Lost intrauterine devices during pregnancy - Wiley Online Library

Ultrasound Obstet Gynecol 2004; 23: 486–489 Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/uog.1036

Lost intrauterine devices during pregnancy: maternal and fetal outcome after ultrasound-guided extraction. An analysis of 82 cases M. SCHIESSER, O. LAPAIRE, S. TERCANLI and W. HOLZGREVE University Women’s Hospital Basel, Basel, Switzerland

K E Y W O R D S: lost IUD; pregnancy; ultrasound-guided extraction

ABSTRACT Objectives It is generally agreed that intrauterine devices (IUDs) with visible strings in pregnancy should be removed because of the increased risk of miscarriage, septic complications and premature delivery. The precise management of pregnancies in association with so-called ‘lost IUDs’, and especially the technique of their removal, has remained controversial. We present our experience of the management of intrauterine pregnancies with a lost IUD. Methods Ultrasound-guided extraction of a lost IUD was performed in 82 intrauterine pregnancies. The subsequent outcome of the pregnancies is described. Results There were no intra- or post-procedure maternal complications. Although the miscarriage rate in the first 3 weeks after the procedure was higher than that in normal pregnancy, the complication rate approached that of normal pregnancy as the pregnancies progressed. The total miscarriage rate of 22% was comparable to that following extraction of IUDs with visible filaments. The rate of live births was 77.0%. Delivery before 37 weeks occurred in 13.5% of cases. Conclusions Ultrasound-guided extraction is a minimally invasive and inexpensive procedure that is associated with few postoperative complications. It has a high success rate and is associated with a moderate miscarriage rate and no maternal complications. Copyright  2004 ISUOG. Published by John Wiley & Sons, Ltd.

INTRODUCTION Intrauterine devices (IUDs) are one of the most widely used contraceptive methods with an estimated 130 million women using them worldwide1 . The prevalence of IUD

users in most European countries and North America is lower (range, 3–24%) than that in developing countries such as China (33%) or Korea (49%)2 . On account of the widespread use of IUDs, unexpected pregnancies are not uncommon. Some risk factors of IUD failure, for example, displacement of the device, are well known. Controversy exists about the role of anti-inflammatory drugs, which may have a negative effect on the Pearl index of IUDs3 . In the event of a pregnancy occurring with an IUD in situ, the patient may be concerned not only about having an unwanted pregnancy but also, if she decides to continue with the pregnancy, about its outcome and the effect of the IUD on her health and especially that of the developing fetus. It is generally agreed that an IUD should be removed in the first trimester when its filaments are visible to reduce the likelihood of complications4 . Controversy exists over the management of so-called ‘lost IUDs’. In these cases the tail string is no longer visible due to retraction of the filaments into the growing uterine cavity. Large-scale studies from China comprising in excess of 3300 patients have shown that it is not the IUD’s filaments but the IUD itself that substantially increases the risk of an infection5 . Hucke demonstrated colonization with facultative pathogenic bacteria in all lost IUDs extracted by his group6 . If a pregnant woman with an IUD decides to continue her pregnancy there are two principal options as to how to proceed. First, the IUD can be left in situ. In this scenario the patient should be counseled on the risk of spontaneous miscarriage, which according to Tatum et al. is about 50%7 , one half of these miscarriages occurring in the second trimester, and on the risk of premature rupture of membranes, premature labor and a possibility of infection. Second, the IUD can be removed with hysteroscopy or under ultrasound guidance. The data presented here were collected to determine the maternal and fetal outcome after ultrasound-guided

Correspondence to: Dr M. Schiesser, University Women’s Hospital Basel, Spitalstrasse 21, CH-4031 Basel, Switzerland (e-mail: [email protected]) Accepted: 12 December 2003

Copyright  2004 ISUOG. Published by John Wiley & Sons, Ltd.

ORIGINAL PAPER

Extraction of lost IUDs removal of lost IUDs. To our knowledge, this is the largest reported series of ultrasound-guided extractions during pregnancy.

METHODS During the period 1989–2002, 82 pregnant women decided to undergo ultrasound-guided extraction of their lost IUD after counseling. In these patients ultrasound examination revealed clearly visible copper IUDs (Multiload Cu375) within the uterine cavity and live singleton pregnancies (Figure 1) in all cases with no filaments on speculum examination. All patients were counseled both about the possible hazards of continuing the pregnancy with an IUD in situ and the risks of the removal procedure. After informed consent was obtained we performed ultrasound-guided extraction of the IUDs using Holzgreve’s pliers (prototype from Karl Storz, Tuttlingen, Germany; Figure 2). A total of 6/82 patients originally opted for conservative management but because of maternal bleeding and/or severe abdominal pain it was decided that their IUDs should be extracted and this was done between the 13th and 17th weeks of gestation. Prior to IUD removal, the vagina and portio were disinfected with an antiseptic solution. If necessary, the

Figure 1 Intrauterine pregnancy in the sixth week of gestation with an intrauterine device at the posterior wall of the uterus.

487

anterior lip of the cervix was grasped with a tenaculum without analgesia. Under transabdominal ultrasound guidance the IUD was slowly removed with the pliers (diameter 2.3 mm). When possible, the long arm of the IUD was grasped at the most caudal end of the device. In most cases with an IUD in the fundus it was approached retroplacentally. After the procedure all patients were discharged the same day in good physical condition.

RESULTS IUD removal was successfully performed in 81 women. Only one IUD could not be extracted because of its position in the uterine fundus. This woman miscarried 5 days after the procedure. There were no intra- or postoperative maternal complications in the successful cases. Ninetythree percent of the IUDs (75/81) were removed in the first trimester (Figure 3). Because of prolonged bleeding and severe abdominal pain an extraction was performed between the 13th and 17th weeks of gestation in 7% (6/81) of cases after an initially conservative approach. In 4.9% (4/81) of cases the women decided subsequently to undergo termination of pregnancy because of socioeconomic and/or psychological reasons and these were excluded from the analysis. Only 3.7% (3/81) of cases were lost to follow-up after the 20th week of gestation but all three pregnancies were uneventful up to that time. The total miscarriage rate was 22.1% (17/77). In this group 41.2% (7/17) of miscarriages occurred within 5 days of the procedure, 58.8% (10/17) within 10 days, 82.3% (14/17) within 17 days and 17.6% (3/17) later than 20 days after the extraction (Figure 4). The rate of live births was 77.0% (57/74), 86.5% were at term and 13.5% (10/74) had a preterm delivery. One birth occurred in the 24th gestational week after premature rupture of membranes and chorioamnionitis; in a second case a Cesarean section was performed in the 31st gestational week following the onset of heavy vaginal bleeding. The outcome in the group in which the IUD was removed at the beginning of the second trimester (six women) did not differ significantly from that in the firsttrimester group, probably due to the small number of cases in the second-trimester group. One miscarriage occurred 25 20

n

15 10 5 0

6

7

8

9

10

11

12

13

15

>15

Gestational weeks

Figure 2 Holzgreve’s pliers (diameter 2.3 mm, flexible end) developed in collaboration with Karl Storz, Tuttlingen, Germany.


Figure 3 Gestational week at the time of ultrasound-guided extraction.

Ultrasound Obstet Gynecol 2004; 23: 486–489.

Schiesser et al.

488

In the 1970s in particular, the Dalkon Shield was associated with an elevated risk of maternal death as compared to other IUDs14 . The maternal mortality rate due to septicemia was 50 times greater in cases with this type of IUD in situ. A possible explanation for this was either the polyfilament tail or insufficient experience with the insertion of the new product15 . An increased risk of miscarriage and preterm delivery with an IUD in situ has been found in numerous studies: Tatum et al. found that the risk of a miscarriage is three times higher and the risk of preterm delivery is four times higher when the IUD remains in situ7 . The risk of septic complications in such cases is increased 26-fold16 . The removal of the IUD lowers the miscarriage rate from 48.3% to 29.6%8 . This finding led to the general recommendation of removing those IUDs in which the tails are visible17 . Uncertainty exists about the precise management of lost IUDs and especially the technique of their removal, mainly due to limited experience and data in the literature. Table 1 lists the studies published to date on alternative methods for the removal of lost IUDs. The miscarriage rate in our cohort was 22.1%. Other authors who performed an ultrasound-guided IUD extraction18 – 22 found either no miscarriages, probably because of small case numbers18,21 , or they observed a high miscarriage rate of 46%20 . Studies on hysteroscopic extraction23 – 26 have also generated rather different complication rates. While Assaf et al.23 reported a miscarriage rate of 42.9% (9/21) beyond 24 weeks, Neis et al.24 observed only two miscarriages in their cohort of 26 patients. Because of high rates of subsequently terminated pregnancies, a comparison with the data of other authors25,26 is not feasible.

7 6 5 n

4 3 2 1 0

0–5

6– 10

11– 15

> 15

Days

Figure 4 Time period between ultrasound-guided extraction and miscarriage.

1 week after the removal in the 18th week. One patient had transient rupture of the membranes but delivered a healthy baby. The remaining four pregnancies were uneventful.

DISCUSSION Despite the good Pearl index of IUDs, the widespread use of this contraceptive device makes failures unavoidable. Pregnant women with an IUD in situ are at increased risk of first-, and especially second-, trimester miscarriage mainly due to infectious etiologies7,8 . Not only virulent bacteria, but also harmless pathogens such as Candida albicans, can cause chorioamnionitis, ascending fetal infections or maternal septicemia9 – 11 . Furthermore, women tend to be concerned about fetal well-being in the presence of a foreign body, as demonstrated by data from a teratological advisory center12 , although to date neither teratogenic nor other detrimental effects on embryological and fetal development have been detected13 .

Table 1 Comparison of ultrasound-guided extraction vs. hysteroscopic procedure Preterm Extractions/ Miscarriage delivery total < 24 weeks (< 37th number ( n) week) (n) Live births (n)

Procedure employed

Reference

Year

Ultrasoundguided extraction

Stubblefield et al.18 Sviggum et al.19 Kirkinen et al.20

1988 1991 1992

4/4 8/9 26/26

0 1 12

0 1 1

Sachs et al.21 Holzgreve et al.22

1992 1988

3/3 16/16

0 5

0 —

Current study Assaf et al.23

2004 1988

81/82 19/21

17 9

10 9

Lin et al.25 Neis et al.24

1993 1994

28/33 25/26

6* 2

1 3

Holzgreve et al.22 Wagner et al.26 Hucke et al.6

1988 1980 1991

26/43 42/54† 13/13‡

6 8 + 28* 0

— 1 3

Hysteroscopic extraction

Mixed cohort

4 7 14 3 7 (plus 4 ongoing pregnancies) 57 12 24 18 (plus 3 ongoing pregnancies) 22 6 11

Complications/therapy None PROM, IUGR Placental abruption, pre-eclampsia None None PROM in the 24th week Analgesia, hospitalization 5 days Antibiotics Antibiotics, hospitalization 1 day — None Narcotics

*Termination of pregnancy. †16/24 with pliers and probe, 8/12 ultrasound-guided, 18/18 hysteroscopic. ‡8/8 hysteroscopic, 5/5 ultrasound-guided. IUGR, intrauterine growth restriction; PROM, premature rupture of membranes.



Extraction of lost IUDs Another important aspect of our study is the short time span between the procedure itself and the miscarriage. In 59% of cases miscarriages occurred before 10 days, which means that the miscarriage took place in the first trimester. During this period the emotional stress for the woman is not generally thought to be as great as later in pregnancy, especially after quickening. Of all the live births in our cohort 13.5% occurred prior to 37 weeks. This moderately increased rate of prematurity compared to background data27 corresponds to the situation in any prenatal center with a high percentage of high-risk pregnancies and is applicable to other centers also28 . The overall rate of preterm deliveries associated with lost IUDs reported in the literature is low in the ultrasound extraction group (7.1%, 2/28) compared with that in the hysteroscopy extraction group (24.1%, 13/54). Studies included in this comparison were those that differentiated the outcome of successful vs. unsuccessful removals. The success rate of ultrasoundguided extraction also seems to be slightly higher than that in the hysteroscopic group. This is most likely due to the flexible ultrasound-guided extraction device as compared to the thicker, non-flexible hysteroscope. Further advantages of ultrasound-guided extraction are lower costs (because operating rooms and their staff are not required), less invasiveness and, in some cases, the avoidance of postoperative hospitalization, antibiotics and analgesia23 – 25 . Our high success rate with no maternal complications and a moderate miscarriage rate compared to other reports in the literature supports not only the advantage of an ultrasound-guided procedure, but also shows that practice may have an influence on the selection of the method and the outcome. The clinical expertise gained from chorionic villus sampling was helpful for the transvaginal approach and intrauterine manipulation29 . Clearly the technical development itself helped to improve this method due to refinement of the ultrasound and extraction equipment. Based on the presented data we recommend ultrasoundguided extraction of lost IUDs in all cases as a minimally invasive, well tolerated and easily practicable method with a high degree of safety for the ongoing pregnancy.

REFERENCES 1. Thonneau P, Goulard H, Goyaux N. Risk factors for intrauterine device failure: a review. Contraception 2001; 64: 33–37. 2. Hieu DT, Van HAT, Donaldson PJ, Nga QL. The pattern of IUD use in Vietnam. Int Fam Plann Perspect 1995; 21: 6–10. 3. Papiernik E, Rozenbaum H, Amblard P, Dephot N, de Mouzon J. Intrauterine device failure: relation with drug use. Eur J Obstet Gyn Reprod Biol 1989; 32: 205–212. 4. Tatum HJ, Connell EB. A decade of intrauterine contraception: 1976 to 1986. Fertil Steril 1986; 46: 173–192. 5. Grimes D. Intrauterine device and upper-genital-tract infection. Lancet 2000; 356: 1013–1019. 6. Hucke J, Campo RL, Kozlowski P, De Bruyne F. Erfahrungen mit der hysteroskopischen oder sonographische gesteuerten Entfernung einer Intrauterinspirale ohne sichtbaren Faden in ¨ der Fruhschwangerschaft. Geburtshilfe Frauenheilkd 1991; 51: 31–33.


489 7. Tatum HJ, Schmidt FH, Jain AK. Management and outcome of pregnancies associated with the Copper T intrauterine contraceptive device. Am J Obstet Gynecol 1976; 1: 869–880. 8. Alvior GT. Pregnancy outcome with removal of intrauterine device. Obstet Gynecol 1973; 41: 894–896. 9. Roque H, Abdelhak Y, Young BK. Intra amniotic candidiasis. Case report and meta-analysis of 54 cases. J Perinat Med 1999; 27: 253–262. 10. Segal D, Gohar J, Huleihel M, Mazor M. Fetal death associated with asymptomatic intrauterine Candida albicans infection and a retained intrauterine contraceptive device. Scand J Infect Dis 2001; 33: 77–78. 11. Marelli G, Mariani A, Frigerio L, Leone E, Ferrari A. Fetal Candida infection associated with an intrauterine contraceptive device. Eur J Obstet Gynecol Reprod Biol 1996; 68: 209–212. ¨ ¨ 12. Tews G, Arzt W, Bsteh M. Schadliche Einflusse in der ¨ Fruhschwangerschaft – Ergebnisse eines teratologischen Beratungszentrums. Wien Klin Wochenschr 1990; 102: 466–471. 13. Barash A, Shoham Z, Borenstein R, Nebel L. Development of human embryos in the presence of a copper intrauterine device. Gynecol Obstet Invest 1990; 29: 203–206. 14. Christian CD. Maternal deaths associated with an intrauterine device. Am J Obstet Gynecol 1974; 119: 441–444. 15. Cates W, Ory HW, Rochat RW, Tyler CW. The intrauterine device and deaths from spontaneous abortion. N Engl J Med 1976; 295: 1155–1159. 16. Foreman A, Stadel BV, Schlesselmann S. Intrauterine device usage and fetal loss. Obstet Gynecol 1981; 58: 669–677. ¨ 17. Arzneimittelkommission der deutschen Arzteschaft: Vorsicht bei Schwangerschaften trotz liegendem Intrauterinpessar! Deutsches Arzteblatt 1978; 2: 71. 18. Stubblefield PG, Fuller AF, Foster SC. Ultrasound guided intrauterine removal of intrauterine contraceptive devices in pregnancy. Obstet Gynecol 1988; 72: 961–964. 19. Sviggum O, Skjeldestad FE, Tuveng JM. Ultrasonically guided retrieval of occult IUD in early pregnancy. Acta Obstet Gynecol Scand 1991; 70: 355–357. 20. Kirkinen P, Simojoki M, Kivela¨ A, Jouppila P. Ultrasoundcontrolled removal of a dislocated intrauterine device in the first trimester of pregnancy: a report of 26 cases. Ultrasound Obstet Gynecol 1992; 2: 345–348. 21. Sachs BP, Gregory K, McArdle C, Pinshaw A. Removal of retained intrauterine contraceptive devices in pregnancy. Am J Perinatol 1992; 9: 139–141. 22. Holzgreve W, Westendorp A, Gerlach B. Sonographisch kon¨ trollierte Extraktion von Intrauterinpessaren in der Fruhschwangerschaft. Gyne 1988; 5: 47–49. 23. Assaf A, Tagy AE, El Kady AE, Agezy HE. Hysteroscopic removal of copper-containing intrauterine devices with missing tails during pregnancy. Adv Contracept 1988; 4: 131–135. 24. Neis KJ, Brandner P, Otte C. Hysteroscopic removal of lost intra-uterine devices in pregnancy. Gynaecol Endosc 1994; 3: 233–237. 25. Lin JC, Chen YO, Lin BL, Valle RF. Outcome of removal of intrauterine devices with flexible hysteroscopy in early pregnancy. Gynecol Surg 1993; 9: 195–200. 26. Wagner H, Schweppe KW, Kronholz HL, Beller FK. ¨ Moglichkeiten der Extraktion von Intrauterinpessaren bei eingetretener Schwangerschaft. Med Welt 1980; 31: 1317–1320. 27. Ventura SJ, Martin JA, Curtin SC, Mathews MS. Report of Final Natality Statistics, 1996; Monthly Vital Statistics Report (vol. 46, no. 11; Suppl.). National Center for Health Statistics: Hyattsville, MD, 1998. 28. Svigos JM, Robinson JS, Vigneswaran R. Threatened and actual preterm labor including mode of delivery. In High Risk Pregnancy: Management Options (2nd edn), James DK, Steer PJ, Weiner CP, Gonik B (eds). W. B. Saunders: London, 1999; 999–1014. 29. Holzgreve W, Miny P. Chorionic villi sampling with an echogenic catheter: experiences of the first 500 cases. J Perinat Med 1987; 15: 244–250.
