Clinical Trials in CVD Flashcards
What type of trial is a clinical trial?
Clinical Trials are longitudinal studies designed to assess if an intervention (removal of exposure) changes the incidence of an outcome. Most interventions are expected to decrease the incidence of the outcome. Most of them involve a control group for comparison.
What are benefit of clinical trials?
Clinical trials are the ‘gold standard’ for evidence of causality:
- active change of exposure status
- tightly controlled study environment
They provide the most evidence for Evidence based Practice
What are some key outcomes of clinical trials?
The Key outcomes of clinical trials are:
- Relative Measures of intervention effect:
- relative risks
- hazard ratios
- Absolute measures of intervention effect:
- absolute risk/rate reduction
- number needed to treat
- Survival Analysis
How is confounding overcome typically?
Randomisation
- random allocation of subjects into each arm of a clinical trial
- Objective: treatment groups identical in all aspects other than the intervention
- Rationale: reduce confounding
What is information bias?
- a systematic difference(s) in the way that information is collected between/among groups being compared.
- Differences are partly responsibe for the observed study results
- Arises when there is variability (esp. subjectivity) in methods for collecting information.
How is information bias overcome?
Blinding (Masking)
- non-awareness of intervention allocation
- Single - blind: subjects or investigators are unaware
- Double-Blind subjects and investigators are unaware
- Rationale: reduce Information/observer bias
What is selection bias?
- systematic difference (s) in characteristics of subjects within groups being compared
- these differences are partly responsible for the observed study results
Explain Cross-Over in (Parallel) Clinical Trials.
Source of selection bias if:
- Significant; and reasons likely to influence outcomes - ie. sick subjects cease active drug due to side effects - healthier group on active drug (less outcomes) - perception that active drug is better than placebo
How do you overcome selection bias?
Intention-to-Treat Analysis:
- assumes that subjects remained in randomised group, regardless of cross-over
- Rationale: reduce selection bias
- always under-estimates any treatment effect (ie. provides conservative estimate)
- Reason: cross over introduces overlap in treatment between groups, which is ignored.
What is the hazard?
- continuously updated, instantaneous rate
- measured in longitudinal studies with close follow-up mostly clinical trials
- example:
- follow up of 1000 subjects, outcome = death
- Week 1, 10 die: hazard for week 1 = 10/1000
- Week 2, 15 die: hazard for week 2 = 15/990
- during follow-up, explicit capture of outcomes (events) and their time of occurence
- measure of interest = ‘time to event’
What is a Kaplan- Meier Curve?
- Kaplan-Meier Curve: plot of hazard or survival (1-hazard) versus time
(survival does not refer to avoidance of death, but avoidance of the event)
What is the hazard ratio?
- conceptually similar to relative risk
- ratio Hintervention : Hcontrol
- derived from (more complex) statistical analysis
- applies to the whole period of follow-up
- example: HR = 0.5 (at any given point in time within the period of follow-up, probability of the outcome in intervention group is half of that of the control group)
What is risk/rate reduction?
- most interventions reduce risk/rate of outcome
- reduction measured in relative and absolute terms
- example:
–rate of outcome in control arm: 10/100py
–rate of outcome in intervention arm = 7/100py
–RR = 7/100py ÷ 10/100py = 0.7
–absolute reduction = 10/100py - 7/100py = 3/100py
Explain what is meant by number needed to treat?
- number of people needed to undergo the intervention in order to prevent outcome in one
- marker of the efficiency of the intervention
- NNT = 1 ÷ (absolute risk or rate reduction)
- example:
– absolute rate reduction = 3/100py
– NNT = 100/3 = 33.3 per year - affected by: – relative effect (often constant) – underlying likelihood of outcome
- example 1: – Rcontrol = 10/100py; Rintervention = 7/100py; RR = 0.7 – absolute reduction = 3/100py; NNT = 33.3
- example 2: – Rcontrol = 10/1000py; Rintervention = 7/1000py; RR = 0.7 – absolute reduction = 3/1000py; NNT = 333
Explain what is meant by number needed to harm?
- when interventions increase risk/rate of outcome
- example:
–rate of outcome in control arm: 10/100py
–rate of outcome in intervention arm = 14/100py –RR = 14/100py ÷ 10/100py = 1.4
–absolute reduction = 10/100py - 14/100py = -4/100py
–NNT = -25 per year
–NNH = 25 per year
