By Katherine Rivlin, MD, MSc
Associate Professor, Obstetrics and Gynecology, The Ohio State University Medical Center, Columbus
In a national sample of people seeking abortion, broadening screening questions beyond last menstrual period to self-assess gestational duration improves accuracy in determining eligibility for medication abortion using a 70-day threshold. In-person ultrasound may not be necessary.
Ralph LJ, Ehrenreich K, Barar R, et al. Accuracy of self-assessment of gestational duration among people seeking abortion. Am J Obstet Gynecol 2022;226:710.e1-710.e21.
A two-day regimen of mifepristone followed by misoprostol is approved by the U.S. Food and Drug Administration for medication abortion up to 70 days, or 10 weeks of gestation. Although clinicians commonly use ultrasound or physical examination to establish gestational duration, the COVID-19 pandemic paved the way for care models that forego in-person evaluation. However, clinicians may have concerns regarding the accuracy of self-assessed gestation, and some state restrictions may require in-person ultrasound or restrict the use of telemedicine abortion. Most prior studies of self-assessed gestational duration have evaluated the accuracy of patient-reported last menstrual period (LMP). Results have varied widely, with a 2014 systematic review reporting a range of 2.5% to 11.8% of participants erroneously self-assessing themselves as eligible for medication abortion using LMP.1 A significant number of people seeking abortion also may not know their LMP, and one in five people have irregular menses or intermenstrual spotting, all of which can delay presentation for care.2 This study examines a broader set of pregnancy-dating questions to identify patients eligible for medication abortion more accurately.
The study team recruited patients seeking abortion from seven clinics in Alabama, California, Florida, Illinois, North Dakota, Texas, and Washington, DC, to achieve geographic and policy diversity. Potential participants had to be age 15 years or older, speak and read English or Spanish, and not yet have had an ultrasound evaluation at their abortion appointment. The study included people seeking both medication abortion and procedural abortion. Participants completed survey questions before having their ultrasound evaluation. Gestational duration then was recorded by the ultrasound technician.
The research team developed an initial survey and then refined it based on input both from patients seeking abortion and a community advisory board that included both medical and nonmedical experts. Examples of survey questions included “How many weeks pregnant do you think you are today?” and “Select the day when you think you got pregnant.” Surveys also queried menstrual regularity, use of hormonal contraception, and any previous pregnancy-dating. The analysis examined the screening performance of each question alone and in combination to find a set of screening questions that included the fewest items to impose the lowest participant burden with the highest sensitivity and accuracy. Then, the team estimated the rate at which participants would falsely identify themselves as eligible for medication abortion using each screening test. Surveys also gathered participant demographic information.
Of 1,697 patients approached, 1,312 (77%) were interested in participating. The final analytic sample included 1,089 participants who were eligible, provided consent, and for whom ultrasound dating measurements were available. One-quarter (25%) of participants were more than 70 days’ gestation on ultrasound. Participant demographics and gestational duration reflected those of people seeking abortion nationally. The question “Are you more than 10 weeks pregnant?” with yes or no answer choices had an 84% (79% to 88%; 95% confidence interval [CI]) sensitivity in screening for ineligibility. In other words, the question identified 84% of participants who measured more than 70 days on ultrasound. The same question had a 4% rate at which people falsely identified themselves as eligible for medication abortion, or 4% of participants more than 70 days’ gestation on ultrasound identified themselves as eligible for medication abortion. Adding three additional statements that asked if the participant was 1) more than two months pregnant, 2) had missed more than two periods, and 3) had irregular periods before pregnancy correctly identified 93% (89% to 96%; 95% CI) of participants more than 70 days on ultrasound and reduced the number of participants who incorrectly screened as eligible to 1.7%.
This study is comprised of a nationally representative sample of people seeking abortion in diverse geographic and policy climates, making the findings generalizable. It adds to a growing body of evidence indicating that in-person ultrasound and assessments are unnecessarily burdensome to patients as they seek medication abortion. Policies restricting the provision of telemedicine abortion or requiring in-person ultrasound are similarly unnecessary. Although the study team did not find a perfect question or combination of questions to establish gestational duration, they found a set of questions that minimize the frequency of erroneous self-assessment. These questions could be adapted easily for use in an over-the-counter medication abortion regimen or as part of an online screening tool for telemedicine abortion.
Patient reporting of LMP long has been considered the medical standard for self-assessing pregnancy duration, but this study highlights the limitations of this question. Although LMP may have clinical significance, it may be hard to recall for patients if it has no personal significance. This study uses input from a diverse team, including researchers, a community advisory board, and, most importantly, patients themselves, to generate more patient-centered, easily recalled, and, therefore, more accurate methods of pregnancy self-assessment. The team successfully created a combination of questions that reduces the proportion of participants who incorrectly screened themselves as eligible for medication abortion at 70 days or fewer compared to previous studies using LMP alone. LMP proved hard for many participants to recall, particularly for young people and people with irregular menses, and was not included in the final set of questions. This study highlights that clinicians may have been asking the wrong question all along.
As much as 70% of participants who erroneously screened as fewer than 70 days still were fewer than 84 days, or 12 weeks, pregnant. Medication abortion remains highly effective at ending pregnancy at these gestational durations, and people reliably obtain follow-up care when needed.3 With the landscape of legal abortion poised to change dramatically, patients may be called upon to self-assess gestational duration as they face a loss of access to legal abortion and consider self-managed abortion. In this context, clinicians should consider their role in harm-reduction models of care or ensuring that patients seeking abortion outside the mainstream healthcare system have adequate counseling and care before and after their abortion procedure. A set of questions that improves the accuracy of self-assessed gestational duration with the least burden to patients can be integrated easily into harm-reduction care models.4
- Schonberg D, Wang LF, Bennett AH, et al. The accuracy of using last menstrual period to determine gestational age for first trimester medication abortion: A systematic review. Contraception 2014;90:480-487.
- Bracken H, Clark W, Lichtenberg ES, et al. Alternatives to routine ultrasound for eligibility assessment prior to early termination of pregnancy with mifepristone-misoprostol. BJOG 2011;118:17-23.
- Raymond EG, Tan Y-L, Comendant R, et al. Simplified medical abortion screening: A demonstration project. Contraception 2018;97:292-296.
- Bixby Center for Global Reproductive Health. Self-managed abortion: What healthcare workers need to know. University of California San Francisco. https://bixbycenter.ucsf.edu/sites/bixbycenter.ucsf.edu/files/Self-managed%20abortion-what%20healthcare%20workers%20need%20to%20know.pdf