Is it time for outpatient cervical ripening with prostaglandins?


Inductions of labour have become more common over the same time period during which the necessity of the efficient provision of care and the importance of patient-centered care have become increasingly apparent. Accordingly, investigators have begun to consider which approaches to induction would be “best” – that is to say, which approaches yield the best health outcomes, are most satisfactory to women, and are most cost-effective.

One potential area of improvement is the approach to cervical ripening, which is utilised when a woman undergoing labour induction has an unripe cervix. Different types of agents (e.g. mechanical versus pharmacologic) have been studied, as have different regimens of those agents. Regardless of the specific agent or regimen used, however, most women who undergo cervical ripening do so in an inpatient, and often highly monitored, setting, such as a labour and delivery unit. Given that cervical ripening can take many hours, the possibility that ripening could be performed in less acute-care settings, including at home, has great appeal, provided it is effective and safe.

Toward that end, Wilkinson and colleagues have performed a randomised trial that was designed to compare an outpatient prostaglandin E2 ripening regimen with an inpatient ripening prostaglandin E2 regimen. The sample size of their study was determined in order to be able, with 80% power, to show a 10% reduction in oxytocin use from 50% to 40% in the intervention group. They achieved their targeted sample size and completed the trial with 827 women. Women who were randomised to outpatient ripening were not significantly different than their counterparts who were ripened as inpatients with regard to frequency of oxytocin use, caesarean delivery, or maternal or perinatal complications.

As has been noted, the question the authors have tried to answer is an important one that has implications for the cost-effective and patient-centred care of a frequently used intervention. And, randomising over 800 participants to a novel intervention is not an easy task. Nevertheless, the similarity of the outcomes in this trial should not be interpreted to provide a clear answer as to whether outpatient ripening with prostaglandin agents is an acceptable approach.

Although there was no difference in the primary outcome (oxytocin use), the power to find this difference was ultimately much lower than the 80% utilised in the original sample size calculation given the large number of women (nearly half) who were ultimately ineligible (due to spontaneous labour or cervical change) to receive their assigned treatment. Also, it is not entirely clear that all cross over occurred in a random fashion. For example, many women who still were eligible to receive their assigned treatment of outpatient ripening did not; while some cases were secondary to abnormalities on monitoring, others were because the participants “wished to remain as inpatients.” Also, significantly more women randomised to the outpatient arm were diagnosed on initial monitoring with non-reassuring EFM and hyperstimulation. It seems likely that this difference is not founded on biology (as there is no clear reason that women who are randomised to outpatient management should have less tolerant fetuses or more sensitive uteri) but on provider reluctance to proceed with outpatient treatment, which then manifested through a greater tendency to judge the monitoring as incompatible with outpatient treatment. Crossover after randomisation due to provider and/or participant choice introduces sources of potential bias that the randomisation sought to remove.

Lastly, many of the other outcomes occurred at a lower frequency than the primary outcome. Correspondingly, the power to detect a difference in these outcomes between the two study groups is even lower than that which was present for the primary outcome. Thus, whether outpatient management with prostaglandin agents is “safe” – and not, for example, associated with a higher risk of perinatal complications – remains uncertain.

Even if this study cannot provide us definitive answers, it provides important data that can help to inform future studies of outpatient cervical ripening. The authors should be congratulated for attempting such a study and helping in the ongoing effort to define best practices in labour induction.

Disclosure of interests

I have no conflicts of interest to disclose.