Treatment of acute pyelonephritis in children

A 9 month old girl presents with high fever, vomiting, lethargy, and bacteriologically confirmed urinary tract infection. The diagnosis is clear—acute pyelonephritis. This is a common problem and the cause of about 5% of febrile episodes in children.¹ But how should she be treated? Which antibiotics should be given and by which route? For how long should antibiotics be given? This article summarises what we know about treatment of acute pyelonephritis from randomised trials and what we think we know about treatment, based on clinical experience.

Acute pyelonephritis comprises urinary tract infection with systemic features including fever, vomiting, abdominal or loin pain, and lethargy. Fever is the most useful symptom clinically. Compared with the reference standard for pyelonephritis—technetium-99m dimercaptosuccinic acid scanning—fever is very sensitive but has only moderate specificity. In few afebrile children—except very young infants—the renal parenchyma is affected. Conversely in about 50% of children with clinical pyelonephritis the renal parenchymal is affected.

The major decisions about treatment that are to be made concern the use of antibiotics. Infants aged 1 month or less with urinary tract infection require intravenous antibiotics because of the high prevalence of concomitant bacteraemia (about 10%) and of uropathology, including posterior urethral valves, obstructed duplex systems and high grade vesicoureteric reflux with serious metabolic disturbance such as hyperkalaemia and hyponatraemia. Also young infants have been systematically excluded from randomised controlled trials, which makes the evidence for treatment very weak. Since Escherichia coli and Enterococcus faecalis are the most common pathogens in this age group, empiric treatment with a β lactam antibiotic and an aminoglycoside is indicated. The choice of specific antibiotics should be based on data about local uropathogens. Clinical experience indicates that intravenous treatment should be continued until systemic signs resolve and then oral antibiotics, chosen to match the in vitro sensitivities of the isolated uropathogen, should be given for seven to 10 days.

What about children aged over 1 month with acute pyelonephritis? Here the evidence available to guide decision making is based on 18 randomised controlled trials, of which 16 are summarised in a recent Cochrane review.²

Should antibiotics be given intravenously or orally? Two trials including 306 and 387 children compared oral (cefixime,³ amoxicillin-clavulinic acid⁴) with intravenous (ceftriaxone) treatment for three days or defervescence followed by cefixime or amoxicillin-clavulinic acid. Total duration was 10 or 14 days. No differences in the time to defervescence, recurrence of urinary tract infection, or frequency of renal parenchymal abnormality at 6-12 months were evident between the two groups. Failure of treatment was very low in the orally treated group, and treatment costs were about one half of those in the intravenous group.³ Four trials have compared different intravenous antibiotics given for 7-14 days, with 3-4 days of intravenous and 4-14 days of oral therapy.² No differences in recurrence of urinary tract infection and renal parenchymal abnormality were found. So what do we conclude? Oral antibiotics, carefully chosen to cover local uropathogens, are as safe and effective as intravenous antibiotics in children with a clinical diagnosis of acute pyelonephritis. This is not surprising given the combination of high bioavailability and renal excretion of orally administered antibiotics. Intravenous treatment should be preserved for children who are seriously ill, or who fail oral treatment because of persistent vomiting. There is no evidence to support the practice of giving a single dose of parenteral antibiotics in addition to a standard course of orally administered antibiotics.⁵

Which antibiotic should be given? The five trials comparing different antibiotics are largely uninformative for routine clinical care because the antibiotics evaluated have limited availability and are not routinely used.² Choice of first line oral antibiotics will vary with local antibiotic resistance patterns, but trimethoprim alone or in combination with sulphamethoxazole, cephalexin or amoxicillin-clavulanic acid are standard first line agents. Given that E coli is the causative organism in 90% of cases and that β-lactamase production is present in at least 50%, amoxicillin alone should not be used.

If intravenous antibiotics are required, aminoglycosides or third generation cephalosporins are often given. Aminoglycosides are favoured because of their pharmacokinetic properties, efficacy, widespread availability, and low cost. Three trials, which compared a single daily dose with three times daily doses of aminoglycosides, have shown no differences in persisting bacteriuria, time to resolution of fever, recurrence of urinary tract infections, hearing impairment, or renal dysfunction.^6-8 These results are similar to data from studies in adults, where toxicity tends to favour single dose treatment.⁹ Given the equivalence of the dosing regimens, the ease of administration, and reduced nursing time, once daily dosing seems preferable in general.

How long should antibiotics be given for? In children with urinary tract infection other than pyelonephritis, there is evidence that short course treatment (3-4 days) is as effective as standard course (7-10 days) treatment.¹⁰ However, none of the three trials examining duration of treatment in children with pyelonephritis have compared these clinically relevant alternatives.² Since children with acute pyelonephritis typically take 3-4 days to recover clinically, it seems prudent to continue antibiotic treatment for 7-10 days until further trials examining treatment duration are performed.