A study of horses for methicillin-resistant Staphylococcus aureus after cleared wound infections

Abstract

An outbreak of methicillin-resistant Staphylococcus aureus (MRSA) infection in horses in Sweden raised questions concerning the risk posed by horses to their surroundings outside hospital after MRSA-infected wounds have cleared. This initiated a longitudinal study of the effect of time on natural decolonization and sensitivity of different sampling sites for detection of MRSA. MRSA is notifiable in animals (Swedish legislation SJVFS 2007:90) and rare in horses within Sweden. The study, approved by the Ethical Committee on Animal Experiments, Uppsala, Sweden (C 309/8), continuously enrolled all horses notified with MRSA-infected wounds after the horse owner’s informed consent was obtained (n1⁄410). Study design: Repeated sampling at five anatomical sites (nostrils, corner of mouth, pastern, perineum, and previous infection site) six times at approximately monthly intervals and a seventh 6 to 12 months after the sixth sampling occasion, starting in October 2008. Day zero relative to the sampling was the date when each horse infection was detected. MRSAspecific enrichment broth and agar was used for culture. PCR-verifiedMRSA isolates (mecAand nuc gene) were spatyped. The effect of time on detecting MRSA was assessed by a multilevel mixed-effect logistic regression model. Sensitivity of sampling sites was calculated. Eight of 10 horses completed the sampling period and two were sampled two and six times, respectively. MRSA were detected in 16/333 (4.8%) samples and on 13/67 (19%) sampling occasions. Four horses tested negative throughout the study, two tested positive once, three twice and one five times. The probability of finding positive MRSA samples significantly decreased over time (p < 0.01), with the last finding 205 days after the infection had been diagnosed.