Disparities in the Incidence of Ectopic Pregnancy in a Large Health Care System in California, 2010−2019

Abstract

Introduction

Ectopic pregnancy leads to reproductive health morbidity, including greater risk of another ectopic pregnancy, infertility, and, in rare cases, mortality. Information on trends in the incidence of ectopic pregnancy in the last decade is limited.

Methods

A population-based cross-sectional study of women aged 15−44 years enrolled at Kaiser Permanente Northern and Southern California 2010−2019 was conducted. Electronic health records were used to identify ectopic pregnancies. The crude ectopic pregnancy incidence per 1000 pregnancies (live births, induced abortions, and ectopic pregnancies) and 95% confidence interval (CI) was estimated per study year, overall, and stratified by age group. The age-adjusted incidence and 95% CI was estimated per study year, overall, and stratified by race/ethnicity. Temporal trend was assessed using Poisson regression.

Results

There were 15,537 ectopic pregnancies among 979,027 pregnancies. The overall age-adjusted ectopic pregnancy incidence was 15.8 per 1000 pregnancies, 95% CI: 15.6, 16.1. The annual incidence increased from 15.2, 95% CI: 14.4, 16.1, in 2010 to 16.4, 95% CI: 15.6, 17.2, in 2019 (p < 0.001). The overall incidence was highest among women aged 40−44 years (24.2, 95% CI: 22.7, 25.6) and non-Hispanic Black women (21.9, 95% CI: 21.0, 22.8); compared to 30–34-year-old (16.2, 95% CI: 15.7, 16.6) and non-Hispanic White (14.6, 95% CI: 14.1, 15.0) women, respectively.

Discussion

The increase in ectopic pregnancy incidence during the studied period was largely driven by increasing incidence in younger women. However, disparities in the incidence by age and race/ethnicity persisted.

Conclusion

Ectopic pregnancy remains a significant source of reproductive health morbidity, especially for older ( >40 years) and non-Hispanic Black women.

Introduction

Ectopic pregnancy is defined as the implantation of a fertilized egg outside the uterus and can be an acute, life-threatening condition in the first trimester of pregnancy. Mortality from ectopic pregnancy has declined significantly in the last two decades and is currently estimated to be less than 0.5 deaths per 100,000 live births. ¹ However, affected women can suffer significant morbidity, including greater risk of another ectopic pregnancy and future infertility. ²

Trends in ectopic pregnancy are difficult to examine, as women with ectopic pregnancies are typically managed in the outpatient setting either medically, with a methotrexate injection, or surgically, with laparoscopy. ^3,4 Therefore, surveillance and efforts to monitor national trends remain arduous. The last year for which nationwide data from the US Centers for Disease Control and Prevention (CDC) are available is 1992, where the estimated total number of ectopic pregnancies was 108,800, with an incidence of 19.7 per 1000 reported pregnancies. ⁵

Electronic health records of patient care, in addition to administrative and claims databases from health systems, have been used to conduct epidemiological studies to investigate trends of reproductive health disorders such as ectopic pregnancy; however, the most recent studies using health system data were conducted over a decade ago. Using data from Kaiser Permanente Northern California, Van Den Eeden et al estimated an ectopic pregnancy incidence of 20.7 per 1000 reported pregnancies from 1997 to 2000. ⁶ Most recently, using data from Group Health Cooperative, a mixed-model, managed health care system in Washington State (now Kaiser Permanente Washington) and western Idaho, Trabert et al estimated an incidence of 15.0 per 1000 pregnancies in 2005–2007. ^7,8

In the last decade, various potential risk factors have been linked to ectopic pregnancies, including increased maternal age and use of assisted reproductive technologies. In addition, potential protective factors, such as increased use of highly effective contraceptive methods, including intrauterine devices and implants, have also been identified. ⁹ Thus, it is important to monitor changes in incidence rates and trends continuously. ¹⁰ The current study aimed to describe the incidence and temporal trends in ectopic pregnancies from 2010 to 2019 by using data from a large integrated health care system in California that serves a community-based, racially and ethnically diverse population.

Materials and Methods

Study population

A population-based cross-sectional study of women aged 15−44 years who were enrolled in a Kaiser Permanente Northern California or Kaiser Permanente Southern California health care system for at least 1 month from January 1, 2010 to December 31, 2019 was conducted. Combined, Kaiser Permanente Northern California and Kaiser Permanente Southern California currently provide care to over 9 million health plan members across the state, representing approximately 47% of the commercially insured (purchased by individuals or employers), and 29% of the publicly insured (Medicare and Medi-Cal managed care) population. ¹¹ Members receive their care almost exclusively from Kaiser Permanente physicians and allied staff in the medical centers and medical office buildings owned or operated by the health plan, of which all utilize an electronic health record based on the Electronic Privacy Information Center (EPIC®) platform. This study was approved by the Institutional Review Board of the Kaiser Foundation Research Institute with waiver of consent. Kaiser Permanente Southern California also received approval from the California Health and Human Services Agency and California Department of Public Health Center for Health Statistics and Informatics.

Ectopic pregnancy ascertainment

The identification of ectopic pregnancy cases was primarily based on a previously validated classification algorithm that incorporated International Classification of Diseases (ICD), 9th Revision, Diagnostic and Procedure codes, Current Procedural Technology (CPT-4) codes for ectopic pregnancy (Supplemental Table 1), and medication codes for methotrexate injections. ¹² In the algorithm, all encounters with an ectopic pregnancy diagnostic, procedural, or medication codes occurring within a 180-day period were considered part of the same pregnancy episode, with the diagnosis date defined as the date of the first encounter with an ectopic pregnancy code in an episode. The algorithm for identifying ectopic pregnancies was updated to include ICD 10th Revision codes, and encounter types and medication codes specific to the current health systems and validated. The accuracy of ectopic pregnancy case ascertainment using the updated algorithm was assessed; the sensitivity was 97.6% and the negative predictive value was 94.6%. ¹³

Statistical analysis

Data on age, gender, self-reported race/ethnicity, and health plan enrollment were obtained from administrative databases. The number of pregnancies (live births, induced abortions, and ectopic pregnancies) in women aged 15–44 years in Kaiser Permanente Northern California and Kaiser Permanente Southern California during the study period was used as the denominator to estimate the crude ectopic pregnancy incidence per 1000 pregnancies and 95% confidence interval (CI) per study year and overall. The crude incidence was estimated for all women and stratified by 5-year age groups (15–19, 20–24, 25–29, 30–34, 35–39, and 40–44 years). Changes in the age distribution of women over the study period were assessed and the age-adjusted ectopic pregnancy incidence and 95% CI per study year and overall was estimated. The percent distribution of pregnant women by age group in the middle of the study period in 2014 was used as a standard for the age-adjusted incidence per 1000 pregnancies. The age-adjusted incidence was estimated for all women and stratified by race/ethnicity group (non-Hispanic White, non-Hispanic Black, Hispanic, Asian/Pacific Islander, and other/unknown).

Stillbirths and spontaneous abortions were excluded. Pregnancies were counted in the year they ended, and each pregnancy was counted as a separate observation in women who had more than one pregnancy during the study period. Poisson regression models adjusting for overdispersion, with calendar year fitted as a continuous variable, were used to evaluate linear trends in incidence over the study period. Two-sided p values that were considered statistically significant (< 0.05) are reported. Data were analyzed using Statistical Analysis System (SAS v9.4; SAS Institute, Cary, North Carolina).

Results

From January 1, 2010 to December 31, 2019, 4,021,312 women were enrolled in Kaiser Permanente Northern California and Kaiser Permanente Southern California health plans. The number of women enrolled and the number of women who had a pregnancy steadily increased throughout the observed period. There were 15,537 ectopic pregnancies among 979,027 pregnancies in 693,685 women (Table 1). The overall age-adjusted incidence of ectopic pregnancy for all women was 15.8 per 1000 pregnancies, 95% CI: 15.6, 16.1. Furthermore, there was a significant increase over time; the annual incidence increased from 15.2 (95% CI: 14.4, 16.1) in 2010 to 16.4 (95% CI: 15.6, 17.2) in 2019 (p < 0.001) (Table 1).

Table 1:

Ectopic pregnancy incidence (per 1000 pregnancies) and trend among women ages 15−44 years, Kaiser Permanente Northern and Southern California, 2010−2019

	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Overall
Number of women
	1,566,081	1,620,472	1,662,112	1,688,589	1,772,521	1,893,286	1,976,705	2,060,954	2,136,241	2,177,725	4,021,312
Number of women with pregnancies a
	85,457	88,190	90,226	89,217	92,807	97,251	101,184	102,743	105,235	108,857	693,685
Number of pregnancies a
	87,233	89,786	91,681	90,600	94,121	98,872	103,611	105,039	107,106	110,978	979,027
Number of ectopic pregnancies
	1307	1342	1359	1354	1450	1641	1702	1727	1810	1845	15,537
Age at ectopic pregnancy
Years (SD)	30.6 (6.0)	30.6 (5.9)	30.8 (6.0)	30.8 (5.9)	30.8 (5.8)	31.3 (5.7)	31.5 (5.7)	31.2 (5.5)	31.2 (5.5)	31.3 (5.6)	31.1 (5.8)
Ectopic pregnancy incidence (per 1000 pregnancies a )
Crude	15.0	14.9	14.8	14.9	15.4	16.6	16.4	16.4	16.9	16.6	15.9
95% CI	14.2,15.8	14.2,15.8	14.0,15.6	14.2,15.8	14.6,16.2	15.8,17.4	15.7,17.2	15.7,17.2	16.1,17.7	15.9,17.4	15.6,16.1
Age-adjusted b	15.2	15.2	15.0	15.0	15.4	16.5	16.3	16.2	16.6	16.6	15.8
95% CI	14.4,16.1	14.4,16.0	14.2,15.8	14.2,15.8	14.6,16.2	15.7,17.3	15.5,17.1	15.4,17.0	15.9,17.4	15.6,17.2	15.6,16.1

^a Includes live births, induced abortions, and ectopic pregnancies; linear trend.

^b p < 0.001.

CI, confidence interval; SD, standard deviation.

The mean age at ectopic pregnancy diagnosis was 31.1 years (SD 5.8) (Table 1). When stratifying by age groups, the overall crude incidence of ectopic pregnancy varied. The lowest incidence was observed among adolescents (ie, 15–19 years) who had an incidence of 9.6 (95% CI: 8.7, 10.6) (Table 2). In addition, adolescents also had the lowest number of ectopic (n = 388; 2.5%) and overall pregnancies (n = 40,464; 4.1%) (Supplemental Table 2). Women aged 40–44 years had the second lowest number of pregnancies (n = 44,925; 4.6%). However, women in this age group had the highest overall incidence of ectopic pregnancy (24.2, 95% CI: 22.7, 25.6) (Table 2). Significant increases were seen in the annual ectopic pregnancy incidence for women aged 24 and younger and women aged 30–34 years over the study period. This contributed to the increase in the overall annual age-adjusted ectopic pregnancy incidence (Table 2). The annual incidence doubled in adolescents over the study period, but remained stable for women aged 35 and older, who had the highest overall ectopic pregnancy incidence (Figure 1).

Table 2:

Ectopic pregnancy incidence (per 1000 pregnancies ^a ) and trend by age and race/ethnicity, Kaiser Permanente Northern and Southern California, 2010–2019

Characteristic	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Overall	95% CI
Ectopic pregnancy crude incidence (per 1000 pregnancies a ) by age
15−19 b	7.2	8.0	8.4	8.7	10.3	10.7	12.3	10.1	10.8	14.4	9.6	8.7, 10.6
20−24 c	11.8	12.1	12.9	12.5	12.2	11.5	12.8	12.9	14.0	13.2	12.6	12.0, 13.2
25−29	14.8	13.2	12.5	13.3	15.2	14.6	14.5	13.9	16.5	15.1	14.4	13.9, 14.9
30−34 c	15.4	16.7	15.3	15.1	15.0	16.8	16.6	16.4	16.4	17.3	16.2	15.7, 16.6
35−39	17.7	18.5	18.8	19.6	18.8	22.4	20.0	21.0	19.5	18.1	19.5	18.9, 20.2
40−44	26.4	21.1	25.3	20.9	22.4	25.3	26.1	26.4	23.2	24.0	24.2	22.7, 25.6
Ectopic pregnancy incidence (per 1000 pregnancies a ) by race/ethnicity d
Non-Hispanic White e	14.2	14.1	13.7	13.1	15.1	15.5	15.4	14.2	15.3	15.1	14.6	14.1, 15.0
Non-Hispanic Black	20.6	24.4	24.3	18.5	21.2	25.1	21.6	21.9	21.6	21.3	21.9	21.0, 22.8
Hispanic b	15.8	14.5	14.2	15.4	15.0	15.3	15.4	16.2	16.6	16.5	15.5	15.1, 15.9
Asian/Pacific Islander b	12.1	12.9	13.2	15.6	13.4	16.4	16.9	15.7	15.6	14.7	14.7	14.2, 15.3
Other/unknown b	15.4	19.3	17.4	14.5	19.8	16.4	18.7	22.8	22.3	22.0	19.3	17.5, 21.3

^a Includes live births, induced abortions, and ectopic pregnancies.

^b Linear trend p < 0.001.

^c p = 0.02.

^d Age-adjusted based on the distribution for age in 2014.

^e p = 0.03.

CI, confidence interval.

Figure 1:

Ectopic pregnancy incidence (per 1000 pregnancies) and temporal trend by age, 2010–2019. Linear trend Poisson regression.

When stratifying by race/ethnicity, the overall age-adjusted ectopic pregnancy incidence also differed (Table 2). The largest number of ectopic pregnancies (n = 5,951; 38.3%) occurred in Hispanic women who, in addition, had the highest number of pregnancies overall (n = 381,232; 38.9%) (Supplemental Table 3). However, the overall age-adjusted ectopic pregnancy incidence was highest among non-Hispanic Black women (21.9 per 1000 pregnancies, 95% CI: 21.0, 22.8). Moreover, the incidence for this group was 50% higher than that of non-Hispanic White women (14.6 per 1000 pregnancies, 95% CI: 14.1, 15.0) (Table 2). Lastly, the annual incidence of ectopic pregnancy significantly increased among all racial/ethnic group except for non-Hispanic Black people; however, their overall incidence remained higher than those of non-Hispanic White people throughout the study period (Figure 2).

Figure 2:

Ectopic pregnancy incidence (per 1000 pregnancies) and temporal trend among women ages 15–44 by race/ethnicity, 2010–2019. Linear trend.

Discussion

In this large population-based cross-sectional study, ectopic pregnancies occurred in about 1.6% of pregnancies. A gradual increase in the ectopic pregnancy incidence was observed over the observed period (2010–2019). This increase was found to be largely driven by increasing ectopic pregnancy incidence in younger women. Thus, higher rates of infertility and use of assisted reproductive technology in older women in the United States, factors associated with increased risk for ectopic pregnancy, are unlikely to underlie this trend. ^14–17 Our findings demonstrate that ectopic pregnancy is a persistent source of reproductive health morbidity, in particular for older women ( >40 years) and non-Hispanic Black women who were found to have the highest ectopic pregnancy incidence. Disparities in ectopic pregnancy incidence when stratifying by race/ethnicity and age have been consistently reported in national surveillance studies and Medicaid populations. ^18,19 The disproportionately high incidence of ectopic pregnancy among non-Hispanic Black people in this medically insured population underscores how racial/ethnic disparities and health inequities cut across payer status. The observed racial/ethnic disparity in the incidence of ectopic pregnancy in our population may be due to racial/ethnic disparities in sexually transmitted infections ^20–23 ; however, social, environmental, and structural factors associated with other adverse maternal health outcomes that disproportionately affect Black women may also play a role in the etiology of ectopic pregnancy. ^23,24 The parallel with other disparities in maternal health outcomes as well as the persistence over time suggests that interventions that include structural approaches, those that do not rely on individual behavior change to alter risks, should be evaluated. In addition to data on race/ethnicity, collection of data on social determinants of health that can be correlated with outcomes may elucidate environmental and structural factors that can be targeted for prevention.

Our findings suggest that ectopic pregnancy has increased in the last decade. This is not in line with previous estimates reported by researchers using health system data that indicated stable or declining rates from 1993 to 2007. For instance, Trabert documented a decline in the crude ectopic pregnancy incidence per 1000 pregnancies from 1993 to 2007 however, the change in the age-adjusted incidence from 18.2 in 1993–1995 to 15.3 in 2005–2007 was not statistically significant. ⁸ Furthermore, Hoover et al estimated the incidence of ectopic pregnancy from 2002 to 2007 using data from MarketScan, an administrative claims database of more than 200 US commercial health plans. ²⁵ This study did not document a change in rate over the study period; however, only claims data were utilized and the method used for ectopic pregnancy case ascertainment did not appear to be validated.

Strengths of our study include the large number of ectopic pregnancies and a diverse community-based population. Furthermore, in previous years the CDC estimated ectopic pregnancy incidence using combined data from the National Hospital Discharge Survey and the National Ambulatory Medical Care Survey. In 2002, 10 years after their previous estimates, the CDC reported that reliable estimates of the incidence of ectopic pregnancy could no longer be achieved by combining available nationally representative data sets because of the high likelihood of multiple counting of cases. ²⁶ A validated algorithm for case ascertainment was used, allowing automated case-finding with good capture of the outcome of interest. ¹³ The algorithm considered all codes identified in a 180-day time period to be associated with a single case, which minimized the chance of duplicate cases, which we believe is a major strength of our study.

One of the major limitations of our study is that it may not be nationally representative. The study population is broadly representative of the overall population in California except for individuals at the lower and upper extremes of income, and is generalizable to health care systems with similar patient populations. ²⁷ In addition, our estimates of ectopic pregnancy incidence may overestimate the true incidence because of underreporting of induced abortions that were obtained outside of the system. ²⁸ Spontaneous abortions and the small number of stillbirths were not included because of incomplete reporting of spontaneous abortions, and to maintain consistency with other surveillance studies that did not include spontaneous abortions or stillbirths in their denominator. ^5–7

In conclusion, our findings demonstrate that the ectopic pregnancy incidence increased in the last decade and remains a significant source of reproductive health morbidity. The incidence was particularly higher for older ( > 40 years) and non-Hispanic Black women. In addition, increasing trends in the incidence of ectopic pregnancy in young women were observed. This is disconcerting and should be further investigated in other study populations and prompt providers to consider the diagnosis for patients across the age span. Evidence-based clinical practice guidelines that focus on early identification and standardize care across the range of patients may reduce the impact of racial disparities and improve the quality of care overall. Furthermore, the racial/ethnic disparities in the incidence of ectopic pregnancy should also be considered in the context of other racial/ethnic disparities in maternal health outcomes, such as preterm and low–birth-weight births. Eliminating maternal health disparities, including ectopic pregnancy, will require targeted and multifaceted approaches to prevention.

Supplementary Material

TABLE S1