Midtrimester amniocentesis is safer than we thought

Abstract

Research question What is the risk of pregnancy loss after a midtrimester amniocentesis?

Answer An estimated 0.06% (95% CI −0.26% to 0.49%)

Why did the authors do the study? Women are traditionally advised that the risk of pregnancy loss after amniocentesis is around 0.5%, based on research conducted in the 1970s. These authors wanted to update this figure to take account of modern practices such as ultrasound guidance.

What did they do? They analysed data from 35 003 women with singleton pregnancies who had taken part in a trial of two different types of serum screening for Down's syndrome (first trimester v second trimester screening). The trial finished recruiting in 2002. Women who screened positive in either arm were offered amniocentesis in the middle of the second trimester. Women who screened negative could also request an amniocentesis. All women were followed up until the end of their pregnancy.

These authors compared the 3096 women who had amniocentesis with the 31 907 women who did not to see if there was any difference in the percentages who had unintended pregnancy loss before the 24th week of gestation. They did further analyses to estimate the odds of pregnancy loss in these two groups after accounting for a wide range of confounding variables, including age, body mass index, history of diabetes, pregnancy history, and serum screening results.

What did they find? Of the women who had an amniocentesis, 1.0% lost their babies before 24 weeks, compared with 0.94% of the women who did not have amniocentesis. So an estimated 0.06% of losses were directly due to amniocentesis. The 95% confidence intervals around this estimate went from −0.26% to 0.49%, indicating that there was no statistically significant difference in the rate of pregnancy loss between the two groups. After adjusting for confounding factors, including serum screening results, the overall odds ratio for pregnancy loss after a midtrimester amniocentesis was 0.4 (0.3 to 0.7). In subgroup analyses, amniocentesis was associated with a lower rate of unintended pregnancy loss for women aged over 35 (0.4 (0.2 to 0.6)), and for women with positive serum screening results (0.3 (0.2 to 0.4)).

What does it mean? The risk of losing a baby after an amniocentesis is lower than previously thought and may be no different from the background risk of spontaneous miscarriage. The odds of unintended pregnancy loss were actually lower after amniocentesis in this study, presumably because some women whose babies had chromosomal abnormalities detected by amniocentesis terminated their pregnancies deliberately instead. Results from subgroup analyses support this view. There were certainly more elective terminations among women having amniocentesis (91/3096, 2.9%) than among controls (67/31 907, 0.2%), but the reasons were not recorded.

This study included a large number of unselected women who had their amniocentesis in clinics and hospitals across the US, so the findings should be generalisable to modern practice in other developed countries. Women can be reassured that midtrimester amniocentesis is reasonably safe, and the risk of losing a baby is nearer 1 in 1600 than the more traditionally cited 1 in 200.