Elsevier

The Lancet

Volume 385, Issue 9969, 21–27 February 2015, Pages 698-704
The Lancet

Articles
Clinical follow-up compared with self-assessment of outcome after medical abortion: a multicentre, non-inferiority, randomised, controlled trial

https://doi.org/10.1016/S0140-6736(14)61054-0Get rights and content

Summary

Background

Medical abortion with mifepristone and prostaglandins is well established. We compared clinical assessment with self-assessment of abortion outcome.

Methods

This randomised, controlled, non-inferiority trial was done in four clinics in Austria, Finland, Norway, and Sweden, between Aug 16, 2011, and Jan 31, 2013. Women aged 18 years and older who had requested medical termination of a pregnancy up to 63 days of gestation were eligible. Computer-generated block randomisation (block size ten) assigned women in a 1:1 ratio to attend routine clinical follow-up or to self-assess outcome at home with a semiquantitative urine human chorionic gonadotropin (hCG) test 1–3 weeks after abortion. The primary outcome was the percentage of women with complete abortion not requiring further medical or surgical intervention within 3 months. Analysis was per protocol and by intention to treat. The non-inferiority margin was five percentage points. This trial is registered with ClinicalTrials.gov, number NCT01487213.

Findings

924 women were assigned routine follow-up (n=466) or self-assessment (n=458) and included in the intention-to-treat analysis. 901 were included in the per-protocol analysis (n=446 and n=455, respectively). Complete abortion was reported in 432 (95%) of 455 in the routine follow-up group and 419 (94%) of 446 women in the self-assessment group (crude difference −1·0, 95% CI −4·0 to 2·0). 20 (4%) women in the routine follow-up group and 17 (4%) in the self-assessment group required surgery. No women in the routine follow-up group versus three in the self-assessment group had undetected continuing pregnancies. Eight (1·8%) and one (0·2%) women, respectively, had infections (p=0·038).

Interpretation

Self-assessment was non-inferior to routine follow-up and could save resources.

Funding

Nordic Federation of Obstetrics and Gynaecology, European Society of Contraception, Helsinki University Central Hospital, Helse Finnmark, Swedish Research Council, and Stockholm County Council and Karolinska University Hospital.

Introduction

Medical abortion with combined mifepristone and misoprostol has been established as a highly effective, safe, and acceptable method for medical termination of pregnancy for the past 25 years.1 This method is simpler and requires fewer resources than surgical abortion and provides women with a choice of intervention.

Medical abortion consists of several steps, including misoprostol to take at home, which helps to demedicalise abortion and provides privacy. Women prefer to terminate pregnancy quickly with as few visits to a clinic as possible.2 Despite the proven efficacy and safety of medical abortion, a need is widely perceived for routine clinical follow-up to exclude continuing pregnancy.3 If complete abortion can be confirmed, however, international guidelines do not endorse clinical follow-up.1, 4 Routine follow-up might include ultrasonography, measurement of human chorionic gonadotropin (hCG) in serum or urine, and pelvic examination, which could require more than one clinic visit.5 By contrast, routine follow-up after surgical abortion is not a standard protocol despite a similar failure rate to medical abortion.4 The number of visits to the clinic might be an important factor in the acceptability of medical abortion. Some studies have indicated that medical abortion is associated with more negative experiences of care and lower acceptability than surgical abortion, which has been attributed partly to the need for more follow-up visits.6 Use of urine hCG tests with sensitivity of around 1000 IU/L to enable women to eliminate clinical visits has been suggested.5, 7 If such a test was proven to be safe and effective, many women would not need to attend follow-up clinic visits.7

Women with complications arising from an abortion generally seek health care, especially if they have received adequate counselling. Continuing viable pregnancy after a failed medical abortion, however, might remain unnoticed until after the legal limit for induced abortion has passed. For this reason, a reliable method for excluding continuing pregnancy is important. A review by Grossman and Grindlay8 showed varying sensitivity of self-assessment to identify continuing pregnancy. Assessment of outcomes with ultrasonography or testing for hCG in serum are the most accurate methods but require substantial resources. High-sensitivity urine hCG pregnancy tests, which can detect hCG concentrations lower than 25 IU/L, are of limited value for follow-up because results can remain positive for several weeks after termination of pregnancy.9, 10 Semiquantitative urine hCG tests with upper cutoff concentrations of 1000 IU/L or higher combined with incrementally lower cutoff concentrations down to as low as 5 IU/L have been developed and have the potential to identify complete abortion within a short time; falling hCG concentrations register as a negative result for pregnancy with these tests.11, 12

We did a non-inferiority study to assess whether a commercially available semiquantitative urine hCG test for self-assessment of abortion outcome would be as effective and manageable as outpatient follow-up after medical abortion.

Section snippets

Study design and patients

Between Aug 16, 2011, and Jan, 31, 2013, all women who requested a medical termination of pregnancy of up to 63 days' gestation (crown-to-rump length 22·4 mm or less13) in four clinics (three in teaching hospitals) in Austria (Vienna), Finland (Helsinki), Norway (Oslo), and Sweden (Stockholm) were informed of this study. Eligible women were aged 18 years or older and had a documented confirmation of evolutive intrauterine pregnancy (ie, visible intrauterine yolk sac or fetal heartbeat on

Results

Of 3942 women who requested termination of pregnancy, 1385 (35%) were eligible for inclusion in the study. 456 (33%) eligible women declined to participate and, therefore, 929 were randomised (figure 2). The intention-to-treat population comprised 924 women (n=466 in the routine follow-up group and n=458 in the self-assessment group). Protocol violations were reported in 11 women in the routine follow-up group (two crossovers, five who received misoprostol in the clinic, and four who miscarried

Discussion

Self-assessment of the outcome of medical abortion with a semiquantitative urine hCG test was not inferior to routine clinical follow-up and no differences were seen in complication rates. Our findings indicate that the outcome of medical abortion can be managed by alternative protocols. Satisfaction with follow-up method was significantly greater for self-assessment than for routine follow-up.

A large number of women who were seeking abortion did not want to be included in the study because

References (26)

  • Safe abortion: technical and policy guidance for health systems

  • H Kopp Kallner et al.

    Home self-administration of vaginal misoprostol for medical abortion at 50-63 days compared with gestation of below 50 days

    Hum Reprod

    (2010)
  • D Grossman et al.

    Routine follow-up visits after first-trimester induced abortion

    Obstet Gynecol

    (2004)
  • Cited by (0)

    View full text