In this historical cohort study we identified, located, and, if living, interviewed 10,590 vasectomized men from four cities, along with a paired neighborhood control for each. The times between procedure data and interview or death ranged from under one to 41 years, with median equal to 7.9 years and with 2,318 pairs having ten or more years of follow-up. Participant reports of diseases or conditions that might possibly be related to vasectomy through an immunopathological mechanism were validated by direct contact with physicians and review of medical records. Results of this study do not support the suggestions of immunopathological consequences of vasectomy within the period of follow-up. Except for epididymitis-orchitis, the incidence of diseases for vasectomized men was similar or lower than for their paired controls.
This along with the Br J Urol. 1996 Feb;77(2):328 are the two most cited references for the 0.1% to 0.25% rate for PVPS.
You’ll note that those numbers show up nowhere in this study. In fact, the study is only concerned with immunopathological consequences. They lump together epididymitis and orchitis together.
They also report an incidence rate or 33 per 10,000 man years, but that’s not a useful metric for one time outcomes as a result of surgery. What’s more important is the total occurances which is 187 per 10,000 patients.
This lines up nicely with the modern accepted rate of a 1-2% occurance rate.