What is evidence based medicine?
The conscientious and explicit use of the current best evidence to make decisions regarding healthcare and INDIVIDUAL patients
- so that the best evidence is used, it is gathered in systematic reviews, and the statistical analysis gathered in a meta-analysis
The outcomes of EBM can be diseases based (biochemical, staging) or patient based (pain relief, life-expectancy etc.)
What criteria are used to determine the quality of a trial?
sample size analysis bias valid hypothesis? ITT analysis are drop outs accounted for? multinational randomisation baseline characteristics balanced no payment/kudos appropriate inclusion/exclusion criteria suitable control outcomes appropriate representative sample blinding
Describe parallel study design:
Groups A, B and C all receive the intervention, and there is a control group
Describe dose-ranging study design:
Group A - 50mg
Group B - 25mg
Group C - 10mg
Control group (placebo)
- creates a dose-response curve to allow the most efficacious dose to be calculated
Describe cross-over study design:
Both groups get both interventions at some point, but there has to be wash out period in the middle so that drug A does not impact on the efficacy of drug B and vice versa
Only useful for conditions that return to baseline in wash out e.g. arthritis
Useful as smaller numbers of participants are needed and the variability is only WITHIN the subjects rather than between them
Describe factorial study design:
Group 1 = A + B
Group 2 = A + control
Group 3 = B + control
Group 4 = control + control
- evaluates two interventions at once but can have both SYNERGISTIC and TOXIC effects
Describe the hierarchy of evidence used to create EBM guidelines:
Systematic reviews and meta analyses RCT's Uncontrolled trials/case reports/cohort studies Non-analytical studies Expert opinion
- studies are all ‘graded’ to determine what hierarchy of evidence they are
What are limitations of a RCT?
- results may be STATISTICALLY significant but NOT CLINICALLY significant
- not all evidence can be tested in an RCT e.g. smoking trials/cancer trials (where you cannot have a control)
- you cannot always randomise (for rare diseases)
- often it is difficult to blind for things (e.g. surgery)
Define bias
- bias is artefact that is introduced into a study design leading to inclination for/against a result
- affects the ‘accuracy’ of a result (i.e. how close it is to the truth)
Define variability
- variability describes how similar subsequent results are to one another, and describes their ‘precision’
Describe barriers to participating in clinical trials:
- lack of awareness
- lack of knowledge
- preconceptions
- pressure from friends/family
- fear
- don’t want to go against physician’s wishes
Describe benefits of participating in clinical trials:
- HAWTHORN EFFECT (you will always do better in a clinical trial than not no matter if control or intervention as you are being monitored more closely)
- will receive gold standard treatment at a minimum
- help save lives and develop new treatments long term
Describe risks of participating in clinical trials:
- side effects
- new treatment may not be as good as current standard
- takes up a lot of time (appointments, surveys, clinics etc)
How do we ensure patients in clinical trials are protected?
- ethics committees
- informed consent
- patients can drop out at any time
- review boards
Why do we blind patients and physicians?
Patients
- avoid placebo affects
- avoid preconceptions
- reduce drop out
- achieve true reporting of side effects
Physicians
- stop compensatory treatment in control group
What are the different types of bias that exist?
CLOP To Some Music! Citation bias Language bias Observer bias Publication bias Time lag bias Selection bias Multiple publication bias
