When Research is Flawed:
Planned Vaginal Birth versus Elective Cesarean
for Breech Presentation
by Henci Goer
Commentary on: Hannah, M. E., Hannah, W. J., Hewson, S. A., Hodnett, E. D., Saigal, S., & Willan, A. R. (2000). Planned caesarean section versus planned vaginal birth for breech presentation at term: A randomised multicentre trial. Term breech trial collaborative group. Lancet, 356(9239), 1375-1383. [Abstract]
Study design and results: multicenter randomized controlled trial assigning 2083 women with a frank or complete breech presentation at full-term to either planned cesarean or planned vaginal birth. The study concludes that planned cesarean results in superior outcomes for the baby and equivalent outcomes for the mother.
- combined stillbirth and neonatal mortality rate excluding lethal congenital abnormalities: 3/1039 (0.3%) planned c/sec versus 13/1039 (1.25%) planned vaginal; relative risk 0.23, 95% confidence interval 0.07 – 0.81
- combined perinatal mortality and serious neonatal morbidity: 17/1039 (1.6%) planned c/sec vs. 52/1039 (5.0%) planned vaginal; relative risk 0.33, 95% confidence interval 0.19 – 0.56
- no differences in maternal mortality or morbidity between groups
- cesarean rate in the planned vaginal group: 43%
Problems include but are not limited to the following:
The trial was intended to “give the option of vaginal breech delivery its best, and perhaps last, chance to be proven a reasonable method of delivery.” Inclusion criteria relevant to that goal were:
- woman a “good candidate” for planned vaginal breech birth, that is, she had a live singleton fetus in frank breech (hips flexed, knees extended) or complete breech (hips flexed, knees flexed, feet not below fetal buttocks) presentation, no known lethal fetal anomaly, no hyperextended neck (head tilted back): In violation of these criteria, the trial included an 1150 g twin who died before enrolment, an infant with a large meningomyelocele (protrusion of the spinal membranes and cord through a defect in the spinal column), and a fetal demise “probably before enrolment” in a head-down baby. To these, a subsequent analysis adds at least two more cases of neonatal morbidity after vaginal birth: a footling breech and the surviving twin in the case above (Su 2004). We don’t know how many more unsuitable candidates there may have been because we only have information on those cases where perinatal morbidity or death occurred.
- birth attended by a clinician experienced with vaginal breech birth: This same analysis reveals that of 51 cases where morbidity or mortality was attributed to events in labor or at delivery, 7 (13.7%) were attended by clinicians with “no experience” (Su 2004). An additional 1 of 12 cases where the outcome was judged “unexplained” was attended by an inexperienced clinician as were 2 of 6 cases where outcomes were deemed unrelated to labor or delivery.
In addition to these problems, if the intent was to determine the safety of planned vaginal breech birth, investigators should have:
- eliminated cases where adverse outcomes were unrelated to birth route: Only 4 of the 13 perinatal deaths in the vaginal breech group could be directly related to difficult vaginal birth, and the baby was abnormally formed in one of these. In a follow-up study, two additional cases of serious morbidity in the planned vaginal birth group were determined to be unrelated to delivery (Su 2004). A critique of the Hannah (2000) trial reports that after eliminating mortality and morbidity arising from cases that should not have been enrolled or that should have been excluded from analysis, 16 cases remain of the trial’s 69: 11 in the planned vaginal birth group and 5 in the planned c/sec group (Glezerman 2006). When the corrected numbers are compared, the difference between groups is no longer significant.
- excluded centers that lacked adequate diagnostic resources: In one-third of cases (31.3%), imaging studies were not used to assess the attitude of the baby’s head in women assigned to planned vaginal birth. Hyperextended neck contraindicates planned vaginal birth. In addition, 4% of women allocated to planned vaginal birth and 2% of women allocated to planned c/sec had head-down babies at delivery. Also, as noted, clinicians failed to diagnose twin pregnancy, antepartum fetal demise, and footling breech.
- confined outcome measures to those that could be measured objectively and that were both serious and non-transient: “Serious” morbidity included hypotonia (poor muscle tone) lasting longer than 2 hours, “hyperalert, drowsy, or lethargic,” and 5-minute Apgar score < 4.
- blinded pediatricians to infant’s birth route to eliminate bias: As can be seen by the examples above, some of the morbidity categories involved judgment calls.
- analyzed outcomes according to institution and individual provider and excluded those with excessive rates of adverse outcomes: Perinatal outcomes of vaginal breech birth depend on clinician skill and institutional resources. In point of fact, clinicians with no experience attended 11.5% (6/52) of the planned vaginal births in which there was neonatal morbidity or stillbirth. The trial, however, reports that only 2.7% of planned vaginal births overall were attended by inexperienced practitioners.
This trial also had problems intrinsic to conventional obstetric management or to randomized controlled trials:
- neonatal morbidity and mortality rates depend on management: Conventional obstetric management might have produced some of the adverse effects seen in the trial. For example, umbilical cord prolapse occurred in 1.2% of labors. Artificially rupturing membranes can precipitate this complication.
- maternal morbidity rates also depend on provider management because they depend on rates of cesarean section, vaginal instrumental delivery, and episiotomy.
- neonatal and maternal morbidity are not equivalent: Once the baby recovers from most morbid events (ex. a broken bone), the threat is ended. A uterine scar has, and a torn anal sphincter often has, permanent consequences.
- trial only presented short-term neonatal outcomes: A follow-up trial at 2 years reported almost identical risks of death or neurodevelopmental delay (3.1% planned c/sec vs. 2.8% planned vaginal) between the two groups (Hannah 2004).
- trial only presented short-term maternal outcomes: Investigators failed to consider the potential long-term and future reproductive harms of cesarean surgery.
- investigator bias can affect trial methodology, reporting of results, and conclusions: Mary Hannah, the lead investigator, believes cesareans to be a superior method of delivery as demonstrated by her chairing a conference entitled “Choosing Delivery by Caesarean: Has Its Time Come (MIRU 2002)?”
Comment: Other breech studies contradict the findings of the Hannah study. For example, a systematic review reports that most studies conclude that “vaginal delivery is safe, given a good selection of patients, assistance by qualified staff and careful management during the delivery (Haheim 2004).” Reviewers reach the same conclusion in their analysis of Norwegian data from 1981 to 1998 for singleton breech births with birth weight > 2499 g. In marked contrast to the Hannah trial, the perinatal mortality rate was 0.31% overall after correction for lethal malformations and 0.09% after additionally correcting for death before admission. Data from four Norwegian hospitals verify results of the national data analysis.
Nonetheless, as a result of the Hannah trial, vaginal breech birth, which had been declining in the U.S. and other Western European countries, has virtually disappeared. Despite a stream of articles discrediting the Hannah trial, influential institutions such as the Cochrane collaboration and the American College of Obstetricians and Gynecologists have not rescinded their endorsement of planned cesarean surgery for breech. “This should not have come as a surprise,” writes Marek Glezermen (2006), “. . . The conclusions of the [term breech trial] are perceived today by many obstetricians as a badly needed set of arguments for [planned cesarean delivery], which they would have preferred anyway.” Glezerman continues, “The consequences of this situation are many superfluous cesarean deliveries with consequential morbidity in women and the vanishing of obstetric expertise with increased risks to those breech babies who must be delivered vaginally,” and, one might add, a policy that forces women into surgeries they neither want nor need because they have no other viable option.
Glezerman, M. (2006). Five years to the term breech trial: The rise and fall of a randomized controlled trial. Am J Obstet Gynecol, 194(1), 20-25.
Haheim, L. L., Albrechtsen, S., Berge, L. N., Bordahl, P. E., Egeland, T., Henriksen, T., et al. (2004). Breech birth at term: Vaginal delivery or elective cesarean section? A systematic review of the literature by a norwegian review team. Acta Obstet Gynecol Scand, 83(2), 126-130.
Hannah, M. E., Whyte, H., Hannah, W. J., Hewson, S., Amankwah, K., Cheng, M., et al. (2004). Maternal outcomes at 2 years after planned cesarean section versus planned vaginal birth for breech presentation at term: The international randomized term breech trial. Am J Obstet Gynecol, 191(3), 917-927.
Kierse MJNC. (2002) Evidence-based childbirth only for breech babies? Birth, 29(1), 55-9.
Kotaska, A. (2004). Inappropriate use of randomised trials to evaluate complex phenomena: Case study of vaginal breech delivery. BMJ, 329(7473), 1039-1042.
Maternal, Infant and Reproductive Health Research Unit (MIRU). Choosing Delivery by Caesarean: Has Its Time Come? Conference sponsored by MIRU, University of Toronto, Toronto, Ontario, Canada, Nov 7, 2002.
Su, M., Hannah, W. J., Willan, A., Ross, S., & Hannah, M. E. (2004). Planned caesarean section decreases the risk of adverse perinatal outcome due to both labour and delivery complications in the term breech trial. BJOG, 111(10), 1065-1074.
Van Roosmalen, J., & Rosendaal, F. (2002). There is still room for disagreement about vaginal delivery of breech infants at term. BJOG, 109, 967-969.
Whyte, H., Hannah, M. E., Saigal, S., Hannah, W. J., Hewson, S., Amankwah, K., et al. (2004). Outcomes of children at 2 years after planned cesarean birth versus planned vaginal birth for breech presentation at term: The international randomized term breech trial. Am J Obstet Gynecol, 191(3), 864-871.