Impact of antimicrobial drug usage measures on the identification of heavy users, patterns of usage of the different antimicrobial classes and time- trends evolution C. CHAUVIN* M. QUERREC* A. PEROT* D. GUILLEMOT   & P. SANDERS à *French Agency for Food Safety (AFSSA), Pig and Poultry Research Laboratory, Ploufragan, France;   Institut Pasteur, Pharmacoepidemiology and Infectious Diseases Unit, Paris; INSERM U 657, Paris; and Universite ´ Versailles St Quentin, UFR Paris Ile de France Ouest, Garches, France; à French Agency for Food Safety (AFSSA), Veterinary Research Laboratory for Medicines and Disinfectants, Fouge `res, France Chauvin, C., Querrec, M., Perot, A., Guillemot, D., Sanders, P. Impact of anti- microbial drug usage measures on the identification of heavy users, patterns of usage of the different antimicrobial classes and time-trends evolution. J. vet. Pharmacol. Therap. 31, 301–311. Several measurement units are available to quantify antimicrobial usage in veterinary medicine, to obtain diverse measures such as the weight of active substance used, the live weight (LW) treated, the fraction of animals exposed, the number of treatments recorded or the cost represented. These measures can be applied to study practices variability between farms, to characterize patterns of usage of the different antimicrobial classes or to follow evolution of antimicrobial usage with time. An investigation was carried out to specifically explore the influence of measurement units on the conclusions obtained from such studies. Antimicrobial exposure was explored in a sample of turkey and chicken broiler flocks, using six different units [kg of active compound, treatments, days of administration, kg of LW treated, animal daily dose to treat 1 kg of LW (ADD kg ) and euros] to compare flocks usage variability and patterns of use of the different antimicrobial classes. Time-trends evolutions of macrolides usage in turkey broilers, characterized by percentage of flocks exposed and LW treated, were also compared. In all analyses, the measure gave different results without equivalences, highlighting the necessity for care in choosing the measurement unit and caution in interpreting the figures obtained. (Paper received 25 October 2007; accepted for publication 6 March 2008) Claire Chauvin, French Agency for Food Safety (AFSSA), Pig and Poultry Research Laboratory, Zoopole, BP 53, 22440 Ploufragan, France. E-mail: c.chauvin@afssa.fr INTRODUCTION Antimicrobial usage surveillance has been widely recommended these last years in veterinary medicine, as a key step in antimicrobial resistance containment (Nicholls et al., 2001; WHO, 2003, 2004). The first difficulty encountered is how to acquire this antimicrobial usage data (Nicholls et al., 2001). The second is the measurement unit to choose to express and analyse it. Several units are available to quantify drug usage but vary in their accessibility and present different advantages and properties (Chauvin et al., 2001). The most widely used to date, particularly to express consumption at an aggregated, e.g. national, level is the weight of active compound (in tons, kg or mg). As the weight unit is a basic measure, it is always applicable whatever the country or species concerned, and has been recommended for quantifying antimicrobial usage in animals (Nicholls et al., 2001; WHO, 2003). However, due to the wide variations in dosage between antimicrobials, this measurement unit does not readily reflect the therapeutic pressure. Other measures were therefore developed, based on transcription of the amounts of active compound acquired into the corresponding quantity of live weight (LW), which would have been treated. These are based on the prescribed or recommended doses and named prescribed daily dose or animal daily dose (ADD), whereas the defined daily dose (DDD) widely used in human medicine refers to an international standard, not defined in veterinary medicine (Anonymous, 2006a). These units approach the therapeutic pressure exerted by antimicrobials, which can also be estimated by computing the number of prescriptions or treatments administered, or calculating the proportion of animals exposed. This latter measure is considered a key parameter for risk assessment (McEwen & Singer, 2006) and an ideal unit for pharmaco-epidemiologic research (Lee & Bergman, 2002). Finally, antimicrobial usage for economic analyses can be evaluated from its cost and expressed in monetary units (Burkgren & Vogel, 2006). J. vet. Pharmacol. Therap. 31, 301–311, doi: 10.1111/j.1365-2885.2008.00960.x. Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Publishing Ltd 301