The trial and the insurance records disagree about blood clots
Insurance data on 39,622 men who had a clot found about double the odds after a testosterone prescription. The randomised trial found no overall excess, though lung clots ran 2.0% against 1.5%.
What the study found
Clots ran at double the odds after a testosterone prescription
Among 39,622 men who had a first venous clot, a testosterone prescription in the preceding months went with about twice the odds, with or without a hypogonadism diagnosis on record. The randomised trial found no overall excess, though lung clots ran higher, 2.0% against 1.5%.
The caveat · A new illness could prompt both
Each man was his own comparison, which handles his fixed characteristics but not a new illness or hospital stay that could prompt both a prescription and a clot. Only 7.8% of the men had a hypogonadism diagnosis code at all. A prescription record shows the drug was dispensed, not that it was taken.
What we don't know
Whether testosterone raises clot risk is unsettled. Insurance claims show about double the odds and the randomised trial no overall excess, yet both point the same way on clots in the lung, the part of the signal that has not gone away.