Introduction
Antimicrobial resistance (AMR) is a leading public health threat of the 21st century in human and veterinary medicine with significant clinical and economic consequences.1–6 Compared to non–multidrug-resistant (MDR) bacterial infections in human hospitals, MDR infections have consequently increased hospital times and cost of treatment and resulted in higher morbidity and mortality rates.7–9 In addition to the clinical consequences of AMR infections, there are also marked economic consequences. In 2019, the CDC conducted a broad retrospective study10 across 700 diverse acute-care hospitals in the US and estimated the direct healthcare cost associated with AMR infections to be $20 billion annually. Additional studies with narrower focuses and those confined to the 6 most common MDR infections estimated infection-related costs to be between $1.9 billion and $5.1 billion annually.11–13 Furthermore, it has been speculated that if antimicrobial resistance continues to rise at its current annual projections, these pathogens may lead to 8.22 to 10 million human deaths annually and incur an economic burden of $100 trillion in costs by the year 2050.14–16
Although data in veterinary medicine regarding AMR infections are comparatively limited,17 there are similar reported consequences.6,18–20 In more recent years, there has been an increased recognition in veterinary hospitals regarding the prevalence of MDR infections.21 However, there are few studies available that evaluate the prevalence, risk factors, and consequences of these pathogens on animal healthcare.17,22–24 The reported risk factors for the development of MDR bacterial infections in human and veterinary medicine share similarities, including prior antimicrobial use, corticosteroid or chemotherapy administration, previous hospitalization episodes, chronic comorbidities, prolonged hospitalization periods, the use of invasive medical or surgical devices (ie, urinary catheters, intravenous catheters, implant devices, etc), and surgical procedures.20,23,25,26
In 2020, the AVMA published a report27 on the challenges imposed by MDR pathogens in veterinary hospitals. Similar to the reported consequences of MDR infections in human medicine, the repercussions in veterinary medicine include increased patient hospitalization lengths, delayed patient recoveries, increased patient morbidity and mortality rates, and increased medical treatment costs.4,6,8,17,18,27,28 Two studies19,20 focusing on the economic impact of tibial plateau leveling osteotomy (TPLO) surgical site infections (SSIs) in dogs supported that the total postoperative costs were significantly higher in dogs with SSIs after TPLO surgery. Espinel-Rupérez et al20 reported a 142.2% increase in postoperative costs associated with SSIs due to increased follow-up appointments, additional treatments, longer hospitalizations, and further diagnostics. In contrast, Nicoll et al19 compared postoperative TPLO patients with methicillin-resistant Staphylococcus pseudintermedius (MRSP) SSIs versus non-MDR pathogen SSIs and found no significant differences in the overall cost of care, postoperative follow-ups, hospitalization durations, or overall length of care. However, the small sample size of MRSP infections (n = 6) and surgical costs not being separated from infection-related costs may have limited the detection of a significant difference.
The objective of this study was to retrospectively evaluate the prevalence of MDR wounds (traumatic and surgical) in dogs at a tertiary referral small animal hospital, screen for risk factors for MDR development, and evaluate the cost difference between MDR and non-MDR infections. We hypothesized that MDR infections would have a higher cost of care than antibiotic-susceptible infections.
Methods
Case selection criteria
The electronic medical record database at a tertiary referral small animal hospital was searched to identify canine patients with an aerobic bacterial culture and diagnosis of a wound infection, nonhealing wound, wound dehiscence, surgical wound dehiscence, implant infection, or bite wound between July 1, 2018, and November 29, 2023.
Canine patients with traumatic wound infections (eg, dog bite wounds, gunshot wounds, lacerations, etc) or SSIs were included if they had at least 1 positive bacterial culture with a paired susceptibility report. All bacterial cultures were performed at the Auburn Bacteriology and Mycology Laboratory. Standard susceptibility testing protocols for nonfastidious organisms used Vetik. Fastidious organisms, organisms that were known to do poorly, or ones that failed to grow in the Vetik would be tested via disk diffusion. Supplemental testing may have included disk diffusion (eg, for rifampin) or E-test (eg, for linezolid). Clinical and Laboratory Standards Institute veterinary standards were followed for breakpoints and susceptibility interpretations. Infections were categorized as susceptible, non-MDR, or MDR on the basis of criteria defined by the CDC. Susceptible infections were defined as those in which the bacteria were sensitive to all tested antibiotics. Non-MDR infections were classified as resistant to < 3 antibiotic drug classes. Multidrug-resistant infections were defined as those resistant to 3 or more antibiotic drug classes.29 The intermediate susceptibility results were excluded from data analysis.
Infections were classified as SSIs if the criteria adopted from the CDC Guideline for Prevention of Surgical Site Infections were met.30 A superficial incisional SSI was defined as only involving the skin and subcutaneous tissue at the incision site, developing within 30 days of the surgical procedure, and meeting at least one of the CDC’s inclusion criteria. A deep incisional SSI was defined as involving deeper fascia and muscle tissue, developing within 30 days of surgery if no implant was placed or within 1 year if an implant was involved, and meeting at least one of the CDC’s inclusion criteria.31 Implant-associated SSIs diagnosed beyond 1 year of surgery were categorized as long-term implant-associated infections.
Medical records review
Medical records were reviewed, and data collected included signalment, body weight, body condition score (BCS), wound location, wound type, culture and susceptibility results, systemic and topical antibiotic use both before and after culture results, wound care, hospitalization duration, concurrent illnesses, prior medications (ie, NSAIDs, corticosteroids, and immunosuppressives), overall costs, wound-specific treatment costs, and patient mortality. Concurrent illnesses were categorized as minor, moderate, or severe. Minor illnesses included urinary tract infections, parasites, upper respiratory signs, osteoarthritis, and heart murmurs. Moderate illnesses included trauma, endocrine diseases, immune-mediated diseases, hypoalbuminemia, neurological conditions, and symptomatic neoplasia. Severe illnesses included major trauma, sepsis, and patients in critical condition.
For patients diagnosed with an SSI, additional data were recorded, including the classification as superficial or deep, details of the surgical procedure, anesthesia duration, surgery duration, antibiotic administration and timing relative to the surgery, time from surgery to SSI diagnosis, and whether the patient was hospitalized postoperatively at the time of SSI diagnosis. Data collection for all dogs in this study was standardized to begin at the first episode with a positive culture and susceptibility result. The initial surgical costs were not included.
Statistical analysis
For statistical analysis, the main outcome variable was categorized into 2 groups: MDR infections and non-MDR infections, which included both susceptible and non-MDR infections. Some continuous variables were categorized on the basis of their median values. Total treatment cost was classified as low if ≤ $3,276 or high if > $3,276, hospitalization duration was categorized as low if ≤ 3 days or high if > 3 days, initial wound treatment cost was classified as low if ≤ $113 or high if > $113, and follow-up wound treatment cost was categorized as low if ≤ $375 or high if > $375.
Multivariable logistic regression was used to assess the associations between costs, wound factors, and patient factors with the presence of MDR infections. Factors were selected with a stepwise regression approach, and the best-fitting model was chosen on the basis of the Akaike information criterion. The significance of each factor’s association with the outcome was evaluated, and ORs, 95% CIs, and P values are reported for significant associations. Statistical significance was defined as P ≤ .05.
Results
One hundred eighteen dogs were identified with a wound- or SSI-associated aerobic culture and susceptibility result. Thirty-eight dogs were excluded due to a negative culture result, the absence of a paired susceptibility report, or culture results unrelated to a wound infection, such as a urine culture. Eighty dogs met the study inclusion criteria. Infections were classified as MDR in 23 dogs and non-MDR in 57 dogs (including 11 susceptible).
Signalment and medical background
Among the 80 dogs, 34 were castrated males, 26 were spayed females, 11 were sexually intact males, and 9 were sexually intact females. The most represented breeds were mixed breed (n = 20) and Labrador Retriever (12). The mean age at initial evaluation was 5.4 ± 3.4 years (range, 7 months to 14 years), and the mean weight was 28 ± 16 kg (range, 3 to 106 kg). The mean BCS was 5.4 ± 1.4/9: 8 dogs were underweight (BCS, 2 to 3/9), 35 had an optimal BCS (4 to 5/9), 23 were slightly overweight (BCS, 6/9), and 14 were obese (BCS ≥ 7/9). Data separated by non-MDR and MDR groups are shown in Table 1.
Patient- and wound-related factors for non–multidrug-resistant (MDR) and MDR groups.
| Non-MDR (n = 57) | MDR (n = 23) | P value | |
|---|---|---|---|
| Patient factors | |||
| Mean (SD) age (y) | 5.45 (3.32) | 4.98 (3.66) | .580 |
| Mean (SD) body weight (kg) | 27.85 (12.98) | 28.72 (22.34) | .829 |
| Gender | .888 | ||
| Intact male | 8 (14%) | 3 (13%) | |
| Castrated male | 24 (42%) | 10 (43.5%) | |
| Intact female | 7 (12%) | 2 (8.7%) | |
| Spayed female | 18 (32%) | 8 (34.8%) | |
| Concurrent illness | .073 | ||
| None | 23 (40.4%) | 6 (26.1%) | |
| Minor | 21 (36.8%) | 5 (21.7%) | |
| Moderate | 11 (19.3%) | 11 (47.8%) | |
| Severe | 2 (3.5%) | 1 (4.3%) | |
| Prior diagnosis of skin disease | .063 | ||
| No | 37 (64.9%) | 9 (39.1%) | |
| Yes | 20 (35.1%) | 14 (60.9%) | |
| NSAID/steroid within 1 week | .016* | ||
| None | 35 (61.4%) | 6 (26.1%) | |
| NSAID | 20 (35.1%) | 15 (65.2%) | |
| Steroid | 2 (3.5%) | 2 (8.7%) | |
| NSAID within 1 year | .102 | ||
| No | 28 (49.1%) | 6 (26.1%) | |
| Yes | 29 (50.9%) | 17 (73.9%) | |
| Steroid within 1 year | .023* | ||
| No | 54 (94.7%) | 17 (73.9%) | |
| Yes | 3 (5.3%) | 6 (26.1%) | |
| Immunosuppressive medication | |||
| Within 1 week | 1.000 | ||
| No | 52 (91.2%) | 21 (91.3%) | |
| Yes | 5 (8.8%) | 2 (8.7%) | |
| Within 1 year | 1.000 | ||
| No | 51 (89.5%) | 20 (87.0%) | |
| Yes | 6 (10.5%) | 3 (13.0%) | |
| Wound factors | |||
| Wound location | .888 | ||
| Dorsum | 4 (7.0%) | 2 (8.7%) | |
| Ventrum | 16 (28.1%) | 9 (39.1%) | |
| Thoracic limb | 9 (15.8%) | 3 (13.0%) | |
| Pelvic limb | 22 (38.6%) | 7 (30.4%) | |
| Head and neck | 6 (10.5%) | 2 (8.7%) | |
| Type of wound | .835 | ||
| SSI | 27 (47.4%) | 12 (52.2%) | |
| Laceration/trauma | 3 (5.3%) | 2 (8.7%) | |
| Bite | 11 (19.3%) | 3 (13.0%) | |
| Abscess | 3 (5.3%) | 1 (4.3%) | |
| Gunshot | 1 (1.8%) | 0 (0.0%) | |
| Long-term implant associated | 5 (8.8%) | 1 (4.3%) | |
| IV catheter associated | 2 (3.5%) | 0 (0.0%) | |
| Other | 5 (8.8%) | 4 (17.4%) | |
| Implant associated | .407 | ||
| No | 46 (80.7%) | 21 (91.3%) | |
| Yes | 11 (19.3%) | 2 (8.7%) |
SSI = Surgical site infection.
*Statistically significant.
In the non-MDR group of dogs, 20 had skin diseases (active pyoderma or atopic dermatitis), 21 had minor concurrent illnesses, 11 had moderate illnesses, and 2 had severe illnesses. In the MDR group of dogs, 14 had skin disease (active pyoderma or atopic dermatitis), 5 had minor concurrent illness, 11 had moderate illnesses, and 1 had severe illness.
In the non-MDR group, 5 dogs received steroids: 2 within 1 week and 3 within 1 year before culture results. Forty-nine dogs received NSAIDs: 20 within 1 week and 29 within 1 year before culture results. Eleven dogs received immunosuppressives including chemotherapy agents, mycophenolate mofetil, cyclosporine, and oclacitinib: 5 within 1 week and 6 within 1 year before culture results. In the MDR group, 8 dogs received steroids: 2 within 1 week and 6 within 1 year before culture results. Thirty-two dogs received NSAIDs: 15 within 1 week and 17 within 1 year before culture results. Five dogs received immunosuppressives: 2 within 1 week and 3 within 1 year before culture results.
Patient-related factors, including age, sex, body weight, concurrent illness, skin disease, NSAID use, steroid use, and immunosuppressive medications, were included in the multivariable logistic regression model to assess associations with MDR infections. The use of steroids in the last year and use of an NSAID or steroid in the last week increased the odds of MDR by 6.35 (95% CI, 1.43 to 28.17; P = .016) and 4.51 (95% CI, 1.54 to 13.18; P = .004), respectively. No other factors were statistically significant.
Wound characteristics and preculture treatment
Wound types included 39 SSIs, 14 bite wounds, 9 other wound types, 6 long-term implant-associated wounds, 5 laceration or traumatic wounds, 4 abscesses, 2 intravenous catheter site infections, and 1 gunshot wound. Wound locations included 29 pelvic limb, 25 ventrum, 12 thoracic limb, 8 head and neck, and 6 dorsum. Data separated by non-MDR and MDR groups are shown in Table 1.
The total percentage of dogs that received at least 1 antibiotic within 3 months before culture was 96.3% (77 of 80 dogs). Of the 57 dogs with non-MDR infections, 54 (94.7%) had received at least 1 antibiotic within 3 months before culture, while all dogs with MDR infections had prior antibiotic administration. In the non-MDR group, a median of 1 antibiotic class (range, 0 to 5) was used within the 3 months preceding culture and the median duration was 6 days (range, 0 to 60). For MDR cases, a median of 2 antibiotic classes (range, 1 to 6) were administered and the median duration was 13.5 days (range, 3 to 81 days). The most commonly administered preculture antibiotics in dogs with non-MDR infections were penicillins (52.6% [30 of 57]), cephalosporins (49.1% [28 of 57]), and fluoroquinolones (21.1% [12 of 57]). In dogs with MDR infections, the most commonly administered preculture antibiotics were penicillins (82.6% [19 of 23]), cephalosporins (39.1% [9 of 23]), clindamycin (21.7% [5 of 23]), and fluoroquinolones (17.4% [4 of 23]).
Wound and wound treatment–related factors, including wound location, wound type, presence of an implant, and antibiotic treatment, were included in the multivariable logistic regression model to assess associations with MDR infections. The use of penicillin or clindamycin as the first antibiotic used in the past 3 months increased the odds of MDR by 7.4 or 6.17 (95% CI, 1.71 to 32.09 or 1 to 37.98; P = .008 or .04), respectively. The first antibiotic type chosen for initial use following culture had a protective effect against multidrug resistance development. The use of a cephalosporin or fluoroquinolone as the first antibiotic following culture resulted in an OR of 0.1 for the development of multidrug resistance for both antibiotics (95% CI, 0.02 to 0.65 or 0.001 to 0.98; P = .014 or .048, respectively).
Of the 39 SSIs, 27 were non-MDR infections and 12 were MDR infections. The types of surgeries associated with SSIs varied: 25 involved soft tissue surgeries and 14 involved orthopedic surgeries. Thirteen SSIs were implant associated, including 12 orthopedic implant–associated infections and 1 soft tissue cisplatin bead implant–associated infection. The non-MDR SSI group had 18 superficial incisional infections and 9 deep incisional infections. The MDR SSI group had 11 superficial incisional infections and 1 deep incisional infection.
A specialist (board-certified surgeon or surgery resident) performed 26 of the 27 surgeries associated with non-MDR SSIs, which included 11 orthopedic and 15 soft tissue surgeries. In cases of MDR SSIs, a specialist performed 8 of the 12 surgeries, which included 1 orthopedic and 7 soft tissue surgeries. A primary care veterinarian performed 1 surgery from the non-MDR SSI group and 4 surgeries from the MDR SSI group. For non-MDR SSIs, the mean anesthesia duration was 252.9 minutes and the mean surgical duration was 145.2 minutes. For MDR SSI cases, the mean anesthesia duration was 197.8 minutes and the mean surgical duration was 139.3 minutes. Twenty-three dogs with non-MDR SSIs received perioperative cefazolin, which was administered an average of 41 minutes before the initial skin incision. Seven dogs with MDR SSIs received perioperative cefazolin, which was administered an average of 25.2 minutes before the initial skin incision.
A multivariable logistic regression model evaluating surgical and anesthetic factors showed that surgery performed by a specialist reduced the odds of MDR by 0.08 (95% CI, 0.01 to 0.99; P = .014) compared with a primary care veterinarian. However, the number of surgeries performed by a primary care veterinarian in this population was small (5 performed by a primary care veterinarian versus 34 performed by a specialist). No other significant risk factors were identified.
Bacterial prevalence
A total of 14 bacterial species were identified, yielding 130 bacterial isolates, including 23 MDR isolates (Supplementary Table S1). Staphylococcus spp were the most common pathogen, accounting for 41 of the 130 total isolates (31.5%). Of these, 9 were susceptible and 13 were MDR. All 13 MDR isolates were S pseudintermedius, which accounted for 36 of the 41 Staphylococcus spp isolates. Escherichia coli was the second most common bacteria, with 20 isolates. Of the E coli isolates, 0 were susceptible, 16 were resistant to < 3 classes of antibiotics, and 4 were MDR. Enterococcus spp were represented by 14 isolates, with 6 being susceptible, 6 resistant to < 3 classes of antibiotics, and 2 MDR. Streptococcus spp accounted for 13 isolates, all of which were MDR. Enterobacter spp represented 12 isolates, with 10 being resistant to < 3 classes of antibiotics and 2 MDR. Other pathogens included 11 Pseudomonas aeruginosa isolates, 5 Klebsiella pneumoniae isolates (1 MDR), 4 Proteus mirabilis isolates (1 MDR), 3 Pasteurella spp isolates, 2 Bacillus spp isolates, 2 Serratia marcescens isolates, and 1 each of the following: Aeromonas veronii biovar veronii isolate, Rhodococcus equi isolate, and Brevibacterium spp isolate.
Antimicrobial resistance patterns
Antibiotic sensitivity patterns for non-MDR bacteria are summarized in Table 2. These isolates were found to be most resistant to cephalexin (90.0%). Resistance to the penicillin drug class as a whole was noted at 52%, which included 65.4% resistance to ampicillin, 48.6% resistance to amoxicillin–clavulanic acid, and 38.2% resistance to penicillin. Susceptibility remained high to chloramphenicol (98.9%), fluoroquinolones (97.7%), rifampin (97.7%), macrolides (94.0%), trimethoprim/sulfa (93.3%), third-generation cephalosporins (92.9%), clindamycin (91.5%), aminoglycosides (90.9%), and tetracyclines (90.3%). When comparing the aminoglycosides, gentamicin had 97.8% isolate susceptibility, whereas amikacin had a lower isolate susceptibility of 84.0%. On extended susceptibility testing for select isolates, there was 100.0% susceptibility to imipenem, neomycin, and piperacillin-tazobactam antibiotics.
Antimicrobial resistance patterns of non-MDR bacterial isolates (n = 107).
| Antibiotics | Sensitive | Resistant | Class resistance | ||
|---|---|---|---|---|---|
| n | % | n | % | ||
| Amoxicillin–clavulanic acid | 37 | 51.4 | 35 | 48.6 | Penicillin (52%) |
| Ampicillin | 18 | 34.6 | 34 | 65.4 | |
| Oxacillin resistant | 0 | 0.0 | 6 | 100.0 | |
| Penicillin | 21 | 61.8 | 13 | 38.2 | |
| Piperacillin-tazobactam | 5 | 100.0 | 0 | 0.0 | |
| Cephalexin | 3 | 10.0 | 27 | 90.0 | 1st-gen cephalosporin (90%) |
| Cefpodoxime | 55 | 94.8 | 3 | 5.2 | 3rd-gen cephalosporin (7.1%) |
| Ceftazidime | 41 | 93.2 | 3 | 6.8 | |
| Cefovecin | 29 | 93.5 | 2 | 6.5 | |
| Ceftiofur | 6 | 75.0 | 2 | 25.0 | |
| Enrofloxacin | 92 | 97.9 | 2 | 2.1 | Fluoroquinolone (2.3%) |
| Marbofloxacin | 95 | 96.9 | 3 | 3.1 | |
| Pradofloxacin | 29 | 100.0 | 0 | 0.0 | |
| Doxycycline | 75 | 89.3 | 9 | 10.7 | Tetracycline (9.7%) |
| Minocycline | 27 | 93.1 | 2 | 6.9 | |
| Gentamicin | 90 | 97.8 | 2 | 2.2 | Aminoglycoside (8.8%) |
| Amikacin | 79 | 84.0 | 15 | 16.0 | |
| Neomycin | 6 | 100.0 | 0 | 0.0 | |
| Erythromycin | 46 | 93.9 | 3 | 6.1 | Macrolide (7.8%) |
| Clarithromycin | 1 | 100.0 | 0 | 0.0 | |
| Clindamycin | 43 | 91.5 | 4 | 8.5 | |
| Chloramphenicol | 87 | 98.9 | 1 | 1.1 | Amphenicol (1.1%) |
| Trimethoprim/sulfa | 70 | 93.3 | 5 | 6.7 | Sulfonamide (6.7%) |
| Rifampin | 43 | 97.7 | 1 | 2.3 | Antimycobacterial (2.3%) |
| Imipenem | 18 | 100.0 | 0 | 0.0 | Carbapenem (0%) |
1st-gen = First-generation. 3rd-gen = Third-generation.
The antimicrobial resistance patterns of MDR bacteria are summarized in Table 3. Multidrug-resistant isolates had the highest resistance to penicillins (96.8%), macrolides (83.8%), fluoroquinolones (70.7%), and tetracyclines (59.4%). Of the beta-lactams, this included 100% resistance to amoxicillin–clavulanic acid, ampicillin, and cephalexin. Some sensitivity remained intact to the third generation of cephalosporins: 60% sensitive to cefpodoxime, 66.7% sensitive to ceftazidime, and 50% sensitive to cefovecin. There was 75.0% isolate resistance to trimethoprim/sulfa, 69.2% resistance to clindamycin, 66.7% resistance to ceftiofur, and 62.5% resistance to gentamicin. The MDR-bacterial isolates displayed high susceptibility to amikacin (95.2%) and rifampin (92.3%).
Antimicrobial resistance patterns of MDR bacterial isolates (n = 23).
| Antibiotics | Sensitive | Resistant | Class resistance | ||
|---|---|---|---|---|---|
| n | % | n | % | ||
| Amoxicillin–clavulanic acid | 0 | 0.0 | 8 | 100.0 | Penicillin (96.8%) |
| Ampicillin | 0 | 0.0 | 7 | 100.0 | |
| Oxacillin resistant | 0 | 0.0 | 11 | 100.0 | |
| Penicillin | 1 | 20.0 | 4 | 80.0 | |
| Cephalexin | 0 | 0.0 | 7 | 100.0 | 1st-gen cephalosporin (100%) |
| Cefpodoxime | 6 | 60.0 | 4 | 40.0 | 3rd-gen cephalosporin (66.7%) |
| Ceftazidime | 4 | 66.7 | 2 | 33.3 | |
| Cefovecin | 2 | 50.0 | 2 | 50.0 | |
| Ceftiofur | 1 | 33.3 | 2 | 66.7 | |
| Enrofloxacin | 6 | 27.3 | 16 | 72.7 | Fluoroquinolone (70.7%) |
| Marbofloxacin | 7 | 30.4 | 16 | 69.6 | |
| Pradofloxacin | 4 | 30.8 | 9 | 69.2 | |
| Doxycycline | 8 | 38.1 | 13 | 61.9 | Tetracycline (59.4%) |
| Minocycline | 5 | 45.5 | 6 | 54.5 | |
| Gentamicin | 6 | 37.5 | 10 | 62.5 | Aminoglycoside (29.7%) |
| Amikacin | 20 | 95.2 | 1 | 4.8 | |
| Erythromycin | 1 | 6.7 | 14 | 93.3 | Macrolide (83.8%) |
| Azithromycin | 0 | 0.0 | 1 | 100.0 | |
| Clarithromycin | 0 | 0.0 | 1 | 100.0 | |
| Clindamycin | 4 | 30.8 | 9 | 69.2 | |
| Chloramphenicol | 14 | 60.9 | 9 | 39.1 | Amphenicol (39.1%) |
| Trimethoprim/sulfa | 5 | 25.0 | 15 | 75.0 | Sulfonamide (75%) |
| Rifampin | 12 | 92.3 | 1 | 7.7 | Antimycobacterial (7.7%) |
| Imipenem | 3 | 100.0 | 0 | 0.0 | Carbapenem (0.0%) |
Economic and clinical outcome impact of antimicrobial resistance
Non-MDR wound infections incurred an average overall hospitalization cost of $2,949.70, with wound care–associated costs averaging $2,721.90. The average cost for wound-related antibiotics was $142.20, and the total follow-up cost for wound care was $397.00 on average. In contrast, MDR infections had an average overall cost of $4,323.90 and wound care–associated costs were an average of $3,664.60. The average wound-related antibiotic cost for MDR infections was $188.70, with the total wound-related follow-up cost being $304.20.
When focusing specifically on SSIs, MDR infections had higher overall total postoperative care costs compared to non-MDR infections. Non-MDR SSIs had an average total postoperative care cost of $2,422.65, with wound care–associated costs of $2,211.34. The average wound-related antibiotic cost for non-MDR SSIs was $182.96, and the total wound-related follow-up cost was $495.69 on average. Comparatively, MDR SSIs incurred an average total postoperative care cost of $3,405.71, with wound care–specific costs of $2,531.80. The average wound-related antibiotic cost for MDR SSIs was $161.08, and the wound-care follow-up cost was $107.72.
Hospitalization duration was also longer in MDR cases, with a mean of 7.95 days compared to 4.64 days for non-MDR cases. Two dogs with non-MDR infections died prior to discharge; of these, only 1 death (0.9%) was attributed to infection. Four dogs (17.4%) with MDR infections had infection-associated deaths: 3 died prior to discharge, and 1 died after discharge.
The multivariable logistic regression model revealed that MDR increased the odds of longer hospitalization duration by 2.98 times (95% CI,1.09 to 8.19; P = .034). Additionally, MDR infections increased the odds of higher overall costs by 3.57 times (95% CI,1.3 to 9.83; P = .014) and higher mortality rates by 11.8 times (95% CI, 1.24 to 112.08; P = .03). No statistical significance was identified when focusing only on the surgical patients and surgery-specific variables evaluated that could be associated with MDR.
Discussion
In accordance with the reported consequences of MDR infections in human medicine,7–9,32 this study found that MDR infections in dogs with traumatic wounds and SSIs are associated with significantly longer hospitalization durations, higher mortality rates, and increased treatment costs. Multidrug-resistant infections in all dogs resulted in an average total hospitalization cost increase of 46.6% (an additional $1,374.2) and a 40.6% increase (an additional $983.06) in dogs specifically with MDR SSIs.
The MDR infection prevalence in all dogs was 23.8% and in dogs with MDR SSIs was 30.8%. This was similar to the reported 35% to 38.7% prevalence of MDR infection found among canine populations in other studies.33,34 Six bacterial species were identified with MDR isolates: S pseudintermedius was the most prevalent, followed by E coli, Enterococcus spp, Enterobacter spp, K pneumoniae, and P mirabilis. These bacteria include those referred to as the ESKAPE pathogens: Enterococcus faecium, Staphylococcus aureus, K pneumoniae, Acinetobacter baumannii, P aeruginosa, and Enterobacter. The ESKAPE pathogens are a highly surveilled group of bacterial species in human medicine due to their multidrug resistance and zoonotic potential.35 A meta-analysis36 on the ESKAPE pathogens in canine populations highlighted the concerning prevalence of these species in veterinary hospitals. Furthermore, 84.6% of the S pseudintermedius isolates in this study were oxacillin drug class–resistant or MRSP, and this isolate is a growing concern in SSIs and cutaneous wound infections in dogs.18,37,38
This study identified risk factors for MDR infections, including previous antibiotic use, steroid use within the past year, and NSAID or steroid use within 1 week prior to culture. These findings align with research in both human and veterinary medicine, which has shown increased MDR risks with previous antimicrobial and corticosteroid use.20,26,39 Although no studies have directly assessed NSAID use as a risk factor for MDR, some human studies have reported mixed effects of NSAID use on infection outcomes.40–44 In certain cases, NSAIDs have demonstrated synergistic antibacterial activity when paired with specific antibiotics such as chloramphenicol, cefuroxime, ciprofloxacin, clindamycin, and gentamicin,40,41 while other studies have indicated that concurrent NSAID use during infection treatment may limit antibiotic efficacy, potentially promoting antibiotic resistance, particularly in soft tissue infections.42–44 The use of NSAIDs may represent a greater degree of trauma necessitating pain medications, and greater trauma may be an aspect of infection risk. However, the wound type and concurrent illnesses were not noted as significant risk factors for MDR in this study. Future research is needed to determine whether NSAIDs may contribute to MDR development.
Another identified risk factor for MDR infections was the use of penicillin or clindamycin as the initial preculture antibiotic. Beta-lactams are the most common first-line empirical antibiotic choice in small animals,45,46 which makes it difficult to determine whether this finding reflects causation or correlation. In addition, a culture may not be performed if an infection is clinically cleared with a first-tier empiric antibiotic choice. However, frequent exposure to this antibiotic class could increase the selective pressure for resistance. Two studies47,48 of dogs have reported significant associations between preculture beta-lactam administration and resistance to oxacillin and methicillin in Staphylococcus spp. Similarly, some sources also include clindamycin as a first-tier antibiotic choice, and this may also contribute to the development of antibiotic resistance to this class.45,49
Interestingly, fluoroquinolones were not identified as a risk factor in this study, and postculture use of cephalosporins and fluoroquinolones was associated with a reduced likelihood of MDR infections. In some studies, empirical fluoroquinolone use has been a reported risk factor for antibiotic resistance and specifically for nosocomial methicillin-resistant S aureus infections in both humans and dogs.3,47 Furthermore, fluoroquinolones have been noted to be among the most commonly prescribed empirical antimicrobials of high importance in veterinary medicine, with 3.5% to 18% of dogs having received them as first-choice antimicrobials across multiple studies.48 In this study, 21.3% of dogs received fluoroquinolones prior to culture. Although fluoroquinolones were not identified as a risk factor for MDR, 70.7% of MDR isolates were resistant to the fluoroquinolone drug class, some of which may have been an intrinsic or previously acquired resistance. The authors recommend using fluoroquinolones cautiously and with the guidance of cultures and susceptibility panels. The protective association of postculture cephalosporin use against MDR may reflect selection bias, as these bacteria were likely susceptible to multiple antibiotic classes, including cephalosporins. Cephalosporins remain a recommended first-line empirical antibiotic choice, with subsequent treatment guided by culture and susceptibility results if needed.50
Surgeries performed by specialists were associated with fewer MDR SSI cases compared to those performed by primary care veterinarians. However, there may be a large bias in these data given the tertiary referral patient population and small sample size of primary care veterinarian cases. Future studies with larger and more evenly distributed population groups are necessary to further explore the potential association of surgeon expertise on the incidence of MDR SSIs.
Limitations of this study included those inherent to its retrospective nature, such as gaps within the medical records and data lost to follow-up. Loss of complete follow-up may have led to underestimation of the total cost of infections. Additional limitations included the small population and population at a single tertiary referral center. Future research investigating similar trends across more diverse study populations and hospital settings is recommended for comparison.
In conclusion, this study underscores the significant clinical and economic burdens MDR infections impose on veterinary healthcare. Potential risk factors for MDR infections included steroid use within the past year, NSAID or steroid use within 1 week before culture, and either penicillin or clindamycin use within 3 months before culture. Additional research is needed to confirm these risk factors across other canine patient populations and help guide antimicrobial treatment protocols.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org.
Acknowledgments
None reported.
Disclosures
The authors have nothing to disclose. No AI-assisted technologies were used in the generation of this manuscript.
Funding
The authors have nothing to disclose.
References
- 1.↑
Guardabassi L, Prescott JF. Antimicrobial stewardship in small animal veterinary practice: from theory to practice. Vet Clin North Am Small Anim Pract. 2015;45(2):361-376, vii. doi:10.1016/j.cvsm.2014.11.005
- 2.
Lagana DM, Taylor DD, Scallan Walter EJ. Advancing antimicrobial stewardship in companion animal veterinary medicine: a qualitative study on perceptions and solutions to a One Health problem. J Am Vet Med Assoc. 2023;261(8):1200-1207. doi:10.2460/javma.23.02.0100
- 3.↑
Palma E, Tilocca B, Roncada P. Antimicrobial resistance in veterinary medicine: an overview. Int J Mol Sci. 2020;21(6):1914. doi:10.3390/ijms21061914
- 4.↑
Prestinaci F, Pezzotti P, Pantosti A. Antimicrobial resistance: a global multifaceted phenomenon. Pathog Glob Health. 2015;109(7):309-318. doi:10.1179/2047773215Y.0000000030
- 5.
Huemer M, Mairpady Shambat S, Brugger SD, Zinkernagel AS. Antibiotic resistance and persistence: implications for human health and treatment perspectives. EMBO Rep. 2020;21(12):e51034. doi:10.15252/embr.202051034
- 6.↑
Guardabassi L. Veterinary hospital-acquired infections: the challenge of MRSA and other multidrug-resistant bacterial infections in veterinary medicine. Vet J. 2012;193(2):307-308. doi:10.1016/j.tvjl.2012.04.005
- 7.↑
Tansarli GS, Karageorgopoulos DE, Kapaskelis A, Falagas ME. Impact of antimicrobial multidrug resistance on inpatient care cost: an evaluation of the evidence. Expert Rev Anti Infect Ther. 2013;11(3):321-331. doi:10.1586/eri.13.4
- 8.↑
Serra-Burriel M, Keys M, Campillo-Artero C, et al. Impact of multi-drug resistant bacteria on economic and clinical outcomes of healthcare-associated infections in adults: systematic review and meta-analysis. PLoS One. 2020;15(1):e0227139. doi:10.1371/journal.pone.0227139
- 9.↑
Cosgrove SE. The relationship between antimicrobial resistance and patient outcomes: mortality, length of hospital stay, and health care costs. Clin Infect Dis. 2006;42(suppl 2):S82-S89. doi:10.1086/499406
- 10.↑
US CDC. Antibiotic resistance threats in the United States, 2019. CDC.gov. Updated December 2019. Accessed June 26, 2024. https://www.cdc.gov/antimicrobial-resistance/media/pdfs/2019-ar-threats-report-508.pdf
- 11.↑
Nelson RE, Hatfield KM, Wolford H, et al. National estimates of healthcare costs associated with multidrug-resistant bacterial infections among hospitalized patients in the United States. Clin Infect Dis. 2021;72(suppl 1):S17-S26. doi:10.1093/cid/ciaa1581
- 12.
Thorpe KE, Joski P, Johnston KJ. Antibiotic-resistant infection treatment costs have doubled since 2002, now exceeding $2 billion annually. Health Aff (Millwood). 2018;37(4):662-669. doi:10.1377/hlthaff.2017.1153
- 13.↑
Nelson RE, Hyun D, Jezek A, Samore MH. Mortality, length of stay, and healthcare costs associated with multidrug-resistant bacterial infections among elderly hospitalized patients in the United States. Clin Infect Dis. 2022;74(6):1070-1080. doi:10.1093/cid/ciab696
- 14.↑
O’Neill J. Antimicrobial resistance: tackling a crisis for the health and wealth of nations. Review on Antimicrobial Resistance. December 11, 2014. Accessed January 25, 2024. https://amr-review.org/sites/default/files/AMR%20Review%20Paper%20-%20Tackling%20a%20crisis%20for%20the%20health%20and%20wealth%20of%20nations_1.pdf
- 15.
O’Neill J. Tackling drug-resistant infections globally: final report and recommendations. Review on Antimicrobial Resistance. May 19, 2016. Accessed January 25, 2024. https://amr-review.org/sites/default/files/160518_Final%20paper_with%20cover.pdf
- 16.↑
Naghavi M, Vollset SE, Ikuta KS, et al.; Global Burden of Disease 2021 Antimicrobial Resistance Collaborators. Global burden of bacterial antimicrobial resistance 1990-2021: a systematic analysis with forecasts to 2050. Lancet. 2024;404(10459):1199-1226. doi:10.1016/S0140-6736(24)01867-1
- 17.↑
Ogeer-Gyles JS, Mathews KA, Boerlin P. Nosocomial infections and antimicrobial resistance in critical care medicine. J Vet Emerg Crit Care (San Antonio). 2006;16(1):1-18. doi:10.1111/j.1476-4431.2005.00162.x
- 18.↑
Nelson LL. Surgical site infections in small animal surgery. Vet Clin North Am Small Anim Pract. 2011;41(5):1041-1056, viii. doi:10.1016/j.cvsm.2011.05.010
- 19.↑
Nicoll C, Singh A, Weese JS. Economic impact of tibial plateau leveling osteotomy surgical site infection in dogs. Vet Surg. 2014;43(8):899-902. doi:10.1111/j.1532-950X.2014.12175.x
- 20.↑
Espinel-Rupérez J, Martín-Ríos MD, Salazar V, Baquero-Artigao MR, Ortiz-Díez G. Incidence of surgical site infection in dogs undergoing soft tissue surgery: risk factors and economic impact. Vet Rec Open. 2019;6(1):e000233. doi:10.1136/vetreco-2017-000233
- 21.↑
Weese JS, van Duijkeren E. Methicillin-resistant Staphylococcus aureus and Staphylococcus pseudintermedius in veterinary medicine. Vet Microbiol. 2010;140(3-4):418-429. doi:10.1016/j.vetmic.2009.01.039
- 22.↑
Sanchez S, McCrackin Stevenson MA, Hudson CR, et al. Characterization of multidrug-resistant Escherichia coli isolates associated with nosocomial infections in dogs. J Clin Microbiol. 2002;40(10):3586-3595. doi:10.1128/JCM.40.10.3586-3595.2002. Published correction appears in J Clin Microbiol. 2002;40(12):4806.
- 23.↑
Stull JW, Weese JS. Hospital-associated infections in small animal practice. Vet Clin North Am Small Anim Pract. 2015;45(2):217-233, v. doi:10.1016/j.cvsm.2014.11.009
- 24.↑
Walther B, Tedin K, Lübke-Becker A. Multidrug-resistant opportunistic pathogens challenging veterinary infection control. Vet Microbiol. 2017;200:71-78. doi:10.1016/j.vetmic.2016.05.017
- 25.↑
Stetter J, Boge GS, Grönlund U, Bergström A. Risk factors for surgical site infection associated with clean surgical procedures in dogs. Res Vet Sci. 2021;136:616-621. doi:10.1016/j.rvsc.2021.04.012
- 26.↑
Nienhoff U, Kadlec K, Chaberny IF, et al. Methicillin-resistant Staphylococcus pseudintermedius among dogs admitted to a small animal hospital. Vet Microbiol. 2011;150(1-2):191-197. doi:10.1016/j.vetmic.2010.12.018
- 27.↑
AVMA Committee on Antimicrobials. Antimicrobial resistant pathogens affecting animal health in the United States. AVMA. August 27, 2020. Accessed January 25, 2024. https://www.avma.org/sites/default/files/2020-10/AntimicrobialResistanceFullReport.pdf
- 28.↑
Corsini CMM, Silva VO, Carvalho OV, et al. Emergence of multidrug-resistant bacteria isolated from surgical site infection in dogs and cats. Arq Bras Med Vet Zootec. 2020;72(4):1213-1220. doi:10.1590/1678-4162-10978
- 29.↑
US CDC. Glossary of terms related to antibiotic resistance. CDC.gov. September 9, 2024. Accessed July 16, 2024. https://www.cdc.gov/narms/glossary/?CDC_AAref_Val=https://www.cdc.gov/narms/resources/glossary.html
- 30.↑
Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR; Hospital Infection Control Practices Advisory Committee. Guideline for prevention of surgical site infection, 1999. Infect Control Hosp Epidemiol. 1999;20(4):250-278. doi:10.1086/501620
- 31.↑
Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR; CDC Hospital Infection Control Practices Advisory Committee. Guideline for prevention of surgical site infection, 1999. Centers for Disease Control and Prevention (CDC) hospital infection control practices advisory committee. Am J Infect Control. 1999;27(2):97-132. doi:10.1016/S0196-6553(99)70088-X
- 32.↑
Poudel AN, Zhu S, Cooper N, et al. The economic burden of antibiotic resistance: a systematic review and meta-analysis. PLoS One. 2023;18(5):e0285170. doi:10.1371/journal.pone.0285170
- 33.↑
Nocera FP, Ambrosio M, Fiorito F, Cortese L, De Martino L. On gram-positive- and gram-negative-bacteria-associated canine and feline skin infections: a 4-year retrospective study of the university veterinary microbiology diagnostic laboratory of Naples, Italy. Animals (Basel). 2021;11(6):1603. doi:10.3390/ani11061603
- 34.↑
Dinkova V, Rusenova N. A retrospective study (2019-2023) on the prevalence and antimicrobial resistance of isolates from canine clinical samples submitted to the University Veterinary Hospital in Stara Zagora, Bulgaria. Microorganisms. 2024;12(8):1670. doi:10.3390/microorganisms12081670
- 35.↑
Santajit S, Indrawattana N. Mechanisms of antimicrobial resistance in ESKAPE pathogens. BioMed Res Int. 2016;2016(1):2475067. doi:10.1155/2016/2475067
- 36.↑
Santaniello A, Sansone M, Fioretti A, Menna LF. Systematic review and meta-analysis of the occurrence of ESKAPE bacteria group in dogs, and the related zoonotic risk in animal-assisted therapy, and in animal-assisted activity in the health context. Int J Environ Res Public Health. 2020;17(9):3278. doi:10.3390/ijerph17093278
- 37.↑
Windahl U, Bengtsson B, Nyman AK, Holst BS. The distribution of pathogens and their antimicrobial susceptibility patterns among canine surgical wound infections in Sweden in relation to different risk factors. Acta Vet Scand. 2015;57(1):11. doi:10.1186/s13028-015-0102-6
- 38.↑
Weese JS. A review of multidrug resistant surgical site infections. Vet Comp Orthop Traumatol. 2008;21(1):1-7. doi:10.3415/VCOT-07-11-0106
- 39.↑
Ponyon J, Kerdsin A, Preeprem T, Ungcharoen R. Risk factors of infections due to multidrug-resistant gram-negative bacteria in a community hospital in rural Thailand. Trop Med Infect Dis. 2022;7(11):328. doi:10.3390/tropicalmed7110328
- 40.↑
Chan EWL, Yee ZY, Raja I, Yap JKY. Synergistic effect of non-steroidal anti-inflammatory drugs (NSAIDs) on antibacterial activity of cefuroxime and chloramphenicol against methicillin-resistant Staphylococcus aureus. J Glob Antimicrob Resist. 2017;10:70-74. doi:10.1016/j.jgar.2017.03.012
- 41.↑
Öztürk İ, Eraç Y, Ballar Kırmızibayrak P, Ermertcan Ş. Nonsteroidal antiinflammatory drugs alter antibiotic susceptibility and expression of virulence-related genes and protein A of Staphylococcus aureus. Turk J Med Sci. 2021;51(2):835-847. doi:10.3906/sag-2003-60
- 42.↑
Li X, Xue X, Jia J, et al. Nonsteroidal anti-inflammatory drug diclofenac accelerates the emergence of antibiotic resistance via mutagenesis. Environ Pollut. 2023;326:121457. doi:10.1016/j.envpol.2023.121457
- 43.
Hamilton SM, Bayer CR, Stevens DL, Bryant AE. Effects of selective and nonselective nonsteroidal anti-inflammatory drugs on antibiotic efficacy of experimental group A streptococcal myonecrosis. J Infect Dis. 2014;209(9):1429-1435. doi:10.1093/infdis/jit594
- 44.↑
Verma T, Bhaskarla C, Sadhir I, Sreedharan S, Nandi D. Non-steroidal anti-inflammatory drugs, acetaminophen and ibuprofen, induce phenotypic antibiotic resistance in Escherichia coli: roles of marA and acrB. FEMS Microbiol Lett. 2018;365(22):fny251. doi:10.1093/femsle/fny251
- 45.↑
Guardabassi L, Larsen J, Weese JS, et al. Public health impact and antimicrobial selection of meticillin-resistant staphylococci in animals. J Glob Antimicrob Resist. 2013;1(2):55-62. doi:10.1016/j.jgar.2013.03.011
- 46.↑
Pomba C, Rantala M, Greko C, et al. Public health risk of antimicrobial resistance transfer from companion animals. J Antimicrob Chemother. 2017;72(4):957-968. doi:10.1093/jac/dkw481
- 47.↑
Cain CL, Morris DO, Rankin SC. Clinical characterization of Staphylococcus schleiferi infections and identification of risk factors for acquisition of oxacillin-resistant strains in dogs: 225 cases (2003-2009). J Am Vet Med Assoc. 2011;239(12):1566-1573. doi:10.2460/javma.239.12.1566
- 48.↑
Faires MC, Traverse M, Tater KC, Pearl DL, Weese JS. Methicillin-resistant and -susceptible Staphylococcus aureus infections in dogs. Emerg Infect Dis. 2010;16(1):69-75. doi:10.3201/eid1601.081758
- 49.↑
Gold RM, Lawhon SD. Incidence of inducible clindamycin resistance in Staphylococcus pseudintermedius from dogs. J Clin Microbiol. 2013;51(12):4196-4199. doi:10.1128/JCM.02251-13
- 50.↑
Beco L, Guaguère E, Lorente Méndez C, Noli C, Nuttall T, Vroom M. Suggested guidelines for using systemic antimicrobials in bacterial skin infections: part 2—antimicrobial choice, treatment regimens and compliance. Vet Rec. 2013;172(6):156-160. doi:10.1136/vr.101070
