American Journal of Emergency Medicine (2012) 30, 399-404

Original Contribution

Fetal loss in symptomatic first-trimester pregnancy with documented yolk sac Intrauterine pregnancy^?

Mary Josephine Hessert DO, MPH?, Michael Juliano MD

Department of Emergency Medicine, Naval Medical Center Portsmouth, Portsmouth, VA USA

Received 25 August 2010; revised 14 December 2010; accepted 14 December 2010

Abstract

Background: The possibility of spontaneous miscarriage is a common concern among pregnant women in the emergency department (ED).

Objective: This study sought to determine Fetal outcomes for women following ED evaluation for first- trimester abdominopelvic pain or vaginal bleeding who had an intrauterine pregnancy on ultrasound before a visible fetal pole (“yolk sac IUP”).

Methods: A retrospective chart review of consecutive ED charts from December 2005 to September 2006 identified patients with a yolk sac IUP. Demographic data, obstetric/gynecologic history, and presenting symptoms were obtained. Outcomes were determined via computerized records. Fetal loss was diagnosed by falling ?-human chorionic gonadotropin or pathology specimen. Live birth was diagnosed by viable fetus at 20-week ultrasound or delivery.

Results: A total of 131 patients were enrolled in this study. Of these, 14 were lost to follow-up (12%), leaving 117 patient encounters. Of the 117 women, 82 carried their pregnancies to at least 20-week gestation. Thirty-five patients miscarried. Fetal loss rate by chief complaint were as follows: 8 of 46 patients presenting with pain only, 14 of 34 presenting with vaginal bleeding only, and 13 of 37 with both vaginal bleeding and pain.

Conclusion: Seventy percent of women diagnosed with a yolk sac IUP in the ED carried their pregnancy to at least 20 weeks. The remaining women (30%) experienced fetal loss. Vaginal bleeding (with or without pain) increased the rate of fetal loss compared with women with pain only. These data will assist the emergency physician in counseling women with symptomatic first-trimester pregnancies.

Introduction

Background

Concern regarding the possibility of spontaneous miscar- riage is one of the most common reasons why women seek evaluation in the emergency department (ED) during their first

^? There was no financial support or grant for this research.

* Corresponding author. Tel.: +1 757 289 2848; fax: +1 757 953 0300.

E-mail address: [email protected] (M.J. Hessert).

trimester. According to the obstetrics and gynecology literature, it is estimated that as many as 20% to 25% of all clinically recognized pregnancies result in spontaneous miscarriage [1,2]. Once a viable fetus is diagnosed via ultrasound, however, the rate of spontaneous miscarriage subsequently decreases to 3% to 6% [3,4]. There is limited research on Fetal outcomes of women with a yolk sac intrauterine pregnancy (IUP before the appearance of a fetal pole). We sought to determine fetal outcomes of women diagnosed with a yolk sac IUP following ED presentation for abdominopelvic pain and/or vaginal bleeding during the first trimester.

0735-6757/$ – see front matter. doi:10.1016/j.ajem.2010.12.021

Importance

Symptomatic first-trimester pregnancies are a common presentation to the ED. Many EDs are now equipped with ultrasound capability. With this increased utilization of ultrasound technology via ultrasound-credentialed emer- gency physicians, more women are being sonographically evaluated in the ED to document an IUP [5]. In a low-risk patient (ie, without assisted reproductive techniques or pro- ovulatory medications), the heterotopic rate has been shown to be approximately 1 in 4000 to 1 in 8000 in the general population [6]. Therefore, an IUP generally means that a patient can be discharged with a low risk of Ectopic pregnancy. However, there is little prognostic information to give a patient with a documented yolk sac IUP regarding fetal outcome.

Goals of this investigation

We sought to determine longitudinal fetal outcomes in women who presented to the ED with vaginal bleeding or pain during their first trimester who had a yolk sac IUP diagnosed during their patient encounter. Our aim was to determine the rates of fetal loss following documentation of a yolk sac IUP. These data will assist the emergency physician with counseling women who have symptomatic first- trimester pregnancies. In addition, we sought to determine the risk of spontaneous abortion in women with only a yolk sac among 3 separate presenting symptoms (pain only, vaginal bleeding only, pain and vaginal bleeding together).

Materials and methods

This study was approved by the hospital’s institutional review board (CIP# NMCP.2007.0091).

Study design

retrospective medical record review of prospectively gathered data was performed from December 2005 through September 2006. Our hospital uses a computer-based charting system termed Composite Health Care System (CHCS). A query was performed of CHCS for the study time frame, generating a list of all patients seen in the ED. The query lists patients by chief complaint and final discharge diagnosis. Each patient encounter was evaluated for the presence of 1 or more of the following: pregnancy, abdominal pain, vaginal bleeding, IUP, threatened abortion, spontaneous abortion, complete abortion, incomplete abor- tion, missed abortion, and embryonic demise. The Emer- gency Treatment Record (ETR) was then reviewed for inclusion into the study. The ETR used in the ED is a unique documentation template for women who present with symptomatic first-trimester pregnancy (<=12 weeks estimated

gestational age [EGA] by last menstrual period). This template was created to standardize all patient encounters. Data collected included patient demographics, presenting symptoms, ultrasound data, and outcome of the pregnancy. All CHCS queries and subsequent data collection were performed by the primary investigator using a data collection sheet created for the study.

Setting

The hospital is a tertiary care-level hospital with an annual ED census of 75 000. The ED employs resident and staff physicians credentialed in transvaginal and transab- dominal ultrasound, with a monthly census of approximately 150 symptomatic first-trimester pregnancies (annual inci- dence approximately 2%).

Selection of participants

Criteria for inclusion were as follows: (1) positive pregnancy test result; (2) first trimester (<=12 weeks EGA via last menstrual period); (3) chief complaint of at least 1 of the following: vaginal bleeding, abdominal pain, pelvic pain, or back pain; and (4) documentation of intrauterine Gestational sac and yolk sac without a fetal pole via ultrasound evaluation (either in the ED or radiology department). All ED staff physicians are credentialed in the performance of transvaginal and Transabdominal ultrasound for the determination of an IUP. The ED uses a Sonosite TITAN portable ultrasound system with a 3-MHz transabdominal probe and 7-MHz transvaginal probe. All ED residents are required to perform 25 supervised (via the ED attending) Transvaginal ultrasounds and

25 supervised transAbdominal ultrasounds before they are allowed to scan patients independently. All scans performed independently by residents are presented to the ED attending before patient discharge for evaluation of adequacy and completeness. One hundred percent of the ultrasounds performed in the ED are then catalogued and reevaluated by an independent reviewer as part of quality assurance. If an ultrasound performed in the ED is indeterminate (ie, empty uterus or an empty gestational sac without the presence of a yolk sac), then serum ?–human chorionic gonadotropin levels are obtained. If the serum HCG is above the discriminatory zone of approximately 1500 IU/L (staff dependent), a formal ultrasound is obtained via the radiology department. The use of discriminatory zones for serum HCG levels to correlate with the expected appearance of an IUP is well documented [7]. A formal ultrasound can be obtained for any reason by an ED physician if there is a concern for ectopic pregnancy regardless of ED ultrasound findings or laboratory data. Our hospital has 24-hour capability to perform pelvic ultrasounds. All ultrasounds performed in the radiology department are read immediately by a radiology resident (after hours) with a preliminary report called to the ordering ED physician. Dictated reports are viewable the next day in CHCS.

Methods of measurement

The CHCS tracks all laboratory values obtained from a patient in addition to providing information regarding which clinic/department ordered the test. Patients entered into the study were considered to have miscarried if their quantita- tive serum HCG decreased by at least 50% after their evaluation in the ED or if the pathology report from tissue collected either in the emergency or gynecology department confirmed products of conception. A formal second- trimester ultrasound (performed in the radiology depart- ment) was used as an end point for successful pregnancy outcome. If the ultrasound demonstrated an EGA of 20 weeks or greater, no further investigation was undertaken. If the ultrasound demonstrated an EGA less than 20 weeks, then the patient’s future 20-week date was calculated and CHCS was queried again. Determination that the patient had been seen on or after their 20-week date via completed obstetrics/gynecology (Ob/Gyn) orders (ie, glucose toler- ance testing, repeat ultrasound evaluation, Medication orders, or documentation of a live birth) was made. Patients were considered lost to follow-up if there were no further entries into CHCS after their initial visit to the ED or no ultrasound results confirming a 20-plus-week EGA. Patients presenting more than once during the study time frame were included as a separate patient encounter only if their chief complaint changed.

Data collection and processing

The ETR for symptomatic first-trimester pregnancy was specially designed to facilitate documentation of the medical encounter. Demographic data, Ob/Gyn history, presenting symptoms, physical examination, Laboratory analysis, and results of ED ultrasound were prospectively recorded. Results from consecutive patients with symptomatic first- trimester pregnancies evaluated in the ED from December 2005 to September 2006 were entered onto 2 data collection sheets. One sheet was used for patient identification (name and social security number), and the second sheet was used for primary data collection. The data collection sheets were created to facilitate subsequent result analyses. All data collection was performed by the principle investigator.

Primary data analysis

SPSS (Chicago, IL) version 15 software was used for statistical analysis. Binary logistic regression analysis was used to predict patient outcome (miscarriage vs 20-plus- week EGA) with respect to chief complaint (vaginal bleeding with or without a pain complaint vs pain only). Interrater reliability of patient complaints and outcomes produced a ? score of 0.847 when 10% of all included ETRs were reviewed by an independent physician. Patients were divided into 3 arms: patients presenting with pain and vaginal

bleeding, patients presenting with pain only, and patients presenting with vaginal bleeding only. An omnibus ?2 test was computed to test for differences among groups in proportions of patients that miscarried. To isolate effects of vaginal bleeding, patients with vaginal bleeding and pain were compared with patients with pain only. Effects of pain were assessed by comparing patients with vaginal bleeding and pain to patients with vaginal bleeding only. These pairwise comparisons of groups yielded Yates corrected ?2 with 1 degree of freedom. An ? level of .05 was adopted for all tests.

Results

A total of 1162 patients presented during the study time frame with a first-trimester symptomatic pregnancy. Of these, 135 met the inclusion criteria and had a documented Intrauterine gestational sac and yolk sac without a visible fetal pole on ultrasound. Four of these were excluded because they presented twice during the same pregnancy. Fourteen of 131 women were lost to follow-up (12%), giving a final study population of 117 patient encounters. The mean Maternal age was 26 years, and the median age was 24 years (range, 15-42 years). Demographic data are shown in Table 1.

Thirty-five patients (30%) experienced a miscarriage, and 82 patients carried their pregnancy at least until their 20-week date (70%). In patients presenting with pain only, the rate of fetal loss was 17.4% (95% confidence interval

Table 1 Demographic data

Reviewed (117)		Lost (14)
Race
White	42 (35.6%)	6	(43%)
Black	41 (35%)	5	(36%)
Hispanic	5 (4.2%)	2	(14%)
Asian	5 (4.2%)	0	(0%)
Unknown	24 (21%)	1	(7%)
Age (y)
b20	15 (13%)	1	(7%)
20-24	44 (37.5%)	6	(43%)
25-29	29 (24.8%)	5	(36%)
30-34	17 (14.5%)	1	(7%)
>=35	12 (10.2%)	1	(7%)
Previous miscarriage
0	79 (67.5%)	10	(71.5%)
1	26 (22.2%)	2	(14.3%)
2	4 (3.4%)	1	(7.1%)
>=3	6 (5.1%)	1	(7.1%)
Unknown	2 (1.8%)	0	(0%)
Chief complaint
Vaginal bleeding	34 (29.1%)	4	(28.6%)
Abdominal pain	46 (39.3%)	2	(14.3%)
Both	37 (31.6%)	8	(57.1%)

Presentation (n)	Miscarriage; 95% CI		20	wk; 95% CI
Total (117)	35 (30.0%;	21.7%-38.3%)	82	(70.0%;	61.7%-78.3%)
Vaginal bleeding (71)	27 (38.0%;	26.7%-49.3%)	44	(62.0%;	50.7%-73.3%)
Bleeding only (34)	14 (41.2%;	24.6%-57.7%)	20	(58.8%;	42.3%-75.4%)
Bleeding and pain (37)	13 (35.1%;	19.8%-50.5%)	24	(64.9%;	29.5%-80.2%)
Pain only (46)	8 (17.4%;	6.4%-28.3%)	38	(82.6%;	71.7%-93.6%)

[CI], 6.4%-28.3%). Among patients who presented with vaginal bleeding only, the risk of fetal loss was 41.2% (95% CI, 24.6%-57.8%). Patients presenting with both pain and vaginal bleeding had a 35.1% chance of subsequent miscarriage (95% CI, 19.8%-50.5%). Vaginal bleeding (with or without pain) increased the rate of fetal loss (odds ratio, 2.9; 95% CI, 1.2-7.2) when compared with women presenting with pain only. Results are summarized in Table 2. Table 3 shows the rates of miscarriage in different age groups, with a trend toward higher miscarriage rates in extremes of age (b20 and N35 years). There was no significance in fetal outcome according to previous miscarriage (Table 4). Physical examination was significant for greater than or equal to 5 mL of blood corresponding with higher rate of miscarriage, as seen in Table 5.

Table 2 Outcome of pregnancy

Of the 131 emergency ultrasounds performed, 102 showed a yolk sac; thus, no further ultrasound was performed. Twenty-nine (22%) ED scans showed either an empty uterus or an empty gestational sac and were sent for a formal ultrasound in the radiology department that demon- strated a yolk sac IUP.

Discussion

This research was patient oriented in its design; and we believe that our study is the first of its kind, in which patient outcomes were generated after the visualization of a yolk sac IUP in the ED. We believe that our study was unique and adequate to determine these outcomes because of our relatively large symptomatic first-trimester pregnant patient population.

Stages of development and their Sonographic findings in the first trimester of pregnancy are well known [8]. There is a limited window of time, approximately 7 days, during the first trimester when only a yolk sac with no demonstrable

Table 3 Fetal outcome determined by age

fetal pole is visible on ultrasound [9]. Because of the brevity of this period, it is less common in the ED to see a yolk sac IUP vs a live IUP or an indeterminate ultrasound. There is minimal literature, especially ED-based research, regarding Pregnancy outcomes after diagnosis of a yolk sac IUP.

We agree with Kurtz et al [10] that the visualization of a yolk sac IUP does not predict pregnancy outcome; however, our study was designed to use Patient symptoms to aid in further prognostication. The study of Kurtz et al was smaller in number of subjects (76 with yolk sac IUP), and all patients were referred to the radiology department for ultrasound vs only 22% in our study [10].

Many authors have discussed first-trimester ultrasound as a useful tool for detection of anomalies and predicting outcomes of early pregnancy [7,10,11,12]. Gracia et al [12] specifically discuss bleeding as a risk factor for spontane- ous abortion but do not further stratify the risk with ultrasound findings. Weiss et al [13] further evaluated pregnancy risks with the amount of bleeding and found that the type of adverse outcome varies with heavy, light, and no bleeding. Neither of these studies combined ultrasound findings of a yolk sac IUP with presenting symptoms to determine fetal outcome.

Limitations

Limitations to our study include the expected errors associated with retrospective chart review. Recommended strategies used to minimize such errors were as follows: creating a new standardized ETR facilitating future data extraction, recording data directly onto a computerized data spreadsheet, limiting the number of variables extracted for each patient, and recruiting a physician not involved with the study to review results and obtain a ? statistic [14]. Despite our efforts, however, there is the possibility that patients were inadvertently excluded because of misdiagnosis or

Age range (y) % Total (117) Miscarriage; 95% CI 20 wk; 95% CI
All	117	(100%)	35	(30.0%;	21.7%-38.3%)	82	(70.0%;	61.7%-78.3%)
b20	15	(13%)	6	(40.0%;	15.2%-64.8%)	9	(60.0%;	35.2%-84.79%)
20-24	44	(37.5%)	12	(27.2%;	14.1%-40.4%)	32	(72.7%;	59.6%-85.9%)
25-29	29	(24.8%)	6	(20.7%;	6.0%-35.4%)	23	(79.3%;	64.6%-94.1%)
30-34	17	(14.5%)	4	(23.5%;	3.4%-43.7%)	13	(76.5%;	56.3%-96.6%)
>=35	12	(10.2%)	7	(58.3%;	30.4%-86.2%)	5	(41.7%;	13.8%-69.6%)

No. of miscarriages	% Total (117)		Miscarriage; 95% CI		20	wk; 95% CI
All	117	(100%)	35	(30.0%; 21.7%-38.3%)	82	(70.0%; 61.7%-78.3%)
0	79	(67.5%)	24	(30.4%; 20.2%-40.5%)	55	(69.6%; 59.5%-79.8%)
1	26	(22.3%)	8	(30.8%; 13.0%-48.5%)	18	(69.2%; 51.5%-87.0%)
>=2	10	(8.5%)	3	(30%; 1.6%-58.4%)	7	(70%; 41.6%-98.4%)
Unknown	2	(1.7%)	0	(0%)	2	(100%)

miscoded information. We attempted to overcome this limitation by reviewing all ETRs of women with complaints of either “vaginal bleeding” or “abdominal pain” to determine if these patients were also pregnant.

Table 4 Fetal outcome determined by previous miscarriage

Although we visualized the yolk sac without a fetal pole, there is a possibility that a fetal pole existed that was too small to see on ED ultrasound. The importance of this limitation is minimized by the fact that if a fetal pole is too small to see on ED ultrasound, it is not clinically useful to the ED physician.

Our study population was derived from a single institution whose patients are composed of women either in the military or dependents of military members (spouses or children). All active duty military members and dependents have access to emergency care, as well as primary and specialty care, without financial concern. Patients often seek emergency care at very early stages of their disease processes and, therefore, may not represent the typical population present- ing to a civilian ED with complaints of a symptomatic first- trimester pregnancy. The transient nature of military families greatly contributed to our rate of patients lost to follow-up, which was 12%.

A larger sample size would have served to strengthen our statistically significant results, but also may have allowed us to make further conclusions with respect to age and previous miscarriage on rate of fetal loss.

That the primary investigator also performed data collection may have introduced bias; however, because there were only 2 variables for chief complaint and only 2 possible patient outcomes, this potential bias was most likely minimal. In addition, using a 20-week ultrasound as a surrogate for successful pregnancy outcome is some- what problematic, although the risk of pregnancy loss after 20-week EGA is only a small possibility. Similarly, Ob/Gyn literature frequently uses the 20-week ultrasound as an end point for viability and successful pregnancy outcome. Lastly, because the accuracy of the second- trimester ultrasound lends itself to a standard deviation of

Table 5 Fetal outcome determined by presence of vaginal bleeding on examination

+-7 days, there is a potential that our 20-week ultrasound end point, in fact, represented a fetus at 19 weeks.

Conclusion

These results will allow emergency physicians to counsel patients with a yolk sac IUP regarding fetal outcomes and the chances of spontaneous abortion based on their presenting symptoms. Our results do not impact manage- ment of the early/indeterminate pregnancy but serve to educate patients and physicians. There is a great deal of anxiety when women present to the ED with a symptomatic first-trimester pregnancy; thus, it is essential for physicians to have evidence-based data to assist in determining possible fetal outcomes.

Article summary

1. Why is this topic important? Fetal loss in symptomatic pregnancy is an important topic in emergency medicine because abdominal pain and bleeding in pregnancy are frequent ED chief complaints. Furthermore, the scope of ED ultrasound use is widening and can help diagnose IUP.
2. What does this study attempt to show? This study demonstrates that fetal loss rate can be predicted using symptoms and ultrasound data together.
3. What are the key findings? Thirty percent of women diagnosed with a yolk sac IUP on ultrasound in the ED experienced fetal loss. Vaginal bleeding (with or without pain) increased the rate of fetal loss compared with women with pain only.
4. How is patient care impacted? Women with pain or bleeding in pregnancy are often worried about pregnancy outcome. Having evidence-based data regarding out- comes will assist the emergency physician accurately

Blood (mL)	% Total (117)	Miscarriage; 95% CI		20	wk; 95% CI
All	117 (100%)	35 (30.0%;	21.7%-38.3%)	82	(70.0%;	61.7%-78.3%)
b5	99 (84.6%)	23 (23.2%;	14.9%-31.6%)	76	(76.8%;	68.5%-85.1%)
>=5	14 (12%)	12 (85.7%;	67.4%-100%)	2	(14.3%;	0.0-32.6%)
Unknown	4 (3.4%)	0 (0%)		4	(100%)

counsel women with symptomatic first-trimester preg- nancies regarding pregnancy outcome.

References

1. Zinaman MJ, Clegg ED, Brown CC, et al. Estimates of human fertility and pregnancy loss. Fertil Steril 1996;65:503-9.
2. Ellish NJ, Saboda K, O’Connor J, et al. A prospective study of early pregnancy loss. Hum Reprod 1996;11:406-12.
3. Tongsong T, Srisomboon J, Wanapirak C, et al. Pregnancy outcome of threatened abortion with demonstrable fetal Cardiac activity: a cohort study. J Obstet Gynaecol 1995;21:331-5.
4. Tannirandorn Y, Sangsawang S, Manotaya S, et al. Fetal loss in threatened abortion after embryonic/fetal heart activity. Int J Gynaecol Obstet 2003;81(3):263-6.
5. Jauniaux E, Johns J, Burton GJ. The role of Ultrasound imaging in diagnosing and investigating early pregnancy failure. Ultrasound Obstet Gynecol 2005;25:613-24.
6. Reece FA, Petrie RH, Sirmans MF, et al. Combined intrauterine and extrauterine gestations: a review. Am J Obstet Gynecol 1983;146:323-30.
7. Dogra V, Paspulati RM, Bhatt S. First trimester bleeding evaluation. Ultrasound Q 2005;21:69-85.
8. Dart RG. Role of pelvic ultrasonography in evaluation of symptomatic first-trimester pregnancy. Ann Emerg Med 1999;33(3):310-20.
9. Hurwitz SR. Yolk sac sign: sonographic appearance of the fetal yolk sac in missed abortion. J Ultrasound Med 1986;5(8):435-8.
10. Kurtz A, Needleman L, Pennell R, et al. Can detection of the yolk sac in the first trimester be used to predict the outcome of pregnancy? AJR Am J Roentgenol 1992;158(4):843-7.
11. McKenna KM, Feldstein VA, Goldstein RB, et al. The empty amnion: a sign of early pregnancy failure. J Ultrasound Med 1995;14(2):117-21.
12. Gracia CR, Sammel MD, Chittams J, et al. Risk factors for spontaneous abortion in early symptomatic first-trimester pregnancies. Obstet Gynecol 2005;106(5 Pt 1):993-9.
13. Weiss JL, Nyberg DA, Berkowitz RL, et al. Threatened abortion: a risk factor for poor pregnancy outcome, a population-based screening study. Am J Obstet Gynecol 2004;190:745-50.
14. Worster A, Haines T. Advanced statistics: understanding medical record review (MRR) studies. Acad Emerg Med 2004;11:187-92.