Predictive accuracy of prenatal ultrasound findings for lower urinary tract obstruction: A systematic review and Bayesian meta‐analysis. (5th August 2021)

• Record Type: Journal Article
• Authors: Keefe, Daniel T.; Kim, Jin Kyu; Mackay, Eric; Chua, Michael; Van Mieghem, Tim; Yadav, Priyank; Lolas, Marisol; Santos, Joana Dos; Skreta, Marta; Erdman, Lauren; Weaver, John; Fermin, Antoine Selman; Tasian, Gregory; Lorenzo, Armando J.; Rickard, Mandy
• Is Part Of: Prenatal diagnosis. Volume 41:Number 9(2021)
• Page Start: 1039
• Page End: 1048
• Publication Date: 2021-08-05
• Subjects: Prenatal diagnosis -- Periodicals; Fetus -- Diseases -- Diagnosis -- Periodicals; Electronic journals; 618.32075
• DOI: 10.1002/pd.6025
• Languages: English
• ISSNs: 0197-3851

Abstract

Abstract: Background: Lower urinary tract obstruction (LUTO) is a rare but critical fetal diagnosis. Different ultrasound markers have been reported with varying sensitivity and specificity. Aims: The objective of this systematic review and meta‐analysis was to identify the diagnostic accuracy of ultrasound markers for LUTO. Materials and Methods: We performed a systematic literature review of studies reporting on fetuses with hydronephrosis or a prenatally suspected and/or postnatally confirmed diagnosis of LUTO. Bayesian bivariate random effects meta‐analytic models were fitted, and we calculated posterior means and 95% credible intervals for the pooled diagnostic odds ratio (DOR). Results: A total of 36, 189 studies were identified; 636 studies were available for full text review and a total of 42 studies were included in the Bayesian meta‐analysis. Among the ultrasound signs assessed, megacystis (DOR 49.15, [15.28, 177.44]), bilateral hydroureteronephrosis (DOR 41.33, [13.36, 164.83]), bladder thickening (DOR 13.73, [1.23, 115.20]), bilateral hydronephrosis (DOR 8.36 [3.17, 21.91]), male sex (DOR 8.08 [3.05, 22.82]), oligo‐ or anhydramnios (DOR 7.75 [4.23, 14.46]), and urinoma (DOR 7.47 [1.14, 33.18]) were found to be predictive of LUTO (Table 1). The predictive sensitivities and specificities however are low and wide study heterogeneity existed. Discussion: Classically, LUTO is suspected in the presence of prenatally detected megacystis with a dilated posterior urethra (i.e., the keyhole sign), and bilateral hydroureteronephrosis. However, keyhole sign has been found to have modest diagnostic performance in predicting the presence of LUTO in the literature which we confirmed in our analysis. The surprisingly low specificity may be influenced by several factors, including the degree of obstruction, and the diligence of the sonographer at searching for and documenting it during the scan. As a result, providers should consider this when establishing the differential for a fetus with hydronephrosis as the presence or absence of keyhole sign does not reliably rule in or rule out LUTO. Conclusions: Megacystis, bilateral hydroureteronephrosis and bladder wall thickening are the most accurate predictors of LUTO. Given the significant consequences of a missed LUTO diagnosis, clinicians providing counselling for prenatal hydronephrosis should maintain a low threshold for considering LUTO as part of the differential diagnosis. Key points: What's already known about this topic? Prenatal ultrasound can identify features concerning for lower urinary tract obstruction (LUTO) Classic findings include 'keyhole sign' with dilated posterior urethra and megacystis Studies suggest these findings have low predictive accuracy for diagnosing LUTO What does this study add? Largest systematic review and meta‐analysis assessing prenatal ultrasound findings predictive of LUTO Keyhole sign has low sensitivity for LUTO. Megacystis, bilateral hydroureteronephrosis, and bladder wall thickening are better predictors of LUTO We suggest clinicians lower their threshold for suspecting and working up all fetuses with these findings to prevent missing a prenatal diagnosis