Population‐based trends in invasive prenatal diagnosis for ultrasound‐based indications: two decades of change from 1994 to 2016

ABSTRACT Objective To assess trends in ultrasound‐indicated prenatal diagnostic testing performed over the past two decades in the Australian state of Victoria, in the context of rapidly changing practices in aneuploidy screening and chromosome analysis. Methods This was a retrospective analysis of all ultrasound‐indicated prenatal diagnostic testing (amniocentesis and chorionic villus sampling) performed in the state of Victoria between 1994 and 2016. Ultrasound indications for testing included: fetal structural abnormality, fetal death, fetal growth restriction, abnormal amniotic fluid volume, genetic ‘soft marker’ and unspecified ultrasound abnormality. Maternal age, indication for testing, type of diagnostic procedure, gestational age, type of chromosome analysis (G‐banded karyotyping or chromosomal microarray (CMA)) and test results were obtained. Diagnostic yield (i.e. percentage of tests yielding a major abnormality) was analyzed by year, maternal age and gestational age. Statistical analysis was performed using the χ 2 tests for trend or difference in proportions, as appropriate. Results During the 23‐year study period, 1 533 317 births were recorded and 16 152 diagnostic procedures were performed for the primary indication of ultrasound abnormality. In recent years, ultrasound abnormality became the most common indication for prenatal invasive testing (29.4% of diagnostic tests between 2013 and 2016) due to a steep decline in testing for other indications such as positive result on combined first‐trimester screening or advanced maternal age alone. In 2016, over 95% of ultrasound‐indicated procedures were performed with CMA; among these, pathogenic copy number variant (CNV) was the most common (3.5%) abnormality detected, followed by trisomy 21 (2.8%). The diagnostic yield of ultrasound‐indicated tests performed < 16 weeks was significantly higher than that of tests performed after 20 weeks (31.5% vs 9.0%). Conclusions Ultrasound‐indicated invasive testing is contributing to prenatal diagnosis in new ways in the genomic era. A pathogenic CNV is now the most likely diagnosis after ultrasound‐indicated testing, rather than trisomy 21 or other whole‐chromosome aneuploidy. Despite steady improvements in first‐trimester screening for aneuploidy, the diagnostic yield of ultrasound‐indicated tests > 20 weeks has remained stable due to increased utilization of CMA. Procedures performed for structural abnormalities < 16 weeks continue to have the highest diagnostic yield, supporting the benefits of early fetal structural assessment at 11–13 weeks. Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.