Vaginal breech delivery: results of a prospective registration study

In 2000, the results of a randomized multicenter trial, the Term Breech Trial (TBT) were published in the Lancet [1]. 2083 women were included in the study, and of the 1,042 women assigned to planned vaginal delivery, 591 (56.7%) delivered by this method. The trial reported significantly lower perinatal mortality, neonatal mortality, or serious neonatal morbidity in the planned cesarean section arm (1.6%) versus in the planned vaginal delivery arm (5.0%).

The TBT has had a major impact on clinical practice, and most countries now recommend planned cesarean section in breech births [2]. Later, several methodological aspects of the study have been questioned [3–5]. Despite this, planned cesarean section is considered as the safest and therefore the most widely used delivery method according to several European publications [6–10]. As a response to the TBT, a Norwegian expert group in 2003 reviewed published international literature on breech births and obtained information on perinatal morbidity and mortality in term breech infants in Norway [11]. Their results showed a lower perinatal morbidity among infants born vaginally in breech presentation compared to both study groups of the TBT. The mortality rate was 0.31% when corrected for lethal malformations and 0.09% after the additional correction for death before admission to the maternity clinic [12]. This could be explained by the close fetal monitoring, national procedures, and sufficient skills of obstetric staff, combined with high quality neonatal service, in contrast to many of the participating clinics in the TBT. As a result, Norwegian delivery units have continued a practice with vaginal delivery provided that strict national guidelines are followed. As one of few countries in the western world which has continued to practice planned vaginal delivery in selected women, quality control of own practice is fundamental, and sharing current experience with vaginal breech deliveries has strongly been encouraged.

The purpose of this study is to evaluate prospectively registered neonatal and maternal outcomes in term singleton breech deliveries in a Norwegian hospital ten years after TBT. We aim to compare outcomes in term breech infants after planned vaginal delivery versus after elective cesarean section.

In our prospective single hospital based study, planned vaginal delivery of term breech infants was associated with low perinatal morbidity. There were no neonatal deaths among the breech babies in this period. Statistically significantly more infants in the planned vaginal delivery group had an Apgar score < 7 and were transferred to the NICU. However, only one infant was admitted for more than four days. Cord blood was only sampled in 19 of 279 planned cesarean section cases and 130 of 289 planned vaginal cases, and can therefore not be analyzed. According to follow-up data, none of the infants (including the infant with serious morbidity) have long-term sequelae due to mode of delivery.

Most authors conclude that cesarean section is associated with increased short term maternal morbidity. However, in the present study, we found no statistical significant differences between the groups except for higher blood loss in the surgery group, which was of no clinical significance. The long term effects of cesarean section on the risk of pregnancy and delivery complications are well documented [14–16]. These include placenta previa and accrete, preterm delivery and pre- and post partum bleeding, which may involve a risk for both mother and child.

Most studies of term breech deliveries are retrospective and based on registry data, which make comparisons difficult because of lack of antenatal and postnatal information. Our study is in terms of design comparable to a large prospective observation study from France/Belgium that included 8105 women [17], According to that study, vaginal delivery of breech infants remains standard practice in France. The proportion of planned vaginal deliveries was 51% in 1998 and decreased to 31% in the study period (2001–2002). In our study, vaginal delivery was planned in 51% of breech births throughout the study period (varying from 45% in 2003 to 57% in 2010). In the study by Goffinet et al. [17], 54 (2%) in the planned vaginal group were transferred to the NICU, versus (29) 10% in our study. However, in our study only one of 29 infants was admitted at the NICU for more than 4 days, compared to 23/54 (43%) of the infants in the French/Belgian study. The main reason is probably variations in routines, including a low threshold for transfer to the NICU at our hospital, reflected by the short admission time. Cultural and traditional differences between the countries may also explain variations in management of delivery. In the French/Belgium study, duration of labor (first stage) was ≥ 7 hours in 1.4% versus 36.3% in our study. Norwegian doctors and women seem to accept longer duration of labor before cesarean section is decided. In our study active pushing started after the presenting part had reached the outlet in 93.5%, in accordance with the Goffinet study (96.4%). Furthermore, the percentage of women with active phase of the second stage of labor (i.e. pushing) longer than 60 minutes was only 2/185 (1.1%), in line with the French/Belgian study (0.2%). According to the Norwegian guidelines, active pushing should be awaited until the presenting part reaches the outlet and is thus followed in a majority of the deliveries. In contrast, active phase of second stage of labor longer than 60 minutes was 5.0% in TBT possibly due to earlier start of pushing than in our study. Shorter duration of active pushing is known to increase the chances of a vaginal birth [18].

In contrary to the findings of the TBT and several retrospective studies [1, 6–9], we found no excess risk for neonatal mortality or serious morbidity in the planned vaginal delivery group versus in the planned cesarean section group. As the rate of vaginally delivered breech infants is much higher at our institution (33%) compared to other studies, comparisons are difficult to make. Today approximately 20% of breeches are delivered vaginally in the Netherlands compared to 50% before the TBT [7]. The numbers are even lower in Scotland, Ireland and Denmark; 5 – 10% today versus 20 – 23% before the TBT [10, 19, 20].

There is a risk that reduction in vaginal breech deliveries may lead to less skilled obstetricians and less favorable results in situations where vaginal delivery is unavoidable, such as undiagnosed breech in advanced labor or delivery of second twin. In the prospective French/Belgian study, 32% of the breech deliveries went vaginally and the study included only one neonatal death of a non-malformed infant (in the cesarean section group) among 8105 term breech deliveries [17]. Furthermore, except for a 5-minute Apgar score less than 4 in a total of five infants, no difference in severe outcomes was observed between the two groups. Also, the authors of a Finnish study conclude that selectively vaginal breech deliveries may be safe in hospitals where the obstetricians have a tradition of vaginal deliveries [21]. In their study (years 1995 – 2002), 455 of 986 (46%) term breech infants were delivered vaginally. When infants with congenital malformations were excluded, they reported no perinatal deaths in their material. This is in line with the conclusions of an Austrian observation study comprising 211 breech presentation pregnancies where 46 of 85 planned vaginal deliveries actually went vaginally [22]. Common features of the French/Belgium, the Finnish, the Austrian and our study, include strict selection criteria for whom should be offered vaginal delivery, close fetal monitoring and the presence of skilled obstetricians. The relatively high frequency (15%) of undiagnosed breech presentation at birth in our study has also been found in other studies (17-21%) supporting the importance of being able to handle an unexpected breech delivery vaginally [23–25].

The present study is not sufficiently powered to demonstrate differences in perinatal morbidity or mortality between the groups, as these outcomes are rare. In a Norwegian paper on breech deliveries, sample size calculation was done based on an anticipated perinatal mortality of 0.25% in the planned vaginal group and 0.1% in the cesarean group [12]. The authors concluded that a Nordic study similar to the TBT would require a total number of 10,000 to be recruited in each group for two-sided testing, a number exceeding practical limits [12]. Randomized controlled trials are considered as the gold standard for clinical trials, but the appropriateness of this design for studies of complex clinical procedures like breech deliveries has been questioned [5, 26]. A strength of the present study is the comprehensive prospective registration conducted by midwifes at our institution. Furthermore, we were able to review the hospital records of all the infants that were admitted to the NICU after birth. In this study, the groups were compared according to planned mode of delivery before onset of birth. However, we have no information on fetuses turned to cephalic before delivery. We cannot control for all confounding factors, including lack of information about education level and maternal pregestational morbidity. Further, we can also assume that there are psychological differences between the groups, as maternal request was the reason for planned cesarean section in 41% of the cases of planned cesarean section. However, antenatal care is uniform in Norway, and more than 90% of pregnant women are followed up by their general practitioner and/or community midwife [27]. Furthermore, pre-labor, all women with known breech presentation are counseled by experienced gynecologists and selected to planned vaginal delivery after a thorough assessment.

Our findings reflect a well-functioning health care system where vaginal delivery of breech babies has been practiced even after the results of the TBT was published. The present study does not have enough power to make a definite conclusion that vaginal breech delivery is completely safe. Yet, our results indicate that vaginal delivery of term breech infants is acceptable provided the following conditions: Selection guidelines are followed, the fetal monitoring is of high quality, and the volume of breeches delivered vaginally is sufficient to maintain a high level of competence among obstetricians. The women should also be counseled about the increased risk of short-term NICU admission.