a b s t r a c t

Aim: Bacteremia is an uncommon complication of urinary tract infection . The aim of this study was to identify risk factors for bacteremic UTI in pediatric patients.

Methods: The medical records of all pediatric patients with UTI between 2013 and 2014 were retrospectively reviewed. Pediatric patients with accompanying bacteremia were compared with pediatric patients with no bacteremia. Results: Five hundred twenty-seven cases of UTI were identified. Blood cultures were taken in 464, 26 (5.6%) of which also were bacteremic. Pediatric patients with bacteremia were more likely to be male (58% vs 28%, P b .01), to be younger than 3 months (54% vs 31%, P = .02), and to have higher creatinine (average 0.77 +- 0.97 vs 0.34 +- 0.24, P b .01). Pediatric patients with bacteremia had higher rate of underlying urologic conditions. The following variables were included in multivariate analysis: age b 3 months, sex, ethnicity, method of Urine collection, creatinine, and underlying urologic conditions. Only creatinine (odds ratio, 3.67; 95% confidence interval, 1.69- 8.11) was found as an independent risk factor for bacteremia.

Conclusions: High creatinine at presentation is a risk factor that might aid in early identification of pediatric patients with high risk for bacteremia and its complications.

(C) 2016

Introduction

urinary tract infection is one of the most common serious bacterial infections in children [1]. The rate of concurrent bacteremia has been reported to be up to 31% [2,3]. Some studies demonstrated that it is difficult to distinguish clinically between bacteremic and nonbacteremic UTIs [2-4] because of conflicting data regarding differ- ences in signs and symptoms of bacteremic vs nonbacteremic pediatric patients. However, children with bacteremia usually receive longer par- enteral antibiotic treatment, although there is no clear evidence of its advantage in preventing complications or improving outcome in such cases. The main parameter that was consistently found as a risk factor for bacteremia is young age [2-5]. Most of the very young infants and all of the neonates are admitted for intravenous antibiotic treatment, but older children are often treated orally. It is thus important to identify risk factors for bacteremia also in the older age group which is not auto- matically admitted but is still prone for complications. The purpose of this study was to compare demographic, clinical, and laboratory charac- teristics of bacteremic and nonbacteremic pediatric patients of all ages and to identify risk factors associated with bacteremia in children with UTI. We assumed that this study can verify some of the already known

? The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

?? The author received no financial support for the research, authorship, and/or publication of this article.

* Infectious Diseases Unit, Shaare Zedek Medical Center, POB 3235, Jerusalem, Israel. Tel.: +972 2 6666340, +972 547946690(cell); fax: +972 26666840.

E-mail addresses: [email protected], [email protected].

risk factors, such as young age and male sex, and can possibly identify new risk factors for bacteremia that were not often examined, such as having chills or having underlying urologic conditions. Early identifica- tion of bacteremic patients can possibly reduce complications by prompt administration of Intravenous antibiotics and supportive treatment, Close monitoring of Hemodynamic state, and early transfer to intensive care unit when indicated.

Materials and methods

The computerized database of Shaare Zedek Medical Center, a university-affiliated general hospital, was retrospectively reviewed for all patients aged 3 days to 16 years that were admitted to the pediatric emergency department (ED), including those that were discharged home from the ED, during a period of 2 years (2013-2014) and had a positive urine culture and a discharge diagnosis of UTI. The following data were retrieved from the medical records: patient age and sex, ethnicity, underlying conditions, date of admission, length of hospitalization, presenting signs and symptoms, complete blood count, C-reactive protein, creatinine, antibiotic treatment, imaging data (ultrasonography [US], voiding cystourethrogram [VCUG]), need for intensive care and outcome, presence of coinfection, complications, organism causing the UTI, results of quantitative culture, technique of obtaining the culture, and presence of concurrent bacteremia. Urine specimens were obtained by suprapubic aspiration or transurethral bladder catheterization in infants b2 years and by midstream sampling in older children. At our pediatric ED, blood cultures are routinely taken in febrile pediatric patients. This stems from several aspects including

http://dx.doi.org/10.1016/j.ajem.2016.09.060

0735-6757/(C) 2016

O. Megged / American Journal of Emergency Medicine 35 (2017) 36-38 37

Table 1

Demographic and clinical characteristics of pediatric patients with UTI who had blood cultures taken, with and without bacteremia

Data are presented in table as the number with the percentage in parenthesis for categorical variables, unless indicated otherwise. SPA, suprapubic aspiration; PICU, pediatric intensive care unit.

the attempt to avoid additional venous sampling and lack of data on the prevalence of occult bacteremia in the era of PCV13 in our medical center.

Blood and urine cultures were obtained from all pediatric patients who had fever >=38?`C and at least 1 of the following symptoms/signs: dysuria, frequency, hematuria, suprapubic tenderness and/or costovertebral area tenderness to percussion, and in all infants younger than 60 days with either fever >=38?C or unexplained vomiting, unexplained jaundice, or failure to thrive.

The study was approved by the local institutional review board committee.

Statistical analysis

Data were collected using Microsoft Excel 2003 (Microsoft, Redwood, WA). Statistical analysis was done using WINPEPI (Computer programs for epidemiologists, and their teaching potential. Epidemio- logic Perspectives & Innovations 2011, 8:1). All bacteremic children were compared with all nonbacteremic children. Qualitative variables were compared using the ?² test. Quantitative variables were compared by the t test. Differences were considered significant at the level of P b .05. Logistic regression was used for multivariate analysis, and variables were included in the multivariate logistic regression analysis if they were found to be associated with differences between the groups in the unadjusted analyses (P b .1).

Results

Five hundred twenty-seven cases of UTI were identified. Blood cultures were taken in 464, 26 (5.6%) of which also were bacteremic. One hundred eighty-two (35%) were admitted to the hospital, and the rest were discharged home from the ED. The median age was 51 months, and 27% of the pediatric patients were male. One hundred fifty-four (29%) of the children were younger than 90 days. Escherichia coli was the most common bacterium isolated from urine cultures (405/527, 77%).

There was no statistically significant difference between the bacteremic and nonbacteremic groups in terms of mean age, length of symptoms prior to admission, temperature, white blood cell count, abnormal imaging studies (US and/or VCUG), and having underlying nonurogenital conditions. However, bacteremic children were more likely than nonbacteremic to be of Jewish descent (100% vs 79%; P = .005), to be male (58% vs 28%; P = .0004), to be younger than 3 months (54% vs 31%; P = .0049), to have known underlying urogeni- tal abnormalities (13% vs 5%; P = .04), to have abnormal creatinine at presentation (mean 0.77 mg/dL vs 0.34 mg/dL; P b .01), and to be admit- ted to intensive care unit (11.5% vs 1.4%; P = .0004). More children with bacteremia were hospitalized (88% vs 38% in the nonbacteremic group; P b .0001), and the mean duration of hospitalization for bacteremic patients was significantly longer than that of nonbacteremic patients (6.5 vs 4.8 days; P = .04). None of the children had renal abscess or acute lobar nephronia, and all recovered. Table 1 presents the pediatric patients’ demographic, clinical, and laboratory characteristics. The vari- ables that were included in multivariate analysis were age b 3 months, sex, ethnicity, having underlying urologic conditions, method of culture collection, and creatinine. Only creatinine (odds ratio, 3.67; 95% confidence interval, 1.69-8.11) was found as an independent risk factor for bacteremia.

Discussion

Bacteremia is not rare in children with UTI. Young age is a well- known risk factor for bacteremia in pediatric UTI [2,3,5-7]. In our study 5.6% of pediatric patients at all ages and 9.3% of infants b 90 days were bacteremic, similar to the rate of bacteremia previously de- scribed in larger studies [8,9]. A possible explanation for this higher rate of bacteremia in younger infants is that hematogenous spread of in- fection has been thought to be more common in neonatal UTI, whereas ascending infection is thought to be the cause of UTI in older children. Another possible explanation could have been undiagnosed urinary malformations in the first episode of UTI in neonates, but in our study, the rate of abnormal US and abnormal VCUG was similar in all age

38 O. Megged / American Journal of Emergency Medicine 35 (2017) 36-38

groups. The rate of bacteremia was slightly higher in preterm infants (7.7% vs 5.4% in term infants) but not statistically significant in our study. Prematurity is a known risk factor for bacteremic UTI [10], and the higher proportion of preterm infants in the male sex has been pre- viously reported as a risk factor for UTI in the neonatal period [11] and for bacteremia in neonatal UTI [5]. Circumcision is a risk factor for UTI in the 2 weeks following the procedure [12]. As almost all of our male population undergo circumcision, this can partially explain the higher rate of bacteremia in very young males in our cohort. We have no expla- nation as to why Jewish descent was found to be more frequently asso- ciated with bacteremia in our study, and our finding is in contradiction to the findings of another center that serves the same population [5]. We thus believe that this is an incidental finding due to the relatively small number of cases with bacteremia and that Jewish descent is not a true risk factor for bacteremia. We identified increased creatinine as a factor associated with bacteremia, which supports the findings of Averbuch et al [5]. A possible explanation can be hypotension accompa- nying bacteremia causing decreased renal blood flow with subsequent in- crease in creatinine. Chronic renal failure due to anatomic malformations can explain some of the cases with ascending infection and associated bacteremia, as structural abnormalities and vesicouretheral reflux were found in higher frequency in the bacteremic pediatric patients.

Children with bacteremia often receive longer parenteral treatment

[13], although no clear guidelines support this policy [14]. Most children older than 1 month with UTI were treated orally or received a short intravenous course before switching to oral therapy, based on a study which showed equal efficacy of oral and parenteral therapy [14]. However, identification of infants who are bacteremic may be important to reduce complications which might occur more often in bacteremic patients, such as severe sepsis and mortality [8]. In our cohort, however, all pediatric patients recovered, a finding that is con- sistent with the regular course of UTI in children even when accompa- nied by bacteremia [13]. Still, bacteremic patients more often required treatment in an intensive care unit. Children with bacteremia received longer courses of intravenous antibiotics and had longer average hospitalization. This is probably due to a more severe course or due to abnormal urologic anatomy (previously known or newly detected).

Our study has a number of potential limitations. First, because of its retrospective nature, some of the data were not available for all study patients, thus limiting the power of the analysis for those variables. For example, we do not have data regarding imaging studies that were done outside our medical center; in some cases, blood tests results were not available; and it is possible that data such as underlying condi- tions or having chills were not recorded in the medical records. In these cases, analysis was performed on all available data. Second, the study was conducted in a single medical center and thus may reflect the

local demographics, which do not necessarily apply to other popula- tions. Nevertheless, we believe that our study is important because it sheds some light on the epidemiology of and risk factors for bacteremic UTI in pediatric patients, specifically factors not related to the clinical judgment and personal experience of the treating physician. Most studies focused on risk factors for bacteremia in infants younger than 3 months, whereas in this study, we found risk factors for all age groups. Children with risk factors for bacteremia should be admitted and treated parenterally awaiting culture results.

In conclusion, we showed that bacteremic UTI is associated with high blood creatinine. Larger studies are required to guide the optimal length of Intravenous treatment required to reduce complications and to assess if treatment of bacteremic pediatric patients should be different from that of nonbacteremic pediatric patients.

References

1. Watt K, Waddle E, Jhaveri R. Changing epidemiology of Serious bacterial infections in febrile infants without localizing signs. PLoS One 2010;5:e12448.
2. Crain EF, Gershel JC. Urinary tract infections in febrile infants younger than 8 weeks of age. Pediatrics 1990;86:363-7.
3. Pitetti RD, Choi S. Utility of blood cultures in febrile children with UTI. Am J Emerg Med 2002;20:271-4.
4. Subcommittee on Urinary Tract Infection, Steering Committee on Quality Improvement and Management, Roberts KB. Urinary tract infection: clinical practice guideline for the diagnosis and management of the initial UTI in febrile infants and children 2 to 24 months. Pediatrics 2011;128:595-610.
5. Averbuch D, Nir-Paz R, Tenenbaum A, Stepensky P, Brooks R, Koplewitz BZ, et al. Fac- tors associated with bacteremia in young infants with urinary tract infection. Pediatr Infect Dis J 2014;33:571-5.
6. Bachur R, Caputo GL. Bacteremia and meningitis among infants with urinary tract in- fections. Pediatr Emerg Care 1995;11:280-4.
7. Ginsburg CM, McCracken Jr GH. Urinary tract infections in young infants. Pediatrics 1982;69:409-12.
8. Schnadower D, Kuppermann N, Macias CG, Freedman SB, Baskin MN, Ishimine P, et al. American Academy of Pediatrics Pediatric Emergency Medicine Collaborative Research Committee. Febrile infants with urinary tract infections at very low risk for adverse events and bacteremia. Pediatrics 2010;126:1074-83.
9. Velasco-Zuniga R, Trujillo-Wurttele JE, Fernandez-Arribas JL, Serrano-Carro B, Campo-Fernandez N, Puente-Montes S. Predictive factors of low risk for bacteremia in infants with urinary tract infection. Pediatr Infect Dis J 2012;31:642-5.
10. Downey LC, Benjamin Jr DK, Clark RH, Watt KM, Hornik CP, Laughon MM, et al. Uri- nary tract infection concordance with positive blood and cerebrospinal fluid cultures in the Neonatal intensive care unit. J Perinatol 2013;33:302-6.
11. Bonadio W, Maida G. Urinary tract infection in outpatient febrile infants younger than 30 days of age: a 10-year evaluation. Pediatr Infect Dis J 2014;33: 342-4.
12. Toker O, Schwartz S, Segal G, Godovitch N, Schlesinger Y, Raveh D. A costly covenant: ritual circumcision and urinary tract infection. Isr Med Assoc J 2010;12:262-5.
13. Roman HK, Chang PW, Schroeder AR. Diagnosis and management of bacteremic uri- nary tract infection in infants. Hosp Pediatr 2015;5:1-8.
14. Hoberman A, Wald ER, Hickey RW, Baskin M, Charron M, Majd M, et al. Oral versus initial intravenous therapy for urinary tract infections in young febrile children. Pe- diatrics 1999;104(1 pt 1):79-86.