Objective: To review the human and companion animal veterinary literature on nosocomial infections and antimicrobial drug resistance as they pertain to the critically ill patient. Data sources: Data from human and veterinary sources were reviewed using PubMed and CAB. Human data synthesis: There is a large amount of published data on nosocomially-acquired bloodstream infections, pneumonia, urinary tract infections and surgical site infections, and strategies to minimize the frequency of these infections, in human medicine. Nosocomial infections caused by multi-drug-resistant (MDR) pathogens are a leading cause of increased patient morbidity and mortality, medical treatment costs, and prolonged hospital stay. Epidemiology and risk factor analyses have shown that the major risk factor for the development of antimicrobial resistance in critically ill human patients is heavy antibiotic usage. Veterinary data synthesis: There is a paucity of information on the development of antimicrobial drug resistance and nosocomially-acquired infections in critically ill small animal veterinary patients. Mechanisms of antimicrobial drug resistance are universal, although the selection effects created by antibiotic usage may be less significant in veterinary patients. Future studies on the development of antimicrobial drug resistance in critically ill animals may benefit from research that has been conducted in humans. Conclusions: Antimicrobial use in critically ill patients selects for antimicrobial drug resistance and MDR nosocomial pathogens. The choice of antimicrobials should be prudent and based on regular surveillance studies and accurate microbiological diagnostics. Antimicrobial drug resistance is becoming an increasing problem in veterinary medicine, particularly in the critical care setting, and institution-specific strategies should be developed to prevent the emergence of MDR infections. The collation of data from tertiary-care veterinary hospitals may identify trends in antimicrobial drug resistance patterns in nosocomial pathogens and aid in formulating guidelines for antimicrobial use.

Results suggested that the proportion of rectal E coli isolates obtained from dogs housed for >or= 3 days in a veterinary teaching hospital ICU that were resistant to antimicrobial agents increased as the duration of hospitalization in the ICU increased. Thus, ICU hospitalization time should be as short as possible to prevent development of antimicrobial resistance among rectal E coli isolates.

The ICU-acquired MDR E coli UTI likely originated from the dog's intestinal flora during hospitalization. Dogs that have been referred from a community practice may have MDR E coli UTIs at the time of admission.

Introduction: Urinary tract infections (UTIs) in dogs with urinary catheters in intensive care units (ICUs) are frequent. Historically, multi‐drug resistant (MDR) Escherichia coli account for about 10% of the UTIs. The objectives of this study were to determine the frequency of E. coli infections and of MDR E. coli in dogs with UTIs in our ICU, and to assess whether the MDR E. coli were community‐acquired or nosocomial in origin. Methods: Over a 1‐year period, rectal swabs were taken from all dogs in the ICU on the day of admission (D0) and on days 3 (D3), 6 (D6), 9 (D9) and 12 (D12). Urine was collected on these days from dogs with an indwelling urinary catheter (n=190). Rectal swabs and urine were routinely cultured. E. coli isolates were identified by biochemical tests. Using NCCLS guidelines, antibiotic susceptibility testing was done by disk diffusion method on fecal and urinary E. coli isolates. Twelve antimicrobial agents were used: nalidixic acid, enrofloxacin, cephalothin, cefoxitin, cefotaxime, ceftiofur, trimethoprim‐sulfa, chloramphenicol, gentamicin, tetracycline, ampicillin, and amoxicillin/clavulanate. Pulsed‐field gel electrophoresis (PFGE) was used to compare MDR E. coli UTI strains with fecal E. coli strains from the same patient and with MDR fecal E. coli from patients that were adjacent to, or housed in the same cages. Results: E. coli was cultured from 12 (48%) of 25 UTIs. Two of the E. coli were MDR. For one dog, PFGE showed no similarities among fecal E. coli and the urinary MDR E. coli isolates from the patient or between these isolates and fecal E. coli from a dog housed in the same kennel on the previous day. The MDR E. coli UTI was likely acquired prior to admission to the ICU, as it was present on D0. For the other dog, PFGE showed genetic similarity but not complete identity between the D3 MDR E. coli urinary isolate and the D3, D6, D9 fecal MDR isolates. This suggests that the UTI originated with the fecal E. coli. Using selective plates, fecal MDR E. coli were not found on D0. Selection of the MDR strain in the intestine by the use of antibiotics occurred while the dog was in the ICU and possibly led to the UTI. Conclusions: Multi‐drug resistant E. coli accounted for 2 of 12 E. coli UTIs in dogs in the ICU over a 1‐year period. Genotyping showed that one of the two MDR E. coli infections could possibly be of nosocomial origin.

Introduction: Antibiotic resistance develops in human patients in the intensive care unit (ICU) with increased duration of stay. The major selection factor is the use of antimicrobial agents. The hypothesis tested in this study was that antimicrobial resistance in fecal E. coli isolated from dogs increased with increased length of stay in the ICU. Methods: E. coli were isolated from rectal swabs collected on days 0 (D0, day of admission to the ICU), 3 (D3), 6 (D6), and 9 (D9) from all dogs housed in the ICU for 3 or more days. The E. coli were tested for resistance to nalidixic acid, enrofloxacin, cephalothin, cefoxitin, cefotaxime, ceftiofur, trimethoprim‐sulfa, chloramphenicol, gentamicin, tetracycline, ampicillin and amoxicillin/clavulanate. Thirty‐two patients were included in the study, 17 with D0‐D3 and 15 with D0‐D6 E. coli isolates. Five E. coli strains were isolated each sampling day and tested for antimicrobial susceptibility, resulting in a total of 60 outcomes per day per dog. The proportion of positive outcomes (resistance) was calculated for each day up to D6. Univariate ANOVA analysis was used to analyze the data. Results: The proportion of positive outcomes (resistance) increased from 9.3% on D0 to 14.7% on D3 to 27.9% on D6. Resistance to the frequently used ampicillin increased in a linear manner with increased duration of stay. The proportion of ampicillin‐resistant colonies increased from 19.6% on D0 to 39.1% on D3 to 68.7% on D6 (p=0.001). Resistance to the infrequently used amoxicillin/clavulanic acid, also increased significantly (p=0.026) and linearly. The increase in resistance to enrofloxacin was significant (p=0.046); but the increase in resistance to the precursor, nalidixic acid, was highly significant (p=0.020). No change in resistance to gentamicin was found. The remaining antibiotics tested showed an insignificant increase in resistance. Conclusions: Resistance to antibiotics increased in dogs with duration of stay in the ICU for 3 or more days. This linear trend was most significant with penicillin‐based beta‐lactams, which are frequently used in the ICU. Consequently, critically ill dogs may be at increased risk for developing opportunistic infections, caused by multi‐drug resistant organisms.