Ectopic pregnancy in China during 2011–2020: a single-centre retrospective study of 9499 cases

In the present study, the overall EP incidence showed a downward trend in China during 2011–2020, possibly resulting from the popularization of contraception and the successful control of sexually transmitted diseases [9]. Contraceptive measures can reduce EP [10]. From 1980 to 2010, the use of contraceptives in China, including intrauterine devices, oral contraceptives, and condoms, remained at the highest level in the world, e.g., a prevalence rate of as high as 89.20% in 2010 [11], although the contraceptive rate among women aged 15–49 declined from 86.6% in 2014 to 80.6% in 2017 after the implementation of two-child policy [9]. Sexually transmitted diseases, which have been effectively controlled in China, can increase EP [12]. Although syphilis and gonorrhea showed an increasing trend from 2014 to 2019 globally, the incidence rates of syphilis and gonorrhea were 38.37 and 8.45 per 100,000 person-years in China in 2019 [13, 14], which are far lower than the global incidences of 178.48 and 1124.39 per 100,000 person-years, respectively, in 2019 [15]. Incidence of human immunodeficiency virus infection and acquired immune deficiency syndrome has also declined in China over the past two decades, with a prevalence rate of human immunodeficiency virus infection at 39 per 100,000 person-years compared to 476 per 100,000 person-years globally in 2019 [16].

The maternal age of women with EP or delivery was mainly distributed at 25–29 and 30–34 as seen in Table 1. However, the number of women with EP or delivery aged 25–29 declined while those aged 30–34 rose after 2016, resulting in the number of EP in women aged 30–34 slightly exceeding that of EP in women aged 25–29 years by 2020. The progressive increase in the proportion of EP and births in women aged 30–34 years may be related to late marriage and changes in family planning policy. Since the implementation of the two-child policy, the proportion of advanced-age mothers for childbearing has increased significantly [17], as indicated by the increase from 8.52% in 2013 to 15.82% in 2017 in Zhejiang Province, China [18]. In the present study, we found that the proportion of advanced maternal age increased from 2011 to 2018, whereas the proportion of EP among those ≥ 35 years old showed a decreasing tendency from 2011 to 2016, which was subsequently reversed to an increasing tendency from 2016 to 2020 after the release of the two-child policy in 2015, despite the absence of a statistical significance for the latter.

After the release of the two-child policy in 2015, the proportion of EP among those ≥ 35 years old was reversed from a downward trend ( 2011–2016 APC − 4.13*) to an upward trend (2016–2020 APC 4.04), which suggests that this policy change may have an impact on EP. It is reported that the two-child policy implementation may be correlated with the decreased contraceptive rate [9], increased gonorrhea and syphilis [13, 14], and increased advanced maternal age, all of which are the possible causes for the increase of advanced EP [6, 17, 19]. Given that the short observation period, limited sample size, inadequate geographical representation and negligence of co-effects by the potential confounders might have compromised our results, which possibly makes the upward trend not statistically significant, the impact of the two-child policy on EP needs further investigation.

A single-centre study in France reported that tubal pregnancy accounted for 95.5%, ovarian pregnancy for 3.2%, intra-abdominal pregnancy for 1.3%, and no cervical and caesarean scar pregnancy (CSP) among the 1800 cases of EP from 1992 to 2001[20]. Our study also showed that tubal pregnancy accounted for the majority of 91.55% in our centre from 2011 to 2020. CSP, which refers to the gestational sac attached to the scar part of the previous uterine incision when the patient is pregnant again after the caesarean section. [21], is one of the potential long-term complications of caesarean section. Harb et al [22] concluded that the prevalence of CSP among pregnant women was about 1.5/10,000 pregnancies in the United Kingdom between 2013 and 2015. Maymon et al [23] reported that CSP accounted for 4.2% of patients with EP in a large tertiary hospital in Israel between 2000 and 2009. In China, Jiao et al [24] reported that CSP accounted for 1.05% of EP cases in Peking Union Medical College Hospital from January 1994 to April 2007, and the ratio for CSP versus total pregnancy was 1:1221. CSP in our hospital accounted for 6.73% of EP, with a ratio versus deliveries at 1:254. The prevalence of CSP in our study was highest among those described in the literature mentioned above, which could be attributed to the higher prevalence of caesarean section. [9], as well as the improvements in diagnostic ultrasound and heightened awareness of the condition by the clinician. In addition, the discrepancy also resulted from the comparison by using different denominators including the number of deliveries or the total number of pregnancies, which biases the results.

Our study revealed a decreasing trend for CSP from 2013 onwards, contributing somewhat to the decline in overall EP. The overall caesarean section rate also showed a decreasing trend from 2011 to 2020, consistent with the results from other studies in which a decreased caesarean section rate was found in all age groups, especially in younger women who had a greater decrease than older ones [25]. In addition, studies have found a decreased caesarean delivery rate in primigravida women [17]. Whereas the two-child policy was implemented in 2015, the increased number of high-risk pregnant women, such as women with multiple pregnancies, recurrent caesarean sections, and women beyond 35 years old, contributed to the resurgence of caesarean section. [17, 25]. This subset of high-risk individuals may be unwilling to experience an additional pregnancy. The decrease in the overall caesarean section rate, especially among young age and primiparous women, might contribute to the decreased trend for CSP.

For the management of tubal pregnancies, 66% were treated surgically and 34% non-surgically in our hospital during 2011–2020. Our data are close to those reported by Taheri [26], in which surgery accounted for 62% and non-surgical treatment for 38%.

The constituent ratio for the patients receiving medical therapy remained comparatively high and increased from 31.22% in 2011 to 36.43% in 2020. This may be partly because of the increasing awareness of pregnancy and the development of medical technology to diagnose ectopic pregnancy much earlier, providing early-stage patients more opportunities to receive medical treatment. It is also possible that some of the patients had less desire to accept the removal of their fallopian tubes because of the loosened childbearing policy and preferred more to medical therapy for later pregnancy. In addition, the financial problem might be another reason to choose non-surgical therapy for a few patients. A relatively high proportion of medical treatment versus surgery may also be due to the referral centre status.

The operation method for tubal pregnancies was principally laparoscopic salpingectomy in our hospital, and laparotomy and laparoscopic salpingostomy was significantly reduced, which is consistent with the report by an Australian team [27]. A reduction in open surgery is encouraged as the development of surgical techniques and the preference for small scars. Laparoscopic salpingostomy is acceptable to gynecologists and patients owing to the existence of a normal contralateral fallopian tube and comparable reproductive outcome compared to laparoscopic salpingectomy, although it brings a risk of bleeding and persistent or recurrent EP [28]. Our data showed a short-time but high-cost hospitalization for surgical treatment compared to non-surgical treatment. Although non-surgical treatment has a lower cost and can avoid the potential dangers of surgery, it needs to be carefully evaluated because approximately 15% of the patients with tubal pregnancy receiving methotrexate due to ectopic obstruction or rupture eventually require surgery [29].

CSP can be treated in a variety of ways, including curettage, surgical removal of the lesion (via abdomen or vagina), medical therapy (local or systemic injection of methotrexate), and expectant management. Almost all CSPs were surgically treated in our hospital, which may be attributed to a large number of the cases that suffered from relatively severe conditions in tertiary hospitals. Curettage was performed in 90% of the cases with pre-evaluation for the risk of major bleeding from the uterine artery before curettage, and laparoscopic or transabdominal local lesion resection and uterine repair were conducted in 10% of the cases. Harb et al. [22] reported that 61% (56/92) of CSP were treated surgically, 16% (15/92) medically, and 23% (21/92) by expectant management; the authors proposed substantial benefits of surgical therapy, including a high success rate, low complication rate, and short post-treatment follow-up. However, there have been no consensus and standard treatment for CSP thus far, and the choice of surgical treatment or non-surgical therapy is mainly based on the patient’s and local medical conditions, which may also require a combination of multiple treatment modalities for some patients.

There are limitations in this observational study. The study was a single-centre, retrospective cross-sectional study, and some clinical and missing data could not be assessed. The number of women enrolled and the geographical area covered in the study are limited, which could result in potential bias. In addition, we cannot establish a cause-effect relationship. Given that our hospital is the largest medical unit of Obstetrics and Gynaecology in Fujian Province, China, serving a diverse urban and rural population and diversified ethnics with a large delivery number of over 20,000 cases annually, it does not adequately represent the circumstances throughout the country. The total number of women at childbearing age should have been used as the denominator for the incidence of EP. Unfortunately, we failed to capture the pregnancy trends that changed over time in this way. Although the ratio/ratio calculation method used in this study has its limitations, live birth or deliveries is a suitable indicator for the comparison of EP when the number of cases enrolled increases with time [4, 30].