Knowledge and Attitudes of Small Animal Veterinarians on Antimicrobial Use Practices Impacting the Selection of Antimicrobial Resistance in Dogs and Cats in Illinois, United States: A Spatial Epidemiological Approach

Abstract

Inappropriate antimicrobial use in animals and humans has been associated with the emergence of antimicrobial resistance, which has become a global public health concern. Veterinarians' practice locations and their knowledge and opinions on antimicrobial resistance may influence their antimicrobial prescription practices, which could impact the emergence of antimicrobial-resistant bacteria. This study used a spatial modeling approach to identify areas where veterinarians are knowledgeable about factors that impact the selection of antimicrobial resistance. In addition, we sought to identify regions with higher- and lower-than-expected response rates to our survey to aid future antimicrobial stewardship efforts. A total of 83 veterinarians who treated dogs and/or cats across 34 different Illinois counties responded to our online survey. Most of the responders (90.9%) considered that insufficient doses or duration of antibiotic treatments contribute the most to the selection of antimicrobial resistance. A high proportion of veterinarians (78.7%) attended educational programs on antimicrobial use and resistance; however, only 46.2% were knowledgeable about the current antimicrobial resistance profiles of prevalent bacteria in their area. A mean knowledge score for each county was calculated based on the responses of veterinarians to the survey questions. Local Moran's I statistic was used to identify counties with high and low knowledge scores. A high knowledge score area in the northeast region and a low knowledge score area in the southeast of Illinois were identified. Using scan statistics with a Poisson model that accounted for the estimated number of veterinarians in a county, a higher-than-expected response rate area was identified in central-east Illinois and a lower-than-expected area in the northeast. This study showed the effectiveness of using geographic analysis and spatial statistics to identify locations where future antimicrobial stewardship programs should focus.

Keywords: Illinois; USA; antimicrobial resistance; antimicrobial use; cats; dogs; small animal veterinarian; spatial epidemiology.

Figure 1

Hierarchical clustering dendrogram of small animal veterinarians’ awareness of AMR and belief of practices that contribute to the selection of antimicrobial resistance. Red color, no (0); blue color, yes (1). p_ap_guid: antimicrobial prescription guideline in the practice facility; p_amr_awrns: awareness of local antimicrobial resistance profile; p_amu_extlabl: prescription of antibiotics extra-label; p_cyto: perform cytology before treating bacterial infection; p_amu_infprev: improper infection prevention in the facilities; p_amu_broad: prescription of broad-spectrum antibiotics; p_amu_ast: empirical treatment without antimicrobial susceptibility tests; p_edu_prgm: attending educational programs on AMU and AMR after graduation; p_amr_conc: concerned about AMR bacteria; p_amr_fam: familiar with AMR topic; p_amu_inad: inadequate dose/duration of antibiotics treatments.

Figure 2

Choropleth map of the estimated number of small animal veterinarians in Illinois (a). Choropleth map of response rates of the survey on AMR and AMU practices among Illinois companion animal veterinarians (b). Isopleth map of the spatial interpolation of survey response rates by using empirical Bayesian kriging (c).

Figure 3

Cluster and outlier analysis (local Moran’s I) results of the survey response rates.

Figure 4

Spatial clusters of response rates of Illinois small animal veterinarians identified by scan statistics using a Poisson model.

Figure 5

Choropleth map of the survey knowledge scores (a). Isopleth map of knowledge scores applying spatial interpolation by using empirical Bayesian kriging (b).

Figure 6

Incremental spatial autocorrelation (Global Moran’s I statistics) analysis results of the knowledge score.

Figure 7

Cluster and outlier (local Moran’s I) analysis result of survey knowledge scores.

Figure 8

Spatial epidemiological approaches to evaluate spatial clustering of knowledge on AMU and AMR and response rates of small animal veterinarians in Illinois, USA.