Evidence Based Medicine (3&4) Flashcards
Hierarchy of evidence
- Meta-analysis of RCTs (randomized clinical trials)
- Single large RCT
- Cohort Study
- Case-control study
- Case series
- Expert opinion
- Meta-analysis of randomized clinical trials (RCTs)
Considered the best evidence and the least biased.
Pool of single large RCTs
- Single large RCT
This is the second best evidence, only after a large pool of these
- Cohort Study
(level 2)
Observational study and more biased than RCTs
- Case-control study
(level 3)
Observational study and more biased than RCTs
- Case series
(level 4)
Observational study and without a comparison group
- Expert opinion
The worst type of evidence
What are the common pitfalls in the conduct of clinical research
- Lack of randomization
- Lack of concealment of randomization
- Lack of blinding
- Wrong intention to treat analysis
- Errors in hypothesis tesgting
- Lack of randomization
(common pitfalls in the conduct of clinical research)
Randomization is important because:
- it gives every patient an equal probability of being enrolled to either group.
- It is the single most important technique to limit bias in clinical research. Thus, studies that fail to randomize patients introduce an important bias in the measurement of the outcome measure.
Why is randomization so important? (1)
it gives every patient an equal probability of being enrolled to either group.
Why is randomization so important? (2)
It is the single most important technique to limit bias in clinical research. Thus, studies that fail to randomize patients introduce an important bias in the measurement of the outcome measure.
Concealed randomization ensures that surgeons are unable to ________ the treatment to which their next patient will be allocated.
(common pitfalls in the conduct of clinical research)
predict
The safest manner in which to limit the occurrence of failed concealment is a _________
remote 24-hour telephone randomization service
If randomization is concealed, then the physician cannot ______ the patient will be allocated to at the time of patient consent and application of exclusion/inclusion criteria.
predict to which treatment
Randomization techniques that preserve ________ can include telephone and internet-based randomization.
concealment
_____ also guards against potential bias in a research study.
Blinding
While blinding is feasible in most, if not all, drug trials, it is rarely feasible in a ______trial
Surgical
When ______ is not possible, investigators should take care to ensure _________ assessment of the outcome measures if they require some interpretation (i.e. Radiographic healing of bone, assessment of clinical outcomes, physical examination
blinding is not possible; independent
When blinding is not possible then researchers should ensure that
Independent assessment of the outcome measures if they require some interpretation (i.e. Radiographic healing of bone, assessment of clinical outcomes, physical examination
- Lack of concealment of randomization
Basically being able to predict what treatment the patient will be allocated to
- Lack of blinding
Basically knowing and having bias in a surgical study (ex. giving the more sick some treatment to show more improvement)
- Wrong intention to treat analysis
When some patients undergo the alternative treatment they were not allocated to either by mistake or the result of technical difficulties
Even when proper randomization is performed, some patients undergo the alternative treatment they were not allocated to either by mistake or the result of technical difficulties with one treatment
This approach seems to be COUNTERINTUITIVE at first glance, it PRESERVES the value of randomization: on average known and unknown prognostic factors will be equally distributed across both treatment groups and therefore the treatment effect will be primarily determined by the assigned treatment.
- Errors in hypothesis testing
A common pitfall in clinical trials