Intro To Evidence Based Medicine (EBM) Flashcards
Why is research based medicine important
Research methods teaching is designed to prepare you to understand and critique (for yourself) medical research and evidence
What is evidence based medicine
Conscientious, explicit and judicious use of current
best evidence in making decisions about the care of
individual patients
“Evidence-based medicine is the integration of best research evidence with clinical expertise and patient values”
Therefore a good doctor will look use his clinical experience on the condition and the external evidence to form an appropriate treatment whilst taking the patients concerns and values into consideration
Lead time bias
Mrs A didn’t like the thought of screening, and was diagnosed with cancer of X at 58, died two years later, aged 60
Mrs B, underwent screening and was diagnosed with a cancer of X at 48. She lived 12 years after the diagnosis and died aged 60
As we can see from this - she died at the same time - since diagnosis she it seems she lived 10 more years - in reality there was no difference in death - she just lived in fear/under treatment for more years
you actually need to demonstrate that patients live to be older, and their overall quality of life improves, rather than…
just have a longer period alive with a bad diagnosis, undergoing multiple unpleasant tests and therapies
Types of study design
Quantitative - research that involves numbers
Can be broken into observational and experimental research
Qualitative - experiments that involve “stuff”
Observational study design
Descriptive studies - 3 main types
1) Case reports and case series
2) Cross- sectional studies
3) correlational studies / ecological studies
Case reports and case series
May be vital in identifying new emerging diseases (e.g. HIV )
May be useful in suggesting aetiological associations
(e.g. 3 cases of angiosarcoma identified in Vinyl chloride workers)
Observational study design
Cross-sectional surveys - (often studies of prevalence, i.e the proportion of a given population with a disease)
May try to explore link between disease and possible
“exposure” – but this would be exploratory
(low grade evidence as not clear which came first chicken/egg)
Often used to address questions of “time, place, person”
Time: patterns in disease over time (repeated cross-sections)
Place: geographical distribution
Person: personal characteristics (suggesting possible causation)
Case control and cohort studies - Important designs to identify links between possible causative (aetiological) agents and disease
Classic work of Bradford-Hill and Doll on smoking and lung cancer (looked at Drs born between 1900 and 1930 and looked at survival rate from age 60 of Drs who continued to smoke and lifelong non smokers
Experimental study design
Interventional studies:
Controlled trials e.g comparing results of two treatment pathways one used in one hospital, the other in a different hospital year 1)
Randomised controlled trials [RCTs]
Reduce confounding / bias (you will learn about this in
Compare two treatments
Compare new treatment against placebo/usual care
Systematic reviews
Systematic review combine study results together They look at the combined evidence of: Randomised controlled trials Observational studies Qualitative studies
Hierarchy of evidence for treatment effectiveness
Consensus/ Expert views —> case study/series —> Obervational study/ cohort studies/ case control studies/ cross sectional studies —> Controlled, non randomised study —> randomised control study —> systematic review of randomised trials
Negatives of ‘better designed studies’ e.g. RCT
RCT: – May have poor randomisation technique – Intervention may be poorly reproducible – May have biased outcome measurement – May be analysed poorly – May not be generalizable
Sources of evidence
Books
Journals
Databases - Cochran’s database of systemic reviews
Internet - Pubmed
How do i do a good literature search
Use PICOS
Check for:
1) Population or patient group
(2) Intervention(s) (or investigation) considered
(3) Comparator (control)
(4) Outcome(s) considered - may not need this
(5) Study design – may not need this
(Often only need to combine two elements: Patient group + Intervention; or three elements: Patient group + intervention + study design)
Problems with EBM
Not everything needs to be tested with an RCT (e.g.
replacing blood volume after massive haemorrhage).
– BUT: be very aware that even some obvious treatments do
not work – so remain questioning!
Not everything can be tested with an RCT (e.g.
smoking).
Not everything has yet been answered!
