Quantifying the risk of ectopic pregnancy with a transient diagnosis of pregnancy of unknown location

To quantify the risk of ectopic pregnancy among a transient diagnosis of pregnancy of unknown location (PUL).

is concerning because it may be associated with an undiagnosed EP, which occurs in 2%-3% of all pregnancies. [4][5][6] An EP has a high maternal mortality rate of 6%-9% due to rupture causing abdominal hemorrhage. 4,5 Banerjee et al. showed that EP was the definitive diagnosis in 14% of patients diagnosed with a PUL. 3 Therefore, it is important to determine whether a transient diagnosis of PUL predicts an increased odds of EP. No study to date has been designed to quantify the risk of EP in a diverse, predominantly Black/African American cohort of patients in an urban setting with a transient diagnosis of PUL. 4,5 Thus, the objective of the current study is to quantify the odds of EP among patients with a transient diagnosis of PUL in a large, urban setting. Secondary outcomes include reporting the type and frequency of definitive diagnoses, summarized case reports of those who had a ruptured EP, the odds of a PUL becoming a nonviable pregnancy, and associations between various factors and the definitive diagnoses.

| MATERIAL S AND ME THODS
This is a retrospective cohort study of highly diverse patients with Sciences System (UI Health). The sonographers followed the AIUM-ACR-ACOG-SMFM-SRU Practice Parameter for the Performance of Standard Diagnostic Obstetric Ultrasound Examinations. 7 The UI Health-approved guideline, Diagnosis and Management of Nonviable Pregnancy <10 Weeks Size in the Hemodynamically Stable Patient, was followed for managing the patients. Figure 1 illustrates the stepwise process to obtain a definitive diagnosis.
The symptomatic pregnant patient, estimated at less than 10 weeks of gestational age, based on last menstrual period, 8 was seen in the OBED. The patient was initially assessed for an acute abdomen. If ruled out, a serum quantitative βhCG was drawn and a "limited" transvaginal ultrasound was performed by the Obstetrics and Gynecology (OBGYN) resident with OBGYN or Reproductive Endocrinology and Infertility (REI) Attending supervision. If no IUP or EP was identified, then a transient diagnosis of PUL was made.
In the OBED, patients were managed based on βhCG level and symptoms. If the initial βhCG level was <1000 mIU/mL, and symptoms settled, the patient was discharged with strict ectopic precautions, and a return appointment was made in 2-4 days in the PUL clinic in the Center for Women's Health for a serial serum quantitative βhCG and a formal transvaginal ultrasound, followed by an outpatient visit with an OBGYN resident with REI Attending supervision. If the initial βhCG was between 1000 and 5000 mIU/mL or the patient remained symptomatic with a βhCG <1000 mIU/mL, the patient was admitted overnight for observation with a formal transvaginal ultrasound in the morning. If the transient diagnosis remained a PUL, the patient was seen in 2-4 days in the PUL clinic. If the βhCG was >5000 mIU/mL, the patient was admitted overnight for a formal transvaginal ultrasound in the morning, with a second opinion from the REI Attending if the pregnancy was not visualized.
In the PUL clinic, patients were managed based on their symptoms, physical examination, serial quantitative βhCGs and formal transvaginal F I G U R E 1 Definitive diagnoses for symptomatic patients presenting initially with a transient diagnosis of a pregnancy of unknown location (PUL). hCG, human chorionic gonadotropin; OBGYN, Obstetrics and Gynecology; PPUL, persisting pregnancy of unknown location; US, ultrasound. ultrasounds. A PUL Patient List was reviewed daily by an OBGYN resident with an REI or Gynecology Attending. If an IUP was diagnosed, based on a double decidual sac sign or a gestational sac with a yolk sac and/or embryo, the patient was transferred to an OBGYN for care. 9 Viability was determined using Doubilet criteria. 10 If a visualized EP was diagnosed, based a gestational sac with a yolk sac and/or embryo, the patient was counseled on options, including expectant, medical, and surgical management. Management with methotrexate was followed using a UI Health-approved protocol, which is consistent with the American College of Obstetrician and Gynecologists clinical Practice Bulletin. 6 Methotrexate dosing was followed uniformly as 50 mg/m 2 .
If serial βhCGs rose appropriately and formal transvaginal ultrasound remained negative for pregnancy, the transient diagnosis of PUL remained. Close monitoring with serial quantitative βhCGs with or without formal transvaginal ultrasounds continued until criteria were met for a definitive diagnosis, a transient diagnosis of a persisting PUL (PPUL), or, if serial βhCGs spontaneously decreased to <5 mIU/mL, the definitive diagnosis of resolved PUL was given.
Based on the criteria published for the ACTorNOT (Optimal

Treatment for Women With a Persisting Pregnancy of Unknown
Location) trial, if serial βhCGs plateaued (defined as a <30% increase over 2 days, a <50% increase over 3 days, a <75% increase over 4 days, a <100% increase over 5 days, a <130% increase over 6 days, a <166% increase over 7 days, or a <50% increase between the first and last βhCG measurements) and the formal transvaginal ultrasound remained negative for pregnancy, the transient diagnosis was changed to a PPUL. 11 Management options of PPUL included expectant, medical, or surgical management. Shared decision-making was encouraged. With expectant management of the PPUL, close monitoring with serial quantitative βhCGs with or without formal ultrasounds continued until criteria were met for a definitive diagnosis, or, if serial βhCGs spontaneously decreased to <5 mIU/mL, the definitive diagnosis was a resolved PPUL. With medical management of the PPUL, the two-dose methotrexate protocol was followed. Once the serial βhCGs decreased to <5 mIU/mL, the definitive diagnosis was a treated PPUL.
Surgical management of a PPUL consisted of dilation and curettage and/or a diagnostic laparoscopy. If intrauterine chorionic villi were identified by pathology, the definitive diagnosis was histological IUP. If chorionic villi were not identified, the definitive diagnosis was nonvisualized EP. Expectant management of the nonvisualized EP was followed if serial βhCGs spontaneously decreased to <5 mIU/ mL. If serial βhCGs increased, then the two-dose methotrexate protocol was followed. With a diagnostic laparoscopy, if extrauterine chorionic villi were confirmed by pathology, the definitive diagnosis was visualized EP. If chorionic villi were not identified, the patient continued to be closely monitored as a PPUL.
After University of Illinois at Chicago institutional review board approval was obtained, patients with a transient diagnosis of PUL were identified from the PUL/PPUL Patient List and OBGYN billing records. Electronic medical records were reviewed to determine study eligibility. Patient consent was not obtained as the study was determined to be exempt. Demographics, medical history, visit details, and management plans were collected.
Descriptive and analytical statistics were performed using IBM SPSS version 27 (IBM Corp., Armonk, NY, USA). We performed a two-way analysis of variance (ANOVA) with a P value <0.050 considered statistically significant. In addition, Bonferroni correction in post hoc analysis as well as multinominal regression controlling for measured covariates were performed. Nine patients in the study were included twice, and one patient was included three times, because of repeat PULs. Therefore, the cohort for analysis consisted of 244 patients with 255 PULs. Demographics are presented in Table 1. Age and body mass index were found to be normally distributed in the sample.

| DISCUSS ION
A total 13% of the 255 PULs were subsequently diagnosed as an EP.
The risk of EPs is higher than the 2%-3% risk of EP in the general reproductive population. 3 Therefore, the transient diagnosis of a PUL is an important predictor of EP. Close monitoring until the location of the pregnancy is determined is required to reduce the risk of an undiagnosed EP. In addition, 68% of the PUL diagnoses subsequently became nonviable pregnancies. This risk is greater than the early pregnancy loss rate of 31% in the general reproductive population. 12 Patients with a transient diagnosis of a PUL often have high anxiety associated with the uncertainty of their diagnosis. Therefore, it is of the utmost importance that counseling and management be datadriven, such as shown by this new finding. 13 It should also be noted that despite medical management with a two-dose methotrexate protocol, three (9%) of the EPs in this study ruptured. This illustrates the importance of close monitoring even when the patient is asymptomatic and serial βhCGs are decreasing appropriately. Although the data set was collected retrospectively, patients with an initial PUL were strictly followed according to the hospital-approved guideline. Further interventions to improve visit compliance need to be determined. In addition, further studies with a broader patient population need to be completed to generalize these findings.

| CON CLUS IONS
Overall, this study provides important new information for counseling and management of patients presenting initially with a PUL.

ACK N OWLED G M ENTS
We thank Sydney Agger for formatting assistance. We also thank the Department of Obstetrics and Gynecology at the University of Chicago at Illinois for their support.

FU N D I N G I N FO R M ATI O N
No funding was obtained for this project.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors report no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.