Systematic Review and Meta-analysis
why can general review be subject to bias
because authors can select which information to include
The horizontal line for each study represents ___ ____. The size of the box represents ___ ___ of the individual trial. The bottom diamond represents the ____ ____.
confidence interval; relative weight; overall results
Fail-Safe N
examines impact of publication bias on results by estimating the number of negative studies that would have to be added in order for the results to no longer be statistically significant
Review abstracts
if relevant to research, pull articles
>50% I2
large degree of heterogeneity
heterogeneity
measure of how different studies are from one another
Meta analysis is a ____ synthesis of data
quantatative
Review full-text articles
read and choose literature to help support study
Characteristics of Systematic reviews
reproducible includes all studies that meet criteria Can include a selection flow chart
When too much heterogeneity exists, we need to do
separate analysis
define and conduct literature searches
should be Specific and look everywhere
<25% I2
small degree of heterogeneity
A systematic review includes a ____ (or systematic) process for identifying all relevant literature to answer a question
structured
To interpret out meta analysis result we can use
subgroup analysis or sensitivity analysis
Fixed effect model
used when all the studies (and their data) are pretty similar and difference is due to chance
Define focused key questions
using PICO
When is it useful to preform a systematic review
when multiple strong studies are available but answers are not in perfect agreement
When should meta analysis be used?
when various study results are only slightly different
4 Goals of Meta-analysis
1. Establish statistical significance when studies are conflicting 2. establish best possible estimate of the magnitude of effect 3. evaluate harms or safety when small numbers of side effects are noted in RTC 4. examine subgroups when # in individual studies are not large enough
8 steps of the systematic process
Define focused key questions define eligibility criteria and outcomes define and conduct literature searches review abstracts review full-text articles hand search reference lists and gray literature data abstraction and quality rating data synthesis and answering key questions
I2 statistic
Provides an estimate of the percentage of variability in results across studies that is likely due to true differences in treatment effect, as opposed to chance
General review
a broad article written to summarize a specific topic in a descriptive manner (tertiary literature)
Meta-analysis
a statistical pooling of data from multiple studies that each contribute data
Random-effects Model
assumes differences in original study population resulted in different outcomes-> harder to prove statistical significance
Data abstraction and quality rating
determines if data is bias free/valuable
Define eligibility criteria and outcomes
eligibility= inclusion/exclusion criteria primary endpoints/result of interest
Hypothesis-driven meta-analysis
focus on a specific issue with limited number of studies meeting criteria
Meta-analysis results are presented in
forrest plot
Funnel plot with all data on one side means
gap in result type in analysis
Funnel plots
graphic representation to visually asses publication bias in a meta analysis
25-50% I2
moderate heterogeneity
Exploratory meta-analysis
more studies than can have differing characteristics that can be used as advantagous
Why is meta-analysis considered primary literature
provides a new estimate of effect
Meta-analysis have risk of ____ bias, due to the tendency to publish studies with ___ results and not with ___ or small results
publication; positive; negative
hand search references and gray literature
published references and unpublished studies(ex; poster at a fair)
Data synthesis and answering key questions
put all data together
A systematic review is a ____ synthesis and is considered ____ literature
qualitative; tertiary
