Why use more retrospective data and modelling to support universal third trimester scanning when prospective data suggests the implementation of specialist vaginal breech birth teams is equally likely to impact outcomes?

This a response to a recently published report in PLOS Medicine suggesting that implementation of universal third trimester ultrasound scanning in pregnancy improves outcomes for babies and mothers.

The following contributors have approved this expression of concern:

Researchers and Clinicians:

• Shawn Walker, Researcher in Residence and Honorary Consultant Midwife, Imperial College Healthcare NHS Trust
• Emma Spillane, Deputy Director of Midwifery and OptiBreech Lead, Kingston Hospital NHS Trust, London
• Sabrina Das, Consultant Obstetrician and OptiBreech Lead, Imperial College Healthcare NHS Trust
• Philippa Corson, Consultant Obstetrician and Breech Clinic Lead, Royal London Hospital, Barts Health NHS Trust
• Susan Bewley, Emeritus Professor of Obstetrics & Women’s Health, King’s College London

OptiBreech Patient and Public Involvement Leads:

• Siân Davies
• Nimisha Johnstone

Norfolk and Norwich University Hospital Maternity Voices Partnership Service User Representatives:

• Lisa Brophy
• Marion Frey-Alqurashi
• Rachel Graveling
• Siobhan Ridley
• Evelyn Shadlock

Knights et al¹ confidently demonstrate that routine third trimester, including point-of-care ultrasound (POCUS) performed by midwives, can significantly reduce undiagnosed breech presentation in labour. This is welcome, as women find diagnosis of breech presentation in labour traumatic, regardless of the outcome.² Although the considerable psychological impacts were not discussed, all should support the plan to increase safety, choice and personalised care through better antenatal detection of breech presentation.

However, the authors then assert that, “Short-term adverse perinatal outcomes, including [neonatal unit] admission and low Apgar scores, were significantly lower for the pregnancies with diagnosed breech presentation at term following a policy for screening by either routine third trimester scan or POCUS.”  This was despite no evidence given that any neonatal outcome achieved a statistically significant improvement. Indeed, hypoxic ischemic encephalopathy (HIE) increased from 0.3% to 0.4% in the St Georges University Hospital (SGH) cohort. The authors then ran Bayesian log-binomial regression models mostly using data from a previous evaluation of the same intervention on a different population (Salim et al),³ falsely concluding that there was a high probability the intervention would reduce adverse outcome rates.

Oxford’s implementation data

The publicly available data for the Salim et al study (S1 Data. Study data set)³ indicate that eight cases of serious neonatal morbidity (HIE and/or death) occurred:

Case number	Presentation at birth	Gestation at birth	Mode of birth	Scan after 35 weeks?	Seen in breech clinic?	Before or after universal USS	Preventable with universal scanning?
23475*	Breech	39+0	VB	Yes	Yes	Before	No
19027	Breech	37+2	VB	No	No	Before	Potentially
20535	Breech	38+1	EMC	Yes	Yes	Before	No
22414	Breech	40+0	VB	No	No	Before	Potentially
22548	Breech	38+3	EMC	Yes	Yes	Before	No
4343	Cephalic (after ECV)	42+2	VC	Yes	Yes (+ECV)	Before	No
16054*	Breech	37+0	VB	Yes	Yes	Before	No
19867	Cephalic (after ECV)	42+2	Ventouse	Yes	Yes (+ECV)	After	No

In six (75%) Oxford cases, the breech presentation was identified antenally. In 1/2 (50%) undiagnosed cases, a presentation scan would not have prevented the breech labour, which occurred at 37+2 weeks gestation. In both cases of death, the breech had been diagnosed clinically and the women had been seen in breech clinic. In two additional diagnosed cases, the breech service worked exactly as it was intended; two successful ECVs were performed. Nonetheless, HIE occurred following these cephalic births.

The total potential benefit in Oxford was a reduction of two cases of HIE and two less NICU admissions with Apgar <7 at 5 minutes without HIE, i.e. 44 versus 40 neonatal composite adverse outcomes in 1052 third trimester breech presentations, at a cost of 7,673 additional scans and 65 additional ECV procedures. 

Norwich charity funding

Knights et al¹ do not explain that Norfolk and Norwich University Hospital (NNUH) spent £100,000 of NNUH Hospitals Charity funding⁴ on handheld ultrasound scanners. The results in their cohort (Table 4) indicate two fewer neonatal admissions and one less case of Apgar score <7 at 5 minutes after implementation. Neither of these would have been considered a serious adverse outcome in the Term Breech Trial⁵ nor PREMODA study.⁶ Thus, there is no causative evidence of improvement, nor is this strategy is likely to prove cost-effective for implementation at scale.

St Georges’ specialist service

Knights et al failed to even look for confounding factors, let alone control for them – a serious source of bias in retrospective studies (see item 7 in the STROBE checklist, ‘Variables’).⁷ Yet, during their study time frame, and known to the authorship team, SGH also participated in a prospective multi-centre evaluation of physiological breech birth training, the results of which were already published in 2021.⁸ Both Knights and Mattiolo report a similar number of vaginal breech births, 64 (49 before and 15 after) and 90 (37 before and 53 after) respectively. Mattiolo et al also report outcomes for actual vaginal breech births. Among births where there was no attendant who had completed the enhanced training present, the severe neonatal composite adverse outcome rate was 5/69 (7.2%). Among births attended by someone who had completed the enhanced training, in the same settings, it was 0/21 (0%).

Between 2017-2020, overlapping with the implementation of the scanning programme, a specialist clinic and intrapartum care service for women requesting a vaginal breech birth were implemented at SGH.⁹  After 100 doctors and midwives had received a whole day’s training, the internal guideline was updated to include a physiological breech management algorithm.¹⁰ This was incorporated into monthly mandatory training to all staff, introducing substantial changes to vaginal breech birth practice. All obstetric trainees received half-day training. New joiners received the whole-day training repeated 6-monthly. By 2019, the rate of planned VBB had increased from 1.3% to 12.3% of all births in breech presentation.¹¹ Internal audit of this service demonstrated substantial reduction in the emergency caesarean birth rate, from 42.9% to 24.8% of all births in breech presentation.¹¹

The specialist service at SGH was discontinued when the breech specialist midwife (Spillane) relocated in 2020 and was not replaced. Nevertheless, the potential confounding effects need to be considered. When services invest in staff skill development, those effects extend beyond each individual birth.¹²

Oxford’s breech team

A specialist intrapartum service was also implemented at Oxford during the Salim et al. study,¹³ with the ‘dedicated on-call team’ for vaginal breech births publicly advertised on the hospital website.¹⁴ This is a significant difference in practice compared to most NHS units (except OptiBreech sites). Among the breech presentations >37 weeks with labour, planned vaginal breech births increased from 7.4% (12/162) to 17.6% (21/98) after the introduction of universal scanning. The provision of this team also appeared to improve the safety of the actual vaginal breech births that occurred.

	Admission to NNU	Apgar <7 at 5	HIE	Perinatal mortality
Planned VBB	2/16 (12.5%)	0/16 (0%)	0/16 (0%)	0/16 (0%)
Unplanned VBB	7/42 (16.7%)	3/42 (7.1%)	5/38 (13.2%)	2/42 (4.8%)

The differences between planned and unplanned VBB are comparable to Mattiolo et al.⁸ Could the authors use these data in Bayesian log-binomial regression models to demonstrate the beneficial effect of implementing breech birth teams? We cannot assume that similar results would occur in settings that do not offer a similar service. 

Summary

The opportunity to access a third trimester presentation scan remains important, especially for women planning an out-of-hospital birth. But women find it psychologically distressing and dehumanising to be unable to access skilled support for a vaginal breech birth, both antenatally and in labour.^15–17 Unless an evidence-based plan for improving this support is in place, matters will never change. When we prospectively evaluated the implementation of breech teams,¹⁵ one in five participants transferred from their original booking hospital to access supportive care for a vaginal breech birth.¹⁸ Some of these women came from SGH (since this VBB service is no longer available) and NNUH (approximately three hours car drive from the nearest OptiBreech site). Further unanticipated risks are introduced for women whose babies are diagnosed as breech but who cannot access their preferred mode of birth locally. More inequalities are created among women for whom the required travel and self-advocacy is impossible.

Determining which interventions improve clinical and cost-effectiveness outcomes for term breech pregnancies requires properly powered, prospectively registered, randomised controlled trials with publicly available, pre-specified protocols and anonymised data sets. It is extraordinary that £100,000 of charitable money was spent on equipment alone, outside the context of carefully planned research, and without service user involvement in priority setting. This has merely resulted in a poor-quality publication and plenty of mass media soundbites.^4,19,20 Currently, many NHS sites lack the funding for trained staff to offer all needed care options (ECV, VBB and ELCB) that are currently recommended in RCOG²¹ and NICE guidelines.^22,23 Investing in staff and their skill development will achieve the same, if not better, results and should be the priority.

Investing in staff and their skill development for vaginal breech births will achieve the same, if not better, results than universal third trimester ultrasound scans and should be a priority. @optiBreech @MidwifeSpillane https://t.co/AVmvWbCFsJ

— Dr Shawn Walker (@SisterShawnRM) May 3, 2023

References

1. Knights S, Prasad S, Kalafat E, et al. Impact of point-of-care ultrasound and routine third trimester ultrasound on undiagnosed breech presentation and perinatal outcomes: An observational multicentre cohort study. PLoS Med. 2023;20(4):e1004192. doi:10.1371/journal.pmed.1004192
2. Lightfoot K. Women’s Experiences of Undiagnosed Breech Birth and the Effects on Future Childbirth Decisions and Expectations. DHealthPsych. University of the West of England; 2018. http://eprints.uwe.ac.uk/33278
3. Salim I, Staines-Urias E, Mathewlynn S, Drukker L, Vatish M, Impey L. The impact of a routine late third trimester growth scan on the incidence, diagnosis, and management of breech presentation in Oxfordshire, UK: A cohort study. Myers JE, ed. PLoS Med. 2021;18(1):e1003503. doi:10.1371/journal.pmed.1003503
4. Norfolk and Norwich University Hospitals NHS Foundation Trust. Extra pregnancy scan significantly reduces the number of breech births, new research shows. Website. Published 2023. Accessed April 16, 2023. https://www.nnuh.nhs.uk/news/extra-pregnancy-scan-significantly-reduces-the-number-of-breech-births-new-research-shows/
5. Hannah ME, Hannah WJ, Hewson SA, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. The Lancet. 2000;356(9239):1375-1383. doi:10.1016/S0140-6736(00)02840-3
6. Goffinet F, Carayol M, Foidart JM, et al. Is planned vaginal delivery for breech presentation at term still an option? Results of an observational prospective survey in France and Belgium. Am J Obstet Gynecol. 2006;194(4):1002-1011. doi:10.1016/j.ajog.2005.10.817
7. Vandenbroucke JP, Von Elm E, Altman DG, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. PLoS Med. 2007;4(10):1628-1654. doi:10.1371/JOURNAL.PMED.0040297
8. Mattiolo S, Spillane E, Walker S. Physiological breech birth training: An evaluation of clinical practice changes after a one‐day training program. Birth. 2021;48(4):558-565. doi:10.1111/birt.12562
9. Spillane E, Walker S. Case study supporting continuity of care models for breech presentation at or near term. Pract Midwife. Published online 2019:36-37.
10. Spillane E, Winstanley C, Swer M. Breech. St George’s Hospital Practice Guideline; 2019.
11. Spillane E. St George’s Breech Clinic – Results. In: Physiological Breech Birth Training [Online]. Breech Birth Network; 2020. Accessed April 18, 2023. https://vimeo.com/486516151
12. Walker S, Parker P, Scamell M. Expertise in physiological breech birth: A mixed-methods study. Birth. 2018;45(2):202-209. doi:10.1111/birt.12326
13. Ord C. John Radcliffe midwife, Anita Hedditch, shortlisted for national award | Oxford Mail. Oxford Mail. https://www.oxfordmail.co.uk/news/17420964.john-radcliffe-midwife-anita-hedditch-shortlisted-national-award/. Published February 9, 2019. Accessed April 16, 2023.
14. Oxford University Hospitals. Specialist antenatal clinics – Maternity. Website. Published 2023. Accessed April 16, 2023. https://www.ouh.nhs.uk/maternity/antenatal/care/specialist/
15. Dasgupta T, Hunter S, Reid S, et al. Breech specialist midwives and clinics in the OptiBreech Trial feasibility study: An implementation process evaluation. Birth. 2022;00:1-10. doi:10.1111/birt.12685
16. Morris SE, Sundin D, Geraghty S. Women’s experiences of breech birth decision making: An integrated review. Eur J Midwifery. 2022;6(January):1-14. doi:10.18332/EJM/143875
17. Petrovska K, Watts NP, Catling C, Bisits A, Homer CS. ‘Stress, anger, fear and injustice’: An international qualitative survey of women’s experiences planning a vaginal breech birth. Midwifery. 2017;44(0):41-47. doi:10.1016/j.midw.2016.11.005
18. Walker S, Spillane E, Stringer K, et al. The feasibility of team care for women seeking to plan a vaginal breech birth (OptiBreech 1) – an observational implementation feasibility study in preparation for a pilot trial. BMC Pilot & Feasibility Studies. 2023;In Press.
19. Hall R. Third scan could greatly reduce UK breech birth numbers. The Guardian. Published April 7, 2023. Accessed May 1, 2023. https://www.theguardian.com/society/2023/apr/06/third-scan-reduce-uk-breech-birth-numbers-study-suggests
20. Pickles K. Third scan could cut breech births by 70%. The Daily Mail. Published April 7, 2023. Accessed May 1, 2023. https://www.mailplus.co.uk/edition/health/270217/third-scan-could-cut-breech-births-by-70?collection=16684
21. Impey L, Murphy D, Griffiths M, Penna L, on behalf of the Royal College of Obstetricians and Gynaecologists. Management of Breech Presentation. BJOG. 2017;124(7):e151-e177. doi:10.1111/1471-0528.14465
22. NICE. Intrapartum care for women with existing medical conditions or obstetric complications and their babies. Evidence review for breech presenting in labour. NICE guideline [NG121]. NICE. Published 2019. https://www.nice.org.uk/guidance/ng121/resources/intrapartum-care-for-women-with-existing-medical-conditions-or-obstetric-complications-and-their-babies-pdf-66141653845957
23. NICE. Antenatal Care.; 2022. Accessed November 29, 2022. https://www.nice.org.uk/guidance/ng201