Essure® present controversies and 5 years’ learned lessons: a retrospective study with short- and long-term follow-up

During 5 years, 274 women were admitted to hysteroscopy with intention to treat for Essure® delivery (Table 2).

This was not achieved bilaterally in 7.3% of the patients mainly because of obstruction to the delivery catheter progression but also because of non-identified ostium or ostia (Table 3). The minority of patients with non-identified ostium or ostia were using continuous hormonal contraception (either a combined estroprogestative or a progestative-only pill). Device (manufacturer)-related difficulties were not identified. Women who could not be successfully sterilized with Essure® underwent either laparoscopic tubal ligation or salpingectomy (according to the woman’s informed choice). There were no severe acute complications. The only acute complication was one case of vagal reaction, immediately after successful implant delivery (with complete satisfaction at long follow-up).

During this first trimester, there were no cases of fever, documented infection, or pelvic pain requiring further care.

One allergic reaction to nickel manifested shortly after the 3-month follow-up with a generalized dyshidrotic eczema, and this was confirmed by immunoallergology patch test. This woman had no history of allergy to metals, and the situation resolved completely after hysterectomy (which was the decision taken together with patient, considering the surgical risks of a more conservator approach, including that of incomplete device removal).

Considering the patients with a 3-month follow-up which confirmed correct Essure® placement, a total of 249 patients had the implants for more than 1 year before. Not all these patients could be contacted by telephone at several attempts. At the end, 72% of them (n = 179) answered to the satisfaction survey; 88% (n = 158) were completely satisfied and 9% (n = 17) were very satisfied but reported abnormal menstrual bleeding and/or pelvic pain (related or not to menstruation). Two percent (n = 4) were unsatisfied or very unsatisfied and were the only who would not recommend this method of contraception (two of these women had the same symptoms than the other patients but more severe, requiring the reintroduction of hormonal therapy; the other two women had correct micro-insert placement documented by either X-ray and ultrasound or by hysterosalpingography but reported a pregnancy diagnosed at their private medical assistance). Assuming these two cases of pregnancy, the pregnancy rate at 1 year was 1%.

In this satisfaction survey, two women pointed out that they had been denied magnetic resonance. This question was readdressed and clarified during the satisfaction survey. Some women revealed misconceptions about the procedure (relating it to the posterior appearance of an adnexal mass or to abnormal cervical smear), which was elucidated at that moment.

In our experience, the surgeon perceived the procedure as difficult in 11% of the cases (mainly unilateral difficulty), with an unsuccessful micro-insert delivery (either with or without attempt), comparable to the results found in the literature (7.3 versus 1.5% [9], 2.8% [7], 8% [6], 12% [10]). Systematic use of a non-steroidal anti-inflammatory or a spasmolytic before the procedure (except if contraindicated) could possibly contribute to diminish tubal spasm and subsequent obstruction, but this is yet to be probed. Although we had no event of perforation or abdominal migration, these events could be related to tubal cannulation attempts and, therefore, also benefit from tubal spasm prevention. In what concerns continuous hormonal contraception, it is thought to improve visibility and therefore facilitate the implant delivery. Theoretically, continuous progestins, instead of combined hormonal contraception, may be more favorable due to a higher atrophic effect of the endometrium, but this is not evidence-based. In our sample, however, all women with non-identified ostium/ostia were using continuous hormonal contraception.

Besides acute pain and one self-limited vagal reaction, we had no other incidents like perforation, syncope, or infection. In the literature, there is no evidence about acute complication incidence, but our findings and of others support that this procedure is moderately painful even if brief [5‐8]. An incidence of 4.2% of pain chronification (lasting more than 3 months after the procedure) has been estimated [10]. Because our study is retrospective, we could not identify our incidence of chronic pain.

A good adherence to the 3-month confirmatory tests and back-up birth control after hysteroscopic sterilization is important to avoid unintended pregnancies. Remarkably, we had no loss to follow-up or pregnancy events at 3 months, with a correct placement confirmed in 99% of the tests which is analogous to previous publications (99.5% [7] and 99.6% [9]; even if in these series a lower first control compliance rate is reported 90.8 and 96.8%, respectively). Our 3-month follow-up detected failures were due to contrast leakage or micro-insert expulsion. In the second case, we would now opt for another type of sterilization instead of a second attempt. In view of the fact that the implant expulsed had been delivered with no difficulty, that a correct number of coils trailed in the cavity, and the facility at which it was asymptomatically expulsed, we can imagine that there can be a sort of tubal incompetency.

At 1-year follow-up, we had two pregnancies reported (corresponding to a failure rate of 1%), after documented correct micro-inserts localization. Both, incorrect interpretation of those tests or correct interpretation of localization tests but without obstruction confirmation, are plausible explanations. However, it may also be hypothesized that some women may produce less or slower reaction to the implant, accounting for cases of migration or expulsion, or even for cases of pregnancy after the 3-month correct follow-up. In agreement with this hypothesis, higher than expected tubal patency at intervals less than 1 year [11] and cases of pregnancy after the 3-month ideal follow-up have been published [2, 7]. Failure rates of all types of sterilization failures are estimated in 0.9% [12] but there is uncertainty about specific failure rates for Essure® due to either short follow-up studies, discrepancy between study results, or lack of information on follow-up or on accomplished protocol [13, 14].

Complication rate for laparoscopic sterilization has been estimated in 1.6% [15]. In our study, we had no case of infection, perforation, embolus, or death. Considering the case of nickel hypersensitivity, our post-procedure second surgery was 0.4%. We consider that the suspicion of nickel allergy should contraindicate the procedure even if since 2011 this is no longer contraindicated by the Food and Drug Administration and if some authors would recommend further evaluation but nevertheless consider the procedure [7]. In the manufacturer’s official site, it can be read that no test can reliably predict this adverse effect [16].

Like in our study, studies evaluating women satisfaction have constantly found high levels of early and late satisfaction with levels of moderate or lower pain [6‐8]. The most frequent reported secondary adverse effects are pelvic pain (with or without dysmenorrhea) and increased menses bleeding or irregularity [2]. Unfortunately, we could not assess “de novo” pain cases because most of these women were previously under hormonal contraception, which could have masked previous symptoms. Resuming hormonal therapy ameliorated the symptoms, avoiding the need for second surgery for implant removal (guaranteeing a more effective contraception). Others have enlightened the importance to clearly evaluate the complaints that patients relate to Essure®, at the risk of removing the implants without improvement [2, 4]. With reported second surgery rates of 0.4% [7] to 4.3% [4], according to a recent retrospective cohort study, more than half of surgeries post-Essure® placement were related to pain complaints and the majority of them related to subjacent gynecological conditions [4]. We believe that this finding is of great relevance for clinicians who evaluate women with Essure®.

The most important weakness of our study is its retrospective nature. With our findings, we now look forward to proceed in our investigation in a prospective way and to a longer interval of follow-up.

Finally, we believe that the possibility to inform women about misconceptions of long-term effects of the method contributes to a higher patient satisfaction, though this has not been previously studied. Further investigation is needed to identify which benefits can be obtained by premedication and if a different first follow-up control could have higher sensitivity in the detection of the method failure.