Discussion
In this single-centre randomised trial of partially outpatient IOL with Foley catheter versus inpatient IOL with vaginal PGE2, we found that outpatient Foley catheter was feasible, of comparable overall clinical effectiveness, and acceptable to women compared to inpatient IOL with vaginal PGE2 gel.
Both important advantages and disadvantages of OPC vs. IP were noted. Consistent with its mechanical mode of action few in the OPC group laboured prior to their booked amniotomy, with consequently a greater need for oxytocin IOL and a longer Birthing Unit stay. Despite this, and consistent with previous studies of mechanical methods,[3] Caesarean Section and induction to delivery intervals were similar between groups. As in prior Outpatient vs. Inpatient trials,[7] OPC women spent significantly less time in hospital prior to the birth of their baby, but the overall decrease in length of stay failed to reach statistical significance.
Vaginal birth was achieved in only two-thirds of each group, did not differ between groups, and is consistent with recent local trial and population data on IOL outcome with unfavourable cervix (5, 12). The increased (borderline significance) instrumental delivery rate in OPC has not been noted in inpatient mechanical vs. chemical IOL trials,[3] but was noted in a prior outpatient vs. inpatient PGE2 trial.[11] Although potentially related to increased oxytocin use and subsequent delivery for non-reassuring trace, the rate of oxytocin use remained constant during the trial while instrumental delivery in OPC decreased as the trial progressed (Figure 3). Therefore, initial increased instrumental delivery in OPC may have reflected the uncertainty of clinicians using OPC and a lower threshold for intervention, especially as the neonatal condition of these infants was similar to the overall cohort.
Figure 3.
Comparison of Instrumental Delivery (ID) numbers.
Regarding safety, both the issue of outpatient vs. inpatient IOL, and safety of the specific IOL method used, needs to be considered. Accordingly, catheter was chosen as our outpatient arm due to meta-analysis reporting lower rates of hyperstimulation than with Prostin,[3] women in the OPC arm were not discharged unless post-insertion CTG was reassuring, and OPC women were advised to represent promptly if contracting regularly, membranes ruptured or they had any concerns. As expected hyperstimulation was not seen in the OPC group, however four OPC women (8%) remained inpatients due to CTG concerns post-insertion, compared to 13% of the outpatient arm in a recent report of outpatient vs. inpatient PGE2.[12] One woman had a pathological trace at Birthing Unit admission after erroneously being told to remain at home when she called to report regular contractions. The baby delivered in good condition by Caesarean Section but with a cord pH of 7.03, underscoring the importance of advising women who commence labour after intervention to present for re-assessment. As only a small minority of women in our trial established in labour with OPC alone, this is likely to be an infrequent concern for OPC use.
The other major concern raised regarding safety of catheter IOL regards infectious morbidity, with a previous meta-analysis suggesting increased infectious morbidity with mechanical IOL.[13] These concerns were not borne out by the updated Cochrane review of mechanical methods of induction,[3] although the number of studies reporting these outcomes remained small. Our trial clearly lacks the statistical power to address such rare but important safety outcomes, however it is hoped our data can be incorporated into future meta-analyses addressing these issues.
Maternal satisfaction with obstetric intervention, including induction of labour, is an under-studied area.[14] Only Pennell et al.[5] have reported on maternal satisfaction and pain scores with catheter vs. PGE2. Like their group, we found increased pain at insertion but decreased pain thereafter, and we have additionally shown superior rest and sleep scores for the catheter.
Acceptability of OPC was high, as reflected in the low rates of expressed anxiety about safety of the catheter, and strong agreement amongst the OPC group that they were able to rest/relax at home. Limitations of the satisfaction survey include its postpartum administration, meaning birth outcomes may have influenced patient responses, and the wish of many trial participants (in a hospital where outpatient IOL is not routinely available) to enter the OPC arm, possibly giving a bias towards favourable responses from OPC women.
The strengths of the study are its prospective, randomised design in a field (outpatient vs. inpatient induction) with few published trials. Like preceding outpatient vs. inpatient trials, the major limitations of the study are its small size and single-centre nature. Failure to reach initial planned recruitment targets is another major limitation, with only the lesser sample size for the second of our pre-specified primary outcomes reached. This was due to only two-thirds of the expected number of women presenting for cervical preparation during the study period at the study hospital, and a higher than expected percentage of screened but not recruited women. The study nevertheless represents the first randomised trial comparing outpatient Foley catheter with an inpatient prostaglandin preparation.
Another study limitation is difficulty in assessing the relative contributions of outpatient vs. inpatient IOL and mechanical vs. chemical IOL when the outpatient arm was mechanical and inpatient chemical. As PGE2 is the default local cervical preparation method, but potentially unsuitable for outpatient IOL given greater hyperstimulation rates than mechanical methods,[3] the primary motivation for our study was exploring the role of the Foley catheter as a potentially feasible alternative to our current (chemical) inpatient standard. Although this inevitably reduces the research purity of the trial, as does the inclusion of (a) both nulliparous and parous women (b) a percentage of women with either maternal or fetal risk factors, we believe this pragmatic approach provided the best chance to assess the likely real-world role of OPC and is valid. Likewise, blinding of clinicians or patients to allocation was not practical, but apart from interpretation of traces and satisfaction survey responses, pre-specified outcomes were unlikely to be influenced by knowledge of allocation.
Finally, our unusual pre-specified primary outcomes are open to criticism. With hindsight, feasibility would have been better assessed using total inpatient hours rather than pre-delivery hours, and concerns regarding Birthing Services load, including overnight workload and its safety,[15,16] assessed using a composite of hours in Birthing Services and the proportion of overnight transfers and births in Birthing Services. The pre-specified secondary outcomes, although underpowered, do include the standard IOL outcomes.
Although our study provides evidence of feasibility and patient acceptability of Foley catheter for partial outpatient IOL, there remains a need for larger studies, ideally at multiple sites, to demonstrate whether OPC yields equivalent birth outcomes (without compromising maternal and fetal safety) to inpatient methods, and whether the non-significant decreases in total inpatient hours noted in our trial and the Cochrane review[7] will be borne out. Parous women, who are under-represented in both outpatient vs. inpatient and mechanical vs. chemical trials, are a particular priority for study, as is economic analysis of outpatient vs. inpatient IOL.
BMC Pregnancy Childbirth. 2013;13(25) © 2013 BioMed Central, Ltd.
