Abstract
Background
Benchmarking small animal antimicrobial use is limited by lack of data on factors associated with prescribing of antimicrobial medications.
Hypothesis/Objectives
To identify factors associated with prescription of antimicrobial drugs (PAD) in dogs receiving outpatient care.
Animals
Nine thousand six hundred eighty-five dogs with 19 597 outpatient visits.
Methods
Electronic medical record data from dogs receiving outpatient care at a veterinary teaching hospital in 2022 were retrospectively analyzed. Study outcomes were receipt of any systemic PAD at the visit, and receipt of “Watch” or “Reserve” (WR) PAD using World Health Organization (WHO) Access–Watch–Reserve (AWaRe) classification. Multivariable models including independent variables related to signalment, hospital service, and pet owner sociodemographic factors were built using generalized estimating equations yielding odds ratios (ORs) with 95% CIs.
Results
Approximately 10% of outpatient visit records (1871/19 597) were associated with ≥ 1 PAD. Factors significantly associated with prescription of antimicrobials included age, male sex, intact status, hospital service visited, visit duration, diagnostics performed, lower estimated owner income, visit date, and prescriber training level. Factors most strongly affecting odds of receiving any PAD were hospital service visited (OR 0.08 [0.05-0.13] to 1.55 [1.18-2.05] for visits to services other than emergency) and having urine culture performed (OR 4.32 [3.06-6.08]). Factors most strongly affecting odds of receiving WHO-WR PAD were higher total count of discrete PAD during visit (OR 10.1 [6.41-15.8]) and undergoing computed tomography or magnetic resonance imaging during the visit (OR 3.41 [1.50-7.76]).
Conclusions and clinical importance
This evaluation of companion animal prescribing practices supports facility-level monitoring and development of normalized antimicrobial use benchmarks and One Health antimicrobial stewardship efforts.
Keywords: antimicrobial prescribing, antimicrobial stewardship, hospital antimicrobial usage, small animal, benchmarking
Introduction
Antimicrobial drugs are a critical component of veterinary medicine. While antimicrobial usage (AMU) patterns have been extensively studied in human medicine, there remains a critical gap in knowledge about the frequency of, and factors associated with, veterinary AMU in companion animals. This deficit of knowledge limits the ability to monitor and analyze small animal veterinary AMU as part of antimicrobial stewardship efforts.
Evaluating facility-specific AMU patterns is critical to inform stewardship efforts. Recent work has begun to characterize antimicrobial prescribing patterns in companion animals in both primary and tertiary care settings using various strategies to extract information from electronic medical records (EMRs).1–6 Point prevalence surveys are used to interrogate total burden and indications for AMU in companion animal facilities,7,8 which demonstrate heterogeneity in prescribing practices but rarely evaluate dog- and hospital-level factors associated with higher odds of receiving antimicrobials.9,10 This prescribing heterogeneity is likely driven by differences in regional medical needs, limited standardized recommendations11–15 for companion animal AMU, and pressures to provide optimal care to dogs balancing pet owner preferences and financial considerations.16 In addition, studies consistently report veterinary AMU identified by the World Health Organization (WHO) as important to human health,17,18 including third-generation cephalosporins and fluoroquinolones, which underscores the potential impact of veterinary antimicrobial prescribing on One Health and suggests opportunities for collaboration with human-focused antimicrobial stewardship initiatives.19–21
In this study, we leveraged data extracted using a veterinary-adapted Observational Medical Outcomes Partnership common data model to evaluate factors associated with receipt of antimicrobial prescriptions in a group of dogs receiving outpatient care at a veterinary teaching hospital. We hypothesized that we could identify specific, unique demographic and clinical factors routinely captured in the EMR that are associated with receipt of antimicrobial drug prescriptions during a veterinary outpatient visit. A second aim of our study was the identification of specific factors associated with receipt of prescription for antimicrobials of greater importance to human health based on the WHO Access–Watch–Reserve (AWaRe) antimicrobial classification system.22 Investigating associations with AMU is necessary to create benchmarks of expected use informed by monitorable dog- and facility-level factors.
Materials and methods
Study design
A retrospective cross-sectional observational study was performed using veterinary EMR data for dog visits to a veterinary teaching hospital.
Study cohort
The study cohort consisted of dogs receiving outpatient care at the Tufts University Cummings School of Veterinary Medicine Foster Hospital for Small Animals (VTH) during the 2022 calendar year. The unit of analysis was the outpatient visit, defined as a visit to any hospital service (including primary care, specialty referral, and emergency services) lasting less than 24 h and not requiring inpatient hospitalization.
Data collection and management
The VTH EMR platform StringSoft (StringSoft Inc., Manchester, NH) was queried to extract data from patient records. A database was constructed using the veterinary-adapted Observational Medical Outcomes Partnership Common Data Model framework (OMOPv5+)23 to collect standardized data from all veterinary patient visits occurring from January 1 to December 31, 2022 at all Tufts University-affiliated small animal practice settings. All data were deidentified prior to analysis.
From this database of more than 65 000 visit records, only those records generated at the VTH (the primary academic teaching hospital) were utilized for analysis. Visits requiring inpatient hospitalization and visits resulting in same-day death or euthanasia of the patient were excluded as were any records from alternate companion animal species. A flow diagram of the visit record inclusion and exclusion criteria is shown in Figure 1.
Figure 1.
Schematic representation of selection and exclusion of patient visit records for analysis.
The database was queried to identify instances of multiple records opened for a single animal on the same day; records opened less than 1 h apart were considered replicates and were combined to create a single-visit record.
Study outcomes
The primary study outcome was the receipt of one or more systemic (enteral or parenteral) prescription(s) of antimicrobial drugs (PAD) during or within 7 days following the index outpatient visit. Prescriptions of antimicrobial drugs were identified within the EMR as prescriptions that had been issued and filled at the VTH pharmacy including new PAD and refills of existing PAD. For dogs that received PAD and had multiple outpatient visits to the VTH within 1 week before the date of the PAD, each prescription receipt was linked by date to only one visit corresponding to the closest prior outpatient visit record within the capture window.
For dogs receiving at least 1 PAD, our secondary study outcome was receipt of 1 or more prescriptions for antimicrobials classified as WHO Watch or Reserve (WHO-WR) compared to receipt of WHO Access (WHO-A) drugs only. The WHO-WR drug list was adapted for this study to include drugs available exclusively for veterinary use. The list of antimicrobials prescribed and their WHO AWaRe classification are included in Table 1. For summary data reporting and WHO-WR multivariable model building, the total antibiotic count was corrected to remove replicate prescriptions for different sizes or formulations of the same drug.
Table 1.
World Health Organization (WHO) Access–Watch–Reserve (AWaRe) classification of antimicrobial drugs adapted for veterinary usage. Asterisks mark veterinary-specific drugs; AWaRe status was assigned for veterinary-specific drugs based on the classification of human/multispecies antimicrobials of the same classes (third-generation cephalosporin and fluoroquinolone drugs). Counts of unique prescriptions of antimicrobial drugs (PAD) received during outpatient visits to the Tufts University Cummings School of Veterinary Medicine Foster Hospital for Small Animals occurring in the 2022 calendar year in dog outpatients.
| WHO Access antimicrobials | PAD count (% of total PAD) in outpatient visits |
|---|---|
| Amoxicillin | 62 (3) |
| Amoxicillin/clavulanate | 942 (52) |
| Ampicillin | 1 (0.1) |
| Ampicillin/sulbactam | 30 (2) |
| Cefazolin | 1 (0.1) |
| Cephalexin | 131 (7) |
| Chloramphenicol | 20 (1) |
| Clindamycin | 116 (6) |
| Doxycycline | 219 (12) |
| Gentamicin | 18 (1) |
| Metronidazole | 257 (14) |
| Trimethoprim-sulfamethoxazole | 20 (1) |
| WHO Watch or Reserve antimicrobials | PAD count (% of total PAD) in outpatient visits |
| Azithromycin | 11 (3) |
| Cefovecin* | 18 (5) |
| Cefpodoxime | 116 (34) |
| Ciprofloxacin | 2 (1) |
| Enrofloxacin* | 157 (46) |
| Marbofloxacin* | 31 (9) |
| Meropenem | 1 (0.3) |
| Minocycline | 3 (1) |
| Orbifloxacin* | 1 (0.3) |
| Pradofloxacin* | 1 (0.3) |
| Rifampin | 1 (0.3) |
Independent variables
The independent variables that were evaluated in relation to receipt of PAD, and to receipt of PAD corresponding to WHO-WR antimicrobials, were selected a priori based on clinical knowledge, prior research, and data availability in the EMR. Variables of interest included signalment, clinical, and pet owner-related factors. These variables included dog breed, age, sex, and reproductive status, brachycephalic breed, hospital service visited, length of outpatient visit, date of outpatient visit (by quarter of the calendar year), training status of veterinarian of record (faculty or house officer), number of prior visits to the VTH, history of a visit to the VTH Emergency department during the previous 30 days, receipt of chemotherapy during the previous 30 days, receipt of surgery within the previous year, lab work performed during the outpatient visit (complete blood count, urine culture), imaging performed during the visit (radiograph, ultrasound, computed tomography [CT] or magnetic resonance imaging [MRI], or point-of-care ultrasound), owner billing zip code, visit invoice amount, total number of pets within the household, and total count of PAD received during a single visit. Approximately 2% of antibiotic prescriptions were not linked with a specific provider training status; these were grouped with faculty to avoid including prescribers with more advanced training in the house officer prescriber categories. Median household income was estimated by linking the billing zip code with the United States Census Bureau’s American Community Survey 5-year database for 2022.24
Outside antimicrobial prescription survey
A multiple-choice survey was constructed to estimate the frequency with which VTH clinicians provided dog owners with paper or electronic antimicrobial prescriptions not intended to be filled by the VTH pharmacy. The survey was offered to all VTH clinicians; participation was voluntary and all responses were anonymous. Survey questions included identification of the respondent’s hospital service, training status, and reasons for providing an outside prescription. Surveys were hosted using the Qualtrics XM platform (Qualtrics, Provo, UT).
Statistical analysis
Data were evaluated for missingness and complete case analyses were performed. Visit records were assessed for clustering by individual animal. Dog demographic data and the amount and type of antimicrobials prescribed to outpatients were reported descriptively.
For the primary data analysis, univariable logistic regression models were fitted to individually test the association of each independent variable with the dichotomous outcome of receipt of 0, or at least 1 PAD during or immediately following an outpatient visit. Independent variables were assessed for collinearity using the variance inflation factor to guide variable selection in final multivariable model fitting. All variables with a P-value < .2 during univariable screening were retained as candidates for the multivariable logistic regression analysis.
Models were fit using generalized estimating equations (GEE) to account for clustering due to dogs with multiple outpatient visits during the study period. The GEE models were built using an exchangeable correlation structure, logit link function, and binomial distribution. Model refinement was performed using backward variable selection based on quasi-likelihood under the independence model criterion to select the final multivariable model. It was hypothesized that veterinary training status (faculty vs house officer) could have different effects based on the date of the hospital visit relative to the academic year; therefore, an interaction term was included between these variables in final multivariable adjusted regression models.
Exploration of the secondary study aim was performed by analyzing the subset of visit records associated with receipt of PAD using the dichotomous outcome of receipt of 0, or at least 1 WHO-WR PAD.
Statistical software Excel (Version 2104, Microsoft, Redmond WA) and R Studio (Version 2023.03, R Foundation, Indianapolis IN) were used for all data analyses. A P-value of .05 was considered significant for all final multivariable model analyses.
Ethical approval
Analyses performed retrospectively on previously collected de-identified veterinary patient data did not require Institutional Animal Care and Use (IACUC) or Institutional Review Board (IRB) oversight. The outside antimicrobial prescribing survey received IRB approval as an exempt protocol.
Results
Study cohort
A total of 19 597 visit records from dog outpatients (generated from 9685 unique dogs) were retained for analysis.
Demographic characteristics of outpatient visit records
The majority of dog outpatients had multiple visits during 2022. The total number of outpatient records for a unique dog ranged from 1 to 39 in total (median 3 visits); 4946 records (25%) were for dogs with only 1 visit during the study period.
The median age at the time of visit was 7.7 years. Approximately 48% of records were from females, and 82% were from dogs that were spayed or neutered. Approximately 200 different dog breeds were represented, with the most frequently occurring breeds being Labrador Retriever (2452, 13%), mixed-breed (1219, 6%), Golden Retriever (1139, 6%), German Shepherd (913, 5%), and American Pit Bull Terrier (732, 4%).
Antimicrobial drug prescriptions
A total of 1871 records (9.5%) were associated with PAD. Among visits in which at least 1 antimicrobial drug was prescribed, 1599 records (86%) were associated with receipt of a single PAD, 246 (13%) were associated with receipt of 2 PAD, and 26 (1%) were associated with receipt of 3 or more PAD during a single outpatient visit. For this group of visits associated with receipt of at least 1 PAD, 1551 visits (83%) were associated with receipt of WHO-A antimicrobial(s) only, 245 (13%) were associated with receipt of WHO-WR antimicrobial(s) only, and 75 (4%) were associated with receipt of both WHO-A and WHO-WR antimicrobials within a single visit.
During the study period, dog outpatients received a total of 1817 unique PAD for WHO-A antimicrobials and 342 unique PAD for WHO-WR antimicrobials. The most commonly prescribed WHO-A drugs were potentiated aminopenicillins (942, 52% of WHO-A prescriptions), metronidazole (257, 14% of WHO-A prescriptions), and doxycycline (219, 12% of WHO-A prescriptions) and the most commonly prescribed WHO-WR drugs were fluoroquinolones (192, 56% of WHO-WR prescriptions) and third-generation cephalosporins (134, 39% of WHO-WR prescriptions). Total PAD counts are summarized in Table 1. Approximately 13% of total PAD were associated with receipt of 2 or more prescriptions for the same type of medication during a single-visit capture period, considered likely related to weight-based drug dosing. After manual correction to remove same-drug PAD replicates, 125 visit records demonstrated polypharmacy, with 30 different drug combinations. The most common PAD combinations were (count, percentage of visits demonstrating polypharmacy): potentiated aminopenicillin plus fluoroquinolone (31, 25%); fluoroquinolone plus lincosamide (21; 17%); and ampicillin-sulbactam plus amoxicillin-clavulanic acid (22, 18%).
Factors associated with PAD in all outpatient visits
Compared with visit records in which no antimicrobials were prescribed, outpatient visits resulting in receipt of PAD were in older (median age 7.9 vs 7.6 years), intact dogs (21% vs 18%), and in those visiting the Emergency/Critical Care service. In the group receiving PAD during the outpatient visit, the estimated median household income was slightly lower, the median visit invoice was higher, the visit duration was longer, and dogs had more diagnostics including urine culture or imaging performed (Table 2).
Table 2.
Characteristics of all dog outpatient visit records, stratified by the primary outcome of receipt of ≥ 1 prescription(s) of antimicrobial drugs (PAD) during or within 7 days following the outpatient visit. Data reported as n (%) or median (interquartile range).
| No PAD | PAD | P-value | |
|---|---|---|---|
| (n = 17 726) | (n = 1 871) | ||
| Dog and owner demographics | |||
| Age (years) | 7.6 (3.9, 10.9) | 7.9 (4.0, 11.0) | .20 |
| Spayed or neutered reproductive status | 14 521 (82%) | 1 472 (79%) | <.001 |
| Female sex | 8 613 (49%) | 915 (49%) | .80 |
| Brachycephalic breed | 2 930 (17%) | 293 (16%) | .33 |
| Total number of household pets | 1 (1, 2) | 1 (1, 3) | .08 |
| Median household income | $108 K ($86 K, $134 K) | $104 K ($82 K, $133 K) | <.001 |
| Visit invoice | $250 ($83, $458) | $433 ($243, $693) | <.001 |
| Dog medical history and diagnostics | |||
| Total number of prior visits to the VTH | 3 (1, 9) | 2 (0, 6) | <.001 |
| Visited the VTH ER in prior 30 days | 2 824 (16%) | 286 (15%) | .47 |
| Received chemotherapy in prior 30 days | 653 (3.7%) | 66 (3.5%) | .73 |
| Received surgery in prior 365 days | 777 (4.4%) | 44 (2.4%) | <.001 |
| Complete blood count performed | 3 735 (21%) | 365 (20%) | .11 |
| Urine culture performed | 187 (1.1%) | 78 (4.2%) | <.001 |
| Radiograph performed | 1 643 (9.3%) | 230 (12%) | <.001 |
| Ultrasound performed | 702 (4.0%) | 76 (4.1%) | .83 |
| CT or MRI performed | 145 (0.8%) | 17 (0.9%) | .68 |
| Point-of-care ultrasound performed | 162 (0.9%) | 31 (1.7%) | .002 |
| Hospital and service | |||
| Hospital service visited | |||
| Emergency/Critical Care | 5 267 (30%) | 1 052 (56%) | ref. |
| Cardiology | 863 (4.9%) | 27 (1.4%) | <.001 |
| Clinical Trials | 595 (3.4%) | 149 (8.0%) | .02 |
| Dermatology | 917 (5.2%) | 71 (3.8%) | <.001 |
| Internal Medicine | 2 803 (16%) | 194 (10%) | <.001 |
| Neurology | 1 382 (7.8%) | 34 (1.8%) | <.001 |
| Oncology | 1 602 (9.0%) | 122 (6.5%) | <.001 |
| Ophthalmology | 1 780 (10%) | 143 (7.6%) | <.001 |
| Surgery | 1 763 (9.9%) | 72 (3.8%) | <.001 |
| Other | 754 (4.3%) | 7 (0.4%) | <.001 |
| Veterinarian training level | |||
| Faculty or not recorded | 7 277 (41%) | 704 (38%) | ref. |
| House officer (intern or resident) | 10 449 (59%) | 1167 (62%) | .005 |
| Quarter of the year | |||
| Q1: January-March | 4 468 (25%) | 537 (29%) | <.001 |
| Q2: April-June | 4 499 (25%) | 568 (30%) | <.001 |
| Q3: July-September | 4 525 (26%) | 326 (17%) | ref. |
| Q4: October-December | 4 234 (24%) | 440 (24%) | <.001 |
| Visit duration (h) | 1.9 (1.0, 3.9) | 3.4 (1.9, 6.4) | <.001 |
Abbreviations: CT, computed tomography; ER, emergency room; MRI, magnetic resonance imaging; VTH, Tufts University Cummings School of Veterinary Medicine Foster Hospital for Small Animals.
A univariable screen was performed on 22 candidate variables (summarized in Table S1). Complete data were available for 19 418 records (<1% data missingness overall). In the final multivariable regression model, the following variables remained significantly associated with the outcome of receipt of PAD: older dog age, intact reproductive status, hospital service visited, quarter of the year, interaction of veterinarian training status and quarter of the year, visit duration, radiograph(s) performed during visit, urine culture performed during visit, total number of prior visits to the VTH, and median household income (Table 3).
Table 3.
Factors significantly associated with receipt of ≥ 1 prescription(s) of antimicrobial drugs (PAD) in dog outpatient visits. Adjusted multivariable model fitted using generalized estimating equations and backward variable selection based on quasi-likelihood information criterion (QIC). Data reported are odds ratios (OR) and 95% confidence interval (CI). *For the quarter of the year and the interaction of training level and quarter of the year, statistically significant OR for individual levels are noted with an asterisk. † denotes values in which the upper bound of the 95% CI is < 1.00 but rounds to 1.00 at two decimal places.
| Variable | OR (95% CI) | P-value |
|---|---|---|
| Age (years) | 1.02 (1.01, 1.03) | .004 |
| Spayed or neutered reproductive status | 0.87 (0.75, 1.00†) | .045 |
| Hospital service visited | ||
| Emergency/Critical Care (referent) | — | |
| Cardiology | 0.15 (0.10, 0.23) | <.001 |
| Clinical Trials | 1.55 (1.18, 2.05) | .003 |
| Dermatology | 0.41 (0.31, 0.55) | <.001 |
| Internal Medicine | 0.27 (0.22, 0.34) | <.001 |
| Neurology | 0.08 (0.05, 0.13) | <.001 |
| Oncology | 0.35 (0.27, 0.45) | <.001 |
| Ophthalmology | 0.46 (0.37, 0.57) | <.001 |
| Surgery | 0.21 (0.16, 0.27) | <.001 |
| Other | 0.05 (0.02, 0.10) | <.001 |
| Quarter of the year | ||
| Q1: January-March* | 2.33 (1.79, 3.03) | <.001 |
| Q2: April-June* | 2.00 (1.54, 2.59) | <.001 |
| Q3: July-September (referent) | — | |
| Q4: October-December | 1.19 (0.88, 1.60) | .26 |
| Veterinarian training level* quarter of the year | ||
| Faculty (referent) | — | |
| House officer in Q1* | 0.53 (0.48, 0.61) | — |
| House officer in Q2* | 0.82 (0.66, 1.00†) | — |
| House officer in Q3 | 0.89 (0.68, 1.18) | — |
| House officer in Q4 | 1.22 (0.93, 1.60) | — |
| Urine culture performed during visit | 4.32 (3.06, 6.08) | <.001 |
| Radiograph(s) performed during visit | 0.74 (0.62, 0.88) | .001 |
| Median household income (per $10 K) | 0.98 (0.97, 1.00†) | .011 |
| Total number of prior visits to the VTH | 0.99 (0.98, 1.00†) | .023 |
| Visit duration (h) | 1.08 (1.06, 1.09) | <.001 |
Abbreviation: VTH, Tufts University Cummings School of Veterinary Medicine Foster Hospital for Small Animals.
Multiple factors were associated with significantly lower odds of receiving PAD during outpatient visits. Visits to hospital services other than the Clinical Trials service had markedly lower odds of receiving PAD than visits to the Emergency/Critical Care service (for example: Neurology odds ratio (OR) 0.08 [0.05-0.13]; Internal Medicine OR 0.27 [0.22-0.34]; Ophthalmology OR 0.46 [0.37-0.57]; full service list in Table 3). The total number of outpatient visit records receiving PAD by service type is summarized in Table S2. Visits of spayed or neutered dogs had lower odds of receiving PAD than visits of reproductively intact dogs (OR 0.87 [0.75-1.00]). Visits in which a radiograph was performed had lower odds of receiving PAD than visits in which radiographs were not ordered (OR 0.74 [0.62-0.88]). Visits occurring during quarters 1 and 2 in which the primary veterinarian of record was a house officer had lower odds of receiving PAD than those occurring with a faculty member (OR 0.53 [0.48-0.61] and 0.82 [0.66-1.00], respectively). The total number of outpatient visit records receiving PAD by prescriber training level and quarter of the year is summarized in Table S3. Each unit increase in the total number of prior VTH visits was associated with 1% lower odds of receiving PAD (OR 0.99 [0.98-1.00]).
Several factors were identified with significantly higher odds of receiving PAD. Visits to the Clinical Trials service had higher odds of receiving PAD than visits to the Emergency/Critical Care service (OR 1.55 [1.18-2.05]. During the study period, 98% of the Clinical Trials service visits were for oncology trials (separate from the Oncology service); of 149 Clinical Trials visits receiving PAD, 109 (73%) were associated with receipt of metronidazole. Visits occurring during quarters 1 and 2 (across all provider training levels) had higher odds of receiving PAD compared to visits occurring during quarter 3 (OR 2.33 [1.79-3.03] and 2.00 [1.54-2.59], respectively). Visits in which a urine culture was performed had markedly higher odds of receiving PAD than visits in which a urine culture was not ordered (OR 4.32 [3.06-6.08]). Each additional year of age was associated with 2% higher odds of receiving PAD (OR 1.02 [1.01-1.03]). Each increasing hour of visit duration was associated with 8% higher odds of receiving PAD (OR 1.08 [1.06-1.09]).
Factors associated with receipt of WHO-WR antimicrobials in outpatient visits associated with receipt of ≥ 1 PAD
This analysis was performed on the subset of outpatient visit records associated with receipt of at least 1 PAD during or immediately following the outpatient visit. Compared with visit records in which only WHO-A antimicrobials were prescribed, visits associated with receipt of WHO-WR PAD were in older, male dogs, and those visiting the Neurology, Dermatology, or Internal Medicine services. Dogs receiving WHO-WR PAD had a history of prior visits to the VTH and were more likely to undergo diagnostic imaging during the visit. The total count of PAD received during a single visit was higher in dogs receiving WHO-WR antimicrobials (Table 4).
Table 4.
Characteristics of dog outpatient visit records associated with receipt of ≥ 1 prescription(s) of antimicrobial drugs (PAD), stratified by the secondary outcome of receipt of ≥ 1 World Health Organization Watch or Reserve (WHO-WR) PAD. Data reported as n (%) or median (interquartile range). Count of PAD reflects corrected data, which eliminates replicates of the same antimicrobial drug (in alternate formulations) prescribed during a single-visit capture window.
| No WHO-WR PAD | WHO-WR PAD | P-value | |
|---|---|---|---|
| (n = 1 551) | (n = 320) | ||
| Dog and owner demographics | |||
| Age (years) | 7.5 (3.8, 10.8) | 9.0 (4.8, 12.0) | <.001 |
| Spayed or neutered reproductive status | 1 218 (79%) | 254 (79%) | .74 |
| Female sex | 779 (50%) | 136 (43%) | .01 |
| Median household income | $104 K ($83 K, $133 K) | $105 K ($76 K, $133 K) | .66 |
| Visit invoice | $437 ($249, $694) | $409 ($219, $681) | .05 |
| Dog medical history and diagnostics | |||
| Urine culture performed | 63 (4.1%) | 15 (4.7%) | .61 |
| Radiograph(s) performed | 189 (12%) | 41 (13%) | .76 |
| Ultrasound performed | 53 (3.4%) | 23 (7.2%) | .002 |
| CT or MRI performed | 9 (0.6%) | 8 (2.5%) | .003 |
| Point-of-care ultrasound performed | 25 (1.6%) | 6 (1.9%) | .74 |
| Hospital and service factors | |||
| Hospital service visited | |||
| Emergency/Critical Care | 888 (57%) | 164 (51%) | ref. |
| Cardiology | 24 (1.5%) | 3 (0.9%) | .53 |
| Clinical Trials | 128 (8.3%) | 21 (6.6%) | .64 |
| Dermatology | 48 (3.1%) | 23 (7.2%) | <.001 |
| Internal Medicine | 147 (9.5%) | 47 (15%) | .003 |
| Neurology | 19 (1.2%) | 15 (4.7%) | <.001 |
| Oncology | 107 (6.9%) | 15 (4.7%) | .34 |
| Ophthalmology | 125 (8.1%) | 18 (5.6%) | .35 |
| Surgery | 61 (3.9%) | 11 (3.4%) | .94 |
| Other | 4 (0.3%) | 4 (1.3%) | .07 |
| Veterinarian training level | |||
| Faculty or status not recorded | 542 (35%) | 162 (51%) | ref. |
| Intern | 561 (36%) | 75 (23%) | <.001 |
| Resident | 448 (29%) | 83 (26%) | .001 |
| Quarter of the year | |||
| Q1: January-March | 430 (28%) | 107 (33%) | .44 |
| Q2: April-June | 476 (31%) | 92 (29%) | .54 |
| Q3: July-September | 268 (17%) | 58 (18%) | ref. |
| Q4: October-December | 377 (24%) | 63 (20%) | .19 |
| Visit duration (h) | 3.5 (1.9, 6.4) | 3.1 (1.8, 6.1) | .02 |
| Antimicrobial prescription factors | |||
| Count of PAD prescribed during visit | <.001 | ||
| 1 | 1 506 (97%) | 240 (75%) | |
| 2 | 44 (2.8%) | 77 (24%) | |
| 3 | 1 (<0.1%) | 3 (0.9%) | |
Abbreviations: CT, computed tomography; MRI, magnetic resonance imaging.
A univariable screen was performed on 22 candidate variables (summarized in Table S4). In the final multivariable regression model, the following variables remained significantly associated with the outcome of receipt of WHO-WR antimicrobials: older dog age, male sex, hospital service visited, veterinarian training status, CT or MRI performed during visit, and total count of antimicrobials prescribed during the outpatient visit (Table 5).
Table 5.
Factors significantly associated with receipt of ≥ 1 World Health Organization Watch or Reserve (WHO-WR) drug prescription (PAD) in dog outpatient visits in which at least 1 antimicrobial drug was prescribed. Adjusted multivariable model fitted using generalized estimating equations and backward variable selection based on quasi-likelihood information criterion (QIC). Data reported are odds ratios (OR) and 95% confidence interval (CI). *For the hospital service category, statistically significant OR for individual services are noted with an asterisk.
| Variable | OR (95% CI) | P-value |
|---|---|---|
| Age (years) | 1.06 (1.03, 1.10) | <.001 |
| Male sex | 1.38 (1.03, 1.84) | .03 |
| Hospital service visited* | ||
| Emergency/Critical Care (referent) | — | |
| Cardiology | 0.34 (0.09, 1.20) | .09 |
| Clinical Trials* | 0.38 (0.19, 0.74) | .005 |
| Dermatology | 1.45 (0.75, 2.80) | .27 |
| Internal Medicine | 1.21 (0.75, 1.95) | .44 |
| Neurology | 2.02 (0.90, 4.52) | .09 |
| Oncology | 0.55 (0.25, 1.16) | .12 |
| Ophthalmology* | 0.41 (0.21, 0.79) | .008 |
| Surgery | 0.58 (0.26, 1.29) | .18 |
| Other | 1.19 (0.28, 5.07) | .70 |
| Veterinarian training level | ||
| Faculty (referent) | — | |
| Intern | 0.34 (0.23, 0.51) | <.001 |
| Resident | 0.44 (0.30, 0.64) | <.001 |
| Count of antimicrobials prescribed during visit | 10.1 (6.41, 15.8) | <.001 |
| CT or MRI performed during visit | 3.41 (1.50, 7.76) | .003 |
Abbreviations: CT, computed tomography; MRI, magnetic resonance imaging.
For records associated with receipt of PAD, visits in which the primary veterinarian of record was an intern or resident had lower odds of receiving a WHO-WR PAD compared to visits in which the primary veterinarian was faculty (OR 0.34 [0.23-0.51] and 0.44 [0.30-0.64], respectively). Visits to the Ophthalmology and Clinical Trials services were associated with lower odds of receiving a WHO-WR PAD compared to records from visits to the Emergency/Critical Care service (OR 0.41 [0.21-0.79] and 0.38 [0.19-0.74], respectively). In contrast, visits from male dogs had higher odds of receiving WHO-WR antimicrobials (OR 1.38 [1.03-1.84]), and each additional year of age was associated with 6% higher odds of receiving WHO-WR PAD (OR 1.06 [1.03-1.10]). Visits in which a CT or MRI was performed had higher odds of receiving WHO-WR drugs [OR 3.41 [1.50-7.76]). Each unit increase in the total count of antimicrobial drugs prescribed during an outpatient visit was associated with markedly higher odds of receiving at least 1 WHO-WR PAD (OR 10.1 [6.41-15.8]).
Evaluation of antimicrobial drug prescribing to dog outpatients from a non-VTH pharmacy source
A brief survey was shared with VTH clinicians to help better characterize the frequency of and reasons for prescribing antimicrobial medications to be supplied by a source other than the VTH pharmacy (“outside” PAD) specifically for dog outpatients. The survey was completed by 48 VTH clinicians, including 18 faculty veterinarians, 21 residents, and 9 interns. Of these, 26% reported never providing “outside” PAD, 60% reported providing “outside” PAD comprising up to 10% of their total antimicrobial prescribing for dog outpatients, and the remaining 14% reported providing “outside” PAD more frequently, comprising > 10% of their total. Survey respondents reported multiple reasons for provision of “outside” PAD to outpatients, including delayed decision for antimicrobial treatment occurring after the outpatient visit had concluded, recurrence of a chronic infectious medical issue outside of a scheduled visit, preference for the prescription to be filled by the primary care veterinarian, unavailability of the drug type or formulation at the VTH pharmacy, and drug cost. In dog outpatients, delayed decision for antimicrobial prescription was the most commonly reported reason for providing an “outside” PAD, followed by preference for the drug to be provided by the primary care veterinarian and lack of availability of the preferred drug formulation at the VTH pharmacy.
Discussion
This study provides an evaluation of antimicrobial prescribing characteristics in dogs receiving outpatient care at a veterinary teaching hospital. These analyses identified multiple demographic and clinical factors that were significantly associated with receipt of any PAD and with receipt of WHO-WR PAD. Overall, the factors significantly associated with the strongest effects on the odds of receiving PAD included the hospital service visited, the dog’s sex and reproductive status, undergoing diagnostic imaging during the visit, and having a urine culture performed during the visit.
In our study, spayed or neutered dogs (independent of sex) were less likely to receive PAD. Among those receiving PAD, visits of male dogs were more likely to be associated with receipt of WHO-WR PAD than visits of female dogs; this might be due to differential risk of urogenital tract infections leading to more first-line AMU in female dogs compared to male dogs. Our results align with a prior UK study in which outpatient male dogs had increased odds of receiving PAD compared with females, and in which neutered dogs and cats had reduced odds of antimicrobial prescription.9 Observed differences in receipt of PAD and the classification of PAD tier associated with sex and reproductive status might be attributable to differential disease risk in males compared to females and in dogs with intact reproductive organs, and could be clarified by collecting data on clinical diagnosis and reason for antimicrobial use.
Older age had a significant, though small, effect on both the odds of receiving PAD and receiving WHO-WR PAD, consistent with prior work demonstrating decreased likelihood of PAD in younger dogs.1 Older age has been associated with prevalence of bacteria in multiple body systems in dogs, and can be associated with increased risk of bacterial infections such as urinary tract infections.25,26 Increasing age can increase the likelihood of comorbid conditions (endocrine dysfunction, neoplasia) that heighten the real or perceived risk of developing a secondary bacterial infection.
Estimated pet owner median household income was hypothesized to potentially affect pet owners’ care plan choices. The associations between owner financial status and veterinary care have been explored to a limited degree,27–29 but large-scale evaluation of real-world veterinary patent data is lacking. Higher estimated median household income was associated with lower odds of receiving PAD. This is generally consistent with trends seen in human medicine, in which higher rates of antimicrobial prescription are associated with lower-income patients.30 It is possible that higher income might be associated with increased likelihood of performing diagnostic testing rather than choosing lower-cost treatment options including empiric administration of antimicrobials.
Records with a higher total number of visits to the VTH were associated with slightly lower odds of receiving PAD in all outpatients. This finding could suggest that dogs with multiple visits to the VTH are more likely to be seeking routine care from the facility or through unscheduled visits to the emergency department (common in human health care31), or owner compliance with recommended follow-up appointments.
Several variables related to receipt of diagnostic lab testing and imaging were assessed to augment characterization of the veterinary medical history and disease state at the time of the outpatient visit. Having a urine culture performed during the outpatient visit was strongly associated with receipt of PAD. However, this factor was not significantly associated with receipt of WHO-WR PAD compared to WHO-A PAD suggesting the use of first-line antimicrobials for lower urinary tract disease. Visits in which diagnostic imaging was performed were variably associated with receipt of PAD, depending on the type of imaging performed. Having a radiograph performed was associated with decreased odds of PAD; having a CT or MRI performed was associated with increased odds of receiving WHO-WR PAD. These findings suggest that the type of imaging performed in combination with other factors reflect the types of abnormalities being evaluated as well as the relative likelihood that antimicrobial therapy could be warranted.
In our study, hospital service, veterinarian training level, and date of the outpatient visit were all variably associated with receipt of PAD. Visits to a service other than Emergency/Critical Care were typically associated with much lower odds of receiving PAD. The VTH emergency departments provide outpatient care to dogs with a wide variety of clinical presentations. Compared to specialist services, the Emergency/Critical Care service might be more likely to see cases that would benefit from antimicrobial treatment, such as injuries, animal bite wounds, and acute development or exacerbation of systemic infectious diseases. A study of 3 veterinary teaching hospitals showed increased prescriptions for emergency patients compared to specialty services in 1 hospital, similar rates for another, and higher prescription rates for specialty services than emergency in the third, suggesting the need for additional studies clarifying factors associated with antimicrobial prescription in veterinary emergency centers.3 In contrast, visits to the Clinical Trials service, which at VTH are comprised almost exclusively of oncology trials (separate from the Oncology service) were associated with higher odds of receiving PAD. In these Clinical Trial visits, the most common PAD was for metronidazole, which might have been prescribed per trial protocols as a prophylactic against chemotherapy-induced diarrhea.32 Differences in the odds of receiving WHO-WR PAD during visits to services other than Emergency/Critical Care reflect both the varied types of clinical conditions treated by each service as well as the different prescribing preferences for each hospital service.
Longer visit duration was associated with slightly greater odds of receiving PAD; however, this association was not observed for receipt of WHO-WR PAD. Longer visits might be associated with complex medical cases that require more time to perform diagnostic investigation or to discuss adjunctive therapeutic management strategies with pet owners.
Veterinarian training level, quarter of the year, and the interaction of these 2 variables were significantly associated with receipt of PAD during an outpatient visit. The nadir in the rate of antimicrobial prescribing occurred during quarter 3 of the year; in comparison, at alternate times of year, outpatient visits were associated with higher odds of receiving PAD. This is in contrast to previous studies that found increased AMU during the summer months.33,34 Discrepancies might be due to climate-related differences in which vector-borne infections are most prevalent. In addition, interaction between the effects of training level and the visit date showed that visits in which the primary veterinarian of record was a house officer (intern or resident) had lower odds of receiving PAD in the first half of the calendar year. The association of lower odds of PAD during house officer-led outpatient visits in the later portion of the academic year (quarters 1 and 2) might correlate with the nature of training programs and suggest increased house officer familiarity with antimicrobial prescribing guidelines. Across all time points, records in which the primary veterinarian of record was a house officer were associated with lower odds of receiving WHO-WR PAD, which could indicate updated stewardship training practices for house officers, or reflect distribution of cases with greater illness severity to more senior members of the care team.
Records with a larger number of PAD prescribed during a single outpatient visit were associated with markedly higher odds of receiving WHO-WR PAD. Receipt of multiple PAD suggests presence of a more complex medical condition such as a mixed-population bacterial infection. This might also reflect practitioner drug familiarity or preferences for combinations of antimicrobials based on prior clinical experience.
Addressing concerns about uncaptured non-VTH pharmacy prescribing, the VTH clinician survey revealed that “outside” PAD constitute a minority of prescribing for dogs receiving outpatient care. The most frequently reported reason for providing an “outside” PAD was a delayed decision to prescribe an antimicrobial, consistent with awaiting receipt of laboratory results or persistence of clinical abnormalities before providing PAD, reflecting good antimicrobial stewardship practice.
This study utilizes the veterinary-adapted OMOPv5+ common data model for collection of data from the veterinary EMR, demonstrating the feasibility and utility of performing standardized collection of veterinary EMR data. The use of the OMOPv5+ model, which standardizes data collection to facilitate multicenter data analysis, has the potential to foster collaborative research between veterinary centers and with human healthcare facilities.23 The large number of cases available for analysis enhanced our ability to perform thorough investigation of multiple factors of interest, including analyses of factors associated with receipt of WHO-WR PAD.
One limitation of our study is that outpatient care provided at an academic tertiary care facility can differ from that at privately or commercially owned veterinary practices, and thus the findings of this study might have limited generalizability to other veterinary care settings. In addition, clinical diagnosis data were not captured in the study database and often are not reported in a standardized fashion in veterinary EMR. International Classification of Disease (ICD) codes or other standardized methods for disease diagnosis are not widely used in veterinary healthcare. Consequently, this database did not contain data sufficient to assess the appropriateness of antimicrobial treatment in our study. Median household income was estimated retrospectively based on billing zip code and thus does not perfectly reflect pet owners’ financial resources. Although receipt of diagnostic testing and imaging was often significantly associated with receipt of PAD, these factors more accurately reflect practitioners’ clinical impressions of patient status, as they do not include results of the performed tests. In addition, veterinarian training level must be interpreted conservatively, as a single prescriber’s training status might not accurately reflect the input of a multi-doctor care team. Finally, this study used the WHO AWaRe system which is designed for use in humans and thus required the addition of specific veterinary drug formulations. Although veterinary studies reference the WHO importance designations,17,18 this system was developed largely considering antimicrobial use in production rather than companion animals, and does not always clearly indicate drugs recommended for treatment. The MIA system18 developed out of the CIA framework is similarly challenging to apply, with multiple “importance” subclassifications and limited reflection of veterinary extra-label usage frequency of antimicrobials classified as “authorized for use in humans only,” and furthermore was unavailable as a reference for clinicians or policy makers in 2022 when the data were generated. The AWaRe system was intentionally selected as an available guideline allowing investigation of accepted first-line drugs compared to those of greater medical importance. While this classification system can be easier to interpret and is translatable to One Health colleagues, we acknowledge it is not designed to reflect veterinary influence on antimicrobial resistance.
Our study offers novel characterization of risk factors associated with AMU in dogs receiving outpatient care, and provides insight into factors that could be monitored or modified to aid in ongoing assessment of AMU and refinement of prescribing practices in veterinary healthcare settings. These results greatly enhance our present understanding of both previously identified and novel factors significantly associated with receipt of PAD in dogs receiving outpatient care, including hospital service visited, visit duration, diagnostics performed during the visit, veterinarian training level, and the dog’s age and reproductive status. Furthermore, this study provides both characterization of and identification of factors associated with receipt of higher-tier PAD during an outpatient visit. These findings add to the limited body of knowledge of small animal veterinary AMU and expand our understanding of clinical- and facility-level factors associated with receipt of PAD.
This knowledge will support ongoing development of antimicrobial stewardship metrics to develop risk adjustment strategies to benchmark AMU and establish useful comparisons between care settings with varied patient and facility characteristics. The factors significantly associated with AMU in our study cohort are commonly recorded in most veterinary EMRs. Our study demonstrates that utilizing a data standardization method such as the OMOPv5+ common data model is feasible and could allow for establishment of veterinary antimicrobial benchmarking and comparison of antimicrobial prescribing practices both longitudinally and between different practice settings by adjusting for readily measurable factors that can affect the odds of a patient receiving an antimicrobial prescription. These results can be used in the future to refine local antimicrobial prescribing practices, explore possibilities of developing tools for benchmarking AMU in veterinary patients, and support One Health initiatives for antimicrobial stewardship.
Supplementary Material
Abbreviations
- AMU
antimicrobial usage
- CT
computed tomography
- EMR
electronic medical record
- GEE
generalized estimating equations
- MRI
magnetic resonance imaging
- PAD
prescription of antimicrobial drugs
- OR
odds ratio
- VTH
Tufts University Cummings School of Veterinary Medicine Foster Hospital for Small Animals
- WHO
World Health Organization
- AWaRe
Access–Watch–Reserve
- A
Access
- WR
Watch or Reserve
Contributor Information
Kristen Thane, Department of Clinical Sciences, Cummings School of Veterinary Medicine, Tufts University, North Grafton, MA, United States.
Manlik Kwong, Tufts Clinical and Translational Science Institute, Tufts Medical Center Institute for Clinical Research and Health Policy Studies, Tufts University, Boston, MA, United States.
Ian M DeStefano, Department of Clinical Sciences, Cummings School of Veterinary Medicine, Tufts University, North Grafton, MA, United States.
Benjamin Sweigart, Brigham and Women’s Hospital, Boston, MA, United States.
Kirthana R Beaulac, Department of Pharmacy, Emerson Hospital, Concord, MA, United States.
Shira I Doron, Division of Geographic Medicine and Infectious Diseases, School of Medicine, Tufts Medical Center, Tufts University, Boston, MA, United States.
Claire L Fellman, Department of Clinical Sciences, Cummings School of Veterinary Medicine, Tufts University, North Grafton, MA, United States.
Author contributions
Kristen Thane (Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Software, Validation, Visualization, Writing—original draft, Writing—review & editing), Manlik Kwong (Data curation, Software, Writing—review & editing), Ian DeStefano (Conceptualization, Writing—review & editing), Benjamin Sweigart (Formal analysis, Software, Writing—review & editing), Kirthana Beaulac (Conceptualization, Writing—review & editing), Shira Doron (Conceptualization, Writing—review & editing), and Claire Fellman (Conceptualization, Funding acquisition, Resources, Supervision, Writing—review & editing)
Conflicts of interest
C.L.F. has served as a consultant for Zoetis. The other authors declare no conflicts of interest.
Funding
U.S. Department of Health and Human Services, National Institutes of Health, National Center for Advancing Translational Sciences, T32TR004418.
Off-label antimicrobial declaration
Multiple categories of off-label drug use are reported in this manuscript, as specific doses and durations of treatment were not categorized as part of this analysis. This report details off-label use of the following drugs: azithromycin, ciprofloxacin, doxycycline, minocycline, pradofloxacin, and rifampin. All antimicrobial prescribing reported in this manuscript was performed by licensed veterinarians and, where off-label, in accordance with ELDU practices under AMDUCA.
Institutional animal care and use committee or other approval declaration
Analyses performed retrospectively on previously collected de-identified veterinary patient data did not require Institutional Animal Care and Use or Institutional Review Board (IRB) oversight (IRB review indicated that the clinician survey was classified as exempt).
Human ethics approval declaration
The authors declare human ethics approval was not needed for retrospective analysis of de-identified veterinary data. The clinician survey received IRB approval as an exempt protocol.
References
- 1. Hur BA, Hardefeldt LY, Verspoor KM, Baldwin T, Gilkerson JR. Describing the antimicrobial usage patterns of companion animal veterinary practices; free text analysis of more than 4.4 million consultation records. PLoS One. 2020;15:e0230049. 10.1371/journal.pone.0230049 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Buckland EL, O'Neill D, Summers J, et al. Characterisation of antimicrobial usage in cats and dogs attending UK primary care companion animal veterinary practices. Vet Rec. 2016;179:489. 10.1136/vr.103830 [DOI] [PubMed] [Google Scholar]
- 3. Goggs R, Menard JM, Altier C, et al. Patterns of antimicrobial drug use in veterinary primary care and specialty practice: a 6-year multi-institution study. J Vet Intern Med. 2021;35:1496–1508. 10.1111/jvim.16136 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Singleton DA, Sanchez-Vizcaino F, Dawson S, et al. Patterns of antimicrobial agent prescription in a sentinel population of canine and feline veterinary practices in the United Kingdom. Vet J. 2017;224:18–24. 10.1016/j.tvjl.2017.03.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Granick JL, Beaudoin AL, Nielsen LK, Bollig ER. Measurement of antibiotic use in cats and dogs presenting to US primary care and referral practices provides insights for antimicrobial stewardship. J Am Vet Med Assoc. 2025;263:640–649. 10.2460/javma.24.11.0716 [DOI] [PubMed] [Google Scholar]
- 6. Weese JS, Mosher M, Low R, et al. Evaluation of antimicrobial purchasing by companion animal veterinary facilities in Canada, the United Kingdom, and the United States of America (2019-2021). J Vet Intern Med. 2024;38:1520–1534. 10.1111/jvim.17068 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Beaudoin AL, Bollig ER, Burgess BA, et al. Prevalence of antibiotic use for dogs and cats in United States veterinary teaching hospitals, August 2020. J Vet Intern Med. 2023;37:1864–1875. 10.1111/jvim.16814 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Hsieh ES, Bollig ER, Beaudoin AL, Morrow A, Granick JL. Serial point-prevalence surveys to estimate antibiotic use in a small animal veterinary teaching hospital, November 2018 to October 2019. J Vet Intern Med. 2022;36:244–252. 10.1111/jvim.16314 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Singleton DA, Pinchbeck GL, Radford AD, et al. Factors associated with prescription of antimicrobial drugs for dogs and cats, United Kingdom, 2014-2016. Emerg Infect Dis. 2020;26:1778–1791. 10.3201/eid2608.191786 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Makita K, Sugahara N, Nakamura K, et al. Current status of antimicrobial drug use in Japanese companion animal clinics and the factors associated with their use. Front Vet Sci. 2021;8:705648. 10.3389/fvets.2021.705648 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Weese JS, Blondeau J, Boothe D, et al. International Society for Companion Animal Infectious Diseases (ISCAID) guidelines for the diagnosis and management of bacterial urinary tract infections in dogs and cats. Vet J. 2019;247:8–25. 10.1016/j.tvjl.2019.02.008 [DOI] [PubMed] [Google Scholar]
- 12. Lappin MR, Blondeau J, Boothe D, et al. Antimicrobial use guidelines for treatment of respiratory tract disease in dogs and cats: Antimicrobial Guidelines Working Group of the International Society for Companion Animal Infectious Diseases. J Vet Intern Med. 2017;31:279–294. 10.1111/jvim.14627 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Hillier A, Lloyd DH, Weese JS, et al. Guidelines for the diagnosis and antimicrobial therapy of canine superficial bacterial folliculitis (Antimicrobial Guidelines Working Group of the International Society for Companion Animal Infectious Diseases). Vet Dermatol. 2014;25:163-e43. 10.1111/vde.12118 [DOI] [PubMed] [Google Scholar]
- 14. Loeffler A, Cain CL, Ferrer L, et al. Antimicrobial use guidelines for canine pyoderma by the International Society for Companion Animal Infectious Diseases (ISCAID). Vet Dermatol. 2025;36:234–282. 10.1111/vde.13342 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Jessen LR, Werner M, Singleton D, et al. European Network for Optimization of Veterinary Antimicrobial Therapy (ENOVAT) guidelines for antimicrobial use in canine acute diarrhoea. Vet J. 2024;307:106208. 10.1016/j.tvjl.2024.106208 [DOI] [PubMed] [Google Scholar]
- 16. Scarborough RO, Sri AE, Browning GF, Hardefeldt LY, Bailey KE. “Brave enough”: a qualitative study of veterinary decisions to withhold or delay antimicrobial treatment in pets. Antibiotics (Basel). 2023;12:12. 10.3390/antibiotics12030540 [DOI] [Google Scholar]
- 17. WHO Advisory Group on Integrated Surveillance of Antimicrobial Resistance (AGISAR) . Critically Important Antimicrobials for Human Medicine. World Health Organization; 2018. https://www.who.int/publications/i/item/9789241515528 [Google Scholar]
- 18. WHO Advisory Group on Integrated Surveillance of Antimicrobial Resistance (AGISAR) . WHO’s List of Medically Important Antimicrobials: A Risk Management Tool for Mitigating Antimicrobial Resistance due to Non-human Use. World Health Organization; 2024: https://cdn.who.int/media/docs/default-source/gcp/who-mia-list-2024-lv.pdf [Google Scholar]
- 19. Singleton DA, Rayner A, Brant B, et al. A randomised controlled trial to reduce highest priority critically important antimicrobial prescription in companion animals. Nat Commun. 2021;12:1593. 10.1038/s41467-021-21864-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Weese JS, Taylor-Rakocevic ME, Topdjian K, et al. Antimicrobial dispensing for common conditions in dogs and cats at a large veterinary practice network. Vet J. 2023, 2025;312:106374. 10.1016/j.tvjl.2025.106374 [DOI] [Google Scholar]
- 21. Weese JS, Battersby I, Morrison J, Spofford N, Soltero-Rivera M. Antimicrobial use practices in canine and feline dental procedures performed in primary care veterinary practices in the United States. PLoS One. 2023;18:e0295070. 10.1371/journal.pone.0295070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Sharland M, Cappello B, Ombajo LA, et al. The WHO AWaRe Antibiotic Book: providing guidance on optimal use and informing policy. Lancet Infect Dis. 2022;22:1528-1530. 10.1016/s1473-3099(22)00683-1 [DOI] [PubMed] [Google Scholar]
- 23. Kwong M, Gardner HL, Dieterle N, et al. Optimization of electronic medical records for data mining using a common data model. Top Companion Anim Med. 2019;37:100364. 10.1016/j.tcam.2019.100364 [DOI] [Google Scholar]
- 24. U.S. Census Bureau . American community survey. In: 2022 American Community Survey 5-Year Estimates, Table S1901. Vol. 7. 2023. https://www.census.gov/data/developers/data-sets/acs-5year.html [Google Scholar]
- 25. Moon DC, Choi JH, Boby N, et al. Bacterial prevalence in skin, urine, diarrheal stool, and respiratory samples from dogs. Microorganisms. 2022;10:10. 10.3390/microorganisms10081668 [DOI] [Google Scholar]
- 26. Ling GV, Norris CR, Franti CE, et al. Interrelations of organism prevalence, specimen collection method, and host age, sex, and breed among 8,354 canine urinary tract infections (1969-1995). J Vet Intern Med. 2001;15:341-347. https://pubmed.ncbi.nlm.nih.gov/11467591/ [Google Scholar]
- 27. Tompson AC, Mateus ALP, Brodbelt DC, et al. Understanding antibiotic use in companion animals: a literature review identifying avenues for future efforts. Front Vet Sci. 2021;8:719547. 10.3389/fvets.2021.719547 [DOI] [Google Scholar]
- 28. Lavigne SH, Louis S, Rankin SC, Zaoutis TE, Szymczak JE. How companion animal veterinarians in the United States perceive financial constraints on antibiotic decision-making. Vet Rec. 2021;188:e62. 10.1002/vetr.62 [DOI] [Google Scholar]
- 29. Mateus AL, Brodbelt DC, Barber N, et al. Qualitative study of factors associated with antimicrobial usage in seven small animal veterinary practices in the UK. Prev Vet Med. 2014;117:68-78. 10.1016/j.prevetmed.2014.05.007 [DOI] [Google Scholar]
- 30. Zanichelli V, Tebano G, Gyssens IC, et al. Patient-related determinants of antibiotic use: a systematic review. Clin Microbiol Infect. 2019;25:48-53. 10.1016/j.cmi.2018.04.031 [DOI] [Google Scholar]
- 31. Lines LM, Li NC, Mick EO, Ash AS. Emergency department and primary care use in Massachusetts 5 years after health reform. Med Care. 2019;57:101-108. 10.1097/MLR.0000000000001025 [DOI] [Google Scholar]
- 32. Fournier Q, Serra JC, Williams C, Bavcar S. Chemotherapy-induced diarrhoea in dogs and its management with smectite: results of a monocentric open-label randomized clinical trial. Vet Comp Oncol. 2021;19:25-33. 10.1111/vco.12631 [DOI] [Google Scholar]
- 33. Hopman NEM, Portengen L, Heederik DJJ, Wagenaar JA, van Geijlswijk I, Broens EM. Time trends, seasonal differences and determinants of systemic antimicrobial use in companion animal clinics (2012-2015). Vet Microbiol. 2019;235:289-294. 10.1016/j.vetmic.2019.07.016 [DOI] [Google Scholar]
- 34. Hardefeldt LY, Selinger J, Stevenson MA, et al. Population wide assessment of antimicrobial use in dogs and cats using a novel data source—a cohort study using pet insurance data. Vet Microbiol. 2018;225:34-39. 10.1016/j.vetmic.2018.09.010 [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.