Long-term outcomes of urinary tract infection (UTI) in childhood (LUCI)study: What is the prevalence of renal scarring following childhood UTI?

Talk Code: 
2C.1
Presenter: 
Kathryn Hughes
Co-authors: 
Rebecca Cannings-John, Hywel Jones, Fiona Lugg-Widger, Mandy Lau, Shantini Paranjothy, Nick Francis, Alastair Hay, Christopher Butler, Judith Van der Voort,Margaret Heginbothom, Kerenza Hood
Author institutions: 
Cardiff University, Oxford University, Bristol University, Public Health Wales

Problem

Guidelines recommend the prompt diagnosis and treatment of UTI in young children to reduce the risk of renal scarring and possible long-term complications. A systematic review in 2010 found that the prevalence of renal scarring following first childhood UTI was 15%. However, the evidence for the association between childhood UTI, renal scarring and long-term complications is weak and these associations have been questioned. There is an urgent need to clarify this as the correct approach to urine sampling and diagnosis of UTI in children hinges on this association. We therefore aimed to determine outcomes following UTI in childhood (<5years) using routinely collected data.

Approach

We used the Secure Anonymised Information Linkage (SAIL) Databank to access demographic, hospital, GP and microbiology data from children in Wales. Children with at least one microbiologically confirmed UTI (mcUTI) aged less than five, were compared to those with no mcUTI. The primary outcome was renal scarring and secondary outcomes included hospital admissions, hypertension, chronic kidney disease, renal failure, GP consultations, antibiotic prescriptions and subsequent UTI. Risk factors and covariates including VUR and renal/urinary system congenital malformations were defined. The primary analysis used multinomial regression and time to event models.

Findings

A cohort of 159,207 children was defined. 88,595 urine samples were received from children aged under five. 11,099 (7.0%) children had at least one mcUTI; 32,485 (20.4%) had at least one urine sample submitted to a microbiology laboratory, but no mcUTI; and 115,617 (72.6%) had no urine samples submitted. The overall prevalence of renal scarring in the cohort by age 7 was 0.15%; 1.32% in those with at least one mcUTI and 0.07% in those with no mcUTI. The unadjusted odds ratio (OR) was 19.23 (15.17-24.39). After adjusting for confounders (including gender, congenital malformations, VUR, comorbidities and surveillance), the OR was 3.96 (2.86-5.47). Among the group with at least one mcUTI, VUR, congenital malformations and comorbidities were highly associated with renal scarring. Having at least one resistant UTI was also associated with renal scarring (unadjusted OR 3.18).

Consequences

UTI was associated with renal scarring even after adjusting for confounding variables. The prevalence of renal scarring in children with at least one mcUTI is much lower than in the 2010 systematic review. Possible explanations include poor coding in routine data, different patient populations or missed renal scarring diagnoses. We are currently validating our results using radiology data. If renal scarring diagnoses are being missed, a more proactive urine sampling and imaging strategy may be indicated. However, if renal scarring rates are low following UTI, this may not be necessary. Further studies with systematic imaging may be necessary to determine the true rate of renal scarring.

Submitted by: 
Kathryn Hughes
Funding acknowledgement: 
Health and Care Research Wales