L31. Clinical Trials in CVD Flashcards
What are clinical trials?
Longitudinal Studies
Assess whether an INTERVENTION changes the incidence of an outcome
Must always include a CONTROL group
They are GOLD STANDARD for evidence of causality
What are the key outcomes of clinical trials?
RELATIVE:
Relative Risks
Hazard Ratios
ABSOLUTE:
Absolute risk/rate reduction
Number needed to treat
What is a major method to reduce confounding?
Randomisation of subjects into each arm of the trial
Can also match by confounders
What is blinding? What does it deal with?
Blinding is a non-awareness of intervention allocation at multiple levels (subjects, assessors and examiners) for single or double blinding
This is to reduce information bias
What is intention-to-treat analysis? What does it deal with?
Treating every subject regardless of what they ended up doing (Eg. switching drugs or dying or dropping out of the study) were initially assigned to
Minimises selection bias and gives a conservative estimate of the results
What is a hazard?
A rate that is continuously updated
Gives the instantaneous rate of incidence that gives exactly when events occur
Eg. Hazard for outcome is 10/1000 in one week then 15/990 in the next week
What is the hazard Ratio?
A specific type of Relative Risk
It is a ratio of Hazard of intervention: Hazard Control
It gives the probability of an outcome in intervention group at any given point in time
What is survival analysis?
Is a measure of the outcomes and their time of occurance giving a “time to event” measurement
Eg. Survival curves
What is the number needed to treat analysis?
The number of people needed to undergo the intervention in order to stop the outcome occurring in just one person.
Marks the EFFICIENCY of the intervention
NNT = 1 / absolute risk or rate raduction
What is the number needed to harm?
Used when interventions increase a risk/rate of outcome (ie. don’t work)
It is a negative number to treat
