Diagnosing Urinary Tract Infections in children. Value of signs, symptoms, and point-of-care tests in primary care.

Introduction Urinary tract Infections (UTIs) occur in 1 of 20 children presenting to general practice. Diagnosis is challenging because the signs and symptoms of UTI in children are non-specific. Ruling out UTI requires urine culture, however these results are only available after 2 to 3 days. Obtaining a urine sample is challenging, especially in very young children, and the available non-invasive techniques result in high rates of contamination, leading to inadequate diagnoses. There is insufficient evidence on the diagnosis of paediatric UTIs in Flemish general practices. Previous research has shown that the diagnosis of UTIs in children in primary care might be inaccurate, resulting in a large number of missed infections. Therefore, in this PhD thesis, we investigated how early and accurate diagnosis of UTIs in children might be improved in general practice. 

Methods In part A, we conducted semi-structured interviews to explore the opinions and perspectives of general practitioners concerning diagnosis and management of UTIs in children. In Part B, we examined the incidence rates and trends of UTIs in Flemish general practices from 2000 to 2020; using a registry-based study (Intego database). In part C, we assessed the diagnostic accuracy of clinical features and clinical prediction rules for UTIs in children in ambulatory care using a systematic review and meta-analysis. In part D, we externally validated existing clinical prediction rules for UTIs in children, using a prospective diagnostic accuracy study in 39 Flemish general practices and two paediatric emergency departments (ERNIE4 study). In part E, we assessed the diagnostic test accuracy of urine collection methods in ambulatory care, using a systematic review. In part F, we assessed the diagnostic test accuracy of biomarkers and point-of-care tests for UTIs in children in ambulatory care, using a systematic review and meta-analysis. Finally, in part G, we calculated the diagnostic test accuracy of three novel point-of-care tests for UTIs, compared to urine culture as reference standard (ERNIE4 study).

Results According to GPs, barriers for early and accurate diagnosis of childhood UTIs are the assumption of low UTI prevalence in general practice, the nonspecific presentation, difficulties in urine collection, and diagnostic uncertainty concerning testing (Part A). General practitioners indicated that they sample urine in children that have 1) clinical features of UTI, or 2) fever without source (Part A). There was a statistically significant increase in the number of UTI diagnoses in Flemish general practices from 2000 to 2020 in each age group (p<0.05), except in boys 2-4 years. However, the overall change in number of UTI diagnoses from 2000 to 2020 was low (Part B). No single clinical feature or biomarker was sufficiently sensitive to rule out UTI (Part C, F) and only a few features might lower the probability of UTI [circumcision in boys, wheezing, stridor and nappy rash] (Part C). Fever and duration of fever had limited diagnostic value for UTI. Alarming features for UTI were changes in urine appearance such as cloudy urine, malodorous urine, and haematuria; or no fluid intake, suprapubic tenderness and loin tenderness (Part C). The diagnostic accuracy of the urine sampling strategy following current routine care in primary care was not sufficiently sensitive to detect all UTIs (Part D). Existing clinical prediction rules for UTIs [DUTY score, UTIcalc, Gorelick score] (Part F) were either very specific or very sensitive in our setting, potentially leading to many missed diagnoses or high urine sampling rates (72-92%) (Part D). To obtain urine in young children, the clean-catch method might be sufficiently accurate for ruling out UTI compared to catheterization (Part E). Three novel point-of-care-tests [Utriplex, Rapidbac and Uriscreen test] and the dipstick test were not sensitive enough to replace urine culture for ruling out paediatric UTI in general practice (Part G). 

Conclusion This PhD thesis added to existing literature to improve the diagnosis of UTIs in children in general practice. We described the incidence rates, prevalence, diagnostic approach and the diagnostic test accuracy of clinical features, urine collection methods, biomarkers, point-of-care tests and clinical prediction rules for paediatric UTIs in general practice.