Prospective evaluation of the influence of sampling method on bacterial culture results and antimicrobial selection in 52 dogs with infected wounds.

Abstract

OBJECTIVE: To determine the influence of sampling methods on culture results and selection of antimicrobials for treating infected wounds in dogs. DESIGN: Prospective study from January to July 2016. SETTING: Three private multispecialty referral centers. ANIMALS: Fifty-two client-owned dogs with infected wounds. INTERVENTIONS: Each wound was sampled for culture by 3 methods: swabbing prior to preparation (dirty swabs), swabbing after debridement and sterile lavage (clean swabs), and biopsy. Bacterial species and phenotypes were compared. Three clinicians unaware of patient, wound, and sampling information selected antimicrobial drugs based on culture and sensitivity reports. Antimicrobials were divided into class I, II, or III based on established guidelines. The number, highest class of antimicrobial chosen, and inter-investigator agreement were examined. MEASUREMENTS AND MAIN RESULTS: Identical populations of bacteria were isolated for all 3 sampling techniques in only 31% of wounds. Significantly fewer bacterial species were isolated from biopsy samples (1.87 bacterial species per wound ± 1.14) than from clean swab samples (2.29 ± 1.18; P = 0.009) but not dirty swab samples (2.29 ± 1.29; P = 0.06). The recovery frequency for gram-positive bacteria was lower for biopsy compared to either swabbing technique (P = 0.001 for both comparisons). No difference was observed between clean and dirty swabbing techniques for any parameter examined. Sampling technique did not affect the proportion of wounds with anaerobic, gram-negative, or multi-drug resistant bacteria. The number (P = 0.28) and highest class of antimicrobial (P = 0.9) selected per wound did not differ between the 3 sampling techniques (P = 0.28). Clinician agreement was 83-90% depending on sampling technique. CONCLUSION: Although there were some differences in bacteria isolated from biopsy samples compared to swab samples from infected wounds, technique did not influence the number and highest class of antimicrobial selected by clinicians. Wound debridement prior to sampling by swabbing did not alter the number or type of bacteria isolated, nor the number or the highest class of antimicrobial selected by clinicians.