Objective. To study awareness of urinary tract infections (UTIs) by determining the diagnostic rate of first UTI in children <2 years of age and to estimate the minimum incidence.
Methods. Twenty-six of a total of 43 pediatric centers participated in the study. Sixty-four percent of the total childhood population <2 years of age was covered. The number of all children included in the study was used to calculate the diagnostic rate. Only UTI confirmed by either suprapubic aspiration (any growth) or midstream or bag samples with ≥100 000 bacteria/mL, together with a positive nitrite reaction, was used to estimate the minimum incidence.
Results. The mean diagnostic rate was 1.5% for boys (range, 0.7%–3.0%) and 1.7% for girls (range, 0.7%–2.9%). The diagnostic rate was significantly higher in June than in December. The mean incidence was 1.0% for both boys and girls (range, 0.3%–3.0% and 0.4%–2.9%, respectively). The minimum cumulative incidence at 2 years of age was estimated to be 2.2% for boys and 2.1% for girls.
Conclusion. This study suggests a high UTI awareness in Sweden as indicated by a higher diagnostic rate and, despite stricter diagnostic criteria, a higher incidence of UTI in children <2 years of age than previously reported. It is suggested that a high UTI awareness may reduce chronic renal failure because of pyelonephritic renal scarring.
- UTI =
- urinary tract infection
Recent reports suggest that renal failure because of pyelonephritic renal scarring or reflux nephropathy is becoming rare.1 ,2 However, in some countries, this is still a major cause of end-stage renal disease.3 The reason for these differences is not known. Circumstantial evidence suggests that early diagnosis and adequate treatment of urinary tract infections (UTIs) are the most important measures to prevent renal scarring.4Increased UTI awareness results in more frequent diagnosis and a higher detection rate.5 To study the UTI awareness, it is of interest to estimate the diagnostic rate, ie, the number of children given a diagnosis of UTI, true or false, and consequently being investigated. To our knowledge, this matter has not been studied previously. It is also important, because costs generated by treatment, primary investigation, and follow-up of children with UTI correlate with the diagnostic rate rather than with the incidence.
The true incidence of UTI in children is difficult to estimate. Published reports on symptomatic UTI reflect differences in general treatment, referral policy, and health care traditions. Winberg et al5 performed the first epidemiologic study in the 1960s and calculated a cumulative risk of 3% in girls and 1% in boys at 10 years of age. The highest incidence rates in the literature were obtained in retrospective studies from the late 1970s and the early 1980s, with a UTI risk of 1% in both boys and girls during the first year of life and a cumulative risk of 1.7% in boys and 7.8% in girls up to 7 years of age.6–8
The aim of the present article was to study the UTI awareness by assessing the number of children with a final diagnosis of UTI, ie, the diagnostic rate, to compare the diagnostic rate between different pediatric centers, to evaluate the basis for diagnosis, and to estimate from these data the minimum incidence of UTI in children <2 years of age.
PATIENTS AND METHODS
This study was part of a quality assurance project initiated by the Swedish Pediatric Nephrology Association.9 All pediatric departments in the country were invited to participate. Twenty-six of a total of 43 pediatric departments participated in the project. For the period September 1, 1993 through August 31, 1995, all children <2 years of age with the first recognized symptomatic UTI were prospectively included in the program. Children with myelomeningocele or overt malformations of the genitals were excluded. There was no information on the frequency of circumcision in the protocol. This procedure is rarely performed in Sweden.
Criteria for UTI Diagnosis
No predetermined criteria for inclusion were given. The responsible pediatrician at each participating center included all children considered to have UTI according to the results of urinalysis, culture, and clinical judgment. The diagnosis of UTI was considered to be definitely confirmed when the urine sample was obtained by either a suprapubic bladder aspiration (any growth), or by catheter, midstream, or bag samples with ≥100 000 bacteria/mL of one species together with a positive nitrite reaction.
Population at Risk
The yearly census in each center area, according to official statistics for the study period, was used to estimate the population at risk. On December 31st each year, the exact number of children <2 years of age was available for each center. The three largest urban regions (regions 1–3) in the country accounted for 39% of the population at risk. In these regions, the health care service is organized in a uniform manner. Moreover, most previous epidemiologic studies on children with UTI in Sweden have originated from one of these urban regions (region 2).5–8 Therefore, the results from these regions are shown separately.
The total population of children <2 years of age at risk was 145 528 during the first year and 137 747 during the second year. The average number of children at risk during the study period was 141 638 (72 660 boys and 68 978 girls). This corresponds to 64% of all Swedish children in that age group. Altogether 1111 boys, 967 <1 year and 144 between 1 and 2 years of age, and 1198 girls, 827 <1 year and 371 between 1 and 2 years of age, were detected with their first UTI. The age at first UTI is shown in Fig 1. All children were diagnosed because of acute symptoms, mainly fever. A temperature of ≥38.5°C was noted in 77% of the children and ≥38.0°C in 86%.
The basis for the diagnosis and the criteria for the diagnostic rate and the minimum incidence are shown in Table 1.
The mean diagnostic rate for all 26 centers was 1.3% (range, 0.6–2.6) for boys 0 to 1 year old and 0.2% (range, 0.0–0.6) for boys 1 to 2 years old. For girls the corresponding figures were 1.2% (range, 0.5–2.1) and 0.5% (range, 0.2–1.2), respectively. The diagnostic rates in the three largest urban regions and in the remaining participating centers are shown in Table 2. The diagnostic rate was similar in boys and girls <1 year of age, but in children between 1 and 2 years of age the rate was higher in girls. The diagnostic rate was higher in region 2 than in the other regions.
There was a significant decrease in diagnostic rate during the study period, the total diagnostic rate being 1.8% during the first year and 1.4% during the second year (P < .001) (Fig 2). There was also a significant seasonal variation in diagnostic rate during the study period (P< .05). For example, the estimated risk of having UTI was 14% higher in June than in December.
The mean minimum incidence for all 26 centers was 1.0% (range, 0.3–2.6) for boys 0 to 1 year old and 0.1% (range, 0.0–0.4) for boys 1 to 2 years old. For girls, the corresponding figures were 0.8% (range, 0.3–2.1) and 0.3% (range, 0.0–0.9), respectively. The results for the three urban regions and the remaining participating centers are shown in Table 3. The cumulative incidences were similar for boys and girls, but were higher in region 2 than in the other two urban regions.
This study was part of a quality assurance project, aiming to provide each pediatric center in Sweden with an instrument to evaluate the care of children with the diagnosis of UTI. Because so many pediatric centers participated in the project, it was of interest to study the diagnostic rate of UTI in each region. All children with the diagnosis of UTI consume resources in the health care system, without respect to the reliability of the diagnosis. Moreover, in regions where the number of children with a reliable diagnosis is high, the diagnostic rate will correspond to the minimum figures for the incidence.
The variation in diagnostic rate and incidence between the different regions suggests differences in UTI awareness, although differences in referral policy and health care organization cannot be excluded. The figures in region 2 suggest high UTI awareness. This is in accordance with the long tradition of research on the subject in this area.5–8 We therefore believe that the figures reported from region 2 more accurately reflect the minimum incidence of UTI in the country. They indicate a minimum incidence of 2.2% for boys and 2.1% for girls <2 years of age.
In this study, the diagnostic criteria used to calculate the minimum incidence are stricter than in previous studies. Still, the minimum incidence reported in the present study is higher. In a study from Finland,12 an incidence of 0.9% was found in boys and girls <1 year of age with symptomatic UTI (recurrent infections included). In a recent retrospective analysis from Sweden in an area corresponding to region 2, covering the years 1978 to 1981, an incidence of 1.1% was reported for both boys and girls during the first year,8 a higher rate than in the older study by Winberg et al.5 Now, almost 20 years later, we report an incidence, using stricter criteria for the diagnosis of UTI, that is considerably higher. This increase in incidence in boys cannot be explained by a change in the frequency of circumcision.
The diagnostic rate in the present study was higher than the incidence, indicating that the reliability of the urine sampling technique was not optimal in a substantial number of patients. Still, many of the children not meeting our criteria for a definitely confirmed diagnosis may have had UTI and, what is more important, they underwent primary investigation and follow-up in the same manner as those with a confirmed diagnosis. In a previous report,9 we found that the frequencies of vesicoureteral reflux in the children with confirmed diagnosis were similar to those diagnosed with a less reliable urine sampling technique. This suggests that when less reliable techniques were used, such as bag samples, other efforts to secure the diagnosis were made, for example by requiring more than one urine sample before start of treatment. Moreover, urine samples with a bacterial count <100 000 bacteria/mL may be associated with UTI.13 In the present study, these cases were excluded by the criteria used to estimate the minimum incidence. This may have further lowered the incidence in relation to the diagnostic rate. The best way to minimize the risk of bacterial contamination and to detect UTI with low bacterial counts is to obtain urine samples with a suprapubic bladder puncture.
In a recent report from England,14 the mean annual referral rate for children 0 to 16 years of age with UTI was 0.46% and for boys and girls <1 year of age the referral rate was 0.7%. The annual referral rate was defined as first referral for UTI and not all were bacteriologically proved. The annual referral rate in the above study is therefore comparable with the diagnostic rate in the present study. The mean diagnostic rate was, in the present study, approximately twice as high in infants and in some centers almost three times higher than in the above study from England. This cannot be explained by overdiagnosis of UTI in Sweden. It seems more likely that the two studies reflect differences in detection rate and UTI awareness. In this regard, it is of interest to note that in Sweden no new cases of end-stage renal disease because of pyelonephritic scarring or reflux nephropathy have been detected since 1986,1although this is still frequently reported in the United Kingdom.3 Taken together, these findings suggest that there may be an inverse relationship between a high UTI awareness and the presence of pyelonephritic scarring as a cause of end-stage renal failure. However, one should be aware that end-stage renal disease may be classified differently in the two countries.
We found a higher risk of attracting UTI in June than in December. In a study on symptomatic UTI in women there was a definite seasonal fluctuation with peaking in August.15 In contrast to these findings, Elo et al16 in a series of almost 1000 verified episodes of UTI in children, including also teenagers, collected during a 10-year period, found the highest frequency in November and the lowest during the summer months. Differences in selection and size of study populations may contribute to these conflicting results. The clinical and epidemiologic significance of seasonal variation remains to be determined.
In accordance with previous studies,8 ,12 we observed a significant decrease in the diagnostic rate during the study period. It is not unlikely that, in this prospective study, the awareness of UTI may have diminished during the 2-year-long study period. This is supported by the fact that this tendency was not found in the three largest centers which are continuously engaged in UTI research.
In summary, this study suggests a high UTI awareness in Sweden, indicated by a higher diagnostic rate of UTI compared with previous studies and, despite stricter diagnostic criteria, a higher incidence of UTI in children <2 years of age than has previously been reported. It is possible that a high UTI awareness in infants and children may reduce chronic renal failure because of pyelonephritic renal scarring.
The study was supported by grants from the Karolinska Institute and the Swedish Medical Research Council.
We thank Anders Odén, PhD, for assistance with the statistics.
The following members of the Swedish Pediatric Nephrology Association contributed to the study: Eva Nyberg, MD, Danderyd; Mats Hölke, MD, Eskilstuna; Gun Forsander, MD, Falun; Åke Stenberg, MD, Gällivare; Björn Wettergren, MD, Gävle; Mats Aili, MD, Halmstad; Dorthe Jensen, MD, Helsingborg; Kerstin Berglund, MD, Hudiksvall; Per Brandström, MD, Jönköping; Per-Olof Lübeck, MD, Karlstad; Ingrid Sjöberg, MD, Malmö; Staffan Mårild, MD, Mölndal; Per Lewander, MD, Norrköping; Marianne Sannum, MD, NÄL; Svante Swerkersson, MD, Skövde; Ingela Bollgren, MD, Stockholm; Tomas Almström, MD, Uddevalla; Anders Dannaeus, MD, Uppsala; Inger Fosdal, MD, Visby; Margaret Aldman, MD, Västervik; Wanda Sikorska, MD, Västerås; Kerstin Abelson-Storby, MD, Växjö; Solveig Richter, MD,Örnsköldsvik; and Monica Jonsson, MD, Östersund.
- Received November 30, 1998.
- Accepted February 16, 1999.
- Address correspondence to Birgir Jakobsson, MD, Department of Pediatrics, Huddinge University Hospital, S-141 86 Huddinge, Sweden. E-mail:
- Imam A,
- Roberts R,
- Verrier Jones K
- Smellie JM,
- Poulton A,
- Prescod N
- ↵Winberg J, Andersen H, Bergström T, Jacobsson B, Larsson H, Lincoln K. Epidemiology of urinary tract infection in childhood. Acta Paediatr Scand. 1974;252:1–20. Supplement
- Hellström A-L,
- Hanson E,
- Hansson S,
- Hjälmås K,
- Jodal U
- ↵Breslow NE, Day NE. Statistical Methods in Cancer Research, II. Lyon, France: IARC Scientific Publications; 1987;32:131–135
- Uhari M,
- Nuutinen M
- Coulthard MG,
- Lambert HJ,
- Keir MJ
- Anderson JE
- Copyright © 1999 American Academy of Pediatrics