Forest Plots
Publication Bias
PICO
Article
Article Part Dos
100
Forest plots can be used in what type of research design?
Meta-analysis
100
How is publication bias assessed?
Funnel plot
100
What does the "O" in PICO represent?
Outcome
100
What was the objective of this study?
To evaluate the literature on the effectiveness of aquatic exercise in relieving pain in adults with neurologic or MSK disease
100
What type of studies did the authors include in their review of the literature?
RCTs
200
What can you infer from a tight horizontal line on a forest plot?
Tight confidence interval: larger sample size, smaller standard error and more precise
200
What is publication bias?
When studies get omitted from publication because they do not have a large effect. The studies are being published because they have a large effect, therefore we do not see that some research had a null or small effect because they were not published.
200
What was the "P" in this article?
Adults with MSK and neurological pain
200
How many studies were included in the meta-analysis?
Five
200
How many articles were accepted for review? and what were the 2 reasons they were excluded from the meta-analysis?
19 High risk of bias; inappropriate raw data
300
Regarding the forest plot in figure 2, what does P=.65 and P=.91 tell us?
There is a significant amount of homogeneity among the populations
300
With a funnel plot, how can you tell that publication exists? In the absence of bias, what should a funnel plot look like?
the plot is asymmetrical the small studies scatter more widely at the bottom and aren’t shown if there is a publication bias. Therefore, resulting in asymmetry within the plot. symmetrical inverted funnel
300
What was the "I" in this study?
Aquatic therapy
300
To be used in a meta-analysis, what is assumed?
Homogeneity
300
What was the overall effect for aquatic exercise versus no-treatment control?
Small significant difference favoring aquatic therapy over the no-treatment control
400
What do the weights in figure 2 mean?
They are based on the sample size and quality of the methods. The higher the weight is, the larger the sample size.
400
A gap in the plot results in asymmetry which means that what studies are missing and where do these studies usually scatter?
Small studies More widely at the bottom of the graph
400
What was the "C" in this study?
Land and no treatment
400
How did the authors assess this quality of the studies and what are some examples of questions are included in this?
SIGN Criteria See table 2
400
What is the P value for comparing land-based exercise with aquatic exercise? What does this P value tell us? How can we interpret this finding by simply looking at the graph?
P=.56; Which means that there is no significant difference between the means. The point estimates are overlapping the null indicating that there is no difference. Although, they are somewhat to the right of the line, indicating that the difference might slightly favor land-based therapy. But this is not significant.
500
What is I^2 a test for and what does it mean that it is 0% in this figure?
It is a test for homogeneity If I^2 is less than 50%, then our samples are homogeneous Being 0% confirms that the samples are significantly homogeneous
500
In the powerpoint, Dr. Goode mentioned a way that we can now find articles that were not published due to a null or small effect, what is this means?
RCT registry
500
How was the "O" in this study measured?
Pain was measured on a continuous scale Some examples: VAS being the most common, McGill pain questionnaire, arthritis impact measurement scale, health assessment questionnaire, health-related quality of life, western ontario and mcmaster universities osteoarthritis index (WOMAC) and 36-Item short form health survey.
500
What does SMD mean in Figure 2?
Standardized mean difference, because the studies did not all use the same outcomes, the authors had to standardize the numbers in order to compare them. If the outcomes were the same, they would just use a mean difference rather than a SMD.
500
What does the symbol size in the forest plot in figure 2 tell us?
It tells us the weight of the given study. (Sample size)
M
e
n
u