Objective
The objective of this study was to describe ultrasound findings, clinical symptoms, and outcomes of first-trimester pregnancies with intrauterine devices (IUDs).
Study Design
This was a retrospective review of 42 women with history of IUD placement and positive serum human chorionic gonadotropin in the first trimester.
Results
There were 31 intrauterine pregnancies (IUPs), 3 ectopic pregnancies, and 8 pregnancies of unknown location. Of 36 IUDs visualized, 15 were normally positioned and 21 malpositioned. Of 31 IUPs, 8 IUDs were within the endometrium, 17 were malpositioned, and 6 were not seen. Indications included bleeding (14 of 31), pain (12 of 31), and missing strings (5 of 31); 11 had no symptoms. Of 26 IUPs with known pregnancy outcomes, 20 were term deliveries and 6 had failed pregnancies of 20 weeks or less.
Conclusion
More than half of IUDs identified in the first trimester were malpositioned. IUP was 3 times as likely with a malpositioned or missing IUD. Three quarters of the IUPs with known outcomes had term deliveries. Symptoms were not predictive of IUD malposition.
Although the intrauterine device (IUD) is a highly effective form of birth control, complications, including pregnancy, do occur. Long-term cumulative pregnancy rates with IUD use are extremely low but vary slightly with the IUD type: 1.4-1.9 per 100 women for the copper containing ParaGard T 380A (Duramed Pharmaceuticals, Inc, Montvale, NJ) and 0.5-1.1 per 100 women for the levonorgestrel-releasing Mirena (Bayer Schering Pharma Oy, Turku, Finland).
Timely evaluation of the pregnant woman with an IUD is important in an effort to reduce subsequent complications if the woman chooses to continue her pregnancy. Evaluation typically includes transvaginal sonography to establish the pregnancy location and to identify IUD position and retrievability.
The purpose of our study was to describe the ultrasound findings, clinical characteristics, and pregnancy outcomes of first-trimester pregnancies with IUDs.
Materials and Methods
From July 1, 2008, through June 30, 2009, a retrospective review of our ultrasound database was performed for women with history of IUD placement and a positive serum beta-human chorionic gonadotropin. Ultrasound images and reports were reviewed for identification and location of pregnancy, identification, type and position of IUD, and presenting symptoms. Outcomes of the pregnancies were reviewed. All ultrasound procedures were performed in the obstetrics and gynecology ultrasound unit at our county-based teaching hospital. The study was approved by our institutional review board (#072009-063).
The 2-dimensional (2D) and/or 3-dimensional (3D) images of the uterine cavities were obtained with an Antares (Siemens, Redmond, WA), an Elegra (Siemens), or a Voluson 730 (General Electric Medical Systems, Milwaukee, WI) scanner and 5 MHz or multifrequency transvaginal transducers. The method of coronal view reconstruction was derived from the Z-plane technique as described by Abuhamad et al.
The location of the pregnancy was determined to be either intrauterine, as defined by the minimum criteria of a double decidual sign within the endometrial cavity, ectopic (no gestational sac within the uterus and an adnexal mass), or pregnancy of unknown location (PUL).
The IUD position was considered to be appropriately endometrial if both the arms and shaft were seen within the fundal or midportion of the cavity ( Figure 1 ). The IUD was considered malpositioned if any part extended into the lower uterine segment, myometrium, or endocervical canal, as depicted in Figure 2 , A and B.
Statistical significance was established at a P ≤ .05 for the analyses. A Fisher’s exact test was used to test the data from 2 independent variables consisting of the measurements classified above. All tests were performed using SAS version 9.2 (SAS Institute, Cary, NC).
Results
As a reference, 4157 IUDs were placed in our health care system from July 1, 2008, through June 30, 2009. Forty-two patients with biochemical or sonographic evidence of pregnancy and history of IUD placement were evaluated. The principal race/ethnicity of our patient population was Hispanic (36 of 42 or 85%), followed by equal numbers of African or African American (2 of 42 or 5%), white (2 of 42 or 5%), and other ethnicities (2 of 42 or 5%). The patients were multiparous, with a mean gravidity of 3 (range, 2–7) pregnancies and a mean parity of 2 (range, 1–5) deliveries. The mean age was 26 years (range, 18–38 years). Data on the time of IUD usage were available in 32 of the 42 patients, with an average of 24.7 months (range, 2–60 months). The mean estimated gestational age at presentation and diagnosis of pregnancy and IUD location was 8.0 weeks (range, 3.4–15.0 weeks).
Of these 42 patients, 31 (74%) had intrauterine pregnancies, 3 (7%) had ectopic pregnancies, and 8 (19%) were diagnosed with PUL.
The most common type of IUD was the copper T 380A or ParaGard (30 of 42 or 72%) ( Figure 1 ), followed by the levonorgestrel-containing IUD or Mirena (5 of 42 or 12%) and 1 Lippes loops (2%) ( Figure 3 ). The distribution of IUD types in this series of gravid patients was similar to that in our nongravid population in a series of symptomatic women. Of the 36 IUDs seen, 15 patients (42%) had IUDs within the endometrial cavity and 21 patients (58%) had malpositioned IUDs. In all cases, the determination of IUD type and location was made on the initial sonographic evaluation by 2D ultrasound, and in 6 patients in whom 2D imaging was equivocal, 3D imaging was performed to confirm the diagnosis.
Among the 31 intrauterine pregnancies (IUPs), 8 (26%) had IUDs within the endometrial cavity ( Figure 1 ), 17 (55%) had malpositioned IUDs ( Figure 2 , A and B), and 6 (19%) had expelled their IUDs as described above. In all 3 ectopic pregnancy patients, an appropriately positioned copper IUD was visualized, as shown in Figure 4 , A and B. Of the PULs, 5 IUDs (62.5%) were within the endometrium and 3 (37.5%) were malpositioned.
Pregnancy outcomes were known in 37 of 42 cases (88%). All 8 pregnancies of unknown location resulted in spontaneous abortions. The 3 ectopic pregnancies were successfully treated; 2 received methotrexate and the third underwent a laparoscopic salpingectomy for rupture of the tubal pregnancy. The Table depicts the intrauterine pregnancy outcomes as they related to sonographic findings. Of 31 intrauterine pregnancies, 5 (16%) were lost to follow-up and outcomes were available in the remaining 26 women (84%), who all desired to continue their pregnancies.
