Which 8 groups have the fewest trials performed on them?
What are observational studies?
How do you select a cohort?
Need to be free of outcome under study at the start
Can all join at the start = fixed cohort
Or can join during time period = dynamic cohort
What is incidence rate and how do you calculate it?
Incidence rate takes changes in population counts into account
The number of new cases of the disease within a specific time period
Incidence rate = Number of new cases / Disease-free person-time at risk
What is risk?
Risk: probability that an individual will be diagnosed with outcome over the follow up period
Assumes that they don’t have the outcome already
What is risk ratio?
Risk Ratio = incidence (exposed) / incidence (unexposed)
Follow-up within cohort study
Length of follow-up required will be dependent on the outcome you are wishing to evaluate
Follow-up is a potential weakness of a cohort study
Loss to follow-up can bias the results if those who are followed-up differ from those who are lost to follow-up
What are covariates (controlled variables)?
Age of those included
Whether more males or females included
Other diagnoses
Other medications
Smoking, BMI, socioeconomic status
What are the 5 strengths of cohort studies?
What are the 2 limitations of cohort studies?
What is matching in studies?
Method used to try to overcome confounding
Often match on age and sex but other matching possible
What are the 3 advantages of case-control studies?
What are the 4 disadvantages of case-control studies?
What are the 3 comparisons of cohort and case control studies?
Cohort - group without outcome
Case - group with outcome
Cohort - calculate relative risk
Case - calculate odds ratio
Cohort - exposure known
Case - event status known
What are the two types of bias?
Information bias
- recall
- measurement
- misclassification
Selection bias
- volunteer
- selection of controls and cases
What is confounding?
The distortion of a risk estimate due to the mixture of the people in the study population
A confounder is a risk factor for the disease and is correlated with the exposure independent of disease
Confounding by indication
A factor’s ability to be a confounder is entirely dependent on whether it is unevenly distributed between the study groups
The indication for prescription is one of the most important factors to consider when evaluating medication exposures
What is channelling?
Where a drug is prescribed to a group of patients because of
a characteristic of the drug
a characteristic of that group of patients
usually is a new drug
What are the 4 ways to control for confounding?
Randomisation - even distribution of confounders
Restriction - gives more control but cannot study variation between levels of that factor
Matching - each case is paired with a control subject for specified constraint/confounder
- Can be difficult to find a match
- Only controls for the criteria that have been matched on
Stratified analysis - separating out factors so any mixture of their effect is removed
What is an effect modifier?
Effect modification occurs when the effect of the exposure is different in different groups of the population
A factor that modifies the effect of a putative causal factor under study
There is no average ‘true value’
What are phase I trials?
Safety and dosage
20 – 80 healthy volunteers
Several months
What are phase II trials?
Efficacy and side effects
~ few hundred people with disease of interest
Several months to 2 years
What are phase III trials?
Efficacy and monitoring adverse reactions
300-3000 people with disease of interest
1 to 4 years
What are phase IV trials?
Required by regulator to follow up drug once released to marked
Trying to identify adverse events
Trying to determine any long term harm