Page 44 - PCC13

Meta‐analysis
• Same patients, same treatment
• Unit of analysis = unit of randomisation
• uses forest plot to summarize studies − fixed effect versus random effect model
• clearer picture
• precision of treatment effect
Averaging studies
A simple average gives each study equal weight
BUT
Some studies may give an answer closer to the ‘true’ effect than others
44
Page 38 of 84