Induced abortion and placenta complications in the subsequent pregnancy

Background. To study the risk of placenta complications following an induced abortion as a function of the interpregnancy interval. Methods. This study is based on three Danish national registries; the Medical Birth Registry, the Hospital Discharge Registry, and the Induced Abortion Registry. All primigravida women from 1980 to 1982 were identified in these three registries. A total of 15,727 women who terminated the pregnancy with a first trimester induced abortion were selected to the abortion cohort, and 46,026 women who did not terminate the pregnancy with an induced abortion constituted the control cohort. By register linkage all subsequent pregnancies were identified from 1980 to 1994. Only women who had a non-terminated pregnancy following the index pregnancy were selected to the study. Placenta complications were identified using either the Hospital Discharge Registry ICD-8 codes or the Medical Birth Registry records. Results. A slightly higher risk of placenta complications following an abortion was found. Retained placenta occurred more frequently in women with one, two or more previous abortions, compared with women without any previous abortion of similar gravidity. Adjusting for maternal age and residence at time of pregnancy, the interpregnancy interval, and the number of previous miscarriages (control cohort only), the odds ratios of retained placenta in deliveries of singleton live births in women with one previous abortion was 1.17 (95%CI=1.02-1.35), and for women with two or more previous abortions it was 1.68 (95%CI=1.23-2.30), respectively, compared with the control cohort of similar gravidity. Only for women who had one abortion did the results follow the predicted pattern of a higher risk of retained placenta after a short pregnancy interval. No association with placenta previa was seen. Conclusions. The findings suggest a positive association between abortions and retained placenta in subsequent singleton live births, but the association was weak and confounding cannot be ruled out.