)


“Meta-analysis” is a method for taking the results reported in a group of studies and averaging them in a certain way to come up with a single summary result. It takes the review a step further by using statistical procedures to combine the results from different studies and measure each of the following: Relative Risk; Odds Ratio; Mean; Standard Deviation; Correlation Coefficient.

Two software packages were recommended for statistical analysis:

  • Review Manager (RevMan) from the Cochrane Collaboration
  • Comprehensive Meta-Analysis (CMA)

The presenters demonstrated how to analyze and read each of the above mentioned values using each of the software packages, and how to present and interpret the forest plot.

It was noted that the value of heterogeneity between studies may determine which model to use:

  • Fixed model will be used when I² ≤ 25- 40%. It involves an assumption that the effects being estimated in the different studies are identical.
  • Random model will be used if the heterogeneity is high I² ≥ 75%. It involves an assumption that the effects being estimated in the different studies are not identical, but follow some distribution.