We will investigate if testosterone is preferentially given to patients at higher risk of cardiovascular events. If this is the case, then observational studies may be prone to selection bias resulting in overestimation of the cardiovascular risk associated with testosterone when compared to a healthier population not taking testosterone

Inclusion Criteria:

- All male individuals in the source population with a first testosterone use during the study period will be included as cases. The date of the GP consultation associated with the first testosterone prescription will be designated as the index day.

Exclusion Criteria:

- Patients of less than 18 years on the index day, with less than 2-year contribution to the CPRD-HES link before the index day, or with use of testosterone before the study period will be excluded from the group of cases.

The source population is the subset of male patients in the Clinical Practice Research Datalink (CPRD) that has been linked to the Hospital Episodes Statistics (HES) database (CPRD-HES).

Cases will be defined by first testosterone use between 1st January 2001 and 31st March 2016 (study period). The date of the GP consultation associated with the first testosterone prescription will be designated as the index day. All cases will be required to be over 18 years on the index day, to have contributed for at least 2 years to the CPRD-HES link and to have no history of testosterone before the index day.