Agreement between two uroflowmetry tests in children with lower urinary tract symptoms

Abstract

Uroflowmetry is a non-invasive examination considered as a first-line assessment for children with lower urinary tract symptoms (LUTS). Currently, the performance of two uroflowmetry tests is recommended by ICCS for all patients with LUTS. This study aims to evaluate the differences between two uroflowmetry tests in children with lower urinary tract symptoms and their impact on the patient's diagnostic workup and clinical outcome. Forty patients with LUTS, aged 4–17 years, were evaluated prospectively with two consecutive uroflowmetry tests and ultrasonography. They were classified based on the ICCS criteria for curve pattern and divided into two groups based on the SPU classification of fractionated (staccato and intermittent) and smooth (bell, tower and plateau). They were also separated into three groups based on estimated bladder capacity (voided volume + post-void residual on ultrasound): high (>115%), low (<60%) or normal capacity. The mean age was 8 (IQR 6.0–10.0) years and 25 (62.5%) patients were female. There was an increase in nonbell-shaped curves from the first (32.5%) to the second test (52.5%). The curve shape between the uroflowmetries based on ICCS classification showed a Kappa value of 0.349 (fair). Classifying curves as fractionated or smooth yielded a Kappa value of 0.714 (substantial) (table). The Bland–Altman test showed disagreement in the parameter of time to Qmax. There was a significant difference in categorizing bladder capacity as high, low, or normal: it was concordant in 63.6% for high, 68.4% for normal and 50% for low capacity (p = 0.001). Despite the disagreement found classifying the curves based on the ICCS pattern, also demonstrated in other studies, there was a substantial agreement using the SPU criteria. It reinforces the greater reliability of the SPU system and it may be the key to reduce the subjectivity of uroflowmetry. Even though this classification being associated with a higher agreement in interpretation of the curves, repeating uroflowmetry does not present a clinically significant divergence that changes the patient's diagnostic workup. Our study is limited by the lack of EMG and larger sample. Considering the flow curve pattern, the agreement between two uroflowmetries was substantial according to the SPU and only reasonable by the ICCS classification. Regardless of some differences found between the flows, a second uroflowmetry test might not have clinical relevance that justifies its recommendation for all patients.