Subgroup analyses

1
Subgroup analyses

• Reza Yousefi Nooraie

2
Subgroup analyses

• To compare effect estimates in different
  subgroups by considering the meta-analysis
  results from each subgroup separately.

3
(No Transcript)
4
Attention!

• Either the effect or the test for heterogeneity
  in one subgroup is statistically significant
  whilst that in other subgroup is not does not
  indicate that the subgroup factor explains
  heterogeneity.

5
Is the effect different in different subgroups?

• Comparing the subgroups with each other

6
Meta-regression
7
Meta-regression

• If studies are divided into subgroups this may be
  viewed as an investigation of how a categorical
  study characteristic is associated with the
  treatment effects in the meta-analysis.

8

• Meta-regression is an extension to subgroup
  analyses
• allows the effect of continuous, as well as
  categorical, characteristics to be investigated,
  and allows the effects of multiple factors to be
  investigated simultaneously

9
Meta-regression

• Multivariate approach
• Use the study characteristics as independent
  variables
• Design, age, population source, quality score etc
  etc
• Use effect size or other outcome as the dependent
  variable
• Identify significant study characteristics
• Unit of observation study
• Can be useful to identify sources of
  heterogeneity, clarify importance of quality
  scores

10

• Meta-regression should generally not be
  considered when there are fewer than 10 trials in
  a meta-analysis.

11
Meta VS MultipleREGRESSION

• Meta-regressions are similar in essence to
  multiple regressions, in which an outcome
  variable is predicted according to the values of
  one or more explanatory variables.

12
Meta VS MultipleREGRESSION

• Differences
• Larger studies have more influence on the
  relationship than smaller studies, since studies
  are weighted by the precision of their respective
  effect estimate.
• It is wise to allow for the residual
  heterogeneity among treatment effects not
  modelled by the explanatory variables.

13

• The regression coefficient
• describe how the outcome variable changes with a
  unit increase in the explanatory variable
• The statistical significance of the regression
  coefficient
• whether there is a linear relationship between
  treatment effect and the explanatory variable.

14
For study subgroups

• The regression coefficients will estimate how the
  treatment effect in each subgroup differs from a
  reference subgroup.
• The P-value of each regression coefficient will
  indicate whether this difference is statistically
  significant.

15
Meta-regression Example(Phillips 1991 26 HIV
studies, Dependent var Specificity)
16
When do a Meta-regression

• Ensure that there are adequate studies to justify
  meta-regressions
• It is very unlikely that an investigation of
  heterogeneity will produce useful findings unless
  there is a substantial number of studies.
• at least ten observations (i.e. ten studies in a
  meta-analysis) should be available for each
  characteristic modelled.

17
When do a Meta-regression

• Specify characteristics in advance
• Reviewers should, whenever possible, pre-specify
  characteristics in the protocol that later will
  be subject to subgroup analyses or
  meta-regression.

18
When do a Meta-regression

• Select a small number of characteristics
• The likelihood of a false positive result among
  subgroup analyses and meta-regression increases
  with the number of characteristics investigated.

19
When do a Meta-regression

• Ensure there is scientific rationale for
  investigating each characteristic
• Selection of characteristics should be motivated
  by biological and clinical hypotheses, ideally
  supported by evidence from sources other than the
  included studies.

20
When do a Meta-regression

• Be aware that the effect of a characteristic may
  not always be identified
• Many characteristics that might have important
  effects on how well an intervention works cannot
  be investigated using meta-regression.
• These are characteristics of participants that
  might vary substantially within studies, but
  which can only be summarised at the level of the
  study.

21
An Example

• Consider a collection of clinical trials
  involving adults ranging from 18 to 60 years old.
  
• There may be a strong relationship between age
  and treatment effect that is apparent within each
  study.
• However, if the mean ages for the trials are
  similar, then no relationship will be apparent by
  looking at trial mean ages and trial-level effect
  estimates.
• The problem is one of aggregating individuals
  results and is variously known as aggregation bias

22
When do a Meta-regression

• Think about whether the characteristic is closely
  related to another characteristic (confounded)
• Two characteristics are confounded if their
  influences on the treatment effect cannot be
  disentangled
• Computing correlations between trial
  characteristics

23
An Example

• If those studies implementing an intensive
  version of a therapy happened to be the studies
  that involved patients with more severe disease

24

• Relationship between treatment effect and a
  single covariate

25

• A scatter plot with the covariate along the
  horizontal axis and the treatment effect along
  the vertical axis

26
(No Transcript)