Approximately 14% of dogs will have at least 1 UTI during their lifetime.1,2 The gold standard for diagnosis of bacterial UTI is QBC of urine, which allows veterinarians to make informed therapeutic decisions.3,4 Current guidelines set forth by the International Society for Companion Animal Infectious Diseases recommend that aerobic bacterial culture and susceptibility testing be performed on urine collected by cystocentesis from animals with simple or complicated UTIs.5
Sample handling affects culture results, and optimal QBC results are consistently obtained when urine samples are processed (ie, inoculated onto agar plates) immediately after collection.6 However, immediate inoculation is impractical for private practitioners, given that urine samples are routinely submitted to external laboratories for bacterial culture. Samples are often collected and held for courier pickup or shipped overnight. Such delays may impact culture results. Sample storage temperature is another critical variable that can affect QBC results. Indeed, a study6 revealed that urine samples stored in plain sterile tubes at room temperature for 24 hours prior to plate inoculation were diagnostically unreliable, with false-positive results obtained for 50% (13/26) of samples.
To mitigate the deleterious effects of storage duration and temperature on QBC results, tubes containing chemical preservatives can be used, providing an attractive alternative to plain sterile tubes. Commercially available tubes containing a buffered boric acid mixture (sodium formate, sodium borate, and boric acid) have been validated for storage of human urine samples prior to QBC. According to the manufacturer,7 urine samples may be stored at room temperature in such tubes for 48 hours without compromising test results and refrigeration will not adversely affect sample stability.
Given the many variables that can affect the reliability of QBC of urine and the consequences of inappropriate sample handling on patient care, veterinary practitioners have reason to choose transport options carefully. The objective of the study reported here was to compare bacterial concentrations in sterile urine samples that were spiked with Escherichia coli, transferred to traditional SCTs or commercially available UTTs, and stored for 0, 8, or 24 hours at 25°C (77.0°F; representing room temperature) and 4°C (39.2°F; standard refrigeration temperature).
Materials and Methods
Urine was collected from dogs after euthanasia in an unrelated project via catheterization (1 male) and manual expression (3 females). The dogs had been cared for in accordance with an animal use protocol approved by the Texas A&M University Institutional Animal Care and Use Committee. A total urine volume of approximately 200 mL was collected, pooled, and stored at 4°C overnight. The urine was then filter sterilized twice. Initially, gross debris was removed with filter papera and the filtrate was passed through a 0.2-μm filter sterilization device.b A 7-mL sample of filter-sterilized urine was submitted for routine urinalysis at the Texas A&M Veterinary Clinical Pathology Laboratory. Urine specific gravity was measured by refractometry.c Urine dipstick analysis was performed by use of a commercially available reagent strip.d Microscopic sediment examination was performed by a trained laboratory technician, who centrifuged 3 mL of the sterilized urine for 5 minutes at 500 × g and then evaluated the sediment by means of a standardized microscopic urinalysis system.e The remaining sterilized urine was stored at 4°C overnight, then divided into six 10-mL aliquots the next day.
Each aliquot of sterile urine was inoculated with bacteria from 1 of 6 independent cultures of Escherichia coli (ATCC 25922) to achieve a final concentration of approximately 105 CFUs/mL. Bacterial counts were performed by making a 10-fold dilution series of the urine, inoculating 100 μL of each dilution onto a blood agar plate (trypticase),f incubating the plate overnight (approx 18 hours) at 37°C, and manually counting the colonies. The dilution that yielded between 30 and 300 colonies was used to determine the total bacterial count for each sample by multiplying the number of colonies by the dilution factor. Bacterial concentration for each of the 6 urine aliquots prior to incubation was confirmed through the same process. Bacterial suspensions were vortex mixed, then 1 mL of urine from each of the 6 aliquots was transferred via sterile technique into 5 plain SCTsg and 5 UTTsh that contained a preservative (combination of sodium formate, sodium borate, and boric acid). Tubes were allocated to be processed for QBC immediately (0 hours) or stored for 8 or 24 hours prior to plate inoculation. Tubes allocated to be stored for 8 hours or 24 hours were further separated into 2 groups and stored at room temperature (25°C) or placed in a refrigerator (4°C). At each assessment point, a 100-μL sample was removed from each tube for QBC, yielding a total of 12 tubes for the 0-hour group, 24 tubes for the 8-hour group, and 24 tubes for the 24-hour group.
Bacterial concentrations in each urine sample were estimated by creating a 10-fold dilution series and inoculating 100 μL of each dilution onto trypticase soy agar supplemented with 5% sheep's blood (blood agar plate).f The plates were incubated overnight (approx 18 hours) at 37°C in 5% carbon dioxide.i Colonies were manually counted the next day and used to determine the concentration of viable E coli in urine samples at the time of inoculation.
Statistical analysis
A power calculation was performed, with the intention to detect a difference of 5,000 CFUs/mL between 2 populations with an SD of 1,000 (α = 0.05). That calculation revealed that a sample size of 4 would yield a power of 100%. Data were analyzed with a commercially available software program.j All data were tested for a normal distribution by use of the D'Agostino-Pearson normality test, which revealed a departure from normality. Therefore, median bacterial concentrations for both transport tube types (SCT and UTT) for each of the 6 independent E coli cultures at the various storage durations and temperatures were compared with the Wilcoxon signed rank test. Differences in median bacterial concentrations within each tube type were evaluated with the Friedman test. Post hoc testing to further define the differences identified was performed with the Dunn multiple comparison test.
RESULTS
The specific gravity of the pooled canine urine after filter sterilization was 1.034. No abnormalities were identified by dipstick analysis, and results of a sulfosalicylic acid test were negative for protein. Bacterial concentrations for tubes containing bacteria from 1 of 6 independent E coli cultures ranged from 72,000 to 162,000 CFUs/mL.
No significant differences in bacterial concentration were identified for urine samples stored in SCTs or UTTs at 4°C or 25°C for 0 or 8 hours. However, significant differences were evident for samples stored for 24 hours. Median bacterial concentration in urine samples stored for 24 hours at 4°C in UTTs was significantly (P = 0.03) lower than that in urine samples stored at the same temperature and duration in SCTs (Table 1). Additionally, after 24 hours of storage at 25°C, there was a significant (P = 0.03) decrease in median bacterial concentration in samples stored in SCTs, compared with median bacterial concentration for samples stored in UTTs.
Median (interquartile range [25th to 75th percentile]) bacterial concentrations in canine urine samples* after storage in SCTs or UTTs for various durations and temperatures.
| Storage duration (h) | Temperature (°C) | SCT (CFUs/mL) | UTT (CFUs/mL) | P value |
|---|---|---|---|---|
| 0 | — | 98,500 (51,000–143,000) | 113,500 (50,000–143,000) | 0.10 |
| 8 | 4 | 84,000 (46,000–112,000) | 93,500 (51,000–119,000) | 0.06 |
| 8 | 25 | 101,000 (72,000–119,000) | 86,000 (68,000–95,000) | 0.16 |
| 24 | 4 | 61,000 (55,000–76,000) | 36,500 (27,000–98,000) | 0.03 |
| 24 | 25 | 275 (0–1,000) | 18,400 (10,000–30,000) | 0.03 |
Urine samples used for the study consisted of pooled urine from 4 dogs that was sterilized and then spiked with bacteria from 1 of 6 independent Escherichia coli cultures to achieve final sample bacterial concentrations of 105 CFUs/mL.
— = Not applicable.
The Friedman test revealed a significant (P = 0.002) decrease in urine bacterial concentration associated with storage duration and temperature when SCTs were used. Posttest results indicated a significant (P < 0.05) difference in median bacterial concentration between urine samples stored for 0 hours and those stored for 24 hours at 25°C, as well as between samples stored for 8 hours at 25°C and those stored for 24 hours at 25°C (Figure 1).
Bacterial concentrations in individual SCT-stored canine urine samples immediately after processing (0 hours) or after storage at 4°C or 25°C for 8 or 24 hours (6 samples/condition). Urine samples used for the study consisted of pooled urine from 4 dogs that was sterilized and then spiked with bacteria from 1 of 6 independent Escherichia coli cultures to achieve final sample bacterial concentrations of 105 CFUs/mL. Horizontal lines indicate median concentration for each storage condition. Posttest results indicated a significant (P < 0.05) difference in median bacterial concentration between urine samples stored for 0 hours and those stored for 24 hours at 25°C as well as between samples stored for 8 hours at 25°C and those stored for 24 hours at 25°C.
Citation: Journal of the American Veterinary Medical Association 248, 2; 10.2460/javma.248.2.183
Bacterial concentrations in individual SCT-stored canine urine samples immediately after processing (0 hours) or after storage at 4°C or 25°C for 8 or 24 hours (6 samples/condition). Urine samples used for the study consisted of pooled urine from 4 dogs that was sterilized and then spiked with bacteria from 1 of 6 independent Escherichia coli cultures to achieve final sample bacterial concentrations of 105 CFUs/mL. Horizontal lines indicate median concentration for each storage condition. Posttest results indicated a significant (P < 0.05) difference in median bacterial concentration between urine samples stored for 0 hours and those stored for 24 hours at 25°C as well as between samples stored for 8 hours at 25°C and those stored for 24 hours at 25°C.
Citation: Journal of the American Veterinary Medical Association 248, 2; 10.2460/javma.248.2.183
Bacterial concentrations in individual SCT-stored canine urine samples immediately after processing (0 hours) or after storage at 4°C or 25°C for 8 or 24 hours (6 samples/condition). Urine samples used for the study consisted of pooled urine from 4 dogs that was sterilized and then spiked with bacteria from 1 of 6 independent Escherichia coli cultures to achieve final sample bacterial concentrations of 105 CFUs/mL. Horizontal lines indicate median concentration for each storage condition. Posttest results indicated a significant (P < 0.05) difference in median bacterial concentration between urine samples stored for 0 hours and those stored for 24 hours at 25°C as well as between samples stored for 8 hours at 25°C and those stored for 24 hours at 25°C.
Citation: Journal of the American Veterinary Medical Association 248, 2; 10.2460/javma.248.2.183
Evaluation of urine samples stored in UTTs at the 2 temperatures revealed a significant (P = 0.001) and substantial decrease in bacterial concentration over time. Posttest results indicated significant differences between urine samples stored for 0 hours and those stored for 24 hours at 4°C or 25°C (Figure 2). Similarly, median bacterial concentration in urine samples stored for 24 hours at 25°C was significantly lower than that in samples stored for 8 hours at 4°C.
Bacterial concentrations in individual UTT-stored canine urine samples immediately after processing (0 hours) or after storage at 4°C or 25°C for 8 or 24 hours. Posttest results indicated significant (P < 0.05) differences between urine samples stored for 0 hours and those stored for 24 hours at 4°C or 25°C. Similarly, median bacterial concentration in samples stored for 24 hours at 25°C was significantly (P < 0.05) lower than that in samples stored for 8 hours at 4°C. See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 248, 2; 10.2460/javma.248.2.183
Bacterial concentrations in individual UTT-stored canine urine samples immediately after processing (0 hours) or after storage at 4°C or 25°C for 8 or 24 hours. Posttest results indicated significant (P < 0.05) differences between urine samples stored for 0 hours and those stored for 24 hours at 4°C or 25°C. Similarly, median bacterial concentration in samples stored for 24 hours at 25°C was significantly (P < 0.05) lower than that in samples stored for 8 hours at 4°C. See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 248, 2; 10.2460/javma.248.2.183
Bacterial concentrations in individual UTT-stored canine urine samples immediately after processing (0 hours) or after storage at 4°C or 25°C for 8 or 24 hours. Posttest results indicated significant (P < 0.05) differences between urine samples stored for 0 hours and those stored for 24 hours at 4°C or 25°C. Similarly, median bacterial concentration in samples stored for 24 hours at 25°C was significantly (P < 0.05) lower than that in samples stored for 8 hours at 4°C. See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 248, 2; 10.2460/javma.248.2.183
Discussion
Results of the present study indicated that, regardless of storage temperature, median concentration of E coli (strain ATCC 25922) in urine samples decreased significantly over a 24-hour period when samples were stored in SCTs or UTTs. Most importantly, median bacterial concentration in urine samples stored in SCTs for 24 hours at 25°C was lower than the established limit for clinically important bacteriuria (< 1,000 CFUs/mL), potentially resulting in a failure to identify an active UTI.1 However, for UTTs, the median bacterial concentration was greater than that clinical limit in urine samples stored for 24 hours at either 4°C or 25°C, and no significant difference was identified in median bacterial concentration between urine samples stored at 4°C and those stored at 25°C.
Although urine samples used in the present study were not actual clinical samples, results indicated that canine urine samples stored in SCTs or UTTs may be left at room temperature or kept refrigerated for up to 8 hours without significant changes in bacterial concentration. Results also suggested that when a urine sample is left in an SCT at room temperature for 24 hours, QBC findings may not reflect original bacterial concentration. However, urine samples stored in SCTs and refrigerated for 24 hours may be reliably used for QBC, as might occur when the QBC is requested following arrival of a urine sample at a reference laboratory. Research regarding the effect of storage conditions on QBC results for urine samples from dogs with bacterial UTIs is needed to validate these clinical suggestions.
Quantification of the amount of bacteria in a urine sample allows veterinary practitioners to distinguish a true infection from specimen contamination. It also allows objective differentiation of UTI relapse from reinfection.5 The clinical importance of a positive culture result is influenced by the sample collection method. For example, a QBC result of 3,000 CFUs/mL would indicate infection if the associated urine sample had been collected via cystocentesis; however, for a naturally voided sample, this result would most likely represent bacterial contamination, and diagnosis of UTI would require a QBC result of > 10,000 CFUs/mL.
The combination of a boric acid preservative and refrigeration was found in 1 study8 to yield consistent QBC results, compared with results for urine samples evaluated immediately after collection. However, urine samples used in that study were refrigerated immediately, which may differ from everyday practice. In a different study,9 researchers evaluated canine and feline urine samples submitted for QBC in the United Kingdom, taking into account the fact that most urine samples were collected via cystocentesis and submitted via the postal service. They concluded that urine samples submitted for bacterial culture (via the postal service) in a transport tube to which boric acid had been added as a preservative were more likely to yield false-negative results than were those submitted in a plain sterile tube. However, marked differences in climatic conditions (Texas in the present study vs the United Kingdom in the other study9) could affect QBC results for urine samples shipped at ambient temperatures during summer months.
Because urine is a viable culture medium, boric acid is a suitable preservative for urine samples. The borate ion has bacteriostatic properties, which maintain a stable environment for pathogenic bacterial growth while limiting contaminant overgrowth of urine samples.10 However, little veterinary data are available to support widespread use of preservatives. One study11 revealed that boric acid was a satisfactory urine preservative, given that it maintained the existing microbiologic flora without yielding high numbers of false-positive results. On the contrary, in the UK study,9 addition of boric acid preservative resulted in an increase in the likelihood of a false-negative result.
Standard procedures for the storage and transportation of urine samples have not been established in veterinary medicine. Given that the general recommendation of immediate culture is not always feasible in routine veterinary practice, the objective of the present study was to evaluate E coli survival in canine urine by use of 2 types of transport tubes maintained at 2 temperatures prior to QBC.12 Urine transport tubes containing preservatives have been validated for use in veterinary medicine but may cost more than plain sterile tubes and are not necessarily warranted in every scenario.8
The present study had several limitations. Only 1 organism (E coli) was evaluated, and it is possible that other organisms may be affected differently by the storage conditions evaluated. The most common organism isolated from canine urine is E coli, which accounts for a third to a half of all isolates.4 Gram-positive cocci (ie, Staphylococcus spp and Streptococcus spp) are the second most common organisms, with less common organisms including Proteus spp, Pseudomonas spp, and Mycoplasma spp.13 In most circumstances, a single species is responsible for UTI.4 Additional studies would be needed to identify the impact of storage conditions on less common urinary tract pathogens or in urine samples from dogs with UTIs caused by several organisms. Another limitation is that urine samples were collected from dogs without a UTI and therefore did not contain inflammatory mediators and cells such as polymorphonuclear cells that might affect bacterial survival and growth. Additionally, the study findings cannot be extrapolated to other species. Cats appear to be more resistant to UTIs than are dogs, so the definition of clinically important bacteriuria could have been expected to differ between those species.2 It should also be noted that in the present study, UTTs were filled with 1 mL of urine, which is less than the volume recommended by the manufacturer (4 mL) but was necessary because of the limited volume of urine available. The ratio of urine to preservative was therefore greater than would exist when manufacturer's recommendations are followed, which might have impacted bacterial viability and subsequent culture results. Another potential source of error related to the initial bacterial spiking of the 6 urine aliquots. Even small inaccuracies in this process could markedly impact subsequent QBC results. We attempted to minimize this error by having an experienced microbiologist (SDL) perform all pipetting required at that stage.
Regardless of the aforementioned limitations, results of the study reported here suggested that canine urine samples should be promptly placed in an SCT, then stored and shipped at 4°C if immediate processing for QBC is not possible. When canine urine samples are likely to be exposed to ambient temperatures during storage or transport to the laboratory, a UTT may be a more suitable choice.
Acknowledgments
The authors declare that there were no conflicts of interest.
ABBREVIATIONS
| QBC | Quantitative bacterial culture |
| SCT | Silicone-coated clot tube |
| UTI | Urinary tract infection |
| UTT | Urine transport tube |
Footnotes
Whatman paper (grade 1), GE, Fairfield, Conn.
Nalgene Rapid-Flow sterilization filter unit, Thermo Scientific, Grand Island, NY.
ATAGO automatic and water resistant refractometer, ATAGO USA Inc, Bellevue, Wash.
Multistix 10SG, Siemens Healthcare Diagnostics Inc, Tarrytown, NY.
KOVA Petter, Hycor Biomedical Inc, Garden Grove, Calif.
Becton Dickinson & Co, Franklin Lakes, NJ.
Monoject red-stopper blood collection tube (silicon coated), Covidien, Mansfield, Mass.
BD Vacutainer Plus C&S preservative tube, Becton, Dickinson & Co, Franklin Lakes, NJ.
Thermo Electron Corp, Marietta, Ohio.
GraphPad Prism, version 6.05, La Jolla, Calif.
References
1. Bartges JW. Urinary tract infections In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. 7th ed. St Louis: Elsevier, 2010;2036–2047.
2. Bartges JW. Diagnosis of urinary tract infections. Vet Clin North Am Small Anim Pract 2004; 34: 923–933.
3. Lulich JP, Osborne CA. Urine culture as a test for cure: why, when, and how? Vet Clin North Am Small Anim Pract 2004; 34: 1027–1041.
4. Smee N, Loyd K, Grauer GF. UTIs in small animal patients: part 2: diagnosis, treatment, and complications. J Am Anim Hosp Assoc 2013; 49: 83–94.
5. Weese JS, Blondeau JM, Boothe D, et al. Antimicrobial use guidelines for treatment of urinary tract disease in dogs and cats: antimicrobial guidelines working group of the international society for companion animal infectious diseases [serial online]. Vet Med Int 2011; 2011: 263768. Available at: www.hindawi. com/journals/vmi/2011/263768/. Accessed Aug 4, 2014.
6. Padilla J, Osborne CA, Ward GE. Effects of storage time and temperature on quantitative culture of canine urine. J Am Vet Med Assoc 1981; 178: 1077–1081.
7. BD. Product FAQs: Venous blood collection. Available at: www.bd.com/vacutainer/faqs/#urine_faq. Accessed Sep 8, 2015.
8. Allen TA, Jones RL, Purvance J. Microbiologic evaluation of canine urine: direct microscopic examination and preservation of specimen quality for culture. J Am Vet Med Assoc 1987; 190: 1289–1291.
9. Rowlands M, Blackwood L, Mas A, et al. The effect of boric acid on bacterial culture of canine and feline urine. J Small Anim Pract 2011; 52: 510–514.
10. Lum KT, Meers PD. Boric acid converts urine into an effective bacteriostatic transport medium. J Infect 1989; 18: 51–58.
11. Perrin J, Nicolet J. Influence of the transport on the outcome of the bacteriological analysis of a dog urine comparison of three transport tubes. J Vet Med 1992; B39: 662–667.
12. Hindman R, Tronic B, Bartlett R. Effect of delay on culture of urine. J Clin Microbiol 1976; 4: 102–103.
13. Smee N, Loyd K, Grauer G. UTIs in small animal patients: part 1: etiology and pathogenesis. J Am Anim Hosp Assoc 2013; 49: 1–7.
