Model Inputs \(\rightarrow\) Biological Parameters \(\rightarrow\) Ectopic Pregnancy
Ectopic pregnancy is a pregnancy in which the fertilized ovum implants outside the intrauterine cavity, with more than 95% occurring in the fallopian tubes.[1] Because these sites can not accommodate placental attachment or embryo growth, the potential for rupture and hemorrhage make ectopic pregnancy a high risk condition and the leading cause of maternal mortality in the first trimester.[1,2]
Prompt ultrasound evaluation is key in diagnosing ectopic pregnancy.[2] Appropriate treatment for patients with nonruptured ectopic pregnancy may include expectant management, medical management with methotrexate, or surgery. Surgical treatment is appropriate if ruptured ectopic pregnancy is suspected.[2]
Ectopic pregnancies occur in around 2% of pregnancies, with increasing risk for older women.[3,4,5] Andersen (2000)[6] presents rates of ectopic pregnancy by maternal age group from a population-based register linkage study in Denmark. After adjusting the denominator by subtracting all induced abortions from the number of pregnancies, the adjusted risk of ectopic pregnancy was calculated by age group. A population-based study in France found similar effect sizes for increased risk of ectopic pregnancy by maternal age.[7]
Prior ectopic pregnancy is a strong risk factor for recurrent ectopic pregnancy, with recurrence rates reported from 10% to 25%,[8,9] or 10 times the risk of the general population. For example, Bhattacharya 2012[10] reports an adjusted hazard rate of 13.0 (95% CI 11.63-16.86) and Jacob 2017[4] reports an adjusted odds ratio of 8.17 (95% CI 4.87-13.69).
Although many women diagnosed with ectopic pregnancy have no identifiable risk factors, a number of other risk factors have been associated with ectopic pregnancy, including smoking, alcohol, tubal surgery or tubal damage from sexually transmitted infections (STI), prior pelvic infection, IUD use, and pregnancy conceived by assisted reproduction.[4,8,11,12,13,14,15] Some of these risk factors may be associated with maternal age. For example, increased risk of ectopic pregnancies in teenage women is most likely caused by pelvic inflammatory disease.[6]
We model risk by maternal age for first-time ectopic pregnancy. For women with a previous ectopic pregnancy we model a risk of recurrence. Using a hierarchical model with country-specific effects allows us to account for different underlying risks not explicitly modeled, such as STI prevalence and assisted reproduction.
Tubal pregnancies may either diminish in size and spontaneously resolve, or continue to grow and eventually lead to tubal rupture. However, there are no reliable clinical, sonographic, or biological markers (e.g. serum beta hCH or serum progesterone) that can predict rupture of a tubal ectopic pregnancy.[16] In addition, the time from conception to tubal rupture is short, often occurring within 6-8 weeks of conception.[13]
Reported tubal rupture rates range from population-based reports of 18%,[17] to clinic/physician office-based reports of 32%,[18] to hospital-based reports of 80% and higher.[19] For example, an analysis of 80 cases of ectopic pregnancy in Pakistan found a rupture rate over 90%,[20] likely due to selection bias of ruptured cases presenting at hospital.
In contrast, a study of women on Medicaid in the US found that ectopic pregnancy-associated complications occurred in only 11% of cases, with a mortality ratio of 0.48 per 100,000 live births.[21] Another study found that the ectopic pregnancy mortality ratio in the US declined from 1.15 to 0.50 deaths per 100,000 live births between 1980-1984 and 2003-2007.[22] In a review of African developing countries, a majority of hospital-based studies reported ectopic pregnancy case fatality rates (CFR) of 1-3% (10 times higher than reported in developed countries), with late diagnosis and emergency surgical treatment likely accounting for the high fatality rates.[23]
Because CFR estimates are based on diagnosed cases, we condition CFR on rupture in the model. We assume that the probability of rupture is around 10-20%, based on the population-based studies.[17,21] We assume that given rupture, the CFR is around 3-5% in the absence of treatment (i.e. natural history). Although tubal rupture seriously affects the current health of the mother, it seems to have no effect on subsequent fertility.[17] We therefore assume that women who survive a ruptured ectopic pregnancy have no continuing morbidity, besides the increased risk of recurrent ectopic pregnancy in the future.
We fit quadratic B-splines to the midpoints of the age groups, based on data from Andersen 2000.[6] Spline Width: 40
Spline # | Knot Location | Height |
---|---|---|
1 | 8 | \(N(0.03,0.002)\) |
2 | 22 \(^1/_3\) | \(N(0.01,0.002)\) |
3 | 37 \(^1/_3\) | \(N(0.05,0.002)\) |
4 | 52 | \(N(0.11,0.015)\) |
We extended the probability from age 50 to older ages.
Following conception in the model, the risk of ectopic pregnancy is simulated based on a woman’s age and her history of previous ectopic pregnancy. We assume that ectopic pregnancies happen during the first trimester (i.e. month 3) in the model.
GMatH (Global Maternal Health) Model - Last updated: 28 November 2022
© Copyright 2020-2022 Zachary J. Ward
zward@hsph.harvard.edu