Elsevier

Contraception

Volume 90, Issue 2, August 2014, Pages 174-181
Contraception

Original research article
Probability of pregnancy after sterilization: a comparison of hysteroscopic versus laparoscopic sterilization,☆☆

https://doi.org/10.1016/j.contraception.2014.03.010Get rights and content

Abstract

Objective

To compare the expected probability of pregnancy after hysteroscopic versus laparoscopic sterilization based on available data using decision analysis.

Study design

We developed an evidence-based Markov model to estimate the probability of pregnancy over 10 years after three different female sterilization procedures: hysteroscopic, laparoscopic silicone rubber band application and laparoscopic bipolar coagulation. Parameter estimates for procedure success, probability of completing follow-up testing and risk of pregnancy after different sterilization procedures were obtained from published sources.

Results

In the base case analysis at all points in time after the sterilization procedure, the initial and cumulative risk of pregnancy after sterilization is higher in women opting for hysteroscopic than either laparoscopic band or bipolar sterilization. The expected pregnancy rates per 1000 women at 1 year are 57, 7 and 3 for hysteroscopic sterilization, laparoscopic silicone rubber band application and laparoscopic bipolar coagulation, respectively. At 10 years, the cumulative pregnancy rates per 1000 women are 96, 24 and 30, respectively. Sensitivity analyses suggest that the three procedures would have an equivalent pregnancy risk of approximately 80 per 1000 women at 10 years if the probability of successful laparoscopic (band or bipolar) sterilization drops below 90% and successful coil placement on first hysteroscopic attempt increases to 98% or if the probability of undergoing a hysterosalpingogram increases to 100%.

Conclusion

Based on available data, the expected population risk of pregnancy is higher after hysteroscopic than laparoscopic sterilization. Consistent with existing contraceptive classification, future characterization of hysteroscopic sterilization should distinguish “perfect” and “typical” use failure rates.

Implications

Pregnancy probability at 1 year and over 10 years is expected to be higher in women having hysteroscopic as compared to laparoscopic sterilization.

Introduction

Female surgical sterilization is the most popular method of pregnancy prevention worldwide and is the most commonly used method of contraception among women age 35 years and older in the United States (U.S.) [1], [2]. Each year, 345,000 U.S. women undergo sterilization procedures and a total of 10.3 million U.S. women rely on female sterilization for pregnancy prevention [3], [4].

Since the introduction of a hysteroscopic approach in 2001, an increasing number of women are undergoing hysteroscopic sterilization instead of laparoscopic sterilization [5], [6], [7]. Hysteroscopic sterilization has several advantages over laparoscopic sterilization: it avoids abdominal entry, can be performed as an office procedure and may avoid general anesthesia [5]. More than 650,000 hysteroscopic sterilization procedures have been performed worldwide [5]. One U.S. academic center reported that the proportion of interval sterilization performed laparoscopically from 2002 to 2006 decreased by 50% with a corresponding increase in the proportion of procedures performed by hysteroscopic sterilization by 50% [6].

However, hysteroscopic sterilization has limitations as well. The likelihood of achieving successful bilateral coil placement on first attempt varies from 76% to 96% [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22]. In addition, unlike laparoscopic sterilization, hysteroscopic sterilization is not immediately effective; at least 3 months is required for tubal fibrosis and occlusion to occur for the procedure to be effective. During these 3 months, women need to use alternative contraception until they can undergo a post-procedure hysterosalpingogram (HSG) to confirm bilateral tubal blockage [23]. Prior research has shown that some (6–87%) women never return for their HSGs [8], [10], [12], [13], [14], [16], [17], [18], [19], [24] and that blockage does not occur in 5–16% of HSG evaluations 3 months post-procedure [8], [12], [13], [14], [16], [17], [18], [19]. The multiple steps involved in hysteroscopic sterilization, including the 3-month delay in possibly achieving sterilization, can increase the risk of patient non-compliance with this clinical care protocol and subject women to contraceptive failures (unintended pregnancies) during the process [23].

For any new method of contraception or sterilization, the most important variable to scrutinize is effectiveness. Unfortunately, the literature on hysteroscopic sterilization is limited by lack of such data [2], [25], [26], [27]. Most studies of its efficacy have excluded women who failed initial microinsert placement did not return for HSG or who became pregnant before their HSG [23]. Furthermore, there are no prospective studies comparing the effectiveness of hysteroscopic and laparoscopic sterilizations. Most studies that do retrospectively report pregnancies after hysteroscopic sterilization are limited by small study numbers, short follow-up duration, lack of reporting follow-up duration and high loss to follow-up [27], [28].

We sought to gain a more objective and comprehensive understanding of hysteroscopic sterilization outcomes, based on the published literature. In the absence of a prospective study directly comparing short- and long-term probability of pregnancy after hysteroscopic and laparoscopic sterilization procedures, the best way to understand the consequences of the two contraceptive approaches is through a decision analytic model. Decision analysis can account for the complexity of the multi-step process for coil placement and follow-up, can incorporate the variability in clinical outcomes reported in the literature and can compare the expected probability of pregnancy after hysteroscopic and laparoscopic sterilization. Also, by mapping out the clinical pathway of these procedures, it provides a unique opportunity to identify knowledge gaps in the current literature and help set priorities for future research.

Section snippets

Study design

We developed a Markov state-transition model (Fig. 1) to estimate the probability of pregnancy following three sterilization strategies: hysteroscopic, laparoscopic with silicone rubber band application (falope rings) and laparoscopic with bipolar coagulation. Laparoscopic sterilization was chosen as the comparator for the newer hysteroscopic sterilization procedure, since it is the standard of care for interval (not related to pregnancy) female sterilization [2]. Using yearly cycles, the

Results

In the base case analysis at all points in time after the sterilization procedure (ranging from 1 year to 10 years after procedure initiation), the expected cumulative risk of pregnancy after sterilization is higher in women opting for hysteroscopic than laparoscopic sterilization using silicone band application or bipolar coagulation. Pregnancy risk after hysteroscopic sterilization is primarily accrued in the first year after initiating the process. The expected pregnancy rates per 1000 women

Discussion

Based on best data currently available, our model suggests that hysteroscopic sterilization is not as effective as laparoscopic sterilization in preventing pregnancy when the complete clinical pathways of the procedures are considered. Our analysis improves upon prior studies of hysteroscopic sterilization by taking into account uncertainties in successful placement of coils, return for HSG and successful blockage of tubes. Reflecting these real-life circumstances, our base case estimates

Acknowledgment

The authors thank Lisbet Lundsberg, PhD, for her assistance with the literature review.

References (35)

Cited by (0)

Funding: A grant from the Society of Family Planning.

☆☆

Disclosure: The authors report no conflict of interest.

View full text