Flashcards in Clinical Trials Deck (16):
Means that the person randomising the patient does not know what the next treatment allocation will be.
Why is Allocation concealment Key to the success of randomisation
– Protects random sequence– Prevents selection bias (also called allocation bias)
Allocation concealment types
• Assignment by centralised 24 hour hotline (remote telephone )
• Pre-numbered or coded containers administered as per allocation
• Computer allocation • Sequentially numbered, sealed, opaque envelopes
Masks the assigned intervention during the whole trial•Why?–Reduces influence of subjectivity–Minimise observer bias
Intention to treat
ALL participants analysed in the group assigned (irrespective of whether or not they completed, or received the treatment)
Why?•Preserves randomisation•Models real world
Per protocol analysis
Only analyse participants who were compliant with thestudy protocol
Complete case analysis
Only analyse participants with available outcome data.
Differences only explained by ...
1.Random error2.Treatment effect3.Systematic error
Elements of a research question
Types of trials
Parallel group trial
First in humans: Evaluates safety, dose & PK/PD. Not randomised. In healthy volunteers or people with advanced disease (e.g. cancer). N=20-60. Extensive monitoring, clinic or hospital-based
In target population: Small (<100 people), short-term studies of safety, dose ranging & “surrogate”efficacy, sometimes randomised.
In target population: Larger (100s) short-term studies of surrogate & target outcome, Randomised.
In target population: Large (100-1000s), looks at long-term efficacy or effectiveness, and safety. Randomised. Looks at acceptability. Compares to usual care or as an adjunct.
In target population or general popn: Post-marketing surveillance, often pragmatic, and head-to-head comparison. Looks at acceptability, picks up rarer adverse events