Amniocentesis or chorionic villus sampling in multiple gestations? Experience with 500 cases

500 women with multiple pregnancies underwent amniocentesis or chorionic villus (CV) sampling at our department between January 1988 and July 1997. The aim of this retrospective study was to evaluate the laboratory aspects and the consequences of discordant results in these pregnancies in relation to the method of sampling. Uncertain results in one or both samples, requiring further investigation were more frequent in CV samples (eight times in 163 paired samples, 5 per cent) than in amniotic fluid (AF) samples (once in 298 paired samples, 0.3 per cent). Sampling one fetus twice (erroneous sampling) was seen only once among 163 pregnancies with two CV samples in our study. Cross contamination due to mixed sampling was discovered in two of seven pregnancies that underwent DNA diagnosis in CV and might be a rather regular occuring phenomenon. In none of the 500 pregnancies mixed sampling caused diagnostic dilemmas. A third sampling problem, maternal cell contamination caused a diagnostic problem once among the AF samples. Selective fetal reduction appeared safer after CV sampling than after amniocentesis. Subsequently, CV sampling instead of amniocentesis has become the method of choice for prenatal diagnosis in multiple pregnancies in our department.