Original Article
Factors Associated with Abnormal Imaging and Infection Recurrence after a First Febrile Urinary Tract Infection in Children Marko Tapani Ristola1
Eliisa Löyttyniemi2
Timo Hurme1
1 Department of Pediatric Surgery, Turku University Hospital, Turku,
Varsinais-Suomi, Finland 2 Department of Biostatistics, Turku University Hospital, Turku, Varsinais-Suomi, Finland
Address for correspondence Marko Tapani Ristola, MD, Department of Pediatric Surgery, Turku University Hospital, PO box 52, 20521 Turku, Varsinais-Suomi 20100, Finland (e-mail: mtrist@utu.fi).
Abstract
Keywords
► urinary tract infections ► vesicoureteral reflux ► diagnostic imaging ► children
Introduction We determined factors associated with abnormal imaging and recurrent infections after a first febrile urinary tract infection (UTI) in children younger than 3 years. Materials and Methods We retrospectively reviewed the records of all patients treated at our institute during the years 2000–2009, for a first febrile UTI in children younger than 3 years, who underwent ultrasonography and voiding cystourethrography. We evaluated data regarding factors potentially associated with abnormal ultrasonography and voiding cystourethrography results and recurrence of infections, and formulated a risk score system to assess risk of reflux and high-grade reflux. Results There were 282 patients. The only factor predicting abnormal ultrasonogram was non–Escherichia coli infection. Risk factors for vesicoureteral reflux included abnormal ultrasonogram, atypical infection, non–E. coli infection and infection recurrence. Patients with no identified risk factors for vesicoureteral reflux were unlikely to have high-grade reflux. Higher risk scores were associated with a higher risk for reflux. Non–E. coli infection was the only statistically significant predictor of infection recurrence. Conclusion All children younger than 3 years with first febrile UTI should undergo ultrasonography. Thereafter, patients with no predictive factors for vesicoureteral reflux may be followed up without further imaging. A non–E. coli infection is associated with reflux and infection recurrence.
Introduction In the past, small children with a febrile urinary tract infection (UTI) have undergone extensive imaging, including renal and bladder ultrasonography (RBUS), voiding cystourethrography (VCUG), and, in some cases, dimercaptosuccinic acid (DMSA) scintigraphy, to identify urinary tract anomalies including hydronephrosis, obstruction, bladder and urethral anomalies,
received September 23, 2015 accepted after revision December 22, 2015
vesicoureteral reflux (VUR), and renal scarring. The significance of VUR and the effectiveness of antimicrobial prophylaxis have been challenged,1–4 which has in turn lead to the development of guidelines that promote a more restrictive approach to imaging.5–8 Of these guidelines, the ones issued by the American Academy of Pediatrics (AAP)6 and the National Institute for Health and Care Excellence (NICE)5 have raised the most conversation and controversy.9–15 They both limit the use of VCUG
© Georg Thieme Verlag KG Stuttgart · New York
DOI http://dx.doi.org/ 10.1055/s-0036-1572418. ISSN 0939-7248.
Downloaded by: Mahidol University. Copyrighted material.
Eur J Pediatr Surg
to a subpopulation of patients thought to be at highest risk for both VUR and progressive renal damage. While some factors, including family history of VUR, recurrence of a UTI, and abnormal RBUS findings, are generally thought to be associated with a higher incidence of VUR, studies on some of the associations are scarce. In this study, we evaluated these factors potentially associated with abnormal RBUS and VCUG, and recurrent UTIs, to identify both a population that may benefit from further imaging studies and one that may be followed up without extensive imaging. The hypothesis was that in patients with normal RBUS and no atypical or complicating factors after a first febrile UTI, the incidence of high-grade VUR and recurrence of UTIs are low. The factors to be studied were selected based on their potential association determined in previous studies.
Materials and Methods Patients We retrospectively reviewed the medical records of all patients younger than 3 years treated for culture proven, febrile UTI in Turku University Hospital between January 1, 2000, and December 31, 2009. All patients’ prenatal ultrasonography results were reviewed and found normal. All selected patients underwent both RBUS and either VCUG or nuclear VCUG after a UTI. Exclusion criteria included any previously known anatomical or neurological abnormalities associated with UTIs or VUR, having the imaging studies performed at another institution or moving to another health care district during follow-up.
Urinary Tract Infection Diagnosis The UTI was considered certain if there was any growth of a single uropathogen in a suprapubic aspiration (SPA), or growth of a single uropathogen in the amount of 100,000 colony-forming units per milliliter (CFU/mL) in one or more samples of clean catch urine or bag specimen with urinalysis results consistent with UTI (positive leukocyte esterase test or nitrite test or microscopic analysis positive for leukocytes or bacteria). Catheter samples were not used. SPA was often attempted unsuccessfully, resulting in a large amount of young patients’ diagnoses being based on bag specimens. The term “febrile” was determined as 38.0°C (100.4°F).
Urinary Tract Infection Treatment Antibiotics were administered according to national guidelines, and the antibiogram of the uropathogen in each case. If antibiotic treatment was started before the antibiogram was available, the drug of choice was intravenous cefuroxime. Most patients received intravenous cefuroxime for 3 days followed by an oral antibiotic after discharge from the hospital for a total treatment period of 10 days. The most common oral antibiotics were trimethoprim/sulfamethoxazole, amoxicillin/clavulanic acid, and cephalexin.
Data Gathered The patients’ medical records were reviewed and data were gathered regarding fever, plasma C-reactive protein (CRP) levels, plasma creatinine levels, urine cultures, blood bacterial European Journal of Pediatric Surgery
Ristola et al.
cultures, family history of VUR, poor urine flow reported by the parents, findings in RBUS and VCUG, antimicrobial prophylaxis, antireflux procedures and other urological procedures, and recurrence of UTIs (both febrile and afebrile). Abnormal RBUS was defined as any abnormal finding as reported by the radiologist. High-grade VUR was determined as grades III–V. We compared these findings to establish predictors of abnormal imaging results and UTI recurrence. For the purposes of the analyses in this study, the following were defined as “atypical UTI”: infection with a non–Escherichia coli organism, elevated plasma creatinine concentration, septicemia, failure to respond to appropriate antibiotics in 48 hours, poor urine flow (as reported by the child’s parents or attending physician), or a family history of VUR. Elevated plasma creatinine was determined as 44 μmol/L. Septicemia was determined as growth of the causative uropathogen in blood bacterial culture. Family history of VUR was determined as a sibling or parent having been diagnosed with VUR. Phimosis was determined as obstructed urine flow due to constriction of the orifice of the prepuce. For factors associated with VUR, the criteria for “recurrent UTI” were fulfilled if the child had another culture confirmed UTI during the time period between the first febrile UTI and the VCUG. All patients’ medical records were reviewed for a median follow-up time of 9.8 years (range: 3.9–13.9 years).
Vesicoureteral Reflux Diagnosis Where possible, the VUR of patients was graded according to the international system of radiographic grading of vesicoureteric reflux,16 and the highest grade detected was recorded. The common finding in nuclear VCUG, “VUR reaching the level of the kidney,” with no reference to dilation, was defined as grade II. However, data regarding dilation in RBUS were combined with the nuclear VCUG finding to determine the grade of VUR as accurately as possible. This was done in only one patient, who had VUR in nuclear VCUG and hydronephrosis in RBUS. In bilateral VUR, the higher grade detected was used for classification.
Analyses We determined the statistical associations between the aforementioned clinical factors and (1) RBUS results, (2) VCUG results, and (3) UTI recurrence, to define which patients might be at higher risk for abnormal imaging and/or UTI recurrence. We also formulated a scoring system for determining the risk for high-grade VUR, based on these correlations and current knowledge on the predisposing factors for VUR. The system and points afforded were determined according to both the results of previous studies and associations found in this study.
Statistical Methods Data are described as frequencies and proportions. In addition, median was calculated for age. To determine important associated factors for abnormal RBUS, VUR, high-grade VUR, and UTI recurrence, binary logistic regression models for these responses were generated with multiple factors included.
Downloaded by: Mahidol University. Copyrighted material.
Abnormal Imaging and Infection Recurrence after a First Febrile UTI
Abnormal Imaging and Infection Recurrence after a First Febrile UTI
Results Altogether, there were 282 patients, and median age was 5 months. There were 135 boys (median age: 2 months) and 147 girls (median age: 9 months). Fever of 39.0°C (102.2°F) or higher was noted in 190 patients (67%). Fever of 40.0°C (104.0° F) or higher was noted in 62 patients (22%). Plasma CRP levels
were measured in 276 patients (98%), and were higher than 40 mg/L in 203 patients (74%). Blood bacterial cultures were taken in 252 patients (89%), and were positive in 17 patients (6.7%). Plasma creatinine levels were measured in 94 patients (33%), and were raised in 29 patients (31%). During the entire followup period, atypical UTI was noted in 82 patients (29%), of which 77 (27%) had an atypical first UTI. The uropathogen was other than E. coli in 36 patients (13%), of which 31 (11%) had a non–E. coli infection before VCUG (►Table 1). Poor urine flow during the first UTI was reported by the parents of three patients (1.1%). Phimosis was present in two patients (0.71%). Failure to respond to suitable antibiotics during 48 hours was noted in three patients (1.1%). UTI recurrence during follow-up was recorded in 57 patients (20%), of which 38 patients (13%) had a UTI recurrence before VCUG. Family history of VUR was present in four patients (1.4%) (►Table 2).
Renal and Bladder Ultrasonography RBUS findings were abnormal in 90 patients (32%). The only factor with a strong statistical association with abnormal RBUS was non–E. coli infection. Out of the 31 patients with a non–E. coli infection, 18 (58%) had abnormal RBUS results. Gender, age, level of fever, plasma CRP level, positive blood bacterial culture, failure to respond to suitable antibiotics during 48 hours, atypical infection, poor urine flow, phimosis, and family history of VUR were not statistically significantly associated with abnormal RBUS findings. Interestingly, a raised plasma creatinine concentration was associated with a normal RBUS. Out of the 202 patients with no potentially predictive factors for abnormal RBUS, 63 patients (31%) had abnormal RBUS results (see ►Table 3 for p-values).
Fig. 1 Incidence of vesicoureteral reflux (VUR) according to risk score. Higher scores were significantly (p < 0.0001) associated with a higher incidence of VUR. European Journal of Pediatric Surgery
Downloaded by: Mahidol University. Copyrighted material.
The abnormal RBUS model included non–E. coli infection, gender, poor response to antibiotics, positive blood culture, atypical infection, poor urine flow, phimosis, and family history of VUR as categorical factors as well as age, fever, and CRP as numerical factors. The VCUG model included non– E. coli infection, atypical UTI, UTI recurrence, abnormal RBUS, gender, raised plasma creatinine, positive blood culture, poor response to antibiotics, poor urine flow, phimosis, and family history of VUR as categorical factors as well as age, fever, and CRP as numerical factors. The UTI recurrence model included non–E. coli infection, gender, poor response to antibiotics, positive blood culture, and abnormal RBUS as categorical factors as well as age, fever, and CRP as numerical factors. For significant predictors of VUR and high-grade VUR, sensitivity (%), specificity (%), positive predictive value (PPV), and negative predictive value (NPV) were also calculated. For evaluating the risk score system, Cochran-Armitage trend test was performed between the risk score and VUR, and between the risk score and high-grade VUR (►Figs. 1 and 2). p-Values of
Factors Associated with Abnormal Imaging and Infection Recurrence after a First Febrile Urinary Tract Infection in Children Marko Tapani Ristola1
Eliisa Löyttyniemi2
Timo Hurme1
1 Department of Pediatric Surgery, Turku University Hospital, Turku,
Varsinais-Suomi, Finland 2 Department of Biostatistics, Turku University Hospital, Turku, Varsinais-Suomi, Finland
Address for correspondence Marko Tapani Ristola, MD, Department of Pediatric Surgery, Turku University Hospital, PO box 52, 20521 Turku, Varsinais-Suomi 20100, Finland (e-mail: mtrist@utu.fi).
Abstract
Keywords
► urinary tract infections ► vesicoureteral reflux ► diagnostic imaging ► children
Introduction We determined factors associated with abnormal imaging and recurrent infections after a first febrile urinary tract infection (UTI) in children younger than 3 years. Materials and Methods We retrospectively reviewed the records of all patients treated at our institute during the years 2000–2009, for a first febrile UTI in children younger than 3 years, who underwent ultrasonography and voiding cystourethrography. We evaluated data regarding factors potentially associated with abnormal ultrasonography and voiding cystourethrography results and recurrence of infections, and formulated a risk score system to assess risk of reflux and high-grade reflux. Results There were 282 patients. The only factor predicting abnormal ultrasonogram was non–Escherichia coli infection. Risk factors for vesicoureteral reflux included abnormal ultrasonogram, atypical infection, non–E. coli infection and infection recurrence. Patients with no identified risk factors for vesicoureteral reflux were unlikely to have high-grade reflux. Higher risk scores were associated with a higher risk for reflux. Non–E. coli infection was the only statistically significant predictor of infection recurrence. Conclusion All children younger than 3 years with first febrile UTI should undergo ultrasonography. Thereafter, patients with no predictive factors for vesicoureteral reflux may be followed up without further imaging. A non–E. coli infection is associated with reflux and infection recurrence.
Introduction In the past, small children with a febrile urinary tract infection (UTI) have undergone extensive imaging, including renal and bladder ultrasonography (RBUS), voiding cystourethrography (VCUG), and, in some cases, dimercaptosuccinic acid (DMSA) scintigraphy, to identify urinary tract anomalies including hydronephrosis, obstruction, bladder and urethral anomalies,
received September 23, 2015 accepted after revision December 22, 2015
vesicoureteral reflux (VUR), and renal scarring. The significance of VUR and the effectiveness of antimicrobial prophylaxis have been challenged,1–4 which has in turn lead to the development of guidelines that promote a more restrictive approach to imaging.5–8 Of these guidelines, the ones issued by the American Academy of Pediatrics (AAP)6 and the National Institute for Health and Care Excellence (NICE)5 have raised the most conversation and controversy.9–15 They both limit the use of VCUG
© Georg Thieme Verlag KG Stuttgart · New York
DOI http://dx.doi.org/ 10.1055/s-0036-1572418. ISSN 0939-7248.
Downloaded by: Mahidol University. Copyrighted material.
Eur J Pediatr Surg
to a subpopulation of patients thought to be at highest risk for both VUR and progressive renal damage. While some factors, including family history of VUR, recurrence of a UTI, and abnormal RBUS findings, are generally thought to be associated with a higher incidence of VUR, studies on some of the associations are scarce. In this study, we evaluated these factors potentially associated with abnormal RBUS and VCUG, and recurrent UTIs, to identify both a population that may benefit from further imaging studies and one that may be followed up without extensive imaging. The hypothesis was that in patients with normal RBUS and no atypical or complicating factors after a first febrile UTI, the incidence of high-grade VUR and recurrence of UTIs are low. The factors to be studied were selected based on their potential association determined in previous studies.
Materials and Methods Patients We retrospectively reviewed the medical records of all patients younger than 3 years treated for culture proven, febrile UTI in Turku University Hospital between January 1, 2000, and December 31, 2009. All patients’ prenatal ultrasonography results were reviewed and found normal. All selected patients underwent both RBUS and either VCUG or nuclear VCUG after a UTI. Exclusion criteria included any previously known anatomical or neurological abnormalities associated with UTIs or VUR, having the imaging studies performed at another institution or moving to another health care district during follow-up.
Urinary Tract Infection Diagnosis The UTI was considered certain if there was any growth of a single uropathogen in a suprapubic aspiration (SPA), or growth of a single uropathogen in the amount of 100,000 colony-forming units per milliliter (CFU/mL) in one or more samples of clean catch urine or bag specimen with urinalysis results consistent with UTI (positive leukocyte esterase test or nitrite test or microscopic analysis positive for leukocytes or bacteria). Catheter samples were not used. SPA was often attempted unsuccessfully, resulting in a large amount of young patients’ diagnoses being based on bag specimens. The term “febrile” was determined as 38.0°C (100.4°F).
Urinary Tract Infection Treatment Antibiotics were administered according to national guidelines, and the antibiogram of the uropathogen in each case. If antibiotic treatment was started before the antibiogram was available, the drug of choice was intravenous cefuroxime. Most patients received intravenous cefuroxime for 3 days followed by an oral antibiotic after discharge from the hospital for a total treatment period of 10 days. The most common oral antibiotics were trimethoprim/sulfamethoxazole, amoxicillin/clavulanic acid, and cephalexin.
Data Gathered The patients’ medical records were reviewed and data were gathered regarding fever, plasma C-reactive protein (CRP) levels, plasma creatinine levels, urine cultures, blood bacterial European Journal of Pediatric Surgery
Ristola et al.
cultures, family history of VUR, poor urine flow reported by the parents, findings in RBUS and VCUG, antimicrobial prophylaxis, antireflux procedures and other urological procedures, and recurrence of UTIs (both febrile and afebrile). Abnormal RBUS was defined as any abnormal finding as reported by the radiologist. High-grade VUR was determined as grades III–V. We compared these findings to establish predictors of abnormal imaging results and UTI recurrence. For the purposes of the analyses in this study, the following were defined as “atypical UTI”: infection with a non–Escherichia coli organism, elevated plasma creatinine concentration, septicemia, failure to respond to appropriate antibiotics in 48 hours, poor urine flow (as reported by the child’s parents or attending physician), or a family history of VUR. Elevated plasma creatinine was determined as 44 μmol/L. Septicemia was determined as growth of the causative uropathogen in blood bacterial culture. Family history of VUR was determined as a sibling or parent having been diagnosed with VUR. Phimosis was determined as obstructed urine flow due to constriction of the orifice of the prepuce. For factors associated with VUR, the criteria for “recurrent UTI” were fulfilled if the child had another culture confirmed UTI during the time period between the first febrile UTI and the VCUG. All patients’ medical records were reviewed for a median follow-up time of 9.8 years (range: 3.9–13.9 years).
Vesicoureteral Reflux Diagnosis Where possible, the VUR of patients was graded according to the international system of radiographic grading of vesicoureteric reflux,16 and the highest grade detected was recorded. The common finding in nuclear VCUG, “VUR reaching the level of the kidney,” with no reference to dilation, was defined as grade II. However, data regarding dilation in RBUS were combined with the nuclear VCUG finding to determine the grade of VUR as accurately as possible. This was done in only one patient, who had VUR in nuclear VCUG and hydronephrosis in RBUS. In bilateral VUR, the higher grade detected was used for classification.
Analyses We determined the statistical associations between the aforementioned clinical factors and (1) RBUS results, (2) VCUG results, and (3) UTI recurrence, to define which patients might be at higher risk for abnormal imaging and/or UTI recurrence. We also formulated a scoring system for determining the risk for high-grade VUR, based on these correlations and current knowledge on the predisposing factors for VUR. The system and points afforded were determined according to both the results of previous studies and associations found in this study.
Statistical Methods Data are described as frequencies and proportions. In addition, median was calculated for age. To determine important associated factors for abnormal RBUS, VUR, high-grade VUR, and UTI recurrence, binary logistic regression models for these responses were generated with multiple factors included.
Downloaded by: Mahidol University. Copyrighted material.
Abnormal Imaging and Infection Recurrence after a First Febrile UTI
Abnormal Imaging and Infection Recurrence after a First Febrile UTI
Results Altogether, there were 282 patients, and median age was 5 months. There were 135 boys (median age: 2 months) and 147 girls (median age: 9 months). Fever of 39.0°C (102.2°F) or higher was noted in 190 patients (67%). Fever of 40.0°C (104.0° F) or higher was noted in 62 patients (22%). Plasma CRP levels
were measured in 276 patients (98%), and were higher than 40 mg/L in 203 patients (74%). Blood bacterial cultures were taken in 252 patients (89%), and were positive in 17 patients (6.7%). Plasma creatinine levels were measured in 94 patients (33%), and were raised in 29 patients (31%). During the entire followup period, atypical UTI was noted in 82 patients (29%), of which 77 (27%) had an atypical first UTI. The uropathogen was other than E. coli in 36 patients (13%), of which 31 (11%) had a non–E. coli infection before VCUG (►Table 1). Poor urine flow during the first UTI was reported by the parents of three patients (1.1%). Phimosis was present in two patients (0.71%). Failure to respond to suitable antibiotics during 48 hours was noted in three patients (1.1%). UTI recurrence during follow-up was recorded in 57 patients (20%), of which 38 patients (13%) had a UTI recurrence before VCUG. Family history of VUR was present in four patients (1.4%) (►Table 2).
Renal and Bladder Ultrasonography RBUS findings were abnormal in 90 patients (32%). The only factor with a strong statistical association with abnormal RBUS was non–E. coli infection. Out of the 31 patients with a non–E. coli infection, 18 (58%) had abnormal RBUS results. Gender, age, level of fever, plasma CRP level, positive blood bacterial culture, failure to respond to suitable antibiotics during 48 hours, atypical infection, poor urine flow, phimosis, and family history of VUR were not statistically significantly associated with abnormal RBUS findings. Interestingly, a raised plasma creatinine concentration was associated with a normal RBUS. Out of the 202 patients with no potentially predictive factors for abnormal RBUS, 63 patients (31%) had abnormal RBUS results (see ►Table 3 for p-values).
Fig. 1 Incidence of vesicoureteral reflux (VUR) according to risk score. Higher scores were significantly (p < 0.0001) associated with a higher incidence of VUR. European Journal of Pediatric Surgery
Downloaded by: Mahidol University. Copyrighted material.
The abnormal RBUS model included non–E. coli infection, gender, poor response to antibiotics, positive blood culture, atypical infection, poor urine flow, phimosis, and family history of VUR as categorical factors as well as age, fever, and CRP as numerical factors. The VCUG model included non– E. coli infection, atypical UTI, UTI recurrence, abnormal RBUS, gender, raised plasma creatinine, positive blood culture, poor response to antibiotics, poor urine flow, phimosis, and family history of VUR as categorical factors as well as age, fever, and CRP as numerical factors. The UTI recurrence model included non–E. coli infection, gender, poor response to antibiotics, positive blood culture, and abnormal RBUS as categorical factors as well as age, fever, and CRP as numerical factors. For significant predictors of VUR and high-grade VUR, sensitivity (%), specificity (%), positive predictive value (PPV), and negative predictive value (NPV) were also calculated. For evaluating the risk score system, Cochran-Armitage trend test was performed between the risk score and VUR, and between the risk score and high-grade VUR (►Figs. 1 and 2). p-Values of
