Abstract

A retrospective review of the casenotes of 403 children admitted to hospital with febrile convulsions was performed to estimate the frequency of symptomatic urinary tract infection and examine medical practice in making this diagnosis. A total of 228 (56%) children had urine cultured: 150 bag specimens, 76 clean voided samples, and two suprapubic aspirates. There were 13 'probable' and six 'possible' infected urine samples together representing 5% of the whole study population (n = 403), 8% of those having urine cultured (n = 228), and 12% of those providing uncontaminated urine samples (n = 155). Those with first febrile convulsions and those aged under 18 months were more likely to have urine examined. Practices varied significantly between different hospitals. These results suggest that there has indeed been a need for practice guidelines, and that further audit of practice is required to assess their impact.