19 - Critical Appraisals Flashcards
What is a critical appraisal?
The use of explicit, transparent methods to assess the data in published research to judge it’s trust worthiness and its value and relevance in a particular context
How do you develop your question in a critical appraisal?
How do we find evidence to use in our critical appraisals?
- Use things like Pubmed, Cochrane Library, NHS evidence
- Is article relevant to my interests?
- Is there a clear research question?
- When was it published?
- Has it been peer reviewed?
What are some reasons that articles get rejected from being included in a critical appraisal?
What is the process involved in a critical appraisal?
How can you quickly narrow down whether you want to include a paper in your critical appraisal?
- Research question and aim at end of background/introduction and you should be able to answer PICO with this
- Is the primary outcome easy to identify?
- Is the study population easy to generalise or applicale to your population of interest?
Once you have figured out if a paper answers your PICO, what are some other factors you need to consider when deciding whether to include the paper in your critical appraisal?
- Selection bias: is how the study sample was develped described? is there a difference between people who have been selected and who hasn’t?
- Data collection: is the data self reported? does the data rely on recall? were validated (e.g piloted questionnaires) tools used? what was the response rate and what do we know about non-responders? how many dropped out?
- Confounding: are confounders discussed and adjusted for?
What are we most concerned about when appraising a cross sectional study?
Cannot see temporal sequence with these studies as outcome and exposure measured at same time
What are some issues when we appraise ecological studies?
What are some critical appraisal tools we can use?
- CASP programme
- Newcastle-Ottowa Scale
What is the outcome measure of a case control study?
Odds ratio
What are some issues we are concerned about when we appraise a case control study?
- Need a clear case and control definition
- Have they used medical records or self report to record exposure
- All the usual e.g is confounding discussed?
- Do they justify the number of cases and controls used?
What is the outcome measure of a cohort study?
- Allows for temporal sequence
- Odds ratio
- Risk ratio
- Incidence rate ratio
What are some issues we are concerned about when we appraise a cohort study?
- Are the exposed cohort from the same community as the non-exposed?
- If it is a historical cohort how was the data collected and how complete is it?
- Was the follow up period long enough?
- Can we be sure the participants were outcome free at the start of the study?
- Was follow up of participants adequate?
- What is the key finding, how precise is it and how applicable are the results to your population of interest?
What should be the analysis in an RCT?
- Robust randomisation eliminates confounding so do intention to treat analysis to preserve randomisation to allow causal inferences to be made
- Blinding not always possible but if it is it eliminates bias
What are some issues we are concerned about when we appraise a RCT?
What are some of the key questions we need to ask when appraising a RCT?
What do systematic reviews help to do and what are some issues we are concerned about when we appraise them?
Help to identify gaps and areas for further research
What are some of the key questions that we need to ask when appraising a systematic review?
