Quantifying topical antimicrobial use before and during participation in an antimicrobial stewardship programme in Dutch companion animal clinics.

The emergence of bacterial strains resistant to topical antimicrobials in both human and veterinary medicine has raised concerns over retaining the efficacy of these preparations. Yet, little information is available regarding the use of topical antimicrobials in either sector for planning targeted interventions. This study aims to quantify the use of topical antimicrobials in 44 Dutch companion animal clinics before and during their participation in an antimicrobial stewardship programme (ASP), to explore the effect of the intervention on topical antimicrobial use (AMU). Hence, prescription and clinic animal population data, collected from July 2012 until June 2018 were used. Specifically, the period from July 2012 until June 2015 was defined as pre-intervention period, whereas clinics started to participate in the ASP from March 2016 onwards. As quantification metric, the Defined Daily Dose for Animals (DDDA) was used and a mixed effect times series model with auto-regression was applied to monthly topical AMU data. The intervention effect was modelled using a step function with a change in (linear) time trend and clinic characteristics, as potential determinants of topical AMU, were assessed using a multivariable regression model. A seasonal pattern was identified, in the pre-intervention period, where topical AMU was highest in July-August and lowest in February-March. In addition, total topical AMU appeared to significantly decrease over time in the pre-intervention period and the proportion of dogs in the clinic was positively associated with topical AMU. The intervention effect was significant only for second line and for skin product AMU. This study demonstrates that during participation in an ASP, second line and skin product AMU decreased in Dutch companion animal clinics. Additionally, this study demonstrates the existence of a seasonal effect and a decrease in topical AMU over time already before introduction of a targeted intervention.