Journal of Veterinary Diagnostic Investigation

http://vdi.sagepub.com/ Trends in Fluoroquinolone Resistance of Bacteria Isolated from Canine Urinary Tracts
Leah A. Cohn, Anthony T. Gary, William H. Fales and Richard W. Madsen J VET Diagn Invest 2003 15: 338 DOI: 10.1177/104063870301500406 The online version of this article can be found at: http://vdi.sagepub.com/content/15/4/338 Published by:
http://www.sagepublications.com

On behalf of:

Official Publication of the American Association of Veterinary Laboratory Diagnosticians, Inc.

Additional services and information for Journal of Veterinary Diagnostic Investigation can be found at: Email Alerts: http://vdi.sagepub.com/cgi/alerts Subscriptions: http://vdi.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://vdi.sagepub.com/content/15/4/338.refs.html

>> Version of Record - Jul 1, 2003 What is This?

Downloaded from vdi.sagepub.com by guest on December 21, 2013

J Vet Diagn Invest 15:338343 (2003)

Trends in uoroquinolone resistance of bacteria isolated from canine urinary tracts

Leah A. Cohn, Anthony T. Gary, William H. Fales, Richard W. Madsen
Abstract. Fluoroquinolone (FQ) antimicrobial agents are used extensively in human and veterinary medicine. Widespread use of any antimicrobial agent can apply selective pressure on populations of bacteria, which may result in an increase in the prevalence of antimicrobial-resistant isolates. Antimicrobial-susceptibility data on bacteria isolated from the canine urinary tract by the University of MissouriColumbia Veterinary Medical Diagnostic Laboratory, Columbia, MO, were used to determine whether there has been an increase in the prevalence of FQ-resistant bacteria over time. Between January 1992 and December 2001, minimum inhibitory concentrations of either ciprooxacin (19921998) or enrooxacin (19982001) were determined for 1,478 bacterial isolates from the canine urinary tract. The predominant bacterial species isolated were Escherichia coli (547 isolates), Proteus mirabilis (156), and Staphylococcus intermedius (147). In all, there were 13 bacterial species with more than 25 isolates each. A signicant increase in the overall proportion of resistant bacterial isolates was documented from 1992 to 2001 (CochranArmitage test for trend, P , 0.0001). The same increase in resistant isolates was documented when either ciprooxacin or enrooxacin was analyzed separately (P , 0.0001 and P , 0.0002, respectively). No difference was detected in rates of bacterial FQ resistance with regard to the sex of the dog from which the bacteria were isolated. The frequency with which some bacterial species were isolated differed with the sex of the infected dog. Proteus mirabilis was found more often in females (P , 0.0001), whereas beta hemolytic Streptococcus spp., were found more often in males (P 5 0.0003). Although the overall efcacy of FQ antimicrobials remained high with greater than 80% of isolates being susceptible, the data demonstrated an increase in the proportion of resistant bacteria isolated from the urinary tract of the dog.

Fluoroquinolone (FQ) antimicrobial agents function by inhibiting bacterial DNA gyrase, and they are commonly used to treat a variety of bacterial infections in dogs. As broad-spectrum bactericidal antimicrobials, FQ are effective against many gram-negative and gram-positive aerobes but possess little activity against anaerobic bacteria. Fluoroquinolones are particularly effective in treating urinary tract infections (UTI) because of the high drug concentrations attained within the urinary tract.19 In 1988, enrooxacin became the rst FQ to be marketed for veterinary use. Shortly before the introduction of enrooxacin, ciprooxacin was introduced for use in humans. Enrooxacin and ciprooxacin are structurally and functionally very similar, and in fact, ciprooxacin is a metabolite of enrooxacin.6 Because of their efcacy and relative safety, FQ rapidly gained widespread use in veterinary medicine. Simultaneously, there has been a dramatic increase in FQ use in human medicine.4,18,21 Whereas there have been few such veterinary studies, several investigators have documented the development of bacterial resistance asFrom the Departments of Veterinary Medicine and Surgery (Cohn, Gary), and Veterinary Pathobiology and the Veterinary Medical Diagnostic Laboratory (Fales), College of Veterinary Medicine, and the Department of Statistics (Madsen), University of Missouri Columbia, Columbia, MO 65211.

sociated with the growing use of FQ in human medicine.4 For FQ to maintain an appropriate position in the antimicrobial armamentarium of practicing veterinarians, it is important to recognize the development of FQ resistance among bacterial isolates associated with disease. The purpose of this study was to determine if there has been a change in documented FQ resistance among pathogens cultured from the canine urinary tract at the University of MissouriColumbia Veterinary Medical Diagnostic Lab (UMC-VMDL), Columbia, MO. Additionally, the effects of sex on the pattern of resistance as well as the relationship between sex and the specic pathogens isolated from the urinary tract were examined. Materials and methods
Bacterial isolates. Susceptibility results of urinary tract cultures submitted to the UMC-VMDL between January 1992 and December 2001 were analyzed. A vast majority of urine samples were obtained by cystocentesis, with nearly all the remaining samples collected by aseptic catheterization. Urine samples were cultured using loops calibrated to deliver 10 and 100 ml samples to 5% trypticase soy sheep blood agar (SBA)a and MacConkey agar plates.a,20 A portion of each urine sample obtained by cystocentesis was ltered with a sterile 0.22-mm lter, and the lter was cultured aerobically to detect samples with a low bacterial colony form-

338
Downloaded from vdi.sagepub.com by guest on December 21, 2013

FQ susceptibility in UTI

339

ing unit (CFU) count. To detect anaerobes, 100 ml of urine was inoculated on prereduced SBA plates.a They were incubated at 35 C for 2472 hr in a gas pak jar containing an anaerobic gas-generating systemb sachet and an anaerobic indicator. All aerobic plates were incubated at 35 C for 24 48 hr. The CFU count was noted, and gram-positive isolates were identied using standard methods.20 Gram-negative isolates were identied with an automated system.c,14 Information obtained from each urine sample identied in this search included organism(s) isolated, FQ susceptibility or resistance as determined by the minimum inhibitory concentration (MIC) of the antimicrobial, and sex of the infected dog. Fluoroquinolone antimicrobials. During various periods of the study, either ciprooxacin or enrooxacin was used as the representative FQ antimicrobial for microbial-susceptibility testing. Before 1998, ciprooxacin was used for routine FQ-susceptibility testing at the UMC-VMDL. Consequently, the majority of susceptibility data obtained before 1998 report susceptibility to ciprooxacin rather than to enrooxacin. Susceptibility tests run after March 1998 used enrooxacin as the standard representative FQ antimicrobial. Susceptibility testing. The MIC for each bacterial isolate was determined using commercially prepared lyophilized antimicrobial susceptibility trays.d Gram-negative bacterial isolates were picked to sterile, demineralized water adjusted to a 0.5 MacFarland standard for use with the AP-80 Gram Negative Identication Systemt.c,d Ten (10 ml) of the adjusted growth were transferred to 10 ml of MuellerHinton broth supplemented with Ca11, Mg11, and TES buffert.d Using an automated device,c lyophilized traysd were inoculated with 50 ml per well of the broth solution, sealed with plastic tape, and incubated at 35 C for 18 hr. Gram-positive isolates were cultured in trypticase soy brotha adjusted to 0.5 MacFarland standard and incubated in a similar manner. Fastidious isolates were cultured in MuellerHinton broth supplemented with 0.01% sterile, lysed horse blood. Antimicrobial susceptibility trays were processed with an automated read-incubation systeme or read manually.f The MIC was determined to be the least amount of antimicrobial agent required to inhibit bacterial growth. Although the high and low range of antimicrobial concentrations contained in the wells of the lyophilized trays varied somewhat during the time span examined, ciprooxacin was most often tested with doubling dilutions of 0.068.0 mg/ml, and enrooxacin was tested using doubling dilutions of 0.54.0 mg/ml. Minimum inhibitory concentrations of 2 and 1 mg/ml for ciprooxacin and enrooxacin, respectively, were used as interpretive breakpoints to determine resistance or susceptibility based on standards set by the National Committee for Clinical Laboratory Standards.10 Thus, an MIC below the breakpoint identied susceptible bacteria, whereas values above the breakpoints identied resistant organisms. Organisms with an MIC at the breakpoint were classied as having intermediate susceptibility. For purposes of analysis, organisms with intermediate susceptibility were grouped with resistant organisms.8 Statistical analysis. Trends in the proportion of susceptible isolates over time were examined to identify the development of FQ resistance. Because of the large number of statistical tests conducted, a stringent P value of 0.01 was

used to compensate for the increased chance of statistical error. A 1-sided CochranArmitage (CA) test for trend was used to determine whether there was an increasing trend in the proportion of resistant isolates over time, ignoring the variables of antimicrobial agent and organism. For each year, the total number of FQ-resistant isolates was compared with the total number of isolates identied that year to calculate the proportion of FQ-resistant bacteria before the application of the CA test for trend. Additionally, each antimicrobial (ciprooxacin and enrooxacin) was analyzed in a similar manner to determine whether there was an increasing trend in the proportion of isolates resistant to each drug. The CochranMantelHaenszel (CMH) test was also used for a stratied analysis, with strata being the type of drug used. Each of the 13 bacterial species isolated from more than 25 urine cultures was analyzed separately to determine whether there was a change in FQ resistance over time for that organism. The CA test for trend was used (as above) to determine the trend in resistance to ciprooxacin and enrooxacin for each of the 13 organisms. The CMH test was then used to control for the type of drug, accounting for both antimicrobial agents throughout the study period. Chi-square test was used to determine whether there was a difference in resistance to FQ between the sex categories (female intact [FI], female spayed [FS], male intact [MI], male neutered [MN]); the CMH was then used to stratify results for year and organism. In addition, the chi-square test was used to determine whether differences in prevalence of a given bacterial isolate from each sex category were signicant.

Results Bacterial isolates. A total of 1,478 urinary bacterial isolates with available susceptibility data were identied during the study period from the computerized records search. A new computer system was integrated into the UMC-VMDL in November 1992. Because of the transition from one database system to another, some data remained inaccessible, resulting in fewer isolates reported from the rst years of the study than from recent years. Ages of dogs with positive urine cultures ranged from 1 month to 20 years, with a mean of 6.9 years. In this study, the majority of isolates in urine were of Escherichia coli (547), Proteus mirabilis (156), or Staphylococcus intermedius (147). In total, 13 bacterial agents were identied from 25 or more urine cultures (Fig. 1). Susceptibility to FQ. The overall proportion of FQresistant isolates increased signicantly from 1992 to 2001 (CA Trend Test, P , 0.0001, n 5 1,478) (Fig. 2); the trend toward a signicant increase in resistance continued when each representative FQ was tested separately (Fig. 3). For the 695 isolates assayed for ciprooxacin susceptibility or resistance, the increase in resistance was signicant (P , 0.0001), as it was for

Downloaded from vdi.sagepub.com by guest on December 21, 2013

340

Cohn et al.

Figure 1. The proportion of bacterial isolates cultured from canine urine samples submitted to the UMC-VMDL between 1992 and 2001 resistant and susceptible to FQ antimicrobial agents.

Figure 2. Trends in resistance and susceptibility of urinary pathogens to FQ antimicrobial agents from 1992 to 2001.

the 783 isolates tested against enrooxacin (P 5 0.0002). In addition, when controlling for both the type of antimicrobial and organism with the CMH statistic, the increase in resistance was again signicant (P , 0.0001). Susceptibility of individual species isolated. Although small sample sizes limited statistical power, when analyzed individually, the only 2 species of bacterial isolates for which a signicant decrease in FQ susceptibility over time was demonstrated were E. coli and alpha hemolytic Streptococcus spp. (Fig. 4). The overall resistance of E. coli to FQ antimicrobials increased signicantly over time (P , 0.0001). When analysis was done for each type of antimicrobial separately, the increased resistance of E. coli to ciprooxacin was signicant (P , 0.0001), but resistance to enrooxacin failed to reach statistical signicance (P 5 0.022). The resistance of alpha hemolytic Streptococcus spp. to both FQ antimicrobial agents also increased during the study period (P 5 0.004). When each antimicrobial was assayed separately, the increase in resistance to ciprooxacin or enrooxacin failed to reach signicance (P 5 0.02 and P 5 0.015, respectively). Effects of sex on bacterial isolates and microbial susceptibility. Female spayed dogs accounted for 46.5% (664 of 1,463) of all positive cultures for which sex was recorded, which was more than that of any other sex category. Female intact dogs accounted for 18.4% of positive cultures, MI dogs accounted for 16.3%, and MC dogs for 21.4%. No statistically signicant relationship between sex and resistance of urinary bacterial isolates to FQ antimicrobials was identied. Two bacterial species were isolated with signicantly different frequencies between sex categories. In particular, P. mirabilis infection was more prevalent in

females (P , 0.0001), whereas beta hemolytic Streptococcus spp. infection was more prevalent in males (P 5 0.0003). Proteus mirabilis represented 10.6% of all bacterial isolates from the urinary tract of dogs but was responsible for 12.8% of isolates from FS dogs and 16.8% of isolates from FI dogs. In contrast, P. mirabilis was isolated from only 7.3% of all positive urine cultures from MI dogs and from 3.0% of cultures from MC dogs. Beta hemolytic Streptococcus spp. represented only 2.5% of the total isolates from the canine urinary tract, but it accounted for 3.0% of infections in MI dogs and for 5.9% in MC dogs, while accounting for only 1.2% of infections in both FS and FI dogs. Discussion The data demonstrate an increase in FQ resistance of canine urinary tract pathogens assayed by the UMC-VMDL from 1992 to 2001. These results are consistent with a slowly developing, progressive resistance to FQ, as noted by other investigators.4,8,23 Re-

Figure 3. Susceptibility of urinary pathogens either to ciprooxacin or to enrooxacin over time. The relative size of the bubble represents the number of isolates tested against the FQ antimicrobial in a given year.

Downloaded from vdi.sagepub.com by guest on December 21, 2013

FQ susceptibility in UTI

341

Figure 4. Escherichia coli and alpha hemolytic Streptococcus spp. isolates either susceptible or resistant to ciprooxacin and enrooxacin from 1992 to 2001.

sistance to FQ antimicrobials occurs through 2 chromosomally mediated mechanismsalteration in FQ target sites (DNA gyrase, topoisomerase IV) and decreased intracellular drug accumulation.1,11,12,1517 Unlike plasmid-mediated resistance, chromosomally mediated changes in susceptibility arise through mutations in bacterial DNA. Therefore, once resistance develops, the changes are irreversible and may progress to a permanent loss of therapeutic options for a given bacterial isolate.5 Furthermore, resistance is vertically transmitted to all daughter cells of the resistant isolate,

potentially leading to large populations of resistant bacteria. In contrast, horizontally transmitted plasmidmediated resistance is not permanent, often subsiding after the removal of the antimicrobial agent.5 The retrospective nature of this study limited it to the evaluation of MIC concentrations routinely used for susceptibility testing. Detection of an increased resistance to an antimicrobial agent and detection of a decreased susceptibility to that agent are distinctly different propositions.23 Although the methods used were adequate to detect an increase in FQ resistance, they

Downloaded from vdi.sagepub.com by guest on December 21, 2013

342

Cohn et al.

were not adequate to detect a decrease in FQ susceptibility. Decreased susceptibility is detected by documenting increases in MIC concentrations, particularly modal MIC as well as geometric mean MIC, MIC50 and MIC90. Most such incremental changes in susceptibility would occur at dilutions of antimicrobial agents that are several fold below the breakpoint MIC for that drug. Because the lowest concentration of enrooxacin used in susceptibility testing in this study was 0.5 mg/ ml, only 1 dilution less than the breakpoint MIC, changes over an appropriate range of dilutions could not be examined. More than 80% of the urinary isolates were susceptible to enrooxacin or ciprooxacin, with MICs less than or equal to 1 and 2 mg/ml, respectively. It is likely that changes in susceptibility that were more dramatic than changes in resistance were not detected because of the limited range of drug concentrations used for routine susceptibility testing. In fact, for ciprooxacin, the range began at 0.06 mg/ ml, and the authors were able to detect increases in geometric mean MIC as well as in MIC50 and MIC90 each year from 1993 to 1997 (data not shown). Susceptibility changes may have occurred in a large portion of the isolates, but they remained undetected because they occurred below the ranges used in the antimicrobial-susceptibility testing. There was an increase in resistance to FQ in urinary tract pathogens over all isolates, but when each isolate was analyzed individually, only E. coli and alpha hemolytic Streptococcus spp. demonstrated a signicant increase in resistance. Each of these organisms was found in relatively high numbers, ranking rst and sixth in overall prevalence. For bacterial species isolated less frequently, statistical power of analysis was limited. This numerical limitation might prevent detection of biologically relevant changes in resistance for those organisms. Alternatively, organisms with an exquisite initial susceptibility to FQ antimicrobial agents may experience a decrease in FQ susceptibility without an increase in resistance, and this trend could not be recognized by the methods used in this study. In fact, others have demonstrated such a phenomenon among canine pathogens.23 Development of bacterial resistance to commonly used antimicrobial agents is an expected phenomenon because bacterial mutations give rise to resistant bacterial populations through natural mutation or through plasmid-mediated transfer of resistance. An increase in FQ resistance of bacterial pathogens, as was demonstrated in this study, has been similarly demonstrated in other veterinary and human studies. Increases in resistance or decreases in susceptibility, or both, to enrooxacin among E. coli and other canine pathogenic bacteria, including P. mirabilis, S. intermedius, and Klebsiella pneumoniae, have been clearly demonstrat-

ed.23 Increased enrooxacin resistance of E. coli isolates from the urinary tract of dogs at the University of California VMTH Davis, California, was likely caused by a general increase in enrooxacin resistance among uropathogenic E. coli in dogs rather than from a single hospital strain clone of E. coli.2 Animal species other than dogs have likewise been found to harbor increasing proportions of FQ-resistant pathogens since the introduction of FQ to their therapeutic armamentarium.9 Numerous studies of human pathogens provide evidence supporting the development of FQ resistance.3,4,8,9,18,21,22 Data from a large-scale study conducted in the Netherlands showed that increased frequency of FQ prescriptions could be correlated with increased resistance of urinary tract pathogens to FQ (from 1.3% in 1989 to 5.8% in 1998).4 The role of sex was explored with regard to FQ resistance and bacterial species isolated. Urinary tract infections in men are more likely to be resistant to antimicrobial agents than are UTI in women, probably because infection in men is often associated with defective protective mechanisms and repeated antimicrobial treatment courses.4 As has been previously described, a higher incidence of UTI in female than in male dogs was documented in this study.7,19 In contrast to the case with human UTI, there was no difference in FQ susceptibility of pathogens found between sex categories for canine UTI. Although not a primary objective of this study, signicant sex differences in the rate of infection with certain species of bacteria were demonstrated. As in previous studies, P. mirabilis was isolated from a disproportionate number of females, but this study did not conrm that Enterococcus spp. were more likely to be isolated from females than from males.7 Additionally, this study documented an increased incidence of Streptococcus spp. beta hemolytic isolates from males compared with females. Reasons for these differences in urinary pathogens by sex remain unknown. Although an increase in bacterial resistance was documented by this study, the efcacy of FQ antimicrobials remained high. In fact, FQ were effective against more than 80% of microbial isolates tested in vitro. This is similar to the gures from another recent veterinary study in which 87.5% of nonenteric E. coli isolates from dogs analyzed over a period of 8 years were susceptible to enrooxacin.13 Although this study measured susceptibility in vitro, the in vivo bactericidal activity of FQ depends on the concentration of the drug at the site of action. Because FQ are concentrated from 100- to 300-fold within the urinary tract, they may still be effective against isolates found to be resistant in vitro with a known MIC.19 However, when the MIC exceeds the highest concentration on the susceptibility panel, alternative antimicrobial agents

Downloaded from vdi.sagepub.com by guest on December 21, 2013

FQ susceptibility in UTI

343

should be considered. Although this study only examined changes in FQ resistance among urinary tract pathogens, most of the same bacterial species can be found as canine pathogens in sites other than the urinary tract. It is incumbent on veterinarians to be cognizant of the role antimicrobial agents play in the development of resistant bacterial strains and to limit use of FQ in a judicious manner to maintain antimicrobial potency. Acknowledgements
The authors would like to thank Irene Ganjam, Simon Davies, and Yangjin Kim for their contributions. This project was supported by a grant from the Phi Zeta Chapter of the University of MissouriColumbia College of Veterinary Medicine.

Sources and manufacturers

a. Remel Laboratories, Santa Fe Trail Drive, Lenexa, KS. b. Mitsubishi Pack-Anaeroy, Mitsubishi Gas Chemical America, Inc., New York, NY. c. AP-80 Sensititre Gram Negative Identication Systemt, TREK Diagnostic Systems, Inc., Westlake, OH. d. Sensititret, TREK Diagnostic Systems, Inc., Westlake, OH. e. Sensititre Arist automated read-incubation system, TREK Diagnostic Systems, Inc., Westlake, OH. f. SensiTouchy Reader, TREK Diagnostic Systems, Inc., Westlake, OH.

References
1. Brown SA: 1996, Fluoroquinolones in animal health. J Vet Pharmacol Therap 19:114. 2. Cooke CL, Singer RS, Jang SS, Hirsh DC: 2002, Enrooxacin resistance in Escherichia coli isolated from dogs with urinary tract infections. J Am Vet Med Assoc 220:190192. 3. Giraud E, Brisabois A, Martel JL, Chaslus-Dancla E: 1999, Comparative studies of mutations in animal isolates and experimental in vitro- and in vivo-selected mutants of Salmonella spp. suggest a counterselection of highly uoroquinolone-resistant strains in the eld. Antimicrob Agents Chemother 43:2131 2137. 4. Goettsch W, van Pelt W, Nagelkerke N, et al.: 2000, Increasing resistance to uoroquinolones in Escherichia coli from urinary tract infections in the Netherlands. J Antimicrob Chemother 46: 223228. 5. Hoffman SB: 2001, Mechanisms of antibiotic resistance. Compend Contin Educ Pract Vet 23:464473. 6. Kung K, Riond JL, Wanner M: 1993, Pharmacokinetics of enrooxacin and its metabolite ciprooxacin after intravenous and oral administration of enrooxacin in dogs. J Vet Pharmacol Ther 16:462468.

7. Ling GV, Norris CR, Franti CE, et al.: 2001, Interrelations of organism prevalence, specimen collection method, and host age, sex, and breed among 8,354 canine urinary tract infections (19691995). J Vet Intern Med 15:341347. 8. Livermore DM, James D, Reacher M, et al.: 2002, Trends in uoroquinolone (ciprooxacin) resistance in Enterobacteriaceae from bacteremias, England and Wales, 19901999. Emerg Infect Dis 8:473478. 9. Malorny B, Schroeter A, Helmuth R: 1999, Incidence of quinolone resistance over the period 1986 to 1998 in veterinary Salmonella isolates from Germany. Antimicrob Agents Chemother 43:22782282. 10. National Committee for Clinical Laboratory Standards (NCCLS): 1997, Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 4th ed., approved standard M7-4, Wayne, PA. 11. Neu HC: 1988, Bacterial resistance to uoroquinolones. Rev Infect Dis 10:S57S63. 12. ODonnell JA, Gelone SP: 2000, Fluoroquinolones. Infect Dis Clin N Am 14:489513. 13. Oluoch AO, Kim CH, Weisiger RM, et al.: 2001, Nonenteric Escherichia coli isolates from dogs: 674 cases (19901998). J Am Vet Med Assoc 218:381384. 14. Patten VH, Shin SJ, Cole J, et al.: 1995, Evaluation of a commercial automated system and software for the identication of veterinary bacterial isolates. J Vet Diagn Invest 7:506508. 15. Piddock LJ: 1998, Fluoroquinolone resistance: overuse of uoroquinolones in human and veterinary medicine can breed resistance. Br Med J 317:10291030. 16. Piddock LJ: 1999, Mechanisms of uoroquinolone resistance: an update 19941998. Drugs 58S2:1118. 17. Piddock LJ, Jin YF, Ricci V, Asuquo AE: 1999, Quinolone accumulation by Pseudomonas aeruginosa, Staphylococcus aureus and Escherichia coli. J Antimicrob Chemother 43:6170. 18. Pieroni P, Goodfellow J, Reesor L, et al.: 1997, Antimicrobial susceptibilities of blood culture isolates obtained before and after the introduction of ciprooxacin. J Antimicrob Chemother 39:419422. 19. Polzin DJ: 1999, Therapy of canine and feline urinary tract infections with enrooxacin. Compend Contin Educ Pract Vet 21: 6572. 20. Quinn PJ, Carter ME, Markey BK, Carter GR: 1994, Microscopy, culture, and identication. In: Clinical veterinary microbiology, ed. Quinn PJ, Carter ME, Markey BK, Carter GR, pp. 2167. Mosby Year Book Europe, Ltd., London, UK. 21. Thomson CJ: 1999, The global epidemiology of resistance to ciprooxacin and the changing nature of antibiotic resistance: a 10 year perspective. J Antimicrob Chemother 43(Suppl.)A:31 40. 22. Vromen M, van der Ven AJ, Knols A, Stobberingh EE: 1999, Antimicrobial resistance patterns in urinary isolates from nursing home residents. Fifteen years of data reviewed. J Antimicrob Chemother 44:113116. 23. Walker RD, Thornsberry C: 1998, Decrease in antibiotic susceptibility or increase in resistance? J Antimicrob Chemother 41:14.

Downloaded from vdi.sagepub.com by guest on December 21, 2013