Levels of Evidence Flashcards
Characteristics of a systematic review and meta-analysis
Analysis of multiple studies (usually RCTs)
Appraise the
- quality of studies
- ways bias and confounders have been minimised
Often use statistical techniques
Characteristics of an RCT
Subjects are randomly assigned to a treatment or control group
Random allocation and blinding reduce bias
Characteristics of cohort studies
Prospective study
Individuals with exposure of interest compared over time with individuals without the exposure
Characteristics of case control studies
Retrospective study
Individuals with a condition of interest are retrospectively compared to a group of people without the condition
Aim to determine exposures
Characteristics of animal and cell studies
Pre-clinical research used to predict the effects of an intervention in humans
What questions would you need to ask to understand the content of a paper
What is the
- research question
- hypothesis
- population
- intervention
- comparator
- outcome
Are the methods appropriate
Have sources of bias been addressed
What are the confounders
Are the statistical analyses correct?
Does the data justify the conclusion?
Are the conclusions inline with the research question and methodology?
What are the questions you want to ask when assessing an RCT
Research question?
Is it justified?
Blinding?
Random allocation?
-was everyone accounted for in final analysis
Experimental conditions?
What outcomes were measured
Limitations?
What are the questions you want to ask when assessing an animal/cell study
Hypothesis?
Is it justified?
Experimental conditions
-how were they selected and how do they compare to human conditions
No of subjects/repeats per group
What was measured
How appropriate is the animal model or cell type used?
Limitations?
Good practice for writing
-where should you put the hypothesis/research question
Stated in the introduction/background
A summary of the study design may be found in the last paragraph of the introduction
Good practice of writing
-where should you put the primary outcomes
Usually stated in the abstract and methods