AUDIT OF A SCREENING SERVICE FOR FETAL ABNORMALITIES USING EARLY ULTRASOUND SCANNING AND MATERNAL SERUM ALPHA-FETOPROTEIN ESTIMATION COMBINED WITH SELECTIVE DETAILED SCANNING

The objective of the study was the evaluation of a screening service f or fetal abnormality using early ultrasound scanning to date and estab lish viability of the pregnancy combined with maternal serum alpha-fet oprotein estimation at 16 weeks and thereafter selective detailed scan ning at 18-20 weeks based on clinical indication. The study was design ed to be retrospective, emphasizing the possibility of diagnosis. Earl y scans were performed by radiographers and radiologists, detailed one s by radiologists and obstetricians with subspeciality training in pre natal diagnosis. The study was carried out in the ultrasound departmen t of a large University Hospital. The subjects studied were all cases of abnormality identified pre- or postnatally in women delivering in t he hospital over 4 years and to whom the screening service applied. In formation was obtained from ultrasound and maternity records, birth no tifications, pathology and autopsy reports and a special needs registe r. Cases were classified as not detectable, potentially detectable and usually detectable. Major fetal anomalies numbered 255 (total deliver ies 19 497), a prevalence of 1.3%; 130 (51%) were diagnosed by ultraso und with 64% before 24 weeks and 36% later. Eleven chromosomal anomali es (4%) were diagnosed by genetic methods and 114 (45%) anomalies were not diagnosed antenatally. The sensitivity of the program was 37% bef ore 24 weeks. Chromosomal anomalies, lesions of the central nervous sy stem and cardiac defects were most common, followed by lesions of the gastrointestinal and urinary tracts. A large proportion of the detecte d and detectable anomalies were lethal. The implementation of a policy of routine scans at 18-20 weeks for fetal abnormality would be justif ied if it led to improved obstetric decision-making, reduction in chil d handicap and less parental suffering. A routine scan applied to our population might not have achieved these except in a comparatively sma ll number of cases.