Assessing antibiotic usage data capture accuracy on dairy farms in England and Wales.

Background: Accurate farm-level data on antibiotic usage (ABU) are needed for the surveillance of antibiotic resistance. Therefore, this study aimed to determine the accuracy of ABU data capture by dairy farmers in South West England and Wales.

Methods: Through a cross-sectional survey of 48 dairy farmers, the accuracy of ABU recording was measured by farmers' assessment of the completeness and timeliness of ABU recording ('perceived accuracy') and the completeness and correctness of on-farm ABU records ('actual accuracy'). Completeness and correctness were compared for paper and software recording methods.

Results: Perceived accuracy was higher than actual accuracy. Antibiotic names, withdrawal periods and dates that products were fit for human consumption were often incomplete or incorrect. More inaccuracies were seen with paper than software. In some software platforms, the date that milk would be fit for human consumption was frequently rounded down by half a day, increasing the risk of residue failures.

Limitation: The small number of on-farm records assessed limits the generalisability of the results.

Conclusions: Electronic recording of ABU should be encouraged. However, functionality needs improvement, alongside consultation with dairy farmers to increase awareness of inaccuracies.