Drug trials

Drug trials may be

Trial design	Arrangement	Notes
Double-blind	Neither patient nor researcher assessing the patient knows which treatment the patient has been randomised to	Patients must be randomised
Single-blind	Either the patient or researcher does not know the treatment (usually the patient)
Unblind / Open	Both the researcher and the patient knows the treatment
Cross-over studies	Each subject receives both the treatment and placebo in random order	Has the advantage in that each patient can act as their own control Removes many between confounders Requires fewer subjects Can only be used for chronic diseases that are not cured by the treatment. They should not carry-over effect from each treatment
Parallel studies	Two or more separate groups of subjects receive just one treatments being compared, one of them acting as the control group

Intention to treat analysis

Individuals are analysed according to the group they were randomised to regardless of whether they complied or completed the study or not

Basic Statistics

Averages:

Mean: The arithmetic mean is the average value
Mode: The most frequent value
Median: The middle value when all the values are ranked. It is a more robust measure of the centre of a distribution as it is not affected by outliers as the mean

Standard deviation

Standard error

Confidence Interval

Range of values within which the "true" population parameter is believed to be found, within a given level of confidence

Sensitivity

A measure of how good a test is in detecting those individuals that have the disease

Specificity

A measure of how good a test is in detecting those individuals that do not have the condition

Null hypothesis

States that there is no difference between groups#
Significance tests are carried out on the assumption that the null hypothesis is correct
If P-values are large (>0.05), then there is insufficient evidence to reject the null hypothesis
Statistical significance implies that the associations are unlikely to be due to chance and hence the null hypothesis can be rejected

Type I error (false positive)

Probability of failing to accept the null hypothesis when it is correct
The probability of making a type I error (alpha) is the level of statistical significance

Type II error (false negative)

Probability of failing to reject the null hypothesis when it is false
The probability of making a type 2 error (B) is determined by the power of the study (1-B)
If B is set at 10%, then there is 10% hance of failing to reject a false null hypothesis
Most studies have a power of 80-90%
The power of a study is particularly increased by increasing the sample size

Drug studies

Pre-clinical: form the basis of the first in man studies

Clinical trials: Phase I - safety assessment studies

Gradually increasing the dose based on the PK until the desired effect or adverse effects are observed
Performed on healthy volunteers
Data help to define the candidate dosage forms and regimens for evaluation in phase 2 in small numbers of patients
Placebo controlled and usually double blind

Clinical trials: Phase II - defining likely safe and efficacious dose for the large phase 3 trials

Clinical trials: Phase III - large scale trials comparing the new treatments to existing treatments

Population PK/PD studies - this can also improve computer modelling
Large efficacy trials
PK/PD in special populations
Once complete, the drug is usually submitted to the regulatory authorities for a product licence

Clinical trials: Phase IV - post marketing surveillance

www.surgicalnotes.co.uk