Contraception 89 (2014) 426 – 430

Original research article

Outcome of intrauterine pregnancies with intrauterine device in place and

effects of device location on prognosis☆
A. Seval Ozgu-Erdinc a,⁎, Ufuk Goker Tasdemir b , Dilek Uygur a ,
Ayla Aktulay a , Nicel Tasdemir c , H. Cavidan Gulerman a
a
Zekai Tahir Burak Women Health Care, Education and Research Hospital, Ankara, Turkey
b
Hayrabolu State Hospital, Tekirdag, Turkey
c
Namik Kemal University, Faculty of Medicine, Tekirdag, Turkey
Received 4 May 2013; revised 26 November 2013; accepted 4 January 2014

Abstract

Objectives: This study aimed to compare the outcome of pregnancies with retained or removed intrauterine devices (IUDs) and the effect of
IUD location on pregnancy outcome.
Study design: In a retrospective cohort study, we searched 27,578 records of women who had CuT380 IUD inserted, and 144 pregnancies
with IUD were analyzed. IUDs were removed from 114 patients and retained for 30 patients.
Results: The combined risk of adverse pregnancy outcomes (miscarriage, intrauterine fetal death, intrauterine growth retardation, preterm
birth and preterm premature rupture of membranes) was 36.8% in the IUD-removed group and 63.3% in the IUD-retained group [pb.01;
relative risk (RR)=2.0; 95% confidence interval (CI) 1.3–3.3]. Newborns of the IUD-retained women had significantly lower Apgar scores
and significantly higher admission rate to the neonatal intensive care unit (p=.01; RR=10.8; 95% CI 1.04–111.6 and pb.01; RR=4.5; 95% CI
1.5–12.9, respectively). There were more miscarriages and adverse pregnancy outcome when the IUD was retained (16.9% vs. 66.7%) in
patients with an IUD in low-lying position (pb.01; RR=3.9; 95% CI 1.8–8.6).
Conclusion: Women who conceived with an IUD in place and chose to continue the pregnancy without removing the IUD need close follow-
up, as there appears to be higher risk of adverse pregnancy and neonatal outcome. Furthermore, when the IUD is retained in the low-lying
position, there is increased risk of miscarriage and adverse pregnancy outcome compared to removal of the IUD. Future randomized
controlled studies are needed to determine the outcome of pregnancies with retained or removed IUD.
Implications: In this study, we have evaluated the IUD location and its effect on pregnancy outcome in women with a retained or removed
IUD. This study is the first to investigate the relationship between IUD location and pregnancy outcome in women who conceived with an
IUD. We need evidence from a collaborative multicenter randomized trial to answer the question of whether the IUD should be removed in
case of pregnancy.
© 2014 Elsevier Inc. All rights reserved.

Keywords: Pregnancy; Intrauterine device; Adverse pregnancy outcome; Location

1. Introduction first most used modern method (17%) after the traditional
withdrawal method (26%) in Turkey [2]. Pregnancy can
Intrauterine devices (IUDs) are widely used, safe and occur rarely, despite the presence of an IUD. The failure rate
effective contraceptives. IUDs are the fifth most used of this contraceptive method ranges from 0.8% to 2.3%
modern contraception method in the United States [1] and [3–6]. Studies investigating why pregnancies occur despite
the second most used method worldwide, while IUDs are the the presence of an IUD suggested composition of IUD
(copper surface area), duration of use, IUD position, age of
☆
women, history of expulsion and failure of the IUD as risk
Financial disclosure: The authors report no conflict of interest.
⁎ Corresponding author. Zekai Tahir Burak Women Health Care,
factors for the efficacy of IUD [7–9]. Some authors have
Training and Research Hospital, Talatpasa Bulvari, 06230, Ankara, Turkey. suggested removing the IUD during the first trimester of
Tel.: + 90 312 306 50 00. pregnancy to prevent septic complications and miscarriages
E-mail address: [email protected] (A.S. Ozgu-Erdinc). [10,11]. A retrospective study has shown that ongoing
0010-7824/$ – see front matter © 2014 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.contraception.2014.01.002
 A.S. Ozgu-Erdinc et al. / Contraception 89 (2014) 426–430 427

pregnancies with a retained IUD have risks, such as preterm were evaluated by using the Kruskal–Wallis test. For
delivery and chorioamnionitis [12]. However, randomized categorical comparisons, two-tailed Pearson χ 2 tests were
controlled studies comparing different management strate- used. Statistical significance was calculated using the t test
gies of pregnancies with an IUD in place are needed. for differences in continuous variables. A p value less than
The World Health Organization (WHO) has made a .01 was considered to be statistically significant.
recommendation for pregnancies in the presence of an IUD
which is mostly based on studies from the 1970s and 1980s [13].
WHO concluded that removing the IUD improves 3. Results
pregnancy outcome if the IUD strings are visible or can
be retrieved safely from the cervical canal and that the risks During the study period, 27,578 patients were seen in the
of miscarriage, preterm delivery and infection are sub- family planning unit for examination of an IUD. The cohort
stantial if the IUD is left in place [3]. As there are no consisted of 174 patients who had an IUD at the time of
randomized controlled studies about pregnancies with pregnancy diagnosis and did not want the termination of the
IUD; studies with large case series are useful to estimate pregnancy. Fourteen patients were excluded after the
the outcome of such pregnancies. diagnosis of an ectopic pregnancy. Patients with IUD
This study aimed to evaluate the pregnancy outcomes of expulsion before the diagnosis of pregnancy were not
women who conceived despite the presence of CuT380A included in the study. Except for 16 patients, all the remaining
IUD and decided to continue the pregnancy. We have patients were routinely followed up at our hospital. The
evaluated the IUD location and its effect on pregnancy outcomes for these 16 patients could not be retrieved. Thus,
outcome in both IUD-retained and -removed women. A the cohort consisted of the remaining 144 patients. All of the
detailed search of literature showed that there are no studies pregnancies were singleton. Moreover, 114 patients had
investigating the relationship between IUD location and chosen the IUD to be removed during the first trimester, and
pregnancy outcome in women who conceived with an IUD. the IUD was retained in the remaining 30 patients. The
clinical characteristics are summarized in Table 1. There was
no significant difference between the groups in terms of age
2. Materials and methods and parity. The mean gestational age at the time of diagnosis
was 7.4±2.6 weeks for the IUD-removed group and 8.7±3.8
A retrospective cohort study was conducted between 01 weeks for IUD-retained group (p=.1).
January 2005 and 01 January 2012 in the Family Planning Table 1 shows 73 patients (64%) in the IUD-removed
Unit of Zekai Tahir Burak Women Health Care, Education group and 11 (36.7%) patients in the IUD-retained group
and Research Hospital, Ankara, Turkey. The study was delivered at term. The difference between the groups for
approved by the institutional review board. In our clinical term pregnancies (deliveries of gestational week ≥ 37
practice, patients found to be pregnant during the weeks) was significant (pb.01) [relative risk (RR): 0.6,
examination are offered to remove the IUD if the strings 95% confidence interval (CI): 0.4–0.9]. The combined risk
are visible. Routine ultrasound examination is performed of adverse pregnancy outcomes (miscarriage, intrauterine
for IUD location. The IUD is removed by gently pulling fetal death, intrauterine growth retardation, preterm birth
the thread (in accordance with WHO recommendations) and preterm premature rupture of membranes) was 36.8%
from the patients in whom the IUD tail is still visible [3]. (n= 42) in the IUD-removed group and 63.3% (n= 19) in the
If the thread is inaccessible, no attempt is made to remove IUD-retained group (pb.01) (RR: 2.0, 95% CI: 1.3–3.3).
the IUD. Clinical records of family planning unit during There were fewer miscarriages in the IUD-removed group
the study period have been searched. All of the IUDs were than the IUD-retained group (pb.01) (RR: 2.0, 95% CI:
CuT380A. Clinical patient characteristics, such as maternal 1.3–3.3) (Table 1). Eighteen (15.8%) pregnancies in the
age, obstetric history, gestational age and medical history IUD-removed group and 8 (26.7%) pregnancies in the IUD-
were evaluated. Early pregnancy losses were recorded, and retained group were complicated by vaginal bleeding
the ongoing pregnancy outcomes were evaluated for mode during the first trimester (p=.2) (RR: 1.7, 95% CI: 0.8–
of delivery, birth weight, 5-min Apgar scores, admission 3.5). The outcome of patients experienced first trimester
to neonatal intensive care unit (NICU) and obstetric bleeding is presented in Table 1. There was no statistically
complications. We also analyzed the maternal serum C- significant difference in terms of intrauterine growth
reactive protein (CRP) level, white blood cell count retardation (IUGR), oligohydramnios and preterm prema-
(WBC) and fibrinogen level as markers of the inflamma- ture rupture of membranes (PPROM) between the two
tory response. groups (Table 1). There were no women with clinically
The statistical software package SPSS (SPSS Inc., diagnosed chorioamnionitis.
Chicago, IL, USA) was used for the statistical analyses. The median gestational week at birth for the IUD-
Whether the distributions of continuous and discrete vari- removed group was 39 weeks and was 37.5 weeks for the
ables were normal or not was determined by using the IUD-retained group (pb.01). There was no statistically
Kolmogorov–Smirnov test. Differences between the groups significant difference between the groups in terms of mode
428 A.S. Ozgu-Erdinc et al. / Contraception 89 (2014) 426–430

Table 1
Comparison of clinical and laboratory characteristics and obstetric outcome of patients with regard to removal of IUD
IUD-removed group IUD-retained group p value RR(95% CI) for
n= 114 (79.2%) n=30 (20.8%) removed IUD
Age (mean, year) 27.5±4.9 29±5.4 .2
Parity (median) 2 (1–6) 2 (1–5) .6
Gestational week at diagnosis (mean) 7.4±2.6 8.7±3.8 .1
Gestational week at birth (median) 39 (24–41) 37.5 (22–40) b.01
Birth weight (median, g) 3400 (730–5140) 3360 (500–4000) .7
Laboratory findings
Median CRP (mg/L), n=32 6.3 (0.9–40) 16.1 (2.2–27.5) .9
Mean WBC, n= 144 10,916±3109 11,193±4190 .7
Mean fibrinogen (mg/dL) n=46 446.3±144.9 485.9±90.3 .3
Term birth, n (%) 73 (64.0%) 11 (36.7%) b.01 0,6 (0.4–0.9)
Preterm birth, n (%) 13 (11.4%) 5 (16.7%) .4 1.5 (0.6–3.8)
Miscarriage, n (%) 28 (24.6%) 15 (50%) b.01 2.0 (1.3–3.3)
First-trimester bleeding, n (%) 18 (15.8%) 8 (26.7%) .2 1.7 (0.8–3.5)
First-trimester bleeding outcome, n
Miscarriage 7 4
Preterm birth 4 0
Term birth 7 4
Oligohydramnios, n (%) 2 (2.3%) 1 (6.3%) .4 2.7 (0.3–27.9)
Intrauterine growth retardation, n (%) 2 (2.3%) 0 .5 1.0 (0.9–1.1)
Preterm premature rupture of membranes, n (%) 2 (2.3%) 2 (12.5%) .05 5.4 (0.8–35.4)
Adverse pregnancy outcome, n (%) 42 (36.8%) 19 (63.3%) b.01 2.0 (1.3–3.3)
Cesarean section, n (%) 38 (44.2%) 8 (50%) .7 1.1 (0.7–1.9)
Birth weight under 2500 g, n (%) 8 (9.3%) 4 (25.0%) .1 2.7 (0.9–7.9)
NICU, n (%) 6 (7.0%) 5 (31.3%) b.01 4.5 (1.5–12.9)
Apgar at 5 min b7, n (%) 1 (1.2%) 2 (12.5%) .01 10.8 (1.04–111.6)
Statistical significance (pb0.01) is stated as bold and underlined.

of delivery, median birth weight and delivery of a baby less outcome was more in the IUD-retained women compared to
than 2500 g (Table 1). Two babies had an Apgar score below the IUD-removed women (pb.01, RR=13.5, 95% CI: 1.9–
7 at 5 min after delivery in the IUD-retained group, while one 94.1 and p=0.01, RR=2.6, 95% CI: 1.6–4.3, respectively).
baby had a low Apgar score in the IUD-removed group (p= The mean WBC, mean fibrinogen levels and median of
0.01) (RR: 10.8, 95% CI: 1.04–111.6). Six (7%) babies born CRP levels were statistically insignificant between the IUD-
from IUD-removed women and 5 (31.3%) babies born from retained and -removed groups (Table 1). Although WBC data
IUD-retained women were admitted to the NICU (pb.01) were available for all of the patients, CRP and fibrinogen
(RR: 4.5, 95% CI: 1.5–12.9). There was no newborn with levels were available only for some of the women as it is not
congenital anomalies in either group. There were no obstetric our routine practice to measure CRP or fibrinogen. Thus, the
or postpartum complications noted in the records. results of the statistical analysis have limitations.
The comparison of obstetric outcomes between removed
and retained IUDs with regard to IUD position is presented
in Table 2. When the IUD was located in the uterine cavity, 4. Discussion
there were no statistically significant differences in any of the
outcome measures between the removed and retained IUD Pregnancies in the presence of an IUD have a greater risk
groups. When the IUD was in low-lying position, there were of adverse obstetric outcomes than pregnancies in the
no statistically significant differences between the removed absence of an IUD [12,14,15]. Spontaneous miscarriage,
and retained IUD groups in terms of gestational week at preterm delivery and chorioamnionitis are the described
birth, birth weight, mode of delivery, number of patients with adverse outcomes for these patients. In the present study, the
oligohydramnios, IUGR, first-trimester vaginal bleeding and overall incidence of adverse pregnancy outcomes was higher
preterm births (Table 2). In women with a low-lying IUD, in the IUD-retained group than the IUD-removed group. The
there were more miscarriages in the IUD-retained women two groups were similar in terms of age and parity, which
(66.7%) than IUD-removed women (16.9%) (pb.01, RR: would be important confounding factors for many pregnancy
3.9, 95% CI: 1.8–8.6). In the low-lying IUD group, the outcomes. This finding and the incidences were similar to the
number of patients delivered at term was more in the IUD- results presented by other authors [16,17].
removed women (p=.01, RR: 0.2, 95% CI: 0.0–1.5) than We found a statistically significant increased risk of
IUD-retained women. In the low-lying IUD group, the miscarriages in women with a retained IUD compared to
number of patients with PPROM and adverse perinatal those with a removed IUD. Most of the observational studies
 A.S. Ozgu-Erdinc et al. / Contraception 89 (2014) 426–430 429

Table 2
Impact of IUD removal on obstetric outcome with regard to IUD position
Uterine cavity Low lying
Removed Retained p value RR (95% CI) Removed Retained p value RR (95% CI)
IUD n= 49 IUD n= 24 IUD n= 65 IUD n= 6
Gestational week at birth (median) 40 (31–41) 37.5 (22–40) 0.1 39 (24–41) 35.5 (32–39) .9
Birth weight (median, g) 3495 3360 0.2 3330 2995 1
(1210–4650) (500–4000) (730–5140) (2090–3900)
Preterm birth, n (%) 4 (8.2%) 4 (16.7%) 0.3 2.0 (0.6–7.5) 9 (13.8%) 1 (16.7%) .9 1.2 (0.2–8.0)
Term birth, n (%) 28 (57.1%) 10 (41.7%) 0.2 0.7 (0.4–1.2) 45 (69.2%) 1 (16.7%) .01 0.2 (0.0–1.5)
Miscarriage, n (%) 17 (34.7%) 11 (45.8%) 0.4 1.3 (0.7–2.4) 11 (16.9%) 4 (66.7%) b .01 3.9 (1.8–8.6)
First-trimester bleeding, n (%) 11 (22.4%) 7 (29.2%) 0.5 1.3 (0.6–2.9) 7 (10.8%) 1 (16.7%)
Birth weight under 2500 g, n (%) 2 (6.3%) 3 (21.4%) 0.1 3.4 (0.6–18.3) 6 (11.1%) 1 (50%) .1 4.5 (0.9–21.8)
Apgar at 5 min b 7, n (%) 0 2 (14.3%) 0.03 1 (1.9%) 0 .9
Cesarean section, n (%) 16 (50%) 6 (42.9) 0.7 0.9 (0.4–1.7) 22 (40.7) 0 .3
Oligohydramnios, n (%) 0 1 (7.1%) 0.1 2 (3.7%) 0 .8
IUGR, n (%) 0 0 – – 2 (3.7%) 0 .8
PPROM, n (%) 0 1 (7.1%) 0.1 2 (3.7%) 1 (50%) b .01 13.5 (1.9–94.1)
Adverse pregnancy outcome, n (%) 21 (42.9%) 14 (58.3%) 0.2 1.4 (0.9–2.2) 21 (32.3%) 5 (83.3%) .01 2.6 (1.6–4.3)
Statistical significance (pb0.01) is stated as bold and underlined.

and one prospective study have supported this finding adverse perinatal outcome risk when we did not remove a low-
[15,18,19]. lying-positioned IUD. These findings may suggest that IUD
The IUD-retained women delivered babies with a lower removal during pregnancy should be considered especially
gestational week compared to the IUD-removed women. when the IUD is low lying. However, as the number of women
Although not statistically significant, pregnancies in the with retained IUD in low-lying position was small, larger-
presence of an IUD have a higher incidence of preterm population-sized prospective studies are needed to clarify the
delivery and PPROM compared to IUD-removed women. question of whether the IUD should be removed or not if it is
Studies have postulated a high incidence of preterm delivery low lying.
in the IUD-retained group with statistical significance We found higher WBC and CRP and fibrinogen levels in
[12,13,18]. However, the difference between the groups in the IUD-retained group, but the difference did not reach
these studies, in terms of preterm delivery, was not statistical significance. In their small-sample-size prospective
confirmed with statistical analysis in a systematic review study, Deveer et al. found a higher CRP level in IUD-retained
by Brahmi et al. [20]. Also, a study by Kim et al. reported a women and explained this finding as chronic inflammation
higher PPROM incidence in pregnancies with an IUD in situ with a low level of bacteremia [19]. We could only have a
[14]. Prospective controlled studies are needed to clarify this small portion of the women’s CRP results, and that might
question. In contrast to previous studies [20], there were decrease the power of these statistical results. As our study is
more newborns with a low Apgar score in the IUD-retained retrospective, a histological evaluation of inflammation was
women in the present study. The need for the NICU was also not possible. An evaluation of inflammation with histological
significantly greater for the IUD-retained women. and microbiological findings would be more valuable [23].
Pregnancy among IUD users is not uncommon. However, In conclusion, our study confirms that pregnant women
debates over management of such pregnancies continue due with an IUD in situ have a higher risk of an adverse
to absence of randomized controlled trials about the pregnancy outcome, lower gestational week, lower Apgar
prognosis of pregnancies conceived in the presence of an scores and greater need for the NICU at birth. Moreover,
IUD [21]. Therefore, we believe that retrospective cohort when the IUDs in low-lying position were not removed, the
studies investigating the results of pregnancies in the women had more miscarriages and adverse perinatal
presence of an IUD are still important. No previous study outcome. Prospective controlled studies evaluating the
has investigated the effect of IUD removal with regard to pregnancy outcome of women with an IUD in place are
location of the IUD at the time of diagnosis and its effect on needed for counseling and management of such pregnancies.
pregnancy outcome. The only study evaluating the location of We recommend that when the IUD is removed from the
IUD in the presence of pregnancy reported by Moschos et al. uterine cavity or retained in situ during pregnancy, close
[22] concluded that an intrauterine pregnancy was three times maternal and fetal surveillance is needed.
as likely with a malpositioned or missing IUD, but they did not
investigate the effect of IUD location on pregnancy outcome. Acknowledgment
In the present study, when the IUD was in the uterine cavity,
IUD removal did not make any difference on pregnancy The authors wish to thank Michael Jones and Sibel Gunal
outcome. Importantly, we observed a higher miscarriage and Jones for revising the article as native speaker.
430 A.S. Ozgu-Erdinc et al. / Contraception 89 (2014) 426–430

References [13] Chaim W, Mazor M. Pregnancy with an intrauterine device in situ and
preterm delivery. Arch Gynecol Obstet 1992;252:21–4.
[1] Mosher WD, Jones J. Use of contraception in the United States: 1982– [14] Kim SK, Romero R, Kusanovic JP, et al. The prognosis of pregnancy
2008. Vital Health Stat 2010;2010:1–4. conceived despite the presence of an intrauterine device (IUD). J
[2] Akadli Ergocmen BYE, Tunckanat FH. Family Planning. In: & of Perinat Med 2010;38:45–53.
HUI, (HUIPS). PS, editor. Turkey demographic and health survey [15] Tatum HJ, Schmidt FH, Jain AK. Management and outcome of
2008. Ankara,Turkey: Hacettepe University Institute of Population pregnancies associated with the copper T intrauterine contraceptive
Studies, T.R. Ministry of Health General Directorate of Mother and device. Am J Obstet Gynecol 1976;126:869–79.
Child Health, T.R. Prime Ministry State Planning Organization; 2009, [16] von Theobald P, Duchemin JM, Levy G. [The outcome of continuing
pp. 75–93. pregnancies in patients with intrauterine devices. A retrospective study
[3] Organization WH. Selected practice recommendations for contracep- from the Maternity Unit of the University Hospital Center at Caen
tive use. Research DoRHa. Geneva: World Health Organization; 2004. during the period 1985–1988]. J Gynecol Obstet Biol Reprod (Paris)
[4] de Araujo FF, Barbieri M, Guazzelli CA, Lindsey PC. The T 380A 1990;19:863–8.
intrauterine device: a retrospective 5-year evaluation. Contraception [17] Mermet J, Bolcato C, Rudigoz RC, Dargent D. Outcome of
2008;78:474–8. pregnancies with an intrauterine devices and their management. Rev
[5] Rivera R, Chi IC, Farr G. The intrauterine device in the present and Fr Gynecol Obstet 1986;81:233–5.
future. Curr Opin Obstet Gynecol 1993;5:829–32. [18] Inal MM, Ertopcu K, Ozelmas I. The evaluation of 318 intrauterine
[6] Trussell J. Contraceptive failure in the United States. Contraception pregnancy cases with an intrauterine device. Eur J Contracept Reprod
2011;83:397–404. Health Care 2005;10:266–71.
[7] Thonneau P, Almont T, de La Rochebrochard E, Maria B. Risk factors [19] Deveer R, Engin-Ustun Y, Sarikaya E, Aydogan P, Doganay M,
for IUD failure: results of a large multicentre case–control study. Hum Mollamahmutoglu L. Comparison of C-reactive protein levels in
Reprod 2006;21:2612–6. pregnancies with retained and removed intrauterine device. J Matern
[8] Thonneau P, Goulard H, Goyaux N. Risk factors for intrauterine device Fetal Neonatal Med 2011;24:1152–4.
failure: a review. Contraception 2001;64:33–7. [20] Brahmi D, Steenland MW, Renner RM, Gaffield ME, Curtis KM.
[9] Thonneau PF, Almont T. Contraceptive efficacy of intrauterine Pregnancy outcomes with an IUD in situ: a systematic review.
devices. Am J Obstet Gynecol 2008;198:248–53. Contraception 2012;85:131–9.
[10] Skjeldestad FE, Hammervold R, Peterson DR. Outcomes of pregnancy [21] Owen C, Sober S, Schreiber CA. Controversies in family planning:
with an IUD in situ — a population based case–control study. Adv desired pregnancy, IUD in situ and no strings visible. Contraception
Contracept 1988;4:265–70. 2013;88:330–3.
[11] Horn LC, Nenoff P, Ziegert M, Hockel M. Missed abortion [22] Moschos E, Twickler DM. Intrauterine devices in early pregnancy:
complicated by Candida infection in a woman with rested IUD. findings on ultrasound and clinical outcomes. Am J Obstet Gynecol
Arch Gynecol Obstet 2001;264:215–7. 2011;204:427 e1-6.
[12] Ganer H, Levy A, Ohel I, Sheiner E. Pregnancy outcome in women [23] Fulcheri E, di Capua E, Ragni N. Pregnancy despite IUD: adverse
with an intrauterine contraceptive device. Am J Obstet Gynecol effects on pregnancy evolution and fetus. Contraception 2003;68:
2009;201:381 e1–5. 35–8.