Outcomes in term breech birth according to intended mode of delivery—A Swedish prospective single‐center experience of a dedicated breech birth team

Abstract

Introduction

The appropriate mode of delivery for breech babies is a topic of ongoing debate. After the publication of the Term Breech Trial in 2000, the proportion of breech babies delivered vaginally in Sweden rapidly dropped to 7% from 26%. In 2015, international guidelines changed to once again recommend offering vaginal breech deliveries in select cases. In 2017, a Swedish hospital established a dedicated Breech Team to provide safe vaginal breech deliveries according to the new guidelines. The aim of this study is to compare neonatal morbidity in the group planned for cesarean breech delivery with the group planned for vaginal breech delivery treated in accordance with the new guidelines. The study adds to the literature by providing insights into the consequences of reintroducing vaginal breech births in a high‐resource health‐care setting.

Material and Methods

A prospective observational study was conducted at Södersjukhuset's maternity ward with 1067 women who gave birth to a single breech fetus at term. Outcomes were compared between the planned vaginal and planned cesarean delivery groups using intention‐to‐treat analysis and multivariate analysis to control for confounders.

Results

Out of the 1067 women, 78.9% were planned for cesarean delivery and 21.1% were planned for vaginal delivery. The planned vaginal group had a significantly greater risk for neonatal morbidity compared to the planned cesarean group (3.1% vs. 0.7%; OR 4.44, 95% CI 1.48–13.34). The risk difference remained significant after controlling for confounders.

Conclusions

Planned vaginal breech delivery was associated with an increased risk of neonatal mortality and short‐term morbidity compared to planned cesarean breech delivery in accordance with the new guidelines. The potential risks and benefits of planned vaginal breech delivery should be carefully weighed against those of planned cesarean delivery.

Study findings indicate higher neonatal morbidity risk associated with planned vaginal breech delivery compared to planned cesarean delivery. Even in carefully selected cases and in the presence of a dedicated breech birth team. Highlighting the importance of informed decision‐making in breech birth management.

Abbreviations

CI

confidence interval

CS

cesarean section

NICU

Neonatal Intensive Care Unit

OR

odds ratio

PCD

planned cesarean breech delivery

PREMODA

Presentation et Mode d'Accouchement: Presentation and mode of delivery

PVD

planned vaginal breech delivery

TBT

term breech trial

VBB

vaginal breech birth

Key Points.

Even in carefully chosen cases and round‐the‐clock attendance of competent midwives and obstetricians, vaginal breech delivery was associated with a small but significantly increased risk of neonatal mortality and short‐term morbidity, when compared to planned cesarean section.

1. INTRODUCTION

In about 3% of term pregnancies, the fetus is in a breech position with the bottom, feet or knee first. ¹ , ² , ³ If the fetus remains in breech position after an attempted external cephalic version, there are two options available: planned vaginal breech delivery (PVD) or planned cesarean breech delivery (PCD). ⁴ Both options are associated with risks. Vaginal breech delivery is associated with an increased risk of short‐term morbidity and asphyxia, partly related to cord compression, or more seriously, difficulties in delivering the aftercoming head. ³ , ⁵ , ⁶ While these risks can be mitigated with pre‐emptive cesarean section (CS), ⁷ CS is associated with neonatal respiratory morbidity, and increased risk of adverse outcomes in future pregnancies. ⁸

The debate on the optimal management of term breech pregnancies has been shaped by a randomized controlled trial published in 2000. The term breech trial (TBT) recruited 2088 women with breech pregnancies from 121 centers in 26 countries, randomized them either to PVD or PCD, and reported that the risk of adverse neonatal outcomes was significantly lower in PCD as compared to PVD. ¹ The results of TBT are supported by a recent meta‐analysis of 21 observational studies that found that overall PCD reduces neonatal morbidity as compared to PVD. ⁹ However, a follow‐up of TBT and meta‐studies has found no significant difference between PCD and PVD with respect to long‐term child morbidity. ⁹ , ¹⁰

Following the publication of TBT, most obstetric clinics switched from PVD to PCD, making CS the default delivery method, and significantly decreasing the number of attempted vaginal breech births (VBB). ¹¹ , ¹² , ¹³ In Sweden, it is estimated that the proportion of breech babies delivered vaginally is 7%, down from approximately 26% in 1999. ⁹ , ¹⁴ Since TBT, three retrospective cohort studies have been conducted in Sweden. Two of the three found that PCD was safer than PVD, ¹⁴ , ¹⁵ whereas the third found no significant risk reduction. ¹⁶

The results of TBT have been called into question. ⁹ , ¹⁰ , ¹² , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ For example, a large French‐Belgian prospective cohort study (PREMODA) found no significant difference between PVD and PCD with respect to neonatal morbidity. ²⁰ Unlike Sweden, vaginal breech deliveries have remained common in France and Belgium following TBT. ²¹

As a response to the criticism and new results, international guidelines changed in 2015 to once again recommend offering vaginal breech deliveries in select cases. ⁴ , ⁶ In 2017, one of Sweden's largest maternity wards at Södersjukhuset in Stockholm established a dedicated Breech Team to provide safe vaginal breech deliveries for women who meet the criteria outlined in the new international guidelines. ²²

The aim of this study is to investigate whether a dedicated breech team following the new guidelines was able to offer as safe PVD as PCD for select women. The study adds to the body of knowledge by providing a prospective cohort study of reintroducing planned vaginal breech deliveries in a high‐resource health‐care setting.

2. MATERIAL AND METHODS

Prospective data collection took place between January 1 2017 through March 17 2022 at the maternity ward at Södersjukhuset in Stockholm, involving 1067 births. The study included women who gave birth to a single breech baby at term (≥37 weeks' gestation), excluding cases of antenatal intrauterine fetal demise and lethal congenital anomalies. The study was designed to include 350 PVDs but was discontinued prematurely after 225 PVDs because of a perinatal death in the PVD group.

Our registry contains data on maternal characteristics, delivery, neonatal morbidity, and neonatal mortality. Additional data were collected from medical records for all newborns admitted to the NICU in order to confirm their medical history.

The cohorts were classified based on the intended mode of delivery. Women planned for delivery by elective CS were categorized as PCD, while those planned for vaginal delivery were categorized as PVD, irrespective of the final mode of delivery. In some cases, breech presentation was not discovered until the onset of labor. Such cases were categorized as PVD if vaginal delivery was attempted. If vaginal delivery was not attempted, they were categorized as PCD, even though the CS were most often emergency CS.

In order to become a candidate for planned vaginal breech delivery certain selection criteria had to be met (Table 1). ²² After a failed external cephalic version, information was provided about the risks and benefits of PCD and PVD, respectively, and the woman was asked whether she was interested in vaginal breech delivery. If she expressed an interest, computed tomography pelvimetry and fetal weight estimation by growth ultrasound were performed. Pelvimetry was performed by an experienced radiologist. The pelvimetric criteria for becoming a candidate for vaginal breech delivery was that the woman's pelvic inlet sagittal diameter was greater than 11.5 cm and her pelvic outlet dimensions, measured as the sum of the interspinous distance, intertuberous diameter, and sagittal pelvic outlet diameter exceed 32.5 cm. Clinical assessment replaced pelvimetry in cases of advanced labor. ²²

TABLE 1.

Criteria recommended by Region Stockholm as a basis for deciding the mode of breech delivery.

Optimal conditions for planned vaginal breech delivery	Contraindications for planned vaginal breech delivery	Risk factors that require individual assessment
*Well‐motivated patient, informed consent	*Small for gestational age, estimated weight <−2 SD	*Estimated fetal weight >4000 g
*Expected weight at partus, 2500–4000 g	*Estimated fetal weight <2500 g	*Suboptimal pelvimetry measurements
*Normal pelvimetry	*Hyperextension of neck	*Known uterus malformation
*Frank or complete breech		*Previous cesarean section
		*Induction of labor
Intrapartum criteria for continued vaginal delivery
*Good progress during the first stage of labor
*Passive second stage no longer than 1 h for multiparas and 2 h for primiparas
*Active second stage no longer than 30 min for multiparas and 60 min for primiparas
*Cardiotocography same criteria as for cephalic presentation

If the criteria were met and no contraindications were discovered, she received individual counseling from a senior obstetrician from the breech team. If she remained interested, she was planned for vaginal breech delivery.

The Breech Team consisted of approximately 25 experienced midwives and four senior obstetricians who were available around the clock. All team members had completed a specialized breech education program focusing on upright breech delivery. Additionally, they participated in external training programs as well as intercollegiate training.

Traditional maneuvers, as outlined in Reitter et al.'s 2020 analysis of the Physiological Breech Birth Algorithm, ²³ were utilized during labor when necessary. In instances where manual maneuvers were insufficient to resolve the entrapment of the after‐coming head, Piper's forceps were employed.

The primary outcome was a composite of severe short‐term fetal outcomes, including perinatal mortality at less than 28 days of age and severe neonatal morbidity defined as seizures >24 h, intubation and ventilation >24 h, Apgar <4 at 5 min, brachial plexus injury, parenteral feeding >4 days, parietal skull fracture or NICU admission >4 days. This outcome is identical to the primary outcome used by PREMODA that, in turn, is similar to that of TBT.

Secondary outcomes were the following: maternal morbidity defined as postpartum hemorrhage >1000 mL, neonatal morbidity defined as umbilical cord pH <7.0, Umbilical cord BE <−16, Apgar <7 at 5 min, Apgar <6 at 10 min, and any birth injury.

The primary objective was to study how the intended mode of delivery affected neonatal outcomes.

2.1. Statistical Analyses

The two groups were compared for neonatal morbidity criteria. Unadjusted logistic regressions with an intended mode of delivery as exposure and morbidity as outcome were performed to calculate unadjusted odds ratios for planned vaginal delivery. Adjusted logistic regressions were performed to control for possible confounders and adjusted odds ratios were calculated for planned vaginal delivery.

Qualitative analysis was performed to identify possible confounders. It was hypothesized that both exposure and outcomes may be affected by the woman's age, ²⁴ maternal pre‐pregnancy body mass index (based on first‐trimester weight and self‐reported length), ²⁵ previous cesarean sections, ⁸ nulliparity, ²⁶ maternal pregnancy complications such as gestational diabetes mellitus or preeclampsia, ²⁷ , ²⁸ as well as the child's birth weight. ²⁹ The analysis is described in a directed acyclic graph (Figure 1). The groups were compared with a chi‐squared test (or exact Fischer test when necessary) for the analysis of categorical variables, and a two‐sided p‐test value of <0.05 was considered as significant. Statistical comparisons and logistic regressions were made with R (R Core Team, R Foundation for Statistical Computing, Vienna, Austria).

FIGURE 1.

Directed acyclic graph describing possible confounders. The blue ellipse represents the primary outcome, the yellow ellipse represents the exposure, and the pink ellipses represent possible confounders. Arrows represent causal influence.

3. RESULTS

During the study period, from January 2017 to mid‐March 2022, 30 110 singleton term births took place at Södersjukhuset, one of Sweden's largest maternity wards. During this period, 1071 singleton term breech babies were delivered at term, resulting in a rate of 3.6% singleton term breech deliveries. Three cases of intrauterine fetal demise and one case of lethal congenital anomaly were excluded. The latter was a neonate with severe congenital ichthyosis that passed at the neonatal intensive care unit (NICU) 12 days postpartum. Thus, 1067 breech babies were eligible for inclusion in the study (Figure 2). Out of the 1067, 842 (78.9%) were planned for cesarean delivery and 225 (21.1%) were planned for vaginal delivery. The conversion rate from PVD to cesarean delivery was 21% including the cases of cesarean section before onset of labor due to prolonged pregnancy. The annual rate of VBB at the site (more than) tripled compared to preceding years, increasing from approximately 9–16 cases per year to around 45 cases annually.

FIGURE 2.

Study flow chart. Term singleton breech deliveries during the study period elegible for inclusion. Delivery mode and reason for emergency cesarean delivery during labor. Abbreviations: IUFD, intrauterine fetal demise; LCA, lethal congenital anomaly.

Women in the PCD group were significantly more likely to have had a previous cesarean section (p < 0.001) than women in the PVD group (Table 2). Additionally, infants born to women in the PVD group had a significantly higher gestational age and a significantly lower birth weight (p < 0.001) compared to those in PCD.

TABLE 2.

Maternal and obstetric characteristics in the term singleton breech population according to intended mode of delivery; planned cesarean delivery and planned vaginal delivery, during the study period January 2017 to March 2022.

Maternal characteristic	Planned cesarean delivery (n = 842)	Planned vaginal delivery (n = 225)	p
Age at partus, median years (1st quartile, 3rd quartile)	33 (30,36)	32 (30,35)	0.056
BMI, median kg/m2 (1st quartile, 3rd quartile)	23.15 (21.09,25.82)	22.54. (2071,25.06)	0.07
Nullipara, n (%)	564 (66.98%)	159 (70.67%)	0.332
Smoker 3 months before pregnancy, n (%)	73 (8.67%)	15 (6.67%)	0.404
Previous cesarean section, n (%)	132 (15.68%)	3 (1.33%)	<0.001
Maternal complications (PE or GDM), n (%)	40 (4.75%)	3 (1.33%)	0.997
Gestational age at partus, median days (1st quartile, 3rd quartile)	272 (270274)	277 (271282)	<0.001
Birthweight, median g (1st quartile, 3rd quartile)	3400 (3130,3703.75)	3280 (3040,3512)	<0.001
Above median weight, n	444 (52.73%)	91 (40.44%)	0.001
Boys, n	395 (46.91%)	102 (45.33%)	0.729

Note: For continuous variables, medians with interquartile range are expressed and for categorical variables numbers and proportions are expressed, p‐values for testing non‐equivalence of ratios. Maternal complications are defined as either pregnancy‐induced diabetes mellitus or preeclampsia.

Abbreviations: BMI, body mass index; GDM, gestational diabetes mellitus; PE, preeclampsia.

Out of the 225 planned vaginal deliveries, 177 resulted in vaginal births, while out of the 842 planned cesarean deliveries, four ended up as vaginal births, resulting in a total of 181 vaginal births. The occurrence of 0.5% unplanned vaginal breech deliveries in the PCD group consisted of women who arrived at the maternity ward in advanced labor. Difficulty during expulsion, as subjectively defined by the attending obstetrician, was experienced in 25 (15.4%) of the 181 vaginal deliveries (Table 3). Adherence to the established regional guidelines for antenatal criteria was observed in all cases. There were two cases of active pushing lasting more than 1 h violating the guideline's intrapartum criteria. Pelvimetry was conducted in 172 (76.4%) of 225 PVDs. The absence of pelvimetry in some of the PVD occurred exclusively in cases diagnosed late in labor and was replaced by clinical assessment. The fraction of adverse neonatal outcomes was 3% and 2% in the pelvimetry and non‐pelvimetry groups, respectively. Thus, the absence of pelvimetry was not associated with an increased risk for neonatal morbidity.

TABLE 3.

Management of labor and delivery in the 181 vaginal breech births.

Characteristics	Number, %
Pelvimetry performed, n (%)	134 (74.03%)
Epidural, n (%)	77 (42.54%)
Oxytocin, n (%)	143 (79.89%)
Duration of labor >5 cm dilation <4 h	88 (48.62)
Duration of labor >5 cm dilation 4–6 h	49 (27.07)
Duration of labor >5 cm dilation >7 h	44 (24.31)
Duration of the active phase of second stage of labor <30 min, n (%)	135 (75%)
Duration of the active phase of the second stage of labor 30–60 min, n (%)	43 (23.89%)
Duration of the active phase of the second stage of labor >60 min, n (%)	2 (1.11%)
Difficulty during expulsion, n (%)	25 (15.43%)
Maneuvers for delivery of extended arms, n (%)	55 (32.93%)
Maneuvers for delivery of fetal head, n (%)	55 (33.54%)
Episiotomy, n (%)	46 (25.41%)
Rupture grade 1 or 2, n (%)	64 (35.36%)
Rupture grade 3 or 4, n (%)	5 (2.76%)

Note: For categorical variables numbers and proportions are expressed.

As seen in Table 4, there was no significant difference between the groups with respect to children transferred to the NICU (5.8% in the PVD group vs. 5.0% in the PCD group; OR 1.16, 95% CI 0.61–2.20). Medical records of the neonates who were transferred to the NICU were examined. Three infants were diagnosed with Down's syndrome (two of these in the PVD group). These were not excluded from the analyses. Two cases of severe asphyxia leading to hypoxic–ischemic encephalopathy and later cerebral palsy were reported, one in the PVD group and one in the PCD group. The latter case consisted of an infant born vaginally though planned for cesarean delivery, due to a precipitous labor (within 20 min of the mother's arrival to the maternity ward) in gestational Week 38 and 2 days. There was one intrapartum death in the PVD group related to the rare and potentially serious dystocial complication entrapment of the after‐coming head. The autopsy of the infant later showed signs of asymmetrical intrauterine growth restriction and a protruding occiput with relative macrocephaly, a known risk factor for entrapment of the after‐coming head. ³⁰ None of the other four infants in the PCD group and five infants in the PVD group categorized as having serious neonatal morbidity exhibited any long‐term sequelae in follow‐up examinations.

TABLE 4.

Neonatal mortality and morbidity outcomes in the planned breech population according to intended mode of delivery.

Neonatal and maternal morbidity outcomes	Planned vaginal delivery (n = 225)	Planned cesarean delivery (n = 842)	OR
Apgar at 5 min <4, n %	5 (2.2%)	2 (0.2%)	9.55 [1.84, 49.53]
Apgar at 5 min <7, n %	10 (4.4%)	5 (0.6%)	7.79 [2.63, 23.02]
Apgar at 10 min <6, n %	4 (1.8%)	1 (0.1%)	15.22 [1.69, 136.88]
Umbilical cord pH <7, n %	5 (3.1%)	4 (0.5%)	6.04 [1.60, 22.76]
Umbilical cord BE <−16, n%	4 (2.5%)	3 (0.4%)	6.54 [1.45, 29.50]
Any injury, n %	3 (1.3%)	0 (0%)	0 [0, NaN]
Long bone of clavicle fracture, n %	1 (0.5%)	0 (0%)	0 [0, NaN]
Other fractures, n %	0 (0%)	0 (0%)	NaN [NaN, NaN]
Brachial plexus injury, n %	1 (0.5%)	0 (0%)	0 [0, NaN]
Other injuries, n %	1 (0.5%)	0 (0%)	0 [0, NaN]
Transfer to NICU, n %	13 (5.8%)	42 (5.0%)	1.16 [0.61, 2.20]
NICU >4 days, n %	2 (0.9%)	3 (0.4%)	2.49 [0.41, 14.98]
Intubation, n %	2 (0.9%)	0 (0%)	0 [0, NaN]
Intubation persistent after 24 h, n %	0 (0%)	0 (0%)	NaN [NaN, NaN]
Convulsion, n%	1 (0.5%)	1 (0.1%)	3.75 [0.23, 60.17]
Continued after first 24 h, n%	1 (0.5%)	0 (0%)	0 [0, NaN]
Parenteral feeding, n %	4 (1.8%)	7 (0.8%)	2.16 [0.63, 7.43]
Parenteral feeding >4 days, n %	2 (0.9%)	5 (0.6%)	1.5 [0.29, 7.78]
ICH or IVH	0 (0%)	0 (0%)	NaN [NaN, NaN]
Perinatal mortality, n %	1 (0.4%)	0 (0%)	0 [0, NaN]
Neonatal mortality or serious neonatal morbidity, n %	7 (3.1%)	6 (0.7%)	4.44 [1.48, 13.34]

Note: Two possible plans: planned vaginal delivery or planned cesarean delivery. Frequencies, ratios, and odds ratios with 95% confidence interval. Neonatal mortality or serious neonatal morbidity defined as either 5‐min Apgar <4, Brachial plexus injury, Parietal skull fracture, NICU >4 days, Intubation persistent after 24 h, Convulsion continued after first 24 h, Parenteral feeding >4 days, or Perinatal mortality.

Abbreviations: ICH, neonatal intracranial hemorrhage; IVH, neonatal intraventricular hemorrhage; OR, odds ratio.

Among the individual outcomes in the primary composite outcome, a significant difference was observed only with respect to 5‐min Apgar score <4 where infants in the PVD group exhibited a less favorable outcome than infants in the PCD group (2.2% vs. 0.2%; OR 9.55, 95% CI; 1.84–49.53). Among the secondary outcomes, a significant difference was observed between the PVD group and PCD group with respect to the following outcomes: 5‐min Apgar score <7 (OR 7.79, 95% CI 2.63–23.02); 10‐min Apgar score <6 (OR 15.22, 95% CI 1.69–136.88); umbilical cord pH <7 (OR 6.04, 95% CI 1.60–22.76); and umbilical cord BE <−16 (OR 6.54, 95% CI 1.45–29.5).

Finally, infants in the PVD group were significantly more likely than those in the PCD group to experience the primary composite outcome of neonatal mortality or serious neonatal morbidity (3.1% vs. 0.7%, OR 4.44, 95% CI 1.48–13.34). The absolute risk increase associated with PVD over PCD was 2.4%, which corresponds to a number needed to harm of 42.

Adjusted odds ratios were calculated for the planned breech delivery population by controlling for potential confounders: nullipara, previous cesarean section, maternal age ≥35, BMI ≥25, maternal complications, as well as birth weight above the median. After controlling for potential confounders, the risk for neonatal morbidity was significantly greater for infants born to women by PVD than by PCD (adjusted OR 4.72, 95% CI 1.38–16.15). The adjusted OR was not significant for any of the other risk factors (Table 5).

TABLE 5.

Logistic regression for associations between risk factors and the outcome of neonatal mortality or serious morbidity in the breech population.

	Neonatal morbidity, n (%)	Total	Unadjusted OR [95% CI]	Adjusted OR [95% CI]
All breech deliveries	13 (1.2%)	1067	‐	‐
Planned cesarean delivery	6 (0.7%)	842	Ref	Ref
Planned vaginal delivery	7 (3.1%)	225	4.44 [1.48, 13.34]	4.72 [1.38, 16.15]
Nullipara, no	3 (0.9%)	344	Ref	Ref
Nullipara, yes	10 (1.4%)	723	1.58 [0.43, 5.77]	2.87 [0.35, 23.58]
Previous section, no	11 (1.2%)	932	Ref	Ref
Previous section, yes	2 (1.5%)	135	1.28 [0.28, 5.85]	5.77 [0.46, 71.70]
Age at partus <35	10 (1.4%)	697	Ref	Ref
Age at partus ≥35	3 (0.8%)	370	0.57 [0.16, 2.08]	0.69 [0.17, 2.76]
BMI <25	7 (1.0%)	752	Ref	Ref
BMI ≥25	5 (1.6%)	360	1.63 [0.51, 5.17]	1.92 [0.59, 6.27]
Maternal complications, no	13 (1.3%)	1024	Ref	Ref
Maternal complications, yes	0 (0%)	43	0 [0, NaN]	0 [0, Inf]
Child's weight above median, no	7 (1.3%)	532	Ref	Ref
Child's weight above median, yes	6 (1.1%)	535	0.86 [0.29, 2.56]	1.20 [0.38, 3.80]

Note: Adjusted odds ratios for exposure and potential confounder, calculated using a logistic regression model adjusted for the remaining potential confounders, i.e. maternal age, BMI, previous cesarean section, nulliparity, maternal pregnancy complication, child's birth weight, and mode of delivery. Neonatal mortality or serious morbidity defined as either: 5‐min Apgar <4, Brachial plexus injury, Parietal skull fracture, NICU >4 days, Intubation persistent after 24 h, Convulsion continued after first 24 h, Parenteral feeding >4 days, or Perinatal mortality. Maternal pregnancy complications defined as preeclampsia or gestational diabetes mellitus.

Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio.

4. DISCUSSION

The main finding was a significant increase in the primary composite outcome of mainly short‐term fetal morbidity for infants in breech presentation planned for vaginal delivery, as compared to infants planned for elective cesarean delivery, even after strict case selection by a dedicated and specially trained team. This result is in line with previous research that has established that PVD is associated with increased risk for neonatal morbidity. ¹ , ⁸ , ⁹ , ³¹ With respect to neonatal morbidity, the number needed to harm of 42 was greater than TBT's number needed to harm of 14 ¹ ; but less than that reported by PREMODA that reported no statistically significant difference between PCD and PVD. ²⁰

The higher risk associated with PVD in our sample, as compared to that in, for example, PREMODA's, Maier's ³² and Vistad's ³³ samples, may partly be explained by differences in the experience of managing vaginal breech births. Unlike many other countries, France, Belgium, Austria, and Norway have not dramatically decreased the proportion of vaginal breech deliveries following the publication of TBT. In France, Collège National des Gynécologues et Obstétriciens Français reported that one‐third of women with breech pregnancy are planned for vaginal delivery ²¹ ; in Vistad's Norwegian study, 50% of all breech deliveries were planned vaginally ³³ ; and a retrospective analysis of 12 years' experience at a single center in Austria reports a trial of vaginal breech delivery in more than 70% of the cases. ³² In Sweden, on the other hand, the rate of vaginal breech deliveries dropped from approximately 26% in 1999 to 6% in the years following the publication of TBT. ¹⁴ It may be hypothesized that the lower conversion rate from PVD to cesarean delivery seen in our study as compared to other studies ³² , ³³ has had an impact on the outcome. However, this remains uncertain as a liberal approach to conversion from VBB to cesarean delivery is already adopted.

To be considered a candidate, specific criteria must be met, including a pelvimetric measurement within a reference, an estimated fetal weight between 2500 and 4000 g, and being considered well‐motivated. ⁴ , ²² Whereas the first two are established objectively with pelvimetry and ultrasound, the third requires a degree of subjective judgment.

If this is correct, stricter selection criteria may be warranted. It is noteworthy that guidelines of optimal fetal weight considering VBB differ internationally. According to the Collège National des Gynécologues et Obstétriciens Français and The Royal College of Obstetricians and Gynecologists, there is an admitted heightened risk for adverse outcomes if the fetal weight estimation exceeds 3800 g at birth and as such constitutes an independent indication for PCD. ²¹ , ³⁴ The Society of Obstetricians and Gynecologists of Canada, ⁷ Norwegian guidelines, ³⁵ as well as our regional guidelines, ²² on the other hand, use a more liberal upper limit of 4000 (−4500) grams. Applied to our sample, adherence to stricter estimated fetal weight restrictions (eg 3000–3800 g) could have led to the avoidance of one case of cerebral palsy in the PVD group where the birth weight exceeded 4000 g and the case of intrapartum death with a birth weight of 2376 g. The guiding principle being that it is better to err on the side of caution and strictly adhere to more narrow selection criteria.

A possible objection to the primary composite outcome is that it does not capture what is clinically relevant about neonatal morbidity. ³⁶ The composite outcome in our study is mainly driven by short‐term perinatal outcomes such as Apgar <4 at 5 min. It may be argued that since long‐term outcomes are clinically more significant than short‐term outcomes, an outcome that better captures long‐term outcomes should have been used. On this account, it would have been better with a composite outcome that includes HIE, cerebral palsy, and mortality. However, it should be noted that our composite outcome is far from clinically insignificant since Apgar <4 at 5 min is associated with worse long‐term outcomes. Furthermore, parents prefer to give birth to uninjured infants, even if the injuries do not lead to sequelae. Finally, the main reason we have decided to use the composite outcome is that it allows us to compare our results with the results of other influential studies, such as TBT and PREMODA.

It may be tempting to conclude that because PCD is safer than PVD, vaginal breech births should be abandoned in favor of planned CS. However, the increased risks of vaginal breech deliveries have to be weighed against the risks of performing more CS. Notable augmented risks associated with repeat delivery after CS include abnormally invasive placenta, uterine abruption, significant blood loss, abdominal adhesions, and organ injury. ³⁷ , ³⁸ Long‐term risks of uterine scarring may lead to life‐threatening complications in future pregnancies. ¹⁹ , ³⁹ Furthermore, by removing the option of VBB, women's autonomy would be limited. Since patient autonomy is broadly recognized as an ethical value in maternity care, ⁴⁰ not offering the option of PVD would entail an ethical cost that has to be weighed against the gain of lowering the neonatal risk.

A major strength of this study is that it provides a contemporary study population at one of Sweden's largest maternity hospitals within the setting of a consistent and controlled environment. The fact that it is a single‐center study ensures adherence to the same protocol, reducing the variability seen in larger multicenter studies. Another major strength is the detailed information about the deliveries and the decisions leading to the choice of management and mode of delivery. The prospectively collected data is of high quality because of the access to complete and accurate patient records, offering a reliable source of information while reducing the risk of measurement bias and errors. Finally, the study adds to the literature, as it sheds light on the effects of offering vaginal breech deliveries in a health‐care setting with a low rate of such births.

However, despite the sample size being large enough to calculate odds ratios and confidence intervals, it was smaller than originally intended. The original plan was to include 350 planned vaginal deliveries, but the program was stopped following an intrapartum death related to entrapment of the aftercoming head in March 2022, during an ongoing investigation of the event. The premature discontinuation of the study rendered a smaller sample and may have affected the results, although the extent of this effect is uncertain.

Further limitations to our trial are mostly associated with the obvious limitations of the study design. Being a single‐center observational study, the restricted sample size, demographics, and timeframe lead to a lack of generalizability as these may not reflect the diversity found in, or be representative of, other settings. The sample size, though substantial, may not have been large enough to provide enough statistical power to detect significant differences for the rare events and outcomes measured in the study, leading to the lack of statistically significant results.

Finally, it is unfortunate that this study did not take race, ethnic or socioeconomic background into account as a risk factor for adverse neonatal and maternal outcomes. After all, it has been repeatedly pointed out in an American setting that black women and infants are more likely to experience adverse outcomes than non‐black women and infants. ⁴¹ Similarly, it has been pointed out that low socioeconomic status is associated with adverse pregnancy outcomes. ⁴²

A question for future research would be to evaluate whether the risk associated with VBB is comparable to other recognized and legitimized risk deliveries where the trial of labor remains an option. The objective is to estimate whether planned breech delivery is feasible without compromising more than tolerable safety. Although the determination of tolerable risk is of nearly philosophical nature, it becomes pertinent in consideration of pre‐emptive CS in lieu of vaginal delivery.

5. CONCLUSION

When compared to planned cesarean breech delivery, planned vaginal breech delivery was associated with a small but significantly increased risk of neonatal mortality and short‐term morbidity. The increased risk should be communicated, when counseling women about modes of delivery for fetuses in breech presentation.

AUTHOR CONTRIBUTIONS

Emma Debora Bartfai: investigation, writing the original draft. Nicolas Yaouzis Olsson: investigation, data curation, formal analysis, writing the original draft. Hanna Åmark, Tove Wallström: conceptualization, methodology, supervision, writing: review and editing.

CONFLICT OF INTEREST STATEMENT

None.

ETHICS STATEMENT

The study was approved by the Swedish Ethical Review Authority (Approval Number: 2019‐064150414 on 14 April 2020). The study complied with the World Medical Association Helsinki Declaration regarding the ethical conduct of research involving human subjects.

Yaouzis Olsson N, Bartfai ED, Åmark H, Wallström T. Outcomes in term breech birth according to intended mode of delivery—A Swedish prospective single‐center experience of a dedicated breech birth team. Acta Obstet Gynecol Scand. 2024;103:2296‐2305. doi: 10.1111/aogs.14945