clinical trials - core reading Flashcards
what is a clinical trial
planned experiment in humans
measure efficacy or effectiveness of an intervention
benefit of experimental studies
not subject to problems like confounding and bias
types of clinical trials
parallel - individuals assigned to different interventions
cross over trials- each patient is their own control - take treatment and control treatment in a random order - then a wash out period
cluster trials - groups of patients, clinics or areas form basis of intervention, outcome compared between groups rather than between individuals - useful if intervention will have a group or population effect eg testing a vaccine, or if interventions to some people will spill over to the rest of an area eg dietary trial.
why randomise
remove selection bias
open label studies
binding not possible - surgical intervention compared to medical treatment
still need randomisation and controls and people doing outcome measurements are blinded to which treatment the patient received
ethics and consent
strictly regulated to ensure the patients are protected
registered
reviewed by independent scientific committee and adhere to government and national guidelines
absolute risk reduction
difference between the experimental and control groups as rate
numbe rneeded to treat
numbe rof people who need to receive the experimental treatment to prevent the worse outcome - reciprocal of absolute risk reduction 1/ARR
things to look for in a clinical trial
high quality? assessing important qn? how complete was follow up? -ve as well as +ve outcomes blind apply to my practice intervention adequately described intention to treat or per-protocol
intention to treat or per protocol
decision made about people who deviate from protocol - depends on the aims of the study
should do intention to treat - maintains randomisation and reduces confounding
judge efficacy - should look at people who stick to protocol
judge effectiveness- how well actually works in practice
phase 1
safety
side effects
small number
may be healthy
phase 2
larger group
have disease
see if treatment effective - short term
safety
phase 3
with placebo how well new treatment works side effects several 1000 people different countries smaller expected advantage = more people take part
phase 4
after drug marketed information from various populations long term side effects observational confounding and bias - other observational studies real world information
