Intrapartum ultrasound estimation of total bladder volume

Abstract

Objective: To determine whether real-time ultrasound measurement of multiple linear dimensions of the bladder could be used to estimate total bladder volumes of women in labour. Aim: To create a simple reproducible formula using ultrasound techniques to help estimate total bladder volumes of women in labour. This may prevent unnecessary catheterisation and conversely, avoid failure to catheterise when indicated, thereby reducing the complications associated with catheterisation and bladder overdistension. Design: Prospective, observational study of women in labour. Setting: Delivery Suite, Obstetric Hospital, University College London Hospitals. Population: Forty-nine women in labour recommended for catheterisation by their attending midwives. Method: A real-time ultrasound scanner was used to measure bladder diameters in two planes: transverse width and height, and sagittal width and height. A Foley's catheter was inserted and the bladder drained into a graduated bowl under sterile conditions. These values and urine volumes were recorded and a formula for detecting bladder volumes above 300 mL was developed using linear regression methods and the Bland-Altman limits of agreement. Main outcome measure: Sensitivity and specificity of derived formulae in predicting bladder volume from ultrasound measurements. Correctly predicted volumes greater than 300 mL. Results: Among the 49 women, 36 were nulliparous and 13 were multiparous. All had full-term pregnancies with cephalic presentation and were in active labour with cervical dilatations ranging from 3 to 10 cm. There were no previous caesarean sections among the multiparous women. Thirty-nine women had epidural analgesia. Indications for catheterisation included a palpable bladder per abdomen, or the interval from last urine voided. The catheterised volumes ranged from 30 to 720 mL with a median of 150 mL. The new equation (0. 68 x h x d x w) had a sensitivity of 77% (95% CI 46-95%) and specificity of 86% (95% CI 71-95%) for detecting bladder volumes greater than 300 mL. In order to ensure that about 95% of women in need of catheterisation are detected, the above equation was improved using the Bland-Altman limits of agreement. The sensitivity of this final equation (1.74 x h x d x w) was 100% (95% CI 75-100%) and the specificity dropped to 36% (95% CI 21-54%). Conclusion: It is possible to use real-time ultrasound in estimation of total bladder volumes of women in labour. The implications are that all of those normally selected as being in need of catheterisation could be detected, while about a quarter of those women would avoid unnecessary catheterisation.