If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
We assessed whether a low-sensitivity pregnancy test is effective at identifying ongoing pregnancy after medication abortion at 64 to 70 days of gestation.
Study design
From October 2018 to March 2020, we performed a prospective observational study of participants in England and Wales undergoing medication abortion. Participants were scheduled to return to the clinic 14 ± 3 days after mifepristone administration to perform a low-sensitivity pregnancy test (human chorionic gonadotropin threshold of 1000 mIU/mL) and symptom checklist, and state whether they thought the abortion was complete. Clinicians also assessed the low-sensitivity pregnancy test and performed an ultrasound to determine abortion status. We calculated the sensitivity, specificity, negative and positive predictive value of the low-sensitivity pregnancy test (with and without a symptom checklist) for detecting ongoing pregnancy.
Results
We enrolled 757 participants. Thirty-one did not progress to abortion and 558 (76.9%) completed follow-up. Most (79.6%) attended per-protocol; 22 (3.9%) attended earlier than 11 days and 92 (16.5%) later than 17 days. Thirteen participants (2.3%) had an ongoing pregnancy. The low-sensitivity pregnancy test correctly identified all the ongoing pregnancies (sensitivity = 100%; specificity = 84.8%; negative predictive value = 100%; positive predictive value = 13.5%). The symptom checklist alone had a sensitivity of 76.9% and a negative predictive value of 99.4% for identifying ongoing pregnancies. Participants and clinicians agreed on the interpretation of the low-sensitivity pregnancy test 94.6% of the time.
Conclusions
Patient self-assessment of a low-sensitivity pregnancy test after medication abortion between 64- and 70 days’ gestation has high sensitivity and negative predictive value for identification of ongoing pregnancy.
Implications
Patients can be offered a low-sensitivity pregnancy test to assess for ongoing pregnancy after medication abortion up to 70 days of gestation thereby reducing the need for in-person visits. Services should be prepared to provide in-person assessments after positive or inconclusive results to ensure early identification of abortion complications.
Medication abortion with mifepristone and misoprostol is safe and highly effective through 70 days’ gestation with a success rate up to approximately 95% [
Safety, efficacy and acceptability of outpatient mifepristone-misoprostol medical abortion through 70 days since last menstrual period in public sector facilities in Mexico City.
]. Early and accurate identification of treatment failure is important to facilitate appropriate management. Therefore, most clinical guidelines recommend some form of follow-up after a medication abortion, generally 1 to 2 weeks after treatment [
]. Providers can offer follow-up in-person - typically with an ultrasound, by following trends in serum human chorionic gonadotropin (HCG), or via low-sensitivity, high-sensitivity, or multilevel (semi-quantitative) urine pregnancy tests with or without a symptom checklist [
Accuracy of a semi-quantitative urine pregnancy test compared to serum beta-hCG measurement: a possible screening tool for ongoing pregnancy after medication abortion.
A systematic review of follow-up after medication abortion using a low-sensitivity pregnancy test up to 63 days of gestation found that the test had a sensitivity of 67% to 100% for detecting ongoing pregnancies [
]. There are theoretical concerns about extrapolating the effectiveness of this model beyond 63 days because of the natural decline in hCG that typically occurs in the latter part of the first trimester [
]. Therefore, in practice, clinicians often require that patients return for in-person follow-up with ultrasound after medication abortion past 63 days of gestation. While this ensures a reliable method of identifying an ongoing pregnancy, it adds an extra burden for patients who may need to take time off from work and/or secure additional childcare to attend visits and therefore often do not attend.
Our study evaluated the ability of a low-sensitivity pregnancy test to identify ongoing pregnancies following medication abortion at 64 to 70 days of gestation. In addition, we sought to determine the sensitivity and specificity of the test, with and without a symptom checklist, to identify an ongoing pregnancy.
2. Material and methods
We designed a prospective observational study of patients presenting for medication abortion at 64 through 70 days’ gestation at the British Pregnancy Advisory Service (BPAS), a national non-profit provider of abortion and related services with 58 clinics in England and 2 in Wales providing medication abortion. At the time of the study, BPAS offered self-assessment of abortion completion with a low-sensitivity pregnancy test (Quadratech check4-hCG, 1000 mIU/mL) and a symptom checklist after medication abortion up to 63 days’ gestation but required patients from 64 to 70 days’ gestation to return to the clinic for an in-person assessment. Our study aimed to extend these self-assessments up to 70 days of gestation. The BPAS Research & Ethics Committee and the National Research and Ethics Service approved this study.
We initially recruited participants at 14 BPAS clinics in England with the highest numbers of medication abortions in the gestational age range of interest. In August 2019, a further 6 clinics were added (5 in England and 1 in Wales) to boost recruitment. Participants who were at least 16 years old, planning a medication abortion between 64 and 70 days of gestation (by ultrasound), able to consent in English, and willing to comply with study procedures were eligible. Clinicians approached potential participants either at the initial abortion consultation or the in-person follow-up scheduled 14 (±3) days after taking mifepristone. After obtaining study consent, we collected background and demographic data.
We performed the study interventions during the follow-up visit. Clinicians asked participants to complete a symptom checklist based on that used in a study by Cameron et al. [
] for remote follow-up commonly used in British abortion clinics. The checklist asked participants to indicate whether they had no bleeding within 24 hours of taking the misoprostol, bleeding for less than 4 days, or still felt pregnant (such as sore breasts, nausea, etc.). Participants then completed and interpreted a low-sensitivity pregnancy test and indicated on the study form whether they thought their abortion was complete based on their interpretation of the test result and checklist. The clinician, who was not blinded to participant interpretation of the low-sensitivity pregnancy test, also assessed the result of the test but performed a routine post-abortion follow-up visit including transvaginal ultrasound to determine abortion status. If clinicians identified an ongoing pregnancy, they managed it per the standard of care. We did not collect data on those outcomes.
If participants recruited at consultation did not return for scheduled follow-up, clinic staff made one attempt to contact them and reschedule. If a participant was unwilling to return or was unreachable, we considered them lost to follow-up. If a participant presented before the scheduled follow-up due to a concern, clinicians recorded the reason for the visit and used their clinical judgment to determine if an ultrasound was needed. If the clinician felt an ultrasound was needed, they carried out study interventions at that time and discharged the participant from the study. If an ultrasound was not indicated, clinicians asked the participant to return for their scheduled follow-up at which time they performed the study procedures.
We calculated that 35 ongoing pregnancies were needed to determine with 95% confidence that the low-sensitivity pregnancy test would identify 90% of ongoing pregnancies, with a margin of error of 10%. Based on a review of medication abortion studies at this gestational age range, we anticipated that approximately 3% of participants would have an ongoing pregnancy [
]. We, therefore, needed to recruit 1155 participants to capture 35 ongoing pregnancies. Within our population at BPAS, there was approximately a 30% loss to follow-up rate after medication abortion. Based on this, we planned to recruit 1500 participants. We performed data analysis with Stata/SE 12.1 (College Station, TX). We calculated sensitivity/specificity and positive/negative predictive values to assess the accuracy of the pregnancy test in identifying ongoing pregnancies.
3. Results
From November 2018 to March 2020, we assessed 1047 patients for eligibility (Fig. 1). We recruited 757 of which 726 progressed to abortion and 558 (76.9%) completed a low-sensitivity pregnancy test and had an ultrasound scan (Table 1). One participant did not complete the checklist but was retained in the analysis as the primary intervention of interest, the low-sensitivity pregnancy test, had been performed. In March of 2020 when lockdown measures to manage COVID were instituted in the UK, BPAS transformed its practice to a primarily no-test medication abortion model. At this time, we began using a low-sensitivity pregnancy test and symptom checklist as follow-up after all medication abortions up to 70 days of gestation. We stopped recruitment for this study because we no longer performed routine postabortion ultrasounds.
Fig. 1Flowchart of participants using a low-sensitivity urine pregnancy test for identifying ongoing pregnancy after medication abortion at 64 to 70 days of gestation.
Table 1Characteristics of participants having follow-up with a low-sensitivity pregnancy test and symptom checklist after medication abortion at 64 to 70 days of gestation at British Pregnancy Advisory Service from October 2018 to March 2020
Self-assessment of medical abortion using a low-sensitivity pregnancy test, checklist and text messages in the South African public sector: A randomized controlled trial.
Contraception and not Randomized Controlled Trail.2015; 92: 373
The median time between taking mifepristone and performing the pregnancy test was 15 days (range 1−47 days; 25th percentile 14 days and 75th percentile 16 days). Most participants completed the pregnancy test at a scheduled follow-up but 4.1% (23/558) completed it at an unscheduled visit. In some cases, follow-up occurred later than scheduled usually because the participant could not attend the planned appointment. Overall, 114 participants performed the pregnancy test outside of the 14 (±3) day window specified in the protocol: 22 (3.9%) earlier than 11 days and 92 (16.5%) later than 17 days; only 31 (5.6%) took it later than 21 days.
In most cases (528/558, 94.6%), patient and clinician interpretations of the low-sensitivity pregnancy test correlated (Table 2). In 17.2% (96/558) of cases, the patient concluded that the test was positive, or they were unsure of the result. Using the self-assessment model applied under 64 days’ gestation, an ultrasound would have been warranted to assess for ongoing pregnancy in these cases.
Table 2Comparison of patient and clinician interpretation of low-sensitivity pregnancy test after medication abortion at 64 to 70 days of gestation
Of the 557 who completed the checklist, 94.8% (n = 528) reported no ongoing pregnancy symptoms, while the remainder reported some or all the following: no bleeding within 24 hours of taking misoprostol (2/557, 0.4%), bleeding for less than 4 days (16/557, 2.9%), or still feeling pregnant (14/557, 2.5%). Adding the checklist findings to the pregnancy test results would have meant that 10 additional patients would have required an ultrasound according to the self-assessment model.
Ongoing pregnancies were diagnosed by ultrasound in 2.3% of participants (13/558) Table 3. compares the patients’ assessments of the low sensitivity pregnancy test and checklist, as well as their overall interpretation of whether they were still pregnant, not pregnant, or unsure of their pregnancy status based on the results of those interventions, to the ultrasound findings. In Table 4, we present the sensitivity, specificity, positive predictive value, negative predictive value and 95% confidence intervals calculated from those figures. The sensitivity of the low-sensitivity pregnancy test in detecting ongoing pregnancy was 100% (75.3%−100%) with a specificity of 84.8% (81.5%−87.7%), and positive and negative predictive values of 13.5% (7.4%−22.0%) and 100% (99.2%−100%), respectively. The sensitivity of the symptom checklist alone was 76.9% (46.2%−95.0%) with a specificity of 96.5% (94.6%−97.9%), and positive and negative predictive values of 34.5% (17.9%−54.3%) and 99.4% (98.3%−99.9%) respectively. One participant who reported no pregnancy symptoms and was unsure of the pregnancy test result concluded from the combination of these interventions that they were no longer pregnant, which would have been a false negative.
Table 3Patient assessments of a low-sensitivity pregnancy test and symptom checklist as compared to ultrasound to identify ongoing pregnancy after medication abortion at 64 to 70 days of gestation
Table 4Sensitivity, specificity, positive and negative predictive values of patient assessment of a low-sensitivity pregnancy test and symptom checklist as compared to ultrasound to identify ongoing pregnancy after medication abortion at 64 to 70 days of gestation
We also analyzed the sensitivity data per protocol wherein we included only participants who had a low-sensitivity pregnancy test within 14 ± 3 days of taking mifepristone and found similar results (Tables 5 and 6). There were no false-negative results in this subset. False positives were related to the timing of the presentation. Of those who returned at less than 11 days, 55% of tests were falsely positive (11/20), compared to 14.7% who returned at 11 to 17 days (64/436), and 9.0% (8/89) at greater than 17 days. This difference reached statistical significance when comparing early to per-protocol presentations (p < 0.001) but not when comparing per protocol to late presentations.
Table 5Patient assessments of a low-sensitivity pregnancy test and symptom checklist as compared to ultrasound to identify ongoing pregnancy 14 +/- 3 days after mifepristone for medication abortion at 64-70 days of gestation.
Comparator
Ultrasound finding
Ongoing pregnancy (n)
All other outcomes (n)
Low sensitivity pregnancy test
Positive or unsure
8
64
Negative
0
372
Symptom checklist only
One or more symptoms
6
11
No symptoms
2
424
Symptom checklist and low sensitivity pregnancy test
One or more symptoms OR positive test
8
71
No symptoms PLUS negative test
0
364
Interpretation of symptom checklist and low sensitivity pregnancy test
Table 6Sensitivity, specificity, and positive and negative predictive values of patient assessment of a low-sensitivity pregnancy test and symptom checklist as compared to ultrasound to identify ongoing pregnancy 14 +/- 3 days after mifepristone for medication abortion at 64-70 days of gestation.
Our study demonstrated that self-assessment with a low-sensitivity pregnancy test was highly effective at detecting ongoing pregnancy after medication abortion from 64 to 70 days of gestation. In a 2018 systematic review, Raymond et al. analyzed data from 10 studies to assess the accuracy of low sensitivity pregnancy tests for detecting ongoing pregnancy after medication abortion up to 63 days’ gestation [
]. In the three prospective studies that compared low-sensitivity pregnancy tests to ultrasound included, a positive test had a sensitivity of 67% to 100% and a negative predictive value of 100% [
Clark W, Bracken H, Tanenhaus J, Schweikert S, Lichtenberg ES, Winikoff B. Alternatives to a routine follow-up visit for early medical abortion. Obs Gynecol 22010;115:264–72.
Gemzell-Danielsson K. Is self-assessment of medical abortion using a low-sensitivity pregnancy test combined with a checklist and phone text messages feasible in South African primary healthcare settings? A randomized trial.
]. In our study, a low-sensitivity pregnancy test had a sensitivity and negative predictive value of 100% for detecting ongoing pregnancy after medication abortion up to 70 days’ gestation. While the overall number of ongoing pregnancies in our study was small, our results suggest that the use of a symptom checklist might not be needed when using low-sensitivity pregnancy test to assess abortion outcomes, as combined use did not appear to improve diagnostic accuracy. On its own, the symptom checklist had a low sensitivity (76.9%) and is subject to recall bias. It may also introduce uncertainty if the patient has no symptoms and a pregnancy test result that is not conclusive, potentially leading to a false negative result as we saw in one participant.
We calculated a low predictive value for a positive pregnancy test result. False-positive (or inconclusive) test results are a drawback of using a low-sensitivity pregnancy test to assess the outcome of medication abortion and may cause anxiety for patients. Our per-protocol analysis found that false-positive results were more likely with presentation earlier than 2 weeks. Patient counseling should emphasize the importance of carrying out the test at the recommended interval as well as contacting the abortion service as soon as possible to discuss such test results and avoid a delayed diagnosis of a possible ongoing pregnancy. At BPAS, options for management of a positive test result include progressing directly to an in-person visit with ultrasound or, if the patient's pain, bleeding, and symptom history suggests that the pregnancy has passed, informing the client that the test may not signal an ongoing pregnancy and sending another test to perform in a week, returning for a clinic visit if the second test remains positive. Although the negative predictive value of a low-sensitivity pregnancy test is high, we also routinely counsel patients on signs of ongoing pregnancy such as still feeling pregnant, less than expected vaginal bleeding or delayed resumption of menses, which could help to identify ongoing pregnancies missed by the low-sensitivity pregnancy test.
A strength of our study was the inclusion of a population from a range of geographical locations across England and Wales. Of the 726 participants that proceeded to abortion, nearly 80% completed study interventions. We have no additional information on the participants who were lost to follow-up to comment on potential differences between these groups. Given that we did not achieve our proposed sample size of 35 ongoing pregnancies, our results are less robust than we had hoped. As discussed, the study was stopped due to the COVID pandemic and a need to reduce in-person care. However, we found it challenging to identify potential participants from the specific gestation of 64 to 70 days of gestation before this point in time. In efforts to improve recruitment, we added additional sites to our protocol, but enrollment was sluggish.
While our study did not assess participant acceptability of low-sensitivity pregnancy test, other researchers have evaluated this outcome. In a retrospective review of nearly 1800 patients having medication abortion in Scotland, 96% preferred self-assessment via the same protocol as our study vs telephone follow-up with a clinician [
Self-assessment of medical abortion using a low-sensitivity pregnancy test, checklist and text messages in the South African public sector: A randomized controlled trial.
Contraception and not Randomized Controlled Trail.2015; 92: 373
Our study demonstrated that a low-sensitivity pregnancy test has high sensitivity and negative predictive value for identifying ongoing pregnancy after medication abortion between 64 and 70 days of gestation. With the addition of these data into the current field of evidence, self-assessment using a low-sensitivity pregnancy test with or without symptom checklist is a reasonable method of follow-up for patients undergoing abortion up to 70 days’ gestation. The high patient acceptability reported in other studies further supports the use of this follow-up method [
Self-assessment of medical abortion using a low-sensitivity pregnancy test, checklist and text messages in the South African public sector: A randomized controlled trial.
Contraception and not Randomized Controlled Trail.2015; 92: 373
]. By offering a reliable method to assess abortion outcomes at home, most patients can avoid the burden of an in-person evaluation. In most cases, clinicians and patients came to the same conclusion regarding the low-sensitivity pregnancy test result suggesting that there should be no reason for people to be concerned that this test cannot be carried out via self-assessment. Abortion services should consider offering remote follow-up with low-sensitivity pregnancy tests after all medication abortions up to 70 days’ gestation.
Acknowledgments
We are grateful for staff assistance at participating BPAS clinics. We thank Jeanette Taylor and Amelia McInnes-Dean for their help with research administration.
References
Winikoff B
Dzuba IG
Chong E
Goldberg AB
Lichtenberg ES
Ball C
et al.
Extending outpatient medical abortion services through 70 days of gestational age.
Safety, efficacy and acceptability of outpatient mifepristone-misoprostol medical abortion through 70 days since last menstrual period in public sector facilities in Mexico City.
Accuracy of a semi-quantitative urine pregnancy test compared to serum beta-hCG measurement: a possible screening tool for ongoing pregnancy after medication abortion.
Clark W, Bracken H, Tanenhaus J, Schweikert S, Lichtenberg ES, Winikoff B. Alternatives to a routine follow-up visit for early medical abortion. Obs Gynecol 22010;115:264–72.
Gemzell-Danielsson K. Is self-assessment of medical abortion using a low-sensitivity pregnancy test combined with a checklist and phone text messages feasible in South African primary healthcare settings? A randomized trial.
Self-assessment of medical abortion using a low-sensitivity pregnancy test, checklist and text messages in the South African public sector: A randomized controlled trial.
Contraception and not Randomized Controlled Trail.2015; 92: 373
☆Declaration of Competing Interest: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Funding: This study was funded internally by BPAS.
00052-X&id=gr1.jpg){kind=link}