clinical trials Flashcards
Evidence based medicine
use of current evidence in making decisions about the care of individual patients
systematic observation examples (more the better)
meta analysis rct uncontrolled trial case series anecdote
types of outcome measures
patient oriented outcomes > stage or extent of disease > disease surrogate markers
bias
affects accuracy
any influence which makes observed results non representative
variability
affects precision
high variability makes it difficult to discern treatment differences
randomisation and groups
groups must be alike in all aspects except treatments
simplest term each patient has same chance of receiving any of the treatments
problems with simple randomisation
can lead to unequal no. people in each group
or imbalanced factors
blocking and stratification
eg aabb abab abba ect so after each 4 there’s equal numbers
stratification has randomisation in each strata eg <75 and >=75
historical controls
new treatment compared to previous comparable subjects
advantages historical controls
cheap and rapid
disadvantages historical controls
change is due to underlying changes changed criteria for selection quality of old data care and management peripheral to treatment changed diagnostic/evaluation criteria
concurrent controls
non randomised patients compared by different strategies used in same period
selection bias and can’t compare treatment groups
blinding
To avoid conscious or unconscious bias bu masking intervention
single-patients
double-and investigators
triple- and commute monitoring response variable
how to blind
placebo
why to blind
can effect patients report
can affect drop out rates
can affect preconceived notions about therapy
blinding problems
not always possible/ethical
eg surgery or if side effects
parallel groups
randomised and then all groups compared (control group aswell)
dose ranging
eg high medium low and control
cross over
when sequence of treatment randomised
for conditions that return to baseline
eg chronic pain, hearing, gingivitis
washing out to prevent carry over effect
factorial
a and b and control
try see effect of intervention and if it differs in absence or presence of other intervention
problems with poly pharmacy effects
protecting patients rights
scientific reviews
ethics committees
data safety and monitory boards
