Comparison of transcervical and transabdominal chorionic villus sampling loss rates in nine thousand cases from a single center.

Objectives: Our purposes were (1) to compare the safety of transabdominal and transcervical chorionic villus sampling with the use of a consistent technique at one center and (2) to determine whether the training of fellows can be accomplished without an increase in the loss rate.
Study Design: We performed a retrospective comparison of transabdominal and transcervical chorionic villus sampling loss rates from procedures performed by three principal operators between 1984 and 1992. The type of procedure was chosen by the operator at the time of the procedure on the basis of placental location.
Results: Procedures 1 through 2573 were performed solely by transcervical chorionic villus sampling and had an overall fetal loss rate of 5.12%. With the addition of transabdominal chorionic villus sampling the overall fetal loss rate dropped to 3.07% (p < 0.0001). Three and one half years after the start of transabdominal chorionic villus sampling (about 1300 transabdominal chorionic villus sampling procedures), the transabdominal chorionic villus sampling loss rate was significantly better than the transcervical loss rate (p = 0.035), and the difference widened steadily after that. During the same time period seven fellows performed 716 procedures for a fetal loss rate among fellows of 2.72%.
Conclusions: (1) Under optimal circumstances (one center, large numbers, few operators, consistent technique, operator choice of best approach), transabdominal chorionic villus sampling may be inherently safer than transcervical chorionic villus sampling. (2) The addition of transabdominal chorionic villus sampling decreases overall chorionic villus sampling loss rates. (3) Although the number of procedures performed by fellows is small, it appears that with close supervision by experienced operators successful training of fellows can be accomplished without adverse effects on loss rates.