Operator experience and sample quality in genetic amniocentesis

Objective: We sought to relate the frequency of maternal cell contamination in amniotic fluid samples that were submitted to a single laboratory for cytogenetic analysis to the experience and training of the physician who performed the amniocentesis. Study design: We reviewed the database of a single cytogenetics laboratory to compare the number of amniocenteses that were performed annually per physician to the rate of maternal cell contamination in genetic amniocentesis samples. Only samples that resulted in a 46 XY male karyotype were studied so that maternal cell contamination could be identified as having occurred when the karyotype revealed ≥1 cell with 2 X chromosomes. Samples were categorized as being submitted by a physician who submitted ≥50 or more samples annually versus <50 samples to this laboratory. The frequency of maternal cell contamination was compared with annual operator volume with 2 × 2 tables and analyzed by chi-squared testing. Results: Between 2000 and 2004, the laboratory received 6332 mid-trimester amniotic fluid samples that generated a male karyotype result. Fourteen of 2081 samples (0.67%) that were submitted by physicians who submitted <50 samples grew ≥1 46 XX cells, compared with 8 of 4251 samples (0.19%; chi-squared, 9.47; degrees of freedom, 1; P = .0021). Conclusion: Maternal cell contamination occurs more frequently in genetic amniocentesis samples that are obtained by physicians who perform <50 genetic amniocenteses annually.