Sampling size methodologies used in pig welfare audits underestimate animal-based measures: a preliminary exploration

The objective of this observational study was to compare and contrast the occurrences of animal-based measures evaluated in the Common Swine Industry Audit (CSIA) with the use of 2 sampling methods (total farm inventory [TOTAL] vs the CSIA sampling method).

Approximately 240,000 pigs were evaluated across 60 farms between October 2022 and July 2023. A map was created for each farm to identify individual pig location by pen/stall, room, and barn. Ten animal-based measures were assessed per pig, and the total occurrence of each measure was calculated (TOTAL). Utilizing farm maps and the same dataset, the CSIA sampling method was calculated to randomly assess a previously designated sample number of pigs per farm by location. The CSIA occurrences were then compared to TOTAL occurrences.

Sow: occurrences of a body condition score of 1, severe lameness (non-weight-bearing), abscesses, open wounds, severe scratches (> 25% of the body), prolapses, vulva lesions, and shoulder sores were less for the CSIA method compared to TOTAL. Nursery: occurrences of severe lameness, abscesses, open wounds, and hernias were less for the CSIA method compared to TOTAL. Finishers: occurrences of severe lameness, abscesses, open wounds, tail bites, prolapses, and vulva lesions were less for the CSIA method compared to TOTAL.

These findings suggest that the sampling method currently used in the CSIA tool does not necessarily accurately estimate the occurrence of animal-based measures at the farm level.

Future work on swine farms should explore alternative sampling methods that more closely represent welfare conditions to that of the total farm inventory.