The Journal of Emergency Medicine, Vol. 53, No. 6, pp.

819–828, 2017
Ó 2017 Elsevier Inc. All rights reserved.
0736-4679/$ - see front matter

http://dx.doi.org/10.1016/j.jemermed.2017.08.074

Clinical
Reviews in Emergency Medicine

EMERGENCY MEDICINE MYTHS: ECTOPIC PREGNANCY

EVALUATION, RISK FACTORS, AND PRESENTATION

Jennifer J. Robertson, MD, MSED,* Brit Long, MD,† and Alex Koyfman, MD‡
*Emory University School of Medicine, Atlanta, Georgia, †San Antonio Military Medical Center, Fort Sam Houston, Texas, and ‡University of
Texas-Southwestern, Parkland Hospital, Dallas, Texas
Reprint Address: Jennifer J. Robertson, MD, MSED, 3645 Habersham Road NE, Apartment T19, Atlanta, GA 30305

, Abstract—Background: Ectopic pregnancy (EP) is an INTRODUCTION

important cause of morbidity and mortality in females of
reproductive age. Proper diagnosis and treatment are An ectopic pregnancy (EP) is a condition in which a
critical, as complications such as rupture, hemorrhagic fertilized ovum implants outside the endometrium (1).
shock, and even death can occur. Objective: EP is a condition
The largest percentage of EPs occur in the fallopian
emergency physicians are trained to detect, yet there are
tubes, but they can also occur in the ovary, abdomen,
multiple myths concerning its evaluation and diagnosis.
This article reviews several of these myths in order to improve and cervix (2). Although the incidence of EP approxi-
emergency department (ED) evaluation and diagnosis. mates 2% of all pregnancies, it remains an important
Discussion: EP is a difficult diagnosis and may be missed on cause of morbidity and mortality (1,2). From 2003 to
initial ED visit. While the diagnosis is often delayed simply 2007, EP accounted for 0.26 maternal deaths per
due to very early presentations, it can also be missed because 100,000 live births in Caucasian women and 1.75
patients may not have all the same risk factors or demonstrate deaths per 100,000 live births in African-American
the same symptoms. They may also not demonstrate the same women. Undiagnosed or untreated EP can also lead to
serum B-human chorionic gonadotropin levels and trends or complications, such as excessive hemorrhage, shock, or
have the same ultrasound findings at equivalent gestational renal failure (1).
ages. Some patients with early EP may have positive ultra-
Although it is important to diagnose and treat EP early,
sound findings with serum b-hCG levels under a defined
the diagnosis can be challenging. Clinicians commonly
discriminatory zone (DZ). On the other hand, some patients
with an early viable intrauterine pregnancy may have no misdiagnose EP on an initial medical encounter (3,4).
visible findings on initial ultrasound, but have serum b-hCG Several recommendations attempt to simplify decision
(quantitative) levels well above the DZ. Although rare, EP making and facilitate the diagnosis with obtaining
has even been demonstrated in women with negative urine accurate history, specific laboratory tests, and imaging
b-hCG tests or low serum b-hCG levels. Conclusions: While studies (5,6). However, not every patient with EP will
EP may be a challenging diagnosis, understanding the myths present similarly and, unfortunately, not every patient
surrounding EP may help emergency physicians consider it, will demonstrate similar findings on these laboratory or
even when patient risk factors, symptoms, or ED laboratory imaging tests. The current article discusses EP and
or imaging studies do not initially or easily define the diag- some of these discrepancies in the diagnosis of EP in
nosis. Ó 2017 Elsevier Inc. All rights reserved.
the emergency department (ED). Of note, in this
, Keywords—ectopic; pregnancy; serum b-hCG; discrim- particular article, all serum b-human chorionic
inatory zone; ultrasound gonadotropin (b-hCG) levels refer to the quantitative test.

RECEIVED: 4 March 2017; FINAL SUBMISSION RECEIVED: 31 July 2017;

ACCEPTED: 16 August 2017

819
820 J. J. Robertson et al.

DISCUSSION thing for the emergency physician to note for reference

purposes.
Myth 1: If a Pregnant Patient Has a Serum hCG Above
Indeed, there have been cases where patients with an
the Discriminatory Zone and No Visualized Intrauterine
initial PUL and serum b-hCG levels above the DZ have
Pregnancy on Ultrasound, Then an Ectopic Pregnancy
gone on to have viable pregnancies (13,21,22). In 2014,
Must Be Present
Ko et al. reviewed 113 patients with initial PUL and
The discriminatory zone (DZ) describes the lowest serum hCG levels > 1000 mIU/mL. They followed these
b-hCG level where ultrasound (US) should detect visible patients for a specified time period to determine
signs of early pregnancy, such as yolk sac or fetal pole (7). outcomes, including eventual viable IUPs (21). The au-
With improving US technology, over time the DZ has thors evaluated only hemodynamically stable patients
fallen from a level of 6500 mIU/mL in the early 1980s without significant vaginal bleeding who had serum
to 1000–2000 mIU/mL in the present day (8–10). b-hCG levels > 1000 mIU/mL taken within 12 h of US.
Women without evidence of a visible intrauterine Transvaginal US was repeated when the hCG level
pregnancy (IUP) on transvaginal US are considered to showed a rise > 53%. Laparoscopy occurred when there
have pregnancy of unknown location (PUL), but it is was suboptimal rise in hCG, a visualized EP on repeat
often presumed that a patient with a PUL and a serum US, or development of hemodynamic instability or hemo-
b-hCG level above the DZ has an EP, which may not be peritoneum. Overall, in those patients with eventual visu-
a correct assumption (11). alized IUPs (n = 42), 23 were viable. Of these 23 patients,
Recent research has questioned that a single hCG the highest hCG level was 9083 mIU/mL, which occurred
measurement can distinguish IUP from EP, even when in a patient with a triplet pregnancy conceived through
the measurement is above the DZ (12,13). From the in vitro fertilization. Of the total 113 patients studied,
emergency practitioner’s standpoint, the initial goal is to 22 had EP diagnosed on laparoscopy, with initial serum
address a life-threatening condition. If any patient is b-hCG level range of 1155–21,770 mIU/mL. The other
unstable with concerning signs and symptoms, resuscita- 49 patients without confirmed EP or IUP were grouped
tion and urgent obstetrics consultation is paramount. In into ‘‘other,’’ which included resolved PUL (n = 45),
stable patients, a more thorough evaluation can occur. treated PUL (n = 2), and molar pregnancy (n = 2) (21).
The primary goal is to determine whether a viable IUP is The authors suspect several factors contributed to the
present (12). If no IUP is seen on US, then PUL is present, non-visualization of pregnancy on initial US, including
and potential outcomes for this pregnancy may include maternal obesity, adenomyosis, and uterine fibroids.
early viable IUP, early non-viable IUP, or EP (12). They conclude a single initial serum b-hCG level and
If a patient has an hCG level below the DZ, it is often inconclusive initial transvaginal US findings are not pre-
thought an US is unnecessary because nothing can be dictive of eventual IUP or EP (21).
visualized. However, as will be mentioned in Myth 2, Similarly, other studies have documented cases in
this may not be the case. In situations where the hCG is which an embryo was not seen on initial US in patients
above the DZ and no IUP is seen, then it is often with serum b-hCG levels above the DZ but was later
presumed the pregnancy is not viable or an EP is present. seen on subsequent US scans (7,13,22). In 2013,
However, the hCG DZ is a surrogate marker for gesta- Connolly et al. reviewed the records of 651 patients
tional age. A normal gestational sac should be visualized with known pregnancy outcomes who initially
around 5 weeks and 5 days from the last normal menstrual presented with symptoms of EP or miscarriage (7).
period, regardless if it is a singleton or multiple gestation They were evaluated with a serum b-hCG level and a
(14). The b-hCG should be used in conjunction with transvaginal US within 6 h of each other. In this popula-
suspected gestational age, and a single level should not tion, the highest observed hCG where no structure could
be used alone in clinical decision making (15). Serum be visualized was 2317 mIU/mL. Using a logistic regres-
progesterone level is another laboratory test that has sion model, the authors calculated the predicted probabil-
been used in conjunction with serum b-hCG level. It is ity of visualizing a gestational sac as a function of the
thought that low serum progesterone levels (< 5 ng/mL) serum hCG values. They found that in their population
are associated with non-viable pregnancies (4,16–18). of patients with a DZ of 1500 mIU/mL, the predicted
However, single progesterone levels are inadequate for probability of visualizing a gestational sac in a viable
determining EP vs. a non-viable IUP (18,19). Just as pregnancy was 80.4% (7). The authors conclude that
with US, the combination of the serum b-hCG level even with improving US technology, initial structural
with a serum progesterone level may be useful in the signs of pregnancy may not be seen on US in some
diagnosis of EP in those patients who have PUL (20). patients who have initial serum b-hCG levels above the
This review will not discuss this further, but it is some- classically defined DZ of 1000–1500 mIU/mL (7).
Ectopic Pregnancy 821

It is important for the emergency physician to under- hospital course (30). Similarly, Fu et al. reported the cases
stand the limitations of the serum b-hCG. If a patient’s of 2 patients who had serum b-hCG levels < 10 mIU/mL
initial serum b-hCG level is higher than the DZ, but no with EP rupture and hemoperitoneum (26). Other studies
signs of pregnancy are seen on initial US, EP is not the have shown rupture of EP with low or even negative serum
only diagnosis and only suggests the pregnancy is non- b-hCG levels (28,29).
viable (15). However, any outcome is possible. As long The most recent American College of Emergency
as patients are hemodynamically stable, reflexively these Physicians clinical guidelines provides a level B recom-
pregnancies as EPs would result in termination of an mendation on obtaining transvaginal US in patients
otherwise desired pregnancy, as well as exposure to treat- who have b-hCG levels below the DZ and concerning
ment side effects (21). Stable patients may be discharged, signs and symptoms (5). Given the studies and guidelines
but should receive appropriate obstetrics follow-up (5). In mentioned, it is prudent to consider imaging with US in
cases where patients return to the ED instead of their patients with concerning signs and symptoms for EP,
obstetrics physician, repeat hCG may be used along despite low or negative serum hCG levels (32). In addi-
with US as a potentially helpful diagnostic tool. Accord- tion, rupture can occur at any hCG level, and patients
ing to the most recent American College of Obstetricians should be counseled concerning the risks and symptoms
and Gynecologists guidelines, an increase in repeat serum of rupture. It is recommended that serum b-hCG levels
b-hCG < 53% in 48 h is suggestive of abnormal preg- be followed to zero (26).
nancy (23). Using this, the combination of US with repeat
serum b-hCG level is helpful when evaluating patients Myth 3: Ectopic Pregnancies Demonstrate Reliable
who return to the ED (13,15,21,22,24). Trends in Serum b-hCG Levels

Myth 2: If the Serum b-hCG Level Is Less Than DZ, Then In conjunction with clinical history and transvaginal US,
US Is of Little Utility trending serial serum b-hCG levels are frequently utilized
to determine the ultimate outcome of a PUL, whether it is
The entire premise behind the DZ is that it helps deter- a viable IUP, non-viable IUP, or EP (23). In the United
mine whether signs of early pregnancy can be visualized States, standard practice is to follow serial b-hCG levels
on US (9). It has been suggested that obtaining imaging in for a period of at least 48 h (6). Any deviation from a
patients with hCG levels below the DZ is of limited value ‘‘normal’’ rise or fall in serum hCG seems to suggest a
in locating the pregnancy. However, studies have demon- diagnosis other than a viable IUP (6). This idea stemmed
strated that this is not the case, and some EPs can rupture from a small study showing that for viable IUPs, the
in patients with b-hCG levels below the DZ (7,25–31). serum b-hCG should rise at least 66% in 48 h (9). Since
A 2007 study of 74 patients with US findings concern- then, a slower rate of rise has been proposed in abnormal
ing for EP evaluated hCG levels. Of these patients, 47 pregnancies and, as mentioned previously, American
were eventually diagnosed with EP (25). In these patients College of Obstetricians and Gynecologists guidelines
with a final diagnosis of EP, b-hCG levels ranged from 41 state an increase in serum b-hCG < 53% in 48 h is sugges-
to 59,846 mIU/mL. In addition, 17 of these patients had tive of abnormal pregnancy (23).
b-hCG levels < 1000 mIU/mL (25). Similarly, in the Most patients with similar types of pregnancies tend to
study by Connolly et al. mentioned previously, a visible follow similar b-hCG patterns. However, there are excep-
pregnancy was noted in a patient with an hCG value of tions, as up to one-third of normal pregnancies will not
390 mIU/mL (7). Most recently, Simsek et al. retrospec- demonstrate these typical trends (33). Table 1 provides a
tively studied initial ED evaluations of 35 women with summary of the myths mentioned in this article that may
known final diagnoses of EP (27). Overall, the mean make the diagnosis difficult. While emergency physicians
initial serum b-hCG value was 3560 mIU/mL; however, are not necessarily the primary providers who trend b-
there were values much lower than this, including a min- hCG levels in patients with PUL, it is important to under-
imum value of 17 mIU/mL (27). They also found that any stand there is overlap in b-hCG trends among viable IUP,
single level of b-hCG did not predict EP rupture (27). non-viable IUP, and EP (34–37). This can be seen in
Case studies have also demonstrated that a single initial cases of both decreasing and increasing b-hCG levels.
b-hCG value under the DZ is not protective against EP First, an appropriate increase in the b-hCG level is not
rupture (26,28–30). A 2012 paper discussed the case automatically diagnostic of an IUP (38). In a 2006 cohort
with a serum b-hCG level of 364 mIU/mL and a study of 200 women with ultimate diagnoses of EP, 60%
ruptured ectopic ovarian pregnancy and hemoperitoneum had an initial rise in b-hCG over 48 h. Importantly, 20.8%
(30). The patient was unstable, and emergency laparos- of these patients had hCG levels that rose $ 53%, similar
copy and resection of a left ovarian EP were performed. to a developing IUP (35). A study by Horne et al. found
The patient recovered and had an otherwise uneventful that in patients with PUL, 16% with eventual EP had
822 J. J. Robertson et al.

Table 1. Summary

1. Pregnant patients with serum b-hCG levels greater than the DZ and no visualized IUP on US may or may not have an EP.
2. If the serum b-hCG level is less than DZ, US may definitely be helpful in diagnosing EP.
3. Patients with EP do not demonstrate reliable trends in serum b-hCG levels.
4. While rare, patients with EP have been found to have initial negative urine point-of-care b-hCG tests.
5. Patients with EP may or may not present with pain or adnexal tenderness.
6. Patients with a ruptured EP can have low negative serum or even negative urine b-hCG tests.
7. Overall, contraceptives protect against EP because they prevent pregnancy.
8. Many patients with EP do not have even one classic predisposing risk factor.

b-hCG = b–human chorionic gonadotropin; DZ = discriminatory zone; EP = ectopic pregnancy; IUP = intrauterine pregnancy;
US = ultrasound.

normally rising b-hCG levels (defined as $ 66%) in the single episode of rupture and bleeding. It typically
first 48 h (39). In a retrospective cohort study of 179 develops into a hematocele that contains old blood clots
women with initial PUL and eventual EP diagnoses, and gestational tissue (46). The sonographic appearance
60% showed an increase in b-hCG (median 32% in- of a chronic EP is an adnexal mass with surrounding or
crease) over the first two measurements (37). While the adjacent low-resistance arterial vessels and no IUP seen
authors do not discuss the actual number of patients (47).
with trends similar to IUPs, 27% demonstrated b-hCG Most recently, studies have suggested that false-
trends resembling that of a growing IUP or miscarriage negative qualitative urine hCG tests may also be due to
(37). Declining b-hCG levels do not necessarily mean a ‘‘hook like’’ or ‘‘hook-like variant’’ effects that occur
miscarriage is present, just as appropriately increasing from abnormal antigen-antibody binding in the assays
b-hCG levels do not indicate an IUP (37,38). (43). During pregnancy, various subunits of the hCG
Secondly, while any progressive decline in serum hormone can be seen in the serum and urine. The concen-
b-hCG likely indicates a non-viable pregnancy, EP is still tration of each subunit depends on gestational age and
possible (38). In the same 2006 cohort study by Silva whether another pathologic condition, such as gestational
et al., 8% of women with eventual EP had b-hCG levels trophoblastic disease, is present (48). The major hCG
similar to those with completed spontaneous abortions subunit detected in the urine is the hCG b core fragment
(35). A 2012 cohort study by Morse et al. also found (hCGbcf), especially as pregnancy progresses (48).
that of 179 patients with eventual EP diagnoses, 30 Some urine POC tests utilize one-step sandwich assays
were initially misclassified with miscarriage or IUP in which antibodies recognize and bind hCGbcf to form
(34). Of these 30 women, 20% had declining b-hCG a ‘‘sandwich,’’ ultimately leading to a positive test. If this
levels similar to those with miscarriages (34). Another hCGbcf concentration is in excess, it can saturate all of
cohort study also found similar inconsistencies with the antibodies and prevent this sandwich formation. This
b-hCG trends among EP, IUP, and miscarriages (33). is called the ‘‘hook effect’’ and is not typically seen unless
Overall, there is no one way for serum b-hCG patterns hCG concentrations reach 1,000,000 mIU/mL (40,43). For
to accurately diagnose EP (35). While the rise or fall of the most part, modern manufacturers have developed tests
the serum b-hCG can guide clinical management, this that can bind very high concentrations of hCG, and this
technique does have limitations. Every patient will be effect is rarely seen except in those who have
different, and patients with concerning signs and symp- abnormally high hCG concentrations (43). However, a
toms of EP should be closely monitored and considered recent 2014 study by Nerenz et al. found several assays
to have PUL until a final diagnosis is made (38). still demonstrated results similar to the hook effect, result-
ing in false-negative urine pregnancy tests (43).
Myth 4: If a Patient Has Concerning Signs and Symptoms The variant hook effect can also occur where a partic-
of EP but a Urine Point of Care b-hCG Test Is Negative, ular test’s antibodies are unable to detect certain hCG
Then EP Is Not Possible variants also present later in the first trimester
(6–12 weeks) (40,41,43,49). If the concentration of one
While rare, it is possible for a urine b-hCG point-of-care particular hCG fragment is in surplus, it can bind one
(POC) test to be falsely negative (40–44). False negative, antibody well and not another, which results in a false-
including serum, tests can be due to insufficient or diluted negative test (40,41,49). Overall, appropriate dilution is
urine, incorrect reading times, an EP with deficient the key to obtaining an accurate test (43). The variant
production of hCG, or chronic EP with inactive hook effect may occur in modern assays because of fluc-
trophoblasts (40,45). Of note, a chronic EP is an EP tuating hCG fragments present in various concentrations
that undergoes repeated small ruptures instead of one during the stages of pregnancy (40,41).
Ectopic Pregnancy 823

The inability of various urine pregnancy tests to prop- tion of this study is that the studied population included
erly detect hCG fragments in very early pregnancy is women who presented to an early pregnancy clinic and
also seen in both POC and over the counter (OTC) tests not an ED.
(40,50,51). A 2009 study by Cervinsky et al. found some Other studies have emphasized that the presentation of
OTC tests to have better analytical sensitivity than POC EP can be non-specific, and women may have varied
tests in early pregnancy (within 10 days of expected symptoms and examination findings (36,58–60). A
menses) (50). Another 2013 study evaluated two urine recent observational cohort study of 72 patients with EP
POC tests and their sensitivities in detecting pregnancy diagnoses found that the presence of pelvic pain does
at 2–5 weeks (52). The authors measured patients’ serum increase the risk that a patient will have an EP. However,
hCG levels and evaluated the ability of these patients’ the study also found that 16 of 72 (22.2%) patients with
urine POC tests to be appropriately positive. The ability EP presented without pain (58). Another prospective
of one device to detect a positive result in urine in patients observational study demonstrated the large ranges of signs
with serum hCG levels > 20 mIU/mL was 80%. The ability and symptoms patients with EP can experience, including
of the other device to detect a positive urine result in pain severity and abdominal or pelvic tenderness (60).
patients with serum hCG concentrations > 20 mIU/mL Overall, authors concluded that there was no particular
was 90% (52). Overall, the authors concluded that the sen- pattern of historical or examination findings that could reli-
sitivities of these two commonly used urine POC assays ably predict or exclude the diagnosis of EP (60).
were insufficient for detecting early pregnancy (mean While there have been no proven medical or economic
4 weeks) (52). Other studies have found similar results benefits to screening asymptomatic women for EP, it is
that many urine pregnancy tests do not possess 100% likely that women who present to the ED will likely
sensitivity in detecting early pregnancy (41,51,53). have some particular symptom resulting in their presenta-
Regardless of the mechanism, it is important to under- tion to the ED (61). Some patients may present atypically,
stand false-negative urine test results can be seen in the and not all patients will present with pelvic or abdominal
early and later stages of pregnancy. This has been seen pain. Pain may not be present, and vaginal bleeding may
in several case reports and case series (40–42,44,54). In be the only symptom. A painless EP occurs rarely, but
patients with a negative urine pregnancy test but even without pain, emergency providers should consider
concerning signs and symptoms of EP, a serum hCG EP in the differential diagnosis in women of childbearing
level and close monitoring are prudent. age who present with other symptoms of EP, such as
vaginal bleeding or diarrhea.
Myth 5: Patients with EP Always Present with Pain or
Adnexal Tenderness on Examination Myth 6: A Ruptured EP Cannot Be Possible if a Patient
Has a Low or Even Negative Serum or Urine b-hCG Test
The classic symptoms and signs of EP are abdominal or
pelvic pain, vaginal bleeding, and amenorrhea (55). A ruptured EP is an important diagnosis, as it can lead to
While abdominal pain and vaginal bleeding in the first serious consequences, including hemorrhagic shock and
trimester are common symptoms of EP, it should be death (29,31). While higher b-hCG levels (> 5–10,000
kept in mind that these symptoms are not sensitive or spe- mIU/mL) are more commonly seen in cases of ruptured
cific. Many women with normal IUPs also have these EP, a low or even negative b-hCG in the concerning
symptoms, and this should be kept in mind when evalu- patient does not rule out potential rupture (62). Reports
ating any pregnant patient with pelvic pain or vaginal have demonstrated cases of patients with ruptured EPs
bleeding (3,4). with negative urine b-hCG tests or low or negative serum
Patients with EP can also have varied presentations b-hCG levels (29–31,45,63). Trends in hCG from patient
and may arrive with minimal or no symptoms (56). In to patient will vary. One level is not predictive of ruptured
one prospective study of 481 women who presented to or non-ruptured EP (64). In addition, the false negatives
an ED with first-trimester pain or vaginal bleeding, 56 mentioned in Myth 4 may also contribute to the false-
were ultimately diagnosed with EP. Of these 56 patients, negative findings seen in ruptured EPs (45).
9% had no pain and had vaginal bleeding as their only A retrospective chart review of 693 pregnancies found
presenting complaint. In addition, 36% of patients with 234 of these were EPs. Of these 234 EPs, 37.6% (n = 88)
an ultimate diagnosis of EP had no adnexal tenderness ruptured and 62.4% (n = 146) were unruptured (63). Of
on examination (57). the ruptured EPs, 11.4% had serum b-hCG levels < 100
In 2005, another prospective observational study eval- mIU/mL. Another 38.6% of the patients with ruptured
uated 527 women with PUL, and of these, 46 had eventual EPs had serum b-hCG levels of 100–999 mIU/mL (63).
diagnoses of EP (13). Of these EP patients, 62% There were wide ranges of b-hCG levels for unruptured
presented with no lower abdominal pain (13). A limita- and ruptured groups, and the authors concluded that there
824 J. J. Robertson et al.

seemed to be no correlation between serum b-hCG levels There has also been concern that emergency contracep-
and tubal rupture. Rupture can occur even in cases with tion, such as levonorgestrel, ulipristal acetate, and
low serum b-hCG levels (63). mifepristone (not available in the United States), also
In 2006, another study evaluated 183 surgically treated places women at risk for EP (68). While there have been
tubal EPs, 75 of which were ruptured (64). Each patient cases of EP in patients taking emergency contraception,
had her serum hCG level drawn before surgery. Overall, studies have shown low rates of EP after emergency
no significant differences were found between groups. contraception, and it remains an effective method of pre-
As in the study mentioned, there was no specific pattern venting pregnancy (66,68–72). A 2015 case-control study
to the hCG levels with broad ranges. In the ruptured group, found levonorgestrel to be an effective contraceptive,
the b-hCG levels ranged from a high of 75,071 mIU/mL to while not increasing overall risk of EP. However, similar
a low of 8 mIU/mL. The unruptured group was similar, to other contraceptive methods, the study demonstrated
showing a high of 89,504 mIU/mL and a low of 15 mIU/ that if pregnancy did occur, the risk of EP increased
mL. Consequently, the authors conclude that there is no (odds ratio 2.79; 95% confidence interval 2.27–3.43)
lower limit in serum b-hCG levels for ruptured EP (64). (68). A systematic review of 136 studies evaluated preg-
A ruptured EP can be a serious and deadly diagnosis. nancy rates after either mifepristone or levonorgestrel,
However, it can also be a difficult diagnosis, as patients and it demonstrated low rates of EP (66). Of all cases of
may present with low or even negative b-hCG tests. In failed contraception, the authors found a rate of 0.6% of
a case report by Lee et al., a 25-year-old patient with EP in the mifepristone group and 1% in the levonorgestrel
persistent negative serum b-hCG tests presented in group. Compared to the general population, these rates do
hemodynamic shock from an EP after having presented not differ much from the average rate of EP (66). Finally,
previously to six various hospitals (31). Her diagnosis of three clinical trials evaluating levonorgestrel emergency
was made via laparoscopy, demonstrating an EP in a uter- contraception, one EP was found out of 67 women with
ine tube (31). Other authors have reported similar cases of failed pregnancies (73–75).
patients with negative b-hCG tests presenting with hemo- The most recently approved emergency contraceptive
dynamic instability secondary to a ruptured EP (29,45). is ulipristal. According to clinical trials and post-
Given the potential morbidity and mortality of a marketing data, no EPs were recorded after using ulipris-
ruptured EP, emergency providers should keep in mind tal acetate (69). In a later study evaluating ulipristal after
the rare but possible situation of ruptured EP with negative its debut in 2010, there were 376 pregnancies reported,
or low hCG levels. While low or negative hCG tests are not with 4 EPs (1.0%). The limitation to this is that all of
commonly seen in ruptured EP, it may be appropriate for the data are self-reported. However, it does demonstrate
providers to consider the diagnosis in any female of child- the small number of EPs reported in comparison to other
bearing age who is hemodynamically unstable and has types of pregnancy outcomes (71).
concerning history or symptoms. Just as in the myths Contraceptive tools are successful in preventing
mentioned, hCG levels will vary in patients and are poor pregnancy. When compared to not using contraception,
predictors of EP. It is not recommended that one quantita- these methods may help prevent EP simply by preventing
tive hCG level be relied on to guide decision making, pregnancy. Physicians should understand that failure of
including whether to obtain imaging (45,64). these methods can lead to an increased risk of EP. Thus,
emergency providers should consider EP in patients
Myth 7: Contraceptives, Including Tubal Ligation, who are utilizing contraception and present with concern-
Intrauterine Devices, and Emergency Methods Are ing signs and symptoms.
Associated with More Ectopic Pregnancies Than Those in
the General Population Myth 8: The Majority of Patients with EP Have at Least
One Risk Factor That Predisposes to the Condition
This statement is actually not a true myth but it is important
to mention that these methods, if taken properly, do not in- The major risk factors for EP include disorders or proced-
crease a women’s risk of EP. Rather, the risk of EP is only ures that may result in fallopian tube damage (76,77).
higher if the method fails (2,65). As compared to using no Based on available literature, suspected risk factors for
contraception, female sterilization and intrauterine devices EP include a history of tubal surgery or congenital tube
(IUDs) actually reduce the risk of EP because they reduce abnormalities, assisted reproductive technology, a history
the risk of pregnancy overall. When used properly, the of induced abortions, previous EP, older pregnancy age,
same is true for emergency contraceptive methods and pelvic inflammatory disease (PID) (63,78,79). While
(2,65–67). It is important to remember IUDs and women with risk factors are likely more prone to
sterilization decrease the overall risk of EP because they developing EP, approximately 50% of EP cases are in
reduce the absolute risk of pregnancy. women without known risk factors (2,60,63).
Ectopic Pregnancy 825

Concern for EP based on

history and examination?

- Resuscitate with IV blood products

Unstable - Obtain Serum hCG, CBC, T&S
Assess hemodynamic status - Perform US: FAST for free fluid, with
focus on pelvic window and uterus
-Emergent OB/GYN consultation
Stable

Evaluate for other

- Concern for EP still No etiology for symptoms,
Urine hCG present or concern for consider other imaging
reliability of Urine hCG? (Pelvic US, CT)
Yes
+

Serum hCG Perform Pelvic US IUP

EP – Ectopic Pregnancy
IV – Intravenous No IUP
CBC – Complete Blood Cell Count
T&S – Type and Screen
US – Ultrasound
FAST – Focused Assessment with
Sonography for Trauma
OB/GYN – Obstetrics/Gynecology Yes
Is there strong suspicion of Consult OB/GYN for
CT – Computed Tomography EP OR is serum suspected EP
hCG > 1000-2000 mIU/mL?

No

Discuss with OB/GYN and

obtain follow up for patient
at 48 hours for repeat hCG
and US

Fig. 1. Algorithm for diagnosis and management of ectopic pregnancy.

In one retrospective study by Saxon and colleagues, While patients with risk factors such as PID and a his-
risk factors such as previous EP, previous PID, previous tory of EP are at greater risk for EP, the absence of risk
IUD use, a history of infertility, or previous tubal ligation factors is also not protective. Many women with EPs do
or tubal surgery were evaluated in 693 EPs (ruptured and not have concerning histories. In patients with concerning
unruptured). Overall, the authors found 57% of these signs and symptoms without known risk factors for EP,
women had none of these risk factors (63). In addition, EP is still a possibility.
33% had one risk factor, while 9.4% had two risk factors,
2.2% had three, and 0.7% had four of these risk factors. CONCLUSIONS
Other authors have demonstrated similar results
(78,80,81). A recent retrospective analysis of 109 EP can be a challenging diagnosis and is commonly missed
patients with known EP showed 53.21% of patients had on initial evaluation. Not all patients possess the classic
no known risk factors. These risk factors included serum hCG DZ levels or demonstrate reliable b-hCG
previous EP, history of tuboplasty or tubal ligation, a trends. Rupture can occur at low b-hCG levels, and patients
history of infertility, PID, a history of abortion, or with EP may even have initial negative urine pregnancy
history of pelvic surgery (78). Another study of 38 tests. Patients may not present with adnexal pain, and not
women found risk factors such as infertility, previous all patients with EP will have a known risk factor. Knowl-
EP, previous tubal surgery, and history of IUD use were edge of several myths concerning US, hCG, risks factors,
not present in 39.5% of the patients studied (80). and history and examination can provide assistance in
826 J. J. Robertson et al.

the evaluation of EP. Clinicians should consider these with pain or bleeding: meta-analysis of cohort studies. BMJ 2012;
345:e6077.
myths when evaluating all women of reproductive age 20. El Bishry G, Ganta S. The role of single serum progesterone
who present with concerning signs and symptoms of EP, measurement in conjunction with bhCG in the management of
including abdominal or pelvic pain, vaginal bleeding, or suspected ectopic pregnancy. J Obstet Gynaecol 2008;28:413–7.
21. Ko JK, Cheung VY. Time to revisit the human chorionic gonado-
hemodynamic instability of unknown etiology. Figure 1
tropin discriminatory level in the management of pregnancy of
may be helpful toward making the diagnosis when a patient unknown location. J Ultrasound Med 2014;33:465–71.
presents with concerning signs and symptoms of EP. 22. Doubilet PM, Benson CB. Further evidence against the reliability of
the human chorionic gonadotropin discriminatory level. J Ultra-
sound Med 2011;30:1637–42.
REFERENCES 23. American College of Obstetricians and Gynecologists. ACOG
Practice Bulletin No. 94: Medical management of ectopic preg-
1. Creanga AA, Shapiro-Mendoza CK, Bish CL, Zane S, Berg CJ, nancy. Obstet Gynecol 2008;111:1479.
Callaghan WM. Trends in ectopic pregnancy mortality in the United 24. Gracia CR, Barnhart KT. Diagnosing ectopic pregnancy: deci-
States: 1980–2007. Obstet Gynecol 2011;117:837–43. sion analysis comparing six strategies. Obstet Gynecol 2001;
2. Marion LL, Meeks GR. Ectopic pregnancy: history, incidence, 97:464–70.
epidemiology, and risk factors. Clin Obstet Gynecol 2012;55(2): 25. Adhikari S, Blaivas M, Lyon M. Diagnosis and management of
376–86. ectopic pregnancy using bedside transvaginal ultrasonography
3. Abbott J, Emmans LS, Lowenstein SR. Ectopic pregnancy: ten in the ED: a 2-year experience. Am J Emerg Med 2007;25:
common pitfalls in diagnosis. Am J Emerg Med 1990;8:515–22. 591–6.
4. Stovall TG, Kellerman AL, Ling FW, Buster JE. Emergency depart- 26. Fu J, Henne MB, Blumstein S, Lathi RB. Rupture of ectopic
ment diagnosis of ectopic pregnancy. Ann Emerg Med 1990;19: pregnancy with minimally detectable beta-human chorionic gonad-
1098–103. otropin levels: a report of 2 cases. J Reprod Med 2007;52:541–2.
5. Brown MD, Byyny R, Diercks DB, et al. Clinical policy: critical is- 27. Simsek Y, Ay MO. Analysis of ectopic pregnancies admitted to
sues in the initial evaluation and management of patients presenting emergency department. Turk J Emerg Med 2015;15:151–4.
to the emergency department in early pregnancy. Ann Emerg Med 28. Hochner-Celnikier D, Ron M, Goshen R, Zacut D, Amir G, Yagel S.
2017;69:241–50. Rupture of ectopic pregnancy following disappearance of serum
6. Barnhart K, van Mello NM, Bourne T, et al. Pregnancy of unknown beta subunit of hCG. Obstet Gynecol 1992;79:826–7.
location: a consensus statement of nomenclature, definitions, and 29. Grynberg M, Teyssedre J, Andre C, Graesslin O. Rupture of ectopic
outcome. Fertil Steril 2011;95:857–66. pregnancy with negative serum b-hCG leading to hemorrhagic
7. Connolly A, Ryan DH, Stuebe AM, Wolfe HM. Reevaluation of shock. Obstet Gynecol 2009;113:537–9.
discriminatory and threshold levels for serum b-hCG in early preg- 30. Resta S, Fuggetta E, D’Itri F, Evangelista S, Ticino A, Porpora MG.
nancy. Obstet Gynecol 2013;121:65–70. Rupture of ovarian pregnancy in a woman with low beta-HCG
8. Barnhart K, Mennuti MT, Benjamin I, Jacobson S, Goodman D, levels. Case Rep Obstet Gynecol 2012;2012:213169.
Coutifaris C. Prompt diagnosis of ectopic pregnancy in an emer- 31. Lee JK, Lamaro VP. Ruptured tubal ectopic pregnancy with nega-
gency department setting. Obstet Gynecol 1994;84:1010–5. tive serum beta hCG—a case for ongoing vigilance? N Z Med J
9. Kadar N, DeVore EGR, Romero R. Discriminatory hCG zone: its 2009;122(1288):94–9.
use in the sonographic evaluation for ectopic pregnancy. Obstet 32. Kohn MA, Kerr K, Malkevich D, O’Neil N, Kerr MJ, Kaplan BC.
Gynecol 1981;58:156–61. Beta-human chorionic gonadotropin levels and the likelihood of
10. Mehta TS, Levine D, Beckwith B. Treatment of ectopic pregnancy: ectopic pregnancy in emergency department patients with
is a human chorionic gonadotropin level of 2,000 mIU/mL a reason- abdominal pain or vaginal bleeding. Acad Emerg Med 2003;10:
able threshold? Radiology 1997;205:569–73. 119–26.
11. Shaunik A, Kulp J, Appleby DH, Sammel MD, Barnhart KT. Utility 33. Seeber BE, Sammel MD, Guo W, Zhou L, Hummel A, Barnhart KT.
of dilation and curettage in the diagnosis of pregnancy of unknown Application of redefined human chorionic gonadotropin curves for
location. Am J Obstet Gynecol 2011;28(204):130.e1–6. the diagnosis of women at risk for ectopic pregnancy. Fertil Steril
12. Doubilet PM, Benson CB, Bourne T, Blaivas M. Diagnostic criteria 2006;86:454–9.
for nonviable pregnancy early in the first trimester. N Engl J Med 34. Morse CB, Sammel MD, Shaunik A, et al. Performance of human
2013;369:1443–51. chorionic gonadotropin curves in women at risk for ectopic preg-
13. Condous G, Kirk E, Lu C, et al. Diagnostic accuracy of varying nancy: exceptions to the rules. Fertil Steril 2012;97:101–6.
discriminatory zones for the prediction of ectopic pregnancy in 35. Silva C, Sammel MD, Zhou L, Gracia C, Hummel AC, Barnhart K.
women with a pregnancy of unknown location. Ultrasound Obstet Human chorionic gonadotropin profile for women with ectopic
Gynecol 2005;26:770–5. pregnancy. Obstet Gynecol 2006;107:605–10.
14. Goldstein SR, Wolfson R. Endovaginal ultrasonographic measure- 36. Barnhart KT. Ectopic pregnancy. N Engl J Med 2009;361:379–87.
ment of early embryonic size as a means of assessing gestational 37. Dillon KE, Sioulas VD, Sammel MD, et al. How and when human
age. J Ultrasound Med 1994;13:27–31. chorionic gonadotropin curves in women with an ectopic pregnancy
15. Barnhart K. Early pregnancy failure: beware of the pitfalls of mod- mimic other outcomes: differences by race and ethnicity. Fertil
ern management. Fertil Steril 2012;98:1061–5. Steril 2012;98:911–6.
16. Stovall TG, Ling FW, Carson SA, Buster JE. Serum progesterone 38. Seeber BE. What serial hCG can tell you, and cannot tell you, about
and uterine curettage in differential diagnosis of ectopic pregnancy. an early pregnancy. Fertil Steril 2012;98:1074–7.
Fertil Steril 1992;57:456–8. 39. Horne AW, McBride R, Denison FC. Normally rising hCG does not
17. McCord ML, Muram D, Buster JE, Arheart KL, Stovall TG, predict live birth in women presenting with pain and bleeding in
Carson SA. Single serum progesterone as a screen for ectopic preg- early pregnancy. Eur J Obstet Gynecol Reprod Biol 2011;156:
nancy: exchanging specificity and sensitivity to obtain optimal test 120–1.
performance. Fertil Steril 1996;66:513–6. 40. Griffey RT, Trent CJ, Bavolek RA, Keeperman JB, Sampson C,
18. Mol BW, Lijmer JG, Ankum WM, van der Veen F, Bossuyt PM. The Poirier RF. ‘‘Hook-like effect’’ causes false-negative point-of-care
accuracy of single serum progesterone measurement in the diagnosis urine pregnancy testing in emergency patients. J Emerg Med
of ectopic pregnancy: a meta-analysis. Hum Reprod 1998;13:3220–7. 2013;44:155–60.
19. Verhaegen J, Gallos ID, van Mello NM, et al. Accuracy of single 41. Gronowski AM, Cervinski M, Stenman UH, Woodworth A,
progesterone test to predict early pregnancy outcome in women Ashby L, Scott MG. False-negative results in point-of-care
Ectopic Pregnancy 827

qualitative human chorionic gonadotropin (hCG) devices due to 60. Dart RG, Kaplan B, Varaklis K. Predictive value of history and
excess hCGb core fragment. Clin Chem 2009;55:1389–94. physical examination in patients with suspected ectopic pregnancy.
42. Maccato ML, Estrada R, Faro S. Ectopic pregnancy with undetect- Ann Emerg Med 1999;33:283–90.
able serum and urine [beta]-hCG levels and detection of [beta]-hCG 61. Mol BW, Van der Veen F, Bossuyt PM. Symptom-free women at
in the ectopic trophoblast by immunocytochemical evaluation. Ob- increased risk of ectopic pregnancy: should we screen? Acta Obstet
stet Gynecol 1993;81:878–9. Gynecol Scand 2002;81:661–72.
43. Nerenz RD, Song H, Gronowski AM. Screening method to evaluate 62. Latchaw G, Takacs P, Gaitan L, Geren S, Burzawa J. Risk factors
point-of-care human chorionic gonadotropin (hCG) devices for sus- associated with the rupture of tubal ectopic pregnancy. Gynecol
ceptibility to the hook effect by hCG b core fragment: evaluation of Obstet Invest 2005;60:177–80.
11 devices. Clin Chem 2014;60:667–74. 63. Saxon D, Falcone T, Mascha EJ, Marino T, Yao M, Tulandi T. A
44. Sheele MJ, Bernstein R, Counselman FL. A ruptured ectopic study of ruptured tubal ectopic pregnancy. Obstet Gynecol 1997;
pregnancy presenting with a negative urine pregnancy test. Case 90:46–9.
Rep Emerg Med 2016;2016. 64. Galstyan K, Kurzel RB. Serum beta-hCG titers do not predict
45. Brennan DF, Kwatra S, Kelly M, Dunn M. Chronic ectopic preg- ruptured ectopic pregnancy. Int J Fertil Womens Med 2005;51:14–6.
nancy—two cases of acute rupture despite negative bhCG. J Emerg 65. Bartz D, Greenberg JA. Sterilization in the United States. Rev
Med 2000;19:249–54. Obstet Gynecol 2008;1:23–32.
46. Bedi DG, Moeller D, Fagan CJ, Winsett MZ. Chronic ectopic preg- 66. Cleland K, Raymond EG, Westley E, Trussell J. Emergency contra-
nancy. A comparison with acute ectopic pregnancy. Eur J Radiol ception review: evidence-based recommendations for clinicians.
1987;7:46–8. Clin Obstet Gynecol 2014;57:741.
47. Talavera MD, Horrow MM. Chronic ectopic pregnancy. J Diagn 67. Skjeldestad FE. How effectively do copper intrauterine devices pre-
Med Sonogr 2008;24:101–3. vent ectopic pregnancy? Acta Obstet Gynecol Scand 1997;76:684–90.
48. Kovalevskaya G, Birken S, Kakuma T, O’Connor JF. Early 68. Zhang J, Li C, Zhao WH, et al. Association between levonorgestrel
pregnancy human chorionic gonadotropin (hCG) isoforms emergency contraception and the risk of ectopic pregnancy: a multi-
measured by an immunometric assay for choriocarcinoma-like center case-control study. Sci Rep 2015;5:8487.
hCG. J Endocrinol 1999;161:99–106. 69. Gemzell-Danielsson K, Rabe T, Cheng L. Emergency contracep-
49. Madersbacher S, Berger P. Antibodies and immunoassays. Methods tion. Gynecol Endocrinol 2013;29(suppl 1):1–14.
2000;21:41–50. 70. Pereira PP, Cabar FR, Raiza LC, Roncaglia MT, Zugaib M. Emer-
50. Cervinski MA, Lockwood CM, Ferguson AM, et al. Qualitative point- gency contraception and ectopic pregnancy: report of 2 cases.
of-care and over-the-counter urine hCG devices differentially detect the Clinics 2005;60:497–500.
hCG variants of early pregnancy. Clin Chim Acta 2009;406:81–5. 71. Levy DP, Jager M, Kapp N, Abitbol JL. Ulipristal acetate for emer-
51. Kamer SM, Foley KF, Schmidt RL, Greene DN. Analytical gency contraception: postmarketing experience after use by more
sensitivity of four commonly used hCG point of care devices. than 1 million women. Contraception 2014;89:431–3.
Clin Biochem 2015;48:448–52. 72. Trussell J, Hedley A, Raymond E. Ectopic pregnancy following
52. Greene DN, Schmidt RL, Kamer SM, Grenache DG, Hoke C, use of progestin-only ECPs. J Fam Plann Reprod Health Care
Lorey TS. Limitations in qualitative point of care hCG tests for 2003;29:249.
detecting early pregnancy. Clin Chim Acta 2013;415:317–21. 73. Von Hertzen H, Piaggio G, Peregoudov A, et al. Low dose mifepristone
53. Butler SA, Khanlian SA, Cole LA. Detection of early pregnancy and two regimens of levonorgestrel for emergency contraception: a
forms of human chorionic gonadotropin by home pregnancy test WHO multicentre randomised trial. Lancet 2002;360(9348):1803–10.
devices. Clin Chem 2001;47:2131–6. 74. Arowojolu AO, Okewole IA, Adekunle AO. Comparative evalua-
54. Ajibade F, Vijayanand R, Cunningham H. P16. 04: Negative urine tion of the effectiveness and safety of two regimens of levonorges-
HCG and ruptured ectopic pregnancy (EP). Ultrasound Obstet trel for emergency contraception in Nigerians. Contraception 2002;
Gynecol 2013;42(suppl 1):172. 66:269–73.
55. Weckstein LN, Boucher AR, Tucker H, Gibson DN, 75. Ho PC, Kwan MS. A prospective randomized comparison of levo-
Rettenmaier MA. Accurate diagnosis of early ectopic pregnancy. norgestrel with the Yuzpe regimen in post-coital contraception.
Obstet Gynecol 1985;65:393–7. Hum Reprod 1993;8:389–92.
56. National Institute for Health and Clinical Excellence. Ectopic preg- 76. Bouyer J, Rachou E, Germain E, et al. Risk factors for extrauterine
nancy and miscarriage: diagnosis and initial management in early pregnancy in women using an intrauterine device. Fertil Steril 2000;
pregnancy of ectopic pregnancy and miscarriage (Clinical Guide- 74:899–908.
line 154). 2012. http://www.nice.org.uk/CG154. Accessed February 77. Barnhart KT, Sammel MD, Gracia CR, Chittams J, Hummel AC,
1, 2017. Shaunik A. Risk factors for ectopic pregnancy in women with symp-
57. Kaplan BC, Dart RG, Moskos M, et al. Ectopic pregnancy: prospec- tomatic first-trimester pregnancies. Fertil Steril 2006;86:36–43.
tive study with improved diagnostic accuracy. Ann Emerg Med 78. Shah R, Darawade S, Potdar J, Naphade P. Study of risk factors of
1996;28:10–7. ectopic pregnancy and its management. Glob J Res Anal 2016;5:X129.
58. Ayim F, Tapp S, Guha S, et al. Can risk factors, clinical history and 79. Hoover KW, Tao G, Kent CK. Trends in the diagnosis and treatment
symptoms be used to predict risk of ectopic pregnancy in women of ectopic pregnancy in the United States. Obstet Gynecol 2010;
attending an early pregnancy assessment unit? Ultrasound Obstet 115:495–502.
Gynecol 2016;48:656–62. 80. Shah N, Khan NH. Ectopic pregnancy: presentation and risk factors.
59. MacKintosh M, Elkattan E, Mahendran M, Basama F. An asymp- J Coll Physicians Surg Pak 2005;15:535–8.
tomatic ectopic pregnancy in a caesarean section scar. J Obstet 81. Majhi AK, Roy N, Karmakar KS, Banerjee PK. Ectopic pregnancy—
Gynaecol 2010;30:868–9. an analysis of 180 cases. J Indian Med Assoc 2007;105:308–10.
828 J. J. Robertson et al.

ARTICLE SUMMARY
1. Why is this topic important?
Ectopic pregnancy (EP) is an important cause of
morbidity and mortality in women of reproductive age.
Patients may present to the emergency department with
concerning signs and symptoms, but may not demonstrate
all of the classic findings of EP on initial evaluation and
workup.
2. What does this review attempt to show?
This review attempts to show patients with EP may not
always show the classic findings on history or evaluation,
including lack of risk factors, abnormal serum b–human
chorionic gonadotropin (b-hCG) levels or trends, and ul-
trasound (US) findings that do not correlate with the
classic discriminatory zone levels.
3. What are the key findings?
This article describes some key findings in that patients
may still require an US even if their serum b-hCG levels
are under the discriminatory zone (DZ). In addition, a
pregnancy of unknown location (PUL) may still be viable,
even if the serum b-hCG level is greater than the DZ.
Serum b-hCG levels do not rise and fall predictively,
and urine pregnancy tests can be negative in a rare number
of patients with EP. Not all patients will present with pain
or have any known risk factors. Finally, contraceptives do
not increase absolute risk of EP.
4. How is patient care impacted?
In patients without abdominal pain or risk factors, the
clinician may still consider EP in the differential diag-
nosis. In addition, he or she may also consider ordering
US in patients who have serum b-hCG levels under the
DZ because findings can still be visualized. Thus, an
earlier diagnosis of EP can be made. In addition, patients
who are stable with PUL and levels greater than the DZ
may simply require further monitoring as a viable preg-
nancy may still be possible. Finally, patients can also be
counseled that contraception does not increase the abso-
lute risk of EP.