Clinical Trials are constructed to statistically test a hypothesis. Once the study has been completed the results are analysed based upon the original hypothesis (primary outcome) and conclusions are drawn based upon the analysis.
It is generally accepted that the play of chance could have produced the observed results up to 1 time in 20. This is often expressed as a P value of 0.05 or less.
Where studies demonstrate that there is little likelihood that the results occurred by chance the quoted P value will be below 0.05 and the results are said to be statistically significant.
However, statistical significance may not be the same thing as clinical significance. This is because statistical differences are also reliant on the size of the population studied therefore a small difference can be statistically valid if the population size is large enough. A large difference is required if the population size is small.
For example, a 2 point difference on a 60 point depression rating scale was found to be statistically significant in comparing escitalopram (cipralex) to citalopram (cipramil). In a clinical setting detection of this 2 point difference would be virtually impossible to detect.