Reviews Of Evidence

Question 1

Define systematic review

Answer 1

Summary of medical literature on a topic, conducted using explicit methods

Question 2

Define meta-analysis

Answer 2

Quantitative synthesis of results of 2 or more primary studies that addressed the same hypothesis in the same way

Question 3

What is involved in systematic review?

Answer 3

Research question, stating types of: study; participants; interventions; outcome measures

Finding studies, including published and grey literature

Appraising studies, reviews results of study search

Synthesising results, may include meta-analysis

Question 4

What is the purpose of meta-analysis?

Answer 4

Synthesis of a large number of study results

Systematically collate study results

Reduce problems of interpretation, due to variation in sampling

Quantify effect sizes and their uncertainty as a pooled estimate

Question 5

Describe the quality criteria of meta-analysis

Answer 5

Compilation of complete set of studies

Identification of common variable or category definition

Standardised data extraction

Analysis incorporating for sources of variation

Question 6

Interpret this forest plot

Answer 6

Squares = individual odds ratios

Lines = their 95% CI

Size of square = proportion of weight given to study

Diamond = pooled estimate, centre indicating pooled odds ratio, width represents pooled 95% CI

Vertical solid line = null hypothesis OR

Question 7

What are the features of a systematic review?

Answer 7

Explicit

Transparent

Reproducible

Question 8

What are the features of meta-analysis?

Answer 8

Quantitative synthesis of primary data

Summarises effect sizes and their uncertainty

Displayed as a forest plot

Question 9

List difficulties of meta-analysis

Answer 9

Heterogeneity between studies

Variable quality of the studies

Publication bias in selection of studies

Question 10

Define heterogeneity

Answer 10

Observed effects in studies are more different than we would expect by chance

Question 11

What causes heterogeneity?

Answer 11

Methodological heterogeneity

Clinical heterogeneity

Question 12

Define methodological heterogeneity

Answer 12

Differences in methods (study design)

Question 13

Define clinical heterogeneity

Answer 13

Differences in patients, interventions, outcomes

Question 14

What can we do about heterogeneity?

Answer 14

2 approaches when calculating the pooled estimate (OR and its 95% CI):

1) Fixed effect model

2) Random effects model

Question 15

What is fixed effect model?

Answer 15

Used to calculate pooled estimate (OR and its 95% CI)

Assumes studies are estimating exactly the same true effect size

Protects against heterogeneity

Question 16

What is random effects model?

Answer 16

Used to calculate pooled estimate (OR and its 95% CI)

Incorporates heterogeneity by assuming that studies are estimating similar, but not the same, true effect size

Protects against heterogeneity BUT does not explain it

Question 17

Is fixed effect or random effects model better?

Answer 17

Point estimate is similar for both

BUT: random effects model is better

• Wider 95% CI
• More equal weighting of studies

Question 18

What is sub-group analysis?

Answer 18

Explains heterogeneity

Takes into account: study characteristics; participant profile

Question 19

Variable quality can be due to:

Answer 19

Poor study design

Poor design protocol

Poor protocol implementation

Question 20

Order studies into more to less prone to bias:

Answer 20

Case-control

Cohort

NRCT

RCT

Question 21

How to combat variable quality of studies?

Answer 21

Define basic quality standard, and only include studies satisfying this criteria

Score each study for its quality, and then:

• incorporate quality score with weighting of study
• use sub-group analysis

Question 22

What is publication bias?

Answer 22

Studies with ‘favourable’ results are more likely to be published

Any systematic review/ meta-analysis can be flawed by this

Question 23

How do we identify publication bias?

Answer 23

Check meta-analysis protocol for method of identification of studies

Funnel plot

Use statistical test for publication bias

Question 24

What is a funnel plot?

Answer 24

Plot of measure of study size (Y), against effect (X)

If no publication bias: plot will be symmetrical

Question 25

Which shows evidence of bias?

Answer 25

Right funnel plot