Reviwing The Evidence Flashcards
What is a metanaylsis?
A quantitative synthesis of results from two or more primary studies that addressed the same hypothesis.
What is the purpose of a metanaylsis?
Synthesis of a large number of study results, quantify uncertainty and effect sizes to give a pooled estimate, reduce problems of variation due to confounding.
What are the key features of a forest plot?
Indivdual odds ratios ar shown as squares, with the square size proportional to the size of the study, the 95% confidence interval is displayed for each study, the diamond shows the pooled estimate and the width of the diamond shows the 95% confidence interval. Solid line is the odds ratio for the null hypothesis.
What are the two ways you can model for variation between studies in a systematic review?
The fixed effects model and the random effects model.
What is the fixed effects model?
Assumes the studies are measuring the same effect size
What is the random effects model?
Assumes studies are measuring a similar but not the same effect size
What are the key comparison points between the fixed and random effects models?
95% confidence interval is often wider in random effects model, the point estimate is similar if not the same in both, and studies are more equally weighted in the random effects model compared to the fixed effect model, where bigger weighting is given to smaller studies.
What are the two methods of sub group analysis?
By participant profile (more statistically significant) and by study design.
What are some of the criteria for assessing the quality of randomised control trials?
Randomisation, blinding and outcome assessment, approiate statistical analysis, patient attrition less than 10% or intention to treat analysis.
What is publication bias?
Studies with statistically significant and favourable results are more likely tog et published than studies with non statically significant or unfavourable results, especially in the case of smaller studies.
How can funnel plot be used to detect publication bias?
If no publication bias expect studies to have a balanced funnel, with publication bias there will be few small studies indicating a small or negative odds ratio.
What is a systematic review?
On overview of primary study that used expilict and reproducible methods.
