Fundamentals Flashcards
Interventional vs Observational Study
Intervention - give drug, Observation - watch
4 types of observational
Cross section, cohort, case control, ecological
Cross sectional Study example
Survey
Cohort study
Follow a population over time
Case control Study
1 group with outcome, 1 without
Ecological
Measure disease rates in population
Gold Standard of Clinical Trials
Randomised Controlled Trials
Randomisation
Allocate treatment groups at random
What type of bias does randomisation avoid?
selection bias
What is bias?
Tendency of estimate to deviate one way or other
What does bias impact on? (2)
Validity and Reliability
Why do clinical trials follow rules? 4
to make sure they are well designed, safe, right measurements,meaningful results
Nuremberg code
Research ethics for trials on humans - end of WWII
What was first sig effort to regulate trials? year
Declaration of Helsinki 1964
CTIMP vs non CTIMP
CTIMP - fall within law, non CTIMP - NHS governance
Phase I trials - describe
First in man and healthy people or patients
How many in Phase I?
What is Phase1 used for?5
Tolerability, safety, bioavailability, pharmacodyn and kin, max tolerated dose
Phase II who?
Larger group - with condition - narrow inclusion criteria
How many people in phase II?
30 - 50 people
What does phase II test?
Safety and effect
Phase III how many?
100s to 1000s
Phase III describe
New vs Placebo - looking for approval
Phase IV
After it has been approved
Four features of good clinical trials
Randomisation, controlled, blinding, large sample size
Control group - why
existing standard of care or placebo - compare to cancel external factors
Randomisation
Neither doctor nor patients choose, sequence concealed, so treatment and control are sim in respect to prog factors
Blinding why?
Minimise observer bias
Why large sample size?
Stat power
7 trial designs
Parallel, cross over, factorial, group allocation, superiority, equivalence, non inferiority, adaptive
Parallel
Most trials, intervention and control simultaneously
Cross over
A then B or B then A
Cross over design assumptions 2
One effect doest carry over into other, effects are independent of which one first
When is cross over used? 2
2 versions of same drug - bioequivalence , not if conditions change
Factorial design
2x2 - 2 interventions, one control
Group allocation
Group randomised - when risk of contamination
Superiority
New better than old
Equivalence
New more or less equal to old
Non inferiority
new no worse than old
Adaptive
multi -arm, multi stage , add or subtract arms
