This work was completed at North Carolina State University College of Veterinary Medicine, Raleigh, NC. The data were presented as an abstract at the ACVIM Forum, Seattle, 2007.
Corresponding author: Natasha Olby, College of Veterinary Medicine, NCSU, 4700 Hillsborough St, Raleigh, NC 27606; e-mail: firstname.lastname@example.org.
Background: Urinary tract infection (UTI) is a common complication in people with spinal cord injury (SCI). Dogs with acute intervertebral disc extrusion (IVDE) have similar risk factors for UTI when compared with human SCI patients and have a high perioperative prevalence of UTI.
Objectives: Determine the prevalence of UTI in dogs for 3 months after surgery for thoracolumbar IVDE and identify risk factors for development of UTI.
Animals: Twenty-five dogs treated surgically for 26 acute disc extrusions.
Methods: Prospective study. Urinalysis and urine culture were performed perioperatively. At home, owners monitored urine with dipsticks every 48 hours for 1 month then once a week until 3 months. Dogs returned for assessment of motor function, urinalysis, and urine culture at 1 and 3 months after surgery. Presence of UTI over the 3-month period was correlated to potential risk factors.
Results: Ten dogs (38%) developed 12 UTIs over the 3-month period, with the majority occurring between weeks 1 and 6; 60% of the UTIs were occult. Hematuria in the absence of pyuria or UTI was a common finding in the perioperative period. Sex, breed, and ambulatory status influenced the risk of developing a UTI.
Conclusions and Clinical Importance: There is a high prevalence of UTIs, many of which are occult, in the 3 months after surgery for thoracolumbar IVDE. These dogs should be routinely monitored for UTI with urine culture regardless of urinalysis results.
Normal micturition is a complex process mediated by visceral and somatic reflexes in the lumbar and sacral spinal cord. These reflexes operate in concert with voluntary input from the forebrain via pontine and mesencephalic upper motor neuron pathways.1–3 Interruption of these pathways because of spinal cord injury (SCI) causes failure of the vesicourethral unit and may result in bladder paralysis, urine retention, vesicosphincter dyssynergia, and urinary overflow incontinence.2,4 Urinary tract infection (UTI) is the most common medical complication during the rehabilitation period after SCI in people with an incidence of 0.68 UTIs/100 patients per day.5 Mortality caused by urinary tract disease has decreased precipitously in human SCI patients since World War I, but urinary tract morbidity still is the second most common cause of death in this population.4,6 Dysfunctional micturition, high-dose corticosteroid treatment, and indwelling or intermittent urinary catheterization are some of the factors that place people at high risk for developing UTI after SCI.4–6 This high prevalence is complicated by decreased ability to detect clinical signs of UTI because of impaired sensation.4
Similar risk factors are present in dogs with SCI because of acute intervertebral disc extrusion (IVDE), and abnormal micturition because of myelopathy has been implicated as a risk factor for persistent UTI and reinfection in dogs.7 Two prospective studies have reported a prevalence of UTI of 21 and 27% in the 1st week after injury,8,9 but there are no prospective studies looking at the prevalence of UTI in this population over a longer postoperative period. Our impression is that clinical signs of UTI may be absent in dogs with persistent paraparesis or paraplegia or may be mistaken by owners for incontinence because of myelopathy. Routine screening of dogs for UTI after surgery for IVDE may detect clinically silent infections and decrease the risk of ascending infection.
The objective of this study was to determine the perioperative, 1-month, and 3-month cumulative prevalence of UTI in dogs after surgery for acute thoracolumbar IVDE and to correlate the presence of UTI with potential risk factors. A secondary objective was to evaluate urinalysis reagent test strips as a potential screening test for UTI by owners.
Materials and Methods
The study population was client-owned dogs presented to North Carolina State University Veterinary Teaching Hospital (NCSU VTH) for acute thoracolumbar SCI due to acute IVDE. At the time of presentation and subsequent re-evaluation, severity of spinal cord disease was graded as (1) normal or minimal paraparesis and ataxia, (2) ambulatory paraparesis, (3) nonambulatory paraparesis, (4) paraplegia with intact pelvic limb or tail nociception or both, and (5) paraplegia without pelvic limb or tail nociception. On admission, age, sex, weight, body condition score, duration of neurological deficits, and recent drug treatment were recorded. Inclusion criteria for the study were nonambulatory paraparesis or paraplegia (grades 3–5) and surgically confirmed IVDE. Hemilaminectomy, as previously described,10 was performed in all cases and all dogs received intraoperative cefazolin (22 mg/kg IV q2h during surgery). After surgery, voluntary urination was noted in the medical record, and method of bladder evacuation was recorded for all dogs lacking voluntary urination. Dogs were assessed for UTI by urinalysis with urine samples collected by catheterization or cystocentesis on admission, during postoperative hospitalization, or both. Urine culture was performed on samples collected by cystocentesis or catheterization at the discretion of the clinician managing the case. Written client consent for inclusion in the study was obtained, and clients were provided with both written and verbal counseling on the clinical signs of UTI in dogs. Dogs returned approximately 1 and 3 months after surgery for neurologic examination, urinalysis, and quantitative urine culture, and clients completed a short questionnaire regarding urinary tract and systemic health of their pet since the last visit.
Urinalysis and Urine Culture Techniques
Urine was collected routinely by cystocentesis or sterile urinary catheterization and submitted for urinalysis (minimum of 2 mL) and urine culture (minimum of 1 mL). If the volume of urine collected was insufficient for both tests, urine culture was performed. Urinalysis entailed visual evaluation (color, turbidity), sulfosalicylic acid turbimetric testing for urine protein concentration, routine urine sediment analysis and a colorimetric urinalysis reagent test stripa evaluating urine protein, blood/hemoglobin, pH, ketones, and glucose. Hematuria was defined as >5 red blood cells [RBC] per high power field [hpf] and was classified as moderate (>50 RBC/hpf) or severe (>250 RBC/hpf). Pyuria was defined as >5 WBC/hpf. For quantitative urine culture, urine samples were centrifuged (1400 g for 10 minutes) and the pellet was inoculated onto plates containing Columbia agar with 5% sheep blood, phenylethyl alcohol agar, and deoxycholate agar and into thioglycolate broth. Before centrifugation, a 100-fold dilution of urine was mixed with molten agar and poured to quantitatively estimate bacterial numbers. All media were incubated at 35°C for 48 hours at which time bacterial species were determined with an automated system.b Antimicrobial susceptibility patterns for bacterial isolates were determined by the NCSU VTH Clinical Microbiology Laboratory using standard automated techniques. An extended susceptibility panel was used to determine bacterial susceptibility and varied based on the bacterial isolate. The number and species of bacterial isolates were recorded along with an approximate concentration (colony forming units [CFU]/mL) of each isolate in the sample. Lack of bacterial growth after 48 hours was considered a negative urine culture.
Screening for UTI
On discharge, clients were instructed to check their pet's urine with urinalysis test stripsb every 48 hours until the 1 month re-evaluation and then once weekly until the 3 month re-evaluation. Clients were provided with a log sheet and asked to record the results for pH, blood/hemoglobin, leukocytes, and protein along with micturition status (voluntary, expressed or catheterized) and antibiotic treatment. Normal values were considered to be leukocytes: negative, pH: 5–7, protein: negative to trace, blood: negative to trace. Clients were instructed to request urinalysis and urine culture if there were clinical signs of a UTI, if any of the test strip results were abnormal for 2 consecutive readings, or if there was a marked increase in any value (eg, blood/hemoglobin increases from negative to ++).
Criteria for Diagnosis of UTI
A positive urine culture or UTI was defined as (1) growth of ≥100 CFU/mL of a uropathogen from a cystocentesis urine sample, (2) growth of ≥10,000 CFU/mL of any bacteria from a catheterized urine sample,11,12 or (3) clinical signs or laboratory evidence (pyuria of >5 WBC/hpf13) of UTI that resolved with empirical antibiotic treatment. Diagnosis of a resistant UTI was based on one of the following susceptibility patterns: (1) Gram negative isolates resistant to both gentamicin and tobramycin, (2) Gram positive isolates other than Enterococcus sp. resistant to oxacillin, and (3) Enterococcus sp. resistant to penicillin. Occult UTI was defined as a UTI that met one of the first 2 criteria for UTI without evidence of pyuria.13
Continuous data (age at presentation, number of days of hospitalization, time to re-evaluation) were summarized as mean, standard deviation, and range. Ordinal (body condition and neurological score) and categorical data were summarized as proportions for each time period assessed. Categorical data included age (<6 years or ≥6 years), sex (F or M), body condition core (>6 or ≤6), preoperative treatment with corticosteroid (yes or no), neurologic grade before surgery (paraparesis or paraplegia), neurologic grade at re-evaluation (ambulatory or nonambulatory), breed (Dachshund or other breeds), catheterization during the perioperative period (yes or no), presence of perioperative hematuria (yes or no), urinating voluntarily at discharge (yes or no), and hematuria or pyuria at re-evaluation (yes or no).
The relationship between categorical data and development of UTI over the 3-month study period was assessed by constructing contingency tables and performing χ2 or Fisher's exact tests as appropriate based on number of data points. The risk factors for development of UTI were evaluated by logistic regression. Odds ratios of potential risk factors for UTI were calculated as independent factors. Statistical evaluation was performed using JMP 7.0.cP<.05 was considered statistically significant.
Twenty-five dogs were entered in the study for treatment of 26 acute IVDE; 1 dog was enrolled twice because of 2 discrete IVDE with complete recovery between surgeries. Mean age at presentation was 5.6 years (SD, 2.5; range, 2.5–12 years). Dachshunds were the most common breed in the study (12/26 dogs: 46%); other breeds were Beagle (4), Miniature Poodle (2), Shih Tzu (2), Dalmatian (1), mixed breed (1), Boykin Spaniel (1), Cocker Spaniel (1), Pekingese (1), and Pembroke Welsh Corgi (1). There were 12 castrated males, 1 intact male, and 13 spayed females. At presentation, neurologic deficits were scored as grade 3 in 9/26 dogs (35%), grade 4 in 13/26 dogs (50%), and grade 5 in 4/26 dogs (15%).
Eight of 26 dogs (31%) demonstrated consistent voluntary urination within 24 hours of surgery. All dogs that were unable to voluntarily urinate had clinical signs of upper motor neuron failure of micturition. A urinary catheter was passed for urine sampling or bladder evacuation at least once in 8 male dogs. Bladder evacuation was otherwise accomplished by intermittent manual expression. The mean number of days in hospital after surgery was 3.5 (SD, 1.8; range, 1–9). At the time of discharge, 13 dogs were able to urinate on their own; the remaining 13 dogs were still unable to urinate and their owners manually expressed their bladders.
Twenty-five dogs presented for the 1-month evaluation and 20 for the 3-month evaluation. Five dogs were lost to follow-up after the 1st re-evaluation and 1 dog was euthanized because of persistent paraparesis and urinary incontinence. One dog returned only once midway between scheduled re-evaluations and was included with the 3-month dataset. Mean time to the 1st re-evaluation was 4.3 weeks (SD, 0.8; range, 2.0–5.9 weeks) and to the 2nd re-evaluation was 12.4 weeks (SD, 1.3; range, 7.6–14.1 weeks). Two dogs presented to NCSU VTH outside of the study time points because of acute macroscopic hematuria due to UTI. Twenty-four of 26 (92%) dogs were ambulatory and urinating voluntarily at the 1st re-evaluation. Eighteen of 20 dogs (90%) were ambulatory and urinating voluntarily at the 2nd re-evaluation. Of the 2 dogs that were nonambulatory at both re-evaluations, 1 was grade 5 at presentation and remained paraplegic. The other dog was grade 4 at presentation and grade 3 at 3 months, and regained ambulation approximately 6 months after surgery.
Urinalysis and Urine Culture Results
Seventy-seven urinalyses and 75 urine cultures were performed as part of the study. Hematuria was documented in 12/25 dogs (48%) while hospitalized after surgery, in 4/25 dogs (16%) at the 1st re-evaluation, and in 4/19 dogs (21%) at the 2nd re-evaluation (inadequate urine was obtained from 1 dog at the 2nd re-evaluation to perform both a urinalysis and urine culture). At the same time points, 8/25 dogs (32%), 2/25 dogs (8%), and 3/19 dogs (16%) were considered to have moderate or severe hematuria. In contrast, pyuria was documented in 2/25 dogs (8%), 3/25 dogs (12%), and 2/19 (11%) at the 3 study time points (Fig 1).
Ten dogs (38%) developed 12 UTIs during the 3-month period studied. UTI was diagnosed in 0 dogs during the perioperative period (within 1 week of surgery), 9/25 (36%) at the 1st re-evaluation (1–6 weeks postoperatively) and 3/20 (15%) at the 2nd re-evaluation (7–14 weeks postoperatively) (Fig 1). Of the 2 dogs that developed multiple UTIs after surgery, 1 was chronically paraplegic and the other recovered a normal gait by the 3-month re-evaluation. Eleven of the 12 UTIs were diagnosed based on a positive urine culture from a urine sample collected by cystocentesis; 1 UTI was diagnosed by the referring veterinarian based on macroscopic hematuria, malodorous urine, urinalysis test strips (positive for blood, WBC, and protein), and response to empirical antibiotic treatment. In samples with a positive urine culture, bacterial concentrations were >100,000 CFU/mL in 6 dogs, between 10,000 and 100,000 CFU/mL in 2 dogs, between 1,000 and 10,000 CFU/mL in 2 dogs, and <1,000 CFU/mL in 1 dog. All dogs with positive urine cultures had concurrent urinalyses; 6 of these dogs were considered to have occult UTI because there was no evidence of pyuria. Owners of only 3 out of 10 dogs confirmed to have a UTI based on urine culture suspected UTI in their pet at the time of urine collection. Three of the 10 dogs that developed UTIs were treated with a course of antibiotics at time of discharge.
UTI was most commonly caused by Escherichia coli (7/12); other bacterial isolates were Klebsiella pneumoniae (3/12), Enterococcus sp. (3/12), Staphylococcus intermedius (2/12), Streptococcus sp. (1/12), and Proteus mirabilis (1/12). Polymicrobial UTIs were diagnosed in 4 dogs, caused by 2 organisms in 3 dogs, and 3 organisms in 1 dog. Four of the UTIs diagnosed during the study period were considered antibiotic resistant; antibiotic resistant bacteria were E. coli (3/6), K. pneumoniae (1/3) Enterococcus faecium (1/3), and Streptococcus sp (1/1). All 4 resistant infections were the first recorded UTIs although one of the dogs had been treated with antibiotics at discharge based on the presence of hematuria on the urinalysis (before obtaining negative results of urine culture). Antibiotic treatment for resistant infections entailed amikacin (n = 2), cefpodoxime (n = 1), and ultrahigh dose amoxicillin (n = 1).
Urinalysis Test Strip Results
Owners reported that it was difficult to monitor their pet's urine with urinalysis test strips; owners of 14/26 dogs (54%) were able to complete this task consistently for at least 1 month after surgery. Urinalysis test strip results were recorded by owners for 5 of the UTIs diagnosed during the study and in all cases, test strip abnormalities supported the diagnosis of UTI. Malodorous urine, macroscopic hematuria, or both also were noted in 3 of these 5 infections. One dog developed test strip abnormalities consistent with UTI along with macroscopic hematuria and malodorous urine approximately 2 weeks after surgery. Clinical signs and test strip abnormalities resolved without treatment but occult UTI was diagnosed at the 1-month re-evaluation.
Seven dogs were discharged from the NCSU VTH with a short course (≤14 days) of antibiotics for suspected UTI based on the presence of hematuria. Urine culture results were negative in 6 of these dogs; culture was not performed in the 7th dog. One dog was given cephalexin before referral because of suspected UTI and this prescribed course of antibiotics was completed. Two dogs were discharged with a 2- to 4-week course of cephalexin for bacterial dermatitis and 1 dog was discharged with 10 days of metronidazole for colitis.
There was no significant difference in the prevalence of UTI between Dachshunds and other breeds, between the different sexes, or among the different grades of injury. However, females were estimated to be 4.7 times more likely to develop UTI than males (95% confidence intervals, 0.7–30.3), Dachshunds were estimated to be 1.6 times more likely to develop UTI than other breeds (95% confidence intervals, 0.29–8.4), and dogs that were paraplegic at presentation were estimated to be twice as likely to develop UTI as dogs that were nonambulatory paraparetic (95% confidence intervals, 0.3–12.3). There was no relationship between ability to urinate at discharge or catheterization during hospitalization and development of UTI over the 3-month period studied. Likewise, there was no significant relationship between recovery of motor function and development of UTI, but dogs that were nonambulatory at 3 months were estimated to be 2.6 times more likely to develop UTI than dogs that were able to walk (95% confidence intervals, 0.14–5). Logistic regression failed to identify any combination of factors that was significantly associated with development of UTI.
In this study, the prevalence of UTIs was 38% over the 3-month period after surgery for IVDE. The infection rate was 0 in the immediate perioperative period despite a high prevalence of hematuria, in contrast to findings of other studies,8,9 with the majority of infections occurring between 1 and 6 weeks after injury. This finding confirms that SCI predisposes to UTI even after recovery of voluntary urination.7,14,15 Fifty percent of the infections were considered occult using a definition of pyuria as >5 WBC/hpf, decreasing the sensitivity of monitoring urine with a dipstick (a task that also was found to be difficult to perform by owners).
Two previous studies have evaluated the prevalence of UTI in a similar population of dogs for 7 days after decompressive surgery for IVDE.8,9 One of these studies included ambulatory dogs and reported UTIs in 15% of dogs at admission, and an overall prevalence of 27% over the course of 7 days.8 The 2nd study evaluated the effect of method of bladder evacuation on the development of UTI, and excluded dogs that had UTIs at entry into the study, dogs that were ambulatory, and dogs that failed to regain voluntary urination during hospitalization.9 The prevalence of UTIs in this study was 21%, which is much higher than the 0% recorded in our study in this time period.9 Possible reasons for the absence of UTIs in the perioperative period of our study include limited duration of hospitalization when compared with the other studies, use of intraoperative cefazolin, and use of postoperative antibiotics in 10 dogs to treat pyoderma, colitis, and suspected UTI based on hematuria. Although these dogs had negative urine cultures, it is possible that they would have developed UTI in the 1st postoperative week without antibiotic treatment.
We were unable to show a significant relationship between any categorical data and the development of UTI. However, females, Dachshunds, and paraplegic dogs were at higher risk. In a large population of dogs with UTIs, Dachshunds have been reported to be at higher risk of developing infection than other breeds,16 possibly reflecting their high prevalence of IVDE. Additional risk factors previously associated with development of UTI in the perioperative period include treatment with intraoperative antibiotics, body temperature <35°C during surgery, and duration of urinary bladder dysfunction.8,9 Method of bladder evacuation has been demonstrated to be important in humans with SCI,5,6 with indwelling catheters posing the greatest risk, followed by sterile and clean intermittent catheterization and then spontaneous voiding. These studies in humans investigated the chronic SCI patient rather than the acute phase after injury, perhaps explaining why their results differ from the findings in dogs, in which the method of bladder evacuation was not a risk factor for developing UTI.8,9 Studies in dogs have shown that indwelling catheters are associated with an increased risk of infection if the catheter is in place for more than 3 days, with a 27% increase in risk with each additional day of catheterization and increased risk in females compared with males.9,17–20
While hospitalized, dogs had their bladders evacuated by manual expression when possible, and intermittent catheterization was used when increased urethral tone precluded effective expression. There was a high perioperative incidence of hematuria, likely because of retention cystitis or manual bladder expression based on the negative urine cultures. This is consistent with the finding that hematuria without pyuria or bacteriuria was associated with UTI in only 1.7% of dogs in a previous study.7
The infection rate in the 3 months after surgery was high (38%). However, the majority of dogs recovered the ability to walk and urinate voluntarily and the presence of UTI was not significantly associated with failure to recover motor and bladder function. This suggests that even dogs that have made a good neurologic recovery may not be emptying their bladders effectively, predisposing them to UTI.7 This phenomenon has been documented in a study that followed the recovery of bladder function after surgery for IVDE by ultrasonic quantification of bladder volume before and after urination21; 4 of 7 dogs that were urinating voluntarily were not voiding completely as late as 18–25 days after surgery.21 Only 3 owners of the 10 dogs that developed UTIs detected clinical signs of an infection. Unfortunately, urinalysis dipsticks were difficult for owners of chondrodystrophoid dogs to use and did not appear to be a useful screening test for UTI in this context both because of impracticality and because of the high percentage of occult UTIs. This emphasizes the importance of performing urine cultures after surgery for IVDE regardless of clinical signs and urinalysis results. Although the majority of UTIs occurred in the first 6 weeks postoperatively, 15% of the dogs had UTI at 3 months, and a longer study is warranted to determine over what period dogs recovering from SCI should be monitored for UTIs.
The most common bacterial isolate was E. coli, accounting for 58% of UTIs, consistent with other reports of bacterial uropathogens in dogs and in humans.7–9,16,18,22 The other pathogens are all common causes of UTI.16,22 All resistant infections were seen at the 1-month evaluation point and only one of these dogs had been treated previously with antibiotics. Prophylactic measures to prevent UTI include the use of cranberry juice, and drugs such as methenamine hippurate.23 Cranberry juice is believed to act by inhibiting binding of uropathogenic bacteria, primarily E. coli, to the bladder epithelium, in addition to increasing urine acidity. Methenamine hippurate produces formaldehyde that acts as a bacteriostatic agent. Several clinical trials evaluating the efficacy of cranberry juice and methenamine hippurate in decreasing UTI prevalence in humans with SCI have been conducted and findings have been contradictory.24–27 Evaluation of bladder prophylaxis protocols in dogs after SCI is warranted.
Low patient numbers limited the power of this study; the requirement for 2 re-evaluations resulted in case attrition between the 1 and 3 month evaluations. When dogs came for their re-evaluations, they did not always have full bladders, necessitating hospitalization until there was an adequate volume of urine for both urinalysis and urine culture. If this was not possible, urine culture was performed without urinalysis, causing additional data loss. The use of postoperative antibiotics to treat a variety of conditions potentially changed the prevalence of UTIs in the 1st week, but should not have influenced the data collected over the following 3 months.
In conclusion, nonambulatory dogs that undergo surgery for IVDE are at risk for development of UTI in the 3 months after surgery, regardless of recovery of pelvic limb and bladder function. Given the relatively high incidence of UTI in this group of dogs and the frequency of occult infections, these patients should be monitored routinely for UTI by urine culture for at least 3 months after surgery.
aChemstrip 9, Roche, Basel, Switzerland
bVITEK automated system, bioMerieux, St Louis, MO
cSAS, Cary, NC
Thanks are due to Carol Lemons and Megan Fauls for performing the urine cultures, and Dr Mark Papich for advice on antibiotic treatment.
This study was funded by a Department of Clinical Sciences Veterinary Practice Plan Research Award.