HYMR Systematic Reviews and Meta Analysis Flashcards
You are reviewing a meta-analysis and are reading over the forest plot included in the paper. When interpreting the overall summary effect or pooled estimate within the forest plot, which of the following does the letter “A” represent on the image below?
SEE Q1 image
A. Odds ratio
B. Point estimate
C. Hazard ratio
D. Median effect
B. Point estimate
- It is important that a reader be able to interpret the components of a forest plot provided by a meta-analysis.
- The “diamond” shaped object present at the bottom of a forest plot represents the point estimate (i.e., point “A.” on the image) and the confidence intervals (the right and left sided points)
- The type of data the diamond represents is determined by the papers being analyzed and how the meta-analysis is set up. So only seeing a diamond alone, a reader cannot tell. The reader must put the diamond into the context of the reset of the forest plot published. As a result the other answer choices are wrong or not the best answer.
High-Yield Core Concept: - The diamond present at the bottom of a forest plot represents the point estimate and the confidence intervals for that point estimate.
High-Yield Fast Fact: - If the right or left sides of the diamond cross the line of no effect then the overall pooled estimate for the studies included in the meta-analysis is NOT statistically significant
Which of the following best represents a publication that provides an systematic qualitative overview of studies done on a specific subject in the medical literature?
A. Meta-Analysis
B. Systematic Review
C. Review Article
D. Case Series
B. Systematic Review
- A systematic review is an overview of the literature or research about a specific topic that is done in a manner that is systematic and reproducible. The process by which a set of studies are used is usually evidence-based, but is a “QUALITATIVE summary.
- A meta-analysis is a technique that uses statistical analysis (pooling or aggregation of results) to “QUANTITATIVELY* combine data from several studies to provide an overall estimate of the magnitude of effect (ie., the “Pooled Estimate”).
- A review article is a general summary about a topic and may or may not follow an evidence-based, systematic approach to summarizing the literature.
- A case-series is a type of descriptive publication that summarizes a series of cases and is not the best answer.
Which of the following is most accurate about the two pooled results from a meta-analysis?
See Q3 pictures
Pooled Result #1
VS.
Pooled Result #2
A. Pooled result #1 the larger point estimate
B. Pooled result #2 has the larger confidence interval
C. Pooled result #1 has the larger confidence interval
D. Pooled result #2 is more significant
B. Pooled result #2 has the larger confidence interval
- The only answer choice that you can appropriately answer based on the two pooled estimates is the second answer choice.
- The width of the diamond represents the confidence interval (most commonly the 95% CI) whereas the center point is the point estimate.
- Without seeing the diamonds in the context of the entire forest plot done for the analysis, a reader cannot discern significance (statistically or clinically).
Which of the following is/are correct about a meta-analysis? (Select ALL that apply)
A. Quantitative summary of study data
B. Provides a pooled estimate or summary statistic
C. Qualitative summary of the data only
D. Is not subject risk of heterogeneity in studies being assessed
A. Quantitative summary of study data
B. Provides a pooled estimate or summary statistic
- A meta-analysis uses statistical analysis (pooling or aggregation of results) to “QUANTITATIVELY” combine data from several studies to provide an overall estimate of the magnitude of effect (i.e., the “Pooled Estimate”).
Unfortunately, meta-analysis are subject to the disadvantage of having heterogeneous data that can make it difficult to know the true effect. - A systematic review is a qualitative summary of the data.
High-Yield Core Concept:
* A meta-analysis uses statistical analysis (pooling or aggregation of results) to “QUANTITATIVELY” combine data from several studies to provide an overall estimate of the magnitude of effect (i.e., the “Pooled Estimate”). While a meta-analysis is technically separate from a systematic review, a systematic review is usually done with a meta- analysis
Which of the following are common tests utilized in a meta-analysis to determine the degree of heterogeneity in the studies included within the meta-analysis? (Select ALL that Apply)
A. Cochrane Q
B. I-squared Statistic
C. T-test
D. Mann-Whitney U
Correct Answer: Cochrane Q, I-squared Statistic
- The two most common tests done in a meta-analysis to determine the degree of heterogeneity, include the Chi-Squared Test of Homogeneity (also called the Cochrane Q) and the I-squared statistic.
- The Chi-Squared Test of Homogeneity (or Cochrane Q) is used to test or determine if the individual data between each study is similar enough to combine them. In other words, it helps to determine the probability that the set of measurements you are using are more different to one another than what is considered due to chance.
- The I-Squared statistic is more stable than the Chi-Squared and reflects the percentage of total variation among all of the data that is due to heterogeneity vs. chance. The larger the percentage the greater degree of heterogeneity.
- A t-test is used in an individual study (not a meta-analysis) when comparing two independent groups using ratio or interval (ie., continuous or numerical) data endpoint.
A Mann-Whitney U is used in an individual study (not a meta-analysis) when comparing two independent groups using ordinal or non-parametric data.
It is important that a reader can interpret a forest plot from a meta-analysis. The pooled or overall summary of effect is represented as a diamond in the forest plot. Which of the following best reflects the letter “B” in the image below.
SEE Q6 figure
A. Confidence interval
B. Mean
C. Point estimate
D. Highest and lowest numbers
A. Confidence interval
- It is important that a reader be able to interpret the components of a forest plot provided by a meta-analysis.
- The “diamond” shaped object present at the bottom of a forest plot represents the point estimate (i.e., point “A.” on the image) and the confidence intervals (i.e., point “B” on the image that represents the right and left sided points)
- The type of data the diamond represents is determined by the papers being analyzed and how the meta-analysis is set up. So only seeing a diamond alone, a reader cannot tell. The reader must put the diamond into the context of the reset of the forest plot published. As a result the other answer choices are wrong or not the best answer.
