forest plots, funnel plots and heterogeneity Flashcards
forest plots use
odd ratios
size of the square
represents the weight given to each study- the larger the square the bigger the weight
pooled effect size is denoted by
a diamond
vertical line- 1
no effect
if the CI line cross 1
the study didn’t show significance
p value
the probability that if this was repeated you would ge the same result
funnel plots represent
publication bias
larger studies have
higher/ lower odds ratios - smaller standard error
negative studies are
much less likely to be published
due to negative studies being less likely to be published
systematic reviews ill have more positive than negative outcome trials
forest plots show that
studies with equal standard error are subjected to publication bias based on whether they yield positive or negative results
meta-analyses are subsets of
systematic reviews
systematic reviews are
subsets of all reviews- narrative reviews
heterogeneity is
the difference in studies not due to chance
hereogenity reduces their ability to be used in
meta-analyses
clinical heterogeneity
differences in:
- patients/study setting
- study design / quality
- interventions
- outcomes (how they are measured)
statistical heterogeneity
individual trials have results not consistent with each other
- benefits vs harms
- size of benefit and harm
- evaluated statistically
I2
the percentage of variation across studies that is due to heterogeneity and not due to chance
0.25
low heterogeneity
0.5
moderate
0.75
high
when looking at forest plots, evidence of heterogeneity
- if the plots are on different sides of 1
- wide CIs
95% CI’s
we can be 95% certain that the true value lies within this range of values