Objective
We sought to identify factors associated with adverse perinatal outcomes (APO) among term breech neonates with planned vaginal deliveries.
Study Design
We conducted univariable and multilevel multivariable analysis of the data collected in the multicenter prospective observational study PREsentation et MODe d’Accouchement (PREMODA) in women with planned vaginal delivery giving birth to singleton term breech babies. The end point was a composite set of APO.
Results
Of 2502 women with planned vaginal delivery recruited in the 174 participating centers, 1772 (71%) delivered vaginally. Adverse outcomes were observed in 165 cases (6.59%). After adjustment, the factors associated with them were geographic origin, gestational age <39 weeks at birth, birthweight <10th percentile, and annual number of maternity unit births <1500.
Conclusion
When strict conditions governed the selection of delivery route and management of labor was rigorous, APO were not associated with any prenatal or peripartum obstetric factors.
The publication of the Term Breech Trial (TBT) in 2000 abruptly changed clinical practices around the world, dramatically increasing the cesarean rate for breech births in those countries that had not already adopted routine cesareans for breech presentations. The trial’s conclusions, that neonatal mortality and severe morbidity were higher in the births with trial of labor than in those with planned cesarean deliveries, must nonetheless be interpreted in the light of its limitations. Its protocol for selecting women for vaginal delivery and for managing labor resulted in a lack of external validity in the eyes of many. Furthermore, at 2 years of age, the risk of death or neurodevelopmental delay did not differ between the children of the 2 groups, and several retrospective single-center series comparing planned vaginal delivery with planned cesarean delivery have demonstrated the safety of vaginal delivery in settings where the staff is experienced and the selection criteria for planned vaginal delivery strict. In 2006, the PREsentation et MODe d’Accouchement (PREMODA) study, a descriptive prospective multicenter study conducted in 174 centers in France and Belgium ( Appendix ) and including >8000 breech deliveries, reported that risks of perinatal mortality or serious neonatal morbidity were not higher in planned vaginal, compared with planned cesarean, deliveries (1.6% vs 1.45%, odds ratio [OR], 1.1; 95% confidence interval [CI], 0.75–1.61). This combination of reports has led some to reconsider what might have been hasty reactions. The American Congress of Obstetricians and Gynecologists, which stated in 2001 that planned vaginal delivery might no longer be appropriate, revised their guidelines to state that such delivery might be reasonable under hospital-specific protocol guidelines for both eligibility and management. The Society of Obstetricians and Gynaecologists of Canada has reached similar conclusions.
Residual complications for term breech fetuses undergoing trial of labor nonetheless remain possible. A better understanding of the factors associated with adverse perinatal outcomes (APO) in planned vaginal deliveries could improve screening of patients eligible for trial of labor, but few studies have sought to identify these.
The purpose of this secondary analysis of the PREMODA data was to estimate the effect of individual, practice, and institutional factors on the risk of APO in term breech births for which vaginal delivery was planned.
Materials and Methods
Patients
The PREMODA study was designed to compare perinatal outcomes according to planned mode of delivery for term breech births in France and Belgium; its methods and results are reported elsewhere. The secondary analysis presented here focuses on the planned vaginal delivery group. Using specific data collected for this purpose about the process for deciding mode of delivery and the management of labor and delivery, we sought to identify the individual, practice, and institutional factors associated with an increased risk of APO.
The data were prospectively collected in volunteer maternity units from June 1, 2001, through May 31, 2002, in 138 centers in France, for 232,999 births, and from Jan. 1, 2002, through Dec. 31, 2002, in 36 centers in Belgium, for 31,106 births. The study was approved by the French Data Protection Authority.
The study included all women giving birth in a participating maternity unit to a singleton fetus in breech presentation >37 weeks’ gestation, alive or not. Detailed reports were obtained for all deaths before discharge and transfers to neonatal intensive or intermediate care units. All existing autopsy reports were sought and obtained. All congenital anomalies and reasons for hospitalization were coded according to the International Statistical Classification of Diseases, 10th Revision . All deaths before discharge–fetal, neonatal, and postneonatal–were reviewed by an independent expert committee to determine the cause of each death and whether a planned cesarean delivery at 39 weeks might have prevented it.
In 2001, vaginal delivery of fetuses in breech presentation remained an acceptable option in France and Belgium. During the study period, the recommendations by the Collège National des Gynécologues Obstétriciens Français (French national college of obstetricians and gynecologists) served as the reference for determining mode of delivery ( Table 1 ).
| Criteria recommended by CNGOF in 2001 |
|---|
| Normal pelvimetry (x-ray or computed tomography) |
| Anteroposterior diameter of inlet ≥105 mm a |
| Transverse diameter of inlet ≥120 mm a |
| Transverse interspinous diameter of midpelvis ≥100 mm a |
| No hyperextension of fetal head (ultrasonography) |
| Fetal weight estimated between 2500-3800 g |
| Frank breech |
| Continuous electronic fetal heart rate monitoring during labor |
| Patient’s informed consent |
a Measurements commonly accepted as defining normal pelvimetry.
Outcomes
An APO was defined as ≥1 of the following: fetal mortality after the delivery route decision, neonatal mortality before discharge, 5-minute Apgar score <7, tracheal intubation, seizures <24 hours of age, tube feeding for at least 4 days, admission to the neonatal intensive care unit for >24 hours, admission to an intermediate care unit for >4 days, birth trauma including subdural hematoma, intracerebral or intraventricular hemorrhage, spinal cord injury, basal skull fracture, peripheral-nerve injury present at discharge, or clinically significant genital injury.
Factors studied
We studied 3 types of factors: individual, practice, and institutional.
Individual factors included maternal characteristics such as age at delivery, geographic origin (native country), educational level, parity, previous vaginal delivery, and previous cesarean delivery. It also included neonatal characteristics such as gestational age at birth, birthweight, and sex. At birth, small for gestational age was defined as birthweight ≤10th percentile according to the French birthweight standards.
Practice factors included prenatal care, ie, factors related to management of labor and delivery: type of breech (frank or not), premature rupture of membranes (yes/no), induction of labor (yes/no), labor augmentation with intravenous oxytocin (yes/no), lack of progress of cervical dilatation for at least 1 hour between 5-10 cm (yes/no), duration of first stage of labor between 5-10 cm, duration of passive phase of the second stage of labor, duration of active pushing (≤20 minutes, >20 minutes), and status of medical staff present at delivery (midwife alone, junior physician without senior, senior obstetrician). The presence or absence of a formal prenatal decision about mode of delivery, recorded in the medical file, was also collected, as were each of the criteria for breech vaginal delivery listed in the French national guidelines (listed in Table 1 ).
Institutional factors were those related to the institution in which delivery occurred, specifically, the number of annual births and level of care. In France, maternity units are ranked in 4 levels according to the neonatal care they provide: 1 (maternity unit with a pediatrician but no neonatology unit), 2a (maternity unit with a special neonatal care nursery), 2b (maternity unit with an intermediate neonatal care unit able to provide mechanical ventilation), and level 3 (maternity unit with neonatal intensive care unit).
Analysis
All characteristics were studied in patients included in the planned vaginal group except characteristics of the late stages of labor, which were studied only in patients who actually delivered vaginally.
Categorical variables were compared with a χ 2 test or Fisher exact test if required. Continuous variables were compared by Student t test. A P value < .05 was considered significant.
Crude OR and 95% CI were calculated for all factors studied in the univariable analysis.
The numerous candidate variables for multivariable analysis were grouped in 5 categories (maternal, pregnancy, labor, neonatal, and institutional characteristics) and analyzed in 5 separate multivariable models. Factors included in these models were those with P < .25 in the univariable analysis. Because gestational age and birthweight are standard confounders, they were forced in the models. Factors considered significant ( P < .20) were selected for the final multivariable analysis model.
We used multilevel logistic regression models to take into account the hierarchical nature of data, ie, patients (level 1) nested within maternity units (level 2). These models consider correlations that may exist for outcomes at a given center. By taking such “cluster” effects into account, these models can better estimate the effects associated with predictor variables and their variance, particularly those associated with level-2 (here maternity-level) characteristics.
The analysis of factors related to labor was performed on the entire population, while analyses of the factors related to end of labor and delivery were limited to women with vaginal deliveries. Only the final model is presented in this report.
Statistical analysis was performed with Stata software, Version 9 (StataCorp, College Station, TX) and HLM (SSI, Lincolnwood, IL).
Results
During the 12-month study period, PREMODA included 8105 women: 1133 in Belgium and 6972 in France. During this period, 264,105 births occurred in the 174 participating centers, for a rate of singleton term fetuses in breech presentation of 3.1%. The planned vaginal delivery group comprised 2526 women; 24 were excluded because of lethal congenital abnormalities. Among the 2502 women finally analyzed, 1772 (71%) delivered vaginally.
Mean maternal age was 29.9 years (SD 5) and 15.4% were age >35 years. The principal outcome measure, APO, was observed in 165 newborns (6.59%) ( Table 2 ), including 2 fetal deaths. The first death involved a woman with planned vaginal delivery after a previous cesarean section; she delivered a stillborn baby during an emergency cesarean during labor for uterine rupture. The second mother had had 5 previous vaginal deliveries and missed several prenatal visits. Fetal death was diagnosed at 39 weeks and 6 days of gestation, but remained unexplained. There were no neonatal deaths.
| Perinatal outcome | Planned vaginal delivery N = 2502 n (%) |
|---|---|
| Birth trauma a , b | 8 (0.32) |
| Brachial plexus injury | 5 (0.2) |
| Parietal skull fracture | 1 (0.04) |
| Significant genital injury | 2 (0.08) |
| Subdural hematoma a , b | 0 |
| Intraventricular hemorrhage a , b | 1 (0.04) |
| 5-min Apgar <7 b | 37 (1.48) |
| 5-min Apgar <4 a | 4 (0.16) |
| Tracheal intubation b | 26 (1.04) |
| Tracheal intubation and ventilation >24 h a | 10 (0.4) |
| NICU hospitalization >24 h b | 42 (1.68) |
| NICU hospitalization >4 d a | 23 (0.92) |
| Intermediate care unit hospitalization >4 d b | 78 (3.12) |
| Seizure a , b | 4 (0.16) |
| Parenteral or tubal feeding >4 d b | 15 (0.60) |
| Fetal death a , b | 2 (0.08) |
| Neonatal death a , b | 0 |
| Fetal and neonatal mortality or serious neonatal morbidity | 40 (1.60) |
| Adverse perinatal outcome | 165 (6.59) |
a Criteria included in combined outcome “Fetal and neonatal mortality or serious neonatal morbidity”;
b Criteria included in combined outcome “Adverse perinatal outcome.”
Univariable analysis
Maternal age >35 years, geographic origin, and educational level below secondary school were the only individual maternal factors associated with an increased risk of APO ( Table 3 ). No pregnancy or labor factors were associated with increased risk. Among the neonatal factors, gestational age at birth of 37 and 38 weeks, compared with 39-40 weeks, and birthweight <10th percentile were associated with this risk ( Table 3 ). Birthweight >90th percentile was not associated with APO.
| Factors | APO, n = 165 n (%) | OR (95% CI) |
|---|---|---|
| Maternal characteristics | ||
| Maternal age >35 y (n = 380) | 34 (9) | 1.49 (1.02–2.22) |
| Geographic origin | ||
| Europe (n = 2047) | 123 (6) | 1 |
| Sub-Saharan Africa (n = 59) | 6 (10.2) | 1.77 (0.75–4.20) |
| North Africa (n = 182) | 14 (7.7) | 1.30 (0.73–2.32) |
| Other (n = 87) | 13 (14.9) | 2.75 (1.48–5.09) |
| Educational level ≤secondary school (n = 553) | 43 (7.8) | 1.52 (1.02–2.28) |
| Nulliparous (n = 1181) | 84 (7.1) | 1.18 (0.86–1.61) |
| No previous vaginal delivery (n = 1201) | 87 (7.2) | 1.23 (0.90–1.69) |
| Uterine scar (n = 46) | 4 (8.7) | 1.36 (0.48–3.84) |
| Pregnancy care | ||
| Cephalic version trial (n = 790) | 47 (29) | 0.85 (0.60–1.21) |
| No x-ray pelvimetry (n = 426) | 36 (22) | 1.39 (0.94–2.04) |
| Pelvic measurements | ||
| Anteroposterior diameter of inlet <105 mm (n = 25) | 2 (1.6) | 1.30 (0.30–5.55) |
| Transverse diameter of inlet <120 mm (n = 228) | 18 (14.6) | 1.33 (0.79–2.24) |
| Transverse interspinous diameter <100 mm (n = 236) | 13 (11.3) | 0.87 (0.48–1.58) |
| No US assessment of fetal head flexion (n = 874) | 66 (42) | 1.30 (0.93–1.80) |
| No formal decision of delivery route before labor (n = 354) | 31 (19) | 1.42 (0.94–2.14) |
| Labor | ||
| Complete breech (n = 744) | 59 (36.9) | 1.32 (0.95–1.85) |
| PROM (n = 491) | 37 (24.3) | 1.22 (0.83–1.80) |
| Induction of labor (n = 222) | 13 (7.9) | 0.87 (0.49–1.56) |
| No labor augmentation with oxytocin (n = 519) | 39 (36.1) | 1.43 (0.95–2.15) |
| Lack of progress in dilatation a (n = 470) | 33 (22.5) | 1.12 (0.75–1.67) |
| Neonate characteristics | ||
| Gestational age at delivery, wk | ||
| 37 (n = 282) | 33 (20.1) | 2.51 (1.62–2.87) |
| 38 (n = 491) | 40 (24.4) | 1.68 (1.12–2.51) |
| 39-40 (n = 1394) | 70 (42.7) | 1 |
| ≥41 (n = 334) | 21 (12.8) | 1.27 (0.77–2.10) |
| Birthweight percentile | ||
| <10th (n = 142) | 37 (22.7) | 6.15 (4.05–9.36) |
| 10-90th (n = 2142) | 116 (71.2) | 1 |
| >90th (n = 202) | 10 (6.1) | 0.91 (0.47–1.76) |
| Male | 82 (50) | 1.25 (0.91–1.72) |
| Institutional characteristics | ||
| Maternity unit activity, birth/y | ||
| ≤1500 (n = 663) | 54 (32.8) | 1.85 (1.18–2.91) |
| 1500-2000 (n = 700) | 32 (19.4) | 1 |
| 2000-2500 (n = 566) | 36 (21.8) | 1.42 (0.87–2.31) |
| >2500 (n = 573) | 43 (26.1) | 1.69 (1.06–2.71) |
| Level of maternity unit | ||
| 1 (n = 422) | 16 (11.4) | 0.49 (0.28–0.88) |
| 2a (n = 346) | 16 (11.4) | 0.61 (0.35–1.08) |
| 2b (n = 667) | 48 (34) | 0.98 (0.66–1.46) |
| 3 (n = 834) | 61 (43.3) | 1 |
An annual number of births <1500 was an institutional characteristic associated with an increased risk of APO, and birth at a level-1 compared with a level-3 facility was associated with reduced risk ( Table 3 ).
In the univariable analysis limited to the subgroup of women who delivered vaginally, the only factor related to the end of labor and delivery associated with an increased risk of APO was pushing efforts >20 minutes ( Table 4 ).
| Factors | No APO n = 1668 | APO n = 104 | OR (95% CI) |
|---|---|---|---|
| 5-10 cm labor duration, min (SD) | 151 (96) | 155 (99) | |
| Duration of second stage, min (SD) | 30 (36) | 33 (35) | |
| Duration of active phase of second stage, min (SD) | 12 (9) | 14 (10) | |
| >20 min (n = 193) | 176 (10.8) | 17 (17.5) | 1.76 (1.02–3.03) |
| Person present at delivery | |||
| Senior obstetrician (n = 1617) | 1521 (92.5) | 109 (95.1) | 1 |
| Resident without senior obstetrician (n = 41) | 39 (2.4) | 2 (2) | 0.81 (0.19–3.42) |
| Midwife alone (n = 87) | 84 (5.1) | 3 (3) | 0.57 (0.18–1.82) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
