Current assessment of fetal losses as a direct consequence of chorionic villus sampling.

Chorionic villus sampling has been developed as a method of first trimester prenatal diagnosis. In order to evaluate this new approach, accurate risk figures for the procedure must be obtained. This has been difficult for a number of reasons, including establishment of baseline fetal loss rates in the first trimester, procedural "learning curves," and reporting biases. This review will discuss these problems and use data from the chorionic villus sampling program at the University of California, San Francisco, to illustrate difficulties in data interpretation.