Systematic Reviews

Question 1

What does the diamond on a forest plot represent?

Answer 1

Summary statistic/estimate

Middle= pooled estimate calculated from combined averages across different studies

Sides= width of combined CIs

Question 2

What are the main features of meta-analysis in SR?

Answer 2

Estimate the RR of each study
Enables comparison of the CI of the studies by looking to see if bars overlap
Can assess the weighting of the study (based on study size)

Question 3

What is cumulative meta-analysis? What will happen to the CI as more data is added?

Answer 3

Data is shown at different points in the study and updated with new data

The confidence interval will get tighter with increased data

Question 4

What does the I squared represent in meta-analysis?

Answer 4

The degree of inconsistency between the results across the studies which is not accountable by chance

CANNOT SAY HETEROGENEITY ALONE

Question 5

How can would you interpret different % of I squared?

Answer 5

0%= no heterogeneity i.e. homogeneity= any variation due to chance 
>25%= low heterogeneity 
>50%= moderate heterogeneity 
>75%= high heterogeneity

Question 6

Why is it important to calculate I squared?

Answer 6

Determine the inconsistency across the studies so can determine whether the results can be generalised

Determine confidence of recommendations
I.e. high inconsistency= reduces confidence

Question 7

What causes statistical heterogeneity? Give examples of what this might look like in a study

Answer 7

Clinical difference in studies

• difference in condition definitions
• difference in treatment of conditions
• difference in study demographics

Methodological differences

• trial design
• risk of bias
• methods of identify condition

Unknown study characteristics