Interrater reliability of bladder ultrasound measurements in children

Abstract

Bladder ultrasound is becoming pivotal in the management and treatment of lower urinary tract dysfunction. There is a paucity of data regarding intra-observer and interobserver reliability of bladder ultrasound, especially in children. A previous study assessed interobserver agreement for both the postvoid residual volume and measurement of the bladder dimensions in adults showing excellent agreement. To examine the interobserver and intra-observer reliability of bladder wall thickness, bladder urinary and postvoid residual urinary volume using ultrasound evaluated by paediatricians having different training levels. Four sonographers, 3 pediatric trainees and one experienced pediatric urologist measured the full bladder volume, the voiding residual volume, and the bladder wall thickness. Each sonographer made 3 measurements of each parameter. We assessed the interobserver and intra-observer variability by using intraclass correlations (ICCs). ICCs were calculated and tested with a significance level of 5%. The interrater ICC was calculated from the mean of the three measurements of each variable (full bladder volume, postvoid residual, bladder wall thickness). ICC ≥0.75 was considered excellent. Bland-Altman plots were also used to assess the interobserver agreement. Sixty children were recruited (7.3±1.1 years). The interobserver ICCs for bladder volume and voiding residual volume were 0.91 (confidence interval 0.85-0.95) for both. The interobserver ICCs for the bladder wall thickness was satisfactory 0.43, with a minimum detectable difference of 2mm. The observed values for intra-observer analysis showed an excellent (ICC≥0.90) agreement between the three measurements performed by each one of the sonographers. Bland-Altman plots showed that the bias (mean difference) was -0.35 and the limits of agreement were -3.43 and 2.73 for the bladder wall thickness measurements (Figure). The interobserver reliability for both the postvoid residual volume and full bladder volume were excellent even in operators with different levels of training, pointing out the role of bladder ultrasound in the routine clinical practice. Weaker agreement was found for the bladder wall thickness measurement. The intra-observer agreement was excellent for all the measurements. These results showed a good reliability of urinary bladder ultrasound in children aged 7.3±1.1 years as far as bladder volume measurement is concerned. Given the variability of bladder wall thickness, a standardized methodology is desirable to increase its reliability.