critical appraisal of research Flashcards
do you know the Hierarchy of Evidence in research?
I Systematic reviews and meta-analyses
II Double-blind RCTs with conclusive results
III RCTs with non-definitive results
IV Cohort studies
V Case-control studies
VI Cross-sectional studies
VII Case reports
what are features of Cross-sectional studies
To determine prevalence
Relatively quick, easy and cheap
Can study multiple outcomes
Do not differentiate between cause and effect
what are features of Case-control studies ?
People with an outcome e.g. an event or disease are matched with controls
Retrospectively determine the importance of a predictor variable in relation to the outcome
Odds ratios (OR) are calculated
Limited by bias (sampling, recall) and confounders
Useful for generating hypotheses – cannot establish cause and effect
Quick, easy, cheap
what are features of Cohort studies ?
People with a risk factor are followed prospectively along with controls
To analyse the impact of risk factors on disease
Establishment of sequence of events (cause versus effect)
Can establish incidence and natural progression of disease
Relative risk (RR) reported
Confounders can be a problem
Expensive, long durations, loss to follow-up
features of a focused question
PICO format
Patient (demographic and clinical details)
Intervention (one or more alternatives)
Control (placebo, do nothing, best current treatment)
Outcomes (indicators or final objectives)
What are the features of an RCT
A RCT is best for studying the effectiveness of an intervention
RCTs cannot help when looking at screening or diagnostic tools, quality of care
Just because it’s a RCT doesn’t mean it’s good or appropriate evidence
What are 5 disadvantages of an RCT
Expensive and resource intensive
Funding from interested parties
Surrogate end-points
Bias
Inclusion/exclusion criteria may not make the results generalisable to the general population
What are the features of an systematic review ?
PICO question
Protocol with a pre-defined eligibility criteria and outcome measures
Comprehensive search strategy
Independent screening of titles and abstracts by at least two authors
Inclusion of eligible studies
Description of eligible studies
Systematic assessment of bias/quality
Checking for heterogeneity i.e. variability in study design and results
Presentation of results with pooling by meta-analysis (Forest plot) if studies are homogeneous
Discussion with implications for practice and research
