OC16.03: *Improving early detection of facial clefts using first trimester sonographic markers: diagnostic accuracy based on 389,000 examinations

To determine the diagnostic accuracy of first trimester ultrasound for the detection of facial clefts (FC); and understand the role of sonographic markers in improving screening performance, including visualization of the retronasal triangle (RNT) and the super-imposed palatine sign (SPS). We evaluated the diagnostic accuracy of screening for FCs (cleft lip and palate, cleft lip or palate alone, and composite) using data that was extracted from all studies assessing FC detection at 11-14 weeks gestation identified from an electronic search (Jan 2008 – Aug 2022). Meta-analysis (random-effects model) was undertaken and secondary analyses based on imaging protocols used by individual studies. An evaluation of markers for FC was undertaken using pair-wise comparisons (chi-squared tests) of detection rates between studies. There were 389,369 fetuses, of which 392 were affected by FC, (prevalence: 10.9/1000 (95% CI 7.7 – 14.7, 41 studies). Overall detection rate (DR) was 43.4% (28.8 – 58.7), with specificity of 100.0% (100.0-100.0) and positive predictive value of 84.2% (73.5 – 92.6). Of 154 screen positive cases,147 were true positive and 7 were false positive (6 cleft palate only; 1 cleft lip only). Significant differences in DR from six protocol subgroups were demonstrated (p < 0.0001, see table 1). In studies examining facial profile, pair-wise comparison showed that additional coronal RNT views confer diagnostic benefit over that offered by an axial maxillary view (p < 0.0001). Accurate detection of FCs in the first trimester is feasible. and can be optimised by evaluating multiple views of the fetal face (axial, coronal and sagittal) with support of sonographic markers. Implementation of such protocols require high-level sonographer skill, fetal anatomy knowledge and additional screening time. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.