Utility of dilation and curettage in the diagnosis of pregnancy of unknown location

Research

www. AJOG.org

GENERAL GYNECOLOGY

Utility of dilation and curettage in the diagnosis of pregnancy of unknown location Alka Shaunik, MD; Jennifer Kulp, MD; Dina H. Appleby, MS; Mary D. Sammel, ScD; Kurt T. Barnhart, MD, MSCE OBJECTIVE: We sought to determine utility of uterine evacuation for di-

agnosis of nonviable pregnancy of unknown location (PUL). STUDY DESIGN: We conducted a cohort study to assess the prevalence of ectopic pregnancy (EP), overall, and stratified by presenting signs and symptoms in women with a nonviable PUL. RESULTS: Of the 173 women, 66 (38%) had miscarriage (spontaneous

abortion [SAB]) and 107 (62%) had EP. When initial human chorionic gonadotropin (hCG) was ⬍2000 mIU/mL, the odds of an EP were greater (odds ratio, 4.32; 95% confidence interval, 2.04 –9.12). Demographic factors, obstetric history, and clinical presentation were not

useful in distinguishing between EP and SAB. Pre-evacuation hCG increase had strong trend association with EP (odds ratio, 2.14; 95% confidence interval, 0.98 – 4.68). A ⬎30% fall in postcurettage hCG was suggestive, but was not a diagnostic indicator of SAB. CONCLUSION: Uterine evacuation is a useful diagnostic aid for women

with nonviable PUL. Nondiagnostic ultrasound findings and absolute and serial hCG values are associated with, but do not accurately predict final diagnosis. Key words: dilation and curettage, ectopic pregnancy, pregnancy of unknown location

Cite this article as: Shaunik A, Kulp J, Appleby DH, et al. Utility of dilation and curettage in the diagnosis of pregnancy of unknown location. Am J Obstet Gynecol 2011;204:130.e1-6.

W

omen at risk of ectopic pregnancy (EP) may present with pain and bleeding in the first trimester of pregnancy. The early diagnosis of an EP in these women reduces the risk of a ruptured EP (a leading cause of maternal mortality that accounts for 10% of all maternal deaths) and improves the fu-

From the Department of Obstetrics and Gynecology (Drs Shaunik and Barnhart), University of Pennsylvania School of Medicine and the Center for Clinical Epidemiology and Biostatistics (Ms Appleby and Drs Sammel and Barnhart), University of Pennsylvania, Philadelphia, PA; and the Department of Reproductive Medicine (Dr Kulp), Yale University, New Haven, CT. Presented at the 64th Annual Meeting of the American Society for Reproductive Medicine, San Francisco, CA, Nov. 8-12, 2008. Received July 12, 2010; revised Sept. 7, 2010; accepted Nov. 5, 2010. Reprints: Kurt T. Barnhart, MD, MSCE, Department of Obstetrics and Gynecology, University of Pennsylvania Medical Center, 3701 Market St., Suite 810, Philadelphia, PA 19104. [email protected]. Supported by Grant no. R01-HD036455 from the National Institutes of Health. 0002-9378/$36.00 © 2011 Published by Mosby, Inc. doi: 10.1016/j.ajog.2010.11.021

130.e1

ture fertility outcomes in 6% of the women who are more likely to be diagnosed with a future EP.1-5 Currently, the diagnosis of an EP requires multiple office visits, serial blood tests for up to a 6-week period, multiple ultrasound examinations, and may entail surgical procedures such as uterine curettage and laparoscopy before a definitive diagnosis can be established. In practice, ultrasound scans can help predict the location of a pregnancy in 6992% of patients seen in specialty centers.6-9 The concept of a discriminatory zone (DZ) with regard to human chorionic gonadotropin (hCG) level is intimately linked to the accuracy of ultrasound and ranges from 1500-2500 IU/L.10 All cases with a nondiagnostic ultrasound (the absence of visualization of either an intrauterine or extrauterine pregnancy with a transvaginal ultrasound examination) are defined as pregnancy of unknown location (PUL).11 A total of 50-70% PUL resolve spontaneously, about 30% are diagnosed as an intrauterine pregnancy (of which 2/3 are viable intrauterine pregnancies), and almost 7-20% are subsequently diagnosed as EP.12 Controversy exists in the management of women with a nonviable PUL. Uterine evacuation may be used to distinguish an intrauterine pregnancy from an EP.9

American Journal of Obstetrics & Gynecology FEBRUARY 2011

Other authors have suggested that persistent PULs should be treated as ultrasonically missed EP.1 A presumed diagnosis of an EP is often made when a woman has no evidence of an IUP with an hCG above a DZ, or when serial hCG values have plateaued. We have previously reported that up to 40% of women with such a presumed diagnosis of EP actually have a nonviable intrauterine gestation.9 If distinction is not made between these 2 diagnoses, women with a spontaneous abortion (SAB) (miscarriage) may be treated as if they had and an EP. Given improved resolution of ultrasound machines in the past decade, the aim of this study was to validate these findings in a new cohort of women, presenting with first-trimester PUL and pain, bleeding, or both. The objective of this study to determine utility of uterine evacuation for diagnosis of nonviable PUL and evaluate if characteristics at initial presentation (clinical symptoms, historical variables, laboratory results of hCG values, and ultrasound examination) can aid in distinguishing between a woman with an EP or a miscarriage.

M ATERIALS AND M ETHODS Approval to conduct this study was obtained from the Institutional Review

General Gynecology

www.AJOG.org

TABLE 1

Demographic and obstetric history Variable Age, y

a

Miscarriage n ⴝ 66

Ectopic n ⴝ 107

29.4 ⫾ 6.6

29.9 ⫾ 7.0

P value .64

.............................................................................................................................................................................................................................................. b

Race,

.43

.....................................................................................................................................................................................................................................

Caucasian

5 (7.5)

8 (7.5)

.....................................................................................................................................................................................................................................

African American

39 (54.5)

48 (49.5)

Other

1 (2.2)

5 (8.2)

Gravidity

3 [1-5]

3 [1-4]

.19

Parity

1 [0-2]

1 [0-2]

.42

1 [0-2]

.54

..................................................................................................................................................................................................................................... .............................................................................................................................................................................................................................................. b .............................................................................................................................................................................................................................................. b .............................................................................................................................................................................................................................................. b

No. of live births

1 [0-2]

..............................................................................................................................................................................................................................................

History of spontaneous miscarriage

11 (16.7%)

18 (16.8%)

.86

History of voluntary interruption of pregnancy

15 (22.3%)

21 (19.3%)

.42

4 (6.0%)

6 (5.6%)

.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................

History of ectopic pregnancy

1.0

.............................................................................................................................................................................................................................................. a

Mean ⫾ SD, P value calculated via Student t test; n (percent of women with outcome who have given risk factor), P value calculated via Fisher’s exact test or ␹2 test. b

Shaunik. Uterine evacuation and diagnosis of pregnancy of unknown location. Am J Obstet Gynecol 2011.

Board of the University of Pennsylvania. A database is maintained at the University of Pennsylvania of all women in the first trimester of pregnancy (with a positive pregnancy test result or history of a missed period) who present to the emergency department with pain, bleeding, or both. Clinical and demographic data were entered directly into the computerized database by clinical staff caring for the patient. The population for this study included women with a nonviable PUL from De-

cember 2003 through July 2007. Specifically, women evaluated were clinically stable with an initial hCG level ⱖ2000 mIU/mL at presentation and a nondiagnostic ultrasound, or had an abnormal rise/fall/plateau of serial hCG levels with the initial hCG value ⬍2000 mIU/mL. During the study period, it was standard clinical practice for clinicians to perform a uterine evacuation prior to medical treatment for a presumed EP. The technique for uterine evacuation was at the discretion of the treating physician. The

TABLE 2

Characteristics of clinical presentation Miscarriage n ⴝ 66

Ectopic n ⴝ 107

P value

Length of amenorrhea, days

49.8 ⫾ 15.3

43.2 ⫾ 21.4

.04

Pain as chief symptom at presentation

34 (66.7)

58 (65.9)

.93

Bleeding as chief symptom at presentation

43 (86.0)

68 (81.0)

.45

Variable a

.............................................................................................................................................................................................................................................. b,c .............................................................................................................................................................................................................................................. c ..............................................................................................................................................................................................................................................

Bleeding severity

procedure performed was either a manual vacuum aspiration or electric vacuum aspiration with general anesthesia or conscious sedation. The techniques of these procedures have been described previously.13,14 Potential predictors of clinical outcome were identified from the medical and surgical history, clinical presentation, and diagnostic tests. History included patient age, race, gravidity, parity, number of live births, number of SAB, elective abortions, and previous EP. Findings at presentation analyzed as predictors included length of amenorrhea, amount of bleeding, pain, and hCG level. Bleeding was categorized as none, mild, moderate, or severe (no bleeding and less than, equal to, or more than regular menstrual bleeding, respectively) as reported by the patient. Pain was self-reported. Women were followed in the clinical database until they were definitively diagnosed with an EP or a nonviable intrauterine pregnancy. Diagnosis of a nonviable intrauterine pregnancy was confirmed by report of histopathology of products of conception obtained from uterine evacuation or postoperative resolution of serum hCG levels (hCG levels were followed serially until ⬍5 mIU/ mL).15 The diagnosis of an EP was confirmed by the presence of chorionic villi in the fallopian tube, by visualizing an extrauterine gestational sac (with yolk sac or embryonic cardiac activity) with ultrasound, or by postoperative rise in hCG level after dilation and evacuation concomitant with no evidence of chorionic villi in endometrial curettage samples (ie, nonvisualized EP).15 The majority of postoperative hCG values were drawn on days 1 or 2 following the procedure. The values presented have been normalized to 1 day changes.

.83

.....................................................................................................................................................................................................................................

None

7 (30.4%)

16 (69.6%)

Mild

30 (35.3%)

55 (64.7%)

9 (45.0%)

11 (55.0%)

1 (33.3%)

2 (66.7%)

..................................................................................................................................................................................................................................... .....................................................................................................................................................................................................................................

Moderate

.....................................................................................................................................................................................................................................

Severe

.............................................................................................................................................................................................................................................. a

Research

Mean ⫾ SD, P value calculated via Student t test; n (percent of women with outcome who have given risk factor), P value calculated via Fisher’s exact test or via ␹2 test; c A subject could have chief symptom of pain, bleeding, or both–therefore percentages may be ⬎100%. b

Shaunik. Uterine evacuation and diagnosis of pregnancy of unknown location. Am J Obstet Gynecol 2011.

Statistical analysis Comparisons between groups (women with a final diagnosis of EP vs nonviable intrauterine pregnancies) were performed using a Student t test for continuous variables. For variables with nonnormal distributions, medians and interquartile range are reported and the Wilcoxon nonparametric test was uti-

FEBRUARY 2011 American Journal of Obstetrics & Gynecology

130.e2

Research

General Gynecology

lized to determine statistical significance. The ␹2 tests were used to compare categorical variables and Fisher’s exact test was utilized when cell counts were ⱕ5. Statistical analysis was performed using software (SAS; SAS Institute Inc, Cary, NC) and a 2-tailed P value ⬍ .05 was considered statistically significant.

www.AJOG.org

TABLE 3

Final diagnosis of women with specific ultrasound findings Variable

Miscarriage

Ectopic

P value ⬍ .001

a

Initial ultrasound impression

..............................................................................................................................................................................................................................................

Likely nonviable intrauterine gestation

6 (60.0)

4 (40.0)

Likely ectopic pregnancy

3 (10.7)

25 (89.3)

48 (43.2)

63 (56.8)

.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................

Nondiagnostic

R ESULTS A total of 173 women were included in the study. Of these, 66 (38%) women were ultimately diagnosed with SAB and 107 (62%) with an EP. None of the patients had serious adverse outcomes. The first step was to describe and compare the general characteristics of women based on final diagnosis. The demographic information and obstetric histories of the 2 groups are presented in Table 1 and are stratified by ultimate outcome. The mean age, race, and obstetric history were similar in both groups. Obstetric history data (gravidity, parity, history of live birth, past miscarriage, previous voluntary interruption of pregnancy, and history of EP) were not associated with final diagnosis of EP or miscarriage. The clinical characteristics of patients are presented in Table 2. Length of amenorrhea was found to be significantly longer in women with miscarriage. Clinical presentation factors including pain, presence or grade of bleeding, and time to diagnosis were not associated with final location of a PUL. The prevalence of EP in women presenting with pain (63%; 58/92) or bleeding (61%; 68/111) as the chief symptom was not significantly different from that of the entire population. The next step in our analysis was to assesses the prevalence of outcome (EP or SAB) in each risk factor, and to determine if risk factors could be used to predict outcome. Patients were included in this cohort because their initial ultrasound impression was not definitively diagnostic. However, nondefinitive impression at the time of the ultrasound findings was predictive of outcome as shown in Table 3. Among those patients whose ultrasound was labeled suspicious for EP, 89% (25/28) were eventually di130.e3

.............................................................................................................................................................................................................................................. a

⬍ .001

Presence of adnexal mass

.....................................................................................................................................................................................................................................

Not present

55 (45.5)

66 (54.5)

Possibly present

2 (28.6)

5 (71.4)

Present

2 (8.3)

22 (91.7)

..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... .............................................................................................................................................................................................................................................. a

Free fluid in cul-de-sac

.04

.....................................................................................................................................................................................................................................

Absent

28 (50.0)

28 (50.0)

Present

26 (32.1)

55 (67.9)

..................................................................................................................................................................................................................................... ..............................................................................................................................................................................................................................................

Presence of intrauterine gestational saca

.002

.....................................................................................................................................................................................................................................

Not present

42 (33.6)

83 (66.4)

Possibly present

14 (70.0)

6 (30.0)

..................................................................................................................................................................................................................................... .....................................................................................................................................................................................................................................

Present

6 (66.7)

3 (33.3)

.............................................................................................................................................................................................................................................. b

Endometrial thickness, mm

15.0 [9.0-36.0]

10.5 [6.0-18.0]

.01

.............................................................................................................................................................................................................................................. a

n (percent of women with outcome who have given risk factor), P value calculated via Fisher’s exact test or via ␹ test; Mean ⫾ SD, P value calculated via Student t test. 2

b

Shaunik. Uterine evacuation and diagnosis of pregnancy of unknown location. Am J Obstet Gynecol 2011.

agnosed with EP; and 60% (6/10) of patients whose ultrasound indicated miscarriage were eventually diagnosed with SAB. In addition, certain specific ultrasound findings were predictive of the final outcome. Presence of an adnexal mass (P ⬍ .001) and the presence of free fluid in the cul-de-sac (P ⫽ .04) were noted more frequently (but not exclusively) in women ultimately diagnosed with an EP. A suspected intrauterine gestational sac was observed more frequently (but not exclusively) in women ultimately diagnosed with a nonviable intrauterine pregnancy (P ⫽ .002). Endometrial thickness was greater in women with a miscarriage (P ⫽ .01), but there was considerable overlap in the values of interquartile range for both diagnoses. The results of comparisons of initial hCG values are presented in Table 4. Both mean (5698 vs 996 mIU/mL, P ⬍ .001) and median (819 vs 344 mIU/mL, P ⬍ .001) hCG values were higher for women diagnosed with a miscarriage

American Journal of Obstetrics & Gynecology FEBRUARY 2011

compared to those diagnosed with an EP, respectively. There was a strong trend that women with a rising hCG were more likely to have an EP than SAB (P ⫽ .09). The percentage of women with an EP or SAB as final diagnosis differed depending on whether the initial hCG value was above or below the DZ (2000 mIU/mL). If the initial hCG value was below the DZ there was a much greater likelihood of diagnosis of an EP (93/113, 70%), P value ⬍ .001; odds ratio for EP, 4.32 (95% confidence interval [CI], 2.04 –9.12). When stratified by initial hCG, the pattern of preoperative hCG values was not predictive of an outcome of either EP or SAB when hCG value was either above or below the DZ. Results of initial hCG values after uterine evacuation are presented in Table 5. Postuterine evacuation hCG values were significantly different in the SAB and EP population (P ⬍ .001). Any rise or plateau in hCG values was noted only in pa-

General Gynecology

www.AJOG.org

TABLE 4

Final diagnosis of women based on hCG before uterine evacuation Variable a

Median initial hCG (mIU/mL)

Miscarriage

Ectopic

P value

819 [268-3684]

344 [142-926]

⬍ .001

.............................................................................................................................................................................................................................................. b

Overall serial preoperative hCG trend

.09

..............................................................................................................................................................................................................................................

Rise ⬎10%

14 (25.5)

41 (74.5)

Plateau (⫾10%)

27 (42.2)

37 (57.8)

Fall ⬎10%

16 (44.4)

20 (55.6)

.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. b

⬍ .001

Stratification on DZ

..............................................................................................................................................................................................................................................

hCG ⬍2000 mIU/mL

40 (30.1)

93 (69.9)

hCG ⱖ2000 mIU/mL

26 (65.0)

14 (35.0)

.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. b

hCG ⬍2000 mIU/mL

.38

..............................................................................................................................................................................................................................................

Rise ⬎10%

11 (22.5)

38 (77.5)

Plateau (⫾10%)

12 (30.8)

27 (69.2)

Fall ⬎10%

11 (36.7)

19 (63.3)

.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. b

hCG ⱖ2000 mIU/mL

.48

..............................................................................................................................................................................................................................................

Rise ⬎10%

3 (50.0)

3 (50.0)

15 (60.0)

10 (40.0)

5 (83.3)

1 (16.7)

..............................................................................................................................................................................................................................................

Plateau (⫾10%)

..............................................................................................................................................................................................................................................

Fall ⬎10%

.............................................................................................................................................................................................................................................. b

Pain

.84

..............................................................................................................................................................................................................................................

hCG ⬍2000 mIU/mL

22 (31.9)

47 (68.1)

hCG ⱖ2000 mIU/mL

12 (52.2)

11 (47.8)

.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. b

Bleeding

.52

..............................................................................................................................................................................................................................................

hCG ⬍2000 mIU/mL

23 (28.1)

59 (71.9)

hCG ⱖ2000 mIU/mL

20 (69.0)

9 (31.0)

.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................

DZ, discriminatory zone; hCG, human chorionic gonadotropin. a

Mean ⫾ SD, P value calculated via Student t test; b N (percent of women with outcome who have given risk factor), P value calculated via Fisher’s exact test or via ␹2 test.

Shaunik. Uterine evacuation and diagnosis of pregnancy of unknown location. Am J Obstet Gynecol 2011.

tients with an EP, while a decrease in hCG values was not observed exclusively in either EP or SAB patients. Of those patients who experienced a fall in hCG values after uterine evacuation, 19/29 (65.5%) were eventually diagnosed with SAB, and the remainder (10/29; 34.5%) were diagnosed with an EP. A fall in hCG values ⬎50% was observed in 13/14 (93%) women with SAB (P ⬍ .0001). If the initial hCG value (before the evacuation) was below the DZ, a decline postprocedure was noted in 17/25 (68%) patients with SAB and in 8/25 (32%) patients with an EP (P ⬍ .001). A decline in hCG value ⬎50% was seen in significantly more women with SAB (12/13, 92%; P value ⬍ .0001) compared to an EP. A fall in hCG levels ⬎30% was highly

predictive of SAB (12/17, 71%; P ⫽ .002). The sample size was too small to detect any statistical difference between outcomes when the initial hCG values was above the DZ.

C OMMENT “Pregnancy of unknown location” is a descriptive term rather than a diagnosis.11 In women with a PUL, definitive diagnosis is cumbersome and often requires multiple office visits, serial blood tests, ultrasound examinations, and surgical procedures over up to a 6-week period. The main objective of our study was to assess the utility of uterine evacuation and assess if patient characteristics may help predict EP in this population.

Research

These data reaffirm some important clinical observations. We note that neither demographic and obstetric history, nor clinical aspects of initial presentation adequately discriminate women ultimately diagnosed with SAB from those diagnosed with an EP. Moreover, we have demonstrated that a substantial proportion of the population of women with a PUL will ultimately be diagnosed with a nonviable intrauterine pregnancy (in a cohort of women distinct from our previous study9). Combining these data with those previously published9 demonstrates that overall 47.4% (135/285; 95% CI, 41.6 –53.2%) of this patient population will have SAB. Of the women who have an hCG above the DZ, 68.9% (93/135; 95% CI, 61.1–76.7%) have a nonviable gestation (not an EP). Of women with an initial hCG below the DZ, 31.1% (42/135; 95% CI, 24.6 –37.7%) will have a nonviable gestation (and not an EP). Therefore, presumptive treatment with methotrexate (MTX) will result in inappropriate treatment of almost half of women with a “presumed EP.” It is recognized that if uterine evacuation can be avoided, then it may be possible to simplify diagnostic protocols for women with a nonviable PUL. Therefore, we assessed whether certain variables could be used as aids to ultimately diagnose women with PUL. These data demonstrate that although preoperative findings may assist in the prediction of the anatomical location of the pregnancy, they are not definitive of diagnosing either SAB or EP in women with a nonviable PUL. Of note, neither chief symptom (pain or bleeding) nor severity of bleeding at the time of presentation was associated with ultimate diagnosis of EP or SAB outcome. Women with a definitive diagnostic ultrasound report for an intrauterine gestation (either viable or nonviable) or EP were not included in this cohort. Even in women with nondiagnostic ultrasounds, there were ultrasound findings that were significantly associated with the final clinical diagnosis of the PUL. However, there were also a substantial number of false-positive results. In total, 11% (3/28) women suspected to have an EP based on preliminary ultra-

FEBRUARY 2011 American Journal of Obstetrics & Gynecology

130.e4

Research

General Gynecology

www.AJOG.org

TABLE 5

Final diagnosis of women based on postuterine evacuation hCG changes Variable

Miscarriage n ⴝ 66 (38.2%)

Ectopic n ⴝ 107 (61.8%)

P value ⬍ .001

Overall

.....................................................................................................................................................................................................................................

Rise ⬎10%

0 (0)

11 (100)

Plateau (⫾10%)

0 (0)

7 (100)

..................................................................................................................................................................................................................................... .....................................................................................................................................................................................................................................

Fall ⬎10%

19 (65.5)

10 (34.5)

..............................................................................................................................................................................................................................................

Stratification by DZ

..............................................................................................................................................................................................................................................

hCG ⬍2000 mIU/mL

⬍ .001

.....................................................................................................................................................................................................................................

Rise ⬎10%

0 (0)

10 (100)

0 (0)

6 (100)

.....................................................................................................................................................................................................................................

Plateau (⫾10%)

.....................................................................................................................................................................................................................................

Fall ⬎10%

17 (68.0)

8 (32.0)

..............................................................................................................................................................................................................................................

hCG ⱖ2000 mIU/mL

1.0

.....................................................................................................................................................................................................................................

Rise ⬎10%

0 (0)

Plateau (⫾10%)

0 (0)

1 (100)

Fall ⬎10%

2 (50.0)

2 (50.0)

1 (100)

..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... ..............................................................................................................................................................................................................................................

hCG fall ⬎30%

⬍ .002

..............................................................................................................................................................................................................................................

hCG ⬍2000 mIU/mL

12 (75)

4 (25)

..............................................................................................................................................................................................................................................

hCG ⱖ2000 mIU/mL

1 (100)

0 (0)

..............................................................................................................................................................................................................................................

hCG fall ⬎50%

⬍ .0001

..............................................................................................................................................................................................................................................

hCG ⬍2000 mIU/mL

12 (92.3)

hCG ⱖ2000 mIU/mL

1 (100)

1 (7.7)

..............................................................................................................................................................................................................................................

0 (0)

..............................................................................................................................................................................................................................................

Values are row n (percent). All P values calculated via Fisher’s exact test. DZ, discriminatory zone; hCG, human chorionic gonadotropin. Shaunik. Uterine evacuation and diagnosis of pregnancy of unknown location. Am J Obstet Gynecol 2011.

sound findings were ultimately diagnosed with SAB. Similarly, 40% (4/10) women suspected to have SAB based on preliminary ultrasound findings ultimately were diagnosed with an EP. These observations have direct implications on clinical care. Of women noted to have an adnexal mass on ultrasound, 8% were found to have SAB. Of those noted to have free fluid in the cul-de-sac, 32% had SAB. Conversely, 33% of women noted to have an early gestational sac were ultimately diagnosed with an EP and therefore the “early sac” was actually a pseudosac. Diagnosis by ultrasound can be optimized with the use of high-quality ultrasound equipment, experienced sonographers, and prior knowledge of risk factors and symptoms of an EP. In this study, ultrasound scans were performed by a radiologist or technician (under the supervision of an attending physician) staffing an inner city emergency depart130.e5

ment. This is an accepted standard of care and is similar to clinical practice in other medical centers. It is imperative that the radiologic findings are correlated to the clinical situation. We noted that initial hCG value can aid in discrimination of ultimate diagnosis with a 4-fold greater odds of diagnosis of an EP when the hCG is below DZ (odds ratio for EP, 4.32; 95% CI, 2.04 – 9.12). In some cases ⬎1 hCG value was obtained prior to uterine evacuation. We also noted that the pattern of the serial hCG may also aid in discrimination of location of the pregnancy. For example, there was a strong trend toward a greater number of women with an increase in hCG values being diagnosed with an EP. This may represent a subgroup of women who, once a viable gestation has been ruled out, may need intervention with uterine evacuation sooner to avoid rupture. This trend was not as strong

American Journal of Obstetrics & Gynecology FEBRUARY 2011

when data were stratified based on being above or below the DZ. As expected, and by definition, the pattern of postoperative serial hCG was highly associated with diagnosis. When the decline in hCG value is ⬎50% in 48 hours, then 93% of the time the hCG is cleared from the serum and the pregnancy can be considered to have resolved. While any decline in hCG ⬎10% was statistically associated with SAB, some women with an initial postoperative decline of 30% were ultimately diagnosed with an EP. There is clinical merit in making a definitive diagnosis of either an EP or SAB in women with a nonviable PUL. This distinction is between an EP and a failed intrauterine pregnancy. Prior studies have demonstrated that there is no reduction in complications or financial benefit to presumptively treating women with MTX.16 Moreover, presumptive diagnosis and treatment with MTX will limit any prognostic information regarding future reproductive potential. A woman with multiple EP may be best assisted with in vitro fertilization for tubal factor infertility. A woman with multiple miscarriages may be best assisted with a workup for recurrent pregnancy loss. In conclusion, these data suggest that in treatment of women with a high hCG and a nondiagnostic ultrasound, or a plateau in serial hCG values presumptively for an EP would have led to an incorrect diagnosis and treatment in up to 50% of women. Presumptive treatment would have led to unnecessary administration of a chemotherapeutic agent to many women. A uterine evacuation and subsequent follow-up of hCG concentrations accomplishes both confirmation of the location of the PUL as well as treatment of those with SAB. Unfortunately, although clinical signs and symptoms, nondiagnostic ultrasound findings, and preoperative hCG values are associated with final outcome, they are not sufficiently discriminatory to eliminate the need for uterine evacuation in most cases. f REFERENCES 1. Kirk E, Bourne T. Pregnancy of unknown location. Obstet Gynaecol Reprod Med 2008;19:3.

www.AJOG.org 2. Centers for Disease Control and Prevention. Ectopic pregnancy–United States, 1990-1992. MMWR Morb Mortal Wkly Rep 1995;44:46-8. 3. Phillips RS, Tuomala RE, Feldblum PJ Schachter J, Rosenberg MJ, Aronson MD. The effect of cigarette smoking, Chlamydia trachomatis infection, and vaginal douching on ectopic pregnancy. Obstet Gynecol 1992;79:85-90. 4. Abbott J, Emmans LS, Lowenstein SR. Ectopic pregnancy: ten common pitfalls in diagnosis. Am J Emerg Med 1990;8:515-22. 5. Tenore JL. Ectopic pregnancy. Am Fam Physician 2000;61:1080-8. 6. Condous G, Kirk E, Lu C, Van Huffel S, Timmerman D, Bourne T. The accuracy of transvaginal ultrasonography for the diagnosis of ectopic pregnancy prior to surgery. Hum Reprod 2005;20:1404-9. 7. Cacciatore B. Stenman UH, Ylostalo P. Diagnosis of ectopic pregnancy by vaginal ultrasonography in combination with a discrimina-

General Gynecology tory serum hCG level of 1000 IU/L (IRP). Br J Obstet Gynaecol 1990;97:904-8. 8. Shalev E, Yarom I, Bustan M, Weiner E, BenShlomo I. Transvaginal sonography as the ultimate diagnostic tool for the management of ectopic pregnancy: experience with 840 cases. Fertil Steril 1998;69:62-5. 9. Barnhart KT, Katz I, Hummel A, Gracia CR. Presumed diagnosis of ectopic pregnancy. Obstet Gynecol 2002;100:505-10. 10. Barnhart KT, Simhan H, Kamelle SA. Diagnostic accuracy of ultrasound above and below the beta-hCG discriminatory zone. Obstet Gynecol 1999;94:583-7. 11. Sagili H, Mohamed K. Pregnancy of unknown location: an evidence based approach to management. Obstetrician Gynaecologist 2008;10:224-30. 12. Kirk E, Bourne T. Predicting outcomes in pregnancies of unknown location. Womens Health (Lond Engl) 2008;4:491-5.

Research

13. Zhang J, Gilles JM, Barnhart K, et al. A comparison of medical management with misoprostol and surgical management for early pregnancy failure. N Engl J Med 2005;353:761-9. 14. Edwards S, Tureck R, Fredrick M, Huang X, Zhang J, Barnhart K. Improved quality of life after manual vacuum aspiration compared with electric vacuum aspiration for early pregnancy loss. J Womens Health 2007;16:1429-36. 15. Barnhart KT, van Mello N, Bourne T, et al. Pregnancy of unknown location: a consensus statement of nomenclature, definitions and outcome. Fertil Steril 2010 [Epub ahead of print] PMID: 20947073. 16. Ailawadi M, Lorch SC, Barnhart KT. Cost effectiveness of presumptively medically treating women at risk for ectopic pregnancy compared with first performing a dilation and curettage. Fertil Steril 2005;83:376-82.

FEBRUARY 2011 American Journal of Obstetrics & Gynecology

130.e6