Thirty seven volunteer single site farrow to wean herds in a pig dense region of southern Minnesota, 20 filtered and 17 non filtered controls, were compared over a seven year study period. Air filtration reduced the risk of novel PRRSV introduction by approximately 80 percent. Installation cost stated in the paper was approximately 150 to 200 dollars per sow, or roughly 450,000 to 600,000 dollars for a 3,000 sow herd.
This is the source of the 80 percent figure, and it is frequently attributed to the wrong paper. Its financial companion, published in Preventive Veterinary Medicine volume 111, carries the payback analysis; this study is volume 112 and carries the epidemiology. What distinguishes it from the earlier work of Dee and colleagues in 2012 is that its result was tested for robustness across a range of phylogenetically based definitions of what counts as a novel introduction, which addresses the case definition uncertainty in that study. The two should be cited together rather than used interchangeably.
The implication that follows from an 80 percent reduction is the one worth stating plainly. If closing the airborne route removes approximately four fifths of novel introductions in a swine dense region, then roughly that share of introductions in such a region were arriving by air. That is the strongest available justification for treating air as a biosecurity boundary rather than as a ventilation input. The remaining fifth is the reason filtration is one control within a programme and not the programme itself.