A total of 1,068 patients were examined by ultrasound to ensure normality of pregnancy and followed prospectively from booking until 28 weeks. The spontaneous miscarriage rate was 2.7 per cent occurring within the first 16 weeks. Threatened miscarriage was associated with a 38 per cent fetal loss. Miscarriage was less likely as pregnancy advanced. The reduction in subsequent miscarriage rate before 11 weeks and from 11 weeks onwards is statistically significant (p less than 0.001). Gravidity, maternal age and a history of previous fetal loss did not contribute significantly to the miscarriage rate. Patients with a history of fetal loss were more likely to experience a threatened miscarriage. The relevance of these findings to chorion villus sampling is discussed.

Detailed ultrasound examination of the placentae from 293 consecutive women requesting first-trimester chorionic villus sampling (CVS) showed evidence of intraplacental sonolucent spaces with varying density in 42.3 per cent of these placentae. Their presence, however, did not complicate the subsequent course of these women's pregnancies. Their prime significance relates to CVS, where inadvertent entry into these areas can lead to bleeding and contamination of the villus specimens with blood. A search for these spaces should be made before sampling, and when present, they should be avoided wherever possible.

Aims to determine the extent to which women suitable for community-based antenatal and intrapartum care will require hospital contact. Reports on an historical cohort study of low risk women who underwent standard shared care and for whom the records for both pregnancy and delivery were complete at The City Hospital and University Hospital, Nottingham. Concludes that the shift to community-based care aims to bring many improvements to the overall care and satisfaction of pregnant women. However, it may not reduce the workload of hospital-based services to a great extent as the majority of women, even if low risk at booking, will require some hospital input at some time during pregnancy or labour.

A brush delivered to the biopsy site by a metal introducer and cannula set was tested as an alternative implement for transcervical collection of chorionic villi. This implement was easy to use and readily identified by ultrasound. With only one attempt at sampling, the overall collection rate for 83 patients was 65 per cent. With practice 84 per cent successful collection was achieved. Gestations between 8 and 11 weeks was the best time for collection of chorionic villi. A single sampling can produce adequate material (15 mg wet weight) suitable for diagnostic purposes. No gestation sac was perforated but some slight bleeding followed the procedure. Appreciable success following a single sampling attempt coupled with the low complication rate suggests that this technique may have clinical application and deserves further investigation.