Levonorgestrel only emergency contraceptive use and risk of ectopic pregnancy in Eldoret Kenya: a case-control study

Introduction ectopic pregnancy is one of the causes of maternal morbidity and mortality in sub-Saharan Africa. Levonorgestrel (LNG) only emergency contraceptive pill is a well-established emergency contraceptive pill that is administered within 72 hours after unprotected intercourse. This study aimed at determining whether or not there is a significant association between levonorgestrel emergency contraceptive use and the occurrence of ectopic pregnancy. Methods case-control (1:3) study among 79 women with ectopic pregnancy (cases) matched against 237 women without (controls) at Moi Teaching and Referral Hospital in Eldoret, Kenya; Sociodemographic and clinical data were collected using a questionnaire. Association between ectopic pregnancy and LNG-EC was assessed using Pearson chi-square test. The relationship between outcome and exposure (while adjusting for confounders) was assessed using logistic regression model. Results The mean age was 27.15 years. Both cases and controls were similar by age (p = 0.990), educational level (p = 0.850), marital status (p = 0.559), employment status (p = 0.186) and parity (p = 0.999). Seventy-eight (24.7%) participants had a history of miscarriage. A higher proportion of the cases had history of using LNG-EC compared to the controls (32.9% vs. 7.2%, p < 0.001). The use of LNG-EC portended more than nine times increased odds of ectopic pregnancy compared to other contraceptive methods {OR = 9.34 (95% CI: 3.9 - 16.0)}. Conclusion levonorgestrel only emergency contraceptive use was associated with ectopic pregnancy. One of the limitations of this study is that we could not control for all confounders of ectopic pregnancy.


Introduction
Globally, emergency contraception pills (ECPs) are widely used by women after unprotected intercourse to prevent unwanted pregnancies [1]. In many countries including Kenya; most pharmacies stock Levonorgestrel-only pills for emergency contraception (LNG-EC) and sell them over-the-counter form [2][3][4][5]. LNG-EC can prevent unwanted pregnancies with an efficacy of 52-94% when used within 12 hours of unprotected intercourse [5,6]. Like other contraceptive methods, LNG-EC reduces the chance of pregnancy whether intrauterine or ectopic pregnancy [6]. However, cases of ectopic pregnancy following LNG-EC failure have been reported recently by various researchers in different countries [1,[7][8][9]. Previous systematic reviews have drawn the conclusion that the incidence of ectopic pregnancy following LNG-EC failure was 1.6% and this did not exceed the incidence in general female population [10]. However, most studies on contraception failure and pregnancy as the primary endpoint; did not report whether pregnancies following LNG-EC failure were intrauterine or extra-uterine [11]. Ectopic pregnancy is one of the causes of maternal morbidity and mortality in sub-Saharan Africa [12]. The incidence of ectopic pregnancy in Western Kenya continues to be a major cause of admission in health facilities [13].
Studies on the association between LNG-EC pills and ectopic pregnancy have had conflicting results [14,15]. According to the latest Kenya demographic health survey, the prevalence of contraceptive use among women in their reproductive age is increasing [13]. Increased use of emergency contraception among women is due to increase in demand for prevention of unplanned pregnancies [3,16,17]. Ectopic pregnancy is associated with psychological and physical pain among the affected persons [18,19] and is potentially fatal and threatens the achievement of millennium development goal (MDG) number five on maternal mortality in Kenya [20,21]. Various studies on the link between use of emergency contraception and ectopic pregnancy are inconclusive [1,7,9,[22][23][24]. Therefore, this study aimed to establish whether there is a significant association between LNG-EC emergency contraception and the occurrence of ectopic pregnancy. Understanding the relationship between incidences of ectopic pregnancies and emergency contraception is useful in designing sexual and reproductive health policy. This study can also act as a baseline for other related studies.

Results
The mean age of the participants was 27.  (Table 2); with majority (both cases and controls) taking their first dose between 13 to 24 hours. There was no evidence of association between the timing of LNG-EC use and occurrence of ectopic pregnancy ( Table 2). History of family planning could not be retained in the model due to its association with use of LNG-EC. The use of LNG-EC pills adjusting for history of using depo Provera and menarche was associated with 9.4 times increased odds of developing ectopic pregnancy {OR: 9.34 (95% CI: 3.9 -16.0)} (Table 3).

Discussion
All methods of contraception can effectively reduce the number of intrauterine and ectopic pregnancies [25]. However, in the event of contraception failure, the risk of ectopic pregnancy is significantly increased in the women who become pregnant [26]. Previous studies have indicated that progesterone and its analogue Levonorgestrel, could inhibit human tubal activities which are considered as the main cause of impaired embryo-tubal retention and implantation [27].
Although the cases had similar experiences in history of miscarriages (26.6% vs. 24.1%, p = 0.763); they had a significantly higher proportion of use of family planning methods compared to the controls (58.2% vs. 11.0%, p < 0.001). They also had a significantly higher proportion of use of levonorgestrel-only emergency contraceptive pills compared to the controls (32.9% vs. 7.2%, p < 0.001). This significant risk of ectopic pregnancy following emergency contraception could be attributed to contraception failure as was noted in previous studies. When the use of LNG-EC was adjusted for history of use of depo Provera; it was associated with more than 9.34 times increased odds of developing ectopic pregnancy, OR: 9.34 (95% CI: 3.9 -16.0). This study found a significant risk for the development of ectopic pregnancy among LNG-EC users compared to non-users. This could be attributed to a failure rate of 0.2-3.3% making it less effective in preventing pregnancy compared to other contraceptive methods such as oral contraceptives. Due to the easy accessibility to the Levonorgestrel only pills for emergency contraception, there is an increased uptake. However, there could be women who use it without strictly adhering to the guidelines [3,18].
Studies done in china showed 5 and 4 times increased odds of developing ectopic pregnancy in those who had used Levonorgestrel emergency contraceptive pill compared to non-users respectively [12,13]. Although the use of Levonorgestrel only pills for emergency contraception significantly influenced occurrence of ectopic pregnancy (p < 0.001) among the cases; the findings of this study clearly indicate that occurrence of ectopic pregnancy could also be influenced by the use of other contraceptives such as depo Provera (< 0.001). We established that menarche was associated with more than four times increased odds of occurrence of ectopic pregnancy.
Other studies [28] have reported lack of such association, and further demonstrated that menarche is confounded by the age at coital debut. In our study we were unable to capture this data.

Conclusion
One third of those who presented with ectopic pregnancy used levonorgestrel only emergency contraceptive pill. Levonorgestrel only emergency contraceptive pill use was associated with ectopic pregnancy. Women who use LNG-EC pill should be counselled on the increased risk of developing ectopic pregnancy. One of the limitations of this study is that we could not control for all confounders of ectopic pregnancy.

Acknowledgments
The authors would like to thank the staff at Moi Teaching and Referral Hospital Reproductive Health and Records Departments. Table 1: sociodemographic characteristics   Table 2: comparison of gynecological characteristics between the cases and controls