Patients and methods

Results

After exclusion of cases not meeting the criteria, 164 sets of case notes (51 boys, 113 girls) were examined (table 1). Table 2 shows results of imaging investigations for inpatients and outpatients. Table3 shows the abnormalities detected on imaging, carried out in accordance with the guidelines. In addition, four of 20 children (20%) aged 1 to 6 years had VUR on MCUG and two of six children (33%) over 7 years (one of three inpatients and one of three outpatients) had scarring on DMSA scans, although these tests were additional to the minimum recommended in the guidelines. A smaller proportion of infants were diagnosed at home than of other age groups (p < 0.01). The prevalence of renal scarring on DMSA scan in children over 1 year was lower in children diagnosed and treated at home compared to children admitted to hospital (p < 0.01). The time between diagnosis of UTI and DMSA scan was 4.3 months and 3.6 months, respectively, for children under 1 and 1–6 years admitted to hospital, and 5.0 months and 4.7 months for children managed at home. The time interval between diagnosis of UTI and DMSA scan for children with normal scans was 4.4 months and 3.5 months, and for abnormal scans was 4.0 months and 4.0 months, for children under 1 year and children aged 1–6 years, respectively. Children with VUR or renal scarring were advised to continue prophylaxis for one to two years, while children with normal scans were advised to stop prophylaxis.

Discussion

The guidelines on management of childhood UTI were published in 1991 as a consensus statement following a one day meeting of clinicians interested in UTI in childhood.5 There was general agreement that children with UTI benefited from imaging investigations, although evidence that these tests alter the immediate or long term outcome is weak.10 The reason stated for benefit from tests was that “a significant number of important abnormalities will be detected in both sexes”. However, no evidence was provided in support of the hypothesis that doing the tests altered the outcome, other than providing the opportunity to discuss the abnormality and anticipate complications. There was significant disagreement between participants about the extent to which young children should be investigated and the most appropriate tests to show renal scarring at different ages. There was agreement that all children should have an ultrasound scan following UTI and that the youngest children should have the most thorough investigation, but there was no clear consensus on the extent of other investigations. The working group made a set of recommendations, as an interim measure, for the minimum imaging investigations in straightforward cases, recognising that further research would be required to enable definitive guidelines to be established.

Ultrasound was recommended for all children, while DMSA scintigraphy was recommended in addition for all children under 7 years; micturating cystography was recommended in addition for infants under 12 months. Although the guidelines stated that the need for these tests should be reviewed when results of further research became available, there has been little published evidence challenging the guidelines which are now widely used by paediatricians throughout the UK. In contrast to practice in many other countries, in the UK little attempt has been made to distinguish clinically between lower tract infection and acute pyelonephritis beyond the evidence obtained by history and examination, and there was no attempt to make this potentially important distinction in the guidelines. Consequently many children with trivial illnesses, diagnosed and treated at home by general practitioners, are now routinely subjected to DMSA scans in addition to ultrasound, although there is some doubt about the value of this. Many children under 7 years also receive long term low dose prophylaxis while awaiting imaging investigations without clear evidence that this is effective in preventing renal scar formation. Some parents now question the wisdom of long term prophylaxis for several months following a trivial illness and compliance has been shown to be poor.11

These imaging tests are not done to aid the diagnosis or treatment of the acute illness but are carried out later when the child has recovered fully and are therefore a form of screening carried out in a high risk population. These children are perceived to be at risk of having underlying anomalies and of acquiring renal damage. However, the World Health Organisation's criteria for a successful screening programme are not met.12 The detection of anomalies in a third of cases is not sufficient justification for carrying out these tests. There should be some evidence that the new knowledge derived from the investigation can alter management and outcome to justify screening.

Although much is known about the natural history of UTI, VUR, and renal scarring, little evidence exists that the natural history can be altered. Scars remain static or progress in a proportion of children, even when given excellent care and supervision.13 ,14 This may be in part because UTI is easily overlooked in infants at the age when there is the greatest risk of renal scarring.9 Much of the preventable damage may have occurred prior to entry into the two largest controlled studies.13 ,14

Only 23% of infants with UTI included in our study were diagnosed in primary care, compared with 77% managed in hospital; in older children the ratio was reversed. In a pilot study, van der Voortet al showed that children with UTI had visited their general practitioners twice as often as controls prior to diagnosis of the first UTI.15 This suggests that UTI may have been overlooked, thus predisposing to renal scar formation.

We have used a simple process, hospital admission, to identify children who were likely to have had severe or systemic illness and to distinguish them from those who were treated at home and who were more likely to have had a trivial illness. Of the children under 1 year, 13/56 (23%) were diagnosed in primary care compared with 77% diagnosed in hospital. This may reflect a higher rate of upper tract infection in this subgroup or the difficulty general practitioners have in establishing the diagnosis of UTI in infancy.9 The incidence of UTI is highest and the risk of renal damage is greatest in the first year of life.4 ,16 The recommendations in the guidelines that infants should have ultrasound, DMSA, and MCUG were generally accepted by members of the working group. In our study renal scarring was shown in 14% of infants, and 17% of those infants treated in hospital. Although no scars were shown in infants diagnosed at home, the numbers are too small for statistical significance or comment. There is a possibility that there may have been some false positive diagnoses of UTI.17

In children aged 1–6 years, only one child (2%) of the 54 treated at home showed any sign of renal scarring compared with 6/18 (33%) of those admitted to hospital. This difference is both statistically and clinically significant. The time from UTI to DMSA scan was slightly shorter in children managed in hospital. This difference could bias the results in favour of scars in the group managed in hospital but would not affect the conclusions of this study that children over 12 months insufficiently ill to require admission do not benefit from DMSA scans.

DMSA scans are invasive tests that involve an intravenous injection, several hours in hospital, radiation to the child and the gonads, and in some cases, sedation or restraint. There is a risk of psychological trauma to the child and the investigation is also stressful to parents.18 Parents may lose time from work and there are financial costs of travelling with young children. In our study the DMSA was abandoned in four children because of technical failure and in two cases the parents did not wish their children to undergo the investigation. In 12 cases it was not requested by the medical staff, perhaps because they did not feel that it was indicated. The test costs around £120, with additional costs in smaller children who are often admitted as day cases for sedation. Stark has questioned the economic value of routine imaging of the urinary tract in children19 and Chambers has questioned the process from the child's perspective.20 On the basis of £120 per DMSA scan, the cost of detecting one scar in this series was £1100 pounds. If selected at risk groups had been targeted and fewer scans had been carried out, the cost could have been reduced. The cost of detecting each scar in infants admitted to hospital with UTI was £720; in children over 7 years with clinical evidence of acute pyelonephritis in this series, the cost was £360 per scar. In contrast, the cost of detecting one scar in children managed as outpatients was £7680. Perhaps this money would be better spent on resources in primary care to improve the diagnosis and treatment of UTI in infants and toddlers.9

This small study indicates that there are reasonable grounds to question the benefit of carrying out a DMSA scan on every child under 7 years with a proven UTI based on risks and benefits to the patient and on the cost to the NHS. The clinician should be free to recommend a DMSA scan in a child with a history suggestive of acute pyelonephritis, or in cases where there is a history of delay in treatment or frequent recurrences of UTI. The conclusion from this study is that routine use of DMSA scans in children over 1 year with a straightforward simple infection is unhelpful and unnecessary.