Task 4 – designing clinical trials
Learning goals
1) What is the difference in sample sizes between superiority and non-inferiority trials?
Non-inferiority trials need a bigger sample size
Depends on: standard error
2) What are the difference between intention to treat (ITT) and per protocol analysis? (pros/cons)
When to use which analysis and why?
INTENTION TO TREAT ANALYSIS
In the intention to treat (ITT) analysis every randomized participant should enter the primary
analysis. Consequently, patient who drop out or are non-compliant to the study treatment are
included in the primary analysis within the respective treatment group (arm) they have been
assigned to at randomization.
The patient set used for the primary analysis according to the ITT is called full analysis set.
Pros:
- The original randomization and the number of patients in the treatment groups remain
unchanged.
- Potential bias due to exclusion of patients is avoided
Cons:
- Imputation of missing data required
There are some specific reasons that might cause and exclusion of a patient for the full analysis set:
- No treatment was applied at al
- There are no data available after randomization
When to use
In a clinical superiority trial one wants to detect a benefit of treatment A compared to treatment B.
The aim is to disprove that treatment A is not better than treatment B null hypothesis
In clinical trials any over-estimation of the effect needs to be avoided. With respect to prevention of
type I error it is still better to choose a method which underestimates the effect (conservative
approach) than a method which might overestimate the effect.
In a superiority trial you should do the analysis according to the ITT principle
In this kind of analysis, the actual treatment effect is usually underestimates. It can be derived from
the fact that in the full analysis set also non-compliant patients are included and non-compliance is
generally associates with a negative outcome.
PER PROTOCOL ANALYSIS
, Task 4 – designing clinical trials
The aim of per protocol (PP) analysis is to identify a treatment effect which would occur under
optimal conditions.
What is the effect if patients are fully compliant?
Therefore, some patients (from the full analysis set) need to be excluded from the population used
for the PP analysis.
This applies to patient fulfilling any of the following criteria:
- Any major protocol deviations
- Non-availability of measurements of the primary outcome
- Non-sufficient exposure to study treatment
Pros:
- Proof of therapeutic concept
Cons:
- Selection bias
- Reduced power depending on non-compliance
When to use
The per protocol analysis is the primary analysis in non-inferiority trial
Could be the secondary analysis in superiority trials
o Difference between ITT and PP = bias
3) How do you determine the non-inferiority limit?
A non-inferiority trial aims to demonstrate that the intervention treatment is not worse than the
comparator (control) by more than a small pre-specified margin. This amount is knows as the non-
inferiority margin or delta.
The outcome a non-inferiority trial is usually assessed by a two-sided 95% confidence interval.
Two aspects of the results that should attract particular attention:
- Point estimate of the difference = observed difference between the test (intervention) and
reference (control).
o Best estimate of the true difference
- Lower limit of the confidence interval = represents the lower bound and is usually
interpreted as the degree of inferiority to the reference that can be excluded based on the
data presented
Delta has to be smaller than the smallest value that would make a clinical difference
4) How to compute the sample size calculation?
The sample size depends on several factors:
- Alpha = type I error
o Type I error = reject H0, if H0 is correct
- Desired statistical power = 1 – beta
o Type II error = do not reject H0, H0 is incorrect
Learning goals
1) What is the difference in sample sizes between superiority and non-inferiority trials?
Non-inferiority trials need a bigger sample size
Depends on: standard error
2) What are the difference between intention to treat (ITT) and per protocol analysis? (pros/cons)
When to use which analysis and why?
INTENTION TO TREAT ANALYSIS
In the intention to treat (ITT) analysis every randomized participant should enter the primary
analysis. Consequently, patient who drop out or are non-compliant to the study treatment are
included in the primary analysis within the respective treatment group (arm) they have been
assigned to at randomization.
The patient set used for the primary analysis according to the ITT is called full analysis set.
Pros:
- The original randomization and the number of patients in the treatment groups remain
unchanged.
- Potential bias due to exclusion of patients is avoided
Cons:
- Imputation of missing data required
There are some specific reasons that might cause and exclusion of a patient for the full analysis set:
- No treatment was applied at al
- There are no data available after randomization
When to use
In a clinical superiority trial one wants to detect a benefit of treatment A compared to treatment B.
The aim is to disprove that treatment A is not better than treatment B null hypothesis
In clinical trials any over-estimation of the effect needs to be avoided. With respect to prevention of
type I error it is still better to choose a method which underestimates the effect (conservative
approach) than a method which might overestimate the effect.
In a superiority trial you should do the analysis according to the ITT principle
In this kind of analysis, the actual treatment effect is usually underestimates. It can be derived from
the fact that in the full analysis set also non-compliant patients are included and non-compliance is
generally associates with a negative outcome.
PER PROTOCOL ANALYSIS
, Task 4 – designing clinical trials
The aim of per protocol (PP) analysis is to identify a treatment effect which would occur under
optimal conditions.
What is the effect if patients are fully compliant?
Therefore, some patients (from the full analysis set) need to be excluded from the population used
for the PP analysis.
This applies to patient fulfilling any of the following criteria:
- Any major protocol deviations
- Non-availability of measurements of the primary outcome
- Non-sufficient exposure to study treatment
Pros:
- Proof of therapeutic concept
Cons:
- Selection bias
- Reduced power depending on non-compliance
When to use
The per protocol analysis is the primary analysis in non-inferiority trial
Could be the secondary analysis in superiority trials
o Difference between ITT and PP = bias
3) How do you determine the non-inferiority limit?
A non-inferiority trial aims to demonstrate that the intervention treatment is not worse than the
comparator (control) by more than a small pre-specified margin. This amount is knows as the non-
inferiority margin or delta.
The outcome a non-inferiority trial is usually assessed by a two-sided 95% confidence interval.
Two aspects of the results that should attract particular attention:
- Point estimate of the difference = observed difference between the test (intervention) and
reference (control).
o Best estimate of the true difference
- Lower limit of the confidence interval = represents the lower bound and is usually
interpreted as the degree of inferiority to the reference that can be excluded based on the
data presented
Delta has to be smaller than the smallest value that would make a clinical difference
4) How to compute the sample size calculation?
The sample size depends on several factors:
- Alpha = type I error
o Type I error = reject H0, if H0 is correct
- Desired statistical power = 1 – beta
o Type II error = do not reject H0, H0 is incorrect
