Levels of evidence Flashcards
Number studies in terms of levels of evidence
- Meta-analyses
- Systematic reviews
- Critically Appraised Literature/Evidence Based Practice Guidelines
- RCT
- Non-RCT
- Cohort studies
- Case series or studies
- Individual Case Reports
- Background information, Expert Opinion, Non-EBM Guidelines
1 to 3 = critical appraisal
4 to 5 = experimental studies
6 to 8 = observational studies
What is the best reasearch design for the following question type/domain;
A) Therapy (treatment)
B) Prevention
C) Diagnosis
D) Prognosis (forecast)
E) Etiology (causation)
F) Meaning
A)
- Randomised Controlled Trial (RCT)
B)
- RCT or Prospective Study
C)
- RCT or Cohort Study
D)
- Cohort Study and/or Case-Control Series
E)
- Cohort Study
F)
- Qualitative Study
For systematic reviews;
A) What do they synthesise the results from?
B) What do they address?
C) What to use systematic and explicit methods for?
D) Why may a meta-analysis be included?
A)
Synthesise the results from all available studies in a particular area
> provide a thorough analysis of results, strengths, and weaknesses of the collated studies
B)
Adress a focused, clearly formulated question
C)
- Identify, select and critically appraise relevant research
- Collect and analyse data from included studies
D)
- Summarises and analyses the statistical results of the included studies
- Studies may have the same outcome measure
What are the FIVE steps to a systematic review?
- Framing questions for a review
- Identifying relevant work
- Assessing the quality of studies
- Summarising the evidence
- Interpreting the findings
Provide THREE considerations for a systematic review
- A systematic review can take years to complete –> findings may be superseded by more recent evidence
- Methodological rigor and strength of findings must be appraised by the reader
- A large, well conducted RCT may provide more convincing evidence than a systematic review of smaller RCTs
How do systematic reviews differ from a narrative review
What are meta-analysis? What is it based on?
A quantitative, formal, epidemiological study design used to systematically assess the results of previous research to derive conclusions about that body of research
Typically based on RCTs
What are the most common measures of effect for meta-analysis?
- Risk ratio/relative risk and odds ratio (dichotomous data)
- Standardised mean difference (SMD) –> continuous data
What does it mean when studies are ‘weighted’ for meta-anlaysis? What to do to correct this?
- Reflects the value of the evidence from a particular study
- Usually weighted according to the inverse of their variance –> smaller studies contribute less
use either a fixed-effects or random-effects model
Why must the examination of heterogeneity be done in a meta-analysis?
very important
- clinical diversity = variablility in participants, interventions, outcomes studied
- methodological diversity = variability in study design and risk of bias
- statistical heterogeneity (heterogeneity) = variability in intervention effects. I2 > 75%.
also consideration of bias
What are the FOUR levels of NHMRC levels of evidence?
Level 1 –> evidence obtained from a systematic review of all relevant RCTs
Level IV –> evidence obtained from a case series, either post-test or pre-test.
