Anita L. Nelson
Table Of Contents
Anita L. Nelson, MD
IDENTIFICATION OF CANDIDATES FOR IUD USE
MECHANISMS OF ACTION
CLINICAL EVENTS AFFECTING CONTINUATION OF IUD USE
IUD REMOVAL AND MANAGEMENT OF MISSING STRINGS
RETURN TO FERTILITY
Currently there are two types of intrauterine contraceptive devices (IUDs) available in the United States: the copper T 380A intrauterine device (ParaGard T 380A Intrauterine Copper Contraceptive, Ortho Pharmaceutical, Raritan, NJ),1 and the levonorgestrel-releasing intrauterine system (Mirena, Berlex Laboratories, Montville, NJ).2 The copper T 380A intrauterine device (Fig. 1) is a T-shaped device composed of a polyethylene frame measuring 36 mm by 32 mm with 176 mg of electrolytic copper wire wrapped around its vertical stem and two copper sleeves of 68.7 mg of copper, placed on each of the horizontal arms, for a total surface area of 380 ± 23 mm2 of copper. The device has a monofilament polyethylene string tied at the base of the stem to create two tail strings that aid in monitoring device’s position and in its removal. The frame contains barium sulfate to permit radiographic visualization. The copper T 380A IUD is Food and Drug Administration (FDA)-approved for 10 years of use, although clinical studies indicate high efficacy for at least 12 years.3
The levonorgestrel-releasing system is also a T-shaped device with a polyethylene frame measuring 32 mm by 32 mm and containing barium sulfate for radiographic imaging. A reservoir of levonorgestrel mixed with polydimethylsiloxane is molded around the vertical stem of the T. The sleeve is coated with a membrane that regulates the release of levonorgestrel (LNG) directly into the endometrium at a rate of 20 μg per day initially and 14 μg per day at the end of 5 years. The system is approved in the United States for 5 years of use, but the system may protect against pregnancy for at least 7 years.4 Systemic absorption of LNG does occur, but because the daily release is only approximately 10% of the amount in an oral contraceptive containing 150 μg of LNG, the mean plasma concentration is approximately 5% of that seen with combined oral contraceptives. Serum levels of LNG with the levonorgestrel-releasing intrauterine system (LNG-IUS) are also lower than those measured with progestin-only pills or with levonorgestrel implants.
Other IUDs may currently be used by women in America. Inert plastic IUDs, such as the Lippes Loop and the Safety Coil, were used in the United States before 1985. Women with these devices may continue to use them as long as they are at risk for pregnancy, because the duration of use of plastic IUDs is unlimited. Other IUDs are available internationally, and women often obtain them before immigrating to this country. The Multiload copper 375 IUD, Nova-T IUD, stainless steel rings, and other stringless IUDs can be obtained in Asia, and other copper IUDs are also available overseas. Sales and production of the progesterone-releasing IUD (Progestasert System, Alza Pharmaceuticals, a Division of Alza Corporation, Palo Alto, CA), which was approved for 1 year of use, were discontinued in September 2001. The focus of this chapter is on the two IUDs marketed in the United States.
Worldwide, IUDs are used by more than 130 million women. In Sweden and Germany, 15% to 20% of reproductive-aged, contraceptive-using women select the IUD. In the United States, however, less than 1% of women use this effective, convenient, and cost-effective method of reversible birth control. In part, the underutilization of IUDs in the United States is because of the medicolegal issues created by the Dalkon Shield IUD, which had a polyfilament tail that increased the risk of upper genital tract infection. Modern IUD designs have eliminated that infection risk. A recent survey of Fellows of the American College of Obstetricians and Gynecologists revealed that although 95% to 98% of respondents believed that IUDs were safe and effective, 20% of them had not inserted any IUDs within the last year, and only 4% had inserted at least 10 IUDs in the last year.5 This muted professional enthusiasm for—and poor patient utilization of—the method can be attributed to (1) unsubstantiated, hypothetical concerns that the IUD may be acting as an abortifacient (20% of survey respondents thought the copper IUD “caused abortion”); (2) concerns that the IUD causes infection (29% of respondents said the IUD increased the risk of pelvic inflammatory disease (PID) over 10 years by 10% or more) or legal liability (16% of survey respondents were concerned); (3) physician inexperience with the modern IUDs6; and (4) lack of insurance coverage for IUDs despite their demonstrated cost effectiveness.7 Some increase in IUD utilization has occurred in recent years as many of these issues have been addressed, but further efforts are needed.
The copper T-380A IUD is a remarkably effective method of reversible contraception that is intended for intermediate to long-term use. In clinical trials, the typical first-year failure rate is 0.7%, the 10-year cumulative failure rate 2.1% to 2.8%, and extended use to 12 years added no pregnancies.3,8 The typical use 10-year failure rate of the copper T-380A IUD is less than the typical use first-year failure rate of oral contraceptive pills; it is also less that than the 10-year failure rate of some methods of permanent sterilization.9 The first-year failure rate of the LNG-IUS is 0.14% with a 5-year cumulative failure rate of 0.71%. Although labeling indicates that the effective life is 5 years, clinical trials have shown that it can be used for up to 7 years.4 French and colleagues10 reviewed the controlled clinical trials and concluded that there is no evidence to suggest that the LNG-IUS is any more or less effective than IUDs with more than 250 mm3 of copper.
|IDENTIFICATION OF CANDIDATES FOR IUD USE|
Modern IUDs are being used for interval contraception (spacing pregnancies) as well as by women who have completed their families. As such, it is important to select women for whom the IUD will be safe and effective and will preserve future childbearing potential. The recommended patient profile listed in the IUD manufacturers’ brochures is a parous woman with no history of PID who is in a stable, mutually monogamous relationship and at a low risk for contracting sexually transmitted diseases (STDs). An estimated 11 million American women meet these strict criteria.
In reality, however, even more women can be considered for IUD use. Nulliparity has never been an absolute contraindication to IUD use. The World Health Organization (WHO) rates nulliparity as a “condition where advantages of using the method generally outweigh the theoretical or proven risks,” commenting only that women who have never been pregnant have higher IUD expulsion rates.11 The real concern with IUD use in nulliparous women, however, has been a medicolegal one. If a woman with previously unproven fertility is unable to conceive after using an IUD, she may blame the IUD rather than focusing on other possible causes. Recent studies have been reassuring. Hubacher12 found no higher risk of tubal infertility in nulliparous women who had previously used copper IUDs compared to control women with other causes of infertility or to pregnant controls. In this study, tubal infertility was not associated with duration of IUD use, the reason for IUD removal, or gynecologic problems relating to the use of the IUD. The only marker of increased tubal infertility was the presence of chlamydia antibodies. Similarly, Doll and colleagues13 measured the return to fertility in nulliparous women who stopped using different contraceptive methods and found that after 1 year, the barrier method users had the highest fertility rate (54%) followed by IUD users (39%) and oral contraceptive users (37%). However, long-term IUD users (longer than 78 months) had lower fertility rates than short-term users.
WHO11 has also evaluated the proscription against offering IUDs to women with previous PID. If the patient has no known current risk for STDs and has had a pregnancy subsequent to the episode of PID, WHO concluded that her history of PID was a “condition for which there is no restriction for the use of the IUD.” In women who have not demonstrated fertility after their infection, WHO rated IUD use by a woman in a stable relationship at low risk for STDs as “one in which the benefits generally outweigh the risks.”
The most important characteristic that a potential IUD user should possess is that she is involved in a stable, mutually monogamous relationship. Cramer and associates14 reported that primary tubal infertility rates were not higher in past copper IUD users with one partner, but were significantly increased in women who had used a copper IUD and had had multiple sexual partners compared to similar groups of women who were not IUD users. Immunocompetent women who are infected with human immunodeficiency virus (HIV) are also possible candidates. Morrison and colleagues15 showed that HIV-infected women had no statistically significant increased risk of infectious complications or overall complications for 2 years when compared to controls. Other than being at risk for STDs, contraindications to IUD use are relatively few. Table 1 lists the contraindications included in product labeling.
Pregnancy, known or suspected
PID, pelvic inflammatory disease; AIDS, acquired immune deficiency disease; IV, intravenous; IUD, intrauterine device; LNG-IUS, levenorgestrel-releasing intrauterine system.
Selection between devices can be individualized on the basis of a woman’s fertility plans, her medical conditions, and physical findings. Often, the copper IUD is selected because of its longer duration of action. The copper IUD has been demonstrated to reduce the risk of ectopic pregnancy; therefore, a history of ectopic pregnancy is not a contraindication for the use of the copper IUD. However, labeling for the LNG-IUS indicates that that system should not be used by women with previous ectopic pregnancies or with risk factors for ectopic pregnancy, although study data do not support such a restriction. The copper IUD is contraindicated in women with copper allergy or Wilson’s disease, and the LNG-IUS is contraindicated for women who are allergic to any of its ingredients.
The major factor in matching a particular woman to one of the available IUDs is the impact each IUD may have on a woman’s menses. The copper IUD increases monthly menstrual blood loss by about 30%. On the other hand, the impact of the LNG-IUS varies over time. During the first 3 to 4 months of use, women experience considerable spotting and bleeding, but later they experience a significant reduction in blood loss compared to their baseline. Women who are looking for long-term contraception, who do not tolerate irregular bleeding, and those who do not tolerate hormonal side effects would be better served by using the copper IUD. Women who have anemia, menorrhagia, or dysmenorrhea may not be as well suited for use of the copper IUD as they would be for the LNG-IUS. Cost should also be considered. Even though it is approved for twice as long an effective useful life as the LNG-IUS, the copper IUD is less expensive.
|MECHANISMS OF ACTION|
The details of all the mechanisms of action of each of the IUDs have not been completely evaluated, but several experiments have unequivocally demonstrated that the IUD does not function as an abortifacient by disrupting an implanted pregnancy. Segal and coworkers16 reported the absence of any detectable postovulatory human chorionic gonadotropin (hCG) in the sera of 30 IUD users during 30 months of use. Wilcox and colleagues17 found one episode of a transient elevation in hCG in the urine tests done during 107 IUD cycles compared to a substantial increase seen in 29% of cycles in women attempting pregnancy.
The primary action of the older, inert IUDs was the establishment of an intense foreign-body reaction within the endometrium, which is spermicidal.18 Tredway and associates19 found that 15 to 30 minutes after artificial insemination, sperm was present in all of the tubal flushings of control women but in none of the tubal flushings of IUD users. Sagiroglu20 found that 2 to 16 hours after sexual intercourse, spermatozoa were phagocytized in the endometrial cavities of women with Lippes loop IUDs; no viable sperm were found there 18 hours after artificial insemination. Both El-Habashi and associates21 and Moyer and colleagues22 reported finding no sperm in the fallopian tubes of women with Lippes loop IUDs.
Copper intensifies this inflammatory endometrial response.9,23 In addition, the copper ions released from the IUD directly reduce forward motility of sperm and block fertilization by interfering with acrosomal enzyme activation and by increasing intraperitoneal prostaglandin levels.9,23,24 Uterine washings yielded eggs in 4 of 115 controls, but in none of the 56 IUD users. Ova were less frequently recovered by flushing from the fallopian tubes of IUD users versus those of controls (39% versus 56%). Furthermore, when the ova recovered from women who had sexual intercourse in the fertile period were examined under light microscopy, none from the IUD group had normally developing ova suggestive of successful fertilization, whereas 50% of the eggs obtained from controls demonstrated normal postfertilization development.25 Detailed evaluation of the ova by electron microscopy showed that none of the IUD users had fertilized eggs.18 Finally, IUD users have a significant reduction in their ectopic pregnancy rates, which implies that IUDs inhibit fertilization.
The LNG-IUS induces a milder inflammatory action. The progestin thickens cervical mucus rapidly to prevent sperm entry into the upper genital tract, especially in women who become amenorrheic.26 The LNG also slows tubal motility and reduces the ability of sperm to fertilize eggs. In a small percent of cycles, ovulation is blocked. Mandelin and coworkers27 reported that all LNG-IUS users had midcycle endometrial expression of glycodelin A, a uterine glycoprotein that inhibits sperm-egg binding and usually is absent during ovulation. The LNG also eventually induces endometrial atrophy, but this impact may be more significant to understanding the impact the LNG-IUS has on bleeding patterns than contributing to its mechanism of action.28
When used as directed, all modern methods of birth control are safer than pregnancy. In every age group from 15 to 44, the IUD has the lowest estimated mortality rates of any method of female contraception.29 The Centers for Disease Control (CDC) report that there have been no reported cases of maternal death as a result of a septic abortion in a pregnant woman with an in situ IUD since 1977.30 All previous reports of life-threatening septic abortion occurred in women with Dalkon Shield IUDs; there is no evidence that either of the modern IUDs increases the risk of septic abortion if pregnancy occurs. After years of monitoring clinical events associated with the modern IUDs, the CDC has concluded: “women desiring long-term, effective contraception and their clinicians should be aware that currently marketed IUDs are highly effective and acceptable and are associated with a low risk for complications in women at low risk for STDs.”31
In a 7-year prospective international study of IUDs, including the copper T-380A, use of the IUD for more than 5 years did not increase the risk of pelvic infection, ectopic pregnancy, anemia, or abnormal Pap smears.32 Duration of IUD use did not affect subsequent pregnancy rates or pregnancy outcome. In the short term, however, there are acknowledged risks that must be discussed with the patient and minimized by careful patient selection and fastidious insertion technique (see below).
The advantages of intrauterine contraception include: it is highly effective; it is convenient; it is cost effective; there is high user satisfaction; it is long lasting; and it is rapidly reversible.
Every patient considering contraception needs to be informed about all of her birth control options and of the risks and benefits of each method tailored to her specific circumstances. Bachman and colleagues33 showed that women who received information about potential side effects prior to insertion were most satisfied with their use of the IUS. It is strongly recommended that each patient sign a formal consent before IUD insertion. The manufacturers’ brochures are useful in providing specific information about each of the devices, in identifying contraindications, and in documenting the informed consent process for the medical record. Those brochures are also an excellent source of information for the patient not only before she has an IUD inserted, but also on an ongoing basis to guide her in follow-up care.
Timing of Insertion
Product labeling for the LNG-IUS advises that that system should be inserted during the first 5 to 7 days of a normal cycle. This early insertion allows time for adequate hormone release to thicken the cervical mucus prior to ovulation.
On the other hand, the copper IUD may be inserted at any time during the normal menstrual cycle when pregnancy can be excluded. Traditionally, copper IUDs had been inserted during menstruation, but White and colleagues34 demonstrated that expulsion rates in the first 2 months are highest when insertion is performed during menses. Jovanovic and coworkers35 found that the optimal time to insert a hormonal IUD is during ovulation, when the cervical os is dilated and the myometrium is calmed by progesterone, but a significant reduction in early expulsion rates can be achieved by delaying insertion until after cycle day 5. Jovanovic and coworkers35 found that infection was also reduced by avoiding insertion during menses.
Women with oligomenorrhea or amenorrhea as a result of breastfeeding, medroxyprogesterone acetate, or other drug use do not need to await menses for insertion of either type of IUD. Once pregnancy is ruled out, the IUD insertion can proceed. Although a recent Swedish survey by Andersson and colleagues36 raised concerns about an increased perforation rate when IUDs are inserted in breastfeeding women, a large-scale study by Chi and colleagues37,38 showed that breastfeeding women appear to be at no higher risk for IUD perforation or pain and are often good IUD candidates. Women who desire removal and reinsertion of another IUD can have those procedures done at the same time any day in the menstrual cycle. Postpartum women in many countries have received IUD insertion immediately after delivery (within 10 to 15 minutes after the delivery of the placenta),39,40 but an earlier WHO-sponsored multicenter study concluded that this practice was associated with higher risk of expulsion and pregnancy.41 In routine situations in the United States, an IUD is inserted after a term pregnancy only after the uterus is completely involuted (at least 4 weeks postpartum). After first-trimester losses or termination, however, immediate IUD insertion is appropriate if there is no evidence of genital tract infection. A recent review by Grimes42 of randomized, controlled studies of immediate postabortal insertion showed that the risk of expulsion and infection were low after first trimester losses; the expulsion rates were higher after second trimester abortions.
Patient Evaluation Before Insertion
On physical examination, it is important to assess uterine size, position, contour, mobility, and tenderness. An immobile or markedly angulated uterus is more susceptible to perforation. A Pap smear should be obtained if the patient does not have documentation of normal results within the last year. She should be clinically assessed for vaginal and cervical infection. An acidic vaginal pH can rule out trichomoniasis and bacterial vaginosis. For those with an abnormal discharge or alkalotic pH, microscopic studies are needed. Because trichomoniasis is usually sexually transmitted, that diagnosis necessitates re-evaluation of the patient’s candidacy for an IUD. Bacterial vaginosis increases infectious complications of intrauterine procedures, so it should be treated systemically before IUD placement. Cervical infection with either gonorrhea or chlamydia would preclude the use of an IUD. In established patients thought to be at low risk for STDs, clinical evaluation of the cervix during the pelvic examination is acceptable to rule out such infections. In new patients or those with questionable risk factors, however, formal laboratory testing may be advisable.
Antibiotic Prophylaxis and Premedication
Routine antibiotic prophylaxis before insertion is not warranted.43 One large randomized, placebo-controlled trial of azithromycin prophylaxis showed that the incidence of PID is very rare (1/1000 insertions) with or without antibiotics in women selected for IUD use following current medical and social screening protocols.44 Only women with cardiac lesions defined by the American Heart Association as putting them at high risk for spontaneous bacterial endocarditis (i.e., those with previous endocarditis, rheumatic heart disease, or prosthetic heart valves) require antibiotic prophylaxis before IUD insertion or removal.45 Nonsteroidal anti-inflammatory drugs (NSAIDs) can be offered 1 hour before insertion to women who have no contraindications to their use; these agents reduce the discomfort caused by the procedure, and increase patient satisfaction. Some investigators have found that pretreatment of women with potentially narrow endocervical canals with misoprostol 400 mg per vagina 4 hours prior to insertion facilitates a more comfortable insertion, but formal follow-up studies of expulsion rates in these patients have not been conducted.
Each IUD has its own loading device and insertion technique, which is described clearly in the manufacturer’s materials. Supervised instruction and experience with each type of IUD is advised before independent insertion. The following general recommendations are germane to both IUDs:
Potential Insertion Complications
Vasovagal reactions occur in fewer than 1% of women during the uterine sounding or during IUD placement. The risk of such reactions is higher in nulliparous women and cases where there is moderate to severe pain on insertion.47 If symptoms do not spontaneously resolve promptly, removal of the instrument will be therapeutic. It is important to rule out other problems, such as perforation, which could present with vasovagal symptoms. A paracervical block may be helpful in preventing recurrence in future cervical manipulations.
Uterine perforation is a relatively rare event, occurring in the United States in only 0.6 in 1000 insertions with the copper IUD. All perforations occur or are initiated at the time of insertion. Complete perforation at the time of insertion should be suspected if there is a sudden, inappropriate loss of resistance to advancement of the instrument, when the uterine depth greatly exceeds uterine size, and when the patient reports severe pain with insertion. Partial uterine perforation, which can ultimately result in complete perforation, occurs when one piece of the IUD is introduced into the myometrium. Three important variables have been identified that increase the risk of perforation: (1) the shape, position, and mobility of the uterus; (2) the technique of insertion; and (3) the experience of the provider. Fixed and extremely angulated uteri are more susceptible to perforation, and health care providers who are less experienced with intrauterine procedures may inadvertently use excessive pressure, especially when advancing instruments through the internal cervical os. Fortunately, the plunging technique used to insert older IUDs (e.g., Lippes Loop) has been replaced by the withdrawal technique of insertion, a more controlled procedure that reduces the risk of uterine perforation.48
However, if there are any suspicions of perforation, the instrument and/or IUD should be gently withdrawn without delay.49 Fortunately, most acute perforations occur transcervically or in the midline of the posterior wall of the uterus, at a distance from the uterine arteries. After suspected uncomplicated perforation, the patient should be closely observed for at least a few hours for signs or symptoms of internal hemorrhage. If there is any suspicion of internal organ damage (involving the bowel or vascular structures), surgical exploration via laparoscopy or laparotomy may be indicated. Andersson and co-workers36 found that most perforations (72%) were diagnosed more than 1 month after insertion. The occurrence of an unexpected pregnancy led to the diagnosis in the majority (56%) of these cases. Most of the remaining cases (42%) were diagnosed as a result of failure to visualize the IUD strings during a routine pelvic examination. If a copper IUD is intraperitoneal, it should usually be removed. Appropriate treatment of the intraperitoneal LNG-IUS is less well defined. Although laparoscopy is the preferred surgical approach, its feasibility is reduced after 5 days with the copper IUD, because that IUD tends to become encased in dense adhesions.48 There are case reports of bowel obstruction and organ perforation caused by intraperitoneal IUDs, but the risks of laparotomy may outweigh the benefits of surgical IUD removal because the overall risk of complications from perforated IUDs is low.49 The decision to operate must be made on a case-by-case basis; however, women with abdominal symptoms definitely require surgery.50
|CLINICAL EVENTS AFFECTING CONTINUATION OF IUD USE|
Not surprisingly, analysis demonstrates that patients who are well informed about potential side effects prior to insertion are most likely to be more satisfied than less well informed women, whether or not they actually experience any problems.33,51 Table 2 summarizes the significant clinical events that led to discontinuation of the copper IUD over the 10-year period of clinical trials. Table 3 displays 5-year cumulative discontinuation rates by cause for the LNG-IUS.
*Rates were calculated by weighing the annual rates by the number of subjects starting each year for each of the Population Council (3536 acceptors) and the World Health Organization (1396 acceptors) trials.
The pregnancy rate in the first year of use in the clinical trial with the copper T-380A was 0.7%. Parous women had lower rates (0.5% to 0.8%), and nulliparous women had slightly higher rates of failure (0.8% to 1.2%). The cumulative 8-year pregnancy rates in various studies ranged from 1.4% to 1.8%. International studies indicate a 10- to 12-year cumulative rate of 2.7% for all women, with no pregnancies occurring after the eighth year of use.3 For the LNG-IUS, the pregnancy rate in the first year of use was 0.14%. The cumulative 5-year pregnancy rate was 0.71%. Meta-analysis shows that there is no significant difference in efficacy between the two types of IUDs.10
If a woman becomes pregnant while using the IUD, it is recommended that the IUD be removed if she is in the first trimester, the IUD tail strings are visible, and the removal can be accomplished without difficulty. Such removal significantly reduces the risk of septic abortion and reduces the risk of spontaneous abortion. However, women who undergo early removal still have a higher risk of preterm delivery (i.e., before 37 weeks’ gestation).53
Women with IUDs without visible tail strings who become pregnant have twice the risk of preterm birth,54,55 but apparently no increased risk for spontaneous abortion. The prevailing recommendation is to provide women information about early signs and symptoms of preterm labor and chorioamnionitis, but not to attempt to remove the IUD in those who desire to continue their pregnancies. Some investigators have reported small series of ultrasound-guided IUD removals56 and of hysteroscopic IUD removals57,58 during the first trimester, but these procedures are still considered experimental.
There is no evidence that IUDs increase the risk of congenital anomalies in children conceived by women using modern IUDs.59 The IUD is generally implanted at the border of the placenta and is physically separated from the fetus by the amniotic sac. Despite the high intrauterine concentration of LNG, in the few pregnancies that have been reported with the LNG-IUS, there is no report of masculinizing effects on the fetus or of any other unexplained congenital anomalies.
The risk of ectopic pregnancy varies by IUD type. Noncontraceptive users have a background ectopic pregnancy rate of 3.0 per 1000 woman-years. The copper IUD significantly reduces a woman’s risk for ectopic pregnancy by a factor of 15, to a rate of 0.2 per 1000 woman-years.60,61 Ectopic pregnancy rates with the LNG-IUS are low. Despite the fact that the IUDs decrease a woman’s overall risk for ectopic pregnancy, if a woman does become pregnant while using an IUD, the risk of an ectopic implantation is approximately 4% to 50% (depending on the type of IUD). Past use of an IUD does not increase a woman’s overall subsequent risk for ectopic pregnancy.62,63 However, Bouyer and colleagues64 found that in IUD users who had an ectopic pregnancy, tubal damage, IUD use for more than 2 years, and a history of problems within the first month after IUD insertion were associated with higher risk of subsequent ectopic pregnancy.
Expulsion rates are highest in the first few months after insertion and depend on insertion technique and timing of insertion within the menstrual cycle.34 Overall, first-year expulsion rates with the copper T-380A are 5.7% among all users and 2.3% among parous women. After the first year, expulsion rates decline. Walsh and coworkers44 found the expulsion rate after 3 months to be 3.5%. Fewer than 10% of the expulsions were complete; most were partial expulsions. Many of these expulsions were clinically silent, so the authors recommended that a routine follow-up visit be scheduled at 3 months to detect this problem and to prevent pregnancies. The 5-year expulsion rate for the LNG-IUS was 5.9%.52 A woman who has expelled one IUD has a 30% chance of expelling a subsequent IUD.65 The current recommendation is to remove IUDs located in the endocervical canal and those that are suspected of having relocated lower in the uterine cavity. If the woman is not pregnant, another IUD can be inserted immediately.
Petta and associates66 concluded that routine ultrasound screening for IUD placement is inappropriate. However, ultrasound can be used to identify the position of an IUD when dislocation is suspected because of elongated strings, thus avoiding unnecessary IUD removals. Faundes and colleagues67 noted that IUDs appear on ultrasound to migrate into the correct position after insertion and move throughout the woman’s cycle, so that if the T is within 1 to 2 cm of the fundus midcycle, it can be assumed to be in the correct position.68
Historically, the risk of infection with IUD use has been a significant concern that resulted in considerable medicolegal activity in the late 1970s and early 1980s, and ultimately led to the virtual loss of IUDs in the United States for many years. With a better perspective today, we can identify the issues relating to serious infection to reduce that risk.
The risk of PID is significantly increased with use of the Dalkon Shield IUD and with use of any IUD by women at risk for STDs. The risk posed by the Dalkon Shield was caused by its polyfilament tail, which facilitated entry of vaginal pathogens into the endometrial cavity by wicking or capillary action through the interstices between the filaments in the tail string.69 Modern IUDs, which have monofilament tails, have eliminated this risk; when monofilament tails are cultured, no growth of either anaerobic or aerobic organisms is observed. Most studies have shown that IUDs with monofilament tails pose no higher risk for PID than stringless IUDs.70,71,72,73
Appropriate selection criteria for IUD candidates and thorough screening for potential infections have significantly reduced the risk of infection among IUD users. In a study by Farley and associates,74 of 23,000 IUD users worldwide, PID risk was inversely related to age and number of live births and varied by region. In the Americas, the rates were the second lowest (1.3/1000 woman-years); China had the lowest rates, with no reported cases of PID among the 4301 Chinese IUD users. Worldwide, Farley and coworkers74 noted that PID rates after 1980 were considerably lower than in previous years. Importantly, Buchan and colleagues75 found no statistically significant increase in the risk of PID in current users of medicated IUDs or in former IUD users from 1981 to 1989 in the Oxford–Family Planning Association study. In the United States, Lee and colleagues76 found that among women with one sexual partner (married or cohabiting), women who used IUDs had no statistically significant increase in the incidence of PID compared to those who used no method of birth control. Tsandis and colleagues77 recently demonstrated that routine culture of removed IUDs and treatment of women with positive cultures are not recommended in women who are asymptomatic for PID.
In women at low risk for STDs, however, there is still a transient increase in the risk for PID due to endometrial contamination at the time of insertion. Older studies reported that an elevated infection risk persisted for 3 to 4 months after insertion.78 More recently, Farley and colleagues74 re-evaluated their data and found that the increased risk for infection was limited to the first 20 days after insertion. The latter findings support the classic observations of Mishell and associates79 that the endometrial cavities of IUD users were contaminated by bacteria at the time of IUD insertion, but they were sterile 30 days after IUD insertion except in women who had developed clinical evidence of infection.
In international studies, Farley and colleagues74 reported that PID developed at a rate of 9.7 per 1000 woman-years during the first 20 days after insertion. Many of these women, however, were not routinely screened for cervical or vaginal infections before insertion. In more recent studies from developed countries, the incidence of PID associated with insertion has been reported to be 1 per 1000 woman-years. No cases of PID were reported in a series of 975 Norwegian women inserted with IUDs.80 Interestingly, these latter authors tested for infection on the day of insertion and found that 5 patients tested positive for chlamydia. The researchers contacted the patients immediately, initiated treatment within 14 days, no cases of clinical PID developed. In the study by Walsh and co-workers44 on antibiotic prophylaxis for insertion of IUDs (primarily the copper T 380A), the absolute incidence of PID in the first 3 months of use was approximately 1 in 1000. They also found that the incidence of removal in the first 3 months for reasons other than expulsion, such as increased bleeding or other possible indications of subclinical infection, was 3.4% to 3.8%. It has been suggested that the rates of infection could be lower with progestin IUDs. In a randomized multicenter European trial, the discontinuation rates for PID were significantly lower in the LNG-IUS users compared to the Nova T users.81,82 However, a study by Sivin and colleagues83 comparing the copper T 380 to the LNG-IUS failed to show any difference in PID rates by IUD type.
If upper genital tract infection develops in an IUD user, she should be treated following CDC guidelines for PID. Most experts would recommend IUD removal after a loading dose of antibiotics to reduce risk of septicemia. If possible, the IUD should be cultured for Actinomyces or its scraping should be examined for cytologic evidence of that organism. If infection with Actinomyces is suspected, penicillin should be added to her antibiotic regimen until the test results are available. Older studies have suggested that removal of the IUD in women with PID may not be mandatory, because the severity of the pelvic infection is not related to the use of an IUD.84 Soderberg and Lindgren85 conducted a small randomized trial of treatment of upper genital tract infection and found that clinical response as measured by erythrocyte-sedimentation rate was not changed by removal of the IUD. However, because of the limited numbers of patients in these older reports, Grimes86 recommends that more evidence is needed to allow routine IUD retention in a woman with PID.
Other Infectious Issues
Actinomyces-like organisms are occasionally detected on routine Pap smears. On careful examination, one study found that the Pap smears of one third of plastic IUD users had sulfur granules indicative of the presence of Actinomyces-like organisms, but copper IUD users had much lower rates of involvement (1% to 6%).87,88,89 On the other hand, Merki-Feld and colleagues90 reported that the Multiload Copper IUD had 20% incidence of Actinomyces-like organisms in Pap smears compared to 2.9% incidence reported in LNG-IUS users. The risk of colonization with Actinomyces increases with the duration of IUD use.
If Actinomyces is detected on the Pap smear of an asymptomatic woman with no signs of upper genital tract disease, it is reasonable to follow her with routine Pap smears and pelvic examinations. The patient should be made aware of her colonization and given PID precautions. Some providers prefer to treat women who have Pap smears suggestive of colonization. Several approaches are acceptable. Women with cytologic evidence of colonization may have cervical cultures to confirm the diagnosis before treatment to reduce the number of women treated, because only 40% of women with Actinomyces-like organisms found on Pap smears are actually colonized. In this approach, treatment with oral penicillin G or tetracycline for 10 to 14 days would be given to those women who test positive. Others have advised that all asymptomatic women whose Pap smears demonstrate the presence of Actinomyces-like organisms be treated with antibiotics and that women whose infections are refractory to treatment undergo IUD removal. This approach reduces the cost of testing, but significantly overtreats unaffected women. Routine IUD removal in the face of cytologic evidence of Actinomyces is not routinely recommended today.
Joesoef and colleagues91 reported that, in a low-risk population, the rate of STDs in IUD users compared to other methods of contraception were no different, but that bacterial vaginitis was more common amount IUD users, even after controlling for age, education, douching practices, and STDs.
Increased menstrual bleeding, often combined with cramping, is the most common medical reason for removing IUDs. On average, copper IUDs increase menstrual flow 35% to 55% above baseline levels, increase menstrual cramps, and extend the duration of menses by 1 to 2 days per cycle. In the first few months after insertion, spotting and cramping can also occur between menses. These changes prompt 5% to 15% of women to discontinue IUD use. Once partial expulsion has been ruled out, treatment with NSAIDs is very effective in reducing these problems.92 NSAIDs should be started at the beginning of menses and continued through the heavy flow days (usually 3 days).
Because of the profound antiproliferative impacts that the LNG-IUS has on endometrial cells,93 on estrogen and progestin receptors in the endometrium,94 and on the basal lamina,95 virtually every woman who uses the LNG-IUS will experience pronounced changes in her menstrual cycling. Women should be counseled to expect more days of spotting and bleeding than bleeding-free days in the early months of use. However, after 4 to 6 months, women experience significant reductions in menstrual blood loss. At 12 months, 20% of women are amenorrheic. The remaining women average 0 to 1 days of bleeding and 1 to 3 days of spotting per month. Overall, at 1 year, women experience an 80% reduction in blood loss compared to baseline. Hemoglobin levels increase and dysmenorrhea decreases at 1 year.
Hormonal Side Effects
Circulating plasma concentrations of LNG with the LNG-IUS are 150 to 200 pg/mL, which is only 10% of the peak concentration of 0.75 mg levonorgestrel-only oral contraceptives and are consistently lower than those seen with LNG implants. Ovarian steroidogenesis is not impacted; estradiol levels remain well within normal range for reproductive aged women, both for amenorrheic LNG users and for those experiencing menstrual bleeding. Other adverse events, such as mastalgia, headache, acne, nausea, nervousness, abdominal pain, vaginitis, back pain, depression, decreased libido, abnormal Pap smears, and sinusitis were reported by at least 5% of women using the LNG-IUS. Most of the hormonally related events were concentrated in the first 3 months of use. Progestin slows atresia of ovarian follicles; 12% of women were diagnosed with persistent functional cysts that generally resolved spontaneously.96 Weight gain attributable to LNG-IUS has not been reported.
|IUD REMOVAL AND MANAGEMENT OF MISSING STRINGS|
Routine removal of IUDs is straightforward. The tail strings are grasped close to the external os with ring forceps or vaginal packing forceps. Gentle traction on the strings generally delivers the IUD with ease. Only women who are at high risk for subacute bacterial endocarditis (SBE) require antibiotic prophylaxis for routine removals.
An international multicenter study found that less than 2% of attempted removals were difficult.97 If a patient presents with missing strings, it is desirable to rule out pregnancy by history and perhaps a pregnancy test. If examination reveals that the strings are curled up in the cervix, a cytobrush can be used to catch and straighten them. Alternatively, an endocervical speculum can open the external os, allowing visualization of the canal. If the strings are not retrievable, ultrasound can localize the IUD to rule out expulsion or perforation. It should be noted that the sonographic appearance of the LNG-IUS is more subtle than that of other IUDs. Generally, it is best to locate the acoustic shadow of the IUS, which will help to define both ends of the device.98 If the device is intrauterine and tail strings cannot be repositioned, alligator forceps can be used by the experienced provider to explore the endometrial cavity with a series of opening, closing, and slight tugging motions to grasp any part of the device to remove it. Real-time ultrasound imaging can also be used to guide the extraction procedure.
Embedded IUDs more frequently occur with prolonged use of certain types of IUDs (e.g., Dalkon Shield, Lippes Loop). The problem can be suspected when the IUD does not deliver with a moderate amount of pulling on the strings or when the strings break with traction on them. In this case, the alligator forceps can be used to grasp the device directly and rock it gently from side to side to release it from the endometrial lining. Embedded IUDs refractory to such manipulations can be removed hysteroscopically.
Patient satisfaction with the IUD is high. Surveys of current contraceptive users early in the 1990s found that IUD users rated their satisfaction higher than users of any other method; 96% of IUD users viewed their method favorably.99 This satisfaction is reflected in the annual continuation rates for the copper IUD (see Table 2). Continuation rates for IUDs in clinical trials were higher than those for oral contraceptives, condoms, or diaphragms. The greatest rate of IUD discontinuation occurred during the first 2 years of use. Every year after that, 80% to 90% of women chose to continue to use the method.
Earlier discontinuation rates are higher for LNG-IUS users, but by 5 years, the continuation rates of the LNG-IUS are essentially equal to those of the copper IUD.
Many of the discontinuation decisions are made early in IUD use. Each of the manufacturers has a program of replacing any unit that is removed or lost within the first 3 months. This helps increase the cost effectiveness of this method.
|RETURN TO FERTILITY|
Fertility returns rapidly after IUD removal, comparable to pregnancy rates after discontinuation of barrier methods or oral contraceptives.100 The cumulative pregnancy rate of the LNG-IUS ranges between 79% and 96% 12 months after removal.101,102,103 Wilson62 reported that first-year pregnancy rates were essentially the same for women who had (copper and inert) IUDs removed in order to conceive and for women who had IUDs removed because of complications (85.8% and 82.8%, respectively), and both were equivalent to expected fertility rates. Both groups had comparably low ectopic pregnancy rates at one year (0.4% and 0.8%). There was no increase in the risk of spontaneous abortion or preterm labor among past IUD users compared to that for the general population.62 These findings were consistent with those of a similar study of former IUD users in Norway.104 Landmark studies in the mid-1980s found that monogamous women who had used only copper IUDs had no increase in the risk of tubal infertility compared to women who had never used IUDs (relative risk [RR], 1.1; confidence interval [CI], 0.5 to 2.7).105 Among women who had had multiple sexual partners, however, copper IUD users had a statistically significant higher risk of tubal infertility compared to those who never used IUDs (RR, 2.8; CI, 1.7 to 4.5; p < .0001).14 These findings reinforce the importance of careful patient selection.
Copper Intrauterine Devices as Postcoital Method of Birth Control
Copper IUDs are extremely effective at preventing pregnancy when inserted within 5 days (120 hours) after unprotected intercourse. Several studies since 1979 have indicated a failure rate of approximately 0.1%,106,107,108 which is significantly lower than that for emergency contraception with the Yuzpe method, which uses oral contraceptive pills. In addition to the extended period for postcoital utilization and the high efficacy, the IUD has the advantage of providing extended contraceptive coverage after insertion. The LNG-IUS is not indicated for use as a postcoital contraceptive agent.109
The copper-bearing and nonmedicated IUDs were found in 6 of 7 studies to reduce the risk of endometrial cancer.110 Hubacher and colleagues110 also analyzed the three articles exploring the possibility that IUDs may reduce the risk of cervical cancer, but none has shown a statistically significant reduction with any of the devices available in the United States.
The noncontraceptive benefits of the LNG-IUS relate primarily to its impacts on the endometrium. Hubacher and colleagues110 also showed that average hemoglobin increases varied from as little as 0.5 g/dL after 2 years to 1.6 g/dL after 5 years. Idiopathic menorrhagia was treated by LNG-IUS in nine studies. In these studies, blood loss was reduced by 79% to 93% by 1 year, and the number of days bleeding in these women was reduced by 50%. Extended use of the LNG-IUS (12 to 13 years) resulted in a 60% rate of amenorrhea.111 In two clinical trials comparing medical management to LNG-IUS use in women awaiting hysterectomy for menorrhagia, 64% to 80% of the LNG-IUS users canceled surgery compared to 9% to 14% of the women assigned to the medical intervention arm.112,113 Nagrani and colleagues114 reported 5-year follow-up of these women; 50% were still using the device and 67.4% avoided surgery. Menorrhagia resulting from uterine leiomyomata was reduced in one clinical study115 and in one case report,116 although the impact on fibroid size was variable. Adenosis-induced menorrhagia was also effectively treated in the study by Fedele and colleagues.117 Overall, Stewart and colleagues118 found in a systematic review of the literature that the LNG-IUS is an effective treatment for menorrhagia and may prove to be cost effective.
Other noncontraceptive benefits of the LNG-IUS reported in small studies are reduction in dysmenorrhea, not only in healthy users, but also in women with endometriosis.119 Endometrial implants in the rectovaginal space shrunk with LNG-IUS use.120 The LNG-IUS has been used as a progestin source for postmenopausal women using hormone therapy.121 Similarly, Gardner and colleagues122 demonstrated that the LNG-IUS protected against endometrial stimulation and polyp formation associated with tamoxifen use. Perino123 and Scarselli124 showed that the LNG-IUS can reverse endometrial hyperplasia. Soderstrom-Anttila and colleagues125 reported the success of using the LNG-IUS in oocyte donors during ovarian stimulation.
3. United Nations Development Progamme/United Nations Population Fund/World Health Organization/World Bank, Special Programme of Research, Development and Research Training in Human Reproduction: Long-term reversible contraception: Twelve years of experience with the TCu380A and TCu220C. Contraception 56:341, 1997
4. Sivin I, Stern J, Coutinho E, et al: Prolonged intrauterine contraception: A seven-year randomized study of the levonorgestrel 20 mcg/day (LNg 20) and the Copper T380 Ag IUDS. Contraception 44:473, 1991
13. Doll H, Vessey M, Painter R: Return of fertility in nulliparous women after discontinuation of the intrauterine device: Comparison with women discontinuing other methods of contraception. Br J Obstet Gynaecol 108:304, 2001
27. Mandelin E, Koistinen H, Koistinen R, et al: Levonorgestrel-releasing intrauterine device-wearing women express contraceptive glycodelin A in endometrium during midcycle: another contraceptive mechanism? Hum Reprod 12:2671, 1997
30. Atrash HK, Frye A, Hogue CJR: Incidence of morbidity and mortality with IUD in situ in the 1980s and 1990s. In Bardin CW, Mishell DR Jr (eds): Proceedings from the Fourth International Conference on IUDs. pp 76, 87 Boston, Butterworth-Heinemann, 1994
32. Sivin I, Stern J: International Committee for Contraceptive Research (ICCR): Health during prolonged use of the levonorgestrel 20 (g/d and the Copper TCu 380Ag intrauterine contraceptive devices: A multicenter study. Fertil Steril 61:70, 1994
38. Chi IC, Wilkens LR, Champion CB, et al: Insertional pain and other IUD insertion-related rare events for breastfeeding and non-breastfeeding women: A decade’s experience in developing countries. Adv Contracept 5:101, 1989
39. Tatum HJ, Beltran RS, Ramos R, et al: Immediate postplacental insertion of GYNE-T 380 and GYNE-T 380 postpartum intrauterine contraceptive devices: Randomized study. Am J Obstet Gynecol 175:1231, 1996
58. Hucke J, Campo RL, Kozlowski P, et al: Experience with hysteroscopy or ultrasound-controlled removal of an intrauterine spiral with no visible thread in early pregnancy. Geburtshilfe Frauenheilkd 51:31, 1991
62. Wilson JC: A prospective New Zealand study of fertility after removal of copper intrauterine devices for conception and because of complications: A four-year study. Am J Obstet Gynecol 160:391, 1989
81. Toivonen J, Luukkainen T, Allonen H: Protective effect of intrauterine release of levonorgestrel on pelvic infection: Three years’ comparative experience of levonorgestrel- and copper-releasing intrauterine devices. Obstet Gynecol 77:261, 1991
83. Sivin I, Stern J, Coutinho E, et al: Prolonged intrauterine contraception: A seven-year randomized study of the levonorgestrel 20 mcg/day (LNg 20) and the Copper T380 Ag IUDS. Contraception 44:473, 1991
90. Merki-Feld GS, Lebeda E, Hogg B, et al: The incidence of actinomyces-like organisms in Papanicolaou-stained smears of copper- and levonorgestrel-releasing intrauterine devices. Contraception 61:365, 2000
93. Rutanen EM, Salmi A, Nyman T: mRNA expression of insulin-like growth factor-I (IGF-I) is suppressed and those of IGF-II and IGF-binding protein-1 are constantly expressed in the endometrium during use of an intrauterine levonorgestrel system. Mol Hum Reprod 3:749, 1997
94. Zhu P, Liu X, Luo H, et al: The effect of a levonorgestrel-releasing intrauterine device on human endometrial oestrogen and progesterone receptors after one year of use. Hum Reprod 14:970, 1999
103. Sivin I, Stern J, Diaz S, et al: Rates and outcomes of planned pregnancy after use of Norplant capsules, Norplant II rods, or levonorgestrel-releasing or copper TCu 380Ag intrauterine contraceptive devices. Am J Obstet Gynecol 166:1208, 1992
111. Ronnerdag M, Odlind V: Health effects of long-term use of the intrauterine levonorgestrel-releasing system. A follow-up study over 12 years of continuous use Acta Obstet Gynecol Scand 78:716, 1999
112. Hurskainen R, Teperi J, Rissanen P, et al: Quality of life and cost-effectiveness of levonorgestrel-releasing intrauterine system versus hysterectomy for treatment of menorrhagia: A randomised trial. Lancet 357:273, 2001
122. Gardner FJ, Konje JC, Abrams KR, et al: Endometrial protection from tamoxifen-stimulated changes by a levonorgestrel-releasing intrauterine system: a randomised controlled trial. Lancet 356:1711, 2000
124. Scarselli G, Tantini C, Colafranceschi M, et al: Levo-norgestrel-nova-T and precancerous lesions of the endometrium. Eur J Gynaecol Oncol 9:284, 1988