The effect of boric acid on bacterial culture of canine and feline urine. M. Rowlands, L. Blackwood, A. Mas, P. Cripps*, C. Crompton and R.

ttp://www.bsava.com/ PETSAVERS PAPER The effect of boric acid on bacterial culture of canine and feline urine M. Rowlands, L. Blackwood, A. Mas, P. Cripps*, C. Crompton and R. Burrow Small Animal Teaching Hospital, University of Liverpool, Chester High Road, Neston, Cheshire CH64 7TE *Department of Epidemiology and Population Health, Institute of Infection and Global Health, Faculty of Health and Life Sciences, University of Liverpool, Leahurst, Neston, Cheshire CH64 7TE Idexx Laboratories, Grange House, Sandbeck Way, Wetherby, West Yorkshire LS22 7DN C. Crompton s current address is CTDS Ltd, Blacksmiths Forge, Brookfield Farm, Selby Road, Garforth, West Yorkshire LS25 1NB OBJECTIVE: To identify the optimal method of submission of canine and feline urine for bacterial culture. METHODS: Cystocentesis samples from 250 animals (200 dogs, 50 cats) suspected of having urinary tract infections were collected. The reference aliquot, without preservative, was processed on site within 2 hours. Two further aliquots (one without preservative, one with boric acid) were stored at room temperature for up to 7 hours and then posted by guaranteed next day delivery to a commercial laboratory for analysis. RESULTS: Forty-seven of the samples were positive on culture in the reference test. There was no significant difference between reference test results and those of samples posted without preservative (P=0 39), but samples posted in boric acid were significantly less likely to give a positive result (P=0 01). Samples posted without preservative had a sensitivity of 82% and a specificity of 98%; for boric acid, sensitivity was 73% and specificity 99%. CLINICAL SIGNIFICANCE: Postal urine samples should be submitted to the laboratory in a plain sterile tube. Journal of Small Animal Practice (2011) 52, 510 514 DOI: 10.1111/j.1748-5827.2011.01102.x Accepted: 31 May 2011 INTRODUCTION Bacterial urinary tract infections (UTIs) affect 14% of all dogs during their lifetime (Ling 1984) and are more common in females (Thomsen and others 1986, Lees 1996, Ling and others 2001). UTIs occur less commonly in cats and are more frequently diagnosed in cats older than 10 years (Lees 1984, Lekcharoensuk and others 2001). In veterinary patients, collection of urine for culture by cystocentesis is considered the gold standard because it avoids bacterial contamination from the distal urogenital tract (Carter and others 1978, Comer and Ling 1981). To avoid false-positive and false-negative results, immediate culture is recommended (Hindman and others 1976). However, in the UK few veterinary practices perform urine culture in-house and samples must usually be posted to the laboratory. The optimal submission method must minimise misleading results, by restricting growth of contaminants while preserving pathogens. The role of refrigeration and preservatives is poorly explored in veterinary patients. Lees and Osborne (1979) showed that following collection of urine and storage at room temperature, bacteria may double every 20 to 45 minutes. Multiplication or destruction of bacteria may occur within 1 hour of collection (Hindman and others 1976, Padilla and others 1981), and for this reason it was recommended that urine samples be refrigerated within 1 or 2 hours of collection if the sample cannot be cultured immediately (Padilla and others 1981). In practice, samples are not usually refrigerated during transportation, which could lead to misleading results (Jefferson and others 1975, Shrestha 1975, Hindman and others 1976, Wheldon and Slack 1977). However, refrigeration can also reduce bacterial growth, resulting in false-negative results (Watson and Duerden 1977, Padilla and others 1981). An effective method of preserving urine at room temperature is therefore desirable. In human medicine, several studies have reported the addition of preservatives, including boric acid, to urine samples obtained by free catch (Porter and Brodie 1969, Lauer and others 1979). Boric acid acts as a chemical preservative and antibacterial agent to reduce overgrowth of bacterial contaminants and thus reduces the number of false-positive results (Porter and Brodie 510 Journal of Small Animal Practice Vol 52 October 2011 2011 British Small Animal Veterinary Association

Effect of boric acid on urine culture 1969, Lauer and others 1979, Guenther and Washington 1981, Lum and Meers 1989, Meers and Chow 1990). However, boric acid may have an inhibitory effect on bacterial growth resulting in false-negative cultures (Watson and Duerden 1977, Nickander and others 1982, Hubbard and others 1983, Meers and Chow 1990, Gillespie and others 1999). There are limited reports on the use of preservatives in veterinary medicine. Allen and others (1987) demonstrated the effectiveness of boric acid in preserving canine urine for bacterial culture collected by various methods including cystocentesis for up to 72 hours before processing. Immediately after collection, however, these samples were refrigerated at 4 C, which may have affected the results. Perrin and others (1992) obtained samples by both catheterisation and free catch and these were stored at room temperature for 24 to 48 hours in plain tubes or with boric acid added. They concluded that the addition of boric acid was a good method of preserving urine at room temperature and avoided the high percentage of false-positive results when urine was stored in the plain tubes. These data cannot be applied to samples collected by cystocentesis, where the risk of contamination is already much reduced. The objective of this study was to determine whether the postal submission of canine and feline urine, obtained by cystocentesis, with boric acid added as a preservative achieves more reliable bacterial culture results compared to the submission of urine in plain tubes. MATERIALS AND METHODS Patients Urine samples were collected from canine and feline patients presenting at site 1 (Small Animal Teaching Hospital, University of Liverpool) between June 2008 and March 2010. All patients were suspected of having an UTI on the basis of history and/or clinical findings. Signalment data, clinical diagnosis and concurrent medication were recorded for each patient. Sampling All samples were collected by cystocentesis. Specimens were immediately divided into three aliquots and transferred from the collection syringe directly into the appropriate collection tubes. For the reference sample, 1 ml of urine was placed into a sterile tube without preservative. Urine culture was performed on site within 1 hour of collection. With the remaining urine, 1 ml was placed in a sterile plain tube and 2 ml in a sterile container containing the preservative boric acid. The amount of boric acid was adjusted accordingly to achieve a concentration of 10 to 20 g/l. These two aliquots were stored at room temperature for up to 7 hours and then posted by guaranteed next day delivery to site 2 [Idexx Laboratories, UK (UKAS reference 17025 accredited)], where culture was performed. For the reference culture, quantitative urine cultures were performed by inoculating 1 µl of urine onto 5% sheep blood agar and CLED (cystine-lactose-electrolyte Deficient) agar followed by aerobic incubation at 37 C for 24 to 48 hours. A bacterial count of greater than 10³ colony-forming units/ml was considered diagnostic of UTI. If no growth occurred after 48 hours, cultures were categorised as negative. Identification of bacterial isolates at genus level was based on cultural and Gram-stained microscopic morphology; species identification was performed using API systems (API-BioMerieux, France) for all significant isolates. For the posted samples, sterile 1 µl urine was inoculated onto 5% sheep blood and MacConkey agar plates. These were incubated aerobically at 35 to 37 C for 18 to 24 hours with a final report at 48 hours. Samples with no growth were reported as negative. Identification of bacterial isolates was as above, with the addition of Vitek 1 (Biomerieux, France) when appropriate. Statistical analysis Data was analysed using Minitab software (Minitab Inc., State College, PA, USA) with the in-house urine culture used as the reference test. Paired samples of the reference test with the external sample without preservative and then with the boric acid sample were compared using a sign test for the median. P<0 05 was considered significant. The sensitivities and specificities for detecting a positive culture using plain and boric acid tubes for delayed urine cultures were derived. RESULTS Culture results Urine specimens were obtained by cystocentesis from 200 dogs (20 female entire, 65 female neutered, 49 male entire and 66 male neutered) and 50 cats (2 female entire, 18 female neutered and 30 male neutered). Forty-seven of the samples were positive in the reference culture. Thirty-eight of these positive samples came from dogs, of which 8 were female entire, 13 female neutered, 8 male entire and 9 male neutered with a median age of seven years ranging from 9 months to 14 years. The breeds represented more than once included: boxer (five), Labrador retriever (four), crossbreed (three), West Highland white terrier (three), Staffordshire bull terrier (two), Bernese mountain dog (two), Border collie (two) and American bulldog (two). Nine of the positive samples came from cats, of which four were female neutered and five male neutered with a median age of five years, ranging from 2 to 11 years. The breeds of cats included five domestic shorthair and one Burmese, Persian, British shorthair and domestic longhair cat. Three of the reference-positive cultures did not have culture results available for the postal samples leaving 247 paired samples with results available from both laboratories. Eight of the 44 samples with positive reference cultures had negative cultures after postal submission without preservative, giving a sensitivity of 82% [95% confidence interval (CI) 68 to 92%]. Two of the false-negative samples were from cats. Four of the 203 negative reference cultures had false-positive results after postal submission, giving a specificity of 98% (95% CI 95 to 99 5%). Two false-positive results were from cats. Journal of Small Animal Practice Vol 52 October 2011 2011 British Small Animal Veterinary Association 511

M. Rowlands and others Twelve of the 44 samples that had positive reference cultures were negative after postal submission with boric acid preservative, giving a sensitivity of 73% (CI 58 to 85%). Two of the falsenegative samples were from cats. Two of the 203 samples that had a negative reference culture had false-positive results after postal submission in boric acid, giving a specificity of 99% (CI 96 to 99 9%). One false-positive result was from a cat. There was no significant difference in positive cultures between reference cultures and those of samples posted without preservative (P=0 39). In contrast, samples posted in boric acid were significantly less likely to give a positive result (P=0 01). Bacteria isolated Bacterial results are listed in Table 1. There were 38 positive results in dogs in the reference cultures. In 32 cases (84%) a single bacterial species was isolated and in 6 (16%) cases two co-infecting species were isolated. In cats, nine of the reference cultures were positive and in all of them a single bacterial species was isolated. The bacterial species most frequently isolated was Escherichia coli (70% of the positive reference cultures). In samples 16, 23, 33, 35, 37, 42 and 45, the reference culture yielded Staphylococcus in two samples, Escherichia in three samples, Bacillus and Streptococcus in one sample each and Escherichia and Streptococcus in one sample, but following overnight transport no bacterial growth was obtained from either plain or boric acid. There were some discrepancies in the bacteria isolated between the reference and posted samples with positive cultures. In three of the posted samples (6, 9 and 49), an additional bacterial species was cultured compared to the reference culture. In two of the reference cultures which contained two bacterial species, only one was isolated in the posted samples (samples 1 and 7). In three cases (10, 14 and 17) a completely different bacterial species was isolated and in one case (22) both postal cultures yielded a mixed population rather than a single species of bacteria. In general, there was good correlation between the bacteria isolated in both the preserved and non-preserved posted samples. In cases 39, 50 and 51 where the reference culture was negative, Escherichia was isolated in the posted samples and in case 40 that was also negative in the reference culture, Enterococcus was isolated after overnight transport. DISCUSSION Although refrigeration of urine samples has been recommended during transportation to the laboratory to avoid false-positive and false-negative bacterial culture results (Hindman and others 1976, Carter and others 1978, Padilla and others 1981), this is rarely performed in practice and is of uncertain value. A reliable and more practical method of preserving urine at room temperature is desirable. Boric acid has been suggested as a cheap, effective and simple method of preservation. Our study, however, showed that there is no benefit in adding boric acid to urine collected by cystocentesis when submitted to the laboratory for bacterial culture with transportation at room temperature and a delay of 24 hours prior to processing. Ling and others (2001) isolated a single pathogen from approximately 75% of dogs with UTI. In a further 20% of dogs, the UTI was caused by two co-infecting species. Similar results were obtained in our study. As previously reported, E. coli was the most frequently cultured pathogen (Wooley and Blue 1976, Ling and others 2001) followed by Gram-positive cocci. A similar frequency in the number of infecting organisms has been previously reported in cats (Davidson and others 1992) and was also found in this study. Posted samples produced a large number of false-negative results compared to the reference cultures, and with the addition of boric acid it was significantly more likely that a false-negative result would be obtained. Indeed, at the concentrations used in our study (10 to 20 g/l), it is reported to be an effective means of preserving human urine (Nickander and others 1982, Lum and Meers 1989, Meers and Chow 1990). However, in the current study, the preservative was associated with an insignificant reduction in false-positive results, but a significant increase in false negatives. This is likely to result from non-specific toxic effects of boric acid on pathogenic bacteria, which has previously been reported (Watson and Duerden 1977, Nickander and others 1982, Hubbard and others 1983, Meers and Chow 1990, Gillespie and others 1999). The toxicity of boric acid is concentration dependent (Nickander and others 1982) and although we attempted to standardise the concentration of boric acid, some variability was unavoidable. We do, however, believe that the variability in the concentration of boric acid was limited, and any variability in this will reflect the situation in practice. Some of the bacteria isolated in the reference culture are not typically considered urinary pathogens (Bacillus species, Raoultella terrigena, Lactobacillus species); therefore, those most likely represent contaminants and do not represent true discrepancies between the laboratories. In our study, there were no significant differences in the number of false-positive results using plain or boric acid tubes for delayed urine cultures. This is in contrast to previous studies where the use of plain tubes resulted in an increase in false positives, so was not recommended (Jefferson and others 1975, Shrestha 1975, Hindman and others 1976, Watson and Duerden 1977, Wheldon and Slack 1977, Padilla and others 1981). In these previous studies, however, the urine samples were obtained by catheterisation or voiding and thus a greater number of contaminants were likely present in the urine at the time of collection. In contrast, the urine samples in our study were obtained by cystocentesis reducing the amount of contaminants present at the time of collection. In most of the false-positive results, the same results were obtained from both the plain and the boric acid tubes, indicating that most likely the samples were contaminated during initial handling and that the bacteria multiplied during transportation. Ideally, we would have used only one laboratory to perform all bacterial cultures. This was not possible at the time because our on-site laboratory, although able to perform immediate urine culture for the reference results, did not perform minimum 512 Journal of Small Animal Practice Vol 52 October 2011 2011 British Small Animal Veterinary Association

Effect of boric acid on urine culture Table 1. Culture results for the reference and posted urine in plain and boric acid Sex Age (years) Breed Reference Posted plain Posted boric 1 FE 7 Cross Escherichia coli,lactobacillus Escherichia coli Escherichia coli species 2 MN 10 WHWT Escherichia coli Escherichia coli Escherichia coli 3 FE 4 Boxer Escherichia coli Escherichia coli Escherichia coli 4 ME 5 OESD klebsiella pneumoniae No results No results 5 FE 2 Dogues de bordeaux Escherichia coli Escherichia coli Escherichia coli 6 MN 3 Boxer Escherichia coli Escherichia coli, Enterococcus species 7 FE 5 American bulldog Raoultella terrigina, Escherichia coli, Enterococcus species 8 FE 9 months Border terrier No growth 9 FE 4 Alaskan malamute Proteus mirabillis, Enterobacter aerogeues 10 FN 5 Greyhound Proteus mirabillis, Escherichia coli Proteus mirabillis, Proteus mirabillis, No growth 11 ME 14 Border collie Escherichia coli No results No results 12 FN 7 Beagle Escherichia coli Escherichia coli No growth 13 FN 11 Border collie Escherichia coli Escherichia coli Escherichia coli 14 FN 6 WHWT Staphylococcus intermedius No growth 15 MN 8 Schnauzer Escherechia coli No results No results 16 FN 2 CKCS Staphylococcus intermedius No growth No growth 17 FE 6 American bulldog Escherichia coli 18 FN 2 Boxer Escherichia coli Escherichia coli No growth 19 ME 12 Labrador Escherichia coli Escherichia coli Escherichia coli 20 FN 8 Labrador Staphylococcus intermedius Staphylococcus intermedius Staphylococcus intermedius 21 FN 7 Labrador Escherichia coli Escherichia coli Escherichia coli 22 FE 1 Labrador Escherichia coli Mixed growth Mixed growth 23 ME 9 SBT Escherichia coli No growth No growth 24 MN 8 Bernese Escherichia coli Escherichia coli Escherichia coli 25 MN 8 Bernese Escherichia coli Escherichia coli No growth 26 FN 11 Springer Escherichia coli Escherichia coli Escherichia coli 27 ME 10 SBT Escherichia coli Escherichia coli Escherichia coli 28 ME 10 Lhasa apso Klebsiella pnemoniae Klebsiella pnemoniae Klebsiella pnemoniae 29 ME 7 Cocker spaniel Staphylococcus intermedius Staphylococcus intermedius Staphylococcus intermedius 30 FN 3 Basset Escherichia coli Escherichia coli Escherichia coli 31 MN 10 Cross Escherichia coli Escherichia coli Escherichia coli 32 MN 11 Cross Escherichia coli Escherichia coli Escherichia coli 33 MN 8 WHWT Escherichia coli No growth No growth 34 MN 5 Newfoundland Escherichia coli Escherichia coli Escherichia coli 35 FN 10 Rottweiler Escherichia, Streptococcus No growth No growth canis 36 FN 9 Boxer Escherichia coli Escherichia coli Escherichia coli 37 ME 3 Bullmastiff Escherichia coli No growth No growth 38 FN 6 Boxer Proteus species Proteus species Proteus species 39 MN 10 CKCS No growth Escherichia coli Escherichia coli 40 FE 5 Rottweiler No growth No growth 41 MN 3 DSH Escherichia coli Escherichia coli Escherichia coli 42 FN 10 Burmese Bacillus species No growth No growth 43 MN 5 DLH Escherichia coli Escherichia coli Escherichia coli 44 MN 5 Persian Escherichia coli Escherichia coli Escherichia coli 45 FN 6 DSH Staphylococcus species No growth No growth 46 FN 5 DSH Escherichia coli Escherichia coli Escherichia coli 47 MN 9 DSH Salmonella species Salmonella species Salmonella species 48 MN 11 BSH Escherichia coli Escherichia coli Escherichia coli 49 FN 2 DSH Enterococcus species Enterococcus species, Staphylococcus species Enterococcus species, Staphylococcus species 50 MN 6 DSH No growth Escherichia coli No growth 51 MN 5 Persian No growth Escherichia coli Escherichia coli BSH Burmese shorthair, CKCS Cavalier King Charles spaniel, DLH Domestic longhair, DSH Domestic shorthair, FE Female entire, FN Female neutered, ME Male entire, MN Male neutered, OESD Old English sheepdog, SBT Staffordshire bull terrier, WHWT West Highland white terrier Journal of Small Animal Practice Vol 52 October 2011 2011 British Small Animal Veterinary Association 513

M. Rowlands and others inhibitory concentration testing required by our hospital clinicians. Also, even though there was a large number of urine samples submitted, a positive reference culture was obtained in only 47 cases. On completion of our study and data analysis, we realised that insufficient cats were included in our study, to allow any conclusions to be made about the addition of boric acid on bacterial growth within urine in this species. Conclusions This study suggests that the addition of boric acid to urine collected by cystocentesis does not increase the reliability of bacterial culture results when culture is delayed by the transportation of urine at room temperature. In fact, the addition of boric acid significantly increases false-negative results. There was no statistically significant difference between reference results and results from plain postal samples, although sensitivity and specificity were reduced compared to immediate culture. Minor differences in the results between laboratories could have been due to the delay caused by postage of the samples. Acknowledgements The authors thank BSAVA and Petsavers for kindly providing the financial support and Idexx Laboratories for facilitating this project. We also acknowledge the diagnostic imagers working at the Small Animal Teaching Hospital who collected the majority of the urine samples under ultrasound guidance and Susan Quinn, Peter Taylor and Dr. Larry Roberts for their assistance with the laboratory work at the Veterinary Pathology Laboratories, University of Liverpool and Idexx laboratories, respectively. Conflict of interest None of the authors of this article has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper. References ALLEN, T. A., JONES, R. L. & PURVANCE, J. (1987) Microbiologic evaluation of canine urine: direct microscopic examination and preservation of specimen quality for culture. Journal of the American Veterinary Medicine Association 190, 1289-1291 CARTER, J. M., KLAUSNER, J. S., OSBORNE, C. A. & BATES, F. Y. (1978) Comparison of collection techniques for quantitative urine culture in dogs. Journal of the American Veterinary Medical Association 173, 296-298 COMER, K. M. & LING, G. V. (1981) Results of urinalysis and bacterial culture of canine urine obtained by antepubic cystocentesis, catheterization, and the midstream voided methods. Journal of the American Veterinary Medical Association 179, 891-895 DAVIDSON, A. P., LING, G. V., STEVENS, F., FRANTI, C. E., JOHNSON, D. L. & LANG, S. S. (1992) Urinary tract infections in cats: a retrospective study 1977-1989. California Veterinarian 46, 32-34 GILLESPIE, T., FEWSTER, J. & MASTERTON, R. G. (1999) The effect of specimen processing delay on borate urine preservation. Journal of Clinical Pathology 52, 95-98 GUENTHER, K. L. & WASHINGTON, J. A., 2ND (1981) Evaluation of the B-D urine culture kit. Journal of Clinical Microbiology 14, 628-630 HINDMAN, R., TRONIC, B. & BARTLETT, R. (1976) Effect of delay on culture of urine. Journal of Clinical Microbiology 4, 102-103 HUBBARD, W. A., SHALIS, P. J. & MCCLATCHEY, K. D. (1983) Comparison of the B-D urine culture kit with a standard culture method and with the MS-2. Journal of Clinical Microbiology 17, 327-331 JEFFERSON, H., DALTON, H. P., ESCOBAR, M. R. & ALLISON, M. J. (1975) Transportation delay and the microbiological quality of clinical specimens. American Journal of Clinical Pathology 64, 689-693 LAUER, B. A., RELLER, L. B. & MIRRETT, S. (1979) Evaluation of preservative fluid for urine collected for culture. Journal of Clinical Microbiology 10, 42-45 LEES, G. E. (1984) Epidemiology of naturally occurring feline bacterial urinary tract infections. Veterinary Clinics of North America Small Animal Practice 14, 471-479 LEES, G. E. (1996) Bacterial urinary tract infections. Veterinary Clinics of North America Small Animal Practice 26, 297-304 LEES, G. E. & OSBORNE, C. A. (1979) Antibacterial properties of urine comparative review. Journal of the American Animal Hospital Association 15, 125-132 LEKCHAROENSUK, C., OSBORNE, C. A. & LULICH, J. P. (2001) Epidemiologic study of risk factors for lower urinary tract diseases in cats. Journal of the American Veterinary Medicine Association 218, 1429-1435 LING, G. V. (1984) Therapeutic strategies involving antimicrobial treatment of the canine urinary tract. Journal of the American Veterinary Medicine Association 185, 1162-1164 LING, G. V., NORRIS, C. R., FRANTI, C. E., EISELE, P. H., JOHNSON, D. L., RUBY, A. L. & JANG, S. S. (2001) Interrelations of organism prevalence, specimen collection method, and host age, sex, and breed among 8,354 canine urinary tract infections (1969-1995). Journal of Veterinary Internal Medicine 15, 341-347 LUM, K. T. & MEERS, P. D. (1989) Boric acid converts urine into an effective bacteriostatic transport medium. Journal of Infection 18, 51-58 MEERS, P. D. & CHOW, C. K. (1990) Bacteriostatic and bactericidal actions of boric acid against bacteria and fungi commonly found in urine. Journal of Clinical Pathology 43, 484-487 NICKANDER, K. K., SHANHOLTZER, C. J. & PETERSON, L. R. (1982) Urine culture transport tubes: effect of sample volume on bacterial toxicity of the preservative. Journal of Clinical Microbiology 15, 593-595 PADILLA, J., OSBORNE, C. A. & WARD, G. E. (1981) Effects of storage time and temperature on quantitative culture of canine urine. Veterinary Clinics of North America Small Animal Practice 178, 1077-1081 PERRIN, J. & NICOLET, J. (1992) Influence of the transport on the outcome of the bacteriological analysis of a dog urine comparison of three transport tubes. Zentralblatt für Veterinarmedizin: Reihe B 39, 662-667 PORTER, I. A. & BRODIE, J. (1969) Boric acid preservation of urine samples. British Medical Journal 2, 353-355 SHRESTHA, T. L. (1975) Effects of delayed culture on semiquantitative urinary bacteriology results. Journal of Clinical Pathology 28, 392-393 THOMSEN, M. K., SVANE, L. C. & POULSEN, P. H. (1986) Canine urinary tract infection. Detection, prevalence and therapeutic consequences of bacteriuria. Nord Veterinary Medicine 38, 394-402 WATSON, P. G. & DUERDEN, B. I. (1977) Laboratory assessment of physical and chemical methods of preserving urine specimens. Journal of Clinical Pathology 30, 532-536 WHELDON, D. B. & SLACK, M. (1977) Multiplication of contaminant bacteria in urine and interpretation of delayed culture. Journal of Clinical Pathology 30, 615-619 WOOLEY, R. E. & BLUE, J. L. (1976) Bacterial isolations from canine and feline urine. Modern Veterinary Practice 57, 535-538 514 Journal of Small Animal Practice Vol 52 October 2011 2011 British Small Animal Veterinary Association