Questions to ask for critical Appraisal Flashcards
What should you ask yourself about a trial?
Who did the research and where was it published?
- Did the study address a clearly focused research question (what was the pico?)?
- Does the methodology fit the objective/questIon?
- Was the assignment of participants to interventions randomised?
What is allocation concealment
- Preventing selection bias by separating the process of randomisation from the recruitment of participants thus concealing the allocation sequence from those recruiting participants to a study.
- Allocation concealment prevents researchers from (unconsciously) influencing which participants are assigned to a given intervention group.
What is baseline comparability?
-The result of successful randomisation. Trialists should report distribution of baseline factors.
What is stratification?
-Ensuring balance between groups for specific baseline factors known to be associated with the outcome of interest. Embedded in the randomisation process. Permuted block can be used as part of stratification. Size of blocks is ideally varied to ensure allocation concealment in maintained.
Why do we need to know about loss to follow-up? Missing outcome data?
It can introduce bias
How do you assess for attrition bias?
Follow through numbers randomised to numbers analysed. Look for accounting; satisfactory audit trail. Cannot assume that no mention of withdrawal=no losses. Cannot assume that those that get lost have similar prognoses to those that don’t.
What is intention to treat analysis?
Participants should be analysed in the groups to which they were randomised, whether or not they completed or ever received the intervention. Prevents bias caused by the loss of participants, which may disrupt the baseline equivalence established by random assignment.
What is blinding in a study?
Want to protect against differential treatment and assessment of the participants. This can be maximised via blinding to prevent performance and detection bias.
What can knowledge of allocation increase the risks of?
Risk of participants using co-interventions based on knowledge of the treatment group (performance bias). Differential treatment of participants by health professionals or other staff. risk of differential withdrawal from the trial.
What is detection bias?
- Systematic differences between groups in the way outcomes are assessed.
- knowlege of assignment can influence outcome
What are odds ratios?
Odds Ratio = (odds of the event in the exposed group) / (odds of the event in the non-exposed group)
What is theoretical sampling?
-The process of sampling is on-going and iterative. Sampling responds to data collection and analysis to allow the researcher to explore areas of understand that may be relevant to developing the area of theory they are focused on.
