9: systematic review and meta-analysis Flashcards
meta-analysis
a type of systematic review that uses statistical techniques to quantitively combine and summarize results of previous research
A review of literature
is a meta-analytic review only if it includes quantitative estimation of the magnitude of the effect and its uncertainty
What is a meta-analysi?
Quantitative approach for systematically combining results of previous research to arrive at conclusions about the body of research.
- Quantitative: numbers
- Systematic: methodological
- Combining: putting together
- Previous research: what is already done
- Conclusion: new knowledge
Rationale for systematic reviews and meta analysis
- MA is an efficient scientific technique usually quicker and less costly than a new study.
- Consistency of relationships across studies can be evaluated
- MA can help to explain data inconsistencies and conflicts in data
- MA increases the statistical power (sample size)
- formulating the research question
a good meta analysis should begin with clearly formulated specific research questions (hypothesis) that are important and testable.
- Obtaining representative studies for review
- Clear inclusion (populations, interventions, outcomes) and exclusion criteria
- Multiple search strategies: manual searches of journals, examining references of each obtained, computer searchers of different databases, searching for unpublished studies
- Coding studies for important information
The goal is to code all study features that might influence outcomes.
Forest plot
The graphical display of results from individual studies on a common scale
Publication bias
Statistically significant effects are more likely to be published.
- what to do? –> search for and include unpublished studies
Funnel plot
- Scatterplot of effect estimates against sample size
- Used to detect publication bias
- If no bias, expect symmetric, inverted funnel. –> if biased, expect asymmetric or skewed shape.
Two basic approaches to quantitative meta-analysis
- Weighted-sum: depending on homogeneity testing
- fixed effect model
- Random effect model
- Cumulative meta-analysis - Meta-regression model
- meta-regression techniques
- Wieghted linear regression
Fixed effect model
All of the observed differences between the studies is due to chance
Basic assumption that there is one true value of the effect.
(weighted sum)
Random effect model
Assumes a different underlying effect for a study.
This model leads to relatively more weight being given to smaller studies and to wider confidence intervals than the fixed effect models
This model has been advocated if there is heterogeneity between study results.
(weighted sum)
statistical heterogeneity
If various results differ too much pooling the results is likely to be misleading since the studies might actually be measuring different effects.
Heterogeneity across studies means that the estimates from individual studies have different magnitude or even different directions.
Meta-regression
It can be either a linear or logistic regression model. In most meta-regression approaches, the unit of analysis, that is each observation in the regression, is a study.
Predictors in the regression at the study level and might include such factors as the treatment protocol, characteristics of the study population such as average age, or other variables describing the study setting.
