All clinical trials are established to answer a clinical question; this is called the primary endpoint. It is to be hoped that the primary endpoint is demonstrated to be statistically significant at the end of the study but this is not always the case.
When a primary endpoint is not significant there is often a tendency to focus upon other endpoints that were investigated during the study or secondary endpoints as they are called. Consideration should be given to the fact hat the study may not have been designed to answer clinical questions based upon the secondary endpoints and therefore these results should be viewed more cautiously.
For example, in the PROactive study, recently covered here, the primary endpoint was not statistically significant while the secondary endpoint of death, non-fatal myocardial infarction and non-fatal stroke was significant. However, this study was designed to assess the efficacy of pioglitazone in secondary prevention based upon a composite of death, non-fatal MI, stroke, acute coronary syndrome, leg amputation, coronary revascularisation and revascularisation of the leg. Forming conclusions based on the secondary endpoint may not valid.