Objective: To investigate births with neural tube defects at a time when most districts were screening for the condition. The objective was to document the circumstances surrounding each affected birth and assess the care provided against given standards.
Design: Retrospective review of antenatal casenotes by the obstetric team.
Population: Three hundred and eight births in England and Wales in 1990 to 1991 were reported to the Office of Population Census Survey (OPCS) to involve neural tube defects. Sufficient information was available to identify both the woman and the obstetrician in 213 pregnancies. Details were obtained from a questionnaire completed by the obstetric team for 168 (79%). In 20/168 cases either the reported outcome was not a live birth/stillbirth or the pregnancy did not involve a neural tube defect: eight resulted in a 'normal' infant, eight were terminated and in four the abnormality was not a neural tube defect. Thus 148 eligible cases were available for analysis.
Results: Of the 148 births, the anomaly was not detected prenatally or detected later than 25 weeks of gestation in 98 cases (66%), diagnosed in a multiple pregnancy in 24 (16%) and diagnosed prenatally but the woman chose to continue the pregnancy in 26 (18%). Of the 98 births not detected prenatally or detected late during pregnancy, the surrounding circumstances were that screening was declined in six cases (4%), screening was not offered due to late booking in 30 (20%), serum alpha-fetoprotein screening gave a false negative result in eight (5%), ultrasound screening gave a false negative result in 29 (20%), both screening methods gave false negative results in 17 (11%) and other reasons in eight (5%). The estimated sensitivity of ultrasound screening for anencephaly was 100%. For spina bifida the estimated sensitivity for singleton pregnancies is higher for serum alpha-fetoprotein screening, 84% to 92%, than ultrasound screening, 70% to 84%, for a range of assumptions regarding the degree of under-reporting to OPCS of live births and terminations.
Conclusions: Late booking precluded the offer of screening tests in a substantial proportion (22%) of cases. The presence of multiple fetuses including one or more with a neural tube defect was a serious additional complication in prenatal screening, diagnosis and counselling. Screening for neural tube defects was widespread in 1990 to 1991, although variations in the services provided were documented. Ultrasound scanning was a major component but was associated with a lower sensitivity than maternal serum alpha-fetoprotein screening for neural tube defects other than anencephaly.