Introduction

Urinalysis and urine culture are common diagnostic tools used to screen dogs for urinary tract infections.¹ Collecting a urine sample from a dog safely, noninvasively, and in a sterile manner is a technical challenge for veterinary professionals. There are several methods used to collect specimens for culture, including cystocentesis (CYS), catheterization, or clean-caught voided. Collection of urine via CYS or clean-caught voided are the most common methods used in general practice, but each has limitations.

CYS is the current preferred method of urine specimen collection in dogs that require urine culture because it yields the lowest number of urethral contaminants.¹ However, CYS requires a compliant or sedated patient, an experienced technician, and a sufficiently filled bladder and even when performed correctly can result in serious complications such as iatrogenic aortic puncture and hemoabdomen.² Patients have been reported to have septic peritonitis, blood clots, and tears of the urinary bladder after CYS.³ This method can also cause spurious hematuria on urinalysis, which is not fully representative of what is present in the urinary bladder.

Clean-caught voided urine specimens are often collected in veterinary medicine owing to their ease and convenience. It is more efficient to walk a patient outside and collect a urine specimen without the need for physical or chemical restraint. This collection method is minimally invasive but has increased bacterial contamination rates when compared with the other urine collection modalities because the collected urine must pass through the nonsterile distal urethra before collection.⁴ Collecting a midstream urine sample where the initial urine stream is discarded mitigates some of the contamination risks because the resident bacteria of the urinary tract may be voided in the initial stream and not collected.¹ This is the preferred method for collecting urine samples in people.

In veterinary medicine, consistently collecting midstream voided urine specimens is difficult. In dogs, clean-caught voided urine specimens are collected by watching for urination and placing a clean container into the urine stream. Ensuring that sufficient urine has been voided before collecting the specimen while also maintaining the urine collection container in the urine stream long enough to collect a sufficient volume for analysis can be challenging. It is not known whether the aforementioned practical clean-caught voided urine collection method used in veterinary medicine yields urinalysis and urine culture results similar to true midstream urine collections.

With urinalysis and culture being common diagnostic tools in veterinary medicine, a urine collection method that requires the least technical staff and is sufficiently sterile would be beneficial. Because midstream urine collections in theory have fewer urethral contaminants, this study aims to determine whether midstream collected urine specimens where the initial void is discarded provide a urine specimen with reduced bacterial contamination as compared with current methods of clean-caught voided urinary collections. To assist in achieving this aim, this study used a midstream collection device (MCD) to obtain a midstream urine sample. The MCD discards the first 10–20 mL of urine before collecting the remaining urine sample in a sterile container. We hypothesized that the MCD collected samples would reveal reduced bacterial colony-forming units (CFUs)/mL and cellular contamination as compared with current clean-caught voided practices.

Materials and Methods

In this descriptive cross-sectional study, 41 client-owned dogs greater than or equal to 15 kg in weight were screened for inclusion. Patients less than 15 kg were excluded from this study because the MCD discards the first 10–20 mL of urine and small dogs do not consistently produce enough urine in a single void for proper use of the device. Other exclusion criteria included antibiotic use within the last 7 days or a history of lower urinary tract signs such as pollakiuria, hematuria, or stranguria over the past 14 days. Dogs that could not have a CYS performed safely without sedation were also excluded. For the standard clean-caught (SCC) and MCD collection groups, patients who did not produce a sufficient volume of urine or had behavioral issues that prohibited collection were excluded as well. The University of Wisconsin-Madison Institutional Animal Care and Use Committee approval was obtained, and all dogs had client consent to be enrolled.

The MCD used in this study is the Peezy Midstream^a device. The sterile Peezy Midstream device captures true midstream samples because the first 10–20 mL of voided urine passes through the device as a cellulose sponge expands. Once the sponge expands, it blocks additional urine from leaving the device, and urine is bypassed into a sterile collection tube. The device is funnel-shaped and had to be modified to a half-funnel so that it could be placed under a female dog as they squat during urination. The MCD device was modified using sterile instruments. The modification did not affect how urine passed through the cellulose sponge or into the sterile collection receptacle.

After enrollment, urine was collected from each patient using SCC, MCD, and CYS during a 7-day period. The order of collections was randomized to control for spurious hematuria detection caused by CYS collection. During the time between collections, owners were questioned regarding the development of lower urinary tract signs (pollakiuria, hematuria, or stranguria), and subjects would have been excluded if any signs did develop. For SCC collections, a clean, flat receptacle was placed into the urine stream and, once obtained, transferred to a sterile collection cup. SCC collections were placed in the urine stream as quickly as possible to simulate the nonstandardized technique of SCC used in veterinary clinics. For MCD collections, the sterile device was placed into the urination stream and the sample was diverted into a sterile collection tube. Voided collections (SCC and MCD) were deemed failed if insufficient volume was collected to complete all testing. CYS collections were performed by placing the dogs in dorsal recumbency. A sterile 1.5-inch, 22-gauge needle was then directed into the urinary bladder as identified by bladder palpation or ultrasound. CYS samples were deemed failed if samples could not be collected safely without sedation or if insufficient volume was collected.

Urine samples were submitted for analysis within 30 min of collection to the University of Wisconsin-Madison Veterinary Care Clinical Pathology and Microbiology Laboratories for urinalysis and quantitative aerobic culture. Urinalysis (urine dipstick and urine sediment) were performed by the University of Wisconsin-Madison Veterinary Care Clinical Pathology Laboratory using standard techniques. Quantitative aerobic urine culture was performed using standard techniques by the University of Wisconsin-Madison Veterinary Care Microbiology Laboratory. The quantitative culture consisted of plating 10 µL of urine collected by clean voiding or 100 µL for urine collected by CYS. Urine cultures were evaluated for growth after 24 and 48 hr of incubation. Growth was quantitated as the number of CFUs/mL. For this project, all bacterial isolates cultivated were quantitated and identified via MALDI-TOF^b.

Urine sample characteristics including the presence/absence of protein and heme were recorded from the urine dipstick; the presence/absence of red blood cells (RBCs), white blood cells, squamous cells, and transitional cells were recorded from the urine sediment; and the presence/absence of bacteria was recorded from the aerobic urine culture. For statistical analysis, any presence of a characteristic was assigned the value 1, whereas the absence of a characteristic was assigned the value 0.

Results

Forty-one dogs were screened for inclusion between June 2021 and April 2022. Of these, two were excluded for safety concerns associated with performing CYS on a noncompliant patient. Seven of the 41 dogs were excluded because of challenges associated with collecting sufficient urine volume for analysis using the MCD. Those excluded were predominantly female dogs that postured too close to the ground to collect urine using the MCD and male dogs that urinated small amounts multiple times rather than producing a consistent stream of urine. Six other dogs did not complete the study because the owners disenrolled or the dogs were found to have aggression or anxiety at the clinic.

Twenty-six dogs fulfilled all inclusion criteria and had the three urine collections accomplished successfully in a 7-day period. Of these 26, 12 were spayed females and 14 were neutered males. The age range of the population was 1–11 yr with a mean (±SD) age of 5.5 ± 2.9. Dogs weighed between 15 and 34 kg with a mean weight of 23.5 kg ± 6.0 kg. Eleven different pure breeds as well as 11 mixed-breed dogs are represented in this population. The most common breeds were mixed breeds (n = 12), greyhounds (n = 4), Australian shepherds (n = 2), and Alaskan huskies (n = 2).

MCD was the first of three collection modes for 14/26 (54%) dogs, with cystocentesis performed on the same day for one of those 14 dogs. SCC was the first mode of collection for 6/26 (23%) of the cohort. When collection by CYS was first (6/26; 23%), the other modes of collection (MCD or SCC) took place at least 1 day later (range 1–3 days later).

To determine whether MCD collected urine samples had decreased cellular components and bacteria compared with SCC, urine sediment, urine dipstick, and aerobic urine culture results were compared (Table 1).

TABLE 1 Prevalence, Expressed as a Percentage, for Each Cell or Sediment Type According to Mode of Collection

View Fullscreen

Among the eight variables assessed from the urinalysis and urine culture results, only the prevalence of squamous cells and of positive aerobic urine culture was noted to differ among modes of collection (P = .001 and .005, respectively). Prevalence of squamous cells appeared not to differ between SCC and MCD (adjusted P value = .109) but showed strong evidence of CYS being lower than MCD (adjusted P value = .048) and SCC (adjusted P value < .009). A similar pattern was found for positive urine cultures that exhibited no difference (adjusted P value = .099) between SCC and MCD and with lower prevalence in CYS compared with each of the other two (adjusted P values of .009 [CYS versus MCD] and <.001 [CYS versus SCC]) (Table 1).

Of the 26 urine samples, 30% (n = 8) of the MCD specimens (6 of the spayed female and 2 of the neutered male) had at least one bacterial species identified and 23% (n = 6) of the samples had more than one isolate. Among the 26 SCC samples, 46% (n = 12) of specimens (7 of the spayed female and 5 of the neutered male) had at least one bacterial species identified and 26% (n = 7) of samples had more than one isolate. Of the CYS specimens, 3% (n = 1) of samples (from a spayed female) had one isolate and no samples had multiple isolates. Table 2 shows the frequency of isolated bacterial species found in CYS, MCD, and SCC samples.

TABLE 2 Comparison of Bacterial Species Isolated from Each Urine Collection Technique

View Fullscreen

Discussion

Overall, these data demonstrate that midstream collections do not differ significantly from urine that is collected using the SCC collection methods currently employed in veterinary medicine. Bacterial contamination frequently occurred with both collection styles. Species isolated from both MCD and SCC were consistent with normal urethral microbiota (Streptococcus canis,⁷Haemophilus sp.⁸), common inhabitants of the skin (Staphylococcus pseudointermedius⁷), and common contaminants from the gastrointestinal tract (Escherichia coli⁹). Notably, none of the MCD and SCC samples had ≥100,000 CFUs/mL of bacteria, which is the recommended cutoff for determining whether bacterial counts are significant in voided samples.¹⁰ Even with midstream collection and a sterile collection device, similar levels of bacterial contamination from the urinary tract and skin were identified when compared with urine collected using the clean-caught technique.

The number of squamous cells was also similar between MCD and SCC samples. Because squamous cells are present in the urogenital system,¹¹ this finding suggests that despite midstream catches being thought to “wash out” cellular contaminants of the urethra, there is still a degree of cellular contamination regardless of how long you wait to begin the urine collection. This is consistent with the bacterial contamination findings as well, showing that regardless of when the collection is initiated, there are alterations to the urine sample that occurs as the sample passes through the urethra and external genitalia.

When comparing the free-catch collection styles with CYS, there were differences in cellular and bacterial contamination noted. Both the SCC and MCD had more squamous cells present than the CYS. This is explained by the CYS bypassing the need for urine to move through the distal urogenital tract,¹² thereby reducing cellular contamination. Moreover, when comparing RBC presence, there were no significant differences between any of the collection styles. This was unexpected as the CYS is often thought to introduce RBCs into the bladder.¹² Some of the RBCs in the free-catch samples could be explained by the CYS being performed before the free catches in six dogs; however, this study did not assess how often or how long blood is detectable in the urine after CYS.

The main study limitation was the number of dogs that needed to be excluded from the study (36% of screened dogs) and, subsequently, the study size. This prevented the authors from obtaining adequate statistical power to stratify the results to assess for additional differences such as bacteriuria detection in the urine of male and female dogs. Many of the dog exclusions were related to challenges using the midstream free-catch device (47% of excluded dogs). This added a potential inadvertent study bias because dogs that urinated small amounts and dogs that urinated close to the ground were excluded from the study because a sufficient urine volume could not be collected using the device (20–30 mL of urine is needed). In this author’s opinion, these difficulties using the midstream urine collection device would prevent its routine use in veterinary clinics regardless of whether the device was able to reduce urine contamination in voided samples.

Conclusion

Collection of urine by CYS remains the most accurate representation of what is present in the urinary bladder and is the recommended urine collection technique when accuracy is required. However, CYS samples are not always practical to obtain during a veterinary appointment for reasons shown in this study, such as a noncompliant patient or a small urinary bladder. This study demonstrated that if a clinician elects to collect urine with SCC or MCD techniques, either option is suitable. Collecting a midstream sterile urine sample did not yield statistically different urinalysis and culture results when compared with collecting a voided urine sample into a clean container before transfer to a sterile cup. However, when any free-catch technique is used, clinicians do need to interpret results considering the patient’s clinical signs, the CFUs/mL detected via culture, and the organism(s) detected because bacterial contamination is common with these collection techniques.