Intrauterine contraception: Management of side effects and complications

INTRODUCTION — Intrauterine contraception is highly effective, very safe, and generally well tolerated by most people. As such, it has become an increasingly popular method of contraception. Despite the overall excellent safety profile, side effects and complications can occur at the time of insertion and at different time points following insertion.

This topic will address diagnosis and management of side effects and complications related to intrauterine contraception in a temporal fashion broken down by typical onset of presentation in relation to timing from intrauterine device (IUD) insertion. Issues related to intrauterine contraception types, patient candidates and device selection, as well as the noncomplicated insertion and removal procedure, are reviewed separately.

●(See "Intrauterine contraception: Candidates and device selection".)

●(See "Intrauterine contraception: Insertion and removal".)

In this topic, the terms "woman/en" or "patient(s)" may be used as described in the studies presented. We encourage the reader to consider the specific counseling and treatment needs of transgender and gender-expansive individuals.

IUD NOMENCLATURE — Alternate terms for intrauterine contraception include the intrauterine device (IUD), intrauterine contraceptive (IUC), or intrauterine system (IUS, used in reference to the progestin-containing devices). In this topic, we use the term IUD for all types of intrauterine contraception [1]. A discussion of the various IUD types and selection is presented elsewhere. (See "Intrauterine contraception: Background and device types" and "Intrauterine contraception: Candidates and device selection".)

For this topic, we will use the following abbreviations:

●Copper 380 mm² IUD (US commercial name Paragard)

●Levonorgestrel 52 mg IUD (US commercial names Mirena or Liletta): LNG 52 mg

●Levonorgestrel 19.5 mg IUD (US commercial name Kyleena): LNG 19.5 mg

●Levonorgestrel 13.5 mg IUD (US commercial name Skyla): LNG 13.5 mg

Other IUDs studied in international research are described where relevant below using IUD shape and dose.

PREVALENCE OF COMPLICATIONS — Complications are rare and include expulsion (3 to 6 percent in the first year), pelvic inflammatory disease (PID) (1 percent in first 20 days, and 0.5 percent in first 3 to 6 months), contraceptive failure (0.1 to 0.6 percent in first year) with increased risk of ectopic pregnancy if failure does occur, and perforation (0.01 percent) [2-10]. Malposition is typically of less significance and occurs in up to 10 percent of patients. The common side effects of pain and irregular bleeding are the most frequently cited reasons for IUD removal within the first 6 months (28 to 35 percent and 10 to 19 percent, respectively) [11]. Thus, recognition and management of these concerns is important for improving patient satisfaction and continuation rates.

RISK FACTORS FOR COMPLICATIONS — Though IUDs can generally be safely used by most individuals, some populations may be at increased risk of complication. As sub-population patterns of IUD use change, recognition of risk factors for side effects and complications becomes increasingly important.

●Young age – Young age, typically defined as age 25 years or less, has been associated with IUD failure (ie, pregnancy) and likely with IUD expulsion. Young age does not appear to increase risk of other IUD complications such as perforation or infection [12]. The mechanisms are not known.

•IUD failure – Young age has been associated with a higher rate of IUD failure, presumably secondary to increased fertility within this age group. In a retrospective claim-based study including over 90,000 users who received either a copper or levonorgestrel-releasing IUD, people age 15 to 19 years were nearly 30 percent more likely to have an IUD failure compared with people age 20 to 24 years and 40 percent more likely to experience IUD failure compared with users age 25 to 44 years [13].

•IUD expulsion – Although there is conflicting evidence, the majority of studies report that an age of 25 years or less is associated with increased risk of IUD expulsion [14-16]. In a systemic review evaluating IUD expulsion among users aged 25 years or younger compared with older IUD users, six studies suggested increased risk of expulsion with decreased age for the copper 380 mm² IUD. Results regarding expulsion risk with LNG IUD were mixed [17]. The risk of IUD expulsion may be particularly increased for very young individuals. In a planned secondary analysis of a prospective cohort study of over 5400 people who were provided with the IUD (copper or levonorgestrel-releasing) of choice at no cost, subjects age 14 to 19 years had more than twice the risk of IUD expulsion compared with subjects age 20 years and older, regardless of parity and IUD type, at up to 36 months of follow-up [4]. For comparison, during the first year after IUD insertion in nonpostpartum subjects of all ages, expulsion occurs following 3 to 10 percent of copper 380 mm² IUD insertions and 3 to 6 percent of levonorgestrel IUD insertions [3,4,18-21]. Of note, nulliparity, which is more common in younger individuals, does not appear to increase the risk of IUD expulsion [3,4].

●Postpartum or postabortion – Risk of IUD expulsion is increased with both immediately and delayed postpartum IUD insertion and following later abortion [22,23]. There is no increased risk of expulsion with insertion after first-trimester abortion compared to interval insertion [24]. Risk of IUD perforation is slightly increased with postpartum insertion [25]. The use of IUDs for postpartum and postabortion contraception is reviewed separately.

•(See "Contraception: Postpartum counseling and methods".)

•(See "Contraception: Postabortion", section on 'Intrauterine device'.)

●Lactating – While older, smaller studies failed to recognize any association between lactation status at time of IUD insertion and perforation risk [26-28], subsequent larger studies suggest a significantly increased risk of IUD perforation associated with lactation [9,29-31]. A large European prospective cohort study of over 61,000 people undergoing IUD insertion found that lactation status at time of IUD insertion was associated with a sixfold increase (RR 6.1, 95% CI 3.9-9.6) risk in perforation among breastfeeding versus nonbreastfeeding users with no difference between LNG and copper 380 mm² IUDs. The overall risk, however, remained small at approximately 1 percent [9]. A systematic review found that other complications related to IUD insertion including expulsion, pain, and removal requests were similar or lower for lactating IUD users compared with nonlactating users [30]. (See "Contraception: Postpartum counseling and methods", section on 'Uterine perforation'.)

●Distorted uterine cavity – Anatomic abnormalities, including müllerian anomalies or large intra-cavitary leiomyomas that severely distort the uterine cavity, are generally accepted relative contraindications to IUD use [32].

•Uterine anomalies – Though there is biologic plausibility to suggest that conditions associated with uterine cavity distortion could be associated with increased risk of complications, there is limited data to support or challenge this hypothesis. There are case reports of successful IUD use in individuals with müllerian anomalies [33,34]. However, based on the limited evidence available we do not routinely advise IUD use in this population, though we recognize that it could be considered in special circumstances with appropriate counseling. (See "Congenital uterine anomalies: Clinical manifestations and diagnosis", section on 'Diagnostic tools'.)

•Uterine leiomyoma – Given the potential contraceptive and noncontraceptive benefits of IUD use (particularly hormonal IUDs) in individuals with leiomyomas and available safety profile data, we support IUD use in this population, which is consistent with the United States Centers for Disease Control and Prevention statement that the "advantages generally outweigh theoretical or proven risks" [32]. That said, we also discuss with patients that the available data suggest a possible increased risk of IUD expulsion in people with leiomyoma. Two cohort studies included in a systemic review of levonorgestrel IUD use among users with leiomyomas reported higher rates of IUD expulsion among users with leiomyomas (11 percent each) than among users without leiomyomas (0 and 3 percent), though the difference was not statistically significant in one study, and the other study did not test for significance [35]. The remaining six prospective noncomparative studies included in this review reported expulsion rates ranging from 0 to 20 percent. These studies, however, did not specify fibroid location or cavitary impact. There are limited data about other potential complications in this population. (See "Uterine fibroids (leiomyomas): Epidemiology, clinical features, diagnosis, and natural history", section on 'Diagnostic evaluation' and "Congenital uterine anomalies: Clinical manifestations and diagnosis", section on 'Diagnostic tools'.)

●Factors that do not increase risk – Although it has been hypothesized that uterine position or size may increase the risk of IUD complication, there are no scientific data to support this claim. In a large international IUD data set of over 5500 users, retroverted uterine position was not associated with higher rates of failure or expulsion as compared with anteverted or mid-positioned uteri [36]. Similarly, in a case-control study of 71 pregnant users with IUDs in-situ compared with 284 matched controls, uterine position was not associated with an increased risk of IUD failure resulting in pregnancy [37]. Similarly, uterine cavity length does not appear to be a risk factor for IUD failure [37] or expulsion [38,39]. Nulliparity does not affect the success rate for IUD placement [3,4].

TIME OF INSERTION

Insertional difficulties

●Prevention – Techniques to reduce the chances of a difficult insertion include using a tenaculum to straighten the angle of the uterus and cervix and use of a paracervical block to reduce patient pain, the goal of reducing potential patient movement during the procedure.

•Straighten angle of uterus and cervix – In order to minimize risk of perforation or misplacement and to maximize provider control over the movement and angles of the cervix and uterus, a tenaculum should be placed around an adequate amount of tissue to reduce the chance of the tenaculum tearing the cervix.

•Paracervical block to reduce pain – Studies report decreased pain with IUD insertion, particularly for nulliparous patients, if a paracervical block is used [40-42]. Therefore, we recommend offering a para cervical block to nulliparous patients and those at higher risk for insertional difficulties (eg, anxiety regarding IUD insertion, cervical scarring) and using shared decision-making to select block use. (See "Intrauterine contraception: Insertion and removal", section on 'Preprocedure'.)

●Difficult insertion – Insertional difficulties can include challenges with identifying the cervical os (eg, individuals after loop electrosurgical excision procedure [LEEP] or cone biopsy) or with passing the sound through the cervix and internal os (eg, cervical stenosis, tortuous path of canal). Potential solutions include use of an os finder, cervical dilators, and ultrasound guidance. We find ultrasound-guided insertion particularly helpful for individuals who have had a previous difficult insertion, malposition, or expulsion or those with known or suspected distortion of the uterine cavity. These issues are addressed in detail separately. (See "Intrauterine contraception: Insertion and removal", section on 'Difficult insertion'.)

●Vasovagal reaction – Vasovagal reactions can occur after IUD insertion and are characterized by presyncope, syncope, nausea, bradycardia, and hypotension. Most reactions are transient and resolve spontaneously. Trendelenburg position, a cool pack or cloth on the head neck or chest, and an alcohol pad under the nose for distraction and to treat nausea can all be of benefit. In rare circumstances, the vasovagal reaction may be severe enough to require an ammonia inhalant (0.3 mL) as a respiratory stimulant, or intravenous (IV) support with fluids and/or atropine (0.4 to 0.6 mL IV) in order to improve symptoms and vital signs. One-on-one continuous monitoring should be performed in these circumstances. (See "Reflex syncope in adults and adolescents: Clinical presentation and diagnostic evaluation" and "Reflex syncope in adults and adolescents: Treatment".)

Unexpected findings at time of placement

●Large or small uterine cavity – Extremes of uterine cavity size, generally agreed upon as size <6 cm or >9 cm, may create problems for IUD function and retention. The mechanism of action for the copper 380 mm² IUD is largely dependent upon approximation of the device with the uterine wall, and, thus, there is concern for decreased efficacy if this device is placed in a cavity that measures beyond 9 cm. The progestin-containing IUDs have more leeway in their positioning against the entire uterine cavity since the primary mechanism of action for these devices is to thicken cervical mucous and to cause a sterile foreign body reaction that makes the uterine cavity inhospitable to proper sperm function. Therefore, we advise patients whose uterine cavity sounds to >9 cm that the copper 380 mm² IUD may be less effective, though the increased relative risk of pregnancy is not known, and the patient may want to consider an LNG 52 mg IUD or subdermal contraceptive implant instead.

If the uterine cavity appears to measure less than 6 cm when assessed with a sound, it is possible that the sound has not traversed the internal os or across the entire length of the uterine cavity. Ultrasound assessment of the uterine cavity with the sound in place can help distinguish between a small cavity and a sound that is stuck at the level of the internal cervical os. Patients with a small uterine cavity diameter appear to have higher risk for IUD misplacement or migration that leads to an embedded device [43]. These patients are still IUD candidates. If a cavity sounds to 5 to 6 cm and the provider truly believes they are through the internal os and at the fundus, we recommend a smaller-framed IUD, such as the LNG 19.5 mg or LNG 13.5 mg devices. If the provider is unsure that they were at the uterine fundus when they measured a small uterine cavity, ultrasound evaluation should be performed to measure the uterine cavity to help with IUD candidacy determination and IUD size considerations [43].

●Infection – If a pelvic infection is suspected (eg, findings of cervicitis, mucopurulent discharge, cervical motion tenderness, and/or fundal tenderness on examination) at the time of IUD insertion, IUD insertion should be delayed [44]. Testing for sexually transmitted infections is performed in at-risk individuals. (See "Screening for sexually transmitted infections".)

Patients diagnosed with asymptomatic gonorrhea or chlamydia cervicitis on screening at the time of IUD insertion should be treated per standard guidelines (eg, CDC Sexually Transmitted Diseases Treatment Guidelines). The IUD does not need to be removed while the infection is treated [44]. While there are no large studies to help guide practice, if bacterial vaginosis, trichomoniasis, or vaginal candidiasis are diagnosed at the time of IUD insertion, the IUD can be inserted and treatment of the infection is started at the same time.

●Incomplete insertion or partial expulsion – Occasionally, partial IUD extraction or downward misplacement (ie, the IUD stem is visible extruding from the cervix) is noticed immediately upon removal of the IUD applicator. Alternately, the clinician can accidently pull the IUD partially out of the uterus when attempting to trim the IUD strings with the scissors. If downward misplacement or partial expulsion is recognized immediately at time of placement, copper 380 mm² IUDs may be removed, reloaded into the applicator, and reinserted with a second attempt under sterile technique. However, the LNG IUDs cannot be reloaded into the applicator. Although a small study suggests a low-lying IUD can be pushed back into place [45], we do not recommend pushing up a low-lying IUD and advise removing and replacing the device instead.

Perforation — Uterine perforation occurs when an instrument passes through the wall of the uterus and can result in visceral injury, hemorrhage, and infection. Risk factors associated with increased risk of perforation at time of IUD insertion include clinician inexperience, stenotic cervix, and immobile or retroverted uterus. The type of IUD does not appear to impact perforation risk [9]. Being postpartum and/or lactating may independently increase the risk for perforation as described above [9,25,29,30,46-48]. The incidence of perforation is reported in 1/1000, most of which likely happen at time of insertion rather than delayed migration of the device causing perforation [9]. (See "Uterine perforation during gynecologic procedures" and "Contraception: Postpartum counseling and methods", section on 'Uterine perforation'.)

●Clinical presentation – IUD perforation can be difficult to diagnose at time of insertion as pain and bleeding symptoms may not significantly differ from those of women who undergo uncomplicated IUD insertion. Signs and symptoms that should raise suspicion for perforation include:

•Lack of anticipated normal fundal resistance of uterine sound or IUD applicator at time of insertion.

•Hemodynamic instability.

•Severe or unrelenting pelvic pain.

•Active bleeding from cervical os (versus tenaculum sites) on speculum examination.

•IUD string not visible or unusually short (unless IUD placed immediately postpartum).

●Management

•Hemodynamically unstable patients – Patients with unstable vital signs, significantly worsening pain, active, heavy vaginal bleeding, or concern for intraperitoneal bleeding should be transferred to an emergency facility for further management, including possible laparoscopic exploration. Of note, hemodynamic instability is rare after a uterine perforation.

•Hemodynamically stable patients

-IUD not deployed – Patients in whom uterine perforation by the sound or IUD is suspected and are hemodynamically stable, but the IUD has not been deployed, are further monitored for signs and/or symptoms of tachycardia, hypotension, syncope, or intraperitoneal bleeding. Patients with normal vital signs and examination can be sent home after a period of prolonged monitoring. We typically monitor vitals and serial exams every 10 to 15 minutes for a total of three assessments. Patients with unchanged examinations are sent home with teaching on worsening symptoms and symptoms that require emergency evaluation. We then follow up by phone within 48 hours. Patients who become unstable are managed as in the above bullet.

-IUD deployed – Patients in whom uterine perforation is suspected immediately following IUD deployment and who are hemodynamically stable can undergo attempted IUD removal by placing gentle traction on the string. If the IUD is not easily removed with gentle traction, operative laparoscopy should be utilized to either visually monitor and potentially correct uterine trauma while the IUD is pulled out from a vaginal approach or, to allow for abdominal localization, dissection, and removal of an IUD that is primarily in the peritoneal cavity or is involving structures, such as bowel or vessels, that may need repair. Intraoperative x-ray can be helpful to find a peritoneal IUD that is encased in adhesions or otherwise not easily visualized, especially as some peritoneal IUDs are encased in omentum that deviates cranially with intra-operative Trendelenburg positioning. Very rare cases necessitate laparotomy for peritoneal IUD localization, dissection, and removal. Given the low risk of infection and limited evidence to support the practice, we do not routinely give prophylactic antibiotics to patients with suspected uterine perforation at time of IUD placement who do not have clinical signs of infection.

•Unclear IUD deployment – Patients in whom uterine perforation is suspected after the IUD has been deployed but placement cannot be confirmed (eg, IUD strings are not visible in the cervical os after deployment or at routine follow-up visit), are monitored as above. If this occurs at the time of IUD insertion, the IUD applicator is examined to ensure the IUD was indeed deployed and the examination field, including vagina, are inspected to ensure that the IUD did not fall out prior to insertion. If the IUD is not found extra-corporally or it is suspected that intraperitoneal placement has occurred, ultrasound is performed without delay (image 1).

-If the ultrasound reveals a correctly positioned intrauterine IUD, then no further action is necessary.

-If the IUD is not found in the uterus on initial ultrasound evaluation, a kidney, ureter, and bladder (KUB) radiograph should be performed. If the IUD is not seen on KUB, it can be assumed that the IUD was expelled. In this case, a pregnancy test and repeat contraceptive counseling are performed. If the KUB demonstrates an intraperitoneal IUD, surgical planning should be initiated for laparoscopic removal. While the surgical retrieval should be expeditious to prevent further injury, emergency removal is not required if the patient is stable. Laparotomy can be required if laparoscopic removal is not possible, if perforation of intra-abdominal organs is suspected, or if ongoing intraperitoneal hemorrhage is suspected.

●Future IUD insertion – Patients who have a perforation with an initial insertion attempt can have a repeat IUD insertion attempted after several weeks (two to four weeks in the authors' practices) with procedural adjuncts to avoid repeat perforation, such as ultrasound guidance. An alternate contraceptive should be used until the IUD is replaced. Given the relatively small size uterine defect associated with IUD perforation, IUD perforation is not a contraindication to labor or vaginal delivery.

IMMEDIATE POST-INSERTION: 0 TO 7 DAYS — We educate patients that vaginal bleeding and cramping uterine pain are common side effects in the first week after IUD insertion. We also discuss warning sides for pelvic infection, although this complication is uncommon.

Post-insertion bleeding — Minor trauma from IUD insertion, including tenaculum placement and abrasion of the cervix by the IUD, commonly results in post-procedural bleeding. Most patients have at least some degree of spotting or bleeding immediately after IUD insertion, though data specific to initial bleeding patterns after IUD insertion are sparse. A study of 160 users who received the copper T-200 reported that 94 percent of users had bleeding immediately after IUD insertion, that mean duration of initial bleeding was 4.5±2.4 days, and that users required a mean of 3.5±2.5 pads per day [49].

Patient education and reassurance regarding normal bleeding are paramount. We counsel patients that post-procedural bleeding should not be much heavier than a normal period, may be prolonged in timing, and can even blend directly into the prolonged or abnormal uterine bleeding that happens in the first three to six months of IUD use as the uterine lining synchronizes to the presence of a new LNG or copper 380 mm² IUD. We generally do not use prophylactic medication as the abnormal bleeding typically resolves quickly and most users do well with advance education and supportive counseling (figure 1). However, some experts treat patients prophylactically with nonsteroidal anti-inflammatory drugs (NSAIDs) or the antifibrinolytic tranexamic acid to reduce the amount of bleeding associated with initial copper 380 mm² IUD use. Pretreatment with tamoxifen has not been shown to be helpful [50]. (See 'Irregular bleeding and/or cramping' below.)

Patients who report heavy bleeding, commonly described as soaking through a pad or more per hour, should be evaluated quickly, as this could be a sign of partial expulsion, uterine perforation, or cervical laceration.

Uterine cramping or pain — Post-procedural cramping is very common after IUD insertion regardless of IUD type. The mild to moderate discomfort or pain that occurs with insertion lessens within minutes for almost all patients [51-54]. Most patients report some ongoing cramping upon leaving the office, though at levels and patterns significantly less than those experienced peri-insertion. In one study of 59 subjects who underwent IUD insertion, 86 percent required over-the-counter analgesics for cramping symptoms, but only 11 to 13 percent required medication for more than five days [54]. We advise NSAIDs (ie, ibuprofen at doses of 600 to 800 mg every six to eight hours) and heating pads as needed for discomfort. Individuals who develop significantly worsening pain not improved with NSAIDs should be seen for urgent evaluation, as this could be a sign of complication such as partial expulsion, malposition, uterine perforation, or pelvic infection.

Infection and/or pelvic inflammatory disease

●Risk – Historically there have been concerns that IUD use increases the risk of pelvic inflammatory disease (PID). However, extensive epidemiologic study, clinical trials and systematic reviews do not support this concern. While gonorrhea and chlamydia infection are associated with PID, rates of PID are similar among asymptomatic infected patients whether or not they have an IUD. Rates of PID in asymptomatic infected individuals range from 0 to 5 percent, while PID rates among uninfected individuals range from 0 to 2 percent [55,56]. Infection attributable to the insertion process is rare, as demonstrated by the following studies and trial:

•In one international study including over 22,000 users, the risk of infection was just under 1 percent within the first 20 days after insertion and rare thereafter with a cumulative risk of 0.1 percent [7]. Moreover, the risk of PID did not increase with prolonged use. Subsequent large United States studies have estimated the risk of PID after IUD insertion to be 0.5 percent in the first 3 to 6 months [5,6].

•In a United States population screened for gonorrhea and chlamydia either before or at the time of IUD insertion, the risk of PID within the first 90 days after insertion was 1.1 per 1000 women [57].

•In a trial of over 2500 IUD users, LNG 52 mg IUD users had a significantly lower rates of PID than copper 380 mm² IUD users over a three-year period, which suggests the LNG IUDs may have a protective effect against PID [58].

●Impact of HIV infection – HIV infection does not appear to affect risk of infection associated with IUD insertion. A 2016 systemic review of six studies found low rates of PID among IUD users with HIV and no increased rate of infection complications, HIV transmission, viral shedding, or disease progress in this population based on limited evidence [59]. The CDC Medical Eligibility Criteria for Contraception classified IUD use among women with HIV as a category 2 (advantages outweigh potential risks) [60].

●Management – Although PID resulting from IUD insertion is rare, given the possible serious sequelae of PID, clinicians should have a low threshold for empiric treatment of PID in individuals who have recently undergone IUD insertion. The diagnosis and treatment of PID are discussed separately. (See "Pelvic inflammatory disease: Clinical manifestations and diagnosis" and "Pelvic inflammatory disease: Treatment in adults and adolescents".)

•Treatment with IUD in situ – Most individuals with an IUD in situ who are diagnosed with PID do not require IUD removal prior to initiation of antibiotic therapy, regardless of interval since IUD insertion (algorithm 1). In a systematic review that included three randomized trials and a prospective cohort study, one cohort study and one trial reported shorter hospitalization duration for the patients who left their IUDs in place, one trial reported no laboratory differences among individuals treated with or without the IUD in situ, and one trial reported that those who underwent IUD removal during treatment "experienced improved recovery in most clinical signs and symptoms" compared with users who retained their IUDs [61]. Based on these data, the Centers for Disease Control and Prevention (CDC) US Selected Practice Recommendations for Contraceptive Use recommend that individuals with an IUD and PID should be started on appropriate antibiotic therapy with the IUD in place [62].

Patients who are treated for PID and have an IUD in-situ should have close clinical follow up within 48 to 72 hours to ensure improvement. Those with persistent or worsening signs or symptoms should undergo ultrasound to evaluate for tubo-ovarian abscess and initiation of parenteral antibiotic therapy.

•Treatment and IUD removal – IUD removal is performed for patients with PID with known actinomycoses or those who do not respond to therapy in the setting of an in situ IUD. As actinomyces flourishes on foreign bodies, such as the IUD [63], possible actinomyces infection requires prompt removal of the IUD to improve treatment and recovery. (See "Treatment of actinomycosis", section on 'IUD associated infections'.)

Patients who are at high risk for severe infection (eg, immunocompromised) may benefit from IUD removal, although the optimal approach for this patient population is not known.

•Role of IUD culture – The IUD can be sent for microbiologic culture at the time of removal. However, we reserve IUD culture for patients who do not adequately respond to routine antibiotic therapy as there are no data to inform the accuracy or clinical significance of this approach.

●Timing of future IUD insertion – For patients who have been treated for documented intra-uterine infection and had their IUD removed, IUD reinsertion timing should be delayed for 12 weeks [32]. We advise that the IUD insertion be delayed until the treatment course is complete, the symptoms are resolved, and the pelvic examination is normal in the setting of cervicitis alone. Due to the high risk of reinfection, the CDC recommends repeat testing for gonorrhea or chlamydia infection at three months from the original infection [44].

EARLY POST-INSERTION: 0 TO 3 MONTHS

Irregular bleeding and/or cramping — Irregular (and/or heavy) bleeding and cramping are the most common patient concerns in the first few months after IUD placement, regardless of IUD type (figure 1). In a self-reported survey study of IUD users, increased bleeding and cramping was experienced in 15 percent and 32 percent of LNG 52 mg users and 71 percent and 63 percent of copper 380 mm² IUD users at three months after IUD insertion [64]. However, by six months following insertion, 62 percent of LNG 52 mg users reported lighter bleeding and 81 percent of copper 380 mm² IUD users reported lighter or no change in bleeding patterns. Despite the altered bleeding patterns, 94 percent of LNG 52 mg IUD users and 93 percent of copper 380 mm² IUD users reported being somewhat or very satisfied with their IUD method. Counseling regarding expected bleeding changes prior to IUD insertion may improve patients' expectations and satisfaction, largely by selecting only those patients who understand and accept potential inconvenient short-term bleeding. Overall, most patients are reassured to know that bleeding does not indicate a decrease in contraceptive efficacy, that the bleeding is not dangerous, and that the symptoms will likely improve with time. For most individuals, supportive counseling is an adequate intervention. (See "Intrauterine contraception: Background and device types", section on 'Impact of device type on bleeding pattern'.)

However, patients who continue to have heavy and/or prolonged bleeding are evaluated to exclude IUD malposition or expulsion, pregnancy, infection, and cervical dysplasia (see 'Anytime post-insertion' below). We offer medical treatment to patients who continue to have bothersome bleeding from IUD insertion, and we discuss that supporting data for these approaches are mostly lacking.

●Levonorgestrel IUD users – For patients with continued bleeding with an LNG IUD, treatment options include a trial of nonsteroidal anti-inflammatory drugs (NSAIDs), combined estrogen/progestin hormonal contraception, and progestin-only pills (table 1). In the absence of supporting data, we take the following approach based on the risks of undesired side effects, ease of use, availability, and cost.

•NSAID – While we typically treat bothersome bleeding with combined hormonal contraceptives, patients with contraindications to estrogen or who desire a non-hormonal treatment option are offered ibuprofen 800 mg every eight hours, which may reduce bleeding and cramping by inhibiting prostaglandin production. A trial of ibuprofen can be continued for one to three cycles or months. Alternate NSAID options include naproxen 500 mg twice daily for five days or diclofenac 50 mg taken three times daily for five days.

•Combined hormonal contraceptive – For patients who accept combined hormonal contraceptives or for those in whom NSAID therapy is not helpful, contraindicated, or declined, we prefer a combined hormonal contraceptive (oral pill, transdermal patch, or vaginal ring). We start with an initial trial of one to three cycles. Patients who respond well can then stop the medication or continue if desired. As there are no data on the optimal pill for this indication, choice of pill is determined by patient preference, drug availability, and insurance coverage. Theoretically, the use of combined hormonal methods help synchronously thin the endometrial lining.

•Tranexamic acid – For individuals who continue to have bothersome bleeding, we offer tranexamic acid 1300 mg three times daily for up to five days (the medication can be stopped sooner than five days if the bleeding stops). Two studies have supported use of 500 mg or 650 mg doses given three to four times a day up to five days, if these are available [65,66].

•Progestin-only therapy – Lastly, if the bothersome bleeding continues or if the patient is not a candidate for estrogen therapy, the last option is a progestin-only pill such as norethindrone 0.35 mg taken daily for one to three months.

●Copper 380 mm² IUD users – For patients with continued bleeding with a copper 380 mm² IUD, treatment options that have been studied include NSAID medications, combined estrogen-progestin contraception or progestin-only pills, the fibrinolysis inhibitor tranexamic acid, and the antidiuretic desmopressin (table 2) [65].

•NSAIDs – While we typically treat bothersome bleeding with combined hormonal contraceptives, patients with contraindications to estrogen or who desire a non-hormonal treatment option are offered ibuprofen 800 mg every eight hours to reduce bleeding and cramping. Alternate NSAID options include naproxen and diclofenac as for LNG IUDs above.

•Hormonal contraception – For patients who accept hormonal contraceptives or for those in whom NSAID therapy is not helpful, contraindicated, or declined, we offer a trial of a hormonal contraceptive method to be used in combination with the IUD. While we prefer combined hormonal contraceptives (oral pill, transdermal patch, vaginal ring), progestin-only methods can be reasonable used. The choice is based on the patient's preferences and risk factors for estrogen use.

•Tranexamic acid – For individuals who continue to have bothersome bleeding, we offer tranexamic acid 1300 mg three times daily for up to five days (the medication can be stopped sooner than five days if the bleeding stops). Two studies have supported use of 500 mg or 650 mg doses, if these are available [65,66].  

●Supporting data – A trial comparing ibuprofen (400 mg three times) daily with placebo in 28 users starting the copper 380 mm² IUD reported that users receiving ibuprofen noted less menstrual blood loss, although both groups reported increased menstrual bleeding relative to pre-IUD insertion (2 percent increase for ibuprofen users versus 74 percent for placebo users) [67]. In a trial of 175 Chinese individuals that compared two doses of tranexamic acid (500 mg and 1000 mg) twice daily with placebo following IUD insertion, both groups receiving tranexamic acid reported less menstrual blood loss [68]. There was no difference in blood loss between the high- and low-dose groups. In a trial of 19 individuals comparing tranexamic acid (1500 mg three times a day for five days), diclofenac (50 mg three times on the first day followed by 25 mg three times daily for four days, for two periods), and placebo, tranexamic acid was associated with the greatest reduction in menstrual blood flow, although a reduction was also noted with diclofenac [69].

Infection — If an infection such as endometritis or pelvic inflammatory disease is going to occur, the most common time of infection is near the time of insertion. However, infection can rarely develop later. Management of infection in a patient with an IUD is the same regardless of the time point from insertion. (See 'Infection and/or pelvic inflammatory disease' above.)

Problems noted at time of follow-up examination — The authors do not routinely recommend a follow-up visit after IUD insertion as studies have not demonstrated a benefit. However, it is reasonable to offer a visit for IUD surveillance several weeks after insertion for patients who desire as it allows the provider an opportunity to assess the patient's satisfaction with the device and identify any unwanted side effects or complications. The following issues can be identified during a routine post-insertion check or at any time a speculum examination is performed.

●Nonpalpable strings – Patients are no longer routinely advised to check their IUD strings on a regular basis because there is no evidence to support this practice [62,70]. However, some people check strings to confirm the IUD is in place. Given the rare but serious sequelae from an unrecognized IUD perforation or early expulsion, patients with concerns for nonpalpable strings should be seen in a timely fashion for pelvic examination to confirm IUD string presence. If the IUD strings are not visible extruding from the cervix, ultrasound examination is performed to identify the IUD location (image 1). (See 'Perforation' above.)

●Nonvisible strings – If the IUD strings are not visible on speculum examination, a cytobrush can be placed in the cervix and gently twisted in attempt to pull the strings down and out of the cervical os. If these maneuvers are not successful, we then perform an ultrasound to localize the IUD (image 1).

•For individuals whose IUD is confirmed as being in the correct location by ultrasound, no further intervention or surveillance is necessary.

•For individuals whose IUD is not visible in the uterus by ultrasound, we perform additional imaging to exclude possible uterine perforation. (See 'Perforation' above.)

●Expulsion – While IUD expulsion can occur at any time, expulsion is more likely to happen within the first months after insertion. In a study of over 2100 users who underwent IUD insertion, the mean expulsion rates were 1 percent at one month of follow-up and 4 percent at one year [3]. The management of IUD expulsion is discussed in detail below. (See 'Expulsion' below.)

●Perforation – While perforation is frequently diagnosed at time of occurrence, delayed diagnosis can be made if the IUD strings are no longer visible at the time of speculum examination. (See 'Perforation' above.)

●Partner feels strings – On occasion, patients report that their sexual partners can feel the IUD strings at time of intercourse, which can be irritating or painful for the partner. At time of insertion, we suggest cutting the strings at least 2 cm long to leave enough length for the strings to curve around the face of the cervix. With time, the IUD strings typically both soften and curve around along the cervix to project towards the cervical-vaginal fornix instead of along the length of the vagina. If the partner is aware of the strings but not bothered by them, no intervention is needed.

•Long strings – For patients whose IUD strings are bothering a sexual partner, a Scopette or clamp can be used to direct the string ends to the posterior vaginal fornix. Additionally, if there is enough length, the strings can be re-cut with curved scissors to make sure that they are cut on the perpendicular; when the strings are cut on the bias with straight scissors there may be greater risk for partner discomfort.

•Short strings or continued partner discomfort – If the partner continues to be irritated by the strings after the above interventions, the strings can instead be cut flush with the external cervical os to completely minimize the risk of the partner feeling them. Prior to such trimming, the patient should be counseled about the potential need for increased intervention (such as dilation and extraction with graspers or office hysteroscopy) at time of IUD removal should the strings be cut flush with the cervix.

ANYTIME POST-INSERTION

Continued bleeding and cramping — While most patients note a reduction in irregular bleeding and cramping with IUD use over time, some level of continued bleeding and cramping is common, but the patterns vary with the IUD type (figure 1). Individuals with copper 380 mm² IUDs maintain menstrual cyclicity and have reported a 55 to 74 percent increase in the amount of menstrual bleeding, likely due to increased endogenous prostaglandin release in response to the IUD [67,71]. The progestin-containing IUDs are typically associated with decreased bleeding and/or amenorrhea, with 44 percent of LNG 19.5 mg users reporting amenorrhea at six months of use and 50 percent reporting amenorrhea at 12 and 24 months [72]. However, unscheduled or heavy bleeding still occurs in individuals with LNG IUDs. By 12 months of use, 4 percent of LNG 19.5 mg users reported frequent or prolonged bleeding [71]. For patients with either type of IUD who continue to have ongoing heavy bleeding and cramping three months or more from insertion, we evaluate them to exclude IUD malposition or expulsion, pregnancy, and cervical dysplasia. (See "Intrauterine contraception: Background and device types", section on 'Impact of device type on bleeding pattern' and 'Malpositioned IUD' below and 'Pregnancy' below and 'Cervical dysplasia' below.)

Once the IUD position is confirmed and other complications have been excluded, we do the following for individuals who continue to have unacceptable bleeding patterns more than three months after IUD insertion:

●Levonorgestrel IUD users – Patients with LNG IUD-induced heavy or unscheduled bleeding are offered medical therapy with NSAIDs, tranexamic acid, and mifepristone (table 1) [65]. Dosing includes naproxen 500 mg twice daily for five days, tranexamic acid 1300 mg three times a day for up to five days (the medication is stopped when the uterine bleeding stops, typically around four days of treatment), or a single monthly dose of mifepristone 100 mg. We do not prescribe mefenamic, estrogen, or ulipristal acetate because these agents have been associated with either no improvement (mefenamic acid) or increased bleeding (estrogen and ulipristal acetate. The matrix metalloproteinase (MMP) inhibitor doxycycline has been studied as a treatment for heavy or irregular bleeding associated with other progestin-only contraceptive methods, but never in patients using the progestin-containing IUD [73,74]. Possible etiologies of the irregular or prolonged bleeding associated with the progestin-containing IUDs include dilation and fragility of superficial capillaries in the setting of a thinned endometrium, changes in the structural support of the endometrium, and changes in the MMP activity [65].

●Copper 380 mm² IUD users – Patients with copper 380 mm² IUD-induced heavy menstrual bleeding are offered medical therapy, including treatment with nonsteroidal anti-inflammatory drugs (NSAIDs), the fibrinolytic agent tranexamic acid, and the antidiuretic desmopressin (table 2) [65]. Of note, aspirin should not be used for these individuals as it is associated with increased menstrual bleeding in copper 380 mm² IUD users [75]. (See 'Irregular bleeding and/or cramping' above.)

Malpositioned IUD — Malposition refers to IUDs that are displaced into the lower uterine segment, partially expelled into the cervix, rotated on their axis, embedded in the myometrium, or partially or completely protruding through the uterine serosa. The true incidence of malpositioned IUDs is not known, but one review of over 17,000 ultrasound evaluations performed at a single academic institution reported a rate of 10 percent [2]. Malpositioned IUDs are more likely to cause symptoms of pain or abnormal bleeding and the abnormal position is typically diagnosed with ultrasound evaluation [76]. The use of coronal plane evaluation with three dimensional ultrasound may provide improved diagnosis of malposition not otherwise noticed with a two-dimensional longitudinal view (image 2 and image 3).

●Regardless of symptoms, for patients whose malpositioned IUD is partially or completely protruding through the uterine serosa or displaced into the cervical canal so far that the IUD stem can be seen on vaginal examination, the IUD should be removed and replaced, if desired. These malpositions may be associated with an increased risk of contraceptive failure and potential sequelae of the malposition (eg, organ perforation). In a study comparing pregnant and nonpregnant users with a copper 380 mm² IUD in place, the risk of malposition was much higher in the pregnant group (64 versus 11 percent) [77]. However, one important limitation is that it is not known if the malpositioned IUDs resulted in pregnancy or if the growth of the pregnancy contributed to the malpositioning of the IUD.

●For patients who are symptomatic (eg, pain and/or bleeding) and with a malpositioned IUD (ie, those displaced in the lower uterine segment, rotated on their axis, or partially embedded in the myometrium), we remove the IUD and replace it with a new IUD (if desired by the patient), often under ultrasound guidance.

●For asymptomatic patients with an incidentally identified malpositioned IUD that is located within the uterus but does not extend into the cervix (ie, those displaced in the lower uterine segment, rotated on their axis, or partially embedded in the myometrium), we do not routinely recommend IUD removal, as the IUD is likely still highly effective based on mechanism of action, although further research is needed in this area. While the displacement could lessen the contraceptive efficacy, the absolute risk of pregnancy when a progestin or copper IUD is malpositioned but within the uterus appears to be low. Supporting data include:

•In a case control study that evaluated copper 380 mm² IUD location in 25 pregnant patients with IUD in situ compared with 97 patients evaluated post device insertion, pregnant patients had a much higher incidence of cervically located IUDs (52 percent [13 in 25] versus 7 percent [7 in 97]) [78].

•A second study of patients with a copper 380 mm² IUD in place compared 318 early pregnant individuals with 300 control patients and reported a higher incidence of malpositioned IUD in the pregnant individuals (64 versus 11 percent) [77]. However, in both of these studies, it is not clear if the malpositioned IUDs resulted in pregnancy or if the growth pregnancy contributed to the IUDs becoming malpositioned.

Additionally, many IUDs in the lower uterine segment will spontaneously move to a proper fundal location over time [79-81]. Furthermore, given the low rate of highly effective contraception reinitiation after malpositioned IUD removal, removal is associated with a higher pregnancy rate than retention of malpositioned IUD [2]. In keeping with the American College of Obstetricians and Gynecologists (ACOG) committee opinion, we advise appropriate patient counseling and a shared-decision making model [82].

Expulsion — IUD expulsion is when the device is expelled from the uterus. Expulsions can be partial (device protruding from the external cervical os or present in the cervical canal on imaging) or complete (device located in the vagina, not present in the uterus or abdomen on imaging, or with documented patient report of the device being found outside the body) [83].

●Symptoms – Symptoms suggestive of IUD expulsion include cramping, dyspareunia, vaginal discharge or bleeding, or pregnancy and its accompanying symptoms.

●Incidence

•Primary expulsion – The incidence of IUD expulsion during the first year of use is [3,19-21,84]:

-Copper 380 mm² IUD – 2 to 10 percent

-Levonorgestrel IUDs – 2 to 6 percent

•Repeat expulsion – Individuals who have previously expelled an IUD are at higher risk of a repeat expulsion should a new IUD be placed. Rates of repeat IUD expulsion after prior expulsion range from 14 to 31 percent [3,85]. The cause(s) for the increased expulsion risk following a prior expulsion is not clearly known, but possible explanations include anatomical factors (eg, severe flexion), abnormally shaped uterine cavity, patulous internal cervical os, or provider insertional technique that put a subset of women at increased risk for serial expulsion. Interestingly, a prior expulsion may be a risk factor for increased contraceptive failure and resultant pregnancy even if the IUD remains in situ [86].

●Risk factors for IUD expulsion

•Increased risk – Reported risk factors for IUD expulsion include prior expulsion, heavy menstrual bleeding, dysmenorrhea, postpartum status, insertion following later abortion, and prior all-vaginal deliveries [83,87-90]. In a retrospective US health record study including over 228,000 patients with an IUD insertion and no delivery in the prior 52 weeks, the factors most associated with increased risk of IUD expulsion (adjusted analysis) were recent or past heavy menstrual bleeding, excess body weight (overweight or obesity), younger age (highest risk in patients aged 20 years or younger), and parity of four or more [83]. In the study, nearly 81 percent of patients had a levonorgestrel IUD while approximately 19 percent had a copper 380 mm² IUD.

•No impact on risk – Nulliparity, length of the endometrial cavity, and placement after a first-trimester abortion do not increase this risk [3,4,18,38,39,90-93].

●Diagnosis – IUD expulsion is typically confirmed with ultrasound imaging demonstrating an empty uterus. In some cases, the patient will see the device in their underwear or the toilet.

●Repeat IUD insertion – Individuals who have had an IUD expulsion can have a new device inserted if desired and are counseled that repeat expulsion rates of over 14 percent have been reported [3,85]. If a second expulsion occurs and the patient wants a third attempt, we suggest evaluating the uterine cavity prior to placement with hysteroscopy or sonography.

If there is suspicion for uterine fibroids or severe ante- or retroversion, we place the IUD using ultrasound guidance to ensure fundal positioning. It is unknown if changing the IUD type reduces the risk of a repeat expulsion. (See "Intrauterine contraception: Candidates and device selection", section on 'Contraindications'.)

●Reinsertion of partially expulsed device – For patients with a partially expulsed IUD (vertical arm extending partially or fully through the cervical canal), limited data suggest reinsertion may be possible using ultrasound guidance. In a small observational study of 55 patients with a partially expelled IUD, 51 (92.7 percent) had the IUD successfully repositioned under ultrasound guidance and nine (17.6 percent) experienced re-expulsion [45]. The reinsertion procedure did not require antibiotics or paracervical anesthesia. While no injuries or uterine infections were reported in the study, the small sample size limits definitive conclusions. We do not use this approach because of concerns for possible infection and/or device fracture.

Pregnancy — IUDs are highly effective contraceptives with a failure rate comparable to permanent sterilization. A database study including over 18 million hospital deliveries reported a retained IUD was present in 12 per 100,000 births [94]. While the overall risk of pregnancy with an IUD is less than 1 percent, the risk of IUD failure resulting in pregnancy is the highest within the first year after IUD insertion and ranges from 0.2 to 0.8 percent [8].

●Risk factors – Risk factors for IUD failure include young age (possibly secondary to increased fertility), IUD malposition, and prior IUD expulsion (even if the current IUD is in the correct location) [13,26,86]. Other risk factors include grand multiparity, obesity, and previous uterine scar [94]. The risk of pregnancy also varies by IUD type [77]. Uterine position and size are not associated with increased risk of pregnancy [37]. (See "Intrauterine contraception: Background and device types".)

●Ectopic – Although the absolute risk of ectopic pregnancy is lower for individuals using an IUD compared with noncontracepting people or those using other reversible contraceptives [95], patients who become pregnant with an IUD in situ are at a higher risk for ectopic pregnancy compared with the general population [96,97]. Diagnosis and management of ectopic pregnancy is reviewed separately. (See "Ectopic pregnancy: Clinical manifestations and diagnosis".)

•Copper versus LNG IUDs – When pregnancies occur, the likelihood of ectopic implantation appears to be lower among copper IUD users than LNG IUD users. Reported rates of ectopic pregnancy among IUD users who conceive ranges from 11 to 15 percent for copper 380 mm² IUD users to 27 to 53 percent for the LNG 52 mg IUD users [9,21,65].

•Risk by LNG IUD type – In a national database study of over 963,000 females using contraception, the adjusted absolute rate of ectopic pregnancy per 10,000 person-years was 7.7 for LNG 52 mg, 7.1 for LNG 19.5 mg, and 15.7 for LNG 13.5 mg [98]. For comparison, patients using other hormonal contraception experienced 2.4 ectopic pregnancies per 10,000 person years.

•Management – Patients with a pregnancy and an IUD in place should undergo ultrasound assessment of IUD location and whatever evaluation is needed for the ectopic pregnancy. If the IUD needs to be removed as part of the ectopic diagnosis, such as for uterine aspiration, the patient can have a new IUD inserted during or after the ectopic management. Patients whose IUD is found to be malpositioned should have it removed, and a new one inserted as desired. Patients with a normally located IUD can leave the IUD in place despite the contraceptive failure or elect a different IUD or contraceptive method, as desired.

●Intrauterine pregnancy

•Risks – Intrauterine pregnancies conceived with an IUD in-situ are associated with both increased maternal and fetal risks compared with the general population [94]. Maternal risk is primarily related to infection and includes septic abortion and chorioamnionitis. Fetal risks include miscarriage, preterm premature rupture of membranes, preterm birth, fetal malpresentation, and intrauterine fetal demise [94]. Though there is theoretical risk of fetal exposure to levonorgestrel, there does not appear to be increased risk of congenital anomalies associated with a retained LNG IUD [99].

-Chorioamnionitis – People who conceive an intrauterine pregnancy with an IUD in place are at increased risk of chorioamnionitis [100,101]. Removal of IUD early in pregnancy appears to reduce, but not eliminate, this risk. A retrospective study comparing nearly 300 pregnant subjects with a copper 380 mm² IUD in situ with pregnant controls without an IUD reported a rising rate of clinical chorioamnionitis dependent on IUD presence and management (chorioamnionitis in 0.7 percent of pregnant controls, 4 percent in pregnant subjects who had the IUD removed, and 7 percent for pregnant subjects who kept the IUD in situ) [101]. (See "Clinical chorioamnionitis".)

-Miscarriage – The miscarriage rate associated with retained IUD has been reported as high as 47 to 57 percent [102-104]. Early IUD removal or expulsion appears to be associated with decreased risk of miscarriage, ranging from 20 to 54 percent [104]. Furthermore retained IUD also appears to be associated with an increased risk of later (after 12 week) miscarriage versus non-IUD users [100]. (See "Pregnancy loss (miscarriage): Terminology, risk factors, and etiology".)

-Preterm birth – Similar to chorioamnionitis and miscarriage, removal of IUD in early pregnancy appears reduce, but not eliminate, the increased risk of preterm birth associated with an IUD and intrauterine pregnancy. A retrospective study of nearly 200 pregnant subjects with retained copper 380 mm² IUDs (n = 196) compared with pregnant control subjects reported a more than fivefold increased risk of preterm birth for subjects with the retained IUD [100]. In the above retrospective study that assessed chorioamnionitis, preterm birth occurred in 14 percent of subjects who had an early IUD removal and 18 percent of subjects who retained the IUD, compared with a 7 percent rate of preterm birth in pregnant controls [101]. (See "Preterm labor: Clinical findings, diagnostic evaluation, and initial treatment" and "Spontaneous preterm birth: Overview of risk factors and prognosis".)

-Abruption – Concomitant pregnancy and intrauterine IUD has been associated with an increased risk of placental abruption [100,101]. (See "Acute placental abruption: Pathophysiology, clinical features, diagnosis, and consequences" and "Acute placental abruption: Management and long-term prognosis".)

•Management – Management of intrauterine pregnancy in the setting of an IUD depends on pregnancy intentions and IUD location.

-Undesired pregnancy – For individuals who desire termination in setting of in situ IUD, IUD removal should be performed prior to medical termination or at the time of surgical termination. The IUD can be replaced as desired immediately after surgical termination. (See "Contraception: Postabortion".)

-Pregnancy continuation – For patients interested in pregnancy continuation, we remove the IUD as early as possible if the IUD strings are visible or if the IUD in within the cervix, which is consistent with American College of Obstetrics and Gynecology, US Centers for Disease Control and Prevention, and US Food and Drug Administration recommendations [62,82,99]. Given increased risk of infection if the pregnancy is being continued, we give azithromycin 500 mg orally for one dose following IUD removal during pregnancy.

There is limited evidence to guide management for women who desire pregnancy continuation when IUD strings are not visible and the IUD is not in the cervix. The 2017 ACOG Practice Bulletin advises counseling patients about risks of IUD retention but not attempting IUD removal in this situation [105]. Given increased maternal and fetal risks associated with continuing a pregnancy with an IUD in situ (eg, sepsis), we favor a shared decision-making model. If the IUD strings are not visible, the patient desires IUD removal, and IUD is inferior to the pregnancy, it is reasonable to try IUD removal under ultrasound guidance with Bozeman uterine packing forceps, a curved Kelly forceps, IUD hook, or alligator forceps as described below. (See 'Issues and complications at time of removal' below.)

Hysteroscopic IUD removal can also be attempted for patients who desire to continue their pregnancy and have the IUD removed. Successful hysteroscopic IUD removal has been described and is a reasonable option for individuals with low-lying IUDs or IUDs that are distal to an intrauterine pregnancy [106,107]. However, there is insufficient evidence to universally support this practice, as it is unclear if this invasive procedure poses greater or less threat of pregnancy loss than retained IUD.

●IUD replacement – For individuals who become pregnant with an IUD in place, the pregnancy and the IUD can be managed as outlined below, and the patient remains a candidate for a repeat IUD as desired.

IUD discontinuation — Bleeding and cramping are the common reasons cited for copper 380 mm² IUD removal. However, with the LNG IUDs, change in bleeding patterns accounts for approximately one-third of discontinuations while two-thirds are due to systemic side effects, with similar rates reported for the LNG 13.5 mg as compared with the LNG 19.5 mg [108,109]. Demographers from the longitudinal National Survey of Family Growth (NSFG) estimate that the typical individual who continuously uses reversible contraceptive methods from their 15^th to their 45^th birthday will discontinue the contraceptive method nearly 10 times due to method-related reasons [110]. While the rate of IUD discontinuation is lower than that for other contraceptive methods, the rate of IUD removal by United States individuals during their first year of IUD use ranges from 20 (LNG IUDs) to 22 (copper 380 mm² IUD) percent [8].

Hormonal side effects — Common progestin-related complaints include acne, weight change, hirsutism, headache, nausea, and mood changes [58,96]. While the progestin effect of the LNG IUD is primarily local, circulating levonorgestrel from the LNG 19.5 mg can be detected at levels of approximately one-fifth that of individuals using a daily combined oral contraceptive pill, with 150 mcg of levonorgestrel [111-113]. While some individuals can be sensitive to even low levels of hormones, this low levonorgestrel level is generally well tolerated and causes few side effects. In a prospective study comparing weight change over 12 months among people electing progestin-based contraceptives (LNG IUD, etonogestrel implant, and depot medroxyprogesterone acetate) or the copper 380 mm² IUD, there was no difference in the weight changes among the groups in the adjusted analysis [114]. In a Danish national register study that reviewed the records of over one million people, the initiation of an antidepressant was more common among users of progestin-only contraceptives, the LNG IUD included, as compared with nonhormonal contraceptive users [115]. However, it is not known if this association is causal. Lastly, one prospective study of new start IUD users and one cross-sectional analysis of hormonal contraceptive users reported no decrease in sexual desire or sexual function in LNG IUD users [116,117].

Whether the systemic side effects reported by individuals are real or perceived by them to be due to the LNG IUD is often irrelevant. All concerns should be validated, education should be provided, and shared decision-making should be performed. Treatment of side effects while leaving the IUD in place is a potential option depending on the patient, their other contraceptive options or concerns, and the degree to which these symptoms affect their daily life.

Cervical dysplasia — People with an IUD in place are not at increased risk of cervical dysplasia, though benign changes of inflammation and metaplasia are seen more commonly. Intrauterine devices do not increase risk of cervical cancer and may potentially even be protective against cervical dysplasia and malignancy [118]. Conditions that increase the risk for cervical cancer, like human immunodeficiency virus, are not contraindications to IUD use [119]. Routine cervical dysplasia and cancer screening should be performed for patients with IUDs as those without as recommended by the American Society for Colposcopy and Cervical Pathology.

We take the following approach for patients with Pap smear abnormalities:

●Cervix biopsy and/or endocervical curettage – For patients who require a biopsy with Kevorkian forceps, the biopsy can usually be done with the IUD in place without additional maneuvers. Similarly, an endocervical curettage can be done while gently moving the IUD string to minimize the risk of their disruption. We counsel patients about risk of accidental IUD removal or expulsion with ECC.

●LEEP procedure – For patients who require a loop electrosurgical excision procedure (LEEP), the procedure can be done with the aid of one of several maneuvers to keep the IUD in place.

•One option is to cut two separate cervical segments while protecting the strings. With this method we use a Scopette to deviate the strings up to the 12 o'clock position of the cervix and cut the first cervical specimen halfway up the cervix on the posterior aspect of the cervix. Then, similarly deviate the IUD strings towards the 6 o'clock position to allow a second specimen to be cut from the anterior aspect of the cervix.

•A second method is to feed the strings through a hollow tube, such as the sheath of a spinal needle in order to protect the strings while a two specimen LEEP is collected.

•Third, the strings may be gently tucked up into the cervix temporarily with the help of a cytobrush, the LEEP specimen can be collected, and then the IUD strings can be teased back down through the external os.

•Lastly, the strings can be disregarded, the LEEP can be performed cutting the strings with the pass of the loop, and the IUD can be managed moving forward like any other intrauterine IUD with missing strings.

Actinomyces and related organisms — Actinomyces is part of the normal gastrointestinal flora. Limited evidence also suggests that asymptomatic female genital colonization is not uncommon. In a review of over 20,000 Pap smears, Actinomyces-like organisms were seen on 0.26 percent of smears [120]. Up to 7 percent of IUD users will have a Pap smear report documenting Actinomyces-like organisms [63,121]. However, the Pap smear is not the most specific test for Actinomyces; of all the Actinomyces diagnoses that made with Pap smear, only half are truly culture positive when cervical culture is performed [63,122]. While culture is the gold standard for identification of Actinomyces, it is not routinely useful [63].

There are no data for or against performing a cervical culture to confirm diagnosis once Actinomyces is identified on Pap smear in asymptomatic individuals. For those with an IUD who are found to have Actinomyces on a Pap smear, we evaluate for signs and/or symptoms of pelvic inflammatory disease (PID) and base our treatment approach on the findings of that evaluation [105,123]:

●For patients with an IUD who have Actinomyces or Actinomyces-like organism on Pap smear but without signs and/or symptoms of PID, we leave the IUD in place and do not perform a cervical culture [105]. We discuss the findings with the patient and make her aware of the limitations of testing and recommend evaluation if she develops pelvic pain or vaginal discharge.

●If Actinomyces is incidentally noted on Pap smear and the patient has signs of symptoms concerning for PID, we remove the IUD, send the IUD for culture, and obtain a pelvic ultrasound to evaluate for abscess. If Actinomyces is cultured from the IUD, we treat the patient with penicillin in addition to empiric PID treatment, as per United States Centers for Disease Control and Prevention 2015 STD Treatment Guidelines. Choice of antibiotic regimen for treatment of abdominopelvic Actinomyces infection is discussed further in detail elsewhere. (See "Management and complications of tubo-ovarian abscess", section on 'Actinomyces or rare pathogens' and "Abdominal actinomycosis".)

ISSUES AND COMPLICATIONS AT TIME OF REMOVAL

●Nonvisualized strings – IUD strings are not visible at the external cervical os on speculum examination in 4 to 18 percent of patients at the time of IUD check or removal [124-127]. The differential diagnosis for nonvisualized IUD strings on speculum examination includes IUD in situ, unrecognized IUD expulsion, malposition such as IUD embedding in the uterine fundus, and IUD perforation.

For people who desire IUD removal but have nonvisible IUD strings, we take the following approach to identifying the strings:

•First, we attempt to tease the strings out of the external os by inserting and rotating a cytobrush into the endocervix. This technique has a reported efficacy of 98 percent and can be performed regardless of concern for pregnancy or intentions for IUD removal versus continuation [68,69]. While using a colposcope to visualize the strings within the cervix has been reported [128-130], the authors do not find this technique helpful.

•If string retrieval with the cytobrush is not successful, there is no concern for pregnancy, and the patient is able to tolerate continued attempts, string identification (and IUD removal) can be attempted using a Bozeman uterine packing forceps, disposable thread finder, a curved Kelly forceps, IUD hook, or alligator forceps. A cervical block may help keep the patient comfortable during the additional instrumentation of the cervix.

•For patients in whom cervical instrumentation does not identify the strings, and depending on the patient's ability to tolerate cervical manipulation, we place a tenaculum and, if needed, dilate the cervix to 21 french (ie, 7mm) with a Pratt dilator to facilitate opening the forceps to grasp the string. A paracervical block (10 to 20 cc 1% lidocaine or chloroprocaine, either without epinephrine) could be employed to improve success and comfort at initial clinic removal.

•If IUD removal remains unsuccessful, pelvic ultrasound should be performed to confirm IUD location (image 1). Once intrauterine position is confirmed, these methods can be attempted again either in the office or outpatient surgical setting, depending on the patient's needs for sedation. In a prospective study of 29 parous individuals who presented for LNG 52 mg IUD removal without visible strings, 100 percent were able to have their IUD removed in office on first visit using an IUD hook, Kelly clamp, or other blunt grasper, or ring forceps with or without adjuvant ultrasound guidance and/or cervical anesthetic [131]. In a second study of 23 subjects with failed initial IUD removal, 83 percent had successful removal with the addition of ultrasound guidance alone [130].

•For patients in whom the above techniques are not successful in identifying the IUD strings or if ultrasound evaluation reveals an IUD embedded in the uterine wall, hysteroscopic evaluation and removal can be necessary, either in an office or outpatient surgical setting. After appropriate cervical anesthesia and dilation, the operative hysteroscope is passed through the cervix and gently advanced into the uterus until the IUD or strings are visualized. Once visualized, the hysteroscopic grasper is used to grasp the strings (preferable if they are seen) or the device, and the IUD, grasper, and hysteroscope can all be withdrawn concurrently from the cervix.

●Difficult removal – Approaches to remove an IUD include using additional traction, pelvic ultrasonography to confirm IUD location, and hysteroscopic removal.

•IUD strings visible – If IUD strings are visible at the external os, but the device cannot be removed with traditional gentle traction, increased moderate traction can be applied to the IUD strings.

-Broken IUD strings – If the IUD still cannot be removed with moderate traction or if the strings either feel like they are breaking or actually do break off the device, a pelvic ultrasound, preferably with 3D images, is obtained to confirm IUD location (if not already performed) (image 1).

-IUD stem visible – If the IUD stem is visible, a Kelly clamp, or similar, can be used to grasp and remove the IUD.

-Role of hysteroscopic removal – If ultrasound confirms intrauterine position but the IUD is unable to be removed, we perform hysteroscopic removal, similar to patients whose IUD strings are not visible. (See 'Problems noted at time of follow-up examination' above.)

•Suspected myometrial IUD – For IUDs known or suspected on ultrasound to be embedded in the fundal myometrium but not perforating through the serosa, we recommend trial of traditional removal in the office, as a case series study of 157 subjects reported that only 10 percent of all IUDs diagnosed as embedded on preoperative imaging were confirmed to be embedded on hysteroscopy [132]. If moderate traction does not remove the embedded IUD, we perform hysteroscopic removal to ensure complete, atraumatic IUD removal. If the embedded IUD portion looks like it might break with traction, it can be teased or dug out of the myometrium with the closed grasper. Once more, if the IUD is visible, it can be regrasped at a portion distal to the weak point in the device and removed.

●Broken IUD – IUD breakage is rare and limited to case reports. Most cases occurred during IUD removal when the IUD fractured and resulted in a retained arm, but there are also rare cases reporting spontaneous expulsion of a portion of a fractured IUD [133]. There are limited data on the consequences of leaving an IUD fragment in situ, including abnormal bleeding, pain, infertility, and infection. For patients who are symptomatic or who want the fragment removed, the location can be diagnosed with ultrasonography, hysteroscopy, or a kidney, ureter, bladder (KUB) x-ray. Asymptomatic patients are counseled that long term sequela of leaving an IUD fragment in utero has unknown possible future consequences. For patients who desire future fertility, we advise fragment removal if the fragments are visible on imaging or hysteroscopy. Options for removal of retained fragments include manual vacuum aspiration, an IUD hook, narrow tip forceps, and hysteroscopic-assisted extraction [133]. For myometrial IUD fragments that cannot be easily removed with hysteroscopy, one author (KP) advises leaving the fragment in place as it is unlikely to impact fertility, although the data to inform this approach are limited.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Contraception".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Intrauterine devices (IUDs) (The Basics)" and "Patient education: Long-acting methods of birth control (The Basics)" and "Patient education: IUD insertion (The Basics)" and "Patient education: IUD removal (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Adverse events – Serious IUD complications are uncommon and include expulsion, pelvic inflammatory disease (PID), contraceptive failure, with increased risk of ectopic pregnancy if failure does occur, and perforation. Less serious side effects include malposition, pain, and irregular bleeding. (See 'Prevalence of complications' above.)

●Risk factors – Risk factors for IUD complications include young age, postpartum or postabortion insertion, breastfeeding, and a distorted uterine cavity. (See 'Risk factors for complications' above.)

●Impact of symptom timing on range of adverse events – The differential of adverse events associated with IUDs varies based on the timing of the event relative to the timing of the IUD insertion. Some symptoms can occur in multiple time settings (ie, vaginal bleeding).

•At time of IUD insertion – Difficult IUD insertion, vasovagal reaction, unexpected findings such as infection, partial expulsion or extraction of the IUD, and perforation can all be encountered at the time of IUD insertion. (See 'Time of insertion' above.)

•Week following IUD insertion – IUD side effects that commonly occur in the first week following insertion include post-insertion bleeding and uterine cramping or pain. Abnormal bleeding and cramping are typically self-limited and resolve within three months. Very rarely, infection such as pelvic inflammatory disease develops. Most women with an IUD in-situ who are diagnosed with PID do not require IUD removal regardless of interval since IUD insertion (algorithm 1). The one exception is users with known actinomyces PID. As actinomyces flourishes on foreign bodies, such as the IUD, actinomyces PID requires prompt removal of the IUD to improve treatment and recovery. (See 'Immediate post-insertion: 0 to 7 days' above.)

•Early postinsertion – Early postinsertion issues include continued bothersome bleeding, infection, and problems that are noted at the time of a follow-up examination, such as missing IUD strings, IUD expulsion, or perforation. (See 'Early post-insertion: 0 to 3 months' above.)

•At any time following IUD insertion – Events that can occur any time after IUD insertion include continued bothersome bleeding, IUD malposition or expulsion, pregnancy, and hormonal side effects. Because of these issues, some individuals will desire IUD removal. In addition, individuals with an IUD can be diagnosed with cervical dysplasia or actinomyces, which requires additional treatment. (See 'Anytime post-insertion' above.)

-Ectopic pregnancy – Patients with an ectopic pregnancy and an IUD in place undergo ultrasound assessment of IUD location and whatever evaluation is needed for the ectopic pregnancy. (See 'Pregnancy' above.)

-Intrauterine pregnancy – Patients with an intrauterine pregnancy and IUD are advised to have the IUD removed at the time of diagnosis based on the increased risk of poor maternal and fetal outcomes associated with a retained IUD and the reduction of risk associated with IUD removal early in pregnancy. (See 'Pregnancy' above.)

●Challenges with IUD removal – Issues and complications that can occur at the time of IUD removal include nonvisible IUD strings, difficult IUD removals, and, rarely, a broken IUD. These can typically be managed with a variety of techniques in the office setting. In some cases, hysteroscopy can be required to remove a retained, embedded, or broken IUD. (See 'Issues and complications at time of removal' above.)

ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges Gillian Dean, MD, MPH, and Alisa B Goldberg, MD, MPH, who contributed to an earlier version of this topic review.