To establish the spontaneous miscarriage rate and compare it with the procedure related miscarriage rate for amniocentesis and chorionic villus sampling (CVS) by experienced operators.
Retrospective audit over a two-year period of all patients having a consultation for prenatal diagnosis before 12 weeks gestation.
A specialised obstetric and gynaecological ultrasound practice.
A total of 2366 patients, mostly over 35 years of age.
Between 1 July 1995 and 30 June 1997, all patients having a prenatal consultation decided between amniocentesis, CVS or no invasive testing. The CVS was performed either transabdominally or transcervically, depending on the position of the uterus and placenta.
Main outcome measures
Delivery, termination of pregnancy for chromosomal abnormality or miscarriage.
Over the two-year period, 2366 patients had a prenatal consultation and outcome data were available for all but 53 patients. After consultation, 346 patients decided not to have any prenatal testing and 29 (8.4%) of these had a spontaneous miscarriage. Amniocentesis was requested by 839 patients; however 10 miscarried before the scheduled procedure. After the amniocentesis, there were 13 terminations for chromosomal abnormality and three miscarriages. CVS was requested by 1128 patients; however, 23 miscarried before the scheduled procedure. Transabdominal CVS was performed in 665 patients, transcervical in 416 and in 24 cases the documentation of the method used was unclear. Eleven patients miscarried after the transabdominal CVS (1.65%) compared with nine patients miscarrying after the transcervical CVS (2.16%), which was not statistically significant (p = 0.27).
In the group studied, the spontaneous miscarriage rate from nine weeks gestation is very high (8.4%). The procedure-related loss rate from amniocentesis was less than 1 in 280. Transabdominal CVS appears to have a lower fetal loss rate than transcervical CVS, but much larger numbers are needed to prove this.