Statistics from Altmetric.com
Editor,—Deshpande and Verrier Jones have recently concluded that it is not worth undertaking dimercaptosuccinic acid (DMSA) scans in children over 1 year of age who present with a “simple” urinary tract infection (UTI).1 Their argument has three strands. First, they interpret their data as indicating a very low chance of children over 1 year having a renal scar, especially if the UTI is diagnosed by a general practitioner at home rather than in hospital. Second, they argue that subjecting children to DMSA scanning is financially and emotionally very costly. Third, they feel there is little value in identifying renal scars. We disagree with all three points.
WHAT IS THE CHANCE OF FINDING A SCAR IN A CHILD AFTER A “SIMPLE” UTI?
Deshpande and Verrier Jones argue that the prevalence of scars is so low in older children that it is not justified to perform DMSA scans after 1 year. Yet their own data show that the prevalence in their 124 patients is very similar under and over 1 year (7/52 (13%)v 7/72 (10%); p=0.52, χ2). This confirms our study of 2842 children who had DMSA scans after a first recognised UTI.2 The prevalence of scars was similar at every age from infancy to 16 years.
Deshpande and Verrier Jones advise using the site of diagnosis of the UTI as well as age to select which children should have a DMSA scan. They seem to be assuming that children diagnosed in hospital are likely to have had a more severe illness (and greater scarring risk) than children diagnosed at home by their general practitioners. Their argument has two flaws. One is that there are many local factors that may influence where the diagnosis is made, but which do not relate to the severity of illness. These will vary but will include the organisation, quality and ease of availability of primary and secondary health care services as well as geographical and social factors. Clearly, it cannot be assumed that their deceptively simple surrogate marker for illness severity will reliably predict scarring risk outside their own centre. A further problem is that their small numbers (54 diagnosed at home, 18 in hospital) give poor predictive power for this association. We also made a crude assessment of illness severity in our study of scar prevalence in children after a UTI,2 noting if they had fever, anorexia, malaise, or required hospital admission (but not who made the diagnosis). Younger patients were much more likely to have a severe illness by any of these criteria (table 1), yet their prevalence of scarring was no greater. We also investigated whether these illness severity markers distinguished between the 92 children who had scars and 232 of the unscarred children who were scanned on the same day. Though scarred children were symptomatic slightly more often, the differences were small, so these criteria would not provide a clinically useful screening tool, either before and after the fourth birthday (table2).
WHAT IS THE COST OF A DMSA?
Deshpande and Verrier Jones' description of DMSA scans bears almost no relation to our experience of them. We use local anaesthetic cream and distraction techniques routinely during venepuncture and find this combination extremely successful. We have not found it necessary to sedate children, nor do we recognise psychological trauma occurring in the children or their parents. A typical comment from one of our families is how interesting their day had been! The effective radiation dose of a DMSA is up to 0.7 mSv.3 This is equivalent to about 4 months extra background radiation in the UK, or 6 weeks in Sweden.
THE VALUE OF IDENTIFYING RENAL SCARS
Deshpande and Verrier Jones seem concerned that “the emphasis on imaging tests” has overshadowed the importance that needs to be given to “the diagnosis and treatment of infection”. Whilst we agree that there is a need for an emphasis on accurate diagnosis and treatment, especially in the very young, we see no conflict between providing a service that delivers prompt diagnosis and treatment, and applying a systematic imaging protocol.
Though the primary aim of imaging children's urinary tracts after a UTI is to identify risk factors that will allow us to prevent renal scarring, there is undoubtedly value in diagnosing scars that have already occurred. Reflux nephropathy is the commonest cause of hypertension in children.4 Children that have their blood pressure monitored because of known renal scarring can receive early treatment. By contrast, children that present unexpectedly with severe hypertension following unrecognised or uninvestigated UTIs may have a high morbidity, and a significant mortality. Similarly, children identified as having extensive renal scarring can have treatments gradually introduced if their renal function begins to decline, rather than presenting with the complications of severe renal impairment such as rickets, poor growth, tiredness and anaemia, or even sudden death from hyperkalaemia.
In summary, whilst we acknowledge the importance of knowing the cost of every intervention and test, we are concerned that their value must also be fully appreciated. Investigations that inform families about their child's condition, and allow monitoring to direct early treatment and prevent unpleasant or permanent sequelae are inherently valuable. Since a DMSA scan performed after a childhood UTI has a similar chance of identifying scarring at any age, we currently advocate undertaking one in every child after their first recognised UTI.