P02.09: The performance of universal Down's syndrome screening using self-derived prediction model specific for Chinese in Hong Kong

To assess the performance of universal Down's syndrome screening using self-derived prediction model specific for Chinese in Hong Kong. This is a population-based prospective study in Hong Kong where the Government launched the universal Down's syndrome screening since July 2010. Four public obstetrics units offered Down's syndrome screening, either as first trimester combined test using nuchal translucency (NT), free beta human chorionic gonadotropin (fb-HCG) and pregnancy-associated plasma protein-A (PAPPA) or second trimester biochemical screening using fb-HCG and alpha-fetoprotein (AFP) when women came late. All blood samples were sent to a central university-based laboratory for maternal serum processing and risk determination. Those with a Trisomy 21 term risk of 1:250 or higher were offered a diagnostic test. 54,417 pregnancies were screened of which 43598 (80.1%) had a first-trimester combined test, 403 (0.74%) had a NT test only and 10416 (19.1%) had a second-trimester double test. There were 113 (0.21%) pregnancies affected by Trisomy 21 (T21). The detection rate of the first-trimester combined tests was 92.5% (87/94) while that of the second-trimester test was 78.6% (11/14), at a screen-positive rate of 5%. Among which, five T21 cases were also picked up by ultrasound because of obvious sonographic abnormalities. Among the 10 screened negative T21 cases, seven resulted in live births while the other three were subsequently detected by morphological scan and terminated. The universal Down's syndrome screening program using our self-derived Chinese-specific prediction model was effective and achieved the expected detection rates with low false-positive rates. To maintain this performance, the current emphasis on training, quality control and regular auditing must continue.