Aims and objectives: The aim of this study was to evaluate the prevalence and management of undiagnosed breech, relating our figures with the national statistics, comparing our management with “standard” and to see if there could be any benefit from introducing additional intervention(s) to enhance breech detection. The author was also interested in finding out if there could have been any associated co-morbidity which could have contributed to failure to diagnose breech and whether relevant co-morbidities were born in mind or noted during clinical assessment of these women.
Methods: This is essentially a retrospective study of deliveries at the Horton Hospital Maternity Unit for a 12- month period, covering the period of October 2003 to September 2004. Delivery summaries of all deliveries were systematically studied twice for consistency. Case notes of breech deliveries were also reviewed twice. Cases of undiagnosed breech deliveries were clearly identified and studied in greater detail.
Results and conclusion: Cases of diagnosed breech were 64, which was 83.1% of the total breech births and 4.12% of the total annual birth. Cases of undiagnosed breech were 13, which was 16.88% of total annual breech deliveries and 0.84% of total annual birth. Interestingly, majority of diagnoses of undiagnosed breech were made during first stage of labour and most of the women had spent hours with staff and each had 2 or more clinical examinations before discovery was made. There was no diagnosis of undiagnosed breech in women with body mass index (BMI) of 40 kg/m2 or higher. As shown in Table 3, contrary to author’s expectation, most cases were diagnosed in women with normal BMI (61.54%).
Co-morbidities like uterine anomaly, polyhydramnios, large fibroids and pelvic mass were not apparent from the study. Most cases (61.54%) of undiagnosed breech were made in primigravidas and most cases of undiagnosed breech (61.54%) were delivered by emergency caesarean section. There was no adverse significant impact on maternal morbidity and short term neonatal outcome. The author’s view is that this study could rekindle the argument that trends of misdiagnoses should ideally prompt reviews rather that few spikes in misdiagnoses, in some aspects of obstetric care. However, one cannot deny the knowledge and reassurance from such reviews prompted by few spikes in misdiagnoses.

Keywords: pregnancy, breech, undiagnosed breech, horton hospital maternity

In September 2004, there were three cases of undiagnosed breech which occurred in succession over a few days. These were subsequently highlighted and reviewed at the interdepartmental Clinical Grand Rounds. This prompted a review of the situation which led to this study (audit).

This is a retrospective study of deliveries at the Horton Hospital Maternity Unit for a 12 month period, covering October 2003 to September 2004. Delivery summaries of all deliveries were systematically studied and reviewed, and case notes reviewed twice in all cases of breech deliveries.

Data from Information Technology Department was requested, reviewed and compared with data from the maternity to ensure consistency and that nothing was missed. There was detailed review of cases identified as undiagnosed breech in discharge summaries and cases. A case report form shown below was prepared to aid the study. More detailed demographic data on cases of undiagnosed breech was collected and analysed. Conversion (feet to centimetre) and calculation of some demographic characteristics like body mass index were undertaken. The author had assistance from Horton Hospital Postgraduate Centre on research methodology and data presentation. Data from the National Statistics and Planning Officer, Cherwell District Council, Oxfordshire, United Kingdom was collected and analysed. Extensive literature search was conducted for comparative analysis and for review of relevant literature. There was liaison with the Clinical Audit Department, John Radcliffe Hospital, Oxford, United Kingdom.

The usual early pregnancy scan and anomaly scan were performed but these are not relevant for breech presentation. The author noted that these scan may reveal known or unknown co-morbidities like large fibroids, pelvic mass, uterine anomaly, polyhydramnios, hydrocephalus, etc. The author also noted that breech presentation is not important before 36 weeks gestations unless the woman is in labour. Antenatal booking data and records may reveal co-morbidities like obesity, nervousness, uterine anomaly, spina bifida. Early third trimester scan was not routine but for cases where there were concerns in the first and second trimesters like twins, growth restriction, etc, and thus breech presentation may be part of other findings in these situations.

Abdominal palpation is usually performed and in cases where there were concerns for instance; not feeling the fetus in the pelvis on abdominal palpation, a presentation scan would be performed. The practice is dependent on skill training and experience of the attending staff and whether during the assessment, the attending staff recognized or realised the interplay of relevant co-morbidities which may be in existence. This is where the undiagnosed breech comes in. Observing current practice was performed retrospectively by collating information from case notes. It is possible there were things that were done but not recorded for which the author recognised as a possibility but would not want to speculate.

From medical literature, our practice appeared to reflected standard practice but the author noted that centres with more skilled staff and who have rigorous screening methods in place are likely to have lower incidence of undiagnosed breech. The author additionally noted external cephalic version (ECV) in early labour was stated to have a success rate of 3-40% in earlier literature.¹ However, currently, Collins S et al.² stated a success rate of 50%. ECV in early labour was not performed in our unit at the time of the study and no reason was given for this practice.

The total number of deliveries from October 2003 to September 2004 was 1554. This is consistent with the annual delivery rate at the Horton Hospital Maternity. The total number of women delivered with breech was 72 and a total of 77 babies (4.95%) were breech deliveries (cf 3-4%). Cased of diagnosed breech were 64 (83.1%) of the total breech birth and is 4.12% of the annual birth. Cases of undiagnosed breech were 13 (16.88%) of total annual breech deliveries and 0.84% of total annual birth. The total population of Banbury in 2001 National Census was 41.³ Males were 20,455 and Females were 21,347 (including female children). However, this hospital serves patients from other communities other than Banbury. The prevalence of undiagnosed breech in women of child bearing age (15-49) is therefore a bit difficult to calculate but estimated to be around 6/10000+/-SD.

Additionally in about 3 cases, the diagnoses were not consistent with other details of the management given. There was a case of caesarean section for undiagnosed breech recorded as spontaneous cephalic version in delivery summary, most likely the mistake of a tired staff. 5 cases (38.46%) of breech presentation were simply recorded as malpresentation. Few surgical notes were only partially completed and in some cases were left blank. In few patients, the height and weight were not recorded. Few records were done retrospective but still could not be followed.

There was a case of diagnosis of undiagnosed breech at second stage of labour, the woman having gone through the first stage of labour in the unit.

1. Diagnoses in advanced first stage of labour (cervical dilatation of 7-8cm) in birthing pool involving 2 women (15.38%) and 1 woman (7.69%) at second stage of labour in the birthing pool.
2. 3 women (23.08%) were admitted in second phase of first stage of labour, 2 women (15.35%) were at the stage of cervical dilatation of 6-7cm and 1 woman (7.69%) was at cervical dilatation of 4-5cm.
3. 2 women (15.38%) were moderately obese and this may or may not have contributed to the missed diagnosis.

Majority of diagnosis were made at first stage of labour and records show most of the women spent hours with staff and each had 2 or more clinical examinations before the discovery was made. No diagnosis was made in any woman with BMI of 40kg/m2 or higher. Co-morbities which may contribute to late diagnosis like uterine anomaly, polyhydramnios, large fibroids and pelvic mass were not apparent from the study and thus unlikely to have made an impact on the cases of missed diagnoses.

Articles published after this study has similar statistics on incidence of breech namely that the incidence of breech presentation decreases from about 20% at 28 weeks of gestation to 3-4% at term, as most babies turn spontaneously to the cephalic presentation.⁴ Breech presentations are mostly diagnosed during the antenatal period but sometimes they are diagnosed in early or established labour as this study found out. Many Obstetricians clearly disagree on whether a vaginal birth or a caesarean section is most appropriate way to manage a breech presentation. ‘The term breech trial’ by Hannah et al.⁵ showed that a surgical delivery by way of caesarean section was less hazardous for the baby compared to vaginal birth for cases of uncomplicated term breech presentation. However, some scholars claimed the study was flawed and called for its recommendations to be withdrawn,⁶ despite this study having substantially influenced obstetrics practice and policies according to Steen & Kingdon⁷ for over a decade.

Tables 1-8 below, illustrates the main findings and had been designed for ease of understanding and referencing. In Table 6, it can be seen that most cases of undiagnosed breech (61.54%) were delivered by emergency caesarean section and most cases of breech occurred in primigravidas. In Table 7, maternal morbidity measured by duration of hospital stay, shows that most cases (38.45%) spent 3 days in hospital. When this was compared with the fact that most cases were delivered by emergency caesarean section, then, it became clear that was no significant deviation from what was expected. Short term neonatal outcome in Table 8, measured by APGAR score at 5 minutes, showed that majority of neonates (61.54%) had very good score of 9 at 5 minutes.

The fact that majority of cases (53.85%) were diagnosed in the first phase of first stage of labour (Table 5), may have contributed to the less maternal morbidity (Table 7) and good neonatal outcome, (Table 8). Table 9 below is the pro forma or case report form used to study each case of undiagnosed breech. However, breech detected at 35-37 weeks of gestation allows time for a planned caesarean section or ECV. Undiagnosed breech in labour may therefore suggest unskilled antenatal care; lack of antenatal care especially in cases of concealed pregnancy or other forms of non-attendance for antenatal care. Undiagnosed breech can put unexpected strain on staff and resources, the extent of this strain vary with the skill and abilities of staff, and the available resources in the unit. 53.85% of women were in early labour at the time of diagnoses but none was offered ECV. It is possible that Obstetricians were not aware it was an option or simply did not want to offer it at first stage of labour for no stated reason.

In cases where the affected unit cannot cope with the pressures, the woman may be transferred to another unit with possible consequent perinatal risk of fetal and maternal morbidity or mortality including poor apgar scores for the newborn and longer hospital stay for the mother. The affected mother also loses the opportunity for and benefits of a planned ECV.

A woman who had been informed that her baby was cephalic presentation and had gone into labour and later told her baby was undiagnosed breech and that she needed delivery by emergency caesarean section, would likely be very upset and distressed. Early diagnosis of breech at or after 36 weeks gestation, and opting for an elective caesarean section leads to less maternal and fetal morbidity compared to undiagnosed breech in labour and subsequent emergency caesarean section.

In a more recent literature, Collins S et al.,² stated that extended breech was 70%, flexed breech, 15% and footling breech (one leg flexed and one extended), 15%, and that breech presentation was commonly undiagnosed before labour in 30%. However, in 2004, we recorded incidence of 16.88% for undiagnosed breech. At that time, there was “clinical furore”, a lot of clinical frenzy and anxieties about the 3 consecutive diagnoses of undiagnosed breech within a week or so. We thought that the incidence of undiagnosed breech was high in our unit and there must be something we could change about our antenatal care to reduce the incidence.

However, other studies have incidence of undiagnosed breech as follows; 21% for Leung et al.⁸ 26% for Nwosu et al.,⁹ 12.9% for Babay et al.¹⁰ and 5.4% for Usta et al.¹¹ Another thought-provoking article showed that significant cases of undiagnosed breech were still possible despite rigorous antenatal surveillance, Nwosu et al.⁹ There was no significance difference in maternal and neonatal outcomes and in the incidence of caesarean section.^9-11

Other studies, although stating there was no significant difference in maternal outcomes between caesarean section and vaginal delivery, had argued that vaginal delivery is associated with excess neonatal morbidity and mortality.¹² Some risk factors have been noted as purely independent e.g. footling breech, hypertension, or failure to attend antenatal care. This study found out that the stage of diagnosis of undiagnosed breech in labour is influenced by the stage of admission in labour. Undiagnosed breech appears to suggest we need to improve our skills and ability to diagnosed breech during the antenatal period or in early labour. Vaginal delivery was more common in cases of undiagnosed breech, 46%, compared with the diagnosed group of breech, 26%, in Leung et al.⁸ The same view was shared by Nwosu et al.⁹ Undiagnosed breech is more likely to deliver vaginally in 42% of cases compared with the diagnosed breech in 11%, according to Leung et al.⁸ The same view was again shared by Nwosu et al..⁹

The author noted that factors such as large uterine fibroids, uterine anomaly, polyhydramnios, and past history of breech were not observed in the cases reviewed during the study. These factors in the opinion of the author could make diagnosis of undiagnosed breech more difficult. This study though limited by the number of women involved, provided insight to one of the puzzles in obstetric care. For instance, it is commonly for an obstetric team raises alarm thinking it may not doing well regarding diagnoses and management of breech presentations, only to discover following audit or review that its practice was not far from acceptable standards. The author’s view is that this study could rekindle the argument that trends or patterns of misdiagnoses should ideally prompt a review or audit rather that few spikes in misdiagnoses. However, one cannot deny the knowledge and reassurance from such reviews or audits prompted by such spikes in misdiagnoses.

1. Mr. Hikmet Naoum, Consultant Obstetrician and Gynaecologist, Oxford Radcliffe Hospital NHS
2. Clinical Audit Department, John Radcliffe Hospitals, Oxford Radcliffe Hospitals NHS Foundation Trust.
3. Postgraduate Library Department staff.
4. Cherwell District Council.
5. Information Technology, Data Analysis Department, John Radcliffe, Oxford Radcliffe Hospitals NHS Foundation Trust.

1. Collier L, Longmore M, Brown TD. Oxford handbook of clinical specialities. 5th edn. Oxford, USA: Oxford University Press; 1999. 124 p.
2. Collins S, Arulkumaran A, Hayes K, et al. Oxford handbook of obstetrics and gynaecology. 3rd edn. Oxford, USA: Oxford University Press; 2013:82–87.
3. National Statistics. People/neighbourhood statistics for Cherwell district (UK) census 2001. 2001.
4. RCOG Green-top Guideline: The management of breech presentation; (20) Apr 2001/20b p1 2006, RCOG, London; 2006.
5. Hannah ME, Hannah WJ, Hewson SA, et al. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet. 2000;356(9239):1375–1383.
6. Glezerman M. Five years to the term breech trial: the rise and fall of a randomised controlled trial. Am J Obstet Gynecol. 2006;194(1):20–25.
7. Steen M, Kingdon C. Vaginal or caesarean delivery? How research has turned breech birth around. Evidence Based Midwifery. 2008;6(3):95–99.
8. Leung CW, Pung TC, Wong WM. Undiagnosed breech revisited. Br J Obstet Gynaecol. 1999;106(7):638–641.
9. Nwosu EC, Walkinshaw S, Chia P, et al. Undiagnosed breech. Br J Obstet Gynaecol. 1993;100(6):531–535.
10. Babay ZA, Al-Nuaim LA, Addar MH, et al. Undiagnosed term breech: management and outcome. Saudi Med J. 2000;21(5):478–481.
11. Usta IM, Nassar AH, Khabbaz AY, et al. Undiagnosed term breech: Impact on mode of delivery and neonatal outcome. Acta Obstetricia et Gynecologica Scandinavica. 2003;82(9):841–844.
12. Ben Aissia N, Youssef A, Said MC, et al. Vaginal delivery or planned caesarean section?. Tunis Med. 2004;82(5):425–430.