The aim of this study was to compare transplacental with non-transplac
ental amniocentesis in terms of related complications. Between January
1991 and December 1992, 4564 genetic amniocenteses were performed in
4527 patients (4491 singleton, 35 twin, and one triplet pregnancy) at
15-16 weeks of gestation. All the procedures were ultrasound-guided an
d performed by the same operator. In 1487 cases, an anterior placenta
was traversed with the needle, whereas in 3077 cases, the needle was i
nserted directly into the amniotic cavity without traversing the place
nta. After the exclusion of patients in whom amniotic cell culture fai
led or in whom an abnormal karyotype was obtained, and of patients los
t to follow-up, a total of 4454 patients (98 per cent) were followed f
or 30 days after amniocentesis. Two spontaneous abortions occurred aft
er a transplacental procedure, and five after a nontransplacental proc
edure (P=not significant). There were no episodes of amniotic fluid le
akage in the first group, whereas 16 ruptures of the membranes that re
solved spontaneously occurred in the second group (P<0.01). Our data s
uggest that transplacental amniocentesis carries a similar abortive ri
sk to and a lower risk of transient rupture of the membranes than non-
transplacental amniocentesis and may therefore be preferred at the ges
tational period examined (15-16 weeks). However, the risk of feto-mate
rnal haemorrhage, which is reported to be higher for a transplacental
procedure, must be considered in the case of an anterior placenta.