Recruitment of pregnant women in randomized trials: What could hold us back?

For many years, members of the Dutch Consortium for Healthcare Evaluation and Research in Obstetrics and Gynecology (NVOG Consortium) have been working on the improvement of care for patients with threatened preterm birth. One of the research topics, the safety and effectiveness of labor-suppressing tocolytic medications, is covered in the APOSTEL trials. In this blog, we would like to outline the APOSTEL trials that have been performed so far, provide an update on the current project, the APOSTEL 8 trial, and provide different stakeholders’ views on recruitment of pregnant women in randomized trials from our recent paper in BMC Pregnancy and Childbirth.

The APOSTEL line of trials has covered APOSTEL I-VII so far. APOSTEL stands for ‘Assessment of Perinatal Outcome with [S….] Tocolysis in Early Labor’ – in this acronym the ‘S’ had different meanings in the different APOSTEL trials. These trials are summarized in Table 1. The most recent publication related to these trials describes stakeholders’ views on recruitment of pregnant women in randomized trials using data gathered during the APOSTEL VI trial.

Overview of APOSTEL trials

APOSTEL trial Research topic Conclusions
I Effectiveness of nifedipine in low-risk women with cervical length of 10-30 mm and negative fetal fibronectin test Nifedipine is not effective in prolonging pregnancy in low-risk women; surprisingly, women in the nifedipine group had significantly lower gestational age births than women in the placebo group.
II Effectiveness of maintenance tocolysis with nifedipine Maintenance tocolysis with nifedipine does not improve adverse neonatal outcome or prolongation of pregnancy.
III Effectiveness and safety of nifedipine and atosiban Nifedipine and atosiban are equally effective with respect to adverse neonatal outcome. A non-significant higher mortality rate was present in the nifedipine group.
IV Nifedipine in women with preterm prelabor rupture of membranes No differences in perinatal outcome and prolongation of pregnancy between nifedipine and placebo groups.
VI Pessary versus no treatment in women with successful treatment of threatened preterm birth (not delivered after 48 hours tocolysis and corticosteroids) Pessary treatment is not effective in reducing adverse perinatal outcome.
V/VII Trials never started Other research groups already investigated the research topic upon completion of design of study.

Table 1. The APOSTEL trials performed to date

For decades, it was thought that tocolytics improved perinatal outcome, since tocolytics could prolong pregnancy. However, in 2015, the World Health Organization (WHO) stated that no effectiveness of tocolytic therapy on perinatal outcome has been proven so far and that randomized placebo-controlled trials on this topic are urgently needed. Our project group completely agreed with this statement and believes that tocolytics should only be administered in routine care when it has been proven to be effective. Prolongation of pregnancy should not be the main goal of tocolytic therapy; the main goal should be improvement of perinatal outcome. Threatened preterm birth could be a sign of a problem within the uterine environment and therefore tocolytic therapy may be harmful by exposing the fetus longer to a suboptimal uterine environment. In several countries it is unusual to administer tocolytic therapy in threatened preterm birth in view of a possible harmful effect or lack of effectiveness.

Therefore we designed the APOSTEL 8 trial, which is currently open for recruitment, to solve the question whether administration of tocolytic drugs improves neonatal outcome in cases of threatened preterm birth. This trial investigates the effectiveness of 48 hours of tocolytics with atosiban versus placebo in women with threatened preterm labor between 30 and 34 weeks. The standard treatment of corticosteroids is administered to all participants. They will be randomized to either 48 hours of atosiban or 48 hours of placebo. The choice for atosiban was made as no significant difference in effectiveness was present between nifedipine and atosiban in the APOSTEL III trial and because of the favorable safety profile of atosiban. In addition, it is registered for the indication of tocolysis in many countries in Europe.

Current status of APOSTEL 8


Power analysis shows that we need to recruit 1514 participants in this study. All perinatal centers in the Netherlands will participate, together with many large secondary hospitals. Since several countries in Western Europe were interested in this trial, the APOSTEL 8 study will also be performed in the UK, Ireland and possibly Belgium.

To date, recruitment is slow due to a number of factors. Doctors are often too busy treating newly admitted patients to be able to take the time to invite them to participate in the trial and provide enough information. And even if there is sufficient time to explain the trial, the pregnant woman and her partner have to make a difficult and urgent decision whether to participate or not at a time which is very scary and insecure for them. To quickly introduce the trial to possible participants, we created a short animation video in which the dilemma on this topic and the necessity for this trial is explained. Furthermore, we approach doctors and midwives to raise awareness of their attitudes when it comes to recruitment of pregnant women for a randomized trial, as these may be hindering recruitment.

Stakeholders’ views on recruitment of pregnant women

For decades, stakeholders involved in the inclusion of pregnant women in clinical research, such as bioethicists, researchers, and clinicians, have argued that research participation of pregnant women is essential in order to increase the evidence base for treatment of pregnant women and fetuses. In this qualitative study, we interviewed various stakeholders, including pregnant women themselves, in order to understand their opinions on including pregnant women in clinical trials.

For this study, we made use of the APOSTEL trials in the Netherlands, specifically the APOSTEL VI study (see Table 1).

Our analysis revealed four themes that characterized stakeholders’ views on inclusion of pregnant women in the APOSTEL VI study:

First, pregnant women participated in the APOSTEL VI study primarily for potential individual benefit and secondarily for altruistic motives, as this quotation illustrates: “I like to participate when it is positive for me, when participation makes me feel like I do something good, but that it is also positive for myself and that nothing can go wrong.” (pregnant woman, participating in APOSTEL VI) This finding is in contrast with hypothetical studies, where altruistic motives are usually mentioned as the primary reason for participation.

While pregnant women may be willing to participate …, an underlying protective sentiment, resulting in gate-keeping and directive counselling, sometimes hampered recruitment.

Second, a gate-keeping tendency hampered recruitment of pregnant women who might be eligible and willing.  Alarmingly, questions about pregnant women’s decision-making capacities sometimes surfaced, as demonstrated in this quotation: “I think they [pregnant women] are behaviorally more vulnerable. I think they have some sort of black, blind spot: everything for the child. […] They are not sufficiently competent.” (gynecologist-in-training)

Third, healthcare professionals may have used the counselling conversation to steer pregnant women towards a particular decision, which was called counseling positive or counseling negative: “Sometimes you know that it is not the right candidate. That it will be a mess. And then you counsel slightly more negatively.” (gynecologist-in-training)

Fourth, all stakeholders were hesitant about inclusion of pregnant women in clinical research in general due to a protective sentiment, although this related mostly to trials that are not pregnancy-specific and in which there are risks: You should not expose pregnant women to medications of which the effects on the baby are unknown, if you have an alternative. It’s different if it is pregnancy-specific. In that case you don’t have an alternative, and then I have fewer objections.” (Research Ethics Committee member and clinician)

What these findings show us is that while pregnant women may be willing to participate for potential individual benefit and altruistic motives, an underlying protective sentiment, resulting in gate-keeping and directive counselling, sometimes hampered recruitment. These results apply to the APOSTEL VI study in particular, but it may be that these views have affected the recruitment in past trials.

In order to improve recruitment, it is therefore important to invest in educating and training of those involved in recruiting participants. This could be done by asking healthcare professionals to assess the logistics and potential risks and benefits of a study before discussing it with potential participants or by making stakeholders aware of their protective attitude and the resulting effects.

View the latest posts on the BMC Series blog homepage