Uva-DARE (Digital Academic Repository)

UvA-DARE (Digital Academic Repository)

Improving management of breech presentation at term

Rosman, A.N.

Publication date 2014 Document Version Final published version

Link to publication

Citation for published version (APA): Rosman, A. N. (2014). Improving management of breech presentation at term.

General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Download date:06 Oct 2021 Improving management of breech presentation at term

A.N. Rosman F. Vlemmix Stellingen bij het proefschrift

Improving management of breech presentation at term

1. Suboptimale implementatie is grotendeels te wijten aan zorgverleners die de uitwendige versie niet aanbieden.

(dit proefschrift)

2. De uitwendige versie is een belangrijke behandeling om het aantal geplande keizersneden terug te dringen.

(dit proefschrift)

3. Tijdens de counseling met betrekking tot de partus moeten de nadelige gevolgen van een geplande keizersnede voor stuitligging voor een volgende zwangerschap worden benoemd, maar deze risico’s wegen niet op tegen de risico’s van een vaginale stuitbevalling.

(dit proefschrift)

4. Een landelijke registratie van alle (aangeboden) versiepogingen is een voorwaarde om tot verbetering van zorg te komen.

(dit proefschrift)

5. Implementatie dient een core business te zijn van elke zorgverlener. Alleen dit leidt tot evidence based practice.

6. De uitkomst van de som van onderzoek door een verloskundige en een arts-assistent in opleiding tot gynaecoloog is een voorbeeld van integrale geboortezorg en kan flink wat opleveren.

(dit proefschrift)

7. De uitwendige versie is zinnige en zuinige zorg.

8. Het is gecontra-indiceerd om te spreken over relatieve en absolute contra-indicaties

9. Lijnen zijn als spoorrails: ze leiden samen ergens naar toe.

357 10. Verloskundige zorg begint voor de conceptie.

11. Aversie van versie van stuiten is stuitend.

(Martijn Ouddijk, 2013)

12. Tweelingen zijn geboren samenwerkers.

13. Sven Kramer is de equivalent van Ben Willem Mol; deze stelling kan uitwendig gekeerd worden.

14. Stuitligging; duidelijker kan de diagnose en eenduidiger kan de behandeling van een medische kwestie niet zijn.

15. Promoveren is balanceren op de rand van onduidelijkheden.

16. Vele handen maken licht werk, vele auteurs maken een artikel zwaar.

17. Verandering roept weerstand op.

(Drs. E.I. Schippers, minister van VWS, april 2014, symposium Gepaste zorg)

18. Probeer niet in de voetstappen van de meesters te treden, maar zoek wat zij zochten.

19. Om het vertrouwen in je promotieonderzoek over implementatie te behouden, moet je beginnen met het negeren van de spelling- en grammaticacontrole van ‘Word‘ die het woord ‘implementatie’ als vaag taalgebruik aanmerkt.

20. Iemand met een koel hoofd, heeft een warm hart nodig.

358 Improving management of breech presentation at term

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam op gezag van de Rector Magnificus prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Agnietenkapel op woensdag 3 september 2014, te 16:00 uur

door Ageeth Nicole Rosman geboren te Groningen

1 Promotiecommissie Promotores: Prof. dr. J.A.M. van der Post Prof. dr. B.W.J. Mol

Co-promotores: Dr. M. Kok Dr. M.A.H. Fleuren

Overige leden: Prof. dr. C.J.M. de Groot Prof. dr. M.L. Essink – Bot Prof. dr. D. Bick Dr. E. de Miranda Dr. M. Zegers

Faculteit der Geneeskunde

2 Dit proefschrift is geschreven in een samenwerking en als duo promotie met F. Vlemmix en heeft derhalve een gelijke inhoud en eenzelfde ISBN nummer. Tijdens de openbare verdediging zal A.N. Rosman de volgende hoofdstukken verdedigen: 1, 4, 6, 7, 9, 10, 12, 13, 14. This thesis is made in collaboration with drs. F. Vlemmix, who will defend her thesis on September 4th at the University of Amsterdam. On behalf of the collaboration this thesis has the same ISBN as the thesis of F. Vlemmix. The content of this thesis is similar to the thesis of F. Vlemmix. A.N. Rosman will defend chapters 1, 4, 6, 7, 9, 10, 12, 13, 14.

Het verschijnen van dit proefschrift werd mede mogelijk gemaakt door de steun van de Koninklijke Nederlandse Organisatie van Verloskundigen en ZonMw (projectnummer 80-82315-98-09011).

Financial support by the Royal Dutch Organisation of Midwives for the publication of this thesis is gratefully acknowledged and ZonMw (The Netherlands Organisation for Health Research and Development - (project number: 80- 82315-98-09011)).

Ontwerp omslag en lay-out: Boudewijn Rosman Printed by: Grafisch Bedrijf de Bruin, Zuidbroek ISBN/EAN: 9789462285040

3 4

CONTENTS

CHAPTER 1 General introduction

PART 1: BREECH DELIVERY CHAPTER 2 Term breech deliveries in the Netherlands: did the increased cesarean rate affect neonatal outcome? A population based cohort-study Accepted, ACTA Obstet Gynecol Scand. CHAPTER 3 Subsequent pregnancy outcome after breech delivery, a population based cohort study

PART 2: DEVELOPMENT OF STRATEGIES TO IMPROVE IMPLEMENTATION OF ECV CHAPTER 4 Contraindications for external cephalic version in breech position at term: a systematic review Acta Obstet Gynecol Scand. 2013 Feb;92(2):137-42 CHAPTER 5 Patient’s willingness to opt for external cephalic version J PSYCHOSOM OBSTET GYNAECOL. 2013 MAR;34(1):15-21. CHAPTER 6 Patients’ and professionals’ barriers and facilitators to exter nal cephalic version for breech presentation at term, a quali tative analysis in the Netherlands. Midwifery. 2014 Mar; 30(3):324-30 CHAPTER 7 Facilitators and barriers to external cephalic version for breech presentation at term among health care providers in the Netherlands: A quantitative analysis Midwifery. 2014 Mar; 30(3): e145-50 CHAPTER 8 Decision aids to improve informed decision-making in pregnancy care: a systematic review BJOG. 2013 FEB;120(3):257-66

PART 3: EVALUATING STRATEGIES TO IMPROVE IMPLEMENTATION OF ECV CHAPTER 9 Implementation of external cephalic version in the Nether lands: a retrospective cohort study CHAPTER 10 Implementation of client versus care provider strategies to improve external cephalic version rate: a cluster randomised controlled trial CHAPTER 11 Cost-effectiveness of improving implementation of external cephalic version for term breech presentation. CHAPTER 12 Mode of delivery and neonatal outcome after external cep halic version: a prospective cohort study CHAPTER 13 Knowledge and decisional conflict scores related to external cephalic version in the Netherlands CHAPTER 14 Summary and general discussion Nederlands samenvatting en algemene discussie

Dankwoord A.N. Rosman PhD portfolio A.N. Rosman List of publications A.N. Rosman Curriculum vitae A.N. Rosman

Dankwoord F. Vlemmix PhD portfolio F. Vlemmix List of publications F. Vlemmix Curriculum vitae F. Vlemmix Authors’ affiliations

ADDENDUM Implementation of the external cephalic version in breech delivery. Dutch national implementation study of external cephalic version BMC Pregnancy Childbirth. 2010 May 10;10:20.

1 General introduction

13 Chapter 1 General introduction

General introduction

Three to four percent of term pregnancies are complicated by a breech presentation. Compared to a delivery of a child in the normal cephalic presentation, vaginal breech delivery is associated with increased risk of neonatal birth trauma, low Apgar scores and neonatal death.1,2 The advantage of less short-term neonatal morbidity and mortality as a result of an elective caesarean delivery for term breech presentation became clear in the publication of the Term Breech Trial (TBT) in 2000.1

Within three months after publication of this randomised controlled trial (RCT), the elective caesarean delivery rate doubled in the Netherlands up to 60% and remained more or less stable afterwards.3,4 Elective caesarean delivery rates in other countries increased even further, up to 90% in the United Kingdom and 96% in Australia.5,6 In 2012, 4,822 (3.1%) of term pregnancies in the Netherlands were in breech presentation, of which 3,258 (68%) were delivered by elective caesarean.3,4 There were a total of 9,811 elective caesarean deliveries for term singletons in the same year. Thus, 33% of elective caesarean deliveries for term singletons were due to breech presentation.

Although a planned caesarean delivery in case of term breech presentation leads to less short-term neonatal morbidity and mortality as compared to a vaginal breech delivery, there are some relevant clinical points to consider. First, delivery by caesarean section leads to increased maternal and perinatal morbidity in future pregnancies. A systematic review of Guise et al. 7 concluded that the rate of uterine rupture for all women with a prior caesarean delivery is 30 per 10,000. This risk was significantly increased with a trial of labour (47 per 10,000 versus 3 per 10,000 elective repeat caesarean delivery). Six percent of uterine ruptures were associated with perinatal death. Perinatal mortality was significantly increased for a trial of labour at 13 per 10,000 deliveries versus 5 per 10,000 for elective repeated caesarean delivery. Furthermore, in the Netherlands four maternal deaths occurred between 2000 and 2002 after elective caesarean delivery for breech presentation.8

External cephalic version (ECV) is an obstetrical intervention that has been proven to reduce the number of breech presentations. It can increase the rate of cephalic presentation at birth by 40% and consequently reduce the caesarean delivery rate.9 ECV is also a relative safe intervention. A meta-

14 Chapter 1 General introduction analysis reporting on 12,955 ECV attempts showed a pooled complication rate of 6.1% (95%CI 4.7 to 7.8%), 0.24% for serious complications (95% CI 0.17 to 0.34%) and 0.35% for emergency caesarean deliveries (95% CI 0.26 to 0.47%).10

The history of external cephalic version: from Hippocrates towards evidence based guidelines

Breech presentation and manoeuvres to turn a foetus in the womb have been known since ancient times. Hippocrates was one of the first men -re porting on breech presentation. The scripts of Hippocrates and later on, the book of Rösslin showed awareness about perinatal risks for mother and foetus, but they could not react on it as they had limited knowledge about the pathophysiology of breech presentation and the process of a vaginal breech delivery. Hippocrates described in his scripts (460-377 BC) a breech delivery as extremely difficult. Hippocrates thought that normal birth could only take place if the foetus was able to push its way out by putting his feet to the fundus uteri, of which a foetus in breech position was incapable to do so. If birth in breech presentation did not progress, they tried to turn the foetus around as they thought that the foetus needed his feet to push to the fundus uteri. In those days it was attempted by shaking the woman ten times during labour and putting her bed into the Trendelenburg position. If these movements did not result in a vaginal birth, Hippocrates advised to clasp the foetus and to turn the head through the birth canal, making a vaginal birth possible. If this treatment did also not succeed or if the birth attendant was convinced that a vaginal birth was not possible, Hippocrates advised an embryotomy (mutilation of a foetus to facilitate removal from the womb when natural delivery is impossible). Till medieval times, the scripts of Hip- pocrates were the most important documents of obstetric care. Thereafter, knowledge of Hippocrates was adopted by the Arabs and transferred to Western Europe in Islamic writings.

Eucharius Rösslin, a German physician, wrote in 1513 a book about childbirth (‘Der Rosengarten’) what became a manual for midwives.11 Rösslin described the possibility to turn a foetus by internal handholds to a cephalic presentation during labour and in case of no success to move the arms towards the legs and to extract the foetus. He described that breech presentation could be associated with prolonged labour and a prolonged or failed second stage of labour.

15 Chapter 1 General introduction

During the Renaissance the knowledge of the anatomy of the human body increased (Vesalius 1514-1654 and Fallopius 1523-1562). Vesalius was a Dutch pathologist- anatomist who wrote the “Humani Corporis Fabrica Libri septem” in which he described the human anatomy. Gabriele Falloppio is nowadays well known from the Fallopian tubes, the two very fine tubes between the ovaries and the uterus. He was born in Italy and was one of the most important anatomists of the sixteenth century. He studied the repro- ductive organs of the human bodies, which provided more insight in the process of vaginal births. Despite of this knowledge, breech delivery remained full of perinatal risks for mother and child and turning a foetus in breech position to cephalic position, was not often applied.

In the 18th century physicians became more involved in the actual process of childbirth leading to a change in obstetric care. Bringing in their knowledge of the human anatomy modernised obstetric care. It took till the 19th century that more attention was paid to the safety of vaginal breech birth and the possibility of external cephalic version (ECV).

Professor Adolphe Pinard (1844-1934), a French obstetrician, is seen as the founder of abdominal palpation. In 1878 he advised to examine every woman before birth to detect abnormal foetal presentations.11 He also recommended to turn a foetus, presenting in transverse presentation to cephalic presentation which was not done or opted before.

In the same time obstetricians (Wigand, 1820; De Snoo 1910) started to publish about manoeuvres for an assisted vaginal breech delivery. Wigand, a German obstetrician described in his book “Die Geburt des Menschen” (1820) a manoeuvre for an assisted breech delivery, known as the Wigand manoeuvre, which made the development of the foetal head safer. De Snoo also described a manoeuvre to develop the foetal head after breech presentation to prevent the foetus for adverse perinatal outcomes.

The first publications on ECV appeared in 1928 by McGuinness.12 McGuin- ness was a lecturer in obstetrics at the University of Manitoba (Winnipeg). In 1927 he gave a lecture for the Winnipeg Medical Society in which he presented the results of breech births and external cephalic version attempts in his clinic of the University of Manitoba. Based on the adverse outcomes of breech births in his clinic, he decided to apply ECV to all his patients with breech presentations. Of the 62 presentations, 57 had a successful ECV -at

16 Chapter 1 General introduction tempt. Three times a contraindication for ECV (twice a multiple pregnancy and once premature labour) was present and two ECV attempts failed. In one case an emergency caesarean delivery was necessary and in one case fetal death occurred after a partial placental abruption during labour. It is not clear if labour started short after ECV. In the same paper he wrote about the technique of ECV and the timing of an ECV attempt. He concluded that while waiting for a spontaneous version the most favourable time for ECV could be lost.

Later on, around the years 1950 and beyond, more publications on ECV and the mode of delivery of breech presentations appeared. It was already in that time that obstetricians discussed about unfavourable conditions for ECV such as oligohydramnion, multiple pregnancies, parity and ruptured membranes.13,14 At the same time ECV was subject of discussion in the Netherlands. In a publication of the Dutch Journal of Medicine in 1940, the safety of ECV was discussed by Van der Hoeven, Treub and Kouwer.13 They had opposing opinions regarding ECV; they feared wrong diagnosis, severe complications like placental abruption and they disagreed on the chances of success. ECV fell into oblivion and it took many years before it became routine practice again. In 1980, two Dutch obstetricians (Mensink and Huisjes) stated that ECV was not useful at 33 weeks gestations and discussed whether it was after 33 weeks gestation.15 In 1987, following international publications that showed that ECV could be useful in modern obstetric care obstetricians, the Dutch obstetricians van Raaij and Van der Lugt reopened the discussion concerning the application of ECV.16 They reviewed international publications of 1980 onwards and formed, based on this review, a list of absolute contraindications for ECV: multiple pregnancies, placenta praevia or low insertion of the placenta, scarred uterus, fetal anomalies, intra uterine growth retardation, necessity for a planned caesarean, ruptured membranes, toxicosis, hypertension, foetal distress, and antepartum bleeding in history. Obesity, oligohydramnion and anterior placenta were mentioned as relative contraindications. They also described the prerequisites for a safe ECV attempt. They concluded that ECV was an important intervention in modern obstetric care and should be routine practice.16

17 Chapter 1 General introduction

Current practice: the implementation of ECV

After publication of the results of the Term Breech Trial in 2000, the professional organizations of obstetric care in the Netherlands (NVOG and KNOV) published guidelines in which they recommended to advise and to offer ECV to all women with a in foetus in breech presentation from 36 weeks onwards.17–19 Guidelines have in common that they recommend a treatment based on evidence but also mention contraindications for the treatment. Guidelines can differ in the definition of contraindications. Contraindications for ECV mentioned by both guidelines are: oligohydramnion and a history of placental abruption. The NVOG also mentions ruptured membranes and uterine anomaly as contraindications. The KNOV adds multiple pregnancies, severe intra uterine growth retardation, severe hypertension, placenta praevia, third trimester bleeding (unknown cause), scarred uterus as contraindications to it. Both guidelines advise a similar routine: to use ultrasound to ascertain foetal position before and after an ECV attempt, to monitor foetal condition before and after the attempt, and to administer anti-rhesus im- munoglobulins in case of a rhesus negative blood of the mother and rhesus positive blood of the foetus. The NVOG advises to perform ECV within a hospital setting, the KNOV describes the possibility to perform ECV in midwives’ practices (out of hospital setting). The KNOV guideline also addresses to the training of midwives on ECV. Only ECV certified midwives who are registered are allowed to perform ECV. Those midwives followed a training on ECV and have to meet several quality requirements yearly in order to keep their competence. All ECV attempts by midwifes are registered by the KNOV in a national database. Data are not published yet. Advice on required professional skills and quality control is lacking in the guideline of the NVOG.

Although the guidelines are quite clear in their advice to counsel for ECV and arrange possibilities for ECV, the ECV uptake by both health care professionals and patients shows wide variation. An inventory survey among all hospitals in the Netherlands in 2007 reported that 5% of the gynaecologic practices did not perform ECV at all.20 In the United Kingdom and Australia the number of eligible women who were not offered an ECV attempt ranged from 4% to 33%.21–23and external cephalic version (ECV An Australian preference study showed that 39% of women did not opt for ECV and a similar study in Israel reported a decrease in women willing to undergo ECV from 54% to 24%.24,25 Based on these figures, we estimated the implementation

18 Chapter 1 General introduction rate of ECV (defined as the amount of patients having an ECV attempt of all breech presentations after 36 weeks gestation) in the Netherlands overall to be around 50 to 60, which we consider too low and important to increase. 26 As there is sound evidence supporting an ECV in case of a term breech presentation, it is clear that there is room for improvement of the implementation of ECV.

Improvement of implementation of the guidelines

Despite of the fact that both professional organizations responsible for obstetrical care updated their guidelines in 2008 and brought the update to the attention of their members, the number of breech presentations at birth remained stable over the years.

Implementation of clinical practice guidelines is a common problem in the health care system; new or revised guidelines are not automatically used by professionals. Several models and frameworks have been developed to improve the uptake of guidelines by paying attention to a systematic introduction within the existing health care system.18,27–32 They all emphasise the importance of a detailed understanding of critical factors, or so-called determinants, that may enhance or impede the use of guidelines. Analysis of determinants before, or even after the introduction of a guideline, can guide the process of designing innovation strategies that will have the potential to create real change in the actual use of the guideline by health care professionals.31,33

In 2008 a grant was obtained for a research project to address the suboptimal implementation of ECV in the Netherlands.26 The project encompassed investigation of determinants of poor implementation (determinant analysis), the development of strategies to improve the implementation of ECV in the Netherlands, and evaluation of the (cost)effectiveness of these strategies. In our project, we used a framework based on several theories that have been used for the introduction and evaluation of innovations in a wide range of domains in (Dutch) health care.31 An implementation process is divided in four main consecutive stages: dissemination (meaning that every professional is actually supplied with the guideline); adoption (the intention of the professional to use or not to use the guideline); implementation (the use of the guideline in daily practice); and continuation (working with the guideline becomes routine practice). Several determinants may positively or

19 Chapter 1 General introduction

negatively influence the transition from one stage to the other and can be divided in four main categories: 1. Characteristics of the guidelines (e.g. pro- cedural clarity, complexity), 2. Characteristics of the health professional (e.g. personal benefits/drawbacks, outcome expectations, self-efficacy, client/ patient cooperation), 3. Characteristics of the organisation (e.g. staff capacity, financial resources, available time), and 4. Characteristics of the socio- political environment (e.g. legislation and regulations.31 Thus knowledge of the critical determinants facilitates the development of implementation strategies with the potential to actually improve the implementation of a guideline. These insights are applied in our project.

AIM OF THIS THESIS

The central aim of this thesis was to evaluate current care for term breech presentation. We have studied the following questions:

1. What is the effect of the increased elective caesarean delivery rates for term breech presentation on maternal and neonatal outcome, and is there still room for improvement left?

2. What is the effect of an elective caesarean sections for breech presentation on maternal and neonatal outcome in the subsequent pregnancy?

3. Which determinants hamper or facilitate the implementation of the guideline on ECV in the Netherlands?

4. What is the effectiveness of a client strategy, a care provider strategy or a combination strategy for the implementation of the guideline on ECV, in terms of:

a. adherence to the three core elements of the ECV guidelines: counselling for ECV, advising ECV and arranging an ECV attempt.

b. mode of delivery and maternal and neonatal outcome

5. What are the costs of the implementation strategies and how do they relate to their effectiveness?

20 Chapter 1 General introduction

OUTLINE OF THE THESIS

In part one of this thesis, we will address the change in mode of delivery for breech presentation at term since publication of the Term Breech Trial and its effect on perinatal and maternal outcome.

Chapter 2 describes the trend in mode of delivery for term breech presentations between 1999 and 2007 and the effect on perinatal outcome of the increased planned caesarean section rate since publication of the Term Breech Trial.

Chapter 3 describes the effect of the increased caesarean section rate due to breech presentation on maternal and perinatal outcome in the subsequent pregnancy.

In part two of this thesis, we identify which issues are potentially related to suboptimal implementation of ECV:

In chapter 4 we systematically reviewed the literature to assess whether the formulation of contraindications in guidelines is based on empiric data.

Chapter 5 evaluates patients’ willingness to opt for EV by means of a vignette study.

Chapter 6 describes a qualitative analysis of patients’ and professionals’ determinants (barriers and facilitators) for ECV, identified by patient interviews and focus group meetings with professionals.

Chapter 7 describes an online survey to evaluate the extent to which the barriers and facilitators for professionals influence the ability to counsel, advice and arrange an ECV. This was done by means of a quantitative analyses of an online questionnaire on the identified determinants.

Chapter 8 presents a systematic review on the potential of decision aids in obstetrics to improve informed decision making.

In part three, we describe and evaluate the implementation strategies we developed, based on the information obtained in part two.

Chapter 9 presents a retrospective description of the implementation rate of ECV from January 1st 2008 till December 31st 2009 in the participating

21 Chapter 1 General introduction

clusters in the randomised controlled trial.

Chapter 10 presents the results of the cluster randomised controlled trial on implementation strategies for ECV in 25 clusters in the Netherlands.

Chapter 11 evaluates the cost-effectiveness of the implementation strategies.

Chapter 12 describes the effect of the implementation strategies for ECV on the decisional conflict patients experience when deciding on ECV.

Chapter 13 describes the effect of ECV on mode of delivery in comparison to patients who do not opt or to whom ECV is not offered.

In the addendum of the thesis the study protocol is presented.

22 Chapter 1 General introduction

REFERENCES

1 Hannah ME, Hannah WJ, Hewson S a, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet [Internet]. 2000 Oct; 356: 1375–1383. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0140673600028403

2 Rietberg C. Term breech presentation in The Netherlands from 1995 to 1999: mortality and morbidity in relation to the mode of delivery of 33,824 infants. BJOG An Int J Obstet Gynaecol [Internet]. 2003 Jun [cited 2014 Mar 3]; 110: 604–609. Available from: http://doi.wiley.com/10.1016/S1470-0328(03)01907-4

3 Brouwers, HAA; Bruinse, HW; Dijs-elsinga, J; Van Huis, AM; De Miranda, E; Ravelli, ACJ; Tamminga, P; de Vries C. Jaarboek zorg in Nederland 2012 (the Netherlands Perinatal Registry, yearly report 2012) [Internet]. 2013. Available from: http://www. perinatreg.nl/uploads/150/141/PRN_jaarboek_2012_09012014_PDF.pdf

4 Rietberg, CCT; Elferink-Stinkens, PM; Visser G. The effect of the Term Breech Trial on medical intervention behaviour and neonatal outcome in The Netherlands: an analysis of 35,453 term breech infants. BJOG [Internet]. 2005 Aug [cited 2014 Mar 3]; 112: 1163; author reply 1163–4. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/16045537

5 Bragg F, Cromwell DA, Edozien LC, Gurol-urganci I, Mahmood TA, Templeton A, et al. Variation in rates of caesarean section among English NHS trusts after accounting for maternal and clinical risk : cross sectional study. BMJ. 2010;341 .

6 Sullivan EA, Moran K, Chapman M. Term breech singletons and caesarean section: a population study, Australia 1991-2005. Aust N Z J Obstet Gynaecol [Internet]. 2009 Oct [cited 2014 Mar 3]; 49: 456–460. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/19780725

7 Marshall N, Walker M, Fu R, Janik R. Vaginal Birth After Cesarean. 2010; 115: 1267– 1278.

8 Schutte JM, Steegers E a P, Santema JG, Schuitemaker NWE, van Roosmalen J. Mater- nal deaths after elective cesarean section for breech presentation in the Netherlands. Acta Obstet Gynecol Scand [Internet]. 2007 Jan [cited 2014 Mar 3]; 86: 240–243. Available from: http://www.ncbi.nlm.nih.gov/pubmed/17364290

9 Hofmeyr, GJ; Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012; . 10 23 Chapter 1 General introduction

10 Grootscholten, K; Kok, M; Oei, SG; Mol, BW; van der Post J. External cephalic version- related risks: a meta-analysis. Obs Gynecol [Internet]. 2008 [cited 2014 Mar 20]; 112: 1143–1151. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/? term=xternal+cephalic+version-related+risks:+a+meta-analysis.

11 Dunn PM. Eucharius Rösslin (c 1470-1526) of Germany and the rebirth of midwifery. Arch Dis Child Fetal Neonatal Ed [Internet]. 1998 Jul; 79: F77–8. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1720814&tool=pmcent rez&rendertype=abstract

12 McGuinness F. Prohylactic External Cephalic Version in Breech Presentation. Can Med Assoc J [Internet]. 1928 Jan; 18: 289–292. Available from: http://www.ncbi.nlm.nih. gov/pubmed/13448357

13 ML M. Clinical lessons about breech presentation. Ned Tijdschr Geneeskd. 1940; 19: 4020–4027.

14 Siegel, I; McNally H. Primary breech presentation and external cephalic version. South Med J. 1951; 44: 942–950.

15 Mensink WFa HH. Heeft uitwendige kering bij stuitligging zin? (Does external cephalic version have any value). NedTijdschrGeneeskd. 1980; 124: 1828–1831.

16 Raaij van R.F.M. L van der B. De plaats van de uitwendige versie in de moderne obst- etrie. (External cephalic version in modern obstetrics). Ned Tijdschr Geneeskd. 1987; 131: 1340–1342.

17 Verburgt, T; Offerhaus P. Uitwendige versie (External Cephalic Version). NTOG. 2007; 120: 4–7.

18 Greenhalgh, T; Robert, G; Macfarlane, F; Bate, P; Kyriakidou O. Diffusion of innovations in service organizations: systematic review and recommendations. Milbank Q [Internet]. 2004 [cited 2014 May 19]; 82: 581–629. Available from: https://c38xrlad- 25d4epd6rje0lpk.sec.amc.nl/pubmed/15595944

19 Otterlo Werkgroep dr. AJ van L en dr. MG van P. NVOG richtlijn - Stuitligging- versie 2.0 (Guideline on breech presentation - version 2.0 of the Dutch Society of Obstet- rics and Gynecology) [Internet]. 2008. Available from: file:///C:/Users/PS13C/Down- loads/Stuitligging (1).pdf

24 Chapter 1 General introduction

20 Feitsma, HA; Middeldorp JM, Oepkes D. De uitwendige versie bij de a terme stuit; een inventariserend onderzoek (ECV for breech presentation at term; an inventory study). NTOG. 2007; 120: 4–6.

21 Leung, TY; Lau, TK; Lo, KW; Rogers M. A survey of pregnant women’s attitude towards breech delivery and external cephalic version. Aust N Z J Obs Gynaecol. 2000; 40: 253–259.

22 Bewley, S; Robson, SC; Smith, M; Glover, A; Spencer J. The introduction of external cephalic version at term into routine clinical practice. Eur J Obs Gynecol Reprod Biol [Internet]. 1993 [cited 2014 Mar 20]; 52: 89–93. Available from: https://c38xrlad- 25d4epd6rje0lpk.sec.amc.nl/pubmed/8157147

23 Caukwell S, Joels L a, Kyle PM, Mills MS. Women’s attitudes towards management of breech presentation at term. J Obstet Gynaecol [Internet]. 2002 Sep [cited 2014 Mar 20]; 22: 486–488. Available from: http://www.ncbi.nlm.nih.gov/pubmed/12521413

24 Raynes-Greenow CH, Roberts CL, Barratt A, Brodrick B, Peat B. Pregnant women’s preferences and knowledge of term breech management, in an Australian setting. Midwifery [Internet]. 2004 Jun [cited 2014 May 19]; 20: 181–187. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15177862

25 Yogev Y, Horowitz E, Ben-Haroush a, Chen R, Kaplan B. Changing attitudes toward mode of delivery and external cephalic version in breech presentations. Int J Gynae- col Obstet [Internet]. 2002 Dec; 79: 221–224. Available from: http://www.ncbi.nlm. nih.gov/pubmed/12445986

26 Vlemmix F, Rosman AN, Fleuren MAH, Rijnders MEB, Beuckens A, Haak MC, et al. Implementation of the external cephalic version in breech delivery. Dutch national implementation study of external cephalic version. BMC Pregnancy Childbirth. 2010; 10: 20.

27 Fullan M. The New Meaning of Educational Change, Fourth Edition [Internet]. Teach- ers College Press; 4 edition; 2007 [cited 2014 May 30]. p. 352. Available from: http:// www.amazon.com/Meaning-Educational-Change-Fourth-Edition/dp/0807747653

28 Guldbrandsson K. From news to everyday use The difficult art of implementation. Stockholm: Strömberg;

29 Grol R, Buchan H. Clinical guidelines: what can we do to increase their use? Med J Aust [Internet]. 2006 Sep 18 [cited 2014 May 30]; 185: 301–302. Available from: http://www.ncbi.nlm.nih.gov/pubmed/16999667

25 Chapter 1 General introduction

30 Heinen MM, Bartholomew LK, Wensing M, van de Kerkhof P, van Achterberg T. Sup- porting adherence and healthy lifestyles in leg ulcer patients: systematic development of the Lively Legs program for dermatology outpatient clinics. Patient Educ Couns [Internet]. 2006 May [cited 2014 May 30]; 61: 279–291. Available from: http:// www.ncbi.nlm.nih.gov/pubmed/15964733

31 Fleuren M, Wiefferink K, Paulussen T. Determinants of innovation within health care organizations: literature review and Delphi study. Int J Qual Health Care[In- ternet]. 2004 Apr; 16: 107–123. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/15051705

32 Grol, R; Wensing, M; Eccles M. Improving Patient Care: The Implementation of Change in Clinical Practice [Internet]. Elsevier; 2005 [cited 2014 May 19]. p. 304. Available from: http://www.amazon.com/Improving-Patient-Care-Implementation- Clinical/dp/075068819X

33 Fleuren M, Paulussen T, Dommelen P V., Buuren S V. Measurement Instrument for Determinants of Innovations (MIDI). BMJ Qual Saf [Internet]. 2013 Aug 15 [cited 2014 May 30]; 22: A42–A42. Available from: http://qualitysafety.bmj.com/lookup/ doi/10.1136/bmjqs-2013-002293.124

26 27

2 Term breech deliveries in the Netherlands: did the increased cesarean rate affect neonatal outcome? A population based cohort study

Floortje Vlemmix

Lester Bergenhenegouwen

Jelle M. Schaaf

Sabine Ensing

Ageeth N. Rosman

Anita C.J. Ravelli

Joris A.M. van der Post

Arno Verhoeven

Gerard H. Visser

Ben Willem J. Mol

Marjolein Kok

ACTA Obstetricia et Gynelogica; accepted for publication.

29 Chapter 2 Term breech deliveries in the Netherlands

ABSTRACT

Objective The aim of this study was to evaluate the effect of the increased cesarean rate for term breech presentation on neonatal outcome. We also investigate whether the clinical case selection for vaginal delivery applied by Dutch obstetricians led to an optimization of neonatal outcome, or if there is still room for improvement in terms of perinatal outcome.

Design Retrospective cohort.

Settingthe Netherlands.

Population Singleton term breech deliveries from 37 0/7 to 41 6/7 weeks, excluding fetus with congenital malformations or antenatal death.

Method We used data from the Dutch national perinatal registry from 1999 up to 2007.

Main outcome measures Perinatal mortality and morbidity.

Results We studied 58,320 women with a term breech delivery. There was an increase in the elective cesarean rate (from 24% to 60%). As a consequence, overall perinatal mortality decreased (1.3 vs. 0.7‰; OR 0.51 (95% CI 0.28 – 0.93)), while it remained stable in the planned vaginal birth group (1.7 vs. 1.6‰; OR 0.96 (95% CI 0.52 – 1.76)). The number of caesareans to prevent one perinatal death was 338.

Conclusions Adjustment of the national guidelines after publication of the Term Breech Trial resulted in a shift towards elective cesarean and a decrease of perinatal mortality and morbidity among women delivering a child in breech at term. Still 40% of these women attempt vaginal birth. The relative safety of an elective cesarean should be weighed against the consequences of a scarred uterus in future pregnancies.

30 Chapter 2 Term breech deliveries in the Netherlands

INTRODUCTION

The term breech trial was the largest randomized controlled trial to investigate the effect of mode of delivery for term breech deliveries on neonatal and maternal outcomes.1 The study reported a highly significant decrease in combined perinatal mortality and morbidity scores among women undergoing a planned caesarean compared to those planned to deliver vaginally (5.0 vs. 16.0‰), RR 0.33; 95%CI 0.19-0.56). The relative benefit of a planned caesarean was strongest in countries with a perinatal mortality rate below 2.0% (57.0 vs. 4.0‰), RR 0.07; 95%CI 0.02-0.29). However, planned caesarean delivery was not associated with a reduction in risk of death or neurodevelop- mental delay in children at 2 years of age.2 The authors suggested that the fact that only 44% of children from the initial trial were investigated in the follow up may have limited of statistical power to detect differences at age 2. Maternal outcomes immediately postpartum and at 2 years were similar after planned caesarean section and planned vaginal birth for the singleton breech fetus at term.1,3

Publication of the term breech trial had a worldwide impact on clinical practice. An Australian population study reported an increase of elective caesarean rate for breech deliveries up to 76.6% and an overall caesarean rate of 96.3% in 2005.4 A Danish study reported an increase of overall caesarean rate up to 94.2% in 2008.5 Within the latter cohort, intrapartum and early neonatal mortality was reduced from 1.3‰ (1997 - 2000) to 0.5‰ in (2001 - 2008) (RR 0.38; 95%CI 0.15-0.98).

In the Netherlands, the overall caesarean rate increased from 50 to 78% within three months after publication of the term breech trial, mainly due to an increase of elective cesareans.6 This resulted in a significant decrease of neonatal death (OR 0.53; 95%CI 0.33-0.83) and neonatal trauma (OR 0.26; 95%CI 0.14-0.50) between 1998 and 2002. However, the most recent ‘Dutch report of perinatal care’ from 2008 showed that after the initial rise in elective caesareans, there was no further increase, and the overall caesarean rate remained stable at 80%.7

The difference in cesarean rate for term breech presentation between the Netherlands and other countries is remarkable and can only be explained in a multifactorial way. Factors potentially influencing the difference in clinical care are; interpretation of research findings presented in published articles

31 Chapter 2 Term breech deliveries in the Netherlands

(reflected by differences in content and style of phrasing of guidelines and translation of guidelines to women by their attending physician), claim cul- ture, jurisdiction and organization of obstetric care.

The aim of this study is to evaluate the effect of the rise in cesarean rate for term breech delivery on neonatal outcome up to 2007. Furthermore, in order to investigate whether the Dutch management of the clinical obstetric issue of safe term breech delivery is justified, we wondered whether the clinical case selection for vaginal delivery applied by Dutch obstetricians, resulting in a relatively large proportion of women (40%) with a term breech being selected for vaginal delivery, led to an optimization of neonatal outcome, or if there is still room for improvement in terms of perinatal mortality and morbidity.

MATERIAL AND METHODS

This study was performed in a population based cohort using the Nether- lands Perinatal Registry (PRN). The PRN consists of population-based data containing information on pregnancies, deliveries and (re)admissions until 7 days after birth. The PRN database is obtained by a validated linkage of three different registries: the midwifery registry (LVR 1), the obstetric registry (LVR 2) and the neonatology registry (LNR) of hospital admissions of newborns.8 The coverage of the PRN registry is about 96% of all deliveries in the Nether- lands. All data contained in the PRN are voluntarily recorded by the caregiver during prenatal care, delivery and the neonatal period. Data on neonatal admissions are registered in 53-58% of hospitals in The Netherlands.7 The neonatal follow up, up to 28 days after birth, is provided by 80% of neonatal care units to the PRN. The data are sent to the national registry office annually, where a number of range and consistency checks are conducted. For this study all births between 1 January 1999 (inception of the PRN registry) and 31 December 2007 (the most recent, validated and linked year available at start of the analyses for this study) were selected. The PRN gave approval for anonymous use of requested data for this analyses (#11.47). No ethical approval or informed consent was required for this retrospective cohort study as the use of the data was within the confines pose by Dutch law on the use of registry data.

We studied pregnant women with singleton breech presentations who delivered between 37 0/7 and 41 6/7 weeks of gestation. Exclusion criteria were

32 Chapter 2 Term breech deliveries in the Netherlands antepartum death and major congenital malformations. Major congenital malformations were defined as lethal congenital malformations (e.g. trisomy 18, Potter’s syndrome, central nervous system abnormalities (meningomy- elocele, exencephaly, anencephaly, hydrocephaly and microcephaly) and infants with multiple congenital malformations (including spina bifida, intesti- nal atresias and congenital heart disease). To validate the perinatal mortality we verified all causes of mortality through the hospital charts.

Planned vaginal birth was defined as initially intended vaginal birth; the combination of actual vaginal birth and emergency caesarean delivery, including both failure to progress and fetal distress as indications for a caesarean delivery. Elective caesarean section was defined as all women opting for a planned caesarean delivery.

The primary outcome was perinatal mortality, defined as intrapartum death or death within 28 days after birth. Secondary outcomes were; a five-minute Apgar score below 7, neonatal trauma (defined as a composite of intracerebral bleeding, cephalic hematoma, facial nerve paresis, brachial plexus lesion, fracture of clavicle, humerus or femur and other trauma), and poor neonatal outcome was defined as perinatal mortality, five-minute Apgar score below 7 and neonatal trauma.

We analysed trends of mode of delivery and neonatal outcomes over the years and differences in neonatal outcomes before and after publication of the term breech trial (before October 2000 and after November 2000). We performed univariate analysis and provided outcomes for different subgroups: parity (nulliparous/multiparous), birth weight (<=3500 / >3500 gram), type of breech (complete/frank), and onset of labor (spontaneous vs. induced or augmented).

Furthermore, we looked at the difference in elective caesarean delivery rate in relation to neonatal outcomes among Dutch hospitals. We not only compared neonatal outcome before and after November 2000, but also the impact of planned vaginal delivery and elective caesarean delivery. All statistical analyses were carried out with SAS 9.2 (SAS Institute, Cary, NC).

33 Chapter 2 Term breech deliveries in the Netherlands

RESULTS

There were 1.4 million singleton term deliveries in the nine year study period, of which 4.4% were breech deliveries. We excluded 161 cases of antenatal death (0.27%) and 642 cases with major congenital abnormalities (1.09%) from the analysis, leaving 58 320 women in the study. Figure 1 shows the trend in the mode of delivery for breech presentation between 1999 and 2007. Publication of the term breech trial led to an increase in elective caesareans from 24% before publication of the study to 60% after.

Figure 1. Trends in vaginal birth and cesarean section rate in women with an infant in breech presentation at term between 1999 and 2007

Publication TBT

Figure 2: Trends in perinatal outcome after breech birth (elective cesarean and planned vaginal birth) at term between 1999 and 2007

2,4 Publication TBT 2,2 2 1,8 1,6 1,4 1,2 1 0,8 0,6 oor neonatal outcome oor

P 0,4 0,2 as percentage ofbirth all breech percentageas 0 1999 2000 2001 2002 2003 2004 2005 2006 2007 Year

34 Chapter 2 Term breech deliveries in the Netherlands

Figure 2 shows the impact on the neonatal outcome. The increase in elective caesarean rate resulted in a decrease in perinatal death (1.3‰ before Oct 2000 vs. 0.7‰ after Nov 2000, OR 0.51; 95%CI 0.28-0.93), a decrease in low Apgar score (20.2 vs. 9.6‰, OR 0.47; 95%CI 0.40-0.55), and a decrease in neonatal trauma (4.8 vs. 2.2‰; OR 0.46 95%CI 0.34-0.63) (Table 1).

Subgroup analysis showed a smaller increase in elective caesarean rate among multiparous compared to nulliparous (31 vs. 39%). The increase of 31% in elective caesarean delivery within the multiparous group did not result in statistically significant decrease of overall perinatal mortality rate among multiparous women, although there was also in this group a positive trend (0.8‰ before Oct 2000 vs. 0.6‰ after Dec 2000, OR 0.69; 95%CI 0.22- 2.20) (Table 2).

Table 1. Neonatal outcome following term breech delivery before and after publication of the Term Breech Trial (TBT)

Before TBT After TBT Odds ratio p-‐value Overall (1999 – Oct 2000) ‰) ( (Dec 2000 – 2007) ‰) ( (95% CI)

n 12,383 45,937

Elective cesarean (%) 24 60 $ Emergency cesarean (%) 34 45 Perinatal death^ 16 (1.3) 30 (0.7) 0.51 (0.28 -‐ 0.93) 0.01 Low Apgar score* 250 (20.2) 439 (9.6) 0.47 (0.40 -‐ 0.55) <0.0001 Neonatal Trauma~ 60 (4.8) 103 (2.2) 0.46 (0.34 -‐ 0.63) <0.0001

Nulliparous

n 7,677 28,871 Elective cesarean (%) 23 62 Perinatal death^ 12 (1.6) 20 (0.7) 0.44 (0.22 -‐ 0.91) 0.03 Low Apgar score* 170 (22.1) 269 (9.3) 0.42 (0.34 -‐ 0.50) <0.0001

Neonatal Trauma~ 40 (5.2) 67 (2.3) 0.44 (0.30 -‐ 0.66) <0.0001

Multiparous n 4,706 17,066 Elective cesarean (%) 26 57 Perinatal death^ 4 (0.8) 10 (0.6) 0.69 (0.22 – 2.20) 0.53 Low Apgar score* 80 (17.0) 170 (10.0) 0.58 (0.45 -‐ 0.76) <0.0001

Neonatal Trauma~ 20 (4.2) 36 (2.1) 0.50 (0.29 -‐ 0.86) 0.01

^Intrapartum death and neonatal mortality up to 7 days *Five-‐minute Apgar score <7 ~composite score of: intracerebral bleeding, cephalic haematoma, facial nerve paresis, brachial plexus lesion, fracture of clavicle, humerus or femur and other trauma $Percentage of pursued vaginal birth

35 Chapter 2 Term breech deliveries in the Netherlands

‰) -‐ -‐ -‐ 2 (0.1) 2 6 (0.2) 6 6 (0.2) 6 9) ( 11 (0.4) 11 1 (0.04) 1 1 (0.04) 1 65 (2.4) 65 27 (1.0) 27 (n=27,54 After TBT After

-‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ ‰) ( caesarean Elective 0 (0.0) 0 13 (4.4) 13 (n=2,954) (n=2,954) Before TBT Before

‰)

-‐ -‐ 3 (0.2) 3 2 (0.1) 2 4 (0.2) 4 30 (1.6) 30 19 (1.0) 19 28 (1.5) 28 30 (1.6) 30 76 (4.1) 76 After TBT After 374 (20.3) 374 avicle, humerus or femur and other trauma. or femur humerus avicle, (n=18,388) ( (n=18,388)

-‐ Pursued vaginal birth vaginal Pursued ‰) TBT ( 9 (1.0) 9 2 (0.2) 2 3 (0.3) 3 3 (0.3) 3 Before Before 22 (2.3) 22 10 (1.0) 10 25 (2.7) 25 16 (1.7) 16 60 (6.4) 60 (n=9,429) (n=9,429) 237 (25.1) 237

‰)

TBT

-‐ -‐ -‐ -‐ -‐ (12.1) 1 (0.1) 1 1 (0.1) 1 12 (1.4) 12 13 (1.6) 13 14 (1.7) 14 After After 100 (n= 8,284) ( 8,284) (n=

caesarean Emergency

-‐ -‐ -‐ -‐ -‐ -‐ ‰) ( 3 (0.9) 3 1 (0.3) 1 5 (1.6) 5 4 (1.2) 4 59 (18.4) 59 (n=3,210) (n=3,210) Before TBT Before

‰) -‐ -‐ traumata and therefore the separate trauma do not always ad up to the composite outcome of children with neonatal trauma. composite the to up ad always not do trauma separate the therefore and traumata TBT 274

After After (27.1) 3 (0.3) 3 1 (0.1) 1 3 (0.3) 3 4) ( to 7 days 7 to 18 (1.8) 18 19 (1.9) 19 28 (2.8) 28 17 (1.7) 17 62 (6.1) 62

(n=10,10

‰)

Vaginal birth -‐ 1 (0.2) 1 9 (1.4) 9 3 (0.5) 3 3 (0.5) 3 10 (1.6) 10 10 (1.6) 10 25 (4.0) 25 11 (1.8) 11 56 (9.0) 56 178 (28.6) 178 Before TBT Before (n=6,219) ( (n=6,219)

Neonatal outcome in case of term breech delivery according to mode of delivery.

minute Apgar score <7 score minute Apgar -‐ Composite of vaginal delivery and emergency caesarean. delivery and vaginal Composite of One child could have suffered multiple neonatal multiple suffered have could One child Table 2. Table death^ Perinatal score* Apgar Low Femur fracture Other mortality up neonatal and death ^Intrapartum ¥ *Five ~composite score of: intracerebral bleeding, cephalic haematoma, facial nerve paresis, brachial plexus lesion, fracture of cl Humerus fracture Neonatal Trauma~ Neonatal Facial nerve lesion nerve Facial fracture Clavicula Brachial plexus lesion Cephalic hematoma Cephalic Intra cerebral bleeding cerebral Intra

36 Chapter 2 Term breech deliveries in the Netherlands

Despite the increased emergency caesarean rate (from 34 to 45%) within the planned vaginal birth group after publication of the term breech trial, the perinatal mortality rate within the planned vaginal breech group remained stable (1.7 before October 2000 vs. 1.6‰ after December 2000, OR 0.96; 95%CI 0.52-1.76). The prevalence of low Apgar score (25.1 vs. 20.3‰, OR 0.81; 95% 0.68-0.95) and of neonatal trauma (6.4 vs. 4.1‰, OR 0.65; 95% 0.46-0.92) improved in women with a planned vaginal delivery after publication of the term breech trial (Table 2). Of the 46 cases of perinatal mortality, breech presentation had not been diagnosed until birth in 9 cases (19.1%).

Table 3 shows the gain in neonatal outcome by performing an elective caesarean delivery. There was no perinatal mortality in the elective caesarean group, compared to 1.6‰ in the planned vaginal birth group. Elective caesarean delivery was associated with a lower risk of poor neonatal outcome compared to planned vaginal birth (OR 0.14, 95% CI 0.11 to 0.18), including a lower risk of low Apgar score (OR 0.12, 95% CI 0.09 to 0.16), and a lower risk of neonatal trauma (OR 0.24, 95% CI 0.15 to 0.37).

Exclusion of the nine women with breech presentation not diagnosed until labor, who were all included in the planned vaginal birth group, did not alter the difference in perinatal mortality between the two groups (1.3‰ in the planned vaginal birth group vs. none in the elective caesarean group).

Table 4 shows the relation between mode of delivery and neonatal- out come for different subgroups. Elective caesarean led to a strong decrease in neonatal morbidity and mortality regardless of parity, type of breech, birth weight, or onset of labor.

Table 3. Neonatal outcomes per mode of delivery: Elective cesarean delivery compared to pursued vaginal breech delivery since publication of the Term Breech Trial.

Pursued vaginal $ birth Elective cesarean delivery Odds ratio p-‐value Poor neonatal outcome# (Dec 424 2000 – 2007) (23.1) (‰) (Dec 92 2000 – 2007) (3.3) (‰) 0.14 (95% (0.11 – CI) 0.18) <0.0001 Perinatal death^ 30 (1.6) 0 -‐ <0.0001 Low Apgar score* 374 (20.3) 65 (2.5) 0.12 (0.09 – 0.16) <0.0001 Neonatal Trauma~ 76 (4.1) 27 (1.0) 0.24 (0.15 – 0.37) <0.0001 #Composite of perinatal death, low Apgar score and neonatal trauma ^Intrapartum death and neonatal mortality up to 7 days *Five-‐minute Apgar score <7 ~Composite score of: intracerebral bleeding, cephalic haematoma, facial nerve paresis, brachial plexus lesion, fracture of clavicle, humerus or femur and other trauma $Composite of vaginal delivery (55%) and emergency cesarean (45%)

37 Chapter 2 Term breech deliveries in the Netherlands

value -‐

0.013 0.0004 0.0002 0.0004 p <0.0001 <0.0001 <0.0001 <0.0001

0.39) 0.54) 0.46) 0.39) 0.42) 0.87) 0.25) 0.65)

– – – – – – – – (95%CI) Neonatal traumata Odds ratio

0.23 (0.13 (0.13 0.23 (0.12 0.25 (0.08 0.19 (0.14 0.23 (0.15 0.25 (0.29 0.50 (0.09 0.15 0.39 (0.23 (0.23 0.39

1 1 1 1 1 1 1 value -‐ 0.008 <0.000 <0.000 <0.000 <0.000 <0.000 <0.000 <0.000 p

–

0.20) 0.21) 0.15) 0.15) 0.15) 0.15) 0.17) 0.20)

– – – – – – – 0.11) (95%CI) Odds ratio Low Apgar score* Low Apgar 0.09 (00.06 (00.06 0.09 0.13 (0.09 (0.09 0.13 (0.08 0.13 0.11 (0.08 (0.08 0.11 0.11 (0.08 (0.08 0.11 0.11 (0.08 (0.08 0.11 0.11 (0.07 (0.07 0.11 (0.12 0.15

cesarean group. cesarean (1.4 vs. 0‰) (0.8 vs. 0‰) (2.1 vs. 0‰) (2.0 vs. 0‰) (1.2 vs. 0‰) (3.2 vs. 0‰) (1.5 vs. 0‰) (1.8 vs. 0‰) -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ death^ Perinatal Odds ratio (Incidences) (‰)

minute Apgar score <7 score minute Apgar -‐ No odds ratio could be calculated due to no events in the elective cesarean delivery group. the in events no to due calculated be could ratio odds No Birth weight Birth

*Five publication of the term breech trial for different subgroups for trial breech term the of publication the planned in mortality no was there cesarean delivery since elective vs. breech vaginal planned for outcomes Neonatal 4: Table Parity Nulliparous Multiparous breech of Type grams >3500 labor of Onset / Induced calculated for perinatal mortality since be could ratios odds No days. 7 to up mortality neonatal and death ^Intrapartum augmented # ≤3500 grams ≤3500 Frank Complete Spontaneous

38 Chapter 2 Term breech deliveries in the Netherlands

In stratified analysis results for parity, type of breech, birth weight and onset of labor, did not change the strong relationship between mode of delivery and poor neonatal outcome. There was no interaction between parity and mode of delivery for poor neonatal outcome. Birth weight below 3500gr was related to a poorer outcome compared to higher birth weight (>3500gr). Complete breech was related to a poorer outcome compared to frank breech presentation and induced or augmented labor to a worse outcome compared to spontaneous birth (table 4).

Data from 99 different hospitals are registered in the PRN registry. The total number of breech births within seven years after the term breech trial varied from 9 to 739 per hospital with a median of 381. The elective caesarean rate in these women varied from 14 to 80% between the hospitals. Perinatal death (n=30) occurred in 24 hospitals (24%), with a maximum of 9.4‰ (3 of 318 breech deliveries in seven years). Correlation analyses did neither demonstrate an association between hospital elective caesarean rate and perinatal death, low Apgar score, neonatal trauma, or poor neonatal outcome, nor was there an association between the volume of breech births and the neonatal outcome.

Since publication of the term breech trial, 1 692 more caesarean deliveries (combined elective and emergency caesarean) were performed annually. This led to an annual reduction of five neonatal deaths (number needed to treat (NNT) 338), 126 neonates with low Apgar scores (NNT 13), and 30 neonates with birth traumata (NNT 56).

If all women who nowadays still undergo a planned vaginal breech birth, would receive an elective caesarean, 6 490 more elective CS would be performed. This would lead to an additional annual reduction of 10 neonatal mortalities, 116 neonates with low Apgar scores and 20 neonates with birth traumata.

Breech deliveries account for 5.8% (10 of 172 per year) of the perinatal and neonatal mortality up to 28 days postpartum among term singleton deliveries in the Netherlands. A policy of elective caesarean for all term breech deliveries could lower the overall term neonatal mortality in term deliveries with 6.8% from 172 to 162 cases per year.

39 Chapter 2 Term breech deliveries in the Netherlands

DISCUSSION

In this analysis of almost a decade of term breech deliveries in the Nether- lands, we found a significant improvement of neonatal outcomes most likely due to the increased elective caesarean rate that occurred after publication of the Term Breech Trial. Nulliparous women benefitted more from the change in policy than multiparous women as elective caesarean rates were higher in the first group. Despite the lower percentage of women opting for or offered a vaginal delivery, and despite a higher emergency caesarean rate during vaginal breech birth, neonatal outcome within the planned vaginal birth group did not improve.

This study on breech birth is unique in its size, and the Dutch setting makes it possible to evaluate the effect of an increase, though not complete chance in daily practice of elective caesarean for breech presentation at term. We showed that based on patient characteristics including parity, onset of labor, type of breech presentation and birth weight, no subgroup of women could be identified with a low risk of poor neonatal outcome during planned vaginal delivery compared to elective caesarean. Several other studies tried to identify a subgroup of patients with low risk at adverse perinatal outcome during vaginal breech delivery. Most of these studies have a lack of power to detect a significant difference on for instance perinatal mortality.8-11 There is one large prospective cohort trial from the PREMODA study group, who evaluated the birth outcome of 2 502 planned vaginal deliveries and 5 573 planned caesarean deliveries and could not find a significant difference in fetal and neonatal mortality or serious neonatal morbidity.12 Women were found eligible for planned vaginal delivery according to the CNGOF guideline and trial of labor was guided according to advices from national guidelines as well. The composite poor neonatal outcome was 16.0‰ in the planned vaginal delivery group vs. 14.5‰ in the planned caesarean group (OR 1.10, (95%CI 0.75-1.61)). Their rates of low Apgar scores and birth trauma are higher in the caesarean section group, but much lower in the planned vaginal breech group, compared to the data presented in this study. Their explanation of the non-significant results is partly the quality advantages of a prospective study design compared to all the significant results in retrospective studies. Similar to this prospective cohort study, we have evaluated the cause of all perinatal mortalities, which improves the reliability of our outcomes compared to those of other retrospective studies. Secondly, the authors state that the

40 Chapter 2 Term breech deliveries in the Netherlands regulated selection procedure for planned vaginal delivery and intrapartum management might explain the non-significant results. Although evidence based proof of these antenatal examinations is lacking, they assume that special attention to the decision on mode of delivery might be an explanation of the low poor neonatal outcome rate in the vaginal delivery group. The potential effectiveness of application of a consensus guidelines is also seen in a study by Vendittelli et al. with less neonatal complications in centers working according to these guidelines compared to other centers (OR 0.27 (95%CI 0.09-0.85)).13 Since vaginal breech delivery is still common practice in the Netherlands, the high poor neonatal outcome rates in comparison to other countries, should be focus of research and training of obstetricians.

In this study, we choose to use a cut off level of birth weight of 3500 grams and not small for gestational age. The correlation of small for gestational age and adverse neonatal outcome is reported in several studies.13-16 Concerning breech presentation, there is discussion among clinicians whether a term fetus below or above a certain birth weight could be more at risk for breech birth related adverse outcome. For instance in the previously mentioned PREMODA study, the estimated fetal weight should not exceed 3800 grams to qualify for planned vaginal breech delivery.

An important strength of this study is the well-maintained, population-based national registry, covering 96% of all births. The missing 4% are deliveries supervised by general practitioners and midwives practices that did not contribute data to the PRN. Since breech presentation is an indication for hospital birth, we do not expect that these missing data would influence our outcomes. Neonatal morbidity of admitted children is not measured in all of the neonatal departments of the 99 hospitals. When analyses were restricted to the hospitals with complete neonatal admission registration, similar results were found.

Analysing data of a population-based national registry presented some disadvantages as well. Only data up to 2007 were available for analysis. How- ever, a recent survey among 24 hospitals in the Netherlands showed similar elective caesarean rates for breech presentation at term in 2011 and 2012. Second, we were not able to perform a complete intention to treat analysis for mode of delivery. Women who opted for elective caesarean, but underwent a vaginal birth (for instance due to unexpected start of delivery and fast progression), could not be included in the caesarean group. In the term

41 Chapter 2 Term breech deliveries in the Netherlands

breech trial, 9.7% of women with a planned caesarean in fact had a vaginal delivery. Evaluation of the perinatal mortalities in hospital files revealed that 19.1% breech presentations was not diagnosed until birth. We did not have this information for the whole study population, consequently there will be a misclassification of women in the planned vaginal birth group, who were simply not aware of the breech presentation. Thus the positive outcome of elective caesarean might be overestimated.

The high percentage of women opting for a planned vaginal breech birth is remarkable compared to other countries. We have no explanation for this high rate of trial of labor in case of breech presentation. It is probably a combination of a doctor and patient driven decision.17 The large intra hospital variation of planned caesarean rates (14-80%) reveal a large variety in management of breech presentation among hospitals. This might be a result of population differences and attitudes of obstetricians towards vaginal breech delivery.

When performing elective caesarean sections, attention should be paid to the gestational age at which these are planned in order to minimize neonatal morbidity related to elective caesarean sections such as respiratory distress syndrome.18-19 The downside of increased caesarean rates is the increased maternal morbidity and mortality. In our database, only two maternal deaths were reported within the study group. A previous study by the Dutch Maternal Mortality Committee, reported four maternal deaths after elective caesarean for term singleton breech delivery (0.47‰) from 2000- 2002.20 Other studies report maternal mortality rates of 0.04‰ for elective caesarean and 0.13‰ for repeat caesarean.20,21 These data support the pre- sumption of underreporting of maternal mortality and morbidity in the database. Therefore we cannot comment on the effects of modes of delivery on maternal outcomes in our cohort.

A frequently mentioned argument against a policy of elective caesarean -de livery is the lower exposure of the health care professional to vaginal breech delivery, and thus a loss of expertise in this field. However, in our opinion, training of the professional is no reason to expose healthy foetuses to high risks. Experience should be gained by simulation training, as is current practice with other uncommon events such as shoulder dystocia, postpartum haemorrhage and newborn life support.

42 Chapter 2 Term breech deliveries in the Netherlands

Our results can be used by clinicians during counselling of women with a term breech presentation for mode of delivery. In the current patient information leaflets of the Dutch Society of Obstetrics and Gynaecology is written that there is no difference in mortality or development at the age of two between vaginal and elective caesarean deliveries (based on the two year follow up of the Term Breech Trial). This information needs to be revised in the light of our findings with the significant difference in mortality rate. To properly inform patients, a combination of risk presentations (absolute risks, relative risks and figures) is necessary to enable individual informed decision making.22,23 For example, absolute numbers on a vaginal breech delivery should be mentioned as well; 55% of all planned breech deliveries lead to a vaginal birth and the chance of a normal neonatal outcome without any additional neonatal care, is 97% in case of planned vaginal breech birth in our study population.

During the study period, the percentage of breech presentation at birth remained stable around 4%. This suggests that there is still room for improvement of the implementation of external cephalic version, as this relative safe treatment can significantly improve neonatal outcome by preventing breech presentation at birth.

Due to a rise in caesarean delivery rate in case of term breech presentation, there was a significant improvement of neonatal outcome. However, 40% of term breech deliveries in the Netherlands are still planned vaginal deliveries and these deliveries generate a tenfold fetal mortality rate compared to elective caesarean delivery. Subgroup analysis could not identify antepartum parameters which could distinguish between women with low versus high risk vaginal breech birth . These facts need to be discussed when women opt for a vaginal breech delivery.

43 Chapter 2 Term breech deliveries in the Netherlands

REFERENCES

2 Whyte H, Hannah ME, Saigal S, Hannah WJ, Hewson S, Amankwah K, et al. 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 [Internet]. 2004 Sep [cited 2014 Mar 2]; 191: 864–871. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15467555

3 Hannah ME, Whyte H, Hannah WJ, Hewson S, Amankwah K, Cheng M, et al. 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 [Internet]. 2004 Sep [cited 2014 Feb 19]; 191: 917–927. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15467565

4 Sullivan EA, Moran K, Chapman M. Term breech singletons and caesarean section: a population study, Australia 1991-2005. Aust N Z J Obstet Gynaecol [Internet]. 2009 Oct [cited 2014 Mar 3]; 49: 456–460. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/19780725

5 Hartnack Tharin JE, Rasmussen S, Krebs L. Consequences of the Term Breech Trial in Denmark. Acta Obstet Gynecol Scand [Internet]. 2011 Jul [cited 2014 Feb 20]; 90: 767–771. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21476999

6 Rietberg C. Term breech presentation in The Netherlands from 1995 to 1999: mortality and morbidity in relation to the mode of delivery of 33,824 infants. BJOG An Int J Obstet Gynaecol [Internet]. 2003 Jun [cited 2014 Mar 3]; 110: 604–609. Available from: http://doi.wiley.com/10.1016/S1470-0328(03)01907-4

7 Van der Leeuw-Harmsen, L; G. Rijninks-van Driel, G; Morshuis, RHL; J. van Lith, J; Tamminga P. Jaarboek zorg in Nederland 2008 (the Netherlands Perinatal Registry, yearly report 2008) [Internet]. 2011. Available from: http://www.perinatreg.nl/uploads/150/122/Jaarboek_Zorg_in_Nederland_2008.PDF

44 Chapter 2 Term breech deliveries in the Netherlands

8 Méray N, Reitsma JB, Ravelli ACJ, Bonsel GJ. Probabilistic record linkage is a valid and transparent tool to combine databases without a patient identification number. J Clin Epidemiol [Internet]. 2007 Sep [cited 2014 Jan 21]; 60: 883–891. Available from: http://www.ncbi.nlm.nih.gov/pubmed/17689804

9 Michel S, Drain a, Closset E, Deruelle P, Ego a, Subtil D. Evaluation of a decision protocol for type of delivery of infants in breech presentation at term. Eur J Obstet Gy- necol Reprod Biol [Internet]. Elsevier Ireland Ltd; 2011 Oct [cited 2014 Mar 3]; 158: 194–198. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21641106

10 Van Loon a J, Mantingh a, Serlier EK, Kroon G, Mooyaart EL, Huisjes HJ. Randomised controlled trial of magnetic-resonance pelvimetry in breech presentation at term. Lancet [Internet]. 1997; 350: 1799–1804. Available from: http://www.ncbi.nlm.nih. gov/pubmed/9428250

11 Alarab, M;Regan, C; O’Connell, MP; Keane, DP; O’Herlihy, C; Foley M. Singleton vaginal breech delivery at term: still a safe option. Obs Gynecol [Internet]. 2004 [cited 2014 Mar 3]; 103: 407–412. Available from: https://c38xrlad25d4epd6rje0lpk.sec. amc.nl/pubmed/14990399

12 Goffinet F, Carayol M, Foidart J-M, Alexander S, Uzan S, Subtil D, et al. Is planned vaginal delivery for breech presentation at term still an option? Results of an obser- vational prospective survey in France and Belgium. Am J Obstet Gynecol [Internet]. 2006 Apr [cited 2014 Mar 3]; 194: 1002–1011. Available from: http://www.ncbi.nlm. nih.gov/pubmed/16580289

13 Vendittelli F, Pons JC, Lemery D, Mamelle N. The term breech presentation: Neonatal results and obstetric practices in France. Eur J Obstet Gynecol Reprod Biol [Internet]. 2006 Apr 1 [cited 2014 Mar 3]; 125: 176–184. Available from: http://www.ncbi.nlm. nih.gov/pubmed/16099580

14 Jacobsson B, Ahlin K, Francis a, Hagberg G, Hagberg H, Gardosi J. Cerebral palsy and restricted growth status at birth: population-based case-control study. BJOG [Inter- net]. 2008 Sep [cited 2014 Mar 3]; 115: 1250–1255. Available from: http://www.ncbi. nlm.nih.gov/pubmed/18715410

15 Boulet SL, Alexander GR, Salihu HM, Kirby RS, Carlo W a. Fetal growth risk curves: defining levels of fetal growth restriction by neonatal death risk. Am J Obstet Gynecol [Internet]. 2006 Dec [cited 2014 Mar 3]; 195: 1571–1577. Available from: http:// www.ncbi.nlm.nih.gov/pubmed/16769013

45 Chapter 2 Term breech deliveries in the Netherlands

16 McCowan, LM; Hardin, JE; Stewart A. Customized birthweight centiles predict SGA pregnancies with perinatal morbidity. BJOG [Internet]. 2005 [cited 2014 Mar 3]; 112: 1026–1033. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/? term=Customized+birthweight+centiles+predict+SGA+pregnancies+with+perinatal+ morbidity

17 Kok M, Gravendeel L, Opmeer BC, van der Post J a M, Mol BWJ. Expectant parents’ preferences for mode of delivery and trade-offs of outcomes for breech presentation. Patient Educ Couns [Internet]. 2008 Aug [cited 2014 Mar 3];72 : 305–310. Avail- able from: http://www.ncbi.nlm.nih.gov/pubmed/18534810

18 Tita ATN, Lai Y, Landon MB, Spong CY, Leveno KJ, Varner MW, et al. Timing of Elective Repeat Cesarean Delivery at Term and Maternal Perioperative Outcomes. Obs Gyne- col. 2011; 117: 280–286.

19 Glavind J, Kindberg SF, Uldbjerg N, Khalil M, Møller a M, Mortensen BB, et al. Elective caesarean section at 38 weeks versus 39 weeks: neonatal and maternal outcomes in a randomised controlled trial. BJOG [Internet]. 2013 Aug [cited 2014 Feb 1]; 120: 1123–1132. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23682628

20 Schutte JM, Steegers E a P, Santema JG, Schuitemaker NWE, van Roosmalen J. Mater- nal deaths after elective cesarean section for breech presentation in the Netherlands. Acta Obstet Gynecol Scand [Internet]. 2007 Jan [cited 2014 Mar 3]; 86: 240–243. Available from: http://www.ncbi.nlm.nih.gov/pubmed/17364290

21 Cunningham FG, Bangdiwala SI, Brown SS, Dean TM, Frederiksen M, Rowland Hogue CJ, et al. NIH consensus development conference draft statement on vaginal birth after cesarean: new insights. NIH Consens State Sci Statements [Internet]. 2010 Mar 10; 27: 1–42. Available from: http://www.ncbi.nlm.nih.gov/pubmed/20869552

22 Ea A, Ad O, Herrin J, Ge V, Terrenato I, Sperati F, et al. Using alternative statistical- for mats for presenting risks and risk reductions ( Review ). Cochrane Database Syst Rev. 2011; 16: CD006776.

23 Connor AO, Jacobsen MJ. Workbook on developing and evaluating patient decision aids [Internet]. 2003. Available from: https://decisionaid.ohri.ca/docs/develop/De- velop_DA.pdf

3 Subsequent pregnancy outcome after breech delivery, a population based cohort study

Floortje Vlemmix

Brenda M. Kazemier

Ageeth N. Rosman

Michel Hof

Jelle M. Schaaf

Anita C.J. Ravelli

Johannes J. Duvekot

Joris A.M. van der Post

Ben Willem J. Mol

Marjolein Kok

Submitted

49 Chapter 3 Subsequent pregnancy outcome after breech delivery

ABSTRACT

Objective To investigate the effect of the mode of delivery for term breech presentation on perinatal and maternal outcome on both the index and the subsequent pregnancy.

Study design We analysed data from the Dutch medical birth registry from 1999 to 2007. Data from 15,558 primiparous women with a singleton term breech delivery were linked with data of their subsequent pregnancy. We compared a strategy of elective cesarean versus a planned vaginal delivery for the first breech delivery. In a scenario analysis, we evaluated the effect of ‘once a cesarean always a cesarean’. The main outcome measures were perinatal and maternal morbidity and mortality.

Results Women who had an elective cesarean delivery in their index pregnancy had an increased risk of adverse neonatal outcome during the second pregnancy, compared to women who had a planned vaginal breech delivery (15.6 vs. 13.4‰, OR 1.3 (95% CI 1.0-1.7)) and an increased risk of adverse maternal outcome (59.9 vs. 43.0‰, OR 1.4, 95%CI 1.2-1.7). There was no significant reduction in perinatal mortality if outcomes of first and second pregnancies were combined. Neonatal morbidity remained significantly reduced (OR 0.47 (95% CI 0.39 – 0.58)). The scenario analysis decreased absolute risks of adverse outcomes, but still not led to a significant reduction in perinatal mortality.

Conclusions Elective cesarean delivery for term breech presentation leads to a significant increase in maternal morbidity and adverse neonatal outcome in subsequent pregnancies. Combining outcomes of the first and the subsequent pregnancy only resulted in a reduction in neonatal morbidity.

50 Chapter 3 Subsequent pregnancy outcome after breech delivery

INTRODUCTION

Breech presentation occurs in 3-4% of term pregnancies and vaginal breech delivery is associated with increased neonatal mortality and morbidity.1–3 Since publication of the results of the Term Breech Trial, breech presentation is a widely accepted indication for elective cesarean delivery.1,4,5 In the Term Breech Trial, Hannah et al. reported a decrease in combined adverse perinatal outcome in case of elective cesarean delivery compared to trial of labor (1.6 vs. 5.0%, RR 0.33; 95%CI 0.19-0.56). This effect was even stronger in countries with a low perinatal mortality rate (0.04 vs. 5.7%, RR 0.07; 95%CI 0.02-0.29). Maternal outcome was found to be comparable between the two groups, with a 43% emergency cesarean delivery rate in the planned vaginal delivery group.

After publication of the Term Breech Trial, the overall cesarean delivery rate in the Netherlands increased within two months from 50% to 80%, but remained stable thereafter, and is still much lower when compared to for instance the United Kingdom (90%), Denmark (94%) or Australia (96%).6–8 Ever since the publication of the Term Breech Trial, there has been discussion whether the gain in neonatal outcome by performing primary cesarean delivery in case of breech presentation, outweighs the neonatal and maternal risks when including outcome of subsequent pregnancies.9–11

Increasing cesarean delivery rates are a major topic of debate in obstetrics all over the world. In 1985, the World Health Organization set the recommended level at 15%.12 Nowadays, many countries face much higher cesarean delivery rates, for example in China (46%), Argentina (36%), USA (33%), Brazil (30%), Canada (24%) and United Kingdom (23%).4,8,13,14 In the Neth- erlands the cesarean delivery rate was 16% in 2009.5 Complication rates of a cesarean delivery reported in literature vary from 10 to 30% including mild complications like cystitis, and severe complications like bladder inju- ry, haemorrhage, thrombo-embolism and even maternal mortality.15-17 The risks of maternal or fetal mortality during a trial of labor after cesarean delivery are 0.04‰ and 1.3‰ respectively, and after elective repeat cesarean delivery 0.13‰ and 0.5‰.4,5 As a result repeat cesarean delivery seems to decrease fetal mortality, at a cost of more maternal mortality.

51 Chapter 3 Subsequent pregnancy outcome after breech delivery

We investigated the effects of the mode of delivery for term breech presentation (elective cesarean delivery versus vaginal delivery) on the maternal and neonatal outcome in the subsequent pregnancy. Secondly, we compared the health of the family as a whole (first and second pregnancy together) after a strategy of elective cesarean delivery as compared to a planned vaginal delivery in the first pregnancy.

MATERIALS AND METHODS

Study population

This study was performed using data from a prospective national cohort registered in the Netherlands Perinatal Registry (PRN). The PRN consists of population-based data containing information on pregnancies, deliveries and (re)admissions until 28 days after birth. The PRN database is obtained by a validated linkage of three different registries: the midwifery registry (LVR 1), the obstetrics registry (LVR 2) and the neonatology registry (LNR) of hospital admissions of new-borns.18 The coverage of the PRN registry is about 96% of all deliveries in the Netherlands. All PRN data are voluntarily recorded by the caregivers during prenatal care, delivery and the neonatal period. Data on neonatal admissions are registered in 99 of the 109 (90%) hospitals in the Netherlands. Data on neonatal follow up to 28 days after birth are provided by 49% (46/94) of neonatal care units in the Netherlands.5 The data are annually sent to the national registry office, where a number of range and -con sistency checks are conducted. For this study all births between January 1st 1999 and December 31st 2007 were selected.

A longitudinal probabilistic linkage procedure was developed to connect the first and subsequent pregnancies of the same mother in the PRN registry, and has been described in detail previously.19 The linkage procedure was necessary as the records included in the PRN were entered at the child’s level and there was no unique maternal identifier available in the registry to follow up on outcomes of subsequent pregnancies of the same mother. The longitudinal linkage was based on the variables birth date of mother, birth date of (previous) child, and zip code of mother. All children from second deliveries (n = 509, 559) registered in the PRN registry were subjected to linkage with their siblings born during a first delivery (nulliparous woman) registered in

52 Chapter 3 Subsequent pregnancy outcome after breech delivery the PRN registry for all births between 1 January 1999 and 31 December 2007. The final linked cohort with complete data on first and second deliveries of the same mother consisted of 272,551 women who had 545,102 (2 times 272,551) deliveries.19,20 From the longitudinal linked cohort, all first pregnancies with singleton breech presentations born between 37 0/7 and 41 6/7 weeks of gestation were selected. Exclusion criteria for the first pregnancies were antepartum death, death caused by placental abruption, and major congenital malformations. Major congenital malformations were defined as abnormalities of the central nervous system (spina bifida, meningo- myelocele, exencephaly, anencephaly, hydrocephaly and microcephaly) and cardiovascular and gastrointestinal system.

Main outcome measures

Outcomes of the first and second pregnancies were compared based on the mode of delivery in the index pregnancy, distinguishing between planned vaginal and planned cesarean breech delivery. Planned vaginal delivery was defined as initially intended vaginal delivery and included true vaginal delivery and emergency cesarean delivery (due to failure to progress or fetal distress).

Perinatal mortality was defined as perinatal and neonatal mortality up to 28 days. Neonatal morbidity was defined as an Apgar score of less than 7 at five minutes, and birth trauma (which included intracerebral bleeding, cephalic hematoma, facial nerve paresis, brachial plexus lesion, fracture of clavicle, humerus or femur and other trauma). Adverse neonatal outcome was defined as a composite measure of mortality and morbidity.

Maternal mortality was defined as mortality up to 28 days. Maternal morbidity was defined as uterine rupture, placental abruption and postpartum haemorrhage (PPH, defined as > 1000 mL, according to Dutch guidelines).21 Adverse maternal outcome was defined as a composite measure of maternal mortality and morbidity.

Statistical analysis

A comparison was made between the outcomes of the second pregnancy, distinguishing between a planned vaginal breech delivery and an elective cesarean delivery in the first pregnancy. Odds ratios and their 95% confidence intervals were calculated, which were adjusted for presentation at

53 Chapter 3 Subsequent pregnancy outcome after breech delivery

birth and trial of labor. We did a subgroup analysis to assess the effect of the mode of delivery on neonatal outcome if all women with a trial of labor in the second pregnancy were excluded. We did this to mimic a scenario of elective cesarean deliveries in case of a previous cesarean delivery. For this subgroup analyses, we excluded the antenatal mortality as the occurrence of a stillbirth influences the chosen mode of delivery (vaginal delivery) and inclusion of stillbirth would thus cause a selection bias. We used an interaction term to test the effect of mode of delivery on perinatal mortality and composite neonatal outcome in the group with trial of labor and repeat cesarean delivery.

To assess the combined outcome of index and subsequent pregnancies we calculated the average maternal and neonatal composite outcome scores over the two pregnancies.

Furthermore, we calculated the number needed to treat (NNT) based on the amount of primary cesarean deliveries are needed to prevent one adverse outcome, and the number needed to harm (NNH), the amount of elective cesarean deliveries that lead to one extra adverse outcome.

All statistical analyses were carried out with SAS 9.2 (SAS Institute, Cary, NC). Permission for record use and analysis of data for the purpose of this study was obtained from the Netherlands Perinatal Registry (registered as data pe- tition 11.44).

RESULTS

Of the 272,551 women in the linked database, 15,558 (5.7%) women had a healthy term singleton fetus in breech presentation during the first pregnancy. Mode of delivery and maternal and neonatal outcome of the first and second pregnancy are presented in Figure I.

In the first pregnancy, 8,116 (52%) women had an elective cesarean delivery for breech presentation. Of the remaining 7,442 (48%) women with a planned vaginal breech delivery, 3,287 (44%) delivered vaginally while 4,155 (56%) had an emergency cesarean delivery.

54 Chapter 3 Subsequent pregnancy outcome after breech delivery

Figure 1. mode of delivery and neonatal outcome of the first and second pregnancy

55 Chapter 3 Subsequent pregnancy outcome after breech delivery

Perinatal outcome in the second pregnancy

Perinatal mortality up to 28 days occurred twice as often in the second pregnancy if the first delivery was an elective cesarean delivery compared to a planned vaginal delivery but this did not reach statistical significance (2.5 vs. 1.3‰; aOR 2.1 (95% CI 0.96 – 4.5)) (Table 1).

A five-minute Apgar score below 7 occurred more often in the second pregnancy if women had an elective cesarean delivery in their first pregnancy compared to a planned vaginal delivery in the first pregnancy (13.8 vs. 11.4‰; aOR 1.5 (95% CI 1.1 – 1.9)), while birth trauma did not differ significantly between both groups (2.1 vs. 1.6‰; aOR 1.4 (95% CI 0.6– 2.8). The adverse neonatal outcome in the second pregnancy was higher in the group of women with an elective cesarean delivery compared to women with a planned vaginal delivery in their first pregnancy (15.6 vs. 13.4‰; aOR 1.3 (95%CI 1.0-1.7)).

Table 1. Outcome of second pregnancies of women with term breech delivery in their first pregnancy

Elective CD Planned vaginal delivery Adjusted OR 95% CI in first pregnancy in first pregnancy b N=8,116 (‰) N= 7,442 (‰) Birth trauma 17 (2.1) 12 (1.6) 1.4 0.65-‐2.8 Five-‐minute Apgar score <7 112 (13.8) 84 (11.3) 1.5 1.1-‐1.9 Perinatal mortality c 20 (2.5) 10 (1.3) 2.1 0.96-‐4.5 Adverse neonatal outcome d 127 (15.6) 99 (13.4) 1.3 1.0-‐1.7

Uterine rupture 18 (2.2) 4(0.5) 4.9 1.6-‐14.7 Placental abruption 7 (0.9) 4 (0.5) 1.3 0.35-‐4.4 PPH (>1000ml) 464 (57.2) 312 (41.9) 1.4 1.2-‐1.6 Maternal mortality 0 0 -‐ -‐ Adverse maternal outcome e 486 (59.9) 320 (43.0) 1.4 1.2-‐1.7 a Vaginal breech birth and emergency caesarean delivery (CD). b Planned vaginal delivery versus elective CD, adjusted for presentation at birth and trial of labour c Perinatal and neonatal mortality until 28 days postpartum d Composite score of perinatal and neonatal mortality up to 28 days, five-‐minute Apgar score 7, < birth trauma. e Composite score of maternal mortality, PPH>1000 ml, placental abruption and uterine rupture

56 Chapter 3 Subsequent pregnancy outcome after breech delivery

Intended mode of delivery of the subsequent pregnancy (trial of labor versus repeat elective cesarean delivery) was a significant effect modifier in the association of mode of delivery in the index pregnancy and neonatal outcome in the second pregnancy (interaction terms were highly significant for mortality (p<0.001) and the composite outcome (p=0.01)). Thus a scenario analysis excluding women with a trial of labor after previous cesarean delivery was justified (Table 2).

In this scenario analysis, the perinatal mortality rates in the second pregnancy were lower in the women who had an elective cesarean delivery compared to a planned vaginal delivery in their first pregnancy, but this was not statistically significant (1.4 vs. 1.6‰; aOR 0.5 (95% CI 0.1 – 2.8)). In Table 1, there are 20 cases of perinatal mortality reported in the elective cesarean group.

After exclusion of all trial of labours, this is reduced to 2 cases (Table 2). Thus the perinatal mortality rate was 2.6‰ in case of a trial of labor after elective cesarean delivery ((20-2) /(8116-1427)), which is similar to the mortality rate in the first pregnancy of the planned vaginal breech delivery group (Table 3).

Table 2. Outcome of second pregnancies of women with term breech delivery in their first pregnancy, after exclusion of women with trial of labour in the second pregnancy after caesarean a delivery in the first pregnancy.

Elective CD Planned vaginal Adjusted 95% CI in first pregnancy delivery ORb N=1,427 (‰) in first pregnancya Birth trauma 2 (1.7) N= 2 3,820 (0.5)(‰) 3.3 0.42-‐25.9 Five-‐minute Apgar score <7 7 (4.9) 33 (8.6) 0.41 0.17-‐0.97 Perinatal mortalityc 2 (1.4) 6 (1.6) 0.52 0.10-‐2.8 Adverse neonatal outcomed 10 (7.0) 39 (10.2) 0.51 0.25-‐1.1

Uterine rupture 1 (0.7) 0 -‐ -‐ Placental abruption 2 (1.7) 2 (0.5) 3.3 0.42-‐25.9 PPH (>1000ml) 44 (30.8) 107 (28.0) 1.04 0.59-‐1.8 Maternal mortality 0 0 -‐ -‐ Adverse maternal outcomee 46 (32.2) 109 (28.5) 1.3 0.90-‐1.9 avaginal breech birth and emergency caesarean delivery (CD). bPlanned vaginal delivery versus elective CD, adjusted for presentation at birth cPerinatal and neonatal mortality until -‐ 28 days post partum dComposite score of neonatal mortality up to 28 days, five-‐minute Apgar score <7 and birth trauma eComposite score of maternal mortality, PPH>1000 ml, placental abruption and uterine rupture

1 Chapter 3 Subsequent pregnancy outcome after breech delivery

A 5-minute Apgar score <7 occurred less frequent in the second pregnancy in the elective cesarean delivery group compared to the vaginal delivery group after exclusion of trial of labor after cesarean delivery, but this did not reach statistical significance (4.9 vs. 8.6‰; aOR 0.4 (95% CI 0.2 – 1.0)). The combined outcome score was also not significantly different (7.0 vs. 7.9‰; aOR 0.5 (95% CI 0.2 – 1.0)). Table 3: Neonatal and maternal outcomes of first and second pregnancy in nulliparous women with a breech presentation in the first pregnancy. NNT is the number of extra caesarean deliveries (CD) needed to perform

a Elective CD Planned vaginal OR (95% CI) NNT / NNH

first pregnancy delivery primary CS (95%CI)

(average risk first pregnancy

per child) (average risk per child)

N(‰) N(‰) Outcome of first pregnancy Number of birth 8116 7442 Perinatal mortality 0 19 (2.6) -‐ 392 (251-‐612) Neonatal morbidity 30 (3.7) 196 (26.3) 0.16 (0.10 – 45 (38-‐53)

Maternal morbidity 111 (13.7) 147 (19.8) 0.69 0.24) (0.53 -‐ 165 (98-‐485)

0.89)

Outcome of first and second pregnancy (of which 66.3% underwent trial of labour) Number of birth 16232 14884 Perinatal mortality 20 (1.2) 29 (1.9) 0.63 (0.34 – 1.2) 1397 Neonatal morbidity 157 (9.7) 301 (20.2) 0.47 (0.39 – (5874†95 (76-‐-‐127)597)

Maternal morbidity 597 (36.8) 467 (31.4) 1.2 0.58) (1.0-‐1.3) 186†

(106†-‐732†)

Outcome of first and second pregnancy, exclusion of trial of labour Number of birth 2854 7640 Perinatal mortality 2 (0.7) 15 (2.0) 0.36 (0.04 -‐ 1.5) 793 (1454†-‐374) Neonatal morbidity 20 (7.0) 193 (25.3) 0.27 (0.16 -‐ 55 (44-‐75)

Maternal morbidity 82 (28.7) 213 (27.9) 1.03 0.43) (0.79 – 1.3) 1174†

(119†-‐169) †number needed to harm perinatal and neonatal mortality up to 28 days in first and second pregnancy

58 Chapter 3 Subsequent pregnancy outcome after breech delivery

Maternal outcome in second pregnancy

No case of maternal mortality was reported in the 15,558 subsequent pregnancies. Uterine rupture occurred five times more often in women with a previous elective cesarean delivery compared to those with a planned vaginal delivery (2.2 vs. 0.5‰; aOR 4.9 (95% CI 1.6-14.7)) (Table 1). The risk of uterine rupture among women with a previous elective, versus emergency cesarean delivery during planned vaginal breech delivery was doubled, although this did not reach statistical significance (2.2 vs. 1.0‰; aOR 2.3 (95% CI 0.8-6.8)). Placental abruption differed not statistically significant between these two groups (0.9 vs. 0.7‰; aOR 1.2 (95% CI 0.3-4.6)). Postpartum haemorrhage (PPH) occurred more often in women with a previous elective cesarean delivery compared to previous vaginal delivery (57.2 vs. 41.9‰; aOR 1.4 (95% CI 1.2-1.6)). The combined poor maternal outcome occurred significantly more often in women with a previous elective cesarean delivery than previous planned vaginal delivery (59.9 vs. 42.9‰; aOR 1.4 (95% CI 1.2- 1.7)) (Table 1).

Table 2 shows the result of the scenario analysis, in which we excluded all women with a trial of labor in the second pregnancy after an elective or emergency cesarean delivery in the first pregnancy. The number of- uter ine ruptures decreased from 18 to 1 in the elective cesarean delivery group (from 2.2 to 0.7‰; aOR 0.3 (0.03-2.0)); and from 4 to none in the women in the vaginal group who underwent an emergency cesarean delivery in their first pregnancy. There was no significant difference in the combined maternal outcome of the second deliveries between the elective cesarean delivery group (both pregnancies) versus the vaginal delivery group (32.2 vs. 28.5‰; aOR 1.3 (95% CI 0.9-1.9)).

Combined outcomes

Table 3 summarizes the combined maternal and neonatal outcomes of the index deliveries and subsequent pregnancies. When only looking at the index pregnancy, the perinatal mortality rate decreased from 2.6 to 0‰ due to the elective cesarean delivery in case of breech presentation (NNT 392). The neonatal morbidity decreased from 26.3 to 3.7‰ (NNT 45)

When the neonatal outcomes of the subsequent pregnancies were taken into account the perinatal mortality rate was 1.2‰ in the elective cesarean

59 Chapter 3 Subsequent pregnancy outcome after breech delivery

delivery group and 1.9‰ in the planned vaginal delivery group, and no significant difference between the two modes of delivery was found (OR 0.63, 95%CI 0.34 – 1.2). The neonatal morbidity rate was 9.7 vs. 20.2‰ respectively (OR 0.47 (95% CI 0.39 – 0.58), NNT 95). The maternal morbidity was 36.8‰ for elective cesarean delivery vs. 31.4‰ for the planned vaginal breech delivery (OR 1.2 (95% CI 1.0-1.3), NNH 186) After exclusion of all women with a trial of labor in the second pregnancy after a cesarean delivery in the index pregnancy, the average risk of perinatal mortality over two pregnancies was 0.7‰ for planned cesarean delivery vs. 2.0‰ for planned vaginal delivery (OR 2.0 (95%CI 0.04-1.5)). The combined neonatal morbidity rate was 7.0 vs. 25.3‰ (OR 0.27 (95%CI 0.16-0.43), NNT 55). There was no significant difference in in maternal morbidity (28.7 vs. 27.9‰) (OR 1.03 (95%CI 0.79 – 1.3)).

DISCUSSION

To investigate the consequences of the mode of delivery for term breech presentation for future offspring, we studied 15,558 women in their second pregnancy who delivered a child in breech presentation in their first pregnancy. The NNT to prevent one case of perinatal morbidity is 95, but the gain in perinatal mortality and maternal morbidity by elective cesarean delivery for breech in the first pregnancy are lost due to an increased perinatal mortality and maternal morbidity rate in the second pregnancy. Exclusion of trial of labor after a previous cesarean section did lower the absolute risks at perinatal mortality, perinatal morbidity and maternal morbidity. However this still did not lead to a significant decrease in neonatal mortality in case of elective cesarean delivery for breech presentation.

Strengths and limitations of our study

Our study was based on data of a large population-based national perinatal registry. The majority of the caregivers contribute to the PRN registry and it therefore comprises approximately 96% of all pregnancy and birth characteristics in the Netherlands. The 4% missing birth data are due to 1–2% non- reporting general practitioners and 2–3% non-reporting midwives.

The fact that almost half of the women opted for planned vaginal breech delivery made it possible to evaluate the effect of a planned vaginal delivery vs. elective cesarean delivery policy. Linkage of index and sub sequent deliveries allowed us to compare maternal and neonatal outcome

60 Chapter 3 Subsequent pregnancy outcome after breech delivery of these pregnancies. When interpreting the NNT, it is important to realize that this is the number of primary cesarean deliveries needed to prevent one complication. The number of extra cesarean deliveries (women who would otherwise have had a vaginal delivery) is actually lower as 56% of the women in the planned vaginal delivery group ended up with an emergency cesarean deliveries.

In the longitudinal probabilistic linkage procedure, 53% of all second deliveries could be linked to a first delivery. The probabilistic linkage method, and limitations to follow up mothers in a subsequent pregnancy, has been discussed in detail by Schaaf et al.20 Non-linkage could be explained because a first child was born before 1999, (which data were not available) in approximately half of the non-linked cases. Other explanations were missing values in the two out of the three linkage variables. It is unlikely that this non-linkage will have affected our results to a large extent, as non-linkage is neither related to fetal presentation in nulliparous singleton delivery, nor to the primary outcome measures.

A limitation of the study is that we were not able to perform a complete intention to treat analysis for mode of delivery. Women who opted for elective cesarean, but underwent a vaginal delivery (for instance due to unexpected start of delivery and fast progression), could not be identified as such and included in the cesarean group. In the Term Breech Trial, 9.7% of women with an elective cesarean underwent a vaginal breech delivery. There also might be ´misclassification´ of emergency cesarean delivery in the database as some doctors have recorded for instance an elective cesarean delivery after PROM or onset of labor as an emergency cesarean delivery. This misclassification might have led to an underestimation of the adverse outcome of planned vaginal delivery, as a small proportion of low risk elective cesarean was added to this group.

A disadvantage of the increased cesarean rates is the increased maternal morbidity and mortality. There was no maternal mortality reported in our study. Among all women (multiparous and nulliparous women) with a breech presentation between 2000 and 2007, two maternal death were reported. Due to the selection of women with a subsequent pregnancy, these were not included in this cohort. A previous study reported four maternal deaths after elective cesarean delivery for term singleton breech delivery between 2000-2002.22 These study results suggest that there is a potential underre-

61 Chapter 3 Subsequent pregnancy outcome after breech delivery

porting of maternal complications and mortality in the Netherlands Perinatal Registry. Therefore, our results on maternal outcome should be interpreted with caution. In literature, maternal mortality rates of 0.04‰ and 0.13‰ are reported for elective and repeat elective cesarean delivery respectively.4,22 The overall risk of 0.17‰ corresponds to one maternal mortality in 5,882 women with two elective caesareans. Our study is to our knowledge the largest study to evaluate the effect of elective cesarean delivery for breech presentation on subsequent pregnancies. The perinatal outcome of the first breech presenting pregnancy, are in line with results of the Term Breech Trial which showed a significant decrease in combined perinatal outcome in case of elective cesarean delivery compared to trial of labor (1.6% vs. 5.0, RR 0.33; 95%CI 0.19-0.56). A large cohort study on breech presentations in Denmark reported a decrease in mortality rate from 1.7‰ in the planned vaginal delivery group to 0.1‰ in the elective cesarean delivery group (RR 19.0 (95%CI 2.3-154.5).1,6 Landon et al. reported on 17,898 women with trial of labor and 15,801 with repeat cesarean delivery and found lower incidences of uterine rupture of 0.7% and 0%, respectively.23 Consequently, the antenatal and neonatal mortality rates were lower than in our study. We cannot explain this risk difference among the cohorts by specific patient characteristics. The management in the United Stated of America of women undergoing a trial of labor is apparently more risk averse than in The Netherlands.

Meaning of the study

Our data facilitate decision making on mode of delivery for women with a singleton breech presentation at term, as they enable quantification of the risks for future pregnancies. Counselling on the mode of breech delivery should indeed start with inventory of a woman’s family plans, followed by shared decision making on mode of delivery. Although neonatal and maternal outcome of both pregnancies are insignificant regarding mode of delivery in the first breech presenting pregnancy, this does not insinuate that the elective cesarean for breech presentation has lost its indication. The high risks of perinatal mortality during a vaginal breech delivery is spread out over two pregnancies and therefor lost its significance in the overall outcome. There is no significant increase of perinatal mortality in the second pregnancy due to the scarred uterus, which is in line with other studies on trial of labor after cesarean section.24–26 Nulliparous women with a breech presentation and a desire for two children would have the lowest risk at adverse neonatal mor-

62 Chapter 3 Subsequent pregnancy outcome after breech delivery bidity with an elective cesarean delivery for breech presentation (NNT 95). The NNT is reduced to 55 if two elective cesarean sections are performed.

As maternal and fetal complications will increase with every repeat cesarean delivery, we feel that our results are not applicable to women who want a family of at least four children or more. Decision analytic modelling including the risk and consequences of subsequent uterine rupture and placenta accreta in future pregnancies should point out when the balance tips over to vaginal breech delivery due to an increase complication rate of repeat cesarean delivery. The current mean birth rate in the Netherlands is 1.76 per women, thus a large majority of women will be helped with the information presented in this study.

The necessity of preventing the first cesarean section to improve obstetric outcome is again demonstrated in this study; the initial gain in neonatal and maternal outcome is largely undone by the consequences of (a trial of labor with) a scarred uterus during the next pregnancy. External cephalic version is the only treatment proven effective to reduce the cesarean delivery rate for breech presentation at term.27 Counselling for external cephalic version is essential in order to truly improve maternal and neonatal outcome in the breech presenting and subsequent pregnancies.

Conclusions

In conclusion, elective cesarean delivery for term breech presentation leads to a significant increase in maternal morbidity, and adverse neonatal outcome in subsequent pregnancies. Combining outcomes of the first breech presenting pregnancy and the subsequent pregnancy only led to a reduction in neonatal morbidity. These data should be used in shared decision making for the mode of delivery in women with a child in breech presentation at term.

63 Chapter 3 Subsequent pregnancy outcome after breech delivery

REFERENCES

3 Hickok, DE; Gordon, DC; Milberg, JA; Williams, MA; Daling J. The frequency of breech presentation by gestational age at birth: a large population-based study. Am J Obs Gynecol [Internet]. 1992 [cited 2014 Mar 3]; 166: 851–852. Available from: https:// c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/1550152

4 Cunningham FG, Bangdiwala SI, Brown SS, Dean TM, Frederiksen M, Rowland Hogue CJ, et al. NIH consensus development conference draft statement on vaginal birth after cesarean: new insights. NIH Consens State Sci Statements [Internet]. 2010 Mar 10; 27: 1–42. Available from: http://www.ncbi.nlm.nih.gov/pubmed/20869552

5 Brouwers, HAA; Bruinse, HW; Dijs-elsinga, J; Van Huis, AM; De Miranda, E; Ravelli, ACJ; Tamminga, P; Schiere, AM; Giesbers H. Jaarboek zorg in Nederland 2009 (the Netherlands Perinatal Registry, yearly report 2009) [Internet]. 2013. Available from: http://www.perinatreg.nl/uploads/150/136/Jaarboek_Zorg_in_Nederland_2009_ Tabellen.pdf

6 Hartnack Tharin JE, Rasmussen S, Krebs L. Consequences of the Term Breech Trial in Denmark. Acta Obstet Gynecol Scand [Internet]. 2011 Jul [cited 2014 Feb 20]; 90: 767–771. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21476999

7 Sullivan EA, Moran K, Chapman M. Term breech singletons and caesarean section: a population study, Australia 1991-2005. Aust N Z J Obstet Gynaecol [Internet]. 2009 Oct [cited 2014 Mar 3]; 49: 456–460. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/19780725

8 Bragg F, Cromwell DA, Edozien LC, Gurol-urganci I, Mahmood TA, Templeton A, et al. Variation in rates of caesarean section among English NHS trusts after accounting for maternal and clinical risk : cross sectional study. BMJ. 2010;341 .

64 Chapter 3 Subsequent pregnancy outcome after breech delivery

9 Biswas A. Term breech trial. Lancet [Internet]. 2001 Jan 20; 357: 225; author reply 227–8. Available from: http://www.ncbi.nlm.nih.gov/pubmed/11213114

10 Erskine J. Term Breech Trial. Lancet. 2001; 357: 228.

11 Bewley, S; Shennan A. Peer review and the Term Breech Trial. Lancet. 2007; 369: 2007.

12 WHO. Appropriate technology for birth. Lancet. 1985; 436–437.

13 Souza JP, Gülmezoglu A, Lumbiganon P, Laopaiboon M, Carroli G, Fawole B, et al. Caesarean section without medical indications is associated with an increased risk of adverse short-term maternal outcomes: the 2004-2008 WHO Global Survey on Maternal and Perinatal Health. BMC Med [Internet]. 2010 Jan [cited 2014 Jan 28]; 8: 71. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=299 3644&tool=pmcentrez&rendertype=abstract

14 Chaillet N, Dumont A. Evidence-based strategies for reducing cesarean section rates: a meta-analysis. Birth [Internet]. 2007 Mar; 34: 53–64. Available from: http://www. ncbi.nlm.nih.gov/pubmed/17324180

15 Van Ham M a, van Dongen PW, Mulder J. Maternal consequences of caesarean section. A retrospective study of intra-operative and postoperative maternal complications of caesarean section during a 10-year period. Eur J Obstet Gynecol Re- prod Biol [Internet]. 1997 Jul; 74: 1–6. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/9243191

16 Rasmussen, SA; Maltau J. Complications following cesarean section. Tidsskr Nor Lae- geforen [Internet]. 1990 [cited 2014 Mar 3]; 110: 351–353. Available from: https:// c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/2309178

17 Nielsen TF, Hökegård KH, Ericson a. Cesarean section and perinatal mortality in Swe- den in 1981. Acta Obstet Gynecol Scand [Internet]. 1986 Jan; 65: 865–867. Available from: http://www.ncbi.nlm.nih.gov/pubmed/3825528

18 Méray N, Reitsma JB, Ravelli ACJ, Bonsel GJ. Probabilistic record linkage is a valid and transparent tool to combine databases without a patient identification number. J Clin Epidemiol [Internet]. 2007 Sep [cited 2014 Jan 21]; 60: 883–891. Available from: http://www.ncbi.nlm.nih.gov/pubmed/17689804

19 Schaaf JM, Hof MHP, Mol BWJ, Abu-Hanna A, Ravelli ACJ. Recurrence risk of preterm

65 Chapter 3 Subsequent pregnancy outcome after breech delivery

birth in subsequent singleton pregnancy after preterm twin delivery. Am J Obstet Gy- necol [Internet]. Elsevier Inc.; 2012 Oct [cited 2014 Mar 3]; 207: 279.e1–7. Available from: http://www.ncbi.nlm.nih.gov/pubmed/22917487

20 Schaaf JM, Hof MHP, Mol BWJ, Abu-Hanna a, Ravelli a CJ. Recurrence risk of preterm birth in subsequent twin pregnancy after preterm singleton delivery. BJOG [Internet]. 2012 Dec [cited 2014 Mar 3]; 119: 1624–1629. Available from: http://www.ncbi.nlm. nih.gov/pubmed/23078576

21 Otterlo Werkgroep dr. HCJS. NVOG richtlijn - Hemorrhagia postpartum (HPP) (Guide- line on post partum hemorrhage- the Dutch Society of Obstetrics and Gynecology) [Internet]. 2013. Available from: http://nvog-documenten.nl/uploaded/docs/NVOG richtlijn HPP 14-11-2013 definitief d.pdf

22 Schutte JM, Steegers E a P, Santema JG, Schuitemaker NWE, van Roosmalen J. Mater- nal deaths after elective cesarean section for breech presentation in the Netherlands. Acta Obstet Gynecol Scand [Internet]. 2007 Jan [cited 2014 Mar 3]; 86: 240–243. Available from: http://www.ncbi.nlm.nih.gov/pubmed/17364290

23 Tita ATN, Lai Y, Landon MB, Spong CY, Leveno KJ, Varner MW, et al. Timing of Elective Repeat Cesarean Delivery at Term and Maternal Perioperative Outcomes. Obs Gyne- col. 2011; 117: 280–286.

24 Gilbert S a, Grobman W a, Landon MB, Spong CY, Rouse DJ, Leveno KJ, et al. Elec- tive repeat cesarean delivery compared with spontaneous trial of labor after a prior cesarean delivery: a propensity score analysis. Am J Obstet Gynecol [Internet]. Else- vier Inc.; 2012 Apr [cited 2014 Apr 19]; 206: 311.e1–9. Available from: http://www. pubmedcentral.nih.gov/articlerender.fcgi?artid=3337034&tool=pmcentrez&rendert ype=abstract

25 Studsgaard A, Skorstengaard M, Glavind J, Hvidman L, Uldbjerg N. Trial of labor compared to repeat cesarean section in women with no other risk factors than a prior cesarean delivery. Acta Obstet Gynecol Scand [Internet]. 2013 Nov [cited 2014 Apr 19]; 92: 1256–1263. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23962339

26 Fagerberg MC, Marsal K, Källen K. Neonatal outcome after trial of labor or elective cesarean section in relation to the indication for the previous cesarean delivery. Acta Obstet Gynecol Scand [Internet]. 2013 Oct [cited 2014 Apr 19]; 92: 1151–1158. Avail- able from: http://www.ncbi.nlm.nih.gov/pubmed/23782390

27 Hofmeyr, GJ; Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012; 10. 66

4 Contraindications for external cephalic version in breech position at term: a systematic review

Ageeth N. Rosman

Aline Guijt

Floortje Vlemmix

Marlies Rijnders

Ben Willem J. Mol

Marjolein Kok

Acta Obstet Gynecol Scand. 2013 Feb;92(2):137-42

69 Chapter 4 Contraindications for external cephalic version in breech position at term

ABSTRACT

Objective External cephalic version (ECV) is a safe and effective intervention that can prevent breech delivery, thus reducing the need for cesarean delivery. It is recommended in national guidelines. These guidelines also mention contraindications for ECV, and thereby restrict the application of ECV. We assessed whether the formulation of these contraindications in guidelines are based on empiric data

Design Systematic review.

Population Pregnant women with a singleton breech presentation from 34 weeks.

Methods We searched the National Guideline Clearinghouse, the Cochrane Central Register of Controlled Trials, MEDLINE (1953-2009), EMBASE (1980- 2009), TRIP database (until 2011), NHS (National Health Services, until 2011), Diseases database (until 2011) and NICE guidelines (until 2011) for existing guidelines on ECV and studied the reproducibility of the contraindications stated in the guidelines. Furthermore, we systematically reviewed the literature for contraindications and evidence on these contraindications.

Main outcome measures Contraindications of ECV.

Results We found five guidelines mentioning 18 contraindications, varying from five to 13 per guideline. The contraindications were not reproducible between the guidelines with oligohydramnios as the only contraindication mentioned in all guidelines. The literature search yielded 60 studies reporting on 39 different contraindications, of which we could only assess evidence of six of them.

Conclusions The present study shows that there is no general consensus on the eligibility of patients for ECV. Therefore we propose to limit contraindications for ECV to clear empirical evidence or to those with a clear pathophysi- ological relevance.

70 Chapter 4 Contraindications for external cephalic version in breech position at term

INTRODUCTION

Breech position of a singleton fetus at term occurs in 3–4% of all term pregnancies, and it imposes an increased risk during delivery for mother and child. Nowadays, about 80–96% of all women with a fetus in breech position opt for a planned cesarean section (CS) as this is the safest way to deliver their child.1–3

External cephalic version (ECV) at or near term is a safe procedure that effectively reduces the risk of a CS in pregnancies with breech presentation.4,5 International guidelines recommend that all women with an uncomplicated breech pregnancy at term should be offered an ECV.6–10 These guidelines also mention contraindications for ECV, restricting the application. Contra- indications for ECV are often based on the safety of the procedure, but also on the fact that the probability of success of ECV in the presence of such contraindications is low.

Prediction models on ECV success rates show us that the outcome is multifactorial and that the presence of one negative prognostic factor should not be a reason to withhold ECV.11 Therefore, excluding patients from ECV in the presence of only one clinical factor that could affect its success rate is not desirable. ECV contraindications should therefore be limited to factors affecting the safety of the procedure.

Considering the fact that ECV is an important obstetric intervention to prevent breech presentation there should be consensus on women considered eligible for the procedure. Therefore, the primary goal of this study was to systematically assess whether the formulation of contraindications for ECV in guidelines is consistent and based on empirical data. Hence, we assessed the reproducibility of the contraindications for the different guidelines and we systematically reviewed the literature on contraindications and the evidence for each contraindication. A secondary goal is to make recommendations for ECV contraindications based on this systematic review.

71 Chapter 4 Contraindications for external cephalic version in breech position at term

MATERIAL AND METHODS

Guidelines

To identify all guidelines, we searched the National Guideline Clearing- house, the Cochrane Central Register of Controlled Trials, MEDLINE (1953– 2009), EMBASE (1980–2009), TRIP database (until 2011), NHS (National Health Services) (until 2011), Diseases database (until 2011) and NICE (until 2011) using a search strategy with keywords: ‘version fetal’, ‘cephalic version’, ‘external version’ and ‘contraindications’.

Literature

We searched the databases UpToDate, PubMed, The Cochrane Central Reg- ister of Controlled Trials, MEDLINE (1953–2011), EMBASE (1980–2011) and the Web of Science (until 2011) for studies reporting on ECV and cited contraindications. Furthermore, we analysed the evidence for each contraindication. References from identified publications were manually searched to identify additional relevant articles. No time or language restrictions were applied. In this review, all studies reporting on ECV in the case of a singleton pregnancy from 36 weeks without a contraindication for vaginal delivery were selected. We included guidelines, case reports, randomized controlled trials and case-control studies.

Studies were selected in a two-stage process. First, two reviewers (A.G., A.R.) scrutinized the titles and abstracts of all references possibly reporting on contraindications for ECV. For all studies that were selected by at least one of the reviewers, full manuscripts were obtained. Secondly, final in-/exclusion decisions were made after independent and duplicate examination of the full manuscripts of selected references by two reviewers. For each included article, data on clinical and methodological study characteristics were extracted independently by two reviewers on piloted data-extraction forms. Any disagreements were resolved by consensus and, if necessary, by a third reviewer (M.K.).

From the included studies reporting on ECV, we extracted a list of all contraindications mentioned in the studies. We classified the contraindications in three predetermined categories: maternal factors, fetal factors and other factors. Maternal factors were defined as contraindications that can affect the safety of the mother or affect the procedure of ECV by maternal charac-

72 Chapter 4 Contraindications for external cephalic version in breech position at term teristics. Fetal factors were defined as contraindications that can affect the safety of the fetus or affect the procedure of ECV by fetal characteristics.

Table 1. Overview of the potential contraindications for external cephalic version in the five international guidelines.

Contraindications Guidelines 1 2 3 4 5 NVOG KNOV ACOG RCOG RANZCOG 1. Oligohydramnios + + + + + 2. Fetal growth restriction − + + + + 3. Uterine anomaly + − + + + 4. Ruptured membranes + − + + − 5. Abnormal CTG + − − + + 6. Preeclampsia, hypertension − + + + − 7. Fetal anomaly − − + + + 8. Antepartum bleeding − − + + + 9. History of placental abruption + + − − − 10. Second trimester bleeding − + − − + 11. Hyperextension fetal head − − + − − 12. Maternal cardiac disease − − + − − 13. Macrosomia >4000 g − − + − − 14. Maternal obesity − − + − − 15. Two or more CS in history − − + − − 16. Active labor − − + − − 17. Unstable lie − − − + − 18. Restrictive nuchal cord − − − − + No. of contraindications per guideline 5 5 13 9 8

1NVOG Dutch Society of Obstetrics and Gynecology. 2KNOV Royal Dutch Organization of Midwives. 3ACOG American College of Obstetricians and Gynecologists. 4RCOG Royal College of Obstetricians and Gynaecologists. 5RANZCOG Royal Australian New Zealand College of Obstetricians and Gynaecologists

73 Chapter 4 Contraindications for external cephalic version in breech position at term

RESULTS

Guidelines

The electronic search yielded three international guidelines on ECV.7–9 Fur- thermore, there are two Dutch guidelines on the topic.6,10 These five guidelines reported on 18 different contraindications varying from five to 13 contraindications per guideline (Table 1).

Literature

We detected 890 studies on ECV, of which 261 were retrieved for complete assessment after reading the title or abstract. Of the 261 primary articles, 201 were excluded after further research: 199 articles because no contraindications were mentioned and two articles which were unobtainable. The 60 articles that were finally included in the systematic review reported on 39 different contraindications for ECV. Forty-three studies (72%) were designed as a cohort study, six (10%) as case-control studies and 11 (18%) as randomized controlled trials. Table 2 shows all 39 contraindicaticontraindications ons cat-categorized by the three predetermined categories: maternal factors, fetal factors and other factors. For each contraindication we performed a literature search to determine whether it was an evidence-based contraindication.

Table 2. Overview of all contraindications for external cephalic version reported in guidelines and literature.

Maternal factors Fetal factors Other factors Guidelines Uterine anomaly 4 ( ) Fetal growth restriction (4) Oligohydramnios 5 ( ) (no. of guidelines) Preeclampsia/ hypertension 3 ( ) Abnormal CTG (3) Restrictive nuchal cord (1) Ruptured membranes (3) Fetal anomaly 3 ( ) Antepartum bleeding (3) Macrosomia >4000 g (1) Second trimester bleeding (2) Hyperextension fetal head (1) Abruptio placentae in history 2 ( ) Unstable lie (1) Active labor (1) Maternal cardiac disease (1) Maternal obesity 1 ( ) Literature Maternal problems Cephalopathy Inexperienced obstetrician ECG abnormalities Doppler abnormalities Anterior placenta Abnormal pelvic Fetal distress Polyhydramnios Age >45 years Positive non stress -‐test Single umbilical artery Allergy Rhesus immunization CS in history Uteroplacental transfusion Diabetic Dilated cervix Grand multipara Hyperthyroidism Irregular thyroxin

CS, cesarean section; cardiotocography; CTG, ECG, electrocardiogram.

74 Chapter 4 Contraindications for external cephalic version in breech position at term

Maternal factors

We found two cohort studies and one review reporting on ECV in women with a previous CS.12–14 The review also included the cohort study of de Meeus et al. The review reported on 166 cases in which one emergency CS was performed with good neonatal outcome.12 The authors concluded that women with a breech-presenting fetus at term and a previous CS, who desire a trial of labor, should be counselled regarding the accumulating evidence about the efficacy and apparent safety of this procedure and possibly offered an ECV attempt. The prospective cohort study of ��Aben- haim et al.13 was published after the review of Sela et al.14 and did not report on complication rates, but only reported success rate in the 36 women with a scarred uterus (which was similar to success in women without a scarred uterus).

We found two case reports on ECV in the case of premature rupture of the membranes (PROM).15,16 The first study reported on one woman who underwent an ECV with ruptured membranes. The ECV was successful, but within two hours after the ECV the umbilical cord prolapsed. The other report concluded that the key indicators for ECV with ruptured membranes might be the absence of labor, normal amniotic fluid index by ultrasound, and a visually closed cervix at the time of the procedure.16

We could not find evidence for the other contraindications mentioned in this category; however, some main risks factors are based on clinical physi- ological knowledge and are therefore discussed. There were no studies reporting on the relationship between ECV and placental abruption; however, it is considered a rare complication of ECV.5 Considering the seriousness of the complication and the higher risk of recurrence in patients with a history of placental abruption, it seems justified to withhold ECV in patients- sus pected of an imminent abruption (for example vaginal blood loss) or with a history of abruption.

Fetal factors

We found one meta-analysis on the relationship between the success rate of ECV and estimated fetal weight.17 Five studies reporting on estimated fetal weight and ECV outcome were included. The weight categories reported on were too heterogeneous to pool results from. Although there were no

75 Chapter 4 Contraindications for external cephalic version in breech position at term

adverse events reported for small-for-gestational age foetuses, there was not enough evidence to conclude that ECV is safe in cases of severe growth restriction with Doppler abnormalities.

Other factors

We found three systematic reviews reporting on the relationship between success rate of ECV and oligohydramnios.17–19 The studies conclude that an amniotic fluid index (AFI) above 10 cm was associated with an increased chance of success, but did not report any adverse events of ECV in those women with low amniotic fluid index.

This is also the case with an anterior placenta. One systematic review reported the relationship between anterior placenta and a lower success rate of ECV, but in none of the included studies were complications reported in this group of women.

DISCUSSION

Our review of international guidelines showed that there was only a slight agreement on contraindications for ECV between the guidelines, oligohydramnios being the only contraindication mentioned in all guidelines. We summarized the available evidence on potential contraindications for ECV. Of the 39 contraindications for ECV mentioned in the guidelines and literature, we could only assess evidence for six. Five could be rejected based on evidence of systematic reviews (one previous CS, fetal growth restriction, macrosomia, oligohydramnios, polyhydramnios) and one was still debat- able due to a low level of evidence (case reports, premature rupture of the membranes). The extensive adherence to contraindications that were not well defined, indicates that ECV may potentially be withheld from a significant proportion of women who could benefit from it without increased risks of adverse events. To our knowledge this is the first review on contraindications for ECV. We conducted this review with a comprehensive search strategy and made a concerted effort to find all the evidence. For each contraindication mentioned in the literature and guidelines, we searched for evidence and thereby exposed gaps in the knowledge about contraindications for ECV. More than three-quarters of the identified contraindications (33 of 39) were not based on any evidence. Publication bias could have influenced the results of our literature search; however, (case) reports on

76 Chapter 4 Contraindications for external cephalic version in breech position at term negative outcomes are more likely to be written down and published compared with reports on cases with good outcomes. Therefore, we think that the chance we have missed adverse outcomes related to suggested contraindications in guidelines and literature is very low. A limitation of this study is the low occurrence rate of complications of ECV, which makes it difficult to prove a (significant) relation between certain patient characteristics and adverse events. A systematic review by Grootscholten et al.20 reported on 60 severe adverse outcomes in 12.955 ECVs (0.5%) from 84 studies. Severe adverse outcomes included stillbirth (20%), placental abruption (20%), prolapsed cord (13%) and abnormal CTG after ECV resulting in emergency CS (47%).Considering the safety and effectiveness of ECV in preventing cesarean delivery in the case of breech presentation, international agreement on contraindications for ECV is important. According to several international studies, we estimate that worldwide ECV is not offered to about 10–15% of eligible women.21–24 Therefore, we would like to propose a reduced list of three contraindications. This list of recommended contraindications is reported in table 3. (table 3 invoegen) Placental abruption is mentioned as a rare complication of ECV in two articles.5,25 Therefore, we think it is justified to withhold ECV from patients with an increased risk of placental abruption (a history of placental abruption and severe preeclampsia or haemolysis el- evated liver enzymes low platelets (HELLP) syndrome. Furthermore, we included signs of fetal distress, as distressed foetuses should not be exposed to more stress. The above-mentioned systematic review on complications5 reports that ECV can result in (passing) fetal distress, as seen on CTGs afterwards. Table 3. Recommended restriction of contraindications (level of evidence).

Maternal factors Fetal factors Placental abruption in history or signs of Signs of fetal distress (abnormal CTG and or placental abruption (level III) abnormal Doppler flow) (level III) Severe preeclampsia or HELLP syndrome (level III)

1. Level I: Evidence obtained from at least one properly designed randomized controlled trial. 2. Level -‐ II 1: Evidence obtained from -‐ well designed controlled trials without randomization. 3. Level -‐ II 2: Evidence obtained from -‐ well designed cohort or -‐ case control analytic studies, 4. preferably from more than one center or research group. 5. Level -‐ II 3: Evidence obtained from multiple time series with or without the intervention. 6. Dramatic results in uncontrolled trials might also be regarded as providing this type of 7. evidence. 8. Level III: Opinions of respected authorities, based on clinical experience, descriptive studies or 9. reports of expert committees. 10. CTG, cardiotocography; HELLP, haemolysis evated el liver enzymes low platelets. 77

Chapter 4 Contraindications for external cephalic version in breech position at term

REFERENCES

1 Rietberg C. Term breech presentation in The Netherlands from 1995 to 1999: mortality and morbidity in relation to the mode of delivery of 33,824 infants. BJOG An Int J Obstet Gynaecol [Internet]. 2003 Jun [cited 2014 Mar 3]; 110: 604–609. Avail- able from: http://doi.wiley.com/10.1016/S1470-0328(03)01907-4

2 Sullivan EA, Moran K, Chapman M. Term breech singletons and caesarean section: a population study, Australia 1991-2005. Aust N Z J Obstet Gynaecol [Internet]. 2009 Oct [cited 2014 Mar 3]; 49: 456–460. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/19780725

3 Hartnack Tharin JE, Rasmussen S, Krebs L. Consequences of the Term Breech Trial in Denmark. Acta Obstet Gynecol Scand [Internet]. 2011 Jul [cited 2014 Feb 20]; 90: 767–771. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21476999

4 Hofmeyr, GJ; Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012; 10.

5 Collaris RJ, Oei SG. External cephalic version: a safe procedure? A systematic review of version-related risks. Acta Obstet Gynecol Scand [Internet]. 2004 Jun; 83: 511– 518. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15144330

6 Verburgt, T; Offerhaus P. Uitwendige versie (External Cephalic Version). NTOG. 2007; 120: 4–7.

7 Committee Opinion No 340 A. Mode of term singleton breech delivery. Obstet Gy- necol 2006. Obs Gynecol. 2006; 108: 235–237.

8 Royal College of Obstetricians and Gynecologists. External cephalic version and reducing the incidence of breech presentation. http://www.rcog.org.uk/files/rcog- corp/uploaded-files/GT20aExternalCephalicVersion.pdf. 2010. p. 1–8.

9 The Royal Australian and New Zealand College of Obstetricians and Gynecologists. Management of Breech Presentation at Term. 2001. p. file:///C:/Users/PS13C/ Downloads/C–Obs%2011%20Mana.

10 Otterlo Werkgroep dr. AJ van L en dr. MG van P. NVOG richtlijn - Stuitligging- versie 2.0 (Guideline on breech presentation - version 2.0 of the Dutch Society of Obstet- rics and Gynecology) [Internet]. 2008. Available from: file:///C:/Users/PS13C/Down- loads/Stuitligging (1).pdf

78 Chapter 4 Contraindications for external cephalic version in breech position at term

11 Kok M, Cnossen J, Gravendeel L, van der Post J, Opmeer B, Mol BW. Clinical factors to predict the outcome of external cephalic version: a metaanalysis. Am J Obstet Gynecol [Internet]. 2008 Dec [cited 2014 May 1]; 199: 630.e1–7; discussion e1–5. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18456227

12 De Meeus JB, Ellia F, Magnin G. External cephalic version after previous cesarean section: a series of 38 cases. Eur J Obstet Gynecol Reprod Biol [Internet]. 1998 Oct; 81: 65–68. Available from: http://www.ncbi.nlm.nih.gov/pubmed/9846717

13 Abenhaim H a, Varin J, Boucher M. External cephalic version among women with a previous cesarean delivery: report on 36 cases and review of the literature. J Perinat Med [Internet]. 2009 Jan [cited 2014 May 1]; 37: 156–160. Available from: http:// www.ncbi.nlm.nih.gov/pubmed/19021458

14 Sela HY, Fiegenberg T, Ben-Meir A, Elchalal U, Ezra Y. Safety and efficacy of external cephalic version for women with a previous cesarean delivery. Eur J Obstet Gynecol Reprod Biol [Internet]. 2009 Feb [cited 2014 May 1]; 142: 111–114. Available from: http://www.ncbi.nlm.nih.gov/pubmed/19019528

15 Berghella V. Prolapsed cord after external cephalic version in a patient with premature rupture of membranes and transverse lie. Eur J Obstet Gynecol Reprod Biol [Internet]. 2001 Dec 1; 99: 274–275. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/11788188

16 Brost, BC; Adams, JD; Hester M. External cephalic version after rupture of me... [Ob- stet Gynecol. 2000] - PubMed - NCBI. Obs Gynecol [Internet]. 2000 [cited 2014 May 1]; 95 (6 Pt 2: 1041. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pu bmed/?term=External+cephalic+version+after+ruptured+membranes.+brost

17 Kok M, Cnossen J, Gravendeel L, Van Der Post J a, Mol BW. Ultrasound factors to predict the outcome of external cephalic version: a meta-analysis. Ultrasound Ob- stet Gynecol [Internet]. 2009 Jan [cited 2014 May 1]; 33: 76–84. Available from: http://www.ncbi.nlm.nih.gov/pubmed/19115237

18 Boucher M. The relationship between amniotic fluid index and successful external cephalic version: a 14-year experience. Am J Obstet Gynecol [Internet]. 2003 Sep [cited 2014 May 1]; 189: 751–754. Available from: http://linkinghub.elsevier.com/ retrieve/pii/S0002937803008469

79 Chapter 4 Contraindications for external cephalic version in breech position at term

19 Haas DM, Magann EF. External cephalic version with an amniotic fluid index < or = 10: a systematic review. J Matern Fetal Neonatal Med [Internet]. 2005 Oct [cited 2014 May 1]; 18: 249–252. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/16318975

20 Grootscholten, K; Kok, M; Oei, SG; Mol, BW; van der Post J. External cephalic version-related risks: a meta-analysis. Obs Gynecol [Internet]. 2008 [cited 2014 Mar 20]; 112: 1143–1151. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/ pubmed/?term=xternal+cephalic+version-related+risks:+a+meta-analysis.

21 Caukwell S, Joels L a, Kyle PM, Mills MS. Women’s attitudes towards management of breech presentation at term. J Obstet Gynaecol [Internet]. 2002 Sep [cited 2014 Mar 20]; 22: 486–488. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/12521413

22 Bewley, S; Shennan A. Peer review and the Term Breech Trial. Lancet. 2007; 369: 2007.

23 Leung, TY; Lau, TK; Lo, KW; Rogers M. A survey of pregnant women’s attitude towards breech delivery and external cephalic version. Aust N Z J Obs Gynaecol. 2000; 40: 253–259.

24 Rijnders M, Offerhaus P, Dommelen P Van, Wiegers T, Buitendijk S. Prevalence , Out- come , and Women ’ s Experiences of External Cephalic Version in a Low-Risk Popu- lation. 2010; 124–133.

25 Sheiner E, Shoham-Vardi I, Hallak M, Hadar a, Gortzak-Uzan L, Katz M, et al. Placen- tal abruption in term pregnancies: clinical significance and obstetric risk factors. J Matern Fetal Neonatal Med [Internet]. 2003 Jan; 13: 45–49. Available from: http:// www.ncbi.nlm.nih.gov/pubmed/12710856

5 Patient’s willingness to opt for external cephalic version

Floortje Vlemmix

Marjon Kuitert

Joke Bais

Brent C. Opmeer

Joris A.M. van der Post

Ben Willem J. Mol

Marjolein Kok

J Psychosom Obstet Gynaecol. 2013 Mar;34(1):15-21.

83 Chapter 5 Patient’s willingness to opt for external cephalic version

ABSTRACT

Objective External cephalic version (ECV) is a relatively simple and safe manoeuvre that reduces the caesarean section (CS) rate for breech presentation. There is professional consensus that ECV should be offered to all women, but only up to 70% of patients opt for this treatment. To improve counselling, we investigated the value patients place on various aspects of ECV.

Methods We studied patient preferences by means of a vignette study. Vary- ing levels of treatment characteristics were investigated in 16 scenarios, all including the ‘opt out’ alternative of an elective CS. The probability that women preferred ECV was estimated using a logistic regression approach.

Results Forty seven women participated in the study. Pain was the most important factor negatively influencing the willingness to opt for ECV (OR 0.11 (95% CI 0.05-0.23) for a pain score of 8-10 compared to 1-2 on a VAS of 0-10). Higher success rates of vaginal delivery after successful ECV increased women’s willingness (OR 3.42 (95% CI 2.04-5.74) if chance of vaginal delivery after successful ECV increased from 24 to 52%). The risk of an emergency CS during ECV did not influence the willingness to opt for ECV (OR 0.83 (95%CI 0.59-1.18) of chance increased from 0 to 1%).

Conclusions We conclude that expected pain during treatment and the success rate are the most important factors influencing the willingness to undergo ECV. Taking this information into account when counselling for ECV and reassuring women that unbearable pain is always a reason to stop ECV, and that the vast majority of women reported that the experienced pain is bearable, might. improve the uptake of ECV and decrease the number of CS due to breech presentation.

84 Chapter 5 Patient’s willingness to opt for external cephalic version

INTRODUCTION

Breech presentation occurs in 3 to 4% of all term pregnancies.1 Vaginal delivery of a breech presentation is associated with higher perinatal mortality and morbidity compared to vaginal delivery of a fetus in cephalic presentation.2,3 An international, multi centre randomized controlled trial (Term Breech Trial), evaluating the neonatal and maternal morbidity and mortality of planned caesarean sections (CS) and planned vaginal breech deliveries was published in 2000. This study proved that a planned CS was the safest mode of delivery for the child leading to less mortality and morbidity. This publication resulted in an increase from 50 to 80% of the overall- caesar ean section (CS) rate for term breech presentations in the Netherlands.3 This makes breech presentation responsible for one fifth of all the at term cesarean deliveries.4

The downside of these cesarean deliveries is the consequences for subsequent pregnancies. There is an increased risk of uterine rupture during a delivery of a next child and placenta accreta (abnormally deep attachment of the placenta, leading to increased blood loss and an increased risk of hyster- ectomy). Although the absolute chances are still small, the relative increase is significant and consequences dramatic causing severe maternal and fetal morbidity or even mortality.5

External cephalic version (ECV) is the only treatment option for women to achieve the safest route of delivery: cephalic vaginal delivery. In the Neth- erlands, ECV can be performed by the gynaecologist either a special trained midwife. Despite the promotion of ECV by national guidelines of gynaecologists and midwives, not all eligible women undergo ECV. In Australia, Raynes- Greenow and others investigated pregnant women’s preferences and knowledge of term breech management; 39% of the women responded that they would or would not choose ECV, and the 22% were uncertain.6 An study in the UK by Bewley et al reported a 50% ECV rate among term breech presentation.7 Yogev and colleagues performed a similar study in Israel.8 They reported a willingness to consider ECV among 53.8% of women in 1995, and only 23.9% in 2001. The suboptimal implementation of ECV might be the result of poor implementation (doctors and midwives do not offer ECV to their patients) or poor communication (doctors and midwives do not succeed in explaining the advantages of ECV to convince patients of the safety and advantages of ECV). On the other hand, the low uptake may also reflect a low

85 Chapter 5 Patient’s willingness to opt for external cephalic version

requirement women feel to avoid a CS or their fear to undergo ECV. In this vignette study, we investigate patients‘ preferences for ECV by means of the value patients place on various aspects of ECV This information can be used during counselling in order to increase the uptake of ECV and decrease the number of CS due to breech presentation.

MATERIALS AND METHODS

Design

A vignette study was designed to investigate if women make trade-offs between various aspects of external cephalic version (ECV), and to elicit the relative weight that women place on these characteristics. Preferences and perceived similarities or differences between choice alternatives canbe evaluated using structured vignettes.9 Vignettes consist of brief structured descriptions of hypothetical scenarios, based on a limited number of characteristics (attributes). A choice set consists of a series of paired vignette- combinations, so that vignettes in each combination systematically differ on one or more attributes. Subsequently, patients are asked to express their preference for either vignette for each vignette-pair in the choice set. Pa- tients preference responses can be statistically analysed to quantify the relative importance of each attribute.

Study population

From June until October 2009, we recruited pregnant women with a breech presentation at term eligible for ECV in one academic and two teaching hospitals in and around Amsterdam, the Netherlands. Women with a poor understanding of the Dutch language were excluded. Women were invited to participate in this study when visiting the hospital for counselling for ECV. All participants were informed about the hypothetical character of the vignettes and the possibility of unrealistic success percentages or complication rates in the choice sets. It was ascertained that the answers would not have any consequences regarding their own clinical situation. If consent was given, women were asked to fulfil the questionnaire before they went home. We aimed to administer the vignette study in 45 women. Earlier studies have

86 Chapter 5 Patient’s willingness to opt for external cephalic version shown that this number of respondents provide sufficient variability to perform the statistical analyses.10–12

Attributes and attribute levels

In a vignette study it is assumed that a given healthcare intervention or treatment can be described by its characteristics (attributes) and that patients preference for a medical intervention is determined by the levels of these attributes. The choice of attributes and attribute levels for this study was based on the literature and expert interviews (n=4).11 The chosen attributes were use of uterine relaxants, risk of emergency CS due to ECV, pain during ECV attempt and chance of successful vaginal cephalic delivery after ECV (table 1).

Table 1. Attributes and levels for external cephalic version

Attributes ECV Beta coefficients in regression analysis ECV attempt Yes, without medication iv β1 Yes, with medication iv

Chance of emergency CS during 3 /1000 β2 ECV attempt 6 /1000 10 /1000

Pain during ECV 1 attempt 1-‐2 β3 3-‐4 5-‐7 8-‐10

Outcome of ECV: 240 /1000 -‐ 37 /1000 β4 Chance of vaginal cephalic 340 /1000 -‐ 36 /1000 delivery and corresponding 420 /1000 -‐ 35 /1000 maternal morbidity in case of CS. 520 /1000 -‐ 34 /1000

1 on a visual analog -‐ scale 0 10 (0= no pain, 10 = worst pain)

87 Chapter 5 Patient’s willingness to opt for external cephalic version

We varied the attribute levels for emergency CS from 3 to 10 per 1000 ECV attempts.13 Experienced pain during ECV was categorized in levels of 0 to 2, 3 to 4, 5 to 7 and 8 to 10 on the visual analogue scale. The benefits of ECV were reported as the increased chance of successful vaginal cephalic delivery and a decreased chance of maternal morbidity due to CS. The success rates of ECV varied from 30 to 70% and the risk of a CS after successful ECV from 10 to 30%.14–16 Combination of these percentages resulted in attribute levels of successful vaginal cephalic delivery ranging from 240-520 per 1000 women undergoing ECV. Maternal morbidity due to CS ranged from 34 to 37 per 1000 women.17 Maternal morbidity decreased if the chance of successful vaginal delivery increased. We assumed that every child in breech presentation at birth would be delivered with CS.

Development of choice sets

The combination of attributes and their levels resulted in 96 hypothetical treatment profiles: one attribute with two levels, one with three levels and two with four levels (21*31*42). We used a fractional factorial design to generate 16 treatment profiles. This number is considered large enough to estimate at least all main effects in a regression analysis, but small enough to be fulfilled by patients within a proper time.

With these sixteen treatment profiles, we designed a similar number of choice sets each containing one of the profiles from the fractional factorial design against the opting out to undergo a primary CS (attributes in the opting out ‘primary CS’ were: no ECV attempt, thus no chance of pain during ECV nor an emergency CS due to complications of ECV, 0,04% (40/1000) chance of a vaginal cephalic delivery). The choice sets are shown in table 2.

88 Chapter 5 Patient’s willingness to opt for external cephalic version

96 90 69 73 88 97 97 72 80 81 97 37 92 93 73 95

-‐ 89 -‐ 69 -‐ 39 -‐ 47 -‐ 60 -‐ 83 -‐ 89 -‐ 45 -‐ 54 -‐ 55 -‐ 81 -‐ 15 -‐ 71 -‐ 74 -‐ 46 -‐ 78 -‐

95% CI

-‐ 92 82 54 61 77 92 95 59 69 70 92 24 85 86 61 89 %women Predicted Predicted choosing ECV choosing

40 420 240 520 240 520 240 240 340 340 420 340 520 340 520 420 420 delivery attempts* / 1000 ECV 1000 /

presented with the corresponding maternal morbidity the corresponding maternal with presented Vaginal cephalic cephalic Vaginal

0 3 0 3 6 0 3 3 6 6 6 0 0 10 10 10 10 Emergency CS /1000 attempts /1000

2 2 7 7 4 4 7 2 4 2 7 4

10 10 10 10 -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ 0 -‐ -‐ -‐ -‐ 1 1 5 5 3 3 5 1 3 1 5 3 ECV 8 8 8 8 VAS score Pain during during Pain

No No No No No No No No No Yes Yes Yes Yes Yes Yes Yes Yes Uterus relaxation

ECV profiles in the choice sets the in profiles ECV

7 9 5 8 4 1 6 3 2 10 12 16 13 14 15 11 (no ECV) (no Table 2. Table cephalic delivery is success vaginal chance of full the profile, every *In in case of CS as described in table 1. described as CS of case in Current order corresponds to figure 1. questionnaire. in set choice of order to corresponds set choice of Number ^ Number of Number choice set^ Comparator

89 Chapter 5 Patient’s willingness to opt for external cephalic version

Development of questionnaire

In the introduction of the questionnaire, the use of uterine relaxants was explained as an intravenous bolus of medication with the possibility of palpi- tations for the mother for at most a couple of hours, but no side effects for the fetus. Emergency CS was described as a CS in the first hours after ECV due to uncertainty of fetal condition. Pain levels were showed on a visual analogue scale from 0 (no pain) to 10 (unbearable pain). The advantage of ECV was expressed as the chance of successful cephalic vaginal birth (the questionnaire is available from the authors on request). Maternal morbidity was described as longer hospital stay and recovery after CS, increased blood loss and increased chance of wound infection. Pictures, graphs and pictograms were included to demonstrate percentages and rates. To assess the understanding of the attributes the questionnaire contained a dominant choice set (rationality test).

Baseline characteristics of participants were documented to explore whether these relate to preferences regarding ECV. This contained parity, mode of previous delivery /deliveries, preferred mode of delivery in case of persisting breech, educational level, and prior knowledge on ECV. After an ECV attempt (or for those women who opted for primary CS, after counselling) we asked for the most important reason (not) to undergo ECV. Women rated the pain during the ECV attempt on a visual analogue scale (VAS).

Statistical analysis

The probability that women preferred an ECV as compared to no ECV for a choice set, given the profile of risks and outcomes described for that choice set, was estimated using a logistic regression approach that accounts for repeated observations from the same individual (ref: generalized estimating equations (GEE).11,18 The overall chance of choosing ECV was calculated by the formula:

V = β0 + β1 (MEDICATION) + β2 (PAIN during ECV)

+ β3 (CHANCE of EMERGENCY CS)

+ β4 (CHANCE of CEPHALIC PRESENTATION)

90 Chapter 5 Patient’s willingness to opt for external cephalic version

Where V represents the chance of choosing in favour of ECV, β0 is the constant reflecting the a priori respondents’ preference for an ECV attempt compared to no treatment for the breech presentation assuming that this choice will be followed by a CS. The Beta coefficients estimated in this model reflect the extent to which the probability that women will prefer ECV to primary CS increases (positive value of beta) or decreases (negative value of beta) when each characteristic on the choice set changes. Thereby the beta for each attribute reflects the independent influence of that attribute on a woman’s preference, given the value of all other attributes. Risk ratios were calculated to assess the relative value of an attribute compared to another attribute by dividing the beta’s of these attributes. The regression model can be used to predict the proportion of women that will prefer ECV based on the characteristics in that choice set. These estimates were compared to the observed proportion of women preferring ECV for each vignette as to indicate the model fit. The analyses were carried out in SPSS 16.0 and SAS 9.1.2. P-values < 0.05 were considered statistically significant.

RESULTS

Study population

Forty-nine respondents participated of which two had to be excluded from analysis due to incomplete questionnaires. The majority of women explained that the choice sets were clear and they had no difficulty in completing the questionnaire. Baseline characteristics are reported in table 3.

None of the respondents had undergone a CS in previous pregnancies. Twen- ty-three women (52%) preferred a CS in case of persisting breech. Only three women (7%) did not undergo ECV for reasons as a traumatic birth experience in the history, preference for primary CS, or the explanation ‘it just doesn’t feel right to physically undergo ECV’. The majority of women who underwent ECV did this to prevent CS. Another argument was ‘to avoid the choice of vaginal breech delivery versus CS’. The ECV success rate in our study population was 26%. The mean level of pain, scored on a VAS of 0 to 10, after ECV was 5.5 (range 2-9). Women with a successful ECV reported slightly lower pain scores as compared to women with an unsuccessful ECV (4.5 vs. 6.0, p=0.067).

91 Chapter 5 Patient’s willingness to opt for external cephalic version

Table 3. Respondent characteristics

All patients N = 47 (%) Mean Age (years) (Range) 30 (21-‐38) Parity Nulliparous 31 (66) Multiparous 16 (34) Mean gestational (weeks) (range) 36 6/7 (36 -‐ 0/7 39 1/7 ) Previous mode(s) of delivery Vaginal delivery 14 (30) Instrumental delivery 3 (6) Caesarean Section 0 Preference of birth in case of breech presentation Vaginal breech delivery 21 (48) Caesarean Section 23 (52) Level of education Low 21 (45) Median 16 (34) High 10 (21) Prior knowledge of breech presentation 34 (72) Doctor / midwife 9 Family /Friends 21 Media (television, magazine) 14 Breech presentation in history 1 Other 6 Prior knowledge of ECV 30 (64) Doctor / midwife 4 Family /Friends 19 Media (television, magazine) 15 ECV in history 1 Other 5 Outcome ECV Successful 12 (26) Not attempted 3 (7)

92 Chapter 5 Patient’s willingness to opt for external cephalic version

Vignette study

The a priori willingness to undergo ECV was 68% (β0= 0.77). Three out of four attributes were found to affect women’s willingness to undergo ECV in order to prevent a breech presentation at birth (table 4).

Table 4. Women’s preference to undergo ECV to prevent breech presentation.

Attribute OR 95%CI p-‐value Beta coefficient Constant (undergoing ECV) 0.056 0.77

Uterus relaxation (β1) No 1.00 -‐ -‐ -‐ Yes 0.57 0.39 -‐ 0.83 0.003 -‐0.56 Pain score (β2) 1-‐2 1.00 -‐ -‐ -‐ 3-‐4 1.01 0.67 -‐ 1.53 0.95 0.01 5-‐7 0.38 0.21 -‐ 0.72 0.003 -‐0.96 8-‐10 0.11 0.05 -‐ 0.23 <0.0001 -‐2.19 Emergency CS (β3) 0% 1.00 -‐ -‐ -‐ 0.3% 1.34 1.02 -‐ 1.78 0.04 0.30 0.6% 1.12 0.77 -‐ 1.62 0.55 0.29 1% 0.83 0.59 -‐ 1.18 0.38 -‐0.18 Vaginal cephalic delivery (β4) 24% 1.00 -‐ -‐ -‐ 34% 1.90 1.18 -‐ 3.01 0.008 0.64 42% 3.06 1.79 -‐ 5.20 <0.0001 1.12 52% 3.42 2.04 – 5.74 <0.0001 1.23

The chance of a vaginal cephalic delivery was the most important attribute increasing the chance of opting for ECV; an increase from 24 to 52% increase the odds of preferring ECV from 1.9 to 3.4. The negative sign of the beta in case of the use of intravenous uterus relaxants indicates the decreased chance of opting for ECV when medication is used. Pain during ECV was also negatively associated with the willingness to opt for ECV if pain exceeded a score of 5 on the VAS score. The willingness to opt for ECV was not related to varying chances of emergency CS during ECV.

To assess the relative value, or importance, of the characteristic of intravenous uterus relaxants (increasing success rate vs. side effects), ratios of the effects were calculated (βsuccess rate / β medication). An increase from a low success rate of 24 to 34% was considered 1.1 times (0.64/0.56) as important as ex-

93 Chapter 5 Patient’s willingness to opt for external cephalic version

posure to side effects of intravenous medication. However, an increase from 42 to 52% was considered of less value than exposure to side effects ((1.23- 1.12)/0.56 = 0.2).

An increased pain score up to 3-4 was never considered as more important than an increased success rate. But a pain score of 5-7 decreased the willingness to undergo ECV more than an increase of 10% in success rate. An increase with 18% from 24% to 42% was considered of more importance than a pain score of 5-7 (1.12 / 0.96 = 1.17), however a similar increase but with a higher initial success rate of 34% did not outweigh the pain score of 5-7 (1.23-.064 / 0.96 = 0.61). A pain score of 8-10 was always of greater influence on the willingness for ECV compared to increased success rates.

The predicted chance of choosing ECV above elective cesarean breech delivery for each presented choice set is shown in figure 1. The choice sets are put in the order of increasing chance of opting for ECV. Attribute levels per choice set and 95% confidence intervals of the predicted chance are reported in table 4. Only one fifth of the women will prefer ECV under circumstances mentioned in choice set seven: intravenous medication, a lot of pain, a moderate chance of emergency CS and a low success rate of ECV. Increasing the chance of a vaginal delivery in cephalic presentation and decreasing exces-

sive pain can raise the number of women willing to undergo ECV up to 90%.

Figure 1: Predicted chance (95% CI) of choosing ECV under given circumstances in the choicesets.

1,0

0,8

0,6

0,4 Chanceof choosing ECV

0,2 under given circumstances in choiceset in circumstances given under

0,0 7 9 12 6 5 16 11 13 8 15 3 2 4 14 10 1 Number of Choiceset

94 Chapter 5 Patient’s willingness to opt for external cephalic version

DISCUSSION

In this study, we assessed to which extend the factors pain, use of uterus relaxants, risk of an emergency CS and chance of a vaginal delivery in cephalic presentation affect women’s preference for ECV. We conclude that fear of considerable pain (VAS>8) and increased chance of vaginal delivery in cephalic presentation are the most important factors influencing the willingness to undergo ECV.

A strength of this study is that the vignette study was administered in a face- to-face setting: the researcher could check a respondent’s understanding of the instrument, provide tailored explanation of the questions, and monitor its completeness. A general limitation of a vignette study is the limitation of number of attributes, due to the random component variability. Other choices of vignettes and attribute level might lead to other outcomes. However, we feel that the chosen samples represent the most important characteristics of ECV. For instance, we limited our attributes to short term consequences, but by offering ECV also the risks for subsequent pregnancies is reduced (less CS due to breech presentation results in less complications due to CS scars for subsequent pregnancies). We only focused on the short term outcomes, as more attributes would make the vignettes more complex and less compre- hensible. Nevertheless, in daily practice, this is an important issue to discuss with parents when counselling for ECV. If the sample of participants reflects the preferences of all women with a breech presentation, might be topic of debate. Only 3 (6%) did not undergo ECV. This indicates that there is a selection bias at the moment of referring patients for counselling of ECV: women refusing ECV the moment the breech presentation is diagnosed are probably not sent in for counselling. Therefore, our population includes more women willing to undergo ECV compared to the general population with a breech presentation at term, which might have influenced the effect size of the -dif ferent attributes and attribute levels. It is reasonable to assume that it will not have affected the negative value of the beta coefficient of the attribute pain or side effects as one might assume that, for instance, those not opting for ECV are even less willing to undergo pain or side effects.

Women were less likely to undergo ECV in the choice sets with a pain level of 8-10 (the first four in figure one). This is an important attribute useful in counselling, as the average pain score of ECV in this study was 5.5 and only 16% of the women scored an 8 or higher in the VAS scale after their

95 Chapter 5 Patient’s willingness to opt for external cephalic version

ECV attempt. Communicating these facts during counselling might decrease fear for pain leading to more women who opt for ECV. Since pain is such an important issue for women, it would be interesting to know if offering pain medication / epidural anaesthesia would increase the uptake of this treatment. We did not incorporate pain relief into our choice sets.

An interesting finding is that the attribute ‘emergency CS’, of which we thought of as a risk which would scare of, did not logically influence the choice women made. The increasing risk did not lead to a corresponding increasing importance (β). An explanation might be that patients consider an absolute risk ranging from 3 to 10 per 1000 as too low to take into account in their decision.

This trial was performed in regions where ECV is performed in the hospital. In some regions in the Netherlands ECV is also performed in out-of-hospital settings by special trained midwives. A Study by Rijnders et al., evaluated women’s experience who’s pregnancies were guided by a midwife in an out- of-hospital setting. Among 167 women with a breech presentation 73.7% received the version. Surprisingly, more women with a failed successful version rated it as a good experience compared to women with a successful version (71 vs. 49%, p= 0.015). Women who reported significant pain or fear during the version experienced the version more negatively (OR: 6.0, 95% CI: 3.3–12.2 and OR: 2.7, 95% CI: 1.1–6.0, respectively).19 Although our trial describes the anticipation of women towards ECV, it is in line with the findings of this trial: pain is a major factor influencing the attitude towards ECV. Since this is a study focusing on a different cohort as our study (out vs. in hospital pregnancy care) we do think our results can be generalized to all women with a breech presentation. Only the item of uterus relaxation use might be scored differently by patients visiting a midwife for ECV as this is not administered in an out-of-hospital setting. As success rates of ECV are an important factor to choose ECV, women need a better insight in these success rates, as these vary across obstetric practices and across individual women.20,21 Many factors are associated with the probability of a successful version, so when formalized in prediction models, this information could be used to counsel women according to their individual profile and preferences.11 Since women put greater value on ECV success rate than side effects of medication, doctors should not hesitate to suggest the use of uterus relaxants in case of a low predicted chance of success. Besides individual patient success, trans-

96 Chapter 5 Patient’s willingness to opt for external cephalic version parency of success rates among clinics is also of value to improve care in this field of medicine, as it enables patients to visit clinics with highest success rates. Besides it might stimulate clinics to improve their expertise in ECV for instance, by organizing special office hours.22

97 Chapter 5 Patient’s willingness to opt for external cephalic version

REFERENCES

2 Evers A, Brouwers H, Hukkelhoven PM, Nikkels PGJ, Boon J, Van A, et al. Perinatal mortality and severe morbidity in low and high risk study. BMJ. 2010; 341: 1–8.

3 Rietberg, CCT; Elferink-Stinkens, PM; Visser G. The effect of the Term Breech Trial on medical intervention behaviour and neonatal outcome in The Netherlands: an analysis of 35,453 term breech infants. BJOG [Internet]. 2005 Aug [cited 2014 Mar 3]; 112: 1163; author reply 1163–4. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/16045537

4 Wilmink F a, Hukkelhoven CWPM, Lunshof S, Mol BWJ, van der Post J a M, Papatso- nis DNM. Neonatal outcome following elective cesarean section beyond 37 weeks of gestation: a 7-year retrospective analysis of a national registry. Am J Obstet Gyne- col [Internet]. Elsevier Inc.; 2010 Mar [cited 2014 May 30]; 202: 250.e1–8. Available from: http://www.ncbi.nlm.nih.gov/pubmed/20207243

5 US National Institutes of Health C. Caesarean section-the first cut isn’t the deepest. Lancet. 2010; 375: 956.

6 Raynes-Greenow CH, Roberts CL, Barratt A, Brodrick B, Peat B. Pregnant women’s preferences and knowledge of term breech management, in an Australian setting. Midwifery [Internet]. 2004 Jun [cited 2014 May 19]; 20: 181–187. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15177862

7 Bewley, S; Robson, SC; Smith, M; Glover, A; Spencer J. The introduction of external cephalic version at term into routine clinical practice. Eur J Obs Gynecol Reprod Biol [Internet]. 1993 [cited 2014 Mar 20]; 52: 89–93. Available from: https://c38xrlad- 25d4epd6rje0lpk.sec.amc.nl/pubmed/8157147

8 Yogev Y, Horowitz E, Ben-Haroush a, Chen R, Kaplan B. Changing attitudes toward mode of delivery and external cephalic version in breech presentations. Int J Gynae- col Obstet [Internet]. 2002 Dec; 79: 221–224. Available from: http://www.ncbi.nlm. nih.gov/pubmed/12445986

98 Chapter 5 Patient’s willingness to opt for external cephalic version

9 Bachmann LM, Mühleisen A, Bock A, ter Riet G, Held U, Kessels AGH. Vignette studies of medical choice and judgement to study caregivers’ medical decision behaviour: systematic review. BMC Med Res Methodol [Internet]. 2008 Jan [cited 2014 May 27]; 8: 50. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcg i?artid=2515847&tool=pmcentrez&rendertype=abstract

10 Bekker-grob EW De, Meerding WJ, Koes BW, Steyerberg EW. Preferences of GPs and Patients for Preventive Osteoporosis Drug Treatment. 2009;27 : 211–219.

11 Kok M, Gravendeel L, Opmeer BC, van der Post J a M, Mol BWJ. Expectant parents’ preferences for mode of delivery and trade-offs of outcomes for breech presentation. Patient Educ Couns [Internet]. 2008 Aug [cited 2014 Mar 3];72 : 305–310. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18534810

12 Watson V, Ryan M, Brown C, Barnett G, Ellis B, Emberton M. Eliciting Preferences for Drug Treatment of Lower Urinary Tract Symptoms Associated With Benign Pros- tatic Hyperplasia. J Urol [Internet]. 2004 Dec [cited 2014 May 30];172 : 2321–2325. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0022534705614048

13 Grootscholten, K; Kok, M; Oei, SG; Mol, BW; van der Post J. External cephalic version-related risks: a meta-analysis. Obs Gynecol [Internet]. 2008 [cited 2014 Mar 20]; 112: 1143–1151. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/ pubmed/?term=xternal+cephalic+version-related+risks:+a+meta-analysis.

14 Hofmeyr, GJ; Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012; 10.

15 Jain S, Mulligama C, Tagwira V, Guyer C, Cheong Y. Labour outcome of women with successful external cephalic version: a prospective study. J Obstet Gynaecol [Inter- net]. 2010 Jan [cited 2014 May 30]; 30: 13–16. Available from: http://www.ncbi. nlm.nih.gov/pubmed/20121496

16 Vézina Y, Bujold E, Varin J, Marquette GP, Boucher M. Cesarean delivery after successful external cephalic version of breech presentation at term: a comparative study. Am J Obstet Gynecol [Internet]. 2004 Mar [cited 2014 May 30]; 190: 763– 768. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15042011

17 Hannah ME, Hannah WJ, Hewson S a, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet [Internet]. 2000 Oct;356 : 1375–1383. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0140673600028403

99 Chapter 5 Patient’s willingness to opt for external cephalic version

18 SH L. Generalized estimating equations for correlated binary data: using the odds ratio as a measure of association. Biometrika. 1991;78 : 153–160.

19 Rijnders M, Offerhaus P, Dommelen P Van, Wiegers T, Buitendijk S. Prevalence , Out- come , and Women ’ s Experiences of External Cephalic Version in a Low-Risk Popu- lation. 2010; 124–133.

20 Kok M, Cnossen J, Gravendeel L, van der Post J, Opmeer B, Mol BW. Clinical factors to predict the outcome of external cephalic version: a metaanalysis. Am J Obstet Gynecol [Internet]. 2008 Dec [cited 2014 May 1]; 199: 630.e1–7; discussion e1–5. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18456227

21 Kok M, Cnossen J, Gravendeel L, Van Der Post J a, Mol BW. Ultrasound factors to predict the outcome of external cephalic version: a meta-analysis. Ultrasound Ob- stet Gynecol [Internet]. 2009 Jan [cited 2014 May 1]; 33: 76–84. Available from: http://www.ncbi.nlm.nih.gov/pubmed/19115237

22 Kuppens SMI, Hasaart THM, van der Donk MWP, Huibers M, Franssen MJ, de Becker BMJ, et al. Fewer caesarean sections for breech presentation following external cephalic version according to a protocol in a special office visit. Ned Tijdschr Geneeskd [Internet]. 2008 Jun 7 [cited 2014 May 30]; 152: 1323–1328. Available from: http:// www.ncbi.nlm.nih.gov/pubmed/18661859

100

6 Patients’ and professionals’ barriers and facilitators to external cephalic version for breech presentation at term, a qualitative analysis in the Netherlands.

Ageeth N. Rosman

Floortje Vlemmix

Margot A.H. Fleuren

Marlies E. Rijnders

Antje Beuckens

Brent C. Opmeer

Ben Willem J. Mol

Myra C. van Zwieten

Marjolein Kok

Midwifery. 2014 Mar;30(3):324-30

103 Chapter 6 Patients’ and professionals’ barriers and facilitators

ABSTRACT

Objective External cephalic version (ECV) is a relatively simple and safe manoeuvre and a proven effective approach in the reduction of breech presentation at term. There is professional consensus that ECV should be offered to all women with a fetus in breech presentation, but only up to 70% of women eligible for ECV undergo an ECV attempt. The aim of the study was to identify barriers and facilitators for ECV among professionals and women with a breech presentation at term.

Design Qualitative study with semi-structured interviews.

Setting Dutch hospitals. Participants: pregnant women with a breech presentation who had decided on ECV, and midwives and gynaecologists tre- ating women with a breech presentation. Measurements On the basis of national guidelines and expert opinions, we developed topic lists to guide the interviews and discuss barriers and facilitators in order to decide on ECV (pregnant women) or advice on ECV (midwives and gynaecologists).

Findings Among pregnant women the main barriers were fear, the preference to have a planned caesarean section (CS), incomplete information and having witnessed birth complications within the family or among friends. The main facilitators were the wish for a home birth, the wish for a vaginal delivery and confidence of the safety of ECV. Among professionals the main barriers were a lack of knowledge to fully inform and counsel patients on ECV, and the inability to counsel women who preferred a primary CS. The main facilitator was an unambiguous policy on (counselling for) ECV within the region.

Conclusion We identified several barriers and facilitators possibly explaining the suboptimal implementation of ECV for breech presentation in the Netherlands. This knowledge should be taken into account in designing implementation strategies for ECV to improve the uptake of ECV by professionals and patients.

104 Chapter 6 Patients’ and professionals’ barriers and facilitators

BACKGROUND

External cephalic version (ECV) is a relatively simple and safe manoeuvre, which is proven to be an effective approach in the reduction of breech presentation at term and consequently reduces the number of caesarean deliveries due to breech presentation.1 After the publication of the Term Breech Trial, which reported an improvement in severe neonatal outcome after planned caesarean section (CS) compared to planned vaginal breech birth, the percentage of caesarean deliveries in the Netherlands increased from 50% to 80% within a year.2,3 However, a CS leads to more maternal and fetal morbidity compared to a vaginal birth of a fetus in cephalic presentation. Thus, ECV is the best answer in the dilemma of short versus long-term consequences of the mode of delivery for breech presentation.

The Royal Dutch Organisation for Midwives (KNOV) and the Dutch Society for Obstetrics and Gynaecology (NVOG) both published guidelines for term breech presentation in 2000 and 2001, which recommend offering ECV to all women without a contra-indication for the procedure.

Despite these guidelines, a significant number of women do not undergo ECV. An inventory survey among all hospitals in the Netherlands in 2007 reported that 5% of the gynaecologic practices did not perform any ECV at all and a prospective cohort study in the Netherlands reported that 26% of eligible women declined an ECV attempt.4,5 This means that a substantial proportion of clients do not receive the appropriate care they should receive according to the guidelines.

One of the main problems with the introduction of guidelines in the healthcare system is that professionals do not – automatically – use the guidelines as intended by the developers. Several models and frameworks exist on how to introduce guidelines effectively.6–11 A detailed understanding of critical determinants, a so called determinant analysis, is a prerequisite for designing an implementation strategy that is adapted to the several critical determinants, in order to achieve real change.6

The aim of this study was to identify professionals’ and patients’ facilitators and barriers for the implementation of ECV at term.

105 Chapter 6 Patients’ and professionals’ barriers and facilitators

METHODS

Setting

In the Netherlands, ECV is performed in either an out-of-hospital setting by special trained midwives or within the hospital by gynaecologists or clinical midwives. There are hospitals with organised ECV office hours where an experienced group of professionals perform ECV, whilst in other hospitals ECV is performed by the professional who is on call at the moment an ECV is scheduled. Patients are informed and counselled by either the professional who diagnosed the breech position, or the professional who performs ECV, or both. The information provided varies widely and may contain an objective brochure, DVD and detailed counselling or only a few minutes explanation about the possibility of the procedure with a directive or non-directive advice from the health-care provider whether or not to undergo an ECV.

Focus group meetings with professionals

To identify potential facilitators and barriers for professionals to offer and perform ECV, we organised four focus group meetings each with a different subgroup of professionals: (1) midwives performing ECV in an out of hospital setting, (2) midwives referring their patients for ECV to a hospital, (3) obstetricians performing ECV, and (4) ambivalent obstetricians who either do not offer ECV or discourage it. The meetings were led by three members of the project group. We opted for a semi-structured interview process to prompt discussion.12. A random sample of hospitals and midwife practices was contacted by telephone and invited to participate, until there were at least five attendants per focus group, who were able to participate at the set date.

In this study, we used a framework developed by Fleuren et al. (Figure 1).13 This framework identifies four main stages in innovation processes: dissemination, meaning that every professional is actually supplied with the guideline; adoption, or the intention of the professional to use or not to use the guideline; implementation, the use of the guideline in daily practice and continuation, where working with the guideline becomes routine practice. The four main stages can be thought of as failure points where the desired change may not occur.

106 Chapter 6 Patients’ and professionals’ barriers and facilitators

Figure 1. Framework for the introduction and evaluation 7 of innovations

Innovation determinants Innovation process

Characteristics of the innovation strategy Characteristics of the innovation Dissemination

Characteristics of the Adoption professional

Characteristics of the Implementation organisation

Continuation Characteristics of the socio-‐political context

The transition from one stage to the next can be affected positively or negatively by various factors or ‘determinants’: (1) characteristics of the guidelines (e.g. relative advantage, complexity), (2) characteristics of the health professional (e.g. skills, knowledge, self-efficacy),. (3) characteristics of the organisation (e.g. available expertise, staff turnover, financial recourses), and (4) characteristics of the socio-political environment (e.g. collaboration with other professionals, client co-operation, reimbursement) (left part of Figure 1).13

Prompts were structured according to these determinants to reveal facilitators and barriers (details of the prompts are available in the supplementary Appendix). Prior to the focus groups confidentiality was assured and the process of the focus group was explained. The focus groups were audio taped and fully transcribed. All attendants were requested to take notice of the Dutch guidelines for their profession before the focus group meeting, as far as they were not familiar with these.

Patient interviews

To identify potential facilitators and barriers for patients to undergo ECV, we organised interviews with women who had made a decision regarding un-

107 Chapter 6 Patients’ and professionals’ barriers and facilitators

dergoing an ECV procedure. Clients were recruited in seven midwife practices and hospitals throughout the Netherlands. We purposely sampled patients with different ethnic and religious backgrounds and education levels. The interviews were carried out by AR and FV and conducted by telephone or face-to-face, according to the preference of the patient. Participation of the partner in the interviews was allowed, but not specifically requested. Again we chose for a semi-structured interview setting to allow free input by the participants. A topic list was based on the expert opinion of all of the authors and the list was extended after the first five interviews with topics indicated by the participants. We analysed the results for the first time after 24 interviews and saturation seemed to be reached after 18 interviews (no new facilitators or barriers were mentioned by the participants). Therefore, no more patients were interviewed and the results are based on the findings in the 24 conducted interviews.

Analysis

The detected determinants in the interviews were structured and analysed in MAXQDA (VERBI GmbH, Marburg, Germany), a software programme to assist qualitative data analysis. The aim of the analysis was to conceptualise the content of the patient interviews and focus group meetings. We started with the first focus group meeting and patient interview respectively, to generate a list of quotations containing all mentioned facilitators and barriers. The list was extended with all new mentioned barriers and facilitators in the subsequent interview. The quotations were categorised in four predefined domains: characteristics of the innovation (ECV), professional, organisation, and socio-political environment (implementation model by Fleuren etal. (Figure 1)13 and a fifth domain containing those quotations that did not fit into the 4 categories. This procedure was carried out by two authors and any inconsistencies were resolved by discussion and consensus and if necessary by consulting a third person. While coding the data it appeared that respondents often did not make a clear distinction between determinants related to the organisation or to the socio-political context. Therefore these two categories were combined into a broad category ‘organisational and political context’. This is in line with later publications of Fleuren et al., in which they specify several determinants which are related to both the level of the organisation and the socio-political context.13 Determinants related to the professional’s anticipated or perceived patient co-operation or patient satis-

108 Chapter 6 Patients’ and professionals’ barriers and facilitators faction, were mentioned apart which is a common category in other qualitative studies as well.12,13 This resulted in the four domains: characteristics of the intervention itself (ECV), of the professional, of the patient, of the organisational and the socio-political context.

FINDINGS

The characteristics of the 20 professionals, who attended the focus group meetings, are summarised in Table 1. Table 2 shows the patient characteristics of the 24 women, and summarised the degree in variety of age, parity, country of origin, and educational level.

Table 1. Professional characteristics

Characteristics Midwives Gynecologists N=12 (%) N= 8 (%) Male 1 (8) 5 (63) Female 11 (92) 3 (37) Regularly performing ECV 5 (42) 5 (63) Positive attitude towards ECV 12 (100) 3 (37)

Table 2. Patient characteristics

Characteristics Number of patients N= 24(%) Parity Nulliparous 13 (54) Multiparous 11 (46) Country of origin The Netherlands 18 (75) Other* 6 (25) Age (median, range) 32 (26 – 36) Educational level Low 3 (12) Medium 9 (38) High 12 (50) ECV attempt 16 (67) Successful 11 (69) Unsuccessful 5 (31) * England, Turkey, Spain, Russia, Ghana

109 Chapter 6 Patients’ and professionals’ barriers and facilitators

Barriers and facilitators related to the implementation of ECV according to professionals

The professionals identified 43 potential barriers and facilitators for ECV. The domain ‘characteristics of the professional’ contained the most barriers and the domain ‘characteristics of the context’ contained the most facilitators.

Domain 1: characteristics of the intervention The ‘procedure of ECV’ was described as the whole process from informing and counselling a patient until the actual turning of the baby. The explanation of ‘fetal monitoring during ECV’ was used to persuade the patient of the safety and therefore mentioned as facilitator.

Domain 2: characteristics of the professional The barriers identified in this domain can be divided into five categories. First of all; self-efficacy, this is the self-assessed skill to inform or counsel patients. Professionals mentioned the inability to convince patients of the safety and effectiveness of ECV who already decided to opt for a planned CS without prior counselling. Second, the lack of knowledge on ECV. This includes knowledge of the procedure (among those professionals who do not perform ECV themselves), the success rates of ECV in general and within their region, the complication risks and the content of the guidelines including the cited literature. Third, an (assumed) lack of skills among the professional performing ECV within that region (due to the idea that the success rates are low). Fourth, a minority of professionals was not convinced of (the safety of) ECV at all, and counselled against it. And last, having witnessed a major complication of ECV was a barrier for those professionals to advise ECV (quotation: ‘I once witnessed severe foetal distress on the foetal heart rate monitor after an ECV attempt, a run for a rapid CS. The foetus passed away. That is why I will never perform an ECV attempt myself.’).

Disagreement on the list of contra-indications in the guidelines was an issue among professionals as well. This means that they tend to use more contraindications than mentioned in the national guidelines (quotation: ‘I once advised against ECV for a patient with a tight belly, foetal growth on the 10th percentile, not too much amniotic fluid and anterior placenta localisation.

110 Chapter 6 Patients’ and professionals’ barriers and facilitators

A doctor should always consider if ECV is safe enough, taking more factors into account then just the list of absolute contra-indications.’).

Some professionals tend to counsel very directive as they were convinced that every client should undergo ECV. Directive counselling was seen as a facilitator.

Domain 3: characteristics of the patientThe professionals mentioned a variety of characteristics of patients which were classified as barriers in order to undergo ECV. First of all, according to professionals not all patients are convinced of the need for ECV and overestimate the chance of a spontaneous version. A second barrier mentioned by professionals was the fear of the patient to the ECV procedure itself. ECV is in conflict with the nature of pregnant women to protect their belly. Not only fear to an unknown procedure, but fear in general was seen as a barrier. Third, professionals assumed that uptake of ECV would decrease if also the complications of ECV, which are rare, were mentioned during counselling. Fourth, the patients’ attitude towards a CS might influence a patient’s choice to opt for ECV. According to the professionals most women focus on short term outcomes and therefore underestimate the risks of a CS. Furthermore, due to the decreased number of children per woman over the last decades, the argument of risks of CS for consecutive pregnancies is considered less relevant as women state to have their last child.

Explanation of risks of ECV was also mentioned as a barrier as some professionals were afraid to frighten the patient (quotation: ‘Especially if it is her last pregnancy, it is even more needless to overcome a CS in her opinion.’).

A facilitator mentioned by the professionals was a good professional–patient relationship. A last factor mentioned was cultural background (other religion, birth rituals, believes etc.) of the patient which could both positively or negatively influence the willingness to undergo ECV.

111 Chapter 6 Patients’ and professionals’ barriers and facilitators

Domain 4: characteristics of the contextHealth-care providers thought there was a lack of adequate patient information materials and mentioned this as a barrier. The time needed for adequate counselling was mentioned as a barrier by all professionals. Lack of reimbursement for the counselling for ECV was mentioned as a barrier for the midwives. In general, a local transparent policy about informing, counselling, referral and performing ECV is considered a facilitator for all professionals (quotation: ‘I am not sure if the professional who performs ECV has the same attitude towards ECV and in- forms the patient in a similar way. I hardly get any feedback of the results.’). And so were standardized information leaflets and the centralisation of ECV attempts in specialized office hours and narrowing the number of professionals involved in ECV in order to increase expertise.

112 Chapter 6 Patients’ and professionals’ barriers and facilitators

Table 3. Barriers (b) and facilitators (f) of ECV according to professionals

Domain 1: Characteristics of ECV procedure Domain 2: Characteristics of the professional Patient information Self -‐efficacy -‐ Lack of adequate patient information materials (b) -‐ Inability to get the message across, meaning difficulties -‐ Written information about ECV (f) explaining the advantages of ECV if couples opt for -‐ Uniform patient leaflet of both organizations primary CS (b)* (hospitals and midwifery practices) (f) -‐ Inability to explain safety of ECV to patients (b)* Procedure -‐ Inability to explain risks of planned CS (b) -‐ Shortage of evidence of safety of ECV (b) Lack of knowledge -‐ Monitoring fetal condition (f)* -‐ about ECV procedure (b) -‐ about success rate (b) -‐ about guidelines and literature of ECV (b) -‐ about complication risk during counselling (overestimation) (b) -‐ Shortage of evidence of safety of ECV (b) Skills/ expertise -‐ Lack of routine in ECV by professional who is supposed to do the ECV (b)* Attitude -‐ Negative attitude towards ECV (b) -‐ Not convinced of importance of ECV (b) -‐ Directive counselling of ECV (f) Outcome expectations/ perceived risk -‐ In daily practice more contra indications compared to guideline (b) -‐ Having witnessed of severe complication of ECV (b)

Domain 3: Characteristics of the patient Domain 4: Characteristics of the context Need for ECV Logistical/ clear procedures -‐ Unawareness of benefits ECV (b) -‐ Regional organisation of care (b) -‐ Cultural background and education (b/f) -‐ Transparency of patient care within the region (f) -‐ Client’s wish to avoid delivery vaginal in general -‐ Explaining ECV takes time (b)* (b) -‐ Subjective information sources (internet, family)(b)* Fear of ECV procedure Available time -‐ ECV is in conflict with the nature of pregnant -‐ Shortage of time in counselling (b)* woman to protect her belly (b) Reimbursement -‐ Fear of losing control (b) -‐ Discrepancy between investment of time and financial -‐ Fear of harm to the fetus (b) support (b) Prevention of CS Collaboration professionals -‐ Patient focus on short term and not on long term -‐ Local consensus about referral policy (f) outcome (b)* -‐ Local consensus about location ECV in and out clinic Outcome expectations/ perceived risk setting (f) -‐ Trivializing risks of ECV (they think: ‘it will not -‐ Transparency of adherence to guideline within the happen to me’. (f)* region (f) -‐ Underestimating risk CS (b) -‐ Offering ECV in specialized center with specialized -‐ Decreasing number of children per woman (b)*^ trained midwives and/or gynaecologists(f)* -‐ Lack of confidence in success of ECV(b)^^ * means the statement is made by more than 50% of the professionals ^ nowadays, the number of children per average woman is on 2, women are less prepared to take risks during their pregnancies and choose more often for safety ^^ if a patient has a confidential relationship with her midwife or gynaecologist, she has more trust in the skills of professional and is illing more w to choose for ECV

113 Chapter 6 Patients’ and professionals’ barriers and facilitators

Barriers and facilitators related to the implementation of ECV according to patients

Patients identified 27 barriers and facilitators for ECV. Most factors were scored under the domain ‘characteristics of the patients’. Compared to the professionals, the patients came up with more facilitators compared to the professionals.

Domain 1: characteristics of the intervention Patients listed pain as a barrier to opt for an ECV attempt. Facilitators were the verification of the fetal condition, the low complication rate (chance of emergency CS below 1%) and mean success rate of 40% (quotation: ‘If the doctor tells me the foetal condition is all right and I have a chance of success of 40% or more to give home birth (as I want), why should I not choose an ECV?’). Lack of information leaflets in different languages made it hard for non-Dutch women to understand the need for ECV (quotation ‘Unfortunately my caregiver did not have an information leaflet in English. I wished she had, because then I should understand better what an ECV was and why it was done. Now I asked a friend of mine to tell me more about ECV.’).

Domain 2: characteristics of the professional Women commented upon the information provided by the health-care professionals. The barriers they mentioned were: the incompleteness of information, conflicting information from different professionals, and a too directive attitude of the professionals, and the misunderstanding of the information they received from the professional, possibly caused by a linguistic barrier between the professional and the patient. Also, some mentioned that the professional emphasised the potential complications of ECV in their opinion (quotation: ‘My professional told me I had to go to the hospital, but I did not understand why. She told me that if the baby was not turned, I could not deliver. I did not understand I had the choice to undergo an ECV attempt or not.’).

A significant proportion of women needed more time than provided to make a decision; the professional gave too much information at once to com- prehend. Patients appreciated involvement of their partner in the decision making process (quotation ‘I was very pleased that my caregiver asked my partner if he understood the process of ECV, the importance of ECV and his opinion to undergo an ECV attempt.’).

114 Chapter 6 Patients’ and professionals’ barriers and facilitators

Domain 3: characteristics of the patientFear was an important barrier, consisting of fear of harm to the fetus during ECV (quotations ‘I was very afraid of a premature birth as I had read on the internet that labour could start after ECV. I asked the midwife if ECV could be done later on in pregnancy. The midwife explained that the baby had not descended yet and that the chance of a preterm birth was very small, but still I was very worried.’ and ‘There are complications connected and it is painful, even though they tell you there are not. I’ve seen enough. And I did not want to disturb the child, it is how it is. And I had no objection to a caesarean. I was well aware of the advantages and disadvantages of a vaginal birth and CS and after I had considered all options, I have chosen to opt for an elective CS.’).

Some women stated that their fear for a CS, or at least the wish to avoid an operation and following longer recovery compared to vaginal delivery), was the incentive to opt for ECV. On the other hand women regularly told they preferred a planned CS in case of breech position to prevent them from pain of a vaginal delivery. ECV was seen as a barrier for not having a planned CS and a more easy way of delivery, according to the patient.

Women, who felt more confident about the chance of success of ECV, were more willing to undergo ECV. Other facilitators were the reassurance that they were in control of stopping the procedure if it would be too painful, and their confidence in the skills of the professional.

Whether a patient is willing to consider ECV is influenced both positively and negatively by cultural background and religion. A barrier was the argument ‘there must be a higher reason why this baby is in breech and therefore, I must accept this.’ Or ‘it is God’s will that the baby is in breech, I have to respect that’.

115 Chapter 6 Patients’ and professionals’ barriers and facilitators

Domain 4: characteristics of the context Since breech presentation is often diagnosed at regular visits earlier in pregnancy, women would also like to be informed earlier (despite the high chance of spontaneous version). The short time between information to decision was mentioned as a barrier by approximately half of the women (quotation ‘I would appreciate if I was told about ECV in an earlier stage. I knew my baby was lying with his head up from 32 weeks onwards. I was informed about ECV around 36 weeks and I had to decide if I would undergo an ECV at the same day my professional told me my baby was in breech position. She made an appointment for the ECV attempt as she did not do it by herself and I could come immediately. Afterwards I was so confused about everything I underwent that day, that I should not choose an ECV a second time. Only if I get more time between the information and the ECV attempt, I will choose ECV another time.’).

Patients often explained that opinions and experiences of friends and family played an important role in their decision making. If a friend/family member had witnessed birth complications or perinatal death, even if there was no relation to breech or ECV, it might be the argument for a patient to not undergo ECV.

On the other hand patients also told that the support of family and friends and the shared decision making were important facilitators for ECV (quotation ‘Well actually I was not totally convinced about ECV, but I have a niece who is pregnant too and when I told her I had also a breech position she told immediately to do an ECV. On the contrary many of my aunts told me not to do an ECV attempt as it was outdated. In the end I decided to listen to my body and to my niece and I opted for an ECV. I was lucky; it succeeded and now I can have a normal birth.’).

All patients who participated in the interviews were asked if travel distance to a specialist in ECV was a facilitator or barrier. Almost all patients stated that they were willing to travel up to 1.5 hours if the specialist had high success rates.

We also asked if patients preferred a familiar health-care provider. Again, this was secondary to the expertise of the professional.

116 Chapter 6 Patients’ and professionals’ barriers and facilitators

Table 4. Barriers (b) and facilitators (f) of ECV according to pregnant women with a breech presentation at 35 weeks gestation onwards

Domain 1: Characteristics of ECV procedure Domain 2: Characteristics of the professional Patient information Self-‐efficacy -‐ Lack of information in foreign languages (b) -‐ Inability to counsel patient for ECV (b) -‐ Involving partner into procedure of ECV (f) -‐ Involving partner into procedure of ECV (f) Procedure Lack of knowledge -‐ Painful treatment (b)* -‐ Professionals emphasize complication risk -‐ Complication risk (risk of CS<1%) (f) factors (b) -‐ Success rate 40% (f) -‐ Incomplete information to patient (b) -‐ Verification of fetal condition (CTG and ultrasound) (f) Skills/ expertise -‐ Inability of providing information or counselling in foreign language (b) Attitude -‐ Professional’s attitude for counselling for ECV, for the procedure of ECV (f) Outcome expectations/ perceived risk -‐ Underestimation of ECV success rates (b)* Domain 3: Characteristics of the patient Domain 4: Characteristics of the context Need for ECV Timing of informing/ counselling -‐ Misunderstanding reason for ECV (b) -‐ Early information in third trimester (f)* -‐ Language barrier (b) -‐ Short period between diagnosis and Fear of ECV procedure procedure (b)* -‐ Assurance of being in control of termination of procedure (f) Social background -‐ Confidence in skills of the (f) professional -‐ Supporting social network (partner, family, -‐ Irrational fear of physical harm to the fetus (b)* friends) (f)* Anticipated patient co-‐operation -‐ Complications of pregnancy within social -‐ Wish to stay in total control of pregnancy (b) network (even if not related to breech Prevention of CS presentation) (b) -‐ Wish to vaginal (home) birth (f) -‐ Fear for vaginal breech delivery or CS (f)* -‐ Underestimating risks of CS (f) Outcome expectations/ perceived risk -‐ Confidence of reasonable chance of success procedure (f) -‐ Believe in natural reason or God of breech presentation (b) * means the statement is made by more than 50% of the patient

117 Chapter 6 Patients’ and professionals’ barriers and facilitators

DISCUSSION

We identified a wide variety of barriers and facilitators influencing the implementation of external cephalic version in breech position. Among professionals the main barriers were a lack of knowledge to fully inform and counsel patients on ECV, and the inability to counsel women who preferred a primary CS. The main facilitators were the monitoring of the fetal condition, and an unambiguous policy in which ECV is offered in specialized centres with specialized trained midwives and/or gynaecologists. The main reasons for not wanting ECV for more than 50% of the interviewed women were fear of physical harm to the fetus and a painful treatment. The main facilitators were fear for a vaginal breech delivery or CS, a supporting social network (partner, family and friends) and early information in the third trimester.

Both patients and professionals identified the lack of (use of) information leaflets and incomplete verbal information as a barrier. Objective, uniform information might also be an important tool to take away fear of pain, fetal injuries, and the fear to lose control, but health-care providers need to be aware of these barriers in order to help the patient. The lack of adequate patient information is a common barrier, reported in different other studies.14 The professional barriers concerning the difficulties of informing and counselling (especially with patients who already made up their minds) indicates a lack of self-efficacy, which is a common barrier in guideline adherence.15–17 Previous research among professionals in communication and counselling showed a better information transfer to patients in professionals who were trained in communication and counselling.18 An Australian study showed that patients using a decision aid could better express their choice for or against an ECV.19

Patients and professionals have a different view on barriers. Professionals’ assumptions about patient’s barriers are certainly not applicable to all patients. For instance, explaining complications of ECV was mentioned asa barrier, because professionals assumed that this might deter patients. But the majority of patients considered ECV to be safe even if the complications were emphasised during counselling. Thus this presumed barrier should not withhold professionals to properly and thus fully inform patients on the risks and benefits of ECV. Also, it is known that people in general tend to underestimate potential risks of their behaviour or the situation they are in (‘there are risks, but it will not happen to me’).20 118 Chapter 6 Patients’ and professionals’ barriers and facilitators

A difficult issue is the opinion of professionals that patients underestimate the risks of a CS. The patients we have interviewed and who opted for a planned CS were fully aware of the complications. Also, a preference study on mode of delivery among couples with a baby in breech presentation showed that women tend to choose the safest option for their baby, while their partners tend to choose the safest option for the mother .21 In the end, it is not very likely that all women will opt for ECV if they are fully informed; our interviews showed that women might choose on less rational but very emotional grounds if for instance they have witnessed major pregnancy or birth complications. The group of women who opt for a planned CS instead of ECV because of fear for a vaginal delivery in general, is a group for which very good counselling skills are needed to diminish this fear. But this is also the group to focus on in order to increase the number of ECV attempts.

We realise there are some limitations in this study that should be considered. The numbers of health-care providers and patients in the groups were small and representativeness of the identified facilitators and barriers need to be quantified in a questionnaire among a significant sample of the professionals and patients. Especially as it was difficult to recruit patients rejecting ECV as this is the minority of patients counselled for ECV. The results of this study show that an implementation strategy needs to focus on training of professionals in objective counselling and information transfer to pregnant women about ECV and on the other hand focus on offering adequate patient information materials to pregnant women with a fetus in breech position. The patient information need to be timed earlier than provided nowadays, for instance from 32 weeks and beyond. With the knowledge obtained in this study, the next step is to design a tailored implementation strategy.

CONCLUSION

In summary, this study gives insight into the barriers and facilitators of the implementation of external cephalic version in the Netherlands. Knowledge of these factors may help to improve implementation of ECV in clinical practice and enlarge the number of pregnant women who choose ECV. The results show that an unambiguous policy on counselling plays an important role in improving the implementation of ECV.

119 Chapter 6 Patients’ and professionals’ barriers and facilitators

REFERENCES

1 Hofmeyr, GJ; Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012; 10.

2 Hannah ME, Hannah WJ, Hewson S a, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet [Internet]. 2000 Oct; 356: 1375–1383. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0140673600028403

3 Rietberg C. Term breech presentation in The Netherlands from 1995 to 1999: mortality and morbidity in relation to the mode of delivery of 33,824 infants. BJOG An Int J Obstet Gynaecol [Internet]. 2003 Jun [cited 2014 Mar 3]; 110: 604–609. Avail- able from: http://doi.wiley.com/10.1016/S1470-0328(03)01907-4

4 Rijnders M, Offerhaus P, Dommelen P Van, Wiegers T, Buitendijk S. Prevalence , Out- come , and Women ’ s Experiences of External Cephalic Version in a Low-Risk Popu- lation. 2010; 124–133.

5 Feitsma, HA; Middeldorp JM, Oepkes D. De uitwendige versie bij de a terme stuit; een inventariserend onderzoek (ECV for breech presentation at term; an inventory study). NTOG. 2007; 120: 4–6.

6 Fleuren M, Wiefferink K, Paulussen T. Determinants of innovation within health care organizations: literature review and Delphi study. Int J Qual Health Care [In- ternet]. 2004 Apr; 16: 107–123. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/15051705

7 Fleuren M, Grol R, de Haan M, Wijkel D, Oudshoorn C. Adherence by midwives to the Dutch national guidelines on threatened miscarriage in general practice: a prospective study. Qual Health Care [Internet]. 1997 Jun;6 : 69–74. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1055455&tool=pmcent rez&rendertype=abstract

8 Grol, R; Wensing, M; Eccles M. Improving Patient Care: The Implementation of Change in Clinical Practice [Internet]. Elsevier; 2005 [cited 2014 May 19]. p. 304. Available from: http://www.amazon.com/Improving-Patient-Care-Implementation- Clinical/dp/075068819X

120 Chapter 6 Patients’ and professionals’ barriers and facilitators

9 Guldbrandsson K. From news to everyday use The difficult art of implementation. Stockholm: Strömberg;

10 Bartholomew, LK; Parcel, GS; Kok, G; Gottlieb N. Planning health promotion programs: an intervention mapping approach. [Internet]. San Francisco: Jossey-Bass; 2011 [cited 2014 May 19]. Available from: http://www.amazon.com/s/ref=nb_sb_ noss?url=search-alias=aps&field-keywords=planning+health+care+promotions+bart holomew&rh=i:aps,k:planning+health+care+promotions+bartholomew

11 Fullan M. The New Meaning of Educational Change, Fourth Edition [Internet]. Teachers College Press; 4 edition; 2007 [cited 2014 May 30]. p. 352. Available from: http://www.amazon.com/Meaning-Educational-Change-Fourth-Edition/ dp/0807747653

12 Gearin, E; Kahle C. Focus Group Methodology Review and Implementation. Mid- wifery [Internet]. 2006 [cited 2014 May 30]; Available from: http://www.midwifery- journal.com/article/PIIS0266613813001046/references

13 Fleuren, M; Wiefferink, K; Paulussen T. Checklist determinanten van innovaties in de gezondheidszorgorganisaties. [Checklist determinants of innovations in health care organisations]. TSG [Internet]. 2010 [cited 2014 May 19];88 : 51–54. Available from: http://kic.nisb.nl/site/catalogus/show/8402

14 Van den Boogaard NM, van den Boogaard E, Bokslag a, van Zwieten MCB, Hompes PG a, Bhattacharya S, et al. Patients’ and professionals’ barriers and facilitators of tailored expectant management in subfertile couples with a good prognosis of a natural conception. Hum Reprod [Internet]. 2011 Aug [cited 2014 May 30]; 26: 2122–2128. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21665873

15 Lugtenberg M, Zegers-van Schaick JM, Westert GP, Burgers JS. Why don’t physicians adhere to guideline recommendations in practice? An analysis of barriers among Dutch general practitioners. Implement Sci [Internet]. 2009 Jan [cited 2014 May 30]; 4: 54. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?arti d=2734568&tool=pmcentrez&rendertype=abstract

16 Haagen EC, Nelen WLDM, Hermens RPMG, Braat DDM, Grol RPTM, Kremer J a M. Barriers to physician adherence to a subfertility guideline. Hum Reprod [Internet]. 2005 Dec [cited 2014 May 30]; 20: 3301–3306. Available from: http://www.ncbi. nlm.nih.gov/pubmed/16055463

121 Chapter 6 Patients’ and professionals’ barriers and facilitators

17 Carbana MD, Rand CS, Powe NR, Wu AW, Wilson MH. Why Don ’ t Physicians Follow A Framework for Improvement. 2014; 0718.

18 Smith MD, McGhan WF, Lauger G. Pharmacist counseling and outcomes of smok- ing cessation. Am Pharm [Internet]. 1995 Aug [cited 2014 May 30];NS35 : 20–9; 32. Available from: http://www.ncbi.nlm.nih.gov/pubmed/7677052

19 Nassar N, Roberts CL, Raynes-Greenow CH, Barratt a, Peat B. Evaluation of a decision aid for women with breech presentation at term: a randomised controlled trial [ISRCTN14570598]. BJOG [Internet]. 2007 Mar [cited 2014 Mar 20]; 114: 325–333. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=240865 8&tool=pmcentrez&rendertype=abstract

20 Reynolds BJ. When the facts are just not enough: credibly communicating about risk is riskier when emotions run high and time is short. Toxicol Appl Pharmacol [Inter- net]. Elsevier B.V.; 2011 Jul 15 [cited 2014 May 30]; 254: 206–214. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21034761

21 Kok M, Gravendeel L, Opmeer BC, van der Post J a M, Mol BWJ. Expectant parents’ preferences for mode of delivery and trade-offs of outcomes for breech presentation. Patient Educ Couns [Internet]. 2008 Aug [cited 2014 Mar 3];72 : 305–310. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18534810

122

Facilitators and barriers to external cephalic version for breechpresentation at term among health care providers in the Netherlands: A quantitative analysis

Ageeth N. Rosman

Floortje Vlemmix

Antje Beuckens

Marlies E. Rijnders

Brent C. Opmeer

Ben Willem J. Mol

Marjolein Kok

Margot A.H. Fleuren

MIDWIFERY. 2014 MAR;30(3):E145-50

125 Chapter 7 Facilitators and barriers to external cephalic version

ABSTRACT

ObjectiveGuidelines recommend that external cephalic version (ECV) should be offered to all women with a fetus in breech presentation at term. How- ever, only 50–60% of the women receive an ECV attempt. We explored the determinants (barriers and facilitators) affecting the uptake of the guidelines among gynecologists and midwives in the Netherlands.

Design National online survey.

Settingthe Netherlands.

ParticipantsGynecologists and midwives.

Measurements In the online survey, we identified the determinants that positively or negatively influenced the professionals adherence to three key recommendations in the guidelines: (a) counselling, (b) advising for ECV, (c) arranging an ECV. Determinants were identified in a previously performed qualitative study and were categorised into five underlying constructs; -at titude towards ECV, professional obligation, outcome expectations, self-efficacy and preconditions for successful ECV. We performed a multivariate analysis to assess the importance of the different constructs for adherence to the guideline.

Findings 364 professionals responded to the survey. Adherence varied: 84% counselled, 73% advised, and 82% arranged an ECV for (almost) all their clients. Although 90% of respondents considered ECV to be an effective treatment for preventing caesarean childbirths, only 30% agreed that ‘every client should undergo ECV’. Self-efficacy (perceived skills) was the most important determinant influencing adherence.

Key conclusions Self-efficacy appears to be the most significant determinant for counselling, advising and arranging an ECV.

Implications for practiceTo improve adherence to the guidelines on ECV we must improve self-efficacy.

126 Chapter 7 Facilitators and barriers to external cephalic version

INTRODUCTION

Clinical problem

Breech presentation occurs in 3–4% of term pregnancies, corresponding to 7450 women annually in the Netherlands. The Term Breech Trial, which compared planned vaginal childbirth to planned caesarean childbirth, showed a significant reduction in poor neonatal outcome and mortality.1 The results of this study had a major impact on the management of the term breech childbirth. The overall caesarean childbirth rate for breech presentation in the Netherlands increased from 50% to 80%.2 This change was accompanied by a substantial decrease in perinatal mortality of breech pregnancies from 0.38% to 0.18% (OR 0.53; 95%CI 0.33–0.83) and neonatal trauma (OR 0.26; 95%CI 0.14–0.50).3

The increased number of caesarean sections (CS) has disadvantages as well: caesarean childbirths are associated with increased maternal morbidity, longer hospital admission and consequences for future pregnancies (increased risk of abnormal placental implantation, uterine rupture and, as an ultimate consequence of these complications, fetal death). External cephalic version (ECV) is a safe obstetrical intervention that has been proven to reduce the number of breech presentations at birth and therefore reduction of the number of caesarean childbirths.4 ECV is worldwide recommended in obstetrical guidelines as the first treatment of choice in case of breech presentation at term, with reported success rates of 40–50%.5

The Royal Dutch Organisation for Midwives (KNOV) and the Dutch Society for Obstetrics and Gynaecology (NVOG) published evidence-based guidelines on the management of women with a fetus in breech position.6,7 According to these guidelines ECV should be advised to all eligible women with a fetus in breech position at 36 weeks and onwards. The three key recommendations in the guidelines are: (1) to counsel all women with a fetus in breech presentation. This means supplying information and taking the woman’s and her partners perspective into account so they are able to make an informed choice; (2) advising an ECV and (3) arranging an ECV; either by performing ECV or referring to a colleague who is experienced in performing ECV.

Irrespective of these guidelines, not all women are offered an ECV. An inventory survey among all hospitals in the Netherlands in 2007 reported that

127 Chapter 7 Facilitators and barriers to external cephalic version

5% of the gynaecologic practices did not perform nor referred for ECV at all, even though women with breech presentations were referred to these hospitals as well.8 A prospective cohort study in the Netherlands reported that 26% of women with a fetus in breech presentation did not undergo ECV; in 48% of these cases, the obstetrician decided not to perform ECV, 37% of women declined a version attempt, and 15% gave birth before the version was performed.9 This means that a substantial proportion of clients are not receiving the intended care in a way that they benefit from the guidelines. As a result, these women will have a breech presentation at birth and most of them will have a caesarean childbirth with all its consequences. As there is sound evidence supporting an ECV in case of breech presentation, it is clear that the problem relates to implementation.

Framework of implementation research

One of the main problems with the introduction of guidelines in the health care system is that professionals do not ‘automatically’ use the guidelines as intended by the developers. 10,11 A detailed understanding of the factors, or so-called determinants, that facilitate or impede the innovation process is a prerequisite for designing an innovation strategy that is adapted to the several critical determinants, in order to achieve real change.12,13 The framework used in the present study represents the main stages in innovation processes and related categories of determinants. Each of the four main stages in innovation processes (dissemination, adoption, implementation, and continuation) can be seen as points at which, potentially, the desired change may not occur. The transition from one stage to the next can be affected by various determinants. This framework is more extensively discussed elsewhere.12

To identify potential determinants for (non)adherence to the guidelines on breech presentation, focus group interviews were conducted with midwives and gynaecologists.14 This resulted in a list of 43 potential determinants, that was further reduced to a shortlist of key determinants, as recognised by implementation experts.12

The aim of this study was to quantify the determinants (facilitators as well as barriers) midwives and gynaecologists perceive in adhering to the three key recommendations in the Dutch ECV guidelines: (a) counselling all eligible women for ECV, (b) advising positively and (c) arranging an ECV.

128 Chapter 7 Facilitators and barriers to external cephalic version

METHODS

Setting

In the Netherlands, obstetric care is organised in regions. A region contains a hospital and several surrounding midwifery practices who initially refer to this hospital. In total, there are 97 hospitals and thus regions, and 503 midwifery practices in the Netherlands.

Study design

The study was designed as a survey among gynaecologists and midwives. As both professions are organised differently, we had a different approach for each. All 1,217 gynaecologists and residents were sent an invitation by e-mail to participate in the online questionnaire. It was not possible to limit the invitations to obstetrical oriented gynaecologists. However, the majority of gynaecologists in the Netherlands are actively involved in obstetrics. There is no mass e-mail listing for midwives, thus a random sample of addresses of 300 midwifery practices were sent an invitation to participate in the online questionnaire. To avoid missing data in the online questionnaire, the questionnaire could not be finished when there were missing answers. If participants wished to explain their answers in detail, they could do so at the end of the questionnaire.

The general outline of the questionnaire was derived from the qualitative determinant analyses.14 Table 2 shows the variables that were measured. First, the potentially relevant determinants of adherence to the three key recommendations were measured: attitude towards EVC (10 items); professional obligation (eight items); outcome expectations (four items); self-efficacy (four items) and preconditions for successful ECV (15 items). Profession- al obligation refers to the degree to which the guideline recommendations fit in with the tasks for which the user feels responsible when doing his/ her work .15 Self-efficacy refers to the perceived competence of users with respect to intended behaviour.16,17

129 Chapter 7 Facilitators and barriers to external cephalic version

Table Ia. Characteristics of respondents; gynaecologists

Gynecologists (%) Number of respondents 298 Gynecologists 203 (68) Residents 69 (23) Unknown 26 (9) Subspecialty of gynaecologists Perinatology 75 (37) Uro-‐gynaecology 28 (14) Fertility 28 (14) Oncology 12 (6) General gynaecologist 57 (28) Unknown 3 (1) Setting Teaching Hospital 270 (91) Other 28 (9) Training in ECV During residency 137 (68) During work as gynaecologist 15 (7) Training of NVOG 15 (7) Other 35 (17) Experience with ECV ECV performed at least once in the past year 255 (85)

Outcome expectations refer to the user’s perceived probability of achieving the client objectives as intended by the guidelines.15 For all items, 5-point Lik- ert scales were used, ranging from ‘totally agree’ to ‘totally disagree’, except for self-efficacy (4-point scale, ranging from ‘feeling totally able to perform’ to ‘feeling totally unable to perform’). The self-reported level of adherence was measured at the level of the three key recommendations in the guidelines (adherence to counsel, advising, and arranging ECV). The respondents were asked to indicate, for each key recommendation, for how many patients they had implemented the activity (7-point Likert scale, ranging from ‘none’ to ‘all’). These three adherence items were used as outcome measures of our analyses. Finally, relevant background variables were assessed, such as type of respondent (consultant gynaecologist/obstetrics, resident, and midwife), type of hospital or practice, and experience with ECV (Table 1).

130 Chapter 7 Facilitators and barriers to external cephalic version

Table Ib. Characteristics of respondents; midwives Midwives (%) Number of respondents 66

Type of practice Solo practice 8 (12) Group practice 54 (82) Temporary helper 4 (6)

Specialist in ECV 3 (5)

Experience with ECV ECV performed at least once in the past year 11 (17)

Data analyses

The items were a priori organised as to reflect clinically plausible constructs related to ECV. We used principal components analyses (PCA) to identify statistical constructs of items based on the inter-item correlations. Items -con tributing <0.50 to the construct were considered not to contribute to the construct and were therefore removed. A Cronbach's alpha of more than 0.70 indicates good reliability.18 If these statistical constructs were found to be consistent with the clinical constructs or otherwise show good clinical face validity, their scale properties were assessed and improved using reliability analysis (internal consistency, inter-item correlation) of the items for each construct. Otherwise the original constructs were used for subsequent scale construction.

For each construct, a summary score reflecting respondents' ratings on that construct was obtained with principal components analysis (PCA). PCA is a way of identifying patterns in data, and expressing the data in such a way as to highlight their similarities and differences. As patterns in data can be hard to find in data of high dimension, PCA is a powerful tool for analysing data.

We explored the extent to which these constructs – the determinants – were associated with the three adherence outcomes. First we took for each construct separately (univariate analyses), and subsequently explaining the outcomes by taking those constructs with a p-value below 0.10 from the univariate model, all into account (multivariable way) to assess their combined association with the outcomes. 131 Chapter 7 Facilitators and barriers to external cephalic version

Table II: Adherence and determinants in adhering to the ECV guideline in daily practice

% of gynaecologists % of midwives agree agree A. Attitude towards ECV1 1. Every client* should be counselled 96 99 2. During counselling should be explained that ECV is the best choice for the client 84 77 3. Every client should undergo ECV 30 22 4. Uterus relaxants increase the success rate of ECV 66 29 5. Uterus relaxants should be offered to all clients 49 25 6. Clinical experience of the professional is leading in the decision to offer a client an 32 22 ECV 7. ECV is a good treatment to increase the number of cephalic presentation at birth 94 69 8. Every client with a failed ECV should be offered a second attempt 28 97 9. ECV is a good treatment to decrease the number of caesarean sections due to 91 94 breech presentation 10. Knowledge of guidelines on ECV improves counselling 91 86 B. Task orientation1 11. It is my responsibility to counsel every client 92 85 12. It is my responsibility to convince every client to opt for ECV 73 66 13. It is my responsibility to refer to the professional with the highest success rate in 52 66 the region 14. It is my responsibility to take care of an unambiguous policy within my hospital / 88 95 practice 15. It is my responsibility to take care of an unambiguous policy within my region 76 59 16. It is my responsibility to discuss reasoning when colleagues do not offer ECV 77 69 17. It is the task of gynaecologists to decide on contra indications of ECV 84 28 18. It is the task of special trained midwives to decide on contra indications of ECV 48 77 C. Outcome expectations1 19. Clients are more content if ECV is always offered 73 82 20. Counselling increases the number of clients opting for ECV 81 86 21. An unambiguous policy increases the number of ECV attempts 80 80 22. Performing ECV leads to more cephalic presentations at birth 93 97 D. Preconditions of successful ECV1 23. Counselling should contain information on breech birth; vaginal vs. caesarean 90 88 section. 24. ECV is safe in an out of hospital setting, under precondition of good agreements 7 77 with the hospital in case of complications 25. To improve quality of care, ECV should be performed in specialized centres. 73 80 26. Registration of ECV in a the nationwide obstetric database is necessary for a good 93 92 evaluation 27. To offer all clients ECV, compromises on reference between midwives and 92 88 gynaecologists is essential 28. Special finance is necessary for counselling 39 78 29. 10 ECV attempts per necessary year are to maintain expertise 64 71 30. Success rates of ECV increase when special office hours are reserved 72 65 31. Success rates of ECV increase when performed in specialized centres 44 52 32. Complication rates of ECV can be reduced specialized when performed in centres. 29 37 33. If ECV could only be performed in hospitals, the number of ECVs decreases 8 23 34. Feedback of results of ECV and consequences of care to the referrer, increases 72 54 number of clients counselled 35. If referrers have a better perception of the expertise and success rates of the 75 72 specialists, more women are referred for ECV. 36. There is shortage of time to counsel clients properly in daily practice 24 56 37. Current cooperation in the region hinders me in offering ECV to clients 6 8 E. Self-‐efficacy2 38. To what extend do you feel able to counsel all clients? 64 51 39. To what extend do you feel able to explain that ECV is the best choice? 46 35 40. To what extend do you feel able to explain the safety of ECV? 41 35 41. To what extend you do feel able to plan all clients opting for ECV an attempt? 55 55 Adherence to guidelines (dependent variables)3 A. How many clients do you counsel? 86 74 B. How many clients do you convince of the 72 74 advantages of ECV? C. For how many clients do you succeed in arranging an ECV attempt? (Either 82 80 referring or performing it yourself) *Client; women with breech presentation at term eligible for ECV 1Percentage of respondents who filled in ‘agree’ to ‘strongly agree’. 2Percentage of respondents who filled in ‘feeling totally able to perform’ 3Percentage of respondents who filled in ‘almost all’ to ‘all’

132 Chapter 7 Facilitators and barriers to external cephalic version

To evaluate whether the association between a determinant and the adherence outcomes was different for gynaecologists and midwives, also interaction variables for all constructs with specialism were included in the multivariable analyses. A backward selection method was applied, and as- sociations with p<0.05 were considered significant. Standardised betas and p-values were reported to quantify the relation between the constructs and the outcome. Statistical Products Services and Solutions (SPSS), version 17.0.1, Chicago, IL, was used for the statistical analyses.

FINDINGS

Respondent characteristics

The survey was online accessible from February 2010 until April 2010. Of 1217 gynaecologists and residents initially contacted, 281 (23%) completed the online survey. The majority (91%) of gynaecologists and residents were working in a teaching hospital and 75 (37%) gynaecologists were specialised in obstetrics. More than half of the gynaecologists were trained in ECV during their residency and 238 (85%) had performed one or more ECV attempts in the past year. Of 300 midwifery practices 66 representatives completed the online survey for their own practice. The vast majority worked in a group practice. Eleven (17%) had performed one or more ECV attempts in the past year. The results of the questionnaire are presented in Table 1. If more than 80% of respondents agreed with an item, it was considered to be a significant facilitating or impeding factor. Table 2 ��presents the respondents'�� characteristics.

Perceived facilitating determinants

Over 90% of professionals acknowledged that ECV is an effective treatment to decrease the number of CS due to breech presentation (Table 2, item 9). The majority of the respondents (over 90%) agreed with the statement ‘every client with a fetus in breech presentation should be counselled for ECV’ and felt it was their responsibility to counsel clients (items 1 and 11). Eighty per cent agreed that counselling would increase the number of women undergoing an ECV (item 20) and thought that knowledge of the guidelines improves counselling (items 10). To be able to offer all clients ECV, good working arrangements between gynaecologists and midwives regarding referral of patients was thought of to be essential (91% of respondents, item 27).

133 Chapter 7 Facilitators and barriers to external cephalic version

Perceived impeding determinants

One-third agreed on the statement ‘every client should undergo ECV’ and only two-third felt responsible to convince the client to opt for ECV (Table 2, items 3 and 12). Just 35–46% of the respondents felt able to successfully explain the importance and safety of ECV to all clients (items 39 and 40). Only 55% of respondents felt able to plan an ECV attempt for all of their clients (item 41). Regarding success rates of ECV, midwives and gynaecologists disagreed on the usefulness of a second version attempt. Only a small majority felt responsible to refer to the professional with the highest success rate within their region (items 8 and 13). Time and lack of finance (for the extra invested time) to inform a client on ECV was a barrier for more than half of the respondents, thereby reported significantly more often by midwives (56% versus 78%) as compared to gynaecologists (24% versus 39%) (items 28 and 36).

Adherence to key recommendations of the guidelines

Adherence to the guidelines is expressed in three key recommendations (Ta- ble 2, items A, B, and C). Of all respondents, 84% stated to counsel all of their clients, and 72% reported to positively advice (almost) all of their clients to opt for an ECV. Over 82% of respondents reported to successfully arrange an ECV attempt for (almost) all of their clients after the client opted for an ECV.

Scale construction

The factors obtained with principal components analyses only partially confirmed our a priori constructs, and also no clinically plausible alternative interpretation could be given to these factors. Face validity of the constructs is essential, as the results need to be translated back to clinical practice; scale construction was therefore based on our initial classification of items. Scale properties were further optimised using reliability analysis. In one construct (attitude), two items (6 and 8) were removed because this substantially improved Cronbach's alpha. Concerning the development of the constructs, the generally recommended critical level of Cronbach's alpha of >0.7 was reached in three of five constructs, but were >0.6 in the other two constructs. This is considered acceptable for newly developed scales and was therefore accepted by the researchers.

134 Chapter 7 Facilitators and barriers to external cephalic version

Determinants predicting the professional’s adherence to the guidelines

We explored whether the five constructs were associated with adherence to guidelines in the univariable and multivariable regression analysis (Table 3). Specialism (gynaecologist working in a hospital versus midwives working in an out of hospital setting) was taken into account as well.

Within the multivariate analysis, midwives had a better reported adherence to the key recommendation ‘advising ECV’ than gynaecologists (unstandardized beta 0.35, p-value 0.004). Attitude towards ECV and a positive self-efficacy contributed evenly. The reported adherence to ‘counselling’ was better among gynaecologists (unstandardized beta −0.28, p-value <0.0001) and was positively influenced by an increased self-efficacy as well. Adherence to arranging an ECV attempt was only influenced by a positive self-efficacy.

We evaluated whether the relation between the five constructs and the three outcomes differed among the gynaecologist and midwives by taking interaction variables into account. None of these resulted in statistically significant in the multivariate analyses of the three outcomes.

DISCUSSION

The results of our national survey show that Dutch gynaecologists and midwives perceive multiple barriers and facilitators for adherence to the guidelines on ECV in clinical practice. The vast majority of professionals consider ECV as a good treatment to prevent CS, agrees that all clients should be offered the treatment and think that counselling could improve the number of clients opting for ECV. However, only two-third of the professionals consider convincing clients to undergo ECV as their task.

The perceived self-efficacy to perform the key recommendations is rather low. The low agreement on the self-efficacy item of ‘successful planning of an ECV attempt’ could indicate that referral for an ECV attempt not necessarily leads to an ECV attempt. The most remarkable finding regarding the construct ‘attitude’ is the low agreement on the item ‘every client should undergo ECV’. An explanation might be found in the free-text comments, where many respondents argued that not all clients are eligible due to contra indications. The introduction of the questionnaire explained that ‘every

135 Chapter 7 Facilitators and barriers to external cephalic version

1 value -‐ <0.000 P

-‐

0.28 Unstan analyses Multivariable dardized beta dardized

-‐

8 P 001 <0.0 0.00 0.91 0.12 0.59 0.87 value

Adherence to arrange ECV attempt arrange to Adherence

0.01 0.02 0.13 0.28 0.08 0.03 -‐ -‐ dardized beta dardized -‐ Univariable analyses Univariable Unstan

-‐

P 001 <0.0 0.03 value

-‐

0.28 0.32 -‐ Unstan analyses Multivariable dardized beta dardized

1 1

value 0.65 0.06 -‐ 0.001 0.006 <0.000 <0.000 P Adherence to counsel ECV counsel to Adherence

dardized dardized -‐ 0.03 0.38 0.18 0.22 0.12 0.33 beta -‐ -‐ Univariable analyses Univariable Unstan

-‐

4 P 001 001 <0.0 <0.0 0.00 value -‐

stand -‐ 0.28 0.29 0.35 Un beta ardized analyses Multivariable

value 0.92 0.16 -‐ P <0.0001 <0.0001 <0.0001 <0.0001 Adherence to advise ECV advise to Adherence -‐

stan -‐ 0.01 0.37 0.28 0.33 0.28 0.07 beta -‐ Univariable analyses Univariable Un dardized dardized

Relationship of the professions and constructs of barriers/facilitators to advising, counselling, and arranging ECV advising, to barriers/facilitators of constructs and professions the of Relationship

efficacy

-‐

e III: l 5. Self 5. 3. Outcome expectations Summary score of constructs 1. Attitude towards ECV 2. ECV as task orientation 4. Preconditions of successful ECV successful of Preconditions 4. Outcome variables Specialism; gynaecologists or midwives Tab 136 Chapter 7 Facilitators and barriers to external cephalic version client’ in the statements referred to ‘every eligible client’ according to contra indications mentioned in the guidelines. Respondents mentioned that in clinical practice additional contra indications are used as compared to those reported in the guidelines. Most examples did not reflect absolute contra indications but (combinations of) clinical factors related to success rate of ECV. Kok et al. have developed a clinical prediction model for the success rate of ECV (based on parity, estimated fetal weight, placenta localisation and amniotic fluid). They concluded that still 20–30% of women with the lowest predicted success rates had a successful ECV. Thus, in our opinion, low predicted success rates are no reason not to offer ECV.19 Organisational factors influencing the adherence to the guidelines were evaluated in the construct ‘precondition of successful ECV’. Time and financial compensation to counsel clients for ECV were reported as a barrier particularly for midwives. To com- ply with this precondition required for a successful implementation of ECV, this barrier could be taken away by the health insurance companies.

The most important determinant for adherence to all three key recommendations is self-efficacy; health care providers feel unable to perform the activities involved in the guidelines due to lack of counselling skills.

This study has some strengths and weaknesses. The importance of this study is that a quantitative assessment of barriers and facilitators enables us to develop an implementation strategy that has the potential to truly affect the adoption and implementation of the guideline and as a consequence, the number of ECV attempts.

The response rate to our online questionnaire might be topic of debate. In quantitative research, the response rate is inferior to the absolute number of respondents and more importantly, the representativeness of these respondents to the total cohort of gynaecologists and midwives. The high number of specialists in favour of ECV could insinuate unwillingness to respond of those health care providers who disagree with (the importance) of ECV. However, we do think this is likely based on the results of a national survey we had performed among all hospital and midwifery practices; there was no region where ECV was not performed at all. The reason we did invite all gynaecologists to participate in this study was a more pragmatic one: the Dutch Society of Obstetrics and Gynaecology was unable to make a random selection of their membership database, and therefore all members received the mass mail.

137 Chapter 7 Facilitators and barriers to external cephalic version

CONCLUSION

In conclusion, the vast majority of professionals consider ECV as a good treatment to prevent CS, agrees that all clients should be offered the treatment and think that counselling could improve the number of clients opting for ECV. However only two-third of the professionals consider convincing clients to undergo ECV as their task. Self-efficacy appears to be the most significant determinant for counselling, advising and arranging an ECV. Thus, improving self-efficacy is an important element of an implementation strategy to improve adherence to the guidelines on ECV.

138

Chapter 7 Facilitators and barriers to external cephalic version

REFERENCES

1 Hannah ME, Hannah WJ, Hewson S a, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet [Internet]. 2000 Oct;356 : 1375–1383. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0140673600028403

2 Rietberg C. Term breech presentation in The Netherlands from 1995 to 1999: mortality and morbidity in relation to the mode of delivery of 33,824 infants. BJOG An Int J Obstet Gynaecol [Internet]. 2003 Jun [cited 2014 Mar 3]; 110: 604–609. Avail- able from: http://doi.wiley.com/10.1016/S1470-0328(03)01907-4

4 Hofmeyr, GJ; Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012; 10.

5 Gj H, Gml G. Interventions to help external cephalic version for breech presentation at term ( Review ). Cochrane Libr. 2010;

6 Otterlo Werkgroep dr. AJ van L en dr. MG van P. NVOG richtlijn - Stuitligging- versie 2.0 (Guideline on breech presentation - version 2.0 of the Dutch Society of Obstet- rics and Gynecology) [Internet]. 2008. Available from: file:///C:/Users/PS13C/Down- loads/Stuitligging (1).pdf

7 Verburgt, T; Offerhaus P. Uitwendige versie (External Cephalic Version). NTOG. 2007; 120: 4–7.

8 Feitsma, HA; Middeldorp JM, Oepkes D. De uitwendige versie bij de a terme stuit; een inventariserend onderzoek (ECV for breech presentation at term; an inventory study). NTOG. 2007; 120: 4–6.

9 Rijnders M, Offerhaus P, Dommelen P Van, Wiegers T, Buitendijk S. Prevalence , Out- come , and Women ’ s Experiences of External Cephalic Version in a Low-Risk Popu- lation. 2010; 124–133.

140 Chapter 7 Facilitators and barriers to external cephalic version

10 Grol, R; Wensing, M; Eccles M. Improving Patient Care: The Implementation of Change in Clinical Practice [Internet]. Elsevier; 2005 [cited 2014 May 19]. p. 304. Available from: http://www.amazon.com/Improving-Patient-Care-Implementation- Clinical/dp/075068819X

11 Guldbrandsson K. From news to everyday use The difficult art of implementation. Stockholm: Strömberg;

12 Fleuren M, Wiefferink K, Paulussen T. Determinants of innovation within health care organizations: literature review and Delphi study. Int J Qual Health Care [In- ternet]. 2004 Apr; 16: 107–123. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/15051705

13 Greenhalgh, T; Robert, G; Macfarlane, F; Bate, P; Kyriakidou O. Diffusion of innovations in service organizations: systematic review and recommendations. Milbank Q [Internet]. 2004 [cited 2014 May 19]; 82: 581–629. Available from: https://c38xrlad- 25d4epd6rje0lpk.sec.amc.nl/pubmed/15595944

14 Rosman a N, Vlemmix F, Fleuren M a H, Rijnders ME, Beuckens a, Opmeer BC, et al. Patients’ and professionals’ barriers and facilitators to external cephalic version for breech presentation at term, a qualitative analysis in the Netherlands. Midwifery [Internet]. Elsevier; 2014 Mar [cited 2014 May 6]; 30: 324–330. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23664312

15 Fleuren M, Paulussen T, Dommelen P V., Buuren S V. Measurement Instrument for Determinants of Innovations (MIDI). BMJ Qual Saf [Internet]. 2013 Aug 15 [cited 2014 May 30]; 22: A42–A42. Available from: http://qualitysafety.bmj.com/lookup/ doi/10.1136/bmjqs-2013-002293.124

16 Nierkens V, Stronks K, de Vries H. Attitudes, social influences and self-efficacy expectations across different motivational stages among immigrant smokers: replication of the Ø pattern. Prev Med (Baltim) [Internet]. 2006 Oct [cited 2014 May 30]; 43: 306–311. Available from: http://www.ncbi.nlm.nih.gov/pubmed/16919321

17 Bartholomew, LK; Parcel, GS; Kok, G; Gottlieb N. Planning health promotion programs: an intervention mapping approach. [Internet]. San Francisco: Jossey-Bass; 2011 [cited 2014 May 19]. Available from: http://www.amazon.com/s/ref=nb_sb_ noss?url=search-alias=aps&field-keywords=planning+health+care+promotions+bart holomew&rh=i:aps,k:planning+health+care+promotions+bartholomew

141 Chapter 7 Facilitators and barriers to external cephalic version

18 Nunnally, JC; Bernstein I. Psychometric Theory [Internet]. McGraw-Hill; 3rd edition; 1994 [cited 2014 May 30]. p. 736. Available from: http://www.amazon.com/Psycho- metric-Theory-Jum-C-Nunnally/dp/007047849X

19 Kok M, van der Steeg JW, van der Post J a M, Mol BWJ. Prediction of success of external cephalic version after 36 weeks. Am J Perinatol [Internet]. 2011 Mar;28 : 103–110. Available from: http://www.ncbi.nlm.nih.gov/pubmed/20661845

142

8 Decision aids to improve informed decision-making in pregnancy care: a systematic review

Floortje Vlemmix

Janna K. Warendorf

Ageeth N. Rosman

Marjolein Kok

Ben Willem J. Mol

Jonathan M. Morris

Natasha Nassar

BJOG. 2013 FEB;120(3):257-66

145 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

ABSTRACT

Background Rapid development in health care has resulted in an increasing number of screening and treatment options. Consequently, there is an urgency to provide people with relevant information about benefits and risks of healthcare options in an unbiased way. Decision aids help people to make decisions by providing unbiased non-directive research evidence about all treatment options.

Objective To determine the effectiveness of decision aids to improve informed decision making in pregnancy care.

Search strategy We searched MEDLINE (1953–2011), EMBASE (1980–2011), CENTRAL (CENTRAL, the Cochrane Library; 2011, Issue 4), Psycinfo (1806– 2011) and Research Registers of ongoing trials (www.clinicaltrials.gov, www. controlled-trials.com).

Selection criteria We included randomised controlled trials comparing decision aids in addition to standard care. The study population needed to be pregnant women making actual decisions concerning their pregnancy.

Data collection and analysisTwo independent researchers extracted data on quality of the randomised controlled trial (GRADE criteria), quality of the decision aid (IPDAS criteria), and outcome measures. Data analysis was undertaken by assessing group differences at first follow up after the interventions.

Main results Ten randomised controlled trials could be included. Pooled analyses showed that decision aids significantly increased knowledge, (weighted mean difference 11.06, 95% confidence interval 4.85–17.27), decreased decisional conflict scores (weighted mean difference −3.66, 95% confidence interval −6.65 to −0.68) and decreased anxiety (weighted mean difference −1.56, 95% confidence interval −2.75 to −0.43).

Conclusions Our systematic review showed the positive effect of decision aids on informed decision making in pregnancy care. Future studies should focus on increasing the uptake of decision aids in clinical practice by identifying barriers and facilitators to implementation.

146 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

INTRODUCTION

Rapid development in healthcare technology and interventions has resulted in an increasing number of screening and treatment options. Many decisions in health care do not have a single best option but rather a number of ‘close call’ decisions that are ultimately influenced by patient preference. In pregnancy, for example, patient preferences are instrumental in decisions on first-trimester screening or analgesia during labour. Consequently, there is an urgency to inform people and provide relevant information about both the benefits and risks of healthcare options in an unbiased way.

Counselling pregnant women is challenging as these women not only need to consider their own health, but also the health of the fetus and the consequences for any subsequent pregnancies. Previous studies have highlighted that pregnant women want to be involved in decision making. 1,2

Decision aids (DA), or decision-support techniques, aim to help people to make these close-call decisions by providing unbiased nondirective research evidence about all treatment options, including the risks and benefits, and assisting people in clarifying their personal values related to corresponding outcomes and adverse effects.3,4 Standard education materials, such as leaflets, help people to understand their diagnosis, treatment and management, but differ from DAs because they do not necessarily facilitate informed decision making by exploring personal values and preferences.5 DAs are intended to be an adjunct to usual care and should not influence intervention uptake.3 A large systematic review of over 500 patient DAs by O’Connor and colleagues demonstrated the overall effectiveness and additional value of DAs for people facing health treatment or screening decisions.4 Although a number of DAs related to aspects of pregnancy care have been developed and evaluated, there has been no review of the overall efficacy of these aids in pregnancy care.6-8 We are specifically interested in the additional value of DAs for pregnant women because of the multiple consequences of their choices, and additional impact on their baby and family, as well as their own health. The aim of this study was to conduct a systematic review of randomised trials to summarise the available decision-support techniques, and to assess their quality and their effectiveness for pregnancy care.

147 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

METHODS

Sources

We searched MEDLINE (1953–2011), EMBASE (1980–2011), CENTRAL (CEN- TRAL, the Cochrane Library; 2011, Issue 4) and Psycinfo (1806–2011) up to March 2011 using keywords: choice behaviour, decision making, decision- support techniques, decisions, (choices or preferences), informed consent, pregnancy, labour, birth. No time or language restrictions were applied. Ref- erence lists of eligible studies, previously published systematic reviews and review articles, as well as the website of the Patient Decision Aids research group (www.ohri.ca/decisionaid/) were checked to identify cited trials not captured by electronic searches. We searched research registers for ongoing trials (www.clinicaltrials.gov, www.controlled-trials.com).

Study selection

We included studies using a randomised controlled trial or cluster randomised design evaluating DAs. Study populations needed to include pregnant women, who were facing the relevant pregnancy care decision in their current pregnancy. DAs were defined as interventions that provide unbiased and nondirective information to help pregnant women make choices based on personal values. They should contain information on all treatment options (including expectant management), and outcomes relevant to a person’s health status. Furthermore, implicit methods to clarify values should also be presented. We included randomised controlled trials that compared pregnancy care DAs with no intervention, usual care, alternative interventions, or a combination of these. We excluded studies where the interventions did not meet the criteria of a DA or where the DA was not available and the article did not provide enough information to determine whether the intervention met the minimum criteria to qualify as a patient DA, according to the International Patient Decision Aid Standards (IPDAS) criteria. These criteria assess the quality of DAs and were developed by over 100 experts from various decision-making fields and representing 14 countries, using the Delphi consensus process.3

148 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

Outcome measures

We studied decisional conflict score (DCS), knowledge and anxiety. Decision- al conflict refers to uncertainty in chosen option and the score ascertains whether individuals had clarity of values and felt informed and supported in their decision making. Knowledge is assessed using specific questions concerning the topic of the DA and usually consists of several true/false or multiple choice questions. Anxiety is often examined in these studies to confirm that DAs do not increase anxiety by providing too much detailed information. Anxiety was usually measured using the State-Trait Anxiety Scale. Secondary outcomes assessed were effectiveness of DA (proportion of individuals undecided, accuracy of risk perception of treatment options, enough information to make decision, involvement in decision making, regret of choice and satisfaction with choice); acceptability of DA (readability of DA and usefulness of information to make choice); decision behaviour outcomes (outcome of decision, uptake of intervention, and adherence to chosen option); health outcomes (neonatal and maternal morbidity and mortality, Apgar score, gestational age at delivery, and depression and self-esteem); and healthcare- system outcomes (cost-effectiveness of the DA, length of stay in hospital and length of consultation).

Quality assessment

Quality of included studies was assessed by examining risk of bias, using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Inter- ventions.9 Quality of evidence was assessed using the GRADE scale.10 The IPDAS Collaboration quality criteria framework was used to assess the quality of the DA.11 Final DAs were retrieved from a number of sources including Internet articles, theses, or by contacting authors. If unavailable, the quality of the DA was assessed using information provided in the articles of the randomised controlled trial, pilot study or protocol. Studies were excluded if the intervention did not qualify as a DA according to the IPDAS criteria used for scoring and where adequate information on the contents could not be obtained. The maximum IPDAS score ranged from 50 to 64 points, depending on the extensiveness of the DA (additional scoring items for patient sto- ries, internet-based DAs and DAs on screening tests). We converted scores to percentages of total scores. At the time of assessment there was no defined score/category for good-quality versus poor-quality DAs, or guidelines on interpretation of the scores. Therefore, we categorised into two categories (high-quality and low-quality aids) based on an arbitrary cut-off value of 60% of total IPDAS score. 149 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

Data extraction and statistical analysis

Two reviewers (FV and JKW) independently assessed all potential studies for inclusion into the review and extracted the data. Inconsistencies were resolved by discussion and consensus and by consulting a third reviewer (NN). Data extracted included information relevant to study characteristics, methodological quality (method of randomisation, allocation concealment and degree of blinding); study population and inclusion/exclusion criteria; topic of decision, type of intervention and comparator and respective formats; number of women in each study group; and quantitative and qualitative data relating to the selected primary and secondary outcomes, where available. Where relevant data were not reported, we contacted corresponding authors for additional information. Studies were categorised according to topic, quality of the DA, format of the DA and type of intervention in the control group.

Data analysis for each primary and secondary outcome was undertaken by assessing group differences at first follow up after the administration of the interventions. We were unable to evaluate differences over time (baseline measurements before intervention and at first follow up), because most studies did not report standard deviations over time. If different scales to measure the same outcome were used among studies, the scales were converted to the scale with the largest range. Continuous measures were assessed by comparing the means and standard deviations between the two treatment groups and calculating pooled weighted mean differences. For dichotomous data we used relative risk to calculate pooled relative risk. We calculated missing values such as standard deviations, mean differences, relative risks and 95% confidence intervals (95% CI) where possible or wrote to authors for additional information on the outcomes. The I2 test was used to assess variability between the studies and where we found strong evidence of heterogeneity (I2 > 50%), we analysed data with a random effects model. Subgroup analyses were conducted to take into account potential differences in the type of intervention applied in the control group (usual care or information program), format of DAs (booklet, counselling and computer program) and quality of DAs (<60% and >60% of IPDAS total scores). For analysing decision behaviour and health outcomes we pooled and assessed outcomes per topic of the DA. We also examined evidence of publication bias using funnel plots and plotting the standard error of the risk difference (or

150 Chapter 8 Decision aids to improve informed decision-making in pregnancy care odds ratio) by the risk difference (or odds ratio) and of these for skewness.9 All analyses were conducted using STATA version 9.2, 2007.

RESULTS

Search results

We critically appraised 6,064 unique citations from the databases, of which we reviewed 92 abstracts and selected 26 articles for further reading and assessment. Of these, ten articles were identified and included in the systematic review. Reasons for exclusion are reported in Figure 1.

Figure 1. F low chart study selection process.

151 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

We identified four possible ongoing trials and approached authors for information on the status of the trials; one trial evaluated a healthcare provider DA, one study was still recruiting (April 2011) and no response was received from the other two research groups, so these were not included.12-15

Characteristics of the ten studies included in this review can be found in Table 1. Scores of methodological quality assessment are presented in Table 2; blinding for this type of intervention is difficult, but most studies did try to limit contamination and blinded care providers by employing a research nurse or research assistant to administer the DA.

Table 2. Methodological quality summary: risk of biases for each included study Randomisation Allocation Patient Assessor Incomplete Intenti concealment blinding blinding data on to treat analysis Raynes-‐Greenow 2010 ⊕ ⊕ ⊕ ∅ ∅ ⊕ Kupperman 2009 ⊕ ⊕ ⊕ ∅ ⊕ ⊗ Nagle, 2007 ⊕ ⊕ ⊗ ⊗ ∅ ⊕ Nassar 2007 ⊕ ⊕ ∅ ∅ ⊕ ⊕ Montgomery, 2007 ⊕ ⊕ ⊗ ⊕ ⊕ ⊕ Wong, 2006 ⊕ ⊕ ⊕ ⊗ ⊕ ⊕ Arimori, 2006 ⊕ ⊕ ⊗ ⊗ ∅ ⊗ Shorten, 2005 ⊕ ⊕ ∅ ⊗ ∅ ⊗ Hunter, 2005 ⊕ ⊕ ⊗ ∅ ⊗ ⊕ Bekker 2004 ⊕ ⊕ ⊗ ∅ ⊕ ⊗ ⊕ Good, ∅ Moderate, ⊗ Missing information and/or data

Most studies did not report if there was blind outcome assessment. Loss to follow-up varied greatly; Raynes–Greenow, Kupperman, Arimori and Short- en and their colleagues had more than 10% loss of respondents at the time of first follow up (comprising an assessment after the intervention).16-19

152 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

-‐ /50)

IPDAS score Total 38.0% (17 45.5% (25/55)* 63.6% (35/55)* 7.3% (15/55)*

-‐

Timing of assessment first follow After intervention After intervention After intervention After intervention

15 weeks

Study population Study gestation <9 weeks gestation 15 weeks <18 weeks gestation 11 -‐ gestation

Information leaflet on contraceptio n control Normal counselling Group counselling care) (usual Usual counselling

guide,

-‐

intervention Information leaflet Structured with counselling decision analysis worksheet Booklet, and audio discuss to option with a genetic counsellor by Counselling on the nurse based Ottawa decision support

Decision Pregnancy termination method diagnosis Prenatal syndrome Down for women with a screening positive test. diagnosis Prenatal for mothers of advanced maternal age testing Prenatal

Desig n RCT RCT RCT RCT

N Control 159 56 110 43

154 50 116 38 N DA

Country UK UK Canada Japan

20 Study characteristics

23 22 19 Table 1. Table Citation Pregnancy termination termination Pregnancy method 2006 Wong, testing Prenatal Bekker, 2004 Hunter, 2005 Arimori, 2006

153 Citation Chapter 8 Decision aids to improve informed decision-making in pregnancy care

74.1% (43/58) 58.2% (32/55)* 75.5% (40/53) 68.0% (34/50)* 58 % (29/50) 83.0% (44/53)

2 weeks after 2 weeks -‐

1 After making at 14 or decision weeks gestation intervention^ after 1 week intervention after 8 weeks intervention 37 weeks gestation after 1 week intervention 18 -‐

35 weeks

<12 weeks gestation, by consultation GP’s <20 weeks gestation >34 weeks gestation 12 Inclusion weeks gestation, 28 at information weeks gestation 19 -‐ gestation >36 weeks gestation,

Standard Standard information pamphlet Computeris ed educational booklet car Usual care Usual care Usual Information leaflet

-‐

framework with Booklet worksheet Interactive program computer worksheet Booklet, and audio value with Booklet clarification exercise Computerized and information decision analysis program and Booklet a. worksheet

ivery ivery

an section

wn after previous after Prenatal testing in testing Prenatal primary care testing Prenatal for Do syndrome Singleton Breech presentation del of Mode section caesarean delivery Mode of previous after caesare Pain relief in labour

Cluste r RCT RCT RCT RCT RCT RCT 172 252 98 92 202 201 167 244 102 99 196 395

Australi a USA Australi a Australi a UK Australi a

7 -‐

18 16 6 Nagle, 2008 Kupperman, 2009 Breech presentation Vaginal birth after CS Shorten, 2005 Montgomery, 2007 in labour relief Pain Raynes Greenow, Nassar, 2007 154 Chapter 8 Decision aids to improve informed decision-making in pregnancy care primiparous primiparous vaginal on planning delivery

guide -‐ b. Booklet, Booklet, b. and worksheet audio

30 weeks of gestation. gestation. of 30 weeks -‐ score based on information in article, full decision aid not available for scoring -‐ *IPDAS at 26 assessed ^regret s atisfaction and

17 2010

155 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

Type of DA and quality assessment

The ten decisions investigated in the included studies were: first-trimester pregnancy termination method,20 prenatal testing, 18,19,21-23 management of breech presentation,8 mode of delivery after previous caesarean section 6,16,24 and labour analgesia.17,24,25 We identified three formats of DAs: six studies used a booklet as intervention7,8,17,20,22,26 (three included an audio-guide), two used an interactive computer program 6,18 and two studies used structured counselling.19,23

Five out of ten studies compared a DA to usual (verbal) counselling. Other studies used group counselling, a placebo intervention (information leaflet on relevant topic), or an information programme as control group.

We obtained access to five out of ten DAs.6,8,19-21,26 The other DAs were scored based on the information provided in the articles. IPDAS scores ranged from 27.3% (15 out of 55 points) to 83.0% (44 out of 53 points) (Table 1).

The lower scores tended to be those where we did not have access to the DA and where there was a lack of information regarding the developmental process.19,20

Decisional conflict score

The overall DCS was the most commonly presented outcome (six studies). 19-21,26 Articles used different DCS scales, so we converted them to a 0–100 scale for pooled data analysis. A lower score is associated with less decisional conflict and a score below 25 is generally considered as good decision making.18There was a significant difference in DCS at first follow up with a weighted mean difference (WMD) of −3.66 (95% CI −6.65 to −0.68) in favour of DAs (Figure 2).

Subgroup analysis showed similar results regardless of the type (booklet, computer program or counselling) or quality (low [<60] or high [>60] IPDAS scores) of the decision-support technique (Table 3). However, the type of control intervention did make a difference to decisional conflict with a stronger effect evident when comparison group was usual care (WMD −5.75, 95% CI 7.75 to −3.75), but no difference was observed when comparing the DA with other information leaflets.17,21

156 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

Figure 2. Forest plot decisional conflict

Knowledge

Five studies assessed knowledge at both baseline and first follow up.6,8,17,20,23 We carried out a pooled analysis for knowledge scores at first follow up, calculated as the percentage of correctly answered questions. The WMD was 11.06% (95% CI 4.85–17.27) in favour of the DA group (Figure 3). There was a similar positive effect of the DA on knowledge regardless of most types of DA, control intervention, or IPDAS quality score (Table 3). However, there was little effect and no difference in knowledge scores when the DA was in the form of counselling (Table 3).

157 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

Anxiety

Anxiety was assessed by six studies at first follow up.6,8,17,20,21,23 Pooled analysis showed significantly lower anxiety in DA versus comparison groups (WMD −1.59, 95% CI −2.75 to −0.43) (Figure 4). Similar results were found in all of the subgroup analyses with the biggest differences reported when DAs were compared with usual care or when the intervention involved some form of counselling or interactive computer program (Table 3). Smaller benefits were observed when DAs were either in the form of a booklet or compared with other information leaflets (Table 3).

Table 3. Results of subgroup analysis Outcome Subgroup analysis Subgroup WMD (95% CI) I2 DCS Type of control Usual care6,8,19,22 -‐5.75 -‐ ( 7.75, -‐ 3.75) 4.3% intervention Information programme17,21 0.0 -‐ ( 2.5, 2.5) NA

Type of intervention Booklet8,17,21,22 -‐3.75 -‐ ( 7.25, -‐ 2.5) 81% Interactive computer programme6 -‐4.2, SE: 1.5 -‐

Counselling19 -‐5.75 SE: 2.75 -‐ IPDAS score >608,17,19,21,22 -‐3.75 -‐ ( 7.25, -‐ 0.25) 81% <606 -‐4.5 -‐ ( 7.25, -‐ 2) NA

Knowledge Type of control Usual care6,8,20,23 11.7 (3.74, 19.65) 90% intervention Information programme17 8.6 SE:2.59 -‐

Type of intervention Booklet8,17,20 9.6 -‐ ( 1.42, 20.67) 94% Interactive computer programme6 10.5 SE: 1.85 -‐

Counselling23 2.5 SE:3.08 -‐ IPDAS score >6014,20 4.59 -‐ ( 3.67, 12.85) 87 <606,20,23 12.62 (1.45, 23.8) 93

Anxiety Type of control Usual care6,8,20,23 -‐2.47 -‐ ( 4.17, -‐ 0.77) 0% intervention Information programme17,21 -‐0.83 -‐ ( 2.42, 0.76) NA Type of intervention Booklet8,17,20,21 -‐1.21 -‐ ( 2.42, -‐ 0.01) 0% Interactive computer programme6 -‐3.4 SE: 1.24 -‐ Counselling23 -‐5.75 SE: 2.75 -‐

IPDAS score >608,16,17 -‐1.37 -‐ ( 2.65, -‐ 0.09) 0% <606,20,23 -‐2.24 -‐ ( 4.39, -‐ 0.05) 15%

158 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

Other secondary outcomes

Various measures were applied to assess the impact of DAs on decision-making. This included assessment of decisional regret, enough information to make a decision, and increase in understanding of treatment options. The studies assessing these outcomes showed significant effects in favour of the DAs. Fewer women were undecided at first follow up (relative risk [RR] 0.42, 95% CI 0.24–0.74) 8,27,28 or regretted their decision (RR 0.58, 95% CI 0.35– 0.97)8,18 and a greater proportion of women felt that they had enough information to make their decision when they were informed with a DA (RR 2.88, 95% CI 2.02–4.10). 58,17,27 Most studies looking at satisfaction found a non- significant positive effect of DAs on satisfaction with decision,6,17,22, decision- making process,17 and experience with birth and pregnancy.16 Kupperman et al.18 found a significant difference in satisfaction with decision-making process at first follow up (scale 0–10, DA 8.1 versus control 7.5; P < 0.001).

Only one study assessed the accuracy of risk perception, by measuring the percentage of women with correct risk perception for intervention related to miscarriage and the risk of having a baby with Down syndrome. In this study, authors found a significantly larger percentage of women in the DA group had correct risk perception (P < 0.001).18

Choice behaviour and adherence to chosen option

In the study by Kupperman et al. (prenatal testing),18 48% of women in the DA group said that the intervention had affected their decision-making pro-

159 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

cess, compared with 28% in the control group (OR 2.42, P < 0.001). Raynes- Greenow et al.17 (pain relief during labour) reported that women with the DA tended to consider their caregiver’s opinion more (37.8% versus 30.7%, P = 0.09) and were more likely to make a shared decision with their care provider (19.3% and 13.8%, P < 0.05).

Seven studies looked at patient preferences for the different treatment options at the time of first follow up after the intervention. We pooled- out comes of the five studies that looked at prenatal testing decisions and found no significant difference between DA and control groups preference for testing (RR 1.04, 95% CI, 0.95–1.14). However, presenting the information in a DA decreased the rate of women who were still undecided after receiving an intervention (RR 0.44, 95% CI 0.26–0.73). 6,8,22,26 The pooled analyses of the studies on prenatal testing preferences did not show a difference in chosen option, but the actual uptake (actual number of women who underwent prenatal screening) was slightly higher among women informed with a DA (RR 1.15, 95% CI 1.04–1.24). In contrast, Nassar et al. reported an increase of women counselled with the DA who intended to undergo external cephalic version (77.1 versus 55.5%, RR 1.38, 95% CI 1.12–1.70); however there was no significant difference in number of women who actually underwent the procedure.

Health and healthcare system outcomes

Two out of nine studies reported on neonatal outcomes8,17 (Apgar score at 1 and 5 minutes postpartum, birth-weight, preterm delivery and cephalic presentation) and found no significant difference between DA and control groups. There was also no difference in maternal mortality and morbidity, maternal length of stay in hospital, self-esteem or depression scores between groups. Assessment of the impact of DAs on healthcare costs revealed no difference in resource-use by mothers and babies in the intervention versus the control group.6,22 Although, Bekker et al.23 reported that the consultation length was slightly longer by 6 minutes in the DA group (MD 5.9, 95% CI 1.15–10.65).

160 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

DISCUSSION

This is the first systematic review of DAs in obstetrics using a critical appraisal of study and DA qualities. We found that DAs in pregnancy care significantly decrease decisional conflict, increase knowledge and decrease anxiety. Fur- thermore, DAs reduce decisional regret, reduce the proportion of women who are undecided and increase accuracy of risk perception. Subgroup analyses highlight that the type of decision-support technique has an impact on outcomes with counselling resulting in less decisional conflict and anxiety, but knowledge is not as enhanced. Results also reveal that the control group did make a difference to results with usual care compared with information leaflets, as the control group was more inferior, and led to the DA having a stronger effect on outcomes.

These findings suggest that DAs improve patient decision making compared with usual care. Furthermore, the greatest benefits were found when the decision-support technique was implemented in the form of counselling from a care provider; involving information, discussion of options and clarification of values, resulting in the greatest benefits to people in the form of less uncertainty and anxiety.23 Written information also resulted in greater recall of information. Hence, although the absolute differences between treatment groups were relatively small for some outcomes and some may question the clinical relevance, these findings suggest that there may be some people who may benefit from increased information and support in their decision making.

Findings were also consistent regardless of the quality of the DA, although there was no explicit ranking or guideline for the interpretation of IPDAS quality scores and our own arbitrary cut-off points of <60 and >60 may not have been sensitive enough. However, the IPDAS criteria for DAs form an evidence-based tool to develop and assess information leaflets and DAs.

In our opinion, it is important to separate these DAs from other topics because decisions in obstetrics do not only concern the mother but also the fetus, which may influence decisional conflict and anxiety. Furthermore, this is the first review that has assessed the quality of pregnancy-related DAs to provide an overview of their effectiveness and utility. One of the strengths of this review is the generalizability of the findings with the additional value of DAs proven in a wide range of pregnant populations and covering the prenatal period from first trimester until birth. 161 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

Of the included studies, only half assessed if people felt that they had ‘made the right choice for them’, a decision consistent with their values. Although, the studies assessing these outcomes showed significant effects in favour of the DAs, measures used varied widely. Three studies used the outcome ‘undecided’, two studies used ‘decision regret’ and two studies used ‘satisfaction with decision’. Generally, a reduction in validated outcome measures such as DCS, knowledge and anxiety make it plausible that people are helped by DAs in making the best choice according to their own values, but there is a need for an overall, uniform outcome measure on which to base this assumption.

Our study does have some limitations. One of the main issues with the review is the heterogeneity among the trials. This may be explained by the variety in topics of the DAs, the different control groups among the studies, the difference in outcome measurement scales, and cultural differences within the countries concerning health care and patient involvement in decision making. We tried to overcome these potential sources of heterogeneity by conducting random effect and subgroup analyses. These analyses revealed that, while some of the differences in outcomes could be explained by the type of control group, the study quality did not have any effect on results. Other weaknesses of the review were the observed asymmetry in the funnel plot for the DCS, indicating possible publication bias (the funnel plot is available online, see Supplementary material, Figure S1) and that evidence in the included studies was graded as being of moderate quality on the GRADE scale. A further limitation of the review is the percentage of loss to follow up, which was over 10% in some of the included studies. However, impact on meta-analyses results were minimised as the proportion of loss to follow up in each study was similar in the intervention and comparison groups.

Surprisingly, of all developed DAs, only three were available on the internet and a further two were obtained on request. This may be because of a lack of resources for implementation or support for the intervention. A number of trials included in the review were funded by grants and it may be that these had funding for development and evaluation, but no further resources to support implementation. This is confirmed by the fact that overall, we only found two studies assessing implementation of DAs in daily practice and both studies found that DAs were poorly used in daily practice and experienced many barriers especially from physicians, despite positive attitudes

162 Chapter 8 Decision aids to improve informed decision-making in pregnancy care to the use of DAs. 29,30 This could be a result of time pressure and ideas of non-productiveness of shared decision making. Other possible barriers may involve the relevance of information presented to the particular setting, lack of availability or relevance of options presented in the DA and whether information is up-to-date. Identification of the barriers and facilitators that may be associated with the implementation of decision-support techniques may increase their application and relevance.

CONCLUSION

Our systematic review highlights the positive effect of DAs on informed decision making in pregnancy care. Furthermore, the most positive benefits were when the decision-support technique was implemented by a care provider. Future studies should focus on increasing the uptake of DAs in clinical practice by both describing the elements of ‘effective DAs’ for their development, and provide advice on how pregnancy care providers should use and implement DAs in their clinical practice. Identifying the barriers and facilitators to implementation is also important in increasing their relevance and application.

163 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

REFERENCES

1 Waldenström U, Rudman A, Hildingsson I. Intrapartum and postpartum care in Swe- den: women’s opinions and risk factors for not being satisfied. Acta Obstet Gynecol Scand [Internet]. 2006 Jan [cited 2014 Jun 3]; 85: 551–560. Available from: http:// www.ncbi.nlm.nih.gov/pubmed/16752233

2 Brown, S; Lumley J. Satisfaction with care in labor and birth: a survey of 790 Aus- tralian women. Birth [Internet]. 1994 [cited 2014 Jun 3]; 21: 4–13. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/?term=Satisfaction+with+car e+in+labor+and+birth%3A+a+survey+of+790+Australian+women.+Birth

3 Elwyn G, O’Connor A, Stacey D, Volk R, Edwards A, Coulter A, et al. Developing a quality criteria framework for patient decision aids: online international Delphi consensus process. BMJ [Internet]. 2006 Aug 26 [cited 2014 May 24]; 333: 417. Avail- able from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1553508&to ol=pmcentrez&rendertype=abstract

4 O Connor A, Bennett C, Stacey D, Barry M, Col N, Eden K, et al. Decision aids for people facing health treatment or screening decisions ( Review ). Cochrane Libr. 2009; CD001431.

5 O Connor AM, Jacobsen MJ, Stacey D. An Evidence-Based Approach to Managing Women ’ s Decisional Conflict. J Obs Gynecol Neonatal Nurs. 2002; 31: 570–581.

6 Montgomery AA, Emmett CL. Two decision aids for mode of delivery among women with previous caesarean section : randomised controlled trial. BMJ. 2007; 334: 1305.

7 Nagle C, Lewis S, Meiser B, Metcalfe S, Carlin JB, Bell R, et al. Evaluation of a decision aid for prenatal testing of fetal abnormalities: a cluster randomised trial [IS- RCTN22532458]. BMC Public Health [Internet]. 2006 Jan [cited 2014 Jun 3]; 6: 96. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=147932 9&tool=pmcentrez&rendertype=abstract

8 Nassar N, Roberts CL, Raynes-Greenow CH, Barratt a, Peat B. Evaluation of a decision aid for women with breech presentation at term: a randomised controlled trial [ISRCTN14570598]. BJOG [Internet]. 2007 Mar [cited 2014 Mar 20]; 114: 325–333. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=240865 8&tool=pmcentrez&rendertype=abstract

164 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

9 Higgins, JPT; Green S. Cochrane Handbook for Systematic Reviews of Interventions. Version 5. Cochrane Collaboration; 2008.

10 Atkins D, Briss P a, Eccles M, Flottorp S, Guyatt GH, Harbour RT, et al. Systems for grading the quality of evidence and the strength of recommendations II: pilot study of a new system. BMC Health Serv Res [Internet]. 2005 Mar 23 [cited 2014 Jun 3]; 5: 25. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=108 4246&tool=pmcentrez&rendertype=abstract

11 Elwyn, G; OÇonnor, AM; Bennett, C; Newcombe, RG; Politi, M; Durand M et al. As- sessing the quality of decision support technologies using the international patient decision aid standards instrument (IPDASi). PLoS One [Internet]. 2009 [cited 2014 Jun 3]; 4: e4705. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubm ed/?term=Elwin+Assessing+the+quality+of+decision+support+technologies+using+t he+international+patient+decision+aid+standards+instrument+(IPDASi).

12 Covic T, Hallab L, Sharpe L, Manolios N, Aspinall D PJ. Motherhood choices: development of a decision aid for women with rheumatoid arthritis. Intern Med J. 2010;40 : 9–10.

13 Luitjes SH, Wouters MG, Franx A, Scheepers HC, Coupé VM, Wollersheim H, et al. Study protocol: Cost effectiveness of two strategies to implement the NVOG guidelines on hypertension in pregnancy: An innovative strategy including a computer- ised decision support system compared to a common strategy of professional audit and feedback, a. Implement Sci [Internet]. 2010 Jan; 5: 68. Available from: http:// www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2940931&tool=pmcentrez&re ndertype=abstract

14 Milne J, Gafni A, Lu D, Wood S, Sauve R, Ross S. Developing and pre-testing a decision board to facilitate informed choice about delivery approach in uncomplicated pregnancy. BMC Pregnancy Childbirth [Internet]. 2009 Jan [cited 2014 Jun 3]; 9: 50. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=277428 6&tool=pmcentrez&rendertype=abstract

15 Segal I, Shahar Y. A distributed system for support and explanation of shared decision-making in the prenatal testing domain. J Biomed Inform [Internet]. Elsevier Inc.; 2009 Apr [cited 2014 Jun 3]; 42: 272–286. Available from: http://www.ncbi. nlm.nih.gov/pubmed/18948227

165 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

16 Shorten A, Shorten B, Keogh J, West S, Morris J. Making choices for childbirth: a randomized controlled trial of a decision-aid for informed birth after cesarean. Birth [Internet]. 2005 Dec; 32: 252–261. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/16336366

17 Raynes-Greenow CH, Nassar N, Torvaldsen S, Trevena L, Roberts CL. Assisting informed decision making for labour analgesia: a randomised controlled trial of a decision aid for labour analgesia versus a pamphlet. BMC Pregnancy Childbirth [Internet]. 2010 Jan; 10: 15. Available from: http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=2868791&tool=pmcentrez&rendertype=abstract

18 Kuppermann M, Norton ME, Gates E, Gregorich SE L LA, Nakagawa S et al. Comput- erized prenatal genetic testing decisionassisting tool: a randomized controlled trial. Obs Gynecol. 2009; 113: 53–63.

19 Arimori N. Randomized controlled trial of decision aids for women considering prenatal testing: the effect of the Ottawa Personal Decision Guide on decisional conflict. Jpn J Nurs Sci. 2006;3 : 119–130.

20 Wong SSM, Thornton JG, Gbolade B, Bekker HL. A randomised controlled trial of a decision-aid leaflet to facilitate women’s choice between pregnancy termination methods. BJOG [Internet]. 2006 Jun [cited 2014 Jun 3]; 113: 688–694. Available from: http://www.ncbi.nlm.nih.gov/pubmed/16709212

21 Nagle C, Gunn J, Bell R, Lewis S, Meiser B, Metcalfe S, et al. Use of a decision aid for prenatal testing of fetal abnormalities to improve women’s informed decision making: a cluster randomised controlled trial [ISRCTN22532458]. BJOG [Internet]. 2008 Feb [cited 2014 Jun 3]; 115: 339–347. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/18190370

22 Hunter a GW, Cappelli M, Humphreys L, Allanson JE, Chiu TT, Peeters C, et al. A randomized trial comparing alternative approaches to prenatal diagnosis counseling in advanced maternal age patients. Clin Genet [Internet]. 2005 Apr [cited 2014 Jun 3]; 67: 303–313. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15733266

23 Bekker HL, Hewison J, Thornton JG. Applying decision analysis to facilitate informed decision making about prenatal diagnosis for Down syndrome: a randomised controlled trial. Prenat Diagn [Internet]. 2004 Apr [cited 2014 Jun 3]; 24: 265–275. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15065100

166 Chapter 8 Decision aids to improve informed decision-making in pregnancy care

24 Montgomery A a. The DiAMOND trial protocol: a randomised controlled trial of two decision aids for mode of delivery among women with a previous caesarean section [ISRCTN84367722]. BMC Pregnancy Childbirth [Internet]. 2004 Dec 10 [cited 2014 Jun 3]; 4: 25. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi? artid=539355&tool=pmcentrez&rendertype=abstract

25 Raynes-Greenow CH, Roberts CL, Barratt A, Brodrick B, Peat B. Pregnant women’s preferences and knowledge of term breech management, in an Australian setting. Midwifery [Internet]. 2004 Jun [cited 2014 May 19]; 20: 181–187. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15177862

26 Shorten A, Chamberlain M, Shorten B, Kariminia A. Making choices for childbirth: development and testing of a decision-aid for women who have experienced previous caesarean. Patient Educ Couns [Internet]. 2004 Mar [cited 2014 Jun 3];52 : 307–313. Available from: http://www.ncbi.nlm.nih.gov/pubmed/14998601

27 Leung KY, Lee CP, Chan HY, Tang MHY, Lam YH, Lee a. Randomised trial comparing an interactive multimedia decision aid with a leaflet and a video to give information about prenatal screening for Down syndrome. Prenat Diagn [Internet]. 2004 Aug [cited 2014 Jun 3]; 24: 613–618. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/15305348

28 Shorten A. Birth choices: what is best for you. Vaginal or caesarean birth? Wollon- gong: Wollongong University Pintery; 2006.

29 Holmes-Rovner M, Valade D, Orlowski C, Draus C, Nabozny-Valerio B, Keiser S. Im- plementing shared decision-making in routine practice: barriers and opportunities. Health Expect [Internet]. 2000 Sep; 3: 182–191. Available from: http://www.ncbi. nlm.nih.gov/pubmed/11281928

30 O’Cathain A, Thomas KJ. Evaluating decision aids--where next? Health Expect [Internet]. 2004 Jun; 7: 98–103. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/15117384

167

9 Implementation of external cephalic version in the Netherlands: a retrospective cohort study

Floortje Vlemmix

Ageeth N. Rosman

Susan te Hoven

Suzanne van de Berg

Margot A.H. Fleuren

Marlies E.B. Rijnders

Antje Beuckens

Brent C. Opmeer

Ben Willem J. Mol

Marjolein Kok

Submitted

169 Chapter 9 Implementation of external cephalic version in the Netherlands

ABSTRACT

Background External cephalic version reduces the rate of elective caesarean sections due to breech presentation at term. The aim of this study was to evaluate the implementation of external cephalic version in the Netherlands and to explain variation in implementation rates by hospital characteristics.

Methods We invited 40 hospitals to participate in this retrospective cohort study. We reviewed hospital charts of singleton breech deliveries from 36 weeks gestation and onwards. We documented whether an external cephalic version attempt was performed, reasons if not, mode of delivery and hospital characteristics.

Results We included 4,770 women from 36 hospitals. External cephalic version was performed in 2,443 women (62.2% of eligible women, range 8.2% to 83.6% in different hospitals). Implementation rates were higher in teaching hospitals, hospitals with special office hours, larger obstetric units and hospitals located in larger cities. Suboptimal implementation is mostly caused by the care provider not offering the treatment and secondly due to women not opting for the offered attempt.

Conclusion External cephalic version implementation rates vary widely among hospitals, and are suboptimal. A prerequisite for designing a proper implementation strategy is a detailed understanding of the reasons for not offering and not opting for external cephalic version.

170 Chapter 9 Implementation of external cephalic version in the Netherlands

INTRODUCTION Breech presentation occurs in 3 to 4% of all term pregnancies.1 Vaginal breech delivery is related to increased neonatal mortality and morbidity.2,3 As a consequence, planned caesarean section (CS) rates for breech delivery have risen up to 94%.4 The downside of this high caesarean rate are risks related to future pregnancies and deliveries with increased maternal and foetal morbidity and mortality.(5,6,7,8) External cephalic version (ECV) reduces the rate of breech presentation at term and leads to a significant reduction in caesarean delivery from almost 80% till just below 40% (RR 0.63 (95%CI 0.44-0.90), without an increased risk for the neonate.9,10 Therefore, ECV is considered as an important obstetrical intervention and is recommended as the first treatment of choice in case of term breech presentation in national and international guidelines.11–14 Despite the recommendations of national guidelines, not all women with a term breech presentation receive an ECV attempt. An inventory survey among all hospitals in the Netherlands in 2007 reported that 5% of the gynaecologic practices did not perform or referred for ECV at all, even though women with breech presentation were seen in these hospitals.15 A small prospective cohort study in the Netherlands reported that 26% of women with a fetus in breech presentation did not receive an ECV attempt; in 48% of these cases, the obstetrician decided not to perform ECV, 37% of women declined a version attempt, and 15% gave birth before the version was performed.16 International studies showed that the number of women eligible for ECV who were not offered an attempt range from 4% to 33% and the number of women declining ECV range from 20-30%.17–19 Insight in factors associated with poor implementation of ECV is needed to design effective strategies to improve implementation. The objective of this study was to evaluate the implementation rate of ECV in a large cohort in the Netherlands and to assess the reasons not to perform ECV.

MATERIAL AND METHODS

We invited 40 randomly selected hospitals in the year 2010, to participate in this retrospective cohort study. Hospitals were asked for permission to review all patient files of singleton breech deliveries from 36 weeks gestation and onwards (as national guidelines advise ECV from 36 weeks). We identified

171 Chapter 9 Implementation of external cephalic version in the Netherlands

all consecutive breech deliveries between the st1 of January 2008 and the 31st of December 2009 from hospital birth registers. We collected data on parity, ECV (attempt, outcome of attempt, gestational age at ECV, reasons if no ECV was performed, and mode of delivery. Unfortunately, we were not able to collect data on cephalic deliveries after a successful ECV attempt as the majority of hospitals do not systematically register ECV attempts. We collected baseline characteristics of participating hospitals: number of birth per year (categorized <1000, 1000-1999, ≥2000), type of hospital (teaching hospital or not), presence of ECV office hour within the region of the hospital (either led by midwives, gynecologists or a collaboration), size of the city (categorized in quartiles based on number of inhabitants in the city where the hospital is located), and localization of the hospital in the north, east, west or south of the Netherlands.20 Our primary outcome was the implementation rate of ECV, for this study defined as the number of ECV attempts performed in women who had a breech presentation at time of delivery and were eligible for ECV, Secondary outcomes were the number of women with a contraindication for ECV according to national guidelines, the number of women who were not offered ECV, and who declined ECV. Statistical analyses were carried out using SPSS 20.0. We used a logistic regression accounting for clustering of observations (generalized estimating equations, or GEE) to analyse the effect of the hospital characteristics on the implementation of ECV. Those characteristics significantly related to ECV in the univariate analyses (p<0.05), were analysed in a multivariate analyses.

RESULTS

Of the 40 invited hospitals, 36 gave approval for participation in this study. We identified 4,770 patients with a singleton fetus in breech presentation at birth. Baseline characteristics are summarized in Table 1. Breech presentation was not diagnosed until birth in 357 women (7.5% ( range 0% to 20%)) (Table 2). A contraindication for ECV according to Dutch guidelines was present in 367 women (7.5%). Most common contraindications were structural uterine anomaly (2.5%, ), and fetal growth restriction (1.6%). Hundred twenty women (2.8%) had an indication for a planned CS other than breech presentation. Thus, 3,926 (82.3%) of women were eligible for ECV.

172 Chapter 9 Implementation of external cephalic version in the Netherlands

Table 1. Baseline characteristics of study population.

Overall Missing values (n) N= 4,770

Parity (%) 8 Nulliparous 64.0 Multiparous 36.0 Gestational age at delivery (mean, range) 38 6/7 (36 – 0/7 43 3/7) 144 Mode of delivery (%) 21 Vaginal delivery 20.0 Planned CS 69.0 Emergency CS 11.0

Missing values: parity 0.2%, gestational age at delivery 3.0%, mode of delivery 3.2%

Counselling for ECV

Of 3,926 women eligible for ECV, 1,483 (37.8%) did not receive an ECV attempt. Reasons for not receiving an ECV attempt are given in Table3. [table 3 invoegen] Of these 1,483 women, 39 (2.6%) cases the women delivered before the scheduled ECV attempt. In 671 (45.2%) cases the women were counselled but they declined ECV while 773 (52.1%) women were not counselled by the health care provider and thus not offered an ECV. Reasons not to offer ECV were reported in only 275 of the 773 cases (35.6%). The most important reasons not to offer ECV were often relative contra indications not mentioned in the national guideline, such as assumed unfavourable factors for a successful ECV (high BMI, macrosomia, anterior placenta, descended breech or advanced gestational age; n= 109 (14.1%)). Other reasons mentioned were possible increased maternal or fetal risk (anticoagulant use, Hepatitis B, HIV, one previous low transverse caesarean, or vaginal blood loss (unknown cause); n = 166 (21.5%)).

Hospital characteristics

Hospitals were located in villages and cities of various sizes; the smallest had 5,000 and the largest 800,000 inhabitants. The categorisation of the hospitals according to the size of the city is shown in Table 4. Twenty six (72%) were teaching hospitals. In 44% of the regions affiliated to the hospitals there was a special ECV office hour, either led by midwives, gynaecologists or a collaboration between these two. Thirteen hospitals had less than 1,000

173 Chapter 9 Implementation of external cephalic version in the Netherlands

births per year, 19 with 1,000-2,000 births per year and four with over 2,000 births per year, regional home deliveries not included.

Table 2. Contraindications for ECV according to national guidelines Overall N= 4,770 (%) Eligible 3,926 (82.3) ECV attempt 2,443 (51.2) No ECV attempt 1,483 (31.1) Not Eligible 844 (17.7) Undiagnosed breech presentation 357 (7.4) Absolute contraindication ECV 367 (7.5) Uterine anomaly 119 (2.5) Fetal growth restriction 75 (1.6) Hypertension /PE / HELLP 46 (1.0) Oligohydramnion 52 (1.1) Non reassuring CTG before ECV 10 (0.2) Placental abruption in previous pregnancy 5 (0.1) (P)PROM 50 (1.0) Elective CS indication 120 (2.8) > 1 previous SC 48 (1.0) Contra indication vaginal delivery 46 (1.0) Placenta praevia 36 (0.8)

Table 3. Eligible patients not having an ECV attempt.

Overall Range among hospitals N=1,483 (%) (%) Not opting 671 (45.2) 0 -‐ 88.9 Birth before planned ECV 39 (2.6) 0 -‐ 9.1 Not offered 773 (52.1) 0 -‐ 96.8 Reasons to not offer ECV other than contraindication in guidelines 275 (18.5) 0 -‐ 62.9 HIV, Hep B, anticoagulants 32 (2.2) Multiple factors decreasing success rate 109 (7.3) One previous CS 52 (3.5) Vaginal blood loss 14 (0.9) Other 68 (4.6) Not offered for unknown reason 498 (33.6) 0 -‐ 96.8

174 Chapter 9 Implementation of external cephalic version in the Netherlands

ECV implementation rates

The ECV implementation rate at the hospital level is visualized in Figure 1. [figure 1 invoegen] The mean implementation rate of ECV among eligible women was 62.2% (range 8.2 to 83.6%). Table 4 shows the results of the univariate and multivariate analysis of hospital characteristics in relation to ECV attempts. [table 4 invoegen] Univariate analysis shows a higher ECV implementation rate in teaching hospitals (66.3 vs. 46.8%, OR 2.35 (95%CI 1.4-4.0), and in regions with an ECV office hour (68.1 vs. 56.1%, OR 1.9 (95% CI 1.2-3.0)). There was a lower implementation rate of ECV in hospitals with less than 1000 annual deliveries compared to those with more than 2000 (44.7 vs. 66.1 vs. 56.1%, OR 0.36 (95% CI 0.22-0.59)), and in hospitals located in smaller villages compared to the largest cities (47.3 vs. 71.0%, OR 0.17- 0.69)). Most of the hospitals were located in the west of the Netherlands and had a higher ECV implementation rate compared to hospitals located in the North of the Netherlands (68.8 vs. 46.4, OR 3.0 (95% CI 1.1-8.3)). Only size of the city (smaller than 49,000 inhabitants) remained significantly related to ECV implementation rate in the multivariate analyse. The other factors (teaching hospital, ECV office hour, birth per year and region) are closely related to size of the city.

Figure 1. Implementation rate of ECV per hospital. Horizontal line represents the mean implementation rate (62.2%). Hospitals are presented from poorest implementation (below 10%) to good implementation (over 80%), with the majority between 50% and 70%.

175

Chapter 9 Implementation of external cephalic version in the Netherlands

DISCUSSION

This study shows that the ECV implementation rate in women with a breech presentation at birth is generally low, and varies widely across hospitals. Poor implementation is mainly caused by health care providers who do not offer ECV to eligible women, followed by women declining an offered ECV. Furthermore, ECV implementation and success rates are poorly registered by hospitals. Considering the high CS rates for breech presentation we can conclude from the above that there is considerable room to improve obstetric care in women with a term fetus in breech presentation. Ideally, improvement activities should focus on those hospitals with low implementation rates, but without a systematic registration of all external cephalic versions in the Netherlands, this is difficult to establish. Our multivariate analyses show that hospitals in small cities more often have poor implementation rates of ECV. This can be explained by the other regional characteristics: the small hospitals are less frequently teaching hospitals and less often have an ECV office hour. Thus either care in these small hospitals need to be improved or centralising care for breech presentation to larger hospitals nearby might be indicated. One reason that ECV is often not performed, is that care providers do not offer an ECV to their patients. In order to address this reason, it is important to explore the underlying rationale why care providers do not offer ECV. In our study, reasons for not offering an ECV were not registered in the majority of patients’ files . If reported, reasons often did not correspond to the contraindication for EVC as mentioned in the current national guidelines. The large amount and range in percentage of women not opting for ECV might be a reflection of the counselling skills of care providers, or explained by different socio-demographic backgrounds of women. Providing standardized information and offering clients the opportunity to discuss their fears or restraints might abate irrational ideas, and improve the number women opting for ECV. However, a study by Nassar et al. evaluating a decision aid for ECV did show a reduction in decisional conflict scores in women deciding on ECV, but did not alter the amount of women opting for ECV.21 The number of undiagnosed breech presentations is an important issue in order to improve care for women with breech presentation. The percentage of undiagnosed breech ranged from 0 to 20.0% among hospitals, partly reflecting poor diagnostic quality of care within that hospital and the referring midwifery practices associated with that hospital). This might be improved by the use

177 Chapter 9 Implementation of external cephalic version in the Netherlands

of ultrasound during the regular pregnancy check-ups. An Australian study showed a sensitivity of 70% (95%CI 68-72%) and specificity of 95% (96-97%) of clinical examination.22 They concluded that clinical examination alone as diagnostic test is not sensitive enough in the management of non-cephalic presentation.

Strengths and limitations

To our knowledge, this is the largest cohort on ECV implementation thus far and it reflects the implementation rate in a random selection of hospitals and their surrounding midwife practices with varying regional characteristics. Our findings are in line with earlier research. Several studies showed that the number of women eligible for ECV who were not offered an attempt range from 4% to 33%.(17,18,19) We reported that 52.1% of women who did not have an ECV attempt, were not offered one. This corresponds to 19.7% of all eligible women (773/3,926). The number of clients refusing ECV in other studies ranged between 20% to 30%. We reported that 45.2% of women not undergoing ECV declined the treatment, which corresponds to 17.1% of all eligible women (671/3,926). The detailed information on reasons why ECV was not provided, and which hospital characteristics are related to better ECV implementation rates provide valuable information to improve care for breech presentation. Although prevalence of breech presentation in a general population is low, only 3-4% of term pregnancies, they represent 33.2% of the planned CS in singleton pregnancies at term in the Netherlands in 2012 (3,258 of 9,811 CS).23 A limitation of this study is the retrospective data collection. Because there is no official registry of ECV attempts, we used data from women with a breech presentation at delivery. Therefore, data on women with a spontaneous version from breech to cephalic presentation and those who delivered in cephalic presentation after a successful ECV are lacking. The average success rate of ECV in the Netherlands is around 40%, thus the 2,443 women with an ECV attempt and breech presentation at birth included in this study probably represent 60% of all women with an ECV attempt.24 After adjusting for this selection bias, the actual ECV implementation rate is estimated to be around 73%.

178 Chapter 9 Implementation of external cephalic version in the Netherlands

CONCLUSIONS

ECV implementation rates vary widely among hospitals. Especially in hospitals with low implementation rates, improvement is urgently needed to reduce the number of planned CS for singleton (breech presenting) pregnancies at term. Suboptimal implementation is mostly caused by care providers who do not offer ECV to eligible women and secondly due to women not opting for the offered ECV attempt. Follow-up studies should focus on the exact reasons for not offering ECV. A detailed understanding of these reasons is a prerequisite for designing implementation strategies that can achieve real change.

179 Chapter 9 Implementation of external cephalic version in the Netherlands

REFERENCES

1. Hickok, DE; Gordon, DC; Milberg, JA; Williams, MA; Daling J. The frequency of breech presentation by gestational age at birth: a large population-based study. Am J Obs Gynecol [Internet]. 1992 [cited 2014 Mar 3];166(3):851–2. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/1550152

2. Hannah ME, Hannah WJ, Hewson S a, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet [Internet]. 2000 Oct;356(9239):1375–83. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0140673600028403

3. Rietberg C. Term breech presentation in The Netherlands from 1995 to 1999: mortality and morbidity in relation to the mode of delivery of 33,824 infants. BJOG An Int J Obstet Gynaecol [Internet]. 2003 Jun [cited 2014 Mar 3];110(6):604–9. Available from: http://doi.wiley.com/10.1016/S1470-0328(03)01907-4

4. Hartnack Tharin JE, Rasmussen S, Krebs L. Consequences of the Term Breech Trial in Denmark. Acta Obstet Gynecol Scand [Internet]. 2011 Jul [cited 2014 Feb 20];90(7):767–71. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/21476999

5. Van Ham M a, van Dongen PW, Mulder J. Maternal consequences of caesarean section. A retrospective study of intra-operative and postoperative maternal complications of caesarean section during a 10-year period. Eur J Obstet Gynecol Reprod Biol [Internet]. 1997 Jul;74(1):1–6. Available from: http:// www.ncbi.nlm.nih.gov/pubmed/9243191

6. Cunningham FG, Bangdiwala SI, Brown SS, Dean TM, Frederiksen M, Rowland Hogue CJ, et al. NIH consensus development conference draft statement on vaginal birth after cesarean: new insights. NIH Consens State Sci Statements [Internet]. 2010 Mar 10;27(3):1–42. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/20869552

7. Rasmussen, SA; Maltau J. Complications following cesarean section. Tidsskr Nor Laegeforen [Internet]. 1990 [cited 2014 Mar 3];110(3):351–3. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/2309178

180 Chapter 9 Implementation of external cephalic version in the Netherlands

8. Brouwers, HAA; Bruinse, HW; Dijs-elsinga, J; Van Huis, AM; De Miranda, E; Ravelli, ACJ; Tamminga, P; Schiere, AM; Giesbers H. Jaarboek zorg in Nederland 2009 (the Netherlands Perinatal Registry, yearly report 2009) [Internet]. 2013. Available from: http://www.perinatreg.nl/uploads/150/136/Jaarboek_Zorg_ in_Nederland_2009_Tabellen.pdf

9. Gj H, Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012;10(CD000083).

10. Grootscholten, K; Kok, M; Oei, SG; Mol, BW; van der Post J. External cephalic version-related risks: a meta-analysis. Obs Gynecol [Internet]. 2008 [cited 2014 Mar 20];112(5):1143–51. Available from: https://c38xrlad25d4epd6rje0lpk.sec. amc.nl/pubmed/?term=xternal+cephalic+version-related+risks:+a+meta-analysis.

11. The Royal Australian and New Zealand College of Obstetricians and Gynecologists. Management of Breech Presentation at Term. 2001. p. file:///C:/Users/PS13C/ Downloads/C–Obs%2011%20Mana.

12. Verburgt, T; Offerhaus P. Uitwendige versie (External Cephalic Version). NTOG. 2007;120:4–7.

13. Otterlo Werkgroep dr. AJ van L en dr. MG van P. NVOG richtlijn - Stuitligging- versie 2.0 (Guideline on breech presentation - version 2.0 of the Dutch Society of Obstetrics and Gynecology) [Internet]. 2008. Available from: file:///C:/Users/ PS13C/Downloads/Stuitligging (1).pdf

14. Royal College of Obstetricians and Gynecologists. External cephalic version and reducing the incidence of breech presentation. http://www.rcog.org.uk/files/rcog- corp/uploaded-files/GT20aExternalCephalicVersion.pdf. 2010. p. 1–8.

15. Feitsma, HA; Middeldorp JM, Oepkes D. De uitwendige versie bij de a terme stuit; een inventariserend onderzoek (ECV for breech presentation at term; an inventory study). NTOG. 2007;120:4–6.

16. Rijnders M, Offerhaus P, Dommelen P Van, Wiegers T, Buitendijk S. Prevalence , Outcome , and Women ’ s Experiences of External Cephalic Version in a Low-Risk Population. 2010;(June):124–33.

181 Chapter 9 Implementation of external cephalic version in the Netherlands

17. Caukwell S, Joels L a, Kyle PM, Mills MS. Women’s attitudes towards management of breech presentation at term. J Obstet Gynaecol [Internet]. 2002 Sep [cited 2014 Mar 20];22(5):486–8. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/12521413

18. Bewley, S; Robson, SC; Smith, M; Glover, A, Spencer J. The introduction of external cephalic version at term into routine clinical practice. Eur J Obs Gynecol Reprod Biol. 1993;52(2):89–93.

19. Leung, TY; Lau, TK; Lo, KW; Rogers M. A survey of pregnant women’s attitude towards breech delivery and external cephalic version. Aust N Z J Obs Gynaecol. 2000;40(3):253–9.

20. Nassar N, Roberts CL, Raynes-Greenow CH, Barratt a, Peat B. Evaluation of a decision aid for women with breech presentation at term: a randomised controlled trial [ISRCTN14570598]. BJOG [Internet]. 2007 Mar [cited 2014 Mar 20];114(3):325–33. Available from: http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=2408658&tool=pmcentrez&rendertype=abstract

21. Nassar, N; Roberts, CL; Cameron, CA; Olive E. Diagnostic accuracy of clinical examination for detection of non-cephalic presentation in late pregnancy: cross sectional analytic study. BMJ [Internet]. 2006 Sep 16 [cited 2014 Mar 20];333(7568):578–80. Available from: http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=1569956&tool=pmcentrez&rendertype=abstract

22. Brouwers, HAA; Bruinse, HW; Dijs-elsinga, J; Van Huis, AM; De Miranda, E; Ravelli, ACJ; Tamminga, P; de Vries C. Jaarboek zorg in Nederland 2012 (the Netherlands Perinatal Registry, yearly report 2012) [Internet]. 2013. Available from: http://www. perinatreg.nl/uploads/150/141/PRN_ jaarboek_2012_09012014_PDF.pdf

23. Kok, M; Bais, JM; van Lith, JM; Papatsonis, DM; Kleiverda, G; Hanny, D; Doornbos, JP; Mol, BW;van der Post J. Nifedipine as a uterine relaxant for external cephalic version: a randomized controlled trial. Obs Gynecol. 2008;112(2):271–6.

182

10 Implementation of client versus care provider strategies to improve external cephalic version rate: a cluster randomised controlled trial

Floortje Vlemmix*

Ageeth N. Rosman*

Margot A.H. Fleuren

Marlies E. Rijnders

Antje Beuckens

Brent C. Opmeer

Ben Willem J. Mol

Marjolein Kok

*( both authors are first listed)

Submitted

185 Chapter 10 Implementation of client versus care provider strategies

ABSTRACT

Objective To determine the effectiveness of a client or care provider strategy to improve the implementation of external cephalic version.

Design Cluster randomized controlled trial.

Setting 25 clusters; hospitals and its referring midwifery practices randomly selected in the Netherlands.

PopulationSingleton breech presentation from 32 weeks gestation onwards.

Methods We randomized clusters to a client strategy (written information leaflets and decision aid) , a care provider strategy (one day counselling course focused on knowledge and counselling skills), a combined client- and care provider strategy and care as usual strategy. We performed an intention to treat analysis.

Main outcome measures Rate of external cephalic version in various strategies. Secondary outcomes were the percentage of women counselled and opting for a version attempt. Results. The overall implementation rate of external cephalic version was 72% (1,169 of 1,613 eligible clients) with a range between clusters of 8-95%. Neither the client (OR 0.78 (95% CI 0.41-1.5)) nor the care provider strategy (OR 1.2 (95% CI 0.64-2.3)) showed significant improvements. Results were comparable when we limited the analysis to those women that actually were offered intervention (OR 0.64 (95% CI 0.30-1.4) and (OR 2.0 (95%CI 0.72-4.5)).

Conclusions Neither a client- nor a care provider strategy improved the external cephalic version implementation rate for breech presentation, either on the number of version attempts offered or on the number of women accepting the procedure.

186 Chapter 10 Implementation of client versus care provider strategies

INTRODUCTION

Breech presentation occurs in 3 to 4% of all term pregnancies.1 The majority of term breech presentation is managed by elective cesarean section, since vaginal breech delivery is associated with significant increased neonatal morbidity and mortality.2–4 Planned cesarean section rates for breech presentation vary from 60% in the Netherlands up to 90% in the UK and 96% in Australia.4–6 Increased maternal en neonatal morbidity and mortality due to these cesarean sections in subsequent pregnancies, is of concern.7–10 External cephalic version (ECV) is an effective and safe treatment to prevent breech presentation at birth, and is recommended in professional guidelines.11–15 One of the main challenges with the introduction of guidelines is that clinicians do not automatically adopt and apply guidelines as intended by the developers.16–19 This applies also to ECV; the number of breech presentations at birth in the Netherlands has remained stable around 4%.20 Furthermore, an inventory survey among all hospitals in the Netherlands in 2007 reported that 5% of the gynaecologic practices did not perform ECV at all.21 In the UK and Australia the number of eligible women who were not offered an ECV attempt ranged from 4% to 33%.22–24 An Australian preference study showed that 39% of women did not choose ECV and a similar study in Israel reported a decrease in women willing to undergo ECV from 54% to 24%.25,26 Based on the figures from above mentioned studies (provided rates of women counselled and opting for ECV), we estimated the implementation rate of ECV to be 50-60%,27 which we consider low and important to improve. Several models have been proposed that describe similar sequences for the systematic planning of guidelines in general terms.18,28–30 First, an analysis of determinants (impeding and enhancing factors) should be performed. A detailed understanding of these determinants is needed to guide the process of designing implementation strategies that will have the potential to produce real change. 28,30,31 We previously performed a determinant analysis showing that lack of knowledge to counsel clients on ECV, is an important barrier mentioned by care providers. Lack of (well-timed) information, fear of pain or harm to the fetus, the preference to have a planned cesarean section, and negative experiences from relatives are important barriers mentioned by clients.32,33 On the basis of these determinants we designed two implementation strategies: a client - and a care provider strategy. We evaluated the effects of these strategies on the actual number of women receiving an ECV attempt.

187 Chapter 10 Implementation of client versus care provider strategies

MATERIAL AND METHODS In the Netherlands prenatal care is supervised by midwives in primary care, and by obstetricians in secondary care. Midwives take care of low risk pregnancies. If specialized care is needed, midwives refer to an obstetrician in an affiliated hospital. We will refer to this stratified care model as ‘cluster’, meaning a hospital including all surrounding midwifery practices. We performed a prospective, open-label, cluster randomized clinical trial. Randomization was performed on clusters level in blocks of 12, in a two- by-two factorial design. Clusters were randomized to a client strategy, a care provider strategy, both implementation strategies, or care as usual. We stratified for clusters with or without ECV office hours as we hypothesized that clusters with an organized office hour might have a better baseline implementation of ECV. An ECV office hour was defined as special reserved time for counselling and ECV attempts during the week, with a select group of experienced care providers who perform ECV attempts either in the hospitals or midwifery practices. Clusters all over the Netherlands were randomly selected and invited to participate in this study, until at least 24 clusters had given approval. The allocation sequence was generated by the clinical research unit of the Academic Medical Center. At the moment a cluster agreed to participate, it was randomized and the implementation of the allocated strategy was planned. Researchers and care providers were not blinded for allocated strategies, but participating clients were kept unaware of the different implementation strategies. Consecutive women with a singleton breech presentation from 32 weeks onwards between January 2011 and August 2012 were eligible for the study. Exclusion criteria were contraindications for ECV according to the national guideline (uterus anomaly, fetal growth restriction, hypertensive disorder, oligohydramnios, (P)PROM), any contraindication for a vaginal delivery and a spontaneous version after 32 weeks.14,15 Calculation of study size was based on the ability to detect a 30% difference in proportion of women undergoing ECV (from 50 to 80%). To detect this difference with a two sided significance level (α) of 0.05, a power (1-ß) of 0.80, adjusting for an intra-cluster correlation (ICC) of 0.10, and taking a dropout rate of individual clients of 15% into account, the calculated sample size was 35 clients per cluster and 6 clusters per strategy. Evidence for an appropriate assumption of the ICC for this trial is limited. As clusters encompass both hospitals and midwifery practices, the ICC was assumed to lie somewhere in between the ICCs for studies clustered by hospitals and those clustered by

188 Chapter 10 Implementation of client versus care provider strategies general practitioners.34 Based on the incidence of breech presentation and the size of the smaller clusters, we calculated that a nine month inclusion period would be appropriate to include at least 35 clients.

Interventions Based upon a previously performed determinant analysis we developed two implementation strategies to improve the implementation of ECV: a client strategy and a care provider strategy.27,32,33 All clients received standard care. In addition, clusters were subjected to one or both of the implementation strategies. In clusters allocated to the client strategy women with a fetus in breech presentation after 32 weeks received a leaflet explaining advantages, benefits and risks of ECV and a decision aid to assist them in clarifying personal values and in choosing the treatment which suited their personal values. The leaflet encouraged clients to involve family and significant others in their deliberation, and to discuss all uncertainties with their care providers. The decision aid was originally developed by an Australian research group of Nassar et al., and translated into Dutch.35 Minor adjustments were made to the leaflet and decision aid after pilot testing the comprehensibility among women with a breech presentation and care providers (n=6). During a kick- off meeting, all participating care providers received the information leaflet and decision aid as well as a short user instruction. They were asked to hand out the leaflet and decision aid to all women with a breech presentation from 32 weeks onwards. In clusters allocated to the care provider strategy, care providers were invited for a one day counselling course. Participants were informed on the guidelines and relevant literature emphasizing the importance, benefits and risks of ECV. Furthermore, participants were trained in counselling and guiding clients in informed decision making by a counselling coach. After this course, client recruitment was started.

Data collection, outcome measures and statistical analysis Data were prospectively documented in study-files. All clients were handed a study-file at the moment breech was diagnosed from 32 weeks onwards, and were requested to bring their file to all subsequent appointments, until ten days after delivery. Care providers were asked to document all provided care in the study-file, including reasons not to offer ECV, items mentioned during counselling, ECV outcomes and birth outcome. Study-files also included

189 Chapter 10 Implementation of client versus care provider strategies

client-administered questionnaires on provided care, in order to assess if the implemented strategy was provided to the client. Care providers within the four strategies received regular feedback on the number of participants, to remind them to work according to the allocated strategy. Apart from the study files, all eligible clients were identified by checking ECV attempt registers and birth registers for breech birth. For women who had not received a study-file, data were retrospectively extracted from the hospital charts. The primary outcome of the study was the ECV implementation rate, defined as the number of women who underwent a version attempt as percentage of all women with a breech presentation from 36 weeks gestation eligible for external cephalic version. Secondary outcomes were the percentage of women counselled and opting for a version attempt. Success of ECV was defined as a fetus being in cephalic presentation 30 minutes after the procedure. Outcome measures pertain to the cluster level. Firstly, we calculated the proportion of women who were offered an ECV, who opted for ECV and who actually underwent ECV. Subsequently, we used a logistic regression accounting for clustering of observations (generalized estimating equations, or GEE) to analyze the effect of the interventions onthese outcomes. This resulted in an Odds ratio, expressing the probability that a client would receive an ECV within a cluster allocated to the implementation strategy, relative to those clusters not exposed to the strategy. An interaction term was included in the model to assess whether the combination of the client and provider strategy results in an additional increased (or decreased) effect. All analyses were performed according to intention-to-treat analyses: clusters were analyzed in the trial arm they were randomized to, regardless of compliance to the implementation strategy; and clients were included regardless of whether or not they were exposed to the implementation strategy (leaflet and decision aid or trained care provider). We set the level of significance at p<0.05. A descriptive analysis was carried out on data from the study files documenting the process of care, and client questionnaires in order to identify whether the implementation strategy was carried out as planned. In a per-protocol analysis, analyses were repeated in the subgroup of women who were actually subjected to the implementation strategy.

190 Chapter 10 Implementation of client versus care provider strategies

RESULTS We invited 40 clusters to participate in the study, of which 26 clusters agreed and were randomized. One cluster randomized to combined client and care provider strategy withdrew before implementation of the strategy because of anticipated lack of time to participate. The remaining 25 clusters consisted of 12 teaching hospitals, 10 general hospitals, and three tertiary hospitals with a total of 179 midwifery practices. The size of the clusters ranged from 6,650 to 11,100 births per year (Table 1). Twelve clusters (48%) had a special ECV office hour at time of randomization. Clusters were randomized in the period January 2011 to December 2011. The follow up of clusters after implementation of the strategies was closed at August 31st 2012. In total 1,793 women with a singleton fetus in breech presentation from 32 weeks onwards or unexpected breech presentation at birth were included, with a range of 13 to 159 per cluster. Of these women, 180 (10%) had an apparent reason to not undergo had ECV: 72 (4.0%) had an absolute contraindication for ECV according to the Dutch guidelines, 46 (2.6%) had an indication for a planned cesarean section, and in 62 (3.5%) of women breech presentation was not diagnosed until onset of labor (Figure 1). Thus 1,613 women were eligible for ECV. The mean age was 31 years (range 15 – 45) (Table 1). The percentage of nulliparous women ranged from 53 to 64% between the four strategies (Table 1).

191 Chapter 10 Implementation of client versus care provider strategies

= 1 ) f o r e t a i n g d d n o a t ( p r o v i d e W i t h d r e w b = 1 7 ) = 8 ) N 4 r s t a e g y = 1 9 ) 2 9 0 ( N 4 9 2 =

2 N 3

c l u s t e r = 3 4

= 4 ) : 6 N 1 l b i r t h N n o m a l y p r a e v i =

n t s w i h b r e c ( p r e s n t a i o N ) = 5 P a t i e v a g i n ) P R O M I e c h n o t d i a g s u l b r l a c e n t P ( O l i g o h y d r a m n s 2 H y p e r t n s i o P U t e r i n a F e t a l g r o w h s i c n > 1 p r e v i o u s S C C r s ( n

o t e l i g b c l i e n t a d r p o v

N -‐ A b s o l u t e C I f r E V -‐ I n d i c a t o f r p m y C S ( -‐ B r e ( N S t u d y p o l a i n : p r o v i d e ) = 9 ) = 6 ) ( n ( N 2 ) e c h r R C T ( N = 8 1 3 2 2

) N = 3 7 6 1 S C c l u s t e r

= 3 9 5

= 1 9 ) : N 0 i n c l u s t e l b i r t h 3 N = 6 p r e s n t a i o N C I f o r E V 4 r e g i o n s = 4 0 ) 1 u t e = P a t i e n s w h b r v a g i n

c a r e p o v i d s t g y 1 p r e v i o u s l i g b e ( I e c h n o t d i a g s u l b r l a c e n t p r v i ( N ( P ) R O M O l i g o h y d r a m n s U t e r i n a o m l y F e t a l g r o w h s i c n H y p e r t n s i o > C P o t e = 2 6 )

N -‐ A b s o l -‐ I n d i c a t o f r p m y C S -‐ B r e ( N = 5 S t u d y p o l a i n : N ( s o p t i m a l e n d c r g o t i m e p a r c ) E C V i N p a t i n g R C T -‐ -‐ = 2 ) l i g b e f o r n c u s ( N R a n d o m i s e ) 6 s e

r o t p a r i c r s g y N 7 5 1 1 3

C l u s t e N = 5 6 2 u s t e 7 : 6 4 7 S C e v i a =

= 8 5 ) : N l b i r t h 4 1 n o m a l y

( N = 7 c l p r a n t s w i h b r e c p r e s n t a i o N c l i e n t s r a p r e v i o u s P a t i e P R O M v a g i n ) l i g b e I e c h n o t d i a g s u l b r l a c e n t U t e r i n a > 1 F e t a l g r o w h s i c n H y p e r t n s i o O l i g o h y d r a m n s ( P C P o t e

I n d i c a t o f r p m y C S ( N B r e A b s o l u t e C I f r E V ( N = 3 2 S t u d y p o l a i n N -‐ -‐ ( N = 3 1 -‐

)

3 = 1 3 ) 1 3 9 4 = 3 8 5 0 = 4 1 7 = 3 2 ) : b i r t h 3 N Flow chart Flow p r a e v i = 7 c l u s t e r n t s w i h b r e c

p r e s n t a i o c a r e s u l N P a t i e v a g i n l c h n o t d i a g s e u l b r 1 p r e v i o u s S C l i g b e ( N c a t i o n f r p m y C S ( N = 1 0 U t e r i n a o m l y F e t a l g r o w h s i c n H y p e r t n s i o O l i g o h y d r a m n s > C I P l a c e n t o t e

-‐ A b s o l u t e C I f r E V ( N N ( P ) R O M 0 -‐ I n d i -‐ B r e ( N = 9 ) S t u d y p o l a i n : N

Figure 1. Figure

192 Chapter 10 Implementation of client versus care provider strategies

9) 9) 121) 44) 10) 17)

2 5 49 (3 -‐ 49 174 (58) 174 29 (4 -‐ 29 29 (2 -‐ 29 112 (1 -‐ 112 31 (17 -‐ 31 Client + care + Client 334 (19 -‐ 334 provider strategy provider

6) 6) 6) 9) 159) 45)

3 6 46 (3 -‐ 46 20 (3 -‐ 20 20 (0 -‐ 20 222 (64) 222 110 (1 -‐ 110 30 (15 -‐ 30 Care Care 395 (19 -‐ 395 strategy provider provider

8) 8) 8) 117) 44) 13)

3 7 Client 41 (3 -‐ 41 41 (4 -‐ 41 strategy 307 (60) 307 40 (4 -‐ 40 181 (1 -‐ 181 31 (17 -‐ 31 647 (47 -‐ 647

6) 8) 8) 111) 42) 12) -‐

4 7 (5 69 (4 -‐ 69 32 (4 -‐ 32 218 (53) 218 184 (1 -‐ 184 31 (17 -‐ 31 417 (13 -‐ 417 Care as usual as Care 58

Characteristics of participating clusters participating of Characteristics ractices (hospital and midwifery)(range) No. of p No. of clusters with special ECV office hours Table 1. Table Cluster characteristics No. of clusters Nulliparous (%)* among clusters value highest and = lowest Range ^Age: 79 (4.4%) missing values *Parity: 214 (11.9%) missing values missing (11.9%) 214 *Parity: No. care providers per cluster (range) gynaecologists midwifes within hospital (range) hospital within midwifes midwives out of hospital (range) hospital of out midwives Client characteristics No of clients (range per cluster) (range)^ (years) Age

193 Chapter 10 Implementation of client versus care provider strategies

Implementation of ECV Overall, the ECV implementation rate attempts was 72%. ECV implementation rates varied considerably across clusters: from 8 to 95% (Figure 2). Four clusters scored below a 50% implementation rate (with a range of 8-46%). Ten (40%) clusters scored an implementation rate of 80% or higher (range 84-95%). Five out of ten clusters with a good implementation rate were clusters in the care as usual group. Sixteen (67%) of the clusters had an implementation rate above the mean. The ECV implementation rate in the clusters with the client strategy was 71% (range 46-95%) compared to 74% (range 8-93%) in the clusters without this strategy (OR 0.78 ((95% CI 0.41-1.5)) (Table 2). The proportion of ECV attempts within the clusters with the care provider strategy was 73% (range 23-94%) compared to 72% (range 8-95%) in the clusters without this strategy (OR 1.2 (95% CI 0.64-2.3)) (Table 3). The combination of both the client and care provider strategy did not result in an additional effect, as the interaction term between ECV implementation rate and the combined intervention was not statistically significant. In order to explain why ECV implementation rates vary, we looked at the percentage of counselled clients and clients opting for ECV. The implementation for ECV counselling did not significantly differ between the groups: 82% in the clusters with the client strategy and 80% in the clusters with the care provider strategy. Counselling rates were not significantly different between.

Figure 2. Range of ECV implementation rate amoung clusters. The horizontal line represents the mean ECV implementation rate.

100

10 % ECV attempts amoung eligible clients eligible amoung attempts ECV %

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Number of Clusters

194 Chapter 10 Implementation of client versus care provider strategies

Also, the percentage of clients opting for ECV (87% within the client strategy and 91% in the care provider strategy) was not significantly different either one of the strategies and the clusters without these strategies.

Of 1,613 eligible clients, 153 (9.5%) women declined an ECV attempt and 291 (18.0%) were not counselled for ECV by their care provider. In the minority of cases, the study-file or hospital chart provided a reason for not offering ECV (Table 4). The percentage of clients who were not counselled without further explanation, was 20 out of 385 eligible clients (5%) in the care as usual cluster, 83 of 562 (15%) within the client strategy cluster, 94 of 376 (25%) within the care provider strategy cluster, and 22 of 290 (8%) in the cluster where both implementation strategies were enrolled ((OR 0.71 (95% CI 0.34-1.5) for the client strategy, and OR 0.84 (95% CI 0.41-1.7) for the care provider strategy).

Table 3. Effect of care provider strategy on counselled clients, clients who opt for ECV and the number of attempts, among all eligible women for ECV

No care provider Care provider strategy Care provider strategy OR strategy (95%CI)

N= 947 N= 666

(%; range*) (%; range*)

Counselled for ECV (of eligible 791 (84; 15-‐100) 531 (80; 35-‐93) 1.1 (0.53-‐2.20) clients)

Opting for ECV (of counselled clients) 692 (88; 51-‐100) 485 (91; 63-‐100) 1.2 (0.64-‐2.2)

ECV attempts (of eligible clients) 686 (72; 8-‐95) 483 (73; 23-‐94) 1.2 (0.64-‐2.3)

*range among clusters included in the strategy

195 Chapter 10 Implementation of client versus care provider strategies

40) 12 ) 20)

0 2 0 3 1 4 4 Care provider Care

strategy +

14 (4; 0 -‐ (4; 14 22 (8; 3 -‐ (8; 22 36 (12; 6 -‐ (12; 36 N= 290 (%, range*) (%, 290 N= Client

) 57) 65) 7 -‐

0 1 0 1 0 3 0 strategy 5 (1; 0 94 (25; 7 -‐ (25; 94 99 (26; 7 -‐ (26; 99 Care provider provider Care N= 376 (%, range*) (%, 376 N=

45) 30) 13 )

1 4 1 0 9 4 strategy 12

31 (5; 1 -‐ (5; 31 83 (15; 0 -‐ (15; 83 114 (20; 5 -‐ (20; 114 Client trateg N= 562 (%, range*) (%, 562 N=

85) 15) 77 ) -‐ 0

1 5 0 0 2 2 12

20 (5; 0 -‐ (5; 20 22 (5; 0 -‐ (5; 22 42 (11; (11; 42 Care as usual as Care N= 385 (%, range*) (%, 385 N=

ed for unknown reason unknown for ed

l Reasons why eligible clients were not counselled for ECV. not were clients eligible why Reasons

*% per strategy and range of % among clusters included in the s

HIV, Hep B, anticoagulants Macrosomia loss blood Vaginal Adipositas Multiple factors decreasing success factors Multiple rate

One previous CS previous One Other Table 4. Table NOT COUNSELED contraindication than other Reasons guidelines in counsel Not

196 Chapter 10 Implementation of client versus care provider strategies

Per-protocol analysis To investigate whether the implementation strategies were not efficacious themselves or did not prove to be effective due to a suboptimal adherence to the strategy-protocols by the care providers, we performed a post-hoc analysis of those clients who were certainly exposed to the implementation strategy. Certain exposure was defined as handing out the client leaflet and decision aid, or fulfilling a list of topics on ECV and breech birth mentioned during counselling, for the client and care provider strategy respectively. We were able to ascertain exposure for 168 out of 852 clients (20%) within the clusters randomized to the client strategy and for 150 out of 666 clients (23%) within the clusters randomized to the care provider strategy. These subgroups did not significantly differ in baseline characteristics compared to the complete study sample (on parity or age). Of these 318 clients, 281 (88%) fulfilled the baseline questionnaire in which 170 (60%) were a priori willing to opt for ECV before counselling. The results of this analysis are shown in Table 5 and 6.

Table 5. Per protocol analysis: Effect of client strategy on counselled clients, clients who opt for ECV and

the number of ECV attempts, among all eligible women for ECV; in those clients who were actually exposed

to the implementation strategies.

No client strategy Client strategy Client strategy

N= 193 (%) N= 168 (%) OR (95%CI)

Counselled for ECV (of eligible clients) 169 (88) 150 (89) 0.76 (0.37-‐1.6)

Opting for ECV (of counselled clients) 166 (98) 143 (95) 0.58 (0.27-‐1.3)

ECV attempts (of eligible clients) 165 (85) 143 (85) 0.64 (0.30-‐1.4)

Table 6. Per protocol analysis: Effect of care provider strategy on counselled clients, clients who opt for ECV

and the number of ECV attempts, among all eligible women for ECV; in those clients who were actually

exposed to the implementation strategies.

No care provider Care provider Care provider strategy OR strategy strategy (95%CI)

N= 211 (%) N=150 (%)

Counselled for ECV (of eligible clients) 186 (88) 133 (89) 1.2 (0.56-‐2.4)

Opting for ECV (of counselled clients) 177 (95) 132 (99) 1.9 (0.83-‐4.4)

ECV attempts (of eligible clients) 176 (83) 132 (88) 2.0 (0.72-‐4.5)

197 Chapter 10 Implementation of client versus care provider strategies

The number of clients counselled, opting for, and having an ECV attempt were higher in all strategies, compared to the intention to treat analyses. Again, no significant difference on ECV attempts was found between the clusters with the client strategy compared to the clusters without this strategy (OR 0.64 (95% CI 0.30-1.4)) or between the clusters with the care provider strategy compared to the clusters without this strategy (OR 2.0 (95% CI 0.72-4.5)). There was no significant interaction when both strategies were implemented; ECV implementation rates were not higher compared to care as usual.

DISCUSSION In this large cluster-randomized clinical trial, we were unable to demonstrate a significant effect of both the client and care provider strategies compared to usual care on the ECV implementation rate. Both in an intention to treat analysis, as well as in a per protocol analysis, in which we only analyzed the women that were counselled for ECV after the implementation strategy was applied correctly. Our study has several strengths and weaknesses. An important strength of the study is that it provides insight in prelabour management of breech presentation in a large cohort, and reasons why ECV is not offered. This knowledge is important to make care providers aware of substandard care with respect to guideline adherence. Withdrawal of one cluster in the client and care provider strategy caused an imbalance in the number of clusters per implementation strategy. Since we used a factorial design to analyze the effect of the implementation strategies, we do not think that, apart from further reduction in power, this baseline imbalance substantially affected the outcomes of the trial. Furthermore, we observed a fair overall ECV attempt rate of 72%, but considerable practice variation resulting in lower power to demonstrate potential effects of the implementation strategies. A main weakness is that many care providers did not adhere to the strategies as intended. In other words, they did not always handed out leaflets, decision aids or discussed all topics on ECV and breech presentation during counselling. However, if we limited our analysis to those women in which the strategies were followed as intended, also no significant differences were found. A first explanation for the low exposure of clients to the implementation strategies was low. An explanation may be that the implementation of the strategies encompassed several steps, and some could have been better

198 Chapter 10 Implementation of client versus care provider strategies controlled then others. All care providers were invited for the day course in the care provider strategy and the kick-off meeting for the client strategy. We aimed to have at least one care provider per practice to attend these meetings. This goal was achieved in almost all clusters. Subsequently, care providers were requested to pass on this knowledge to colleagues within their practices and to send all clients for counselling to those care providers who attended the meeting. Disclosure of the study information to all other care providers was beyond our control. Unrolling strategies to single care providers instead of practices, personal communication on initial implementation and during follow-up, would most certainly have resulted in better implementation rates, which was demonstrated in other studies.36,37 The introduction and actual implementation of innovations like this - involving all care providers in one region - may take up to two years which requires considerably more capacity and financial resources than we had at our disposal. A second explanation of the low implementation rate of the strategies could be found in the low incidence of breech presentation; breech presentation concerns 10% of pregnancies at 32 weeks gestation, but only 3-4% at 37 weeks gestation. Consequently, on average only one to three clients presented with breech presentations to each care provider during the study period. Due to this low incidence, handing out leaflets, decision aids or providing full counselling will not easily become part of routine practice. We sent reminders, feedback on inclusion and ascertained the availability of information leaflets in order to improve this suboptimal adherence to the implementation strategy, apparently with limited success. Concentration of care in a region, i.e. a small group of care providers taking care for women with a baby in breech position, could be a solution for this. Similar strategies have been successful in other areas, for example in Parkinson disease.38

The per protocol analyses showed better ECV attempt rates among the clients exposed to the implementation strategies compared to the intention to treat analysis. Again, differences between strategies were not significant, but as this subgroup has considerably lower power, we cannot rule out that strategies may potentially be effective if better adhered to. Thisper protocol analysis may have suffered from selection bias, resulting in lower contrasts between strategies; care providers potentially may have handed out study files more often to those clients willing to consider ECV instead of those clients requesting a planned cesarean section. In view of odds ratio of 2 that we found for the health care provider strategy, uptake of ECV could

199 Chapter 10 Implementation of client versus care provider strategies

potentially be improved by education and awareness of the care provider. We should take into account that in the per protocol analysis, that was limited to women who were actually informed by the care provider, the uptake of ECV was already very high (88%), thus limiting the room for improvement. The potential of this intervention when the ECV rate is low, is potentially higher. Another explanation for the limited contrasts that were observed in the intention to treat and per protocol analyses could be the so called Hawthorne effect: care providers in the care as usual clusters might have put more effort in the counselling for ECV as they knew they were being monitored. Finally, the results may be the consequence of an imbalance in implementation rate of ECV at baseline. The data showed substantial variability among clusters, which might have existed at start of the trial. We did not have implementation rates per cluster at randomization and were not able to stratify for this. Due to the variation, the ICC will be higher as assumed in our power calculation. This results in lower power to demonstrate potential effects of the implementation strategies.

A strength of our study is that we have also monitored the actual exposure of clients to the implementation strategies, which may be a realistic explanation for our findings. Our study reflects circumstances in daily practice. Just as there is extensive empirical evidence that the level of use of guidelines affects outcomes in clients,39 this is also true for adherence to implementation strategies. We need to take into account whether a strategy has actually been put into practice by professionals in order to determine its benefits. Otherwise, one may incorrectly conclude that an implementation strategy is not efficacious in itself when, in fact, it has not been properly implemented. There might be lack of publications on implementation strategies with insignificant results, potentially indicating a publication bias in favour of those studies with significant results. This might hinder dissemination of knowledge on developed strategies which are effective itself, but which was not shown in the trial due to other influences. Therefore, other research groups may benefit from existing results. Our data showed that the overall ECV implementation rate is 72%. One could argue that this is quite acceptable, as a considerable proportion of women will not opt for ECV, even after counselling. However, we think there is still room for improvement considering the large practice variation and the fact that every elective cesarean section results in an increased maternal and neonatal morbidity and mortality in subsequent pregnancies.40

200 Chapter 10 Implementation of client versus care provider strategies

Another finding in our study related to the implementation of ECV is that breech presentations are not accurately diagnosed by care providers. Four percent (62 / 1,793) of breech presentations were not diagnosed until birth, exposing these foetuses to high risks of intrapartum morbidity and mortality. A study by Lau et al. in 1997 already reported on this problem and an intervention trial by Robberts et al. tried to lower the percentage of undiagnosed breech by increasing awareness among care providers and pregnant women to asses fetal presentation in the last trimester.41,42 The need for routine ultra sound assessment to confirm the presentation of the foetuses might be indicated to rule out this problem. Improvement of implementation of guidelines with the focus on contraindications could also improve ECV implementation rates by 4 to 5 % as more contra-indications are reported in daily practice than given in the guideline. In conclusion, in this cluster randomized controlled trial neither a provider based nor a client based strategy improved ECV implementation. The implementation rates were on average fair, although low in some clusters. Improvement could be realized in those regions with poor implementation rates either by improving care within that region with a specific target at the care provider, or by allocation of management of breech presentation to those centres with high ECV implementation and success rates.

201 Chapter 10 Implementation of client versus care provider strategies

REFERENCES 1 Hickok, DE; Gordon, DC; Milberg, JA; Williams, MA; Daling J. The frequency of breech presentation by gestational age at birth: a large population-based study. Am J Obs Gynecol [Internet]. 1992 [cited 2014 Mar 3]; 166: 851–852. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/1550152 2 Hannah ME, Hannah WJ, Hewson S a, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet [Internet]. 2000 Oct;356 : 1375–1383. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0140673600028403 3 Whyte H, Hannah ME, Saigal S, Hannah WJ, Hewson S, Amankwah K, et al. 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 [Internet]. 2004 Sep [cited 2014 Mar 2]; 191: 864–871. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15467555 4 Rietberg C. Term breech presentation in The Netherlands from 1995 to 1999: mortality and morbidity in relation to the mode of delivery of 33,824 infants. BJOG An Int J Obstet Gynaecol [Internet]. 2003 Jun [cited 2014 Mar 3]; 110: 604–609. Available from: http://doi.wiley.com/10.1016/S1470-0328(03)01907-4 5 Bragg F, Cromwell DA, Edozien LC, Gurol-urganci I, Mahmood TA, Templeton A, et al. Variation in rates of caesarean section among English NHS trusts after accounting for maternal and clinical risk : cross sectional study. BMJ. 2010;341 . 6 Sullivan EA, Moran K, Chapman M. Term breech singletons and caesarean section: a population study, Australia 1991-2005. Aust N Z J Obstet Gynaecol [Internet]. 2009 Oct [cited 2014 Mar 3]; 49: 456–460. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/19780725 7 US National Institutes of Health C. Caesarean section-the first cut isn’t the deepest. Lancet. 2010; 375: 956. 8 Nielsen TF, Hökegård KH, Ericson a. Cesarean section and perinatal mortality in Sweden in 1981. Acta Obstet Gynecol Scand [Internet]. 1986 Jan; 65: 865–867. Available from: http://www.ncbi.nlm.nih.gov/pubmed/3825528 9 Van Ham M a, van Dongen PW, Mulder J. Maternal consequences of caesarean section. A retrospective study of intra-operative and postoperative maternal complications of caesarean section during a 10-year period. Eur J Obstet Gynecol Reprod Biol [Internet]. 1997 Jul; 74: 1–6. Available from: http://www.ncbi.nlm.nih. gov/pubmed/9243191 10 Rasmussen, SA; Maltau J. Complications following cesarean section. Tidsskr Nor Laegeforen [Internet]. 1990 [cited 2014 Mar 3]; 110: 351–353. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/2309178

202 Chapter 10 Implementation of client versus care provider strategies

11 Hofmeyr, GJ; Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012; 10. 12 Grootscholten, K; Kok, M; Oei, SG; Mol, BW; van der Post J. External cephalic version-related risks: a meta-analysis. Obs Gynecol [Internet]. 2008 [cited 2014 Mar 20]; 112: 1143–1151. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/ pubmed/?term=xternal+cephalic+version-related+risks:+a+meta-analysis. 13 The Royal Australian and New Zealand College of Obstetricians and Gynecologists. Management of Breech Presentation at Term. 2001. p. file:///C:/Users/PS13C/ Downloads/C–Obs%2011%20Mana. 14 Otterlo Werkgroep dr. AJ van L en dr. MG van P. NVOG richtlijn - Stuitligging- versie 2.0 (Guideline on breech presentation - version 2.0 of the Dutch Society of Obstetrics and Gynecology) [Internet]. 2008. Available from: file:///C:/Users/PS13C/ Downloads/Stuitligging (1).pdf 15 Verburgt, T; Offerhaus P. Uitwendige versie (External Cephalic Version). NTOG. 2007; 120: 4–7. 16 Fleuren M, Grol R, de Haan M, Wijkel D, Oudshoorn C. Adherence by midwives to the Dutch national guidelines on threatened miscarriage in general practice: a prospective study. Qual Health Care [Internet]. 1997 Jun;6 : 69–74. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1055455&tool=pmcent rez&rendertype=abstract 17 Grimshaw J, Eccles M, Thomas R, MacLennan G, Ramsay C, Fraser C, et al. Toward evidence-based quality improvement. Evidence (and its limitations) of the effectiveness of guideline dissemination and implementation strategies 1966-1998. J Gen Intern Med [Internet]. 2006 Feb [cited 2014 May 7]; 21 Suppl 2: S14–20. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=255713 0&tool=pmcentrez&rendertype=abstract 18 Grol, R; Wensing, M; Eccles M. Improving Patient Care: The Implementation of Change in Clinical Practice [Internet]. Elsevier; 2005 [cited 2014 May 19]. p. 304. Available from: http://www.amazon.com/Improving-Patient-Care-Implementation- Clinical/dp/075068819X 19 Guldbrandsson K. From news to everyday use The difficult art of implementation. Stockholm: Strömberg; 20 Dijs-elsinga, J; Van Huis, AM; De Miranda, E; Ravelli, ACJ; Tamminga, P; Verdonk- Wesenbeek, MJH; de Winter, GA; Hukkelhoven, CWPM; SChiere A. 10 jaar perinatale registratie, de grote lijnen (10 years of Perinatal Care in the Netherlands, the big picture) [Internet]. 2011. Available from: http://www.perinatreg.nl/ uploads/173/123/10_jaar_Perinatale_Zorg_in_Nederland_de_grote_lijnen.PDF

203 Chapter 10 Implementation of client versus care provider strategies

21 Feitsma, HA; Middeldorp JM, Oepkes D. De uitwendige versie bij de a terme stuit; een inventariserend onderzoek (ECV for breech presentation at term; an inventory study). NTOG. 2007; 120: 4–6. 22 Caukwell S, Joels L a, Kyle PM, Mills MS. Women’s attitudes towards management of breech presentation at term. J Obstet Gynaecol [Internet]. 2002 Sep [cited 2014 Mar 20]; 22: 486–488. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/12521413 23 Bewley, S; Shennan A. Peer review and the Term Breech Trial. Lancet. 2007; 369: 2007. 24 Leung, TY; Lau, TK; Lo, KW; Rogers M. A survey of pregnant women’s attitude towards breech delivery and external cephalic version. Aust N Z J Obs Gynaecol. 2000; 40: 253–259. 25 Raynes-Greenow CH, Roberts CL, Barratt A, Brodrick B, Peat B. Pregnant women’s preferences and knowledge of term breech management, in an Australian setting. Midwifery [Internet]. 2004 Jun [cited 2014 May 19]; 20: 181–187. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15177862 26 Yogev Y, Horowitz E, Ben-Haroush a, Chen R, Kaplan B. Changing attitudes toward mode of delivery and external cephalic version in breech presentations. Int J Gynaecol Obstet [Internet]. 2002 Dec; 79: 221–224. Available from: http://www. ncbi.nlm.nih.gov/pubmed/12445986 27 Vlemmix F, Rosman AN, Fleuren MAH, Rijnders MEB, Beuckens A, Haak MC, et al. Implementation of the external cephalic version in breech delivery. Dutch national implementation study of external cephalic version. BMC Pregnancy Childbirth. 2010; 10: 20. 28 Fleuren M, Wiefferink K, Paulussen T. Determinants of innovation within health care organizations: literature review and Delphi study. Int J Qual Health Care [Internet]. 2004 Apr; 16: 107–123. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/15051705 29 Glasgow RE, Vogt TM, Boles SM. Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health [Internet]. 1999 Sep; 89: 1322–1327. Available from: http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=1508772&tool=pmcentrez&rendertype=abstract 30 Greenhalgh, T; Robert, G; Macfarlane, F; Bate, P; Kyriakidou O. Diffusion of innovations in service organizations: systematic review and recommendations. Milbank Q [Internet]. 2004 [cited 2014 May 19]; 82: 581–629. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/15595944

204 Chapter 10 Implementation of client versus care provider strategies

31 Bartholomew, LK; Parcel, GS; Kok, G; Gottlieb N. Planning health promotion programs: an intervention mapping approach. [Internet]. San Francisco: Jossey- Bass; 2011 [cited 2014 May 19]. Available from: http://www.amazon.com/s/ref=nb_ sb_noss?url=search-alias=aps&field-keywords=planning+health+care+promotions+ bartholomew&rh=i:aps,k:planning+health+care+promotions+bartholomew 32 Rosman AN, Vlemmix F, Beuckens A, Rijnders ME, Opmeer BC, Mol BWJ, et al. Facilitators and barriers to external cephalic version for breech presentation at term among health care providers in the Netherlands: a quantitative analysis. Midwifery [Internet]. Elsevier; 2014 Mar [cited 2014 May 6]; 30: e145–50. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24530121 33 Rosman a N, Vlemmix F, Fleuren M a H, Rijnders ME, Beuckens a, Opmeer BC, et al. Patients’ and professionals’ barriers and facilitators to external cephalic version for breech presentation at term, a qualitative analysis in the Netherlands. Midwifery [Internet]. Elsevier; 2014 Mar [cited 2014 May 6]; 30: 324–330. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23664312 34 Cosby RH, Howard M, Kaczorowski J, Willan AR, Sellors JW. Randomizing patients by family practice: sample size estimation, intracluster correlation and data analysis. Fam Pract [Internet]. 2003 Feb; 20: 77–82. Available from: http://www.ncbi.nlm.nih. gov/pubmed/12509376 35 Nassar N, Roberts CL, Raynes-Greenow CH, Barratt a, Peat B. Evaluation of a decision aid for women with breech presentation at term: a randomised controlled trial [ISRCTN14570598]. BJOG [Internet]. 2007 Mar [cited 2014 Mar 20]; 114: 325– 333. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=24 08658&tool=pmcentrez&rendertype=abstract 36 Fleuren M, Dusseldorp E, van den Bergh S, Vlek H, Wildschut J, van den Akker E, et al. Implementation of a shared care guideline for back pain: effect on unnecessary referrals. Int J Qual Health Care [Internet]. 2010 Oct; 22: 415–420. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2939752&tool=pmcent rez&rendertype=abstract 37 Wiefferink CH, Poelman J, Linthorst M, Vanwesenbeeck I, van Wijngaarden JCM, Paulussen TGW. Outcomes of a systematically designed strategy for the implementation of sex education in Dutch secondary schools. Health Educ Res [Internet]. 2005 Jun [cited 2014 May 14]; 20: 323–333. Available from: http://www. ncbi.nlm.nih.gov/pubmed/15385429 38 Bloem, BR; Munneke M. Revolutionising management of chronic disease : the ParkinsonNet approach. BMJ. 2014; 1838: 1–8.

205 Chapter 10 Implementation of client versus care provider strategies

39 Durlak J a, DuPre EP. Implementation matters: a review of research on the influence of implementation on program outcomes and the factors affecting implementation. Am J Community Psychol [Internet]. 2008 Jun [cited 2014 May 12]; 41: 327–350. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18322790 40 Shennan a, Bewley S. How to manage term breech deliveries. BMJ [Internet]. 2001 Aug 4; 323: 244–245. Available from: http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=1120869&tool=pmcentrez&rendertype=abstract 41 Roberts CL, Cameron C a, Nassar N, Raynes-Greenow CH. A simple patient-initiated intervention to increase antenatal detection of breech presentation at term. Paediatr Perinat Epidemiol [Internet]. 2004 Sep; 18: 371–376. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15367324 42 Lau, TK: Lo, KW; Wan DRM. Predictors of successful external cephalic version at term: a prospective study. BJOG [Internet]. 1997 [cited 2014 May 19];104 : 798– 802. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/9236644

206

11 Cost-effectiveness of improving implementation of external cephalic version for term breech presentation.

Floortje Vlemmix

Ageeth N. Rosman

Margot A.H. Fleuren

Marlies E. Rijnders

Antje Beuckens

Ben Willem J. Mol

Joris A.M. van der Post

Marjolein Kok

Brent C. Opmeer

Submitted

209 Chapter 11 Cost-effectiveness of improving implementation

ABSTRACT

Objective To assess cost-effectiveness of two implementation strategies developed to improve the external cephalic version rate compared to care as usual. Design Economic evaluation based on the results of a randomised controlled trial on the implementation of two strategies to improve external cephalic version rates. Clusters were randomised to the client strategy (information leaflets and decision aid), a care provider strategy (one day counselling course focussed on knowledge and counselling skills), or to both strategies combined. For this economic evaluation, the baseline implementation rate of external cephalic version in each cluster was used as comparator. Setting 25 clusters, each consisting of a hospital and referring midwifery practices in the Netherlands. Population Pregnant women with a breech presentation from 32 weeks onwards. MethodsModel based cost-effectiveness analysis from a hospital perspective. Main outcome measures We estimated direct medical costs in euros per client, associated with healthcare utilisation from start of the implementation strategies until the version attempt, and until birth. We calculated the incremental cost-effectiveness ratio as costs per extra version attempt, and per prevented caesarean section. Results The combination of the client and care provider implementation strategy has the best potential to improve external cephalic version rates (mean increase 7%, range among clusters 3-44%). The extra costs per client associated with the implementation strategy were €83, the incremental cost-effectiveness ratio to gain one ECV attempt was €1,119 (range €190 - €2,659). If the effect of the implementation strategy sustains for two years after implementation, the extra costs per patient would decrease to €64 (incremental cost-effectiveness ratio €865 (range €147 – €2,055)). As ECV attempt rates were only modestly improved due to poor adherence to implementation strategies, caesarean section rates were more or less unaffected. Conclusions Implementation of external cephalic version rates can be improved by a strategy focussing on both clients and care providers. Whether this strategy is cost-effective depends on the societal willingness to pay for an additional attempt, and the sustainability of the effect of the implementation strategy.

210 Chapter 11 Cost-effectiveness of improving implementation

INTRODUCTION Breech presentation occurs in 3 to 4% of all term pregnancies. Since the publication of the Term Breech Trial results, the majority of breech presentations at birth are managed by elective caesarean section. This randomised controlled trial reported a significant increased neonatal morbidity and mortality during planned vaginal breech delivery.1 This has led to elective caesarean sections rates up to 90% in the UK and 96% in Australia.2,3 The downside of these caesarean sections is the increased maternal and neonatal morbidity and mortality in subsequent pregnancies due to the scarred uterus.4–6 Whether the gain in neonatal outcome due to planned caesarean section for breech presentation outweighs the neonatal risks in the subsequent pregnancy, is still a topic of debate Thus, prevention of breech presentation at birth should have a high priority in every obstetric unit. External cephalic version (ECV) is an effective and safe treatment to prevent breech presentation at birth.7,8 It is proven to be a cost effective intervention when compared to the planned caesarean section, if success rates of ECV are above the 32% (society’s perspective, USA).9 However, recommendation of ECV in professional guidelines, has not automatically led to an optimal implementation of this intervention.10–13 We have developed two implementation strategies to improve the ECV implementation rate in the Netherlands.14–16 The aim of this economic evaluation study was to estimate the costs associated with these implementation strategies and the incremental costs in order to achieve an extra ECV attempt, and to prevent a caesarean section.

METHODS

Study design A model based economic evaluation was performed, where results from a cluster randomised controlled trial (RCT) of the implementation of ECV were used to estimate parameters in the model.14–16 This trial compared the effectiveness of two implementation strategies (a client and a care provider strategy) on the ECV rate within clusters (a hospital and the referring midwifery practices). Clusters were randomised in a two-by-two factorial design meaning that they were assigned to either the client strategy, the care provider strategy, both implementation strategies, or care as usual. Within the client strategy, clients received an information leaflet on ECV and

211 Chapter 11 Cost-effectiveness of improving implementation

a decision aid on ECV.17 During a two hour kick-off meeting, all participating care providers received the user instruction; they were asked to hand out the leaflet and decision aid to all women with a breech presentation from 32 weeks onwards and to schedule an extra counselling appointment. The care provider strategy contained a one day counselling course. Participants were informed on the contents of national guidelines and relevant literature emphasising the importance, benefits and risks of ECV.18,19 Furthermore, participants were trained in counselling by a counselling coach. After enrolment of the implementation strategies within the clusters, data on all consecutive women with a singleton breech presentation from 32 weeks onwards, eligible for ECV, over a period of nine months were collected; in total 1,613 women from 25 clusters. We will briefly describe the results of this cluster RCT to introduce the choices we made for the model parameters within this economic evaluation. The main outcome of this trial was the effect of the implementation strategy on the ECV rate compared to those clusters who were not exposed to the implementation strategy. The overall implementation rate was 72% (1,169 of 1,613 eligible clients) with a range between clusters of 8-95%. Neither the client (OR 0.64 (95% CI 0.30-1.39)) nor the care provider strategy (OR 1.98 (95% CI 0.72-4.47)) showed significant improvements. As it was unclear, whether contamination occurred, positively affecting the results in the control group (e.g. increased awareness through participation, Hawthorne effect), and because of the large intra-cluster variation of implementation of ECV, we performed a retrospective baseline measurement within the participating clusters. Therefore, we identified all consecutive breech deliveries from 36 0/7 weeks gestation and onwards from hospital birth registers between the 1st of January 2008 and the 31st of December 2009. We reviewed patient files to collect data on ECV attempts or reasons why no ECV was offered or performed. ECV registers are missing in most clusters and therefore we were not able to collect data on successful ECV attempts. Estimates for model parameters are derived from dataon 2,077 breech presentations. In this economic evaluation, we compared changes in the implementation rate of ECV between the baseline measurement and the outcomes of the strategies evaluated in the cluster randomized controlled trial, after adjusting for the missing 37% successful ECV attempts (mean ECV success rate in the cluster RCT).

212 Chapter 11 Cost-effectiveness of improving implementation

Model parameters Outcomes The following outcome parameters were used in our model: the percentage of women counselled for ECV, percentage opting for ECV, the overall ECV attempt rate, the mean ECV success rate, adherence rate of care providers to the implemented strategy and caesarean section rates among women with a cephalic presentation after successful ECV, breech presentation after failed ECV and breech presentation without an ECV attempt. Parameters estimates are provided in Table 1. Estimates and ranges reflecting their uncertainty are based on the results of the baseline measurement and cluster RCT. For those parameters influencing the ECV implementation rate (percentage counselled, opting and overall ECV attempt) we calculated the mean of clusters assigned to one strategy with a range representing the worst and best achieving cluster within that strategy. For all other parameters (ECV success rate and CS rates) we used mean and range values for all clusters, irrespective of strategy.

Table 1. Model parameters of outcomes based on results of study cohort

Baseline Cluster RCT Δ improvement measurement % % % min -‐ max Adherence to strategy Client -‐ 0.24 -‐ -‐ Care provider -‐ 0.21 -‐ -‐ Combined -‐ 0.23 -‐ -‐ ECV attempts Client 0.74 0.68 -‐0.05 -‐0.37 – 0.32 Care provider 0.68 0.66 -‐0.02 -‐0.70 – 0.44 Combined 0.78 0.86 0.07 0.03 – 0.44 CS without ECV 0.91 0.91 -‐ -‐ CS after failed ECV 0.78 0.78 -‐ -‐ CS after successful ECV 0.14 0.14 -‐ -‐ Overall CS rate Client 0.67 0.67 0.01 0.06 – -‐0.05 Care provider 0.68 0.68 0.00 0.11 – -‐0.07 Combined 0.66 0.65 -‐0.01 0.00 – -‐0.07 Ranges represent the worst and best achieving cluster on ECV within one strategy ^Including the successful ECV attempts (37% of all ECV attempts)

213

Chapter 11 Cost-effectiveness of improving implementation

Costs Estimates for cost parameters were calculated from a health care perspective. All costs were expressed in 2012 euros and inflated where appropriate according to the consumer pricing index.20 Costs were divided into implementation costs (costs related to deploying the strategy into clinical practice) and operational costs (costs related to counsel a client according to standard care or the implementation strategy, and the costs of an ECV attempt). The baseline estimates, range and references of cost parameters are presented in Table 2. We accounted for suboptimal adherence of care providers to the implementation strategy. Adherence rates were 24%, 21% and 23% for the client, care provider and combined strategy respectively. Thus we only included extra operational costs of the implementation strategy into account for those clients actually counselled according to the strategy (adherence rate). For the remaining women, we included the baseline costs of counseling.

Analyses

Cost-effectiveness The primary measure of effectiveness was the change in ECV attempts, and expressed as the increase/decrease in the probability that a woman presenting with a breech presentation will have an ECV attempt. The associated costs were estimated as the average costs per woman presenting with a breech presentation. Secondary outcomes were change in caesarean section rate (increased/decreased probability of a CS) and associated costs per client presenting with a breech presentation. While costs of the ECV attempt itself were not included in the cost per change in ECV rate (to avoid double counting), these costs were taken into account in the cost per change in CS rate. The incremental cost-effectiveness ratios (ICER) combines costs and clinical outcomes in a single figure, and is defined as ratio of the difference in effectiveness and the difference in costs, as compared to a reference strategy. ICERs for each of the three implementation strategies reflect the costs needed to obtain one extra ECV or to prevent one caesarean section, as compared to the care as usual strategy.

214 Chapter 11 Cost-effectiveness of improving implementation

Table 2. Estimated costs of implementation strategies and standard care (2012 €)

Units Baseline Estimate € Implementation costs Client Two -‐hour introduction session Midwife (n=89) hour 37` Gynaecologist (n=24) hour 76` Project leaders (n=1) hour 15* Overall costs Per session 10,264

Care provider Training course of 8 hours Midwife (n=41) hour 37` gynaecologist (n=5) hour 76` Project leaders (n=2) hour 15* Consultant (counselling) (n=1) hour 125* Actress (n=1) hour 85* Overall costs Per session 17,096

Combined Training course 9 of hours Midw ife (n=41) hour 37` gynaecologist (n=5) hour 76` Project leaders (n=2) hour 15* Consultant (counselling) (n=1) hour 125* Actress (n=1) hour 85* Overall costs Per session 19,023

Operational costs Client Decision aid Per client 0.84* Information leaflet Per client 0.73* Counselling appointment Per client 254^ Extra counselling appointment Per client 254^ Ov erall costs Per client 510 Care provider Counselling appointment Per client 254^ Overall costs Per client 254 Combined Decision aid Per client 0.84* Information leaflet Per client 0.73* Counselling appointment Per client 254^ Extra counselling appointment Per client 254^ Overall costs Per client 510

ECV attempt Per client 843^

References of costs: monitored * during Cluster RCT, ^Financial department of non-‐academic hospital, `reference 23.

215 Chapter 11 Cost-effectiveness of improving implementation

Sensitivity and scenario analysis Robustness of our findings was evaluated in several sensitivity analyses.The lower and upper limit of the confidence interval of effectiveness and ICER are based on the analyses where the mean probabilities are replaced by the probabilities of the worst and best achieving clusters in the RCT. In addition to the base case model (model 0), using the most probable estimates for all model parameters, we evaluated the consequence of different assumptions in effectiveness and cost estimates in the following seven analyses. In model 1 and 2, we have assessed the impact of respectively an overestimation or underestimation of the overall costs by 10%, which corresponds to the range in costs among different sources. In model 3, we assumed that the observed effects of the implementation strategies sustain for two years (instead of nine months) after introduction of the strategies. As a consequence the implementation costs per client were decreased in this scenario. In the fourth model, we assumed that a second introduction session of the implementation strategy would double the effect on adherence tothe implementation strategy as more care providers are reached. In the fifth analysis, both scenarios were combined; a sustained effect and a repeated implementation. All statistical analyses to calculate model parameters from primary study data were performed using SPSS version 18.0 (Chicago, IL, USA), and MICROSOFT EXCEL 2003 was used to develop the model and perform sensitivity analyses.

RESULTS

Cost-effectiveness of change in ECV attempts Results of our economic evaluation are presented in Table 3. After using the client strategy, on average the percentage ECV attempts decreased by 5%(range - 37% to +32% among different clusters). On average, costs per client in the client strategy increased by €45. After the care provider strategy, the percentage of ECV attempts decreased by 2% (range - 70% to +44%). The care provider strategy resulted in a cost decrease of €5. Combining both strategies increased the ECV rate by 7% (range +3 to +44%). The additional costs per client were €83. The client as well as the care provider strategy led to less ECV attempts against higher costs (inferior strategies, with non-informative ICERs). Only the combined strategy led to increase in ECV attempts against higher costs

216 Chapter 11 Cost-effectiveness of improving implementation

Table 3. results for the base case analysis for improvement of ECV and improvement of CS (2012 €)

Baseline measurement Cluster RCT Δ improvement Δ costs ICER % €/patient % €/patient % € mean min -‐ max ECV attempts Client 0.74 203 0.68 249 -‐0.05 45 -‐842 -‐ Care provider 0.68 196 0.66 191 -‐0.02 -‐5~ 310* -‐ Combined 0.78 224 0.86 307 0.07 83 1,119 190 – 2,659

Caesarean sections Client 0.67 743 0.67 826 0.01 83 9,936^ -‐ Care provider 0.68 676 0.68 749 0.00 73 30,348^ -‐ Combined 0.66 810 0.65 1030 -‐0.01 221 19,119 3,248 -‐ 45,418 The ICER of ECV are the extra costs per ECV attempt. The ICER can be positive or negative valued based on the effect s. on ECV attempts or CS rate If positive, the scenario leads to more ECV attempts against more costs. If negative, the scenario leads to less ECV attempts against more costs. *The ICER for this scenario means that there were less ECV attempts against less costs ^The ICER for this scenario means that there were more CS against more costs. ~Both client and care provider strategies led to less ECV attempts because of less counselled s patients. Due to the higher cost of counselling (e.g. two counselling appointments instead of one) in the client strategy, only the costs in the care ategy provider str decreased due to less counseled clients.

(with ICERs indicating costs associated per unit in health gained). In the base case, the costs to achieve one extra ECV attempt were €1,119 (range among clusters €190 to €2,659).

Sensitivity analyses The differences in outcomes, costs and ICERs for the different sensitivity analyses are presented in Table 4 and Figure 1 and 2. Varying implementation and operational costs (sensitivity analyses, model 1 and 2) led to a range from €1,007 to €1,213 around the mean of €1,119 per extra ECV attempt. Model 3 presents the scenario that the effect of our implementation strategies will reach beyond the nine month follow up of the cluster RCT. If the effect would maintain for two years, the ICER of the combined strategy would be lowered to €865 (range €147 – €2,055). Model 4 evaluate scenarios were ECV implementation rates would be further increased; repeated implementation courses to improve adherence of care providers to the implementation strategies would still not lead to an overall increase of ECV attempts in the client or care provider strategy. The implementation rate within the combined strategy would be improved by 15%, but against costs of €495 per patient, resulting in an ICER of €3,320 (range €564 – €7,888). If both sustained effect and repeated courses were taken into account (model 5), the ICER would be €1,602 (range €272 – €3,806).

217 Chapter 11 Cost-effectiveness of improving implementation 48 418 0 701 960 876 522 , , 8 0 max

45 , 65 , 3 49 , 4 41 , -‐

, 248 -‐ 652 -‐ 768 -‐ 572 -‐ 923 -‐ 969 -‐ min , , , , 3 4, 2, 3 2 2

€

119 382 291 030 207 479 ,

7 ICER ICER 9,936^ 8,942^ 9,033^ 19 , 2 16 , 21, 17, 17 , CS 30,348^ 27,313^ 52,852^ 10,929^ 33,383^ 23,681^ 126,952^ 219,257^ 101,537^

0.14 0.14 0.07 0.14 0.07 0.07 0.05 0.07 0.10 0.05 0.05 0.07 0.07 0.07 0.05 0.07 0.14 0.10 -‐ -‐ max -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐

-‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ Δ improvement Δ min 0.01 0.01 0.01 0.11 0.11 0.05 0.05 0.11 0.11 0.06 0.11 0.03 0.06 0.06 0.06 0.06 0.11 0.11 0.06 0.06 0.05 0.05 0.03 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -‐ -‐

01 01 01 02 01

00 00 01 00 01 01 00 01 00 00 . . . . . % ...... 0 0 0 0.02 0 0 0 0.00 0 0 0 0.00 0 0 0 0 0 0 -‐ -‐ -‐ -‐ -‐ -‐

659 393 888 806 055 , , max 2 2, 2,924 7 3, 2,

-‐ – – – – – –

d based on the effect on ECV attempts or CS rates. If positive attempts ECV on effect the on based d 72 min 190 171 209 56 4 2 14 7

€ 02

942 335

58 26 , 119 007 213 320 7 9 842 702 , 3 865

-‐ -‐ -‐ -‐ 9,644 6,793 310* 279* 341* 1 1, 1, 3, 1,6 2 14. ICER ICER -‐ -‐ * 1,342 -‐ -‐

0.88 0.44 0.44 0.32 0.44 0.64 0.32 0.32 0.44 0.32 0.88 0.64 0.44 0.44 0.44 0.88 0.88 0.44

max

the scenario leads to less ECV attempts against more costs. attempts ECV less to leads scenario the

– – – – – – – – – – – –

-‐ – – – – – – , min 0.35 0.35 0.70 0.70 0.70 0.70 0.37 0.70 0.18 0.37 0.37 0.37 0.37 0.70 0.70 0.37 0.37 0.35 0.35 0.18 0.18 0.03 0.03 0.03 0.03 0.03 0.03 0.06 0.06 0.03 0.03 -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ Δ improvement ECV Δ

02 02 05 02 05 05 02 05 01 03

07 07 07 07 15 ...... % . . . . . 0.01 0 0 0 0 0.03 0 0 0 0 0 0 0 0 0 0.15 0 0 -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ -‐ ICER can be positive or negative value

Combined Combined provider Care provider Care Client Combined Client Care provider Care Combined provider Care Client Client

Combined

Client Care provider Care Combined Care provider Care Client Strategy

plementation and plementation

Base case Lower 10% and implementation costs operational Description

10% Higher Higher 10% im costs operational Repeated implementation, effect sustained

Sustained effect of implementation to two up strategies years Repeated to implementation adherence double

Sensitivity and scenario analyses for the combined strategy (2012 €) combined the for analyses scenario and . Sensitivity 4

Univariate sensitivity analyses sensitivity Univariate 0. 2. Model 1.

Scenario analyses Scenario 3.

4. *The ICER for this scenario means that there were less ECV attempts against less costs more CS against costs. were there that means scenario this for ICER ^The the scenario leads to more ECV attempts against more costs. If negative against attempts ECV more to leads scenario the The ICER of ECV are the extra costs per ECV attempt. The costs extra the are ECV of ICER The Table Table

218 Chapter 11 Cost-effectiveness of improving implementation

The results are also summarized in a cost-effectiveness plane (shown in Figure 1). Points located in the upper right quadrant represent those strategies are more effective (i.e. result in more ECV) but at higher costs. Points inthe upper left quadrant represent inferior strategies, as they are less effective (i.e. decreased ECV rate) but at higher costs. This figure clearly visualises that only the combined strategy could be a cost-effective alternative to improve ECV attempt rates, if society is willing to pay the additional costs to achieve this improvement in ECV attempt rate. If the effect of that strategy last up to two years, the costs are €64 per client to achieve a mean increase of 7% in ECV attempt rates.

Figure 1. Cost -‐effectiveness plane. Each point -‐ in the cost effectiveness plane represent the additional costs per patient and gain in ECV attempts for the different strategies and different sensitivity and scenario analyses ( the points are numbered according to corresponding the analysis).

Additional costs per patient €

Additional increase in ECV attempts

Cost-effectiveness of change in planned caesarean sections As a result of these limited effects of implementation strategies on the ECV attempt rate, caesarean section rates were more or less unaffected. In the client strategy the caesarean section rate increased with 1% (range -5% to 6%), in the care provider strategy the caesarean section rate remained unchanged (0%, range -7 to 11%) and in the combined strategy the caesarean

219 Chapter 11 Cost-effectiveness of improving implementation

section rate was reduced with 1% (range -7 to 0%). The additional costs per client presenting with a breech presentation were €83, €73 and €221 for respectively the client, care provider and combined strategy in the caesarean section model (costs of ECV were taken into account). Due to theminor change in mean caesarean section rates, the incremental cost-effectiveness ratios were very high (Table 3). the ICER for the combined strategy was €19,119 (range €3,248-€45,418) per prevented caesarean section. The difference in direct medical costs between a caesarean section compared to a vaginal breech delivery is approximately €1500,21 and the break-even point for the combined strategy to become cost effective is a 95% increase in ECV rate. The results with planned caesarean sections as an outcome are also visualised in a cost effectiveness plane (Figure 2).

Figure 2. Cost -‐effectiveness plane. Each point -‐ in the cost effectiveness plane represent the additional costs per patient and change in planned caesarean delivery rate for the different strategies and different sensitivity and scenario analyses (the points are numbered according to the corresponding analysis).

Additional costs per patient €

Additional decrease in planned CS

The points in the upper right quadrant, represent those strategies which led to more planned caesarean deliveries are performed against additional costs. The points in the upper left quadrant represent strategies which decreased the caesarean section rate against more costs. If society is willing to pay the considerable costs to prevent one caesarean section, the combined strategy could become cost-effective.

220 Chapter 11 Cost-effectiveness of improving implementation

DISCUSSION This study evaluated the cost-effectiveness of different implementation strategies to improve ECV rates within clusters. The economic analysis was performed from a health care perspective, based on results of the cluster randomised controlled trial on the implementation of ECV, and the baseline measurement performed within the participating clusters. Our analyses showed that the combined strategy (client and care provider strategies together) has the best potential to increase ECV implementation rates. Only in those regions with the highest increase in ECV rates, the incremental cost effectiveness ratios (costs per additional ECV) could be considered acceptable. However, in terms of improving planned caesarean section rates, the costs per prevented planned caesarean section always largely exceed the costs savings resulting from the intervention prevented. The wide range in effect of the client strategy and care provider strategy to improve ECV implementation rates is mainly due to those clusters which contributed the least amount of patients and could be defined as outliers. Costs are estimated on reliable sources, but the costs per ECV attempt and counseling are relatively high compared to the costs per strategy and have a great influence on the outcome of the cost effectiveness analyses to prevent a caesarean section. If these costs could be lowered (for instance by more efficient organization or group-counseling), the strategies might become more cost effective.

Strength and limitations A strength of our study is the accuracy of the model parameters which were based on the prospective data gathering within the cluster randomised controlled trial; the effects of the implementation strategies on ECV attempts and mode of delivery. The baseline measurement, made it possible to deal with the wide range of ECV implementation rates among clusters and evaluate the effect of the implementation strategies within clusters and not only between the clusters assigned to the different strategies. The scenario analyses shed light on the potential opportunities of the strategy, beyond the study protocol of the cluster randomised controlled trial. Sustained effectiveness of the implementation strategies would decrease the costs per gained ECV attempt, but the willingness to pay needed for the combination strategy to be considered cost-effective remained high. The break-even point of a 95% improvement can only be reached in those clusters with very poor baseline implementation rates. The break-even point could be lower

221 Chapter 11 Cost-effectiveness of improving implementation

in other settings (countries) with higher CS rates for breech presentation at birth, as the effect of ECV at the CS rate would be more distinct. (The break-even point might also be lower if downstream costs are taken into account; prevention of CS leads to less repeated CS and less complications in subsequent pregnancies. Repeated implementation efforts to increase the uptake of ECV only led to an increase of the incremental cost effectiveness ratios. Limitations of this study are similar to those of any cost-effectiveness analysis. In our analysis we have assumed that ECV success rates are equal among all clusters (we took the average rates). We made this assumption to reflect the effect of the implementation strategy on planned caesarean section rates in an unbiased way. We know from other studies that assuming a mean ECV success rates of 37% is a reasonable assumption for the Dutch setting.22 But some regions with high ECV implementation rates, also achieved ECV success rates up to 60%. If implementation strategies were developed, not only to increase the quantity (amount of attempts) but also the quality (success rates) of the ECV attempt, these might be more effective in prevention of CS. Our model could be further refined by taking these effects into account, if more data become available in future studies on the topic of improvement of ECV success rates. A potential bias in our study is the time window between the baseline measurement within the clusters (2008-2009) and the cluster randomized controlled trial (2011-2012). During the trial, we did not have any clues to suspect another change in ECV policy among clusters, which could have caused the measured effect of the combined strategy. Potential bias could be caused by the so called Hawthorne effect: care providers might have put more effort in the counselling for ECV as they knew they were being monitored during the study period. If this has played any role in the outcome, it probably would have influenced the results of certain regions more (best performing regions) compared to others (decrease in ECV implementation rates). The mean effect of the strategies and results of best performing clusters might be slightly overrated due to this effect.

222 Chapter 11 Cost-effectiveness of improving implementation

Relation to other studies To our knowledge, this is the first study to evaluate the cost-effectiveness of different strategies to improve the implementation of ECV. Our results show the difficulties to change common practice and the effort and costs needed to achieve minimal change. This is contrary to societies expectations that care providers work according to state of the art guidelines, and only differ if it can be argued.

CONCLUSION

An implementation strategy focussing on both patients and care providers can improve ECV implementation rates, but against considerable costs. Whether this strategy is cost-effective depends on the societal willingness to pay for an additional attempt, and the sustainability of the effect of the implementation strategy. None of the strategies was cost-effective to improve planned caesarean section rates.

223 Chapter 11 Cost-effectiveness of improving implementation

REFERENCES 1 Hannah ME, Hannah WJ, Hewson S a, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet [Internet]. 2000 Oct;356 : 1375–1383. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0140673600028403

3 Bragg F, Cromwell DA, Edozien LC, Gurol-urganci I, Mahmood TA, Templeton A, et al. Variation in rates of caesarean section among English NHS trusts after accounting for maternal and clinical risk : cross sectional study. BMJ. 2010;341.

4 US National Institutes of Health C. Caesarean section-the first cut isn’t the deepest. Lancet. 2010; 375: 956.

5 Nielsen TF, Hökegård KH, Ericson a. Cesarean section and perinatal mortality in Sweden in 1981. Acta Obstet Gynecol Scand [Internet]. 1986 Jan; 65: 865–867. Available from: http://www.ncbi.nlm.nih.gov/pubmed/3825528

6 Van Ham M a, van Dongen PW, Mulder J. Maternal consequences of caesarean section. A retrospective study of intra-operative and postoperative maternal complications of caesarean section during a 10-year period. Eur J Obstet Gynecol Reprod Biol [Internet]. 1997 Jul; 74: 1–6. Available from: http://www.ncbi.nlm.nih. gov/pubmed/9243191

7 Hofmeyr, GJ; Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012; 10.

8 Grootscholten, K; Kok, M; Oei, SG; Mol, BW; van der Post J. External cephalic version-related risks: a meta-analysis. Obs Gynecol [Internet]. 2008 [cited 2014 Mar 20]; 112: 1143–1151. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/ pubmed/?term=xternal+cephalic+version-related+risks:+a+meta-analysis.

9 Tan JM, Macario A, Carvalho B, Druzin ML, El-Sayed YY. Cost-effectiveness of external cephalic version for term breech presentation. BMC Pregnancy Childbirth [Internet]. 2010 Jan; 10: 3. Available from: http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=2826287&tool=pmcentrez&rendertype=abstract

224 Chapter 11 Cost-effectiveness of improving implementation

10 Caukwell S, Joels L a, Kyle PM, Mills MS. Women’s attitudes towards management of breech presentation at term. J Obstet Gynaecol [Internet]. 2002 Sep [cited 2014 Mar 20]; 22: 486–488. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/12521413

11 Leung, TY; Lau, TK; Lo, KW; Rogers M. A survey of pregnant women’s attitude towards breech delivery and external cephalic version. Aust N Z J Obs Gynaecol. 2000; 40: 253–259.

12 Bewley, S; Robson, SC; Smith, M; Glover, A; Spencer J. The introduction of external cephalic version at term into routine clinical practice. Eur J Obs Gynecol Reprod Biol [Internet]. 1993 [cited 2014 Mar 20]; 52: 89–93. Available from: https:// c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/8157147

13 Feitsma, HA; Middeldorp JM, Oepkes D. De uitwendige versie bij de a terme stuit; een inventariserend onderzoek (ECV for breech presentation at term; an inventory study). NTOG. 2007; 120: 4–6.

14 Vlemmix F, Rosman AN, Fleuren MAH, Rijnders MEB, Beuckens A, Haak MC, et al. Implementation of the external cephalic version in breech delivery. Dutch national implementation study of external cephalic version. BMC Pregnancy Childbirth. 2010; 10: 20.

15 Rosman AN, Vlemmix F, Beuckens A, Rijnders ME, Opmeer BC, Mol BWJ, et al. Facilitators and barriers to external cephalic version for breech presentation at term among health care providers in the Netherlands: a quantitative analysis. Midwifery [Internet]. Elsevier; 2014 Mar [cited 2014 May 6]; 30: e145–50. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24530121

16 Rosman a N, Vlemmix F, Fleuren M a H, Rijnders ME, Beuckens a, Opmeer BC, et al. Patients’ and professionals’ barriers and facilitators to external cephalic version for breech presentation at term, a qualitative analysis in the Netherlands. Midwifery [Internet]. Elsevier; 2014 Mar [cited 2014 May 6]; 30: 324–330. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23664312

17 Nassar N, Roberts CL, Raynes-Greenow CH, Barratt a, Peat B. Evaluation of a decision aid for women with breech presentation at term: a randomised controlled trial [ISRCTN14570598]. BJOG [Internet]. 2007 Mar [cited 2014 Mar 20]; 114: 325– 333. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=24 08658&tool=pmcentrez&rendertype=abstract

225 Chapter 11 Cost-effectiveness of improving implementation

18 Otterlo Werkgroep dr. AJ van L en dr. MG van P. NVOG richtlijn - Stuitligging- versie 2.0 (Guideline on breech presentation - version 2.0 of the Dutch Society of Obstetrics and Gynecology) [Internet]. 2008. Available from: file:///C:/Users/PS13C/ Downloads/Stuitligging (1).pdf

19 Verburgt, T; Offerhaus P. Uitwendige versie (External Cephalic Version). NTOG. 2007; 120: 4–7.

20 Statistics Netherlands [Internet]. Statline Consum. Pricing Index. [cited 2014 May 28]. Available from: www.statline.cbs.nl

21 Vijgen SMC, van der Ham DP, Bijlenga D, van Beek JJ, Bloemenkamp KWM, Kwee A, et al. Economic analysis comparing induction of labor and expectant management in women with preterm prelabor rupture of membranes between 34 and 37 weeks (PPROMEXIL trial). Acta Obstet Gynecol Scand [Internet]. 2014 Apr [cited 2014 May 26]; 93: 374–381. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24392746

22 Kok, M; Bais, JM; van Lith, JM; Papatsonis, DM; Kleiverda, G; Hanny, D; Doornbos, JP; Mol, BW;van der Post J. Nifedipine as a uterine relaxant for external cephalic version: a randomized controlled trial. Obs Gynecol. 2008; 112: 271–276. 23 van Baaren GJ, Jozwiak M, Opmeer BC, Oude Rengerink K, Benthem M, Dijksterhuis MG, van Huizen ME, van der Salm PC, Schuitemaker NW, Papatsonis DN, Perquin DA, Porath M, van der Post JA, Rijnders RJ, Scheepers HC, Spaanderman M, van Pampus MG, de Leeuw JW, Mol BW, Bloemenkamp KW. Cost-effectiveness of induction of labour at term with a Foley catheter compared to vaginal prostaglandin E₂ gel (PROBAAT trial). BJOG. 2013 Jul;120(8):987-95.

226

12 Mode of delivery and neonatal outcome after external cephalic version: a prospective cohort study

Ageeth Rosman

Floortje Vlemmix

Sabine Ensing

Brent Opmeer

Susan te Hoven

Joost Velzel

Marcella de Hundt

Suzanne van de Berg

Hanneke Rota

Joris van der Post

Ben Willem Mol

Marjolein Kok

Submitted

229 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version

ABSTRACT

ObjectiveTo assess the mode of delivery and adverse neonatal outcome in women with a breech presentation with or without an external cephalic version (ECV) attempt and to compare the mode of delivery among women with successful ECV to women with a spontaneous cephalic presentation.

Design Prospective matched cohort study.

Setting 25 clusters (hospitals and its referring midwifery practices) in the Netherlands. Data of the Dutch perinatal registry for the matched cohort.

Population Singleton pregnancies from January 2011 to August 2012 with a foetus in breech presentation and a delivery from 36 weeks gestation onwards. Spontaneous cephalic presentations (selected from national registry 2009 and 2010) were matched in a 2:1 ratio to cephalic presentations after a successful version attempt. Matching criteria were maternal age, parity, gestational age at delivery and foetal gender.

Main outcome measures Mode of delivery and neonatal outcome.

Results Of 1,613 women eligible for ECV, 1,169 (72.5%) received an ECV attempt, and their caesarean delivery rate was significantly lower compared to women who did not receive an ECV attempt (57% vs 87%; RR 0.66 (0.62 – 0.70)). Women with a cephalic presentation after ECV had, compared to women with a spontaneous cephalic presentation, an increased risk of caesarean delivery (RR 1.7 (95%CI 1.2 - 2.5)), but a decreased risk for instrumental vaginal delivery (RR 0.52 (95% CI 0.29 – 0.94)).

Conclusion Women who had a successful ECV are at increased risk for a caesarean delivery. Despite this fact, ECV is an effective treatment to reduce the caesarean rate.

230 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version

INTRODUCTION Breech presentation occurs in 3-4% of all term pregnancies.1 The Term Breech Trial reported that a planned caesarean delivery reduced the risk of foetal mortality and morbidity as compared to a planned vaginal breech delivery.2 However, a planned caesarean delivery is associated with an increased risk of maternal morbidity and maternal and foetal complications in subsequent pregnancies.3,4 External cephalic version (ECV) is proven to be effective in prevention of breech presentation, as seven randomised controlled trials indicated a reduction in caesarean delivery rate from 78% to 37%.5 However, there is increasing evidence that women with a foetus in cephalic presentation after ECV have an increased risk of an emergency caesarean delivery, which can potentially lower the effectiveness of ECV. A recent review by de Hundt et al. on mode of delivery after successful ECV showed a twofold risk of overall caesarean delivery compared to spontaneous cephalic presentations at birth (20.7 vs. 10.9%, pooled OR 2.2 (95%CI 1.7 to 2.8)).6 However, this evidence came from several rather smaller cohort studies reporting on a total of 1,547 women who delivered after a successful ECV. The review of de Hundt did not focus on neonatal outcome. The objective of this study was therefore to evaluate the mode of delivery and adverse neonatal outcomes in women with and without an ECV attempt in a large Dutch cohort study and compare the mode of delivery and adverse neonatal outcome to spontaneous cephalic presentations at birth in the Netherlands.

METHODS

Setting For this study we used data from women included in a cluster randomized controlled trial (RCT) in the Netherlands in which the effect of two ECV implementation strategies to improve the implementation of ECV were evaluated.7 This cluster RCT was performed in 25 hospitals and the referring midwife practices liaised with these hospitals.

Participants The trial was performed between January 2011 and August 2012. We collected data from consecutive women with a singleton foetus in breech presentation from 34 weeks gestation and birth from 36 weeks gestation onwards.

231 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version

Exclusion criteria were women with a contraindication for ECV according to national guidelines,8,9 spontaneous version before counselling for ECV, indication for primary caesarean section other than breech presentation, first diagnosis of breech presentation at onset of birth, major congenital malformations related to poor neonatal outcome (e.g. hydrocephalus, spina bifida, cardiac malformations, trisomy 13, 18, 21), and delivery before 36 weeks gestation (as national guidelines recommend ECV from 36 weeks onwards). We documented baseline characteristics including s maternal age, parity and gestational age at delivery.

Matched cohort From the cohort mentioned above, we selected all women with a cephalic presentation at birth after successful ECV. We matched each woman with a cephalic presentation after ECV with two women with spontaneous presentations from the Dutch perinatal registry from January 2009 to December 2010. Matching was based on antenatal detectable risk factors associated with an increased risk of emergency caesarean delivery.10 The risk factors available from both databases were maternal age, parity, gestational age at delivery and foetal gender.

Outcome measures The primary outcome was mode of delivery. We calculated the difference in mode of delivery (overall, planned and emergency caesarean delivery rates) among women with and without an ECV attempt. Secondary outcomes were complications of ECV and neonatal outcome after birth (Apgar score, arterial pH value, admission to neonatal care, perinatal death up to 28 days). We compared mode of delivery of women after a successful ECV attempt to the mode of delivery (instrumental vaginal delivery, and overall, planned and emergency caesarean delivery rates) of women with a spontaneous cephalic presentation at birth.

Statistical analysis To analyse the data for ECV and mode of delivery, we calculated crude and adjusted relative risks. For the adjusted relative risks, we took the significant difference in mode of delivery between nulliparous and multiparous women (p< 0.001) into account. We also used logistic multilevel analyses, adjusting for possible clustering of observations. The included levels were patient and hospital. Analysis was carried out in SPSS 20.0. P-values were calculated with

232 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version double sided independent – samples T-Test for continues data and the χ2 test for binominal data. A p-value of less than 0.05 for a two-tailed test was considered statistically significant.

RESULTS We identified 1,793 women with a breech presentation from 34 weeks onwards (figure 1). We excluded 180 women (10.0%): 72 (4.0%) for reasons of an absolute contraindication for ECV, 46 (2.6%) had an indication for a planned CS other than breech presentation and 62 (3.5%) breech presentations were not diagnosed until birth (figure 1). Of the 1,613 eligible women 1,169 (72.5%) underwent an ECV attempt with a success rate of 38.7%. ECV was more successful in multiparous than in nulliparous (56% versus 24%, p<0.001). Baseline characteristics are shown in Table 1. There was no difference in age among women who underwent ECV (30.8 vs. 31.2 years (p=0.13)). Mean gestational age at delivery was 39 3/7 weeks (36 0/7 – 42 2/7) if ECV was applied and 38 6/7 days (36 0/7 – 42 1/7) if ECV was not applied (Table 1). Of the nulliparous women 73% underwent ECV compared to 74% of multiparous women (p=0.74).

Figure 1. Flow chart eligible patients

233 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version

Table 1 baseline characteristics

Breech presentation p-‐ value

No ECV attempt ECV attempt

N = 444 N = 1169

Age (years) (mean, SD) 31.2 (4.9) 30.8 (4.7) 0.13

Parity (%) 0.74

Nulliparous 224 (59.9) 616 (58.9)

Multiparous 150 (40.1) 430 (41.1)

Gestational age at delivery 38 6/7 39 3/7 <0.001

(weeks, range) (36 0/7 – 42 2/7) ( 36 – 0/7 42 2/7)

Missing values on 61 age n= , parity n = 193, gestational age n= 157.

Mode of delivery Women who had ECV had significantly more cephalic presentations at birth than women who did not have ECV (34.7% vs 7.3%, aRR 5.0 (95%CI 3.2-8.0)), and both the overall CS rate (57.3% vs 84.3% (aRR 0.26 (95%CI 0.17-0.40))) and the planned CS rate were reduced (43.1% to 75.3% (aRR 0.26 (95%CI 0.18-0.38)). The emergency CS rate among women with a trial of labour was also reduced (24.9% to 36.3%, aRR 0.69 (95%CI 0.36-1.3)) (Table 2).

234 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version

Table 2. Relation of ECV attempt to mode of delivery and neonatal outcome.

No ECV attempt ECV attempt Unadjusted Adjusted Relative Relative risk (95% risk (95% CI)” N = 413* N = 1058* CI)

Cephalic presentation at birth 308 (74.6) 367 (34.7) 4.8 (3.6 – 6.8) 5.0 (3.2-‐8.0)

Mode of delivery

Overall CS (%) 348 (84.3) 606 (57.3) 0.68 (0.64 – 0.73) 0.26 (0.17-‐0.40)

Planned CS 311 (75.3) 456 (43.1) 0.57 (0.52 – 0.63) 0.26 (0.18-‐0.38)

Emergency CS 37 (9.0) 150 (14.2) 0.69 (0.51 – 0.92)$ 0.69 (0.36-‐1.3)$

Neonatal outcome

Apgar score ˂7 at 1 minute~ 19 (4.6) 73 (6.9) 1.5 (0.92-‐2.5) 1.3 (0.7-‐2.3)

Apgar score ˂7 at 5 minutes~ 4 (1.0) 16 (1.5) 1.5 (0.53-‐4.6) 1.3 (0.47-‐3.9)

Umbilical artery pH ˂7.05~ 1 (0.2) 3 (0.3) 1.2 (0.12-‐11.2) 0.66 (0.06-‐8.0)

Umbilical vein pH ˂7.20 ~ 1 (0.2) 6 (0.6) 2.3 (0.28-‐19.4) 0.36 (0.02-‐6.4)

Neonatal admission ~^ 3 (0.5) 17 (1.6) 2.2 (0.654-‐7.5) 1.6 (0.79-‐3.3)

Perinatal death’ 0 3 (0.3) -‐ -‐

Foetal death 0 2 (0.2) -‐ -‐

Neonatal death 0 1 (0.1) -‐ -‐

*Missing values presentation at birth and mode of delivery N = 142 of 1,613 cases (8.8%), excluded from analyses.

`` adjusted for parity and clustering of observations within included clusters

$Relative risks calculated over all women with a trial of labour

~Missing values n = 195 of 1,501 cases (13.0%)

^admission to medium care or neonatal intensive care unit.

‘composite of foetal death, perinatal death and neonatal death

235 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version

Complications of ECV There occurred three complications immediately following ECV. Two women had a foetal bradycardia at the start of foetal heart registration after ECV. Both bradycardias recovered rapidly in such a way that an intervention to pursue birth was not necessary. In one woman with a gestational age of 36 2/7 weeks the membranes ruptured during the ECV attempt, after which a primary CS was performed resulting in the birth of a healthy baby. There were two stillbirths (1.8‰) reported in the ECV group versus none in the no-ECV group (Table 2). The stillbirths were diagnosed within 24 hours after ECV and within three days after ECV, respectively. In both cases, autopsy and histological study of the placenta did not reveal a cause of death.

Neonatal outcome at birth Apgar scores at 1 minute and 5 minutes were available in 88% of the cases (Table 2). There was no significant difference in Apgar score below 7 after one and five minutes in the group with an ECV attempt compared to those without an ECV attempt. There were four cases of an umbilical artery pH below 7.05 and 3 cases with an umbilical vein pH between 7.05 and 7.20, without differences between women with and without an ECV attempt. Neonatal admission to a medium or intensive care unit occurred in 20 cases, without a significant difference between the groups with and without ECV. One neonate out of the ECV group was postpartum diagnosed with an uncommon congenital metabolic disease, causing liver failure and died two months after birth.

Matched cohort Of the 460 women who had a successful ECV attempt, presentation at birth and mode of delivery was known in 409 women, of which 346 (85%) remained in cephalic presentation until birth. Complete data on maternal age, parity, gestation age at delivery, and foetal gender (matching variables) were available in 301 of 346 cases, and we could match 301 of them. Mode of delivery and neonatal outcomes are presented in Table 3. Overall CS rate was almost doubled (14.6 vs. 8.5% (RR 1.7 (95%CI 1.2 - 2.5)) and the emergency CS rate was higher among women with a cephalic presentation after ECV, though not significant (12.0 vs. 8.5% (RR 1.4 (95% CI 0.94 – 2.1)). There was a twofold decrease in instrumental vaginal delivery for women with a cephalic presentation after ECV compared to spontaneous cephalic presentation (4.7 vs. 8.8% (RR 0.52 (95% CI 0.29 – 0.94)).

236 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version

The incidence of five minute Apgar below 7 was not statistically significant different and neonatal admission only occurred in the control group (n=3) as did perinatal death (n=1).

Table 3. Mode of delivery and neonatal outcome matched cohort

Cases Controls Relative Risks

N = 301(%) N = 602(%) (95% CI)

Mode of delivery

Spontaneous 243 (80.7) 498 (82.7) 0.98 (0.91 – 1.0)

Instrumental vaginal delivery 14 (4.7) 53 (8.8) 0.52 (0.29 – 0.94)

Overall CS (%) 44 (14.6) 51 (8.5) 1.7 -‐ (1.2 2.5)

Planned CS 8 (2.6) 0

Emergency CS 36 (12.0) 51(8.5) 1.4 – (0.94 2.1)

Neonatal outcome

Apgar score ˂7 at 5 minutes 2 (0.7) 7 (1.2) 0.57 (0.12 – 2.7)

Neonatal admission 0 3

Perinatal death’ 0 1

´ composite of foetal death, perinatal death and neonatal death

237 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version

DISCUSSION

To investigate the effect of ECV on mode of delivery, we studied the ECV rate and mode of delivery in 1,613 women eligible for ECV. Our study has shown that an ECV attempt, with a mean success rate of 38.7%, significantly reduces the number of overall caesarean deliveries and planned caesarean deliveries. However, women with a foetus in cephalic presentation after ECV are at increased risk of an emergency caesarean delivery compared to women with a spontaneous cephalic presentation. To our knowledge, this is the largest cohort on ECV and mode of delivery thus far. Our findings are in line with the Cochrane review of seven randomised controlled trials, representing in total 1,245 women with a breech presentation at or near term regarding the benefits of ECV without major neonatal risks.5 Furthermore, with a mean ECV success rate of 39% the results of this study are also in line with a cost effectiveness study on ECV, stating that from society’s perspective, ECV trial is cost-effective when compared to a scheduled caesarean for breech presentation provided the probability of successful ECV is more than 32%.11 A limitation of this study is that we did not collect data on adverse maternal outcome. Complication rates of CS reported in literature vary from 11.6 to 29% including mild complications as cystitis to severe complications as reoperation12–15 and even greater topics of concern are the long term complications of a caesarean delivery, namely the maternal and foetal morbidity and mortality during subsequent pregnancies. Risk of maternal and foetal death during vaginal birth after caesarean delivery (VBAC) is 0.04‰ and 1.3‰ respectively, and during elective repeat caesarean delivery (ERCD) 0.13‰ and 0.5‰ respectively.16 Thus, data on maternal outcome would strengthen the current results as prevention of caesarean delivery improves maternal outcome and outcome of future offspring which is in line with the recommendation to perform ECV. A limitation of our matched cohort was that we could only match for four antenatal variables related to increased risk of emergency caesarean section (maternal age, parity, gestation age at delivery, and foetal gender). A clinical prediction model on operative vaginal delivery or caesarean delivery by Schuit et al. demonstrated that age, parity, caesarean delivery in history, diabetes, gestational age at delivery, foetal gender, estimated foetal weight and induction of labour are good predictors for the risk on operative vaginal delivery or caesarean delivery.10 Better adjustment for other risk factors related to caesarean section rates, might

238 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version result in different relative risk ratios for cephalic presentation after ECV. The results of our matched cohort are for so far the overall caesarean rates concerned in line with the most recent systematic review on the subject6 but not in line with the instrumental deliveries. Our match showed a smaller risk on an instrumental delivery as compared to spontaneous cephalic presentations. In addition to the data presented in the review, we were able to analyse the risk at an emergency caesarean delivery. We did not find a significant difference. The planned caesarean section rate of 62% is in line with a large Dutch cohort study on mode of delivery for term breech presentation after publication of the results of the Term Breech Trial, but is much lower than elective caesarean rates in other countries.17,18 An interesting finding is the fact that women who did not have an ECV opted for an elective CS in 80% of the cases. Whether this had to do with the perception that a CS would be safer than an ECV or the breech presentation was an opportunity to opt for an elective caesarean delivery remains unclear. The foetal death rate of 1.8‰ within the ECV group is higher than the 0.9‰ (12 / 12,955) reported in the systematic review on ECV related risks by Grootscholten et al.19 However, it corresponds to our national foetal death rate of singleton pregnancies after 37 weeks gestation (1.6‰ in 2012), indicating no increased risk from ECV.20

CONCLUSION In conclusion, after successful ECV, women are at decreased risk ofan instrumental vaginal delivery and at increased risk of caesarean delivery compared to women with a spontaneous cephalic presentation. Nevertheless, our data confirm that ECV is a key tool to reduce the overall, planned and emergency caesarean delivery rate due to breech presentation at term without increasing neonatal risks.

239 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version

REFERENCES

1 Hickok, DE; Gordon, DC; Milberg, JA; Williams, MA; Daling J. The frequency of breech presentation by gestational age at birth: a large population-based study. Am J Obs Gynecol [Internet]. 1992 [cited 2014 Mar 3]; 166: 851–852. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/1550152 2 Hannah ME, Hannah WJ, Hewson S a, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet [Internet]. 2000 Oct;356 : 1375–1383. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0140673600028403 3 Van Ham M a, van Dongen PW, Mulder J. Maternal consequences of caesarean section. A retrospective study of intra-operative and postoperative maternal complications of caesarean section during a 10-year period. Eur J Obstet Gynecol Reprod Biol [Internet]. 1997 Jul; 74: 1–6. Available from: http://www.ncbi.nlm.nih. gov/pubmed/9243191 4 Souza JP, Gülmezoglu A, Lumbiganon P, Laopaiboon M, Carroli G, Fawole B, et al. Caesarean section without medical indications is associated with an increased risk of adverse short-term maternal outcomes: the 2004-2008 WHO Global Survey on Maternal and Perinatal Health. BMC Med [Internet]. 2010 Jan [cited 2014 Jan 28]; 8: 71. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2 993644&tool=pmcentrez&rendertype=abstract 5 Hofmeyr, GJ; Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012; 10. 6 De Hundt M, Velzel J, de Groot CJ, Mol BW, Kok M. Mode of Delivery After Successful External Cephalic Version: A Systematic Review and Meta-analysis. Obstet Gynecol [Internet]. 2014 May 6 [cited 2014 May 17]; 123: 1327–1334. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24807332 7 Vlemmix F, Rosman AN, Fleuren MAH, Rijnders MEB, Beuckens A, Haak MC, et al. Implementation of the external cephalic version in breech delivery. Dutch national implementation study of external cephalic version. BMC Pregnancy Childbirth. 2010; 10: 20. 8 Otterlo Werkgroep dr. AJ van L en dr. MG van P. NVOG richtlijn - Stuitligging- versie 2.0 (Guideline on breech presentation - version 2.0 of the Dutch Society of Obstetrics and Gynecology) [Internet]. 2008. Available from: file:///C:/Users/PS13C/ Downloads/Stuitligging (1).pdf 9 Verburgt, T; Offerhaus P. Uitwendige versie (External Cephalic Version). NTOG. 2007; 120: 4–7.

240 Chapter 12 Mode of delivery and neonatal outcome after external cephalic version

10 Schuit E, Kwee a, Westerhuis MEMH, Van Dessel HJHM, Graziosi GCM, Van Lith JMM, et al. A clinical prediction model to assess the risk of operative delivery. BJOG [Internet]. 2012 Jul [cited 2014 May 1]; 119: 915–923. Available from: http://www. ncbi.nlm.nih.gov/pubmed/22568406 11 Tan JM, Macario A, Carvalho B, Druzin ML, El-Sayed YY. Cost-effectiveness of external cephalic version for term breech presentation. BMC Pregnancy Childbirth [Internet]. 2010 Jan; 10: 3. Available from: http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=2826287&tool=pmcentrez&rendertype=abstract 12 Häger RME, Daltveit AK, Hofoss D, Nilsen ST, Kolaas T, Øian P, et al. Complications of cesarean deliveries: rates and risk factors. Am J Obstet Gynecol [Internet]. 2004 Feb [cited 2014 May 28]; 190: 428–434. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/14981385 13 Snijder C a, Cornette JMW, Hop WCJ, Kruip MJH a, Duvekot JJ. Thrombophylaxis and bleeding complications after cesarean section. Acta Obstet Gynecol Scand. 2012 May; 91: 560–565. 14 Rasmussen, SA; Maltau J. Complications following cesarean section. Tidsskr Nor Laegeforen [Internet]. 1990 [cited 2014 Mar 3]; 110: 351–353. Available from: https://c38xrlad25d4epd6rje0lpk.sec.amc.nl/pubmed/2309178 15 Nielsen TF, Hökegård KH, Ericson a. Cesarean section and perinatal mortality in Sweden in 1981. Acta Obstet Gynecol Scand [Internet]. 1986 Jan; 65: 865–867. Available from: http://www.ncbi.nlm.nih.gov/pubmed/3825528 16 US National Institutes of Health C. Caesarean section-the first cut isn’t the deepest. Lancet. 2010; 375: 956. 17 Rietberg C. Term breech presentation in The Netherlands from 1995 to 1999: mortality and morbidity in relation to the mode of delivery of 33,824 infants. BJOG An Int J Obstet Gynaecol [Internet]. 2003 Jun [cited 2014 Mar 3]; 110: 604–609. Available from: http://doi.wiley.com/10.1016/S1470-0328(03)01907-4 18 Sullivan EA, Moran K, Chapman M. Term breech singletons and caesarean section: a population study, Australia 1991-2005. Aust N Z J Obstet Gynaecol [Internet]. 2009 Oct [cited 2014 Mar 3]; 49: 456–460. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/19780725 19 Grootscholten, K; Kok, M; Oei, SG; Mol, BW; van der Post J. External cephalic version-related risks: a meta-analysis. Obs Gynecol [Internet]. 2008 [cited 2014 Mar

241

13 Knowledge and decisional conflict scores related to external cephalic version in the Netherlands

Ageeth N. Rosman

Floortje Vlemmix

Margot A.H. Fleuren

Brent C. Opmeer

Ben Willem J. Mol

Eline A.F. Dancet

Marjolein Kok

Submitted

243 Chapter 13 Knowledge and decisional conflict scores

ABSTRACT

ObjectiveTo assess whether in women with a foetus in breech presentation considering external cephalic version (ECV), training the counselling skills of professionals or providing patients with written information and a decision aid alters the final decision, women’s knowledge and decisional conflict.

Design Cluster randomized controlled trial.

Setting25 Dutch hospitals and their network of referring midwifery practices.

Population Women with a singleton pregnancy in breech presentation from 32 weeks between January 2011 and August 2012.

Methods Clusters were randomly assigned to the following strategies: a) standard counselling (care as usual), b) counselling after a one day training of professionals on medical aspects of breech presentation and counselling skills, c) counselling facilitated by the use of a patient information leaflet and a decision aid and d) counselling optimized by both strategies. Patients’ knowledge and decisional conflict were measured with questionnaires before and after the strategy. An overall ANOVA and follow-up test was conducted to identify significant differences between strategies in women’s final choice and in changes over time of their knowledge and decisional conflict.

Results 384 women received a set of questionnaires on knowledge, decisional conflict and the choice on ECV. Overall knowledge was high and decisional conflict low at study entry. Knowledge improved and decisional conflict decreased over time. Only in the care provider strategy knowledge improved significantly if compared to care as usual. 9.8% ( N= 4 of 41) altered their final choice on ECV and underwent an ECV attempt.

Conclusion Knowledge on ECV of Dutch women opting for ECV is already high before counselling. Decisional conflict before counselling is low. Informing and counselling women in different ways, increased knowledge not significantly but lowered decisional conflict significantly. Only a small amount of women altered their decision on ECV (9.8%).

244 Chapter 13 Knowledge and decisional conflict scores

INTRODUCTION Breech presentation occurs in 3 to 4% of all term pregnancies.1 Women with term breech presentation face several treatment options: external cephalic version to increase the chance on a vaginal delivery in cephalic presentation, a trial of labour in breech presentation or a planned caesarean delivery. Confronted with these treatment options, women can experience a lack of knowledge on the consequences of each treatment, which can lead to decisional conflict. External cephalic version (ECV) is an effective treatment to prevent breech presentation at birth and it is considered a relatively safe procedure that is recommended in obstetrical practice guidelines.2–5 Women involved (and their partners) should be counselled about ECV to enable informed decision making. Counselling can be done in several ways. Decision aids offer a systematic approach to involve patients in medical decision making by providing them with information about treatment options and possible outcomes, and by clarifying personal values.6–8 A systematic review comparing care as usual to care as usual with a validated decision aid in obstetric care showed that the use of a decision aid increased knowledge, decreased the decisional conflict related to uncertainty about a course of action, and stimulated women to be more active in shared decision making.8 In the Netherlands, decision aids are not commonly used in prenatal care, and currently there is no Dutch decision aid for ECV. At present, patients are therefore mainly informed and counselled in a face to face setting by their care provider. Earlier research showed that in general, professionals find it difficult to counsel patients and prefer to have more training in counselling skills to inform and counsel patients properly.9 We also know that lack of knowledge to fully inform and counsel patients on ECV, and the inability to counsel women who prefer an elective caesarean delivery are important barriers for professionals to counsel for ECV.10 Fear, the preference to have a planned caesarean delivery, incomplete information and negative experiences from family or friends are important barriers for women to opt for ECV.11 Based on this knowledge we developed two strategies to improve the counselling for ECV: 1) a one day training course of professionals on medical aspects of breech presentation and counselling skills and 2) a written information leaflet supplemented with a decision aid. The aim of this study was to assess whether in women with a foetus in breech presentation considering ECV, training professionals counselling skills

245 Chapter 13 Knowledge and decisional conflict scores

or providing patients with written information and a decision aid alters the final decision, women’s knowledge and decisional conflict.

MATERIAL AND METHODS Setting In the Netherlands prenatal care is supervised by midwives in primary care, and by obstetricians in secondary care. Midwives take care of low risk pregnancies. If specialized care is needed, midwives refer to an obstetrician in an affiliated hospital. We will refer to this organisational unit as a cluster or catchment area around a hospital.

Participants Women with a singleton breech presentation from 32 weeks onwards between January 2011 and August 2012 were eligible for the study. Exclusion criteria were contraindications for ECV according to the national guidelines (uterus anomaly, foetal growth restriction, hypertensive disorder, oligohydramnion, (P)PROM), any contraindication for a vaginal delivery and a spontaneous version after 32 weeks.12,13 An (unselected) convenience sample of women were asked to complete questionnaires about knowledge and decisional conflict to assess knowledge and decisional conflict of women with a foetus in breech presentation facing the option of ECV.

Strategies Based on the results of a determinant analysis of facilitators and barriers of the implementation of ECV, we developed a one day training course on knowledge and counselling skills for health care providers.10 We translated the validated decision aid on breech presentation of Nassar et al. into Dutch, supplemented with specific information about the organisation of ECV in the Netherlands.7 Furthermore, we developed a written information leaflet based on existing leaflets of the professional organisations of obstetric care in the Netherlands (KNOV and NVOG) supplemented with results of qualitative interviews held with women with a foetus in breech presentation opting for ECV.11 The information leaflet was presented in a frequently asked question (FAQ) format. Clusters were randomly assigned to the following strategies: care as usual (control group), a one day training of professionals on medical aspects of breech presentation and counselling skills (care provider strategy), providing women with the information leaflet and decision aid (client strategy) and a

246 Chapter 13 Knowledge and decisional conflict scores combination of both strategies (care provider and client strategy). After approval of the study protocol by the local ethical committees of the hospitals within the clusters, women were invited to participate in the study to assess knowledge of ECV and decisional conflict.

Questionnaire Directly after inclusion (T0), women were asked to fill in a baseline questionnaire about their preference to opt for ECV, their feeling of safety of the procedure of ECV, knowledge of the benefits and risks of ECV and decisional conflict about the treatment of breech presentation. Knowledge and decisional conflict were measured using the knowledge and decisional conflict scales of O’Connor.6 After counselling for ECV and/or after a possible ECV attempt but before labour started (T1), the woman was asked to fill in a follow-up questionnaire, including the same scales as at T0 (knowledge and decisional conflict) supplemented with her decision on ECV and the question whether an ECV attempt was actually performed.

Outcome measures We registered baseline maternal characteristics such as age, length, body mass index (BMI), ethnicity, native language, education level, parity, the preference to opt for ECV, the feeling of safety of ECV and an ECV attempt. Preference was measured on a discrete scale from 0 to 10, by which 0 indicates absolute preference not to undergo ECV and 10 absolute preference to undergo ECV. The feeling of safety of the ECV procedure was also measured on a discrete scale from 0 to 10, by which 0 indicated ECV is absolutely not safe to 10 ECV is absolute safe. Knowledge was measured with six questions; answer options yes, no or unsure. Three questions were about the benefits and risks of ECV, two questions about the success rate of ECV and one question about the information given. The scores on knowledge were, conform the User’s Manual of O’Connor, translated into 0 for yes, 4 for no and 2 for unsure. Thereafter item scores were added, and transformed to a scale between 0-100, with 0 indicating absolutely no correct answers and 100 indicating only correct answers on knowledge questions.6 Decisional conflict was measured conform the low literacy format of the decisional conflict scale of O’Connor, containing five constructs: information, support, values clarity, uncertainty and effective decision.6 The respective constructs cover areas related to information provided to the woman

247 Chapter 13 Knowledge and decisional conflict scores

about ECV; feeling of support by others regarding the decision that has to be made; aspects which are important for the woman in decision making; aspects which women are uncertain about regarding ECV; and satisfaction of a woman regarding process of decision making. All constructs together reflect decisional conflict. Each construct was addressed in three questions except for the latter, which was addressed in four questions. Each question was answered on a five point scale from absolute agreement to absolute disagreement with the statement. The scores of the separate constructs were translated in 0 for absolute agreement with the statement, 1 for agreement, 2 for agreement nor disagreement, 3 for disagreement and 4 for absolute disagreement with the statement. Thereafter, the individual scores on the constructs were added and transformed to a score between 0 – 100, by which 0 meant absolute no decisional conflict facing the treatment options and 100 meant absolute decisional conflict facing the treatment options.6

Statistical analysis An overall ANOVA and follow-up tests were conducted to identify significant differences between strategies in women’s final choice and in changes over time of knowledge and decisional conflict. For all analysis we set the level of significance at p<0.05.

RESULTS 384 women received a set of questionnaires assessing knowledge and decisional conflict between January 2011 and August 2012. We could analyse data of 374 (97.4%) women on knowledge and data of 373 (97.1%) women on decisional conflict (tables 2 and 3). Maternal baseline characteristics of the study population are shown in table 1. (table 1 invoegen) The mean age was 31.0 (SD 4.4 years). The mean body mass index before pregnancy (BMI) was 26.0 (SD 4.8). There were more nulliparous than multiparous in the study population (59% vs. 41%). Most participants were Dutch native speakers (77%), had vocational education (62%) and were Caucasian (69%). The preference to opt for ECV at baseline-was 7.3 (SD 2.7) and the safety perception of ECV was 6.7 (SD 1.9). Overall knowledge, changes over time and the interaction effects of the different strategies are shown in table 2. Overall, knowledge improved from 67.5 to 69.7 (mean difference 2.22, 95% CI 0.29 to 4.16).Follow up tests showed that knowledge increased in the care provider strategy, the client and care provider strategy and care as usual

248 Chapter 13 Knowledge and decisional conflict scores but decreased in the client strategy (table 2b Pairwise comparisons of the estimated marginal means showed that only a significant difference of the mean was found if the client strategy was compared to care as usual (table 2c).

Table 1. Maternal baseline characteristics mean (SD) or percentage (n=379)

Characteristic Value Missing values (%)

Age (years)(SD) 31.0 1 (0.3) (4.4)

Length (cm) (%) 169 140 (36.9) (9.1)

BMI before pregnancy (kg/m²)(SD) 26.0 164 (43.3) (4.8)

Parity ( % nulliparous) 59 12 (3.1)

Ethnicity (% Caucasian) 69 80 (21.1)

Education level (% Post-‐secondary) 62 133 (35.1)

Native language (% Dutch) 77 72 (18.9)

Feeling of safety of (SD) 6.7 7 (1.8) ECV^ (1.9)

Preference to opt for (SD) 7.3 98 (25.6) ECV^ (2.7)

^feeling of safety and preference to opt for ECV are measured on a discrete scale from 0 to 10 whereby 0 means ECV is absolute not safe or absolute no preference to opt for ECV and 10 means ECV is absolute safe of absolute o preference t opt for ECV

Table 2. Knowledge (N=373)

Table 2a. Overall knowledge, change over time, based on estimated marginal means

Baseline Follow up Mean difference p-‐value 95% Confidence Interval for difference

Mean 67.50 69.72 2.22 0.025 0.29 to 4.16

249

Chapter 13 Knowledge and decisional conflict scores

Table 2b. Pairwise comparisons of strategies over time, based on estimated marginal means

time strategy Mean Std. Error 95% Confidence Interval

Lower Upper Bound Bound

Client + Provider 68.27 2.52 63.03 73.51

Client 72.98 2.11 68.62 77.33 Baseline Provider 65.48 2.64 60.11 70.85

Care as usual 63.26 2.50 58.19 68.32

Client + Provider 72.70 2.30 67.81 77.60

Follow up Client 72.11 1.87 68.10 76.11

Provider 68.15 2.24 63.44 72.86

Care as usual 65.92 2.20 61.41 70.44

250 Chapter 13 Knowledge and decisional conflict scores

Table 2c. Interaction effects of strategies

(I) strategy (J) strategy Mean Std. Error p-‐value95% Confidence Interval for Difference Difference

(I-‐J) Lower Bound Upper Bound

Client -‐2.06 2.86 0.49 -‐8.24 4.13

Client + Provider Provider 3.67 3.13 0.26 -‐2.99 10.34

Care as usual 5.90 3.04 0.07 -‐0.55 12.34

Client + Provider 2.06 2.86 0.49 -‐4.13 8.24

Client Provider 5.73 2.87 0.06 -‐0.37 11.83

Care as usual 7.95 2.77 0.01 2.10 13.80

Client + Provider -‐3.67 3.13 0.26 -‐10.34 2.99

Provider Client -‐5.73 2.87 0.06 -‐11.83 0.37

Care as usual 2.23 3.04 0.47 -‐4.15 8.60

Client + Provider -‐5.90 3.04 0.07 -‐12.34 0.55

Care as usual Client -‐7.95 2.77 0.01 -‐13.80 -‐2.10

Provider -‐2.23 3.04 0.47 -‐8.60 4.15

251 Chapter 13 Knowledge and decisional conflict scores

Table 3 shows overall decisional conflict and pairwise comparisons of the estimated marginal means over time and interaction effects of the different strategies. Overall, decisional conflict decreased significantly from 17.5 to 10.2 (p<0.001) (table 3a). Pairwise comparison of the overall decisional conflict in time showed a significant decrease (mean difference -7.34,p <0.001). Follow up tests per strategy showed a decrease of decisional conflict in each strategy (table 3b). Pairwise comparison of the estimated marginal means showed that in none of the strategies a significant effect on the mean difference was found (table 3c). 86.2% (N=331) had a preference to opt for ECV at baseline. 90.4% (N=347) underwent an ECV attempt (OR 1.5 (95% CI 0.96 to 2.35)) (table 4a). In each strategy more women underwent an ECV attempt than preferred at baseline (table 4b).

Table 3. Decisional conflict (N = 374)

Table 3a.Overall mean decisional conflict

Baseline(I) Follow up (J) Mean p-‐value 95% Confidence Interval for difference difference

(J-‐I)

Lower Bound Upper Bound

Mean 17.78 10.16 -‐7.34 <0.001 -‐5.89 8.79

252 Chapter 13 Knowledge and decisional conflict scores

Table 3b. Pairwise comparisons of strategies over time, based on estimated marginal means time strategy Mean Std. Error 95% Confidence Interval

Lower Bound Upper Bound

Client + 19.29 1.80 15.59 22.98 Provider

Client 16.82 1.49 13.76 19.88 Baseline Provider 16.60 1.95 12.66 20.54

Care as usual 17.41 1.85 13.69 21.12

Client + 10.78 1.48 7.57 13.99 Provider

Client 8.10 1.20 5.47 10.73 Follow up

Provider 12.28 1.48 9.08 15.48

Care as usual 9.60 1.48 6.54 12.65

253 Chapter 13 Knowledge and decisional conflict scores

Table 3c. Interaction effects of strategies

(I) strategy (J) strategy Mean Std. p-‐ 95% Confidence Interval for Difference (I-‐J) Error value Difference

Lower Bound Upper Bound

Client 2.57 1.93 0.21 -‐1.59 6.74 Client + Provider 0.59 2.14 0.79 -‐3.99 5.18 Provider Care as usual 1.53 2.08 0.47 -‐2.88 5.94

Client + Provider -‐2.57 1.93 0.21 -‐6.74 1.59

Client Provider -‐1.98 1.97 0.33 -‐6.18 2.22

Care as usual -‐1.04 1.90 0.59 -‐5.06 2.97

Client + Provider -‐0.59 2.14 0.79 -‐5.18 3.99

Provider Client 1.98 1.97 0.33 -‐2.22 6.18

Care as usual 0.94 2.12 0.66 -‐3.51 5.39

Client + Provider -‐1.53 2.08 0.47 -‐5.94 2.88

Care as usual Client 1.04 1.90 0.59 -‐2.97 5.06

Provider -‐0.94 2.12 0.66 -‐5.39 3.52

254 Chapter 13 Knowledge and decisional conflict scores

Table 4a. Preference to opt for ECV at baseline and real number of ECV attempts

N=384 (%) yes no missing

Preference to opt for ECV at baseline 331 (86.2) 41 (10.7) 12 (3.1)

ECV attempt 347 (90.4) 37 ( 9.6) 0

OR for ECV attempt (95% CI) 1.5 (0.96 to 2.35)

Table 4b. Preference to opt for ECV at baseline and number of ECV attempts per strategy

N=384 (%) Care as usual Care provider Client Client and strategy provider care provider strategy strategy (N=82) (N=91) (N=125) (N=86)

Preference to opt for ECV at baseline

Yes 83 (91.2) 67 (81.7) 102 (81.6) 79 (91.7) No 6 ( 6.6) 12 (14.6) 18 (14.4) 5 (5.8) Missing 2 ( 2.2) 3 ( 3.7) 5 ( 4.0) 2 (0.5)

ECV attempt

Yes 88 (96.7) 69 (84.1) 110 (88.0) 80 (93.0)

No 3 ( 3.3) 13 (15.9) 15 (12.0) 6 ( 7.0)

Missing 0 0 0 0

OR for ECV attempt 2.83 1.65 1.19 1.18 (95% CI) (0.72 to (0.82 to (0.53 to 2.69) (0.38 to 3.67) 11.02) 3.34)

255 Chapter 13 Knowledge and decisional conflict scores

The chance that women underwent an ECV attempt was the highest in care as usual group (OR 2.83 (95%CI 0.72 to 11.02) and the lowest in the care provider strategy (OR 1.19 (95% CI 0.53 to 2.69)). 9.8 % (N=4 of 41) altered their decision on ECV and underwent an ECV attempt (table 4b, no preference to opt for ECV (N=41) vs. no ECV attempt (N=37).

DISCUSSION We assessed knowledge, decisional conflict and the final decision on ECV of women with a foetus in breech presentation opting for ECV. Knowledge at baseline was already high and decisional conflict low. Applying strategies to improve the implementation of ECV had only small effects on knowledge and decisional conflict. Knowledge increased except for the client strategy, where knowledge decreased (not significant). Overall decisional conflict decreased significantly (p<0.001) over time. In the final decision more women underwent ECV than having the preference to opt for ECV at baseline. To our knowledge, this is the first study to assess knowledge and decisional conflict of women eligible for ECV in the Netherlands. Fear, the preference to have a planned caesarean delivery, incomplete information and negative experiences from family or friends are important barriers for patients to opt for ECV. This is associated with lack of knowledge and may incur decisional conflict.11 On the other hand, in a quantitative analysis on barriers and facilitators for ECV among care providers, we found that a lack of knowledge to fully inform and counsel patients on ECV, and the inability to counsel women who preferred an elective caesarean delivery are important barriers to counsel for ECV.10 A randomised controlled trial comparing usual care supplemented with a decision aid on ECV with only care as usual showed that women’s knowledge increased with the use of a decision aid.7 This study also showed that women who reviewed the decision aid, had significantly lower decisional conflict and felt that they were better informed and more satisfied with the process of decision making. We measured a high level of knowledge at the baseline. There are several possibilities to explain this high level. Firstly, it is possible that we had selection bias in our study population. The set of questionnaires were overhanded to the women by care providers. We cannot exclude that an unintentional selection took place by the care providers. Another explanation of the high level can be that our study population already did a self-supported search for information before study entry as most women are interested in the foetal presentation and regularly ask their care providers to the foetal presentation.

256 Chapter 13 Knowledge and decisional conflict scores

It could be that they gained already sufficient knowledge to answer the knowledge questions correctly. This could also be a possible explanation for the high level of women who preferred to opt for ECV at baseline and their perception of safety of the ECV procedure. Concerning decisional conflict, we measured low levels of decisional conflict at study entry (overall mean17.8 at baseline). Decisional conflict scores below 25 indicate that there is no decisional conflict on the decision.14 This means that our study population had no decisional conflict at study entry. We were not aware of these findings as never before decisional conflict of women opting for ECV was assessed in the Netherlands. After applying the strategies, overall decisional conflict was lower (mean 10.2 (MD -7.34, p<0.001)) (table 3a). Pairwise comparison showed no significant interaction effects nor a significant decrease of decisional conflict per strategy in the follow up. This means that decisional conflict does not play a major role for Dutch women opting for ECV. We could not demonstrate large effects on decisional conflict. As we introduced more than only a decision aid in our study, we cannot compare our results to the results of the trial of Nassar in which she evaluated the use of a decision aid. Unfortunately we did not find a comparable study as ours to compare our results with. Our study has some limitations. We had an inequality of the number of clusters within each strategy (range 5-7). The number of women who received a set of questionnaires was around 25% (range 21-33%) (figure 1). We think the inequality of the cluster size was not problematic to answer our questions as we wanted to assess whether counselling after an one day training of professionals on medical aspects of breech presentation and counselling skills (care provider strategy), counselling prepared by given women written information and a decision aid improved knowledge, decreased decisional conflict and altered the final decision on ECV. Another limitation was that our written information and decision aid were only available in Dutch. We do not think this had a major impact on knowledge nor decisional conflict as we used the low literacy format of O’Connor’s questionnaire to explore knowledge and decisional conflict. If we take the results of our study into account, we see that if women had a preference to opt for ECV before counselling, there is no much to be gained in the number of women undergoing ECV. In our study the majority of women opted for ECV and underwent an ECV attempt. Overall, the number of ECV attempts increased, but none of the strategies proved superior. In our study the number of women not opting for ECV at baseline and the

257 Chapter 13 Knowledge and decisional conflict scores

number of women not undergoing an ECV attempt was too small to draw conclusions from. In the Netherlands the estimated implementation rate of ECV is about 50-60%.18 It should be interesting to evaluate and assess knowledge and decisional conflict of women who prefer not to opt for ECV. For further research we recommend to assess the use of written information and a decision aid of women who do not opt for ECV. Nassar showed that most care providers felt responsible to convince every eligible patient to opt for ECV and that a small part of care providers were reluctant in offering ECV.7 The use of a decision aid could possibly be a tool to help care providers, who hesitate about ECV to provide themselves of evidence based information about ECV. We also know that the majority of care providers in the Netherlands see counselling for ECV as their task and feel responsible to counsel every eligible woman with a foetus in breech presentation for ECV.10 Sometimes care providers are reluctant to offer ECV to eligible women by reason of not supporting ECV by themselves. It should be interesting to examine to what extent decision aids can help these care providers to provide themselves of evidence towards the benefits of ECV.

258 Chapter 13 Knowledge and decisional conflict scores

CONCLUSION Knowledge on ECV of Dutch women opting for ECV is already high before counselling. Decisional conflict before counselling is low. Informing and counselling women with a foetus in breech presentation opting for ECV in different ways, increased knowledge not significantly but lowered decisional conflict significantly. Only a small amount of women altered their decision on ECV (9.8%).

259 Chapter 13 Knowledge and decisional conflict scores

2 Hofmeyr, GJ; Kulier R. External cephalic version for breech presentation at term ( Review ). Cochrane Database Syst Rev. 2012; 10.

3 Committee Opinion No 340 A. Mode of term singleton breech delivery. Obstet Gynecol 2006. Obs Gynecol. 2006; 108: 235–237.

4 Royal College of Obstetricians and Gynecologists. External cephalic version and reducing the incidence of breech presentation. http://www.rcog.org.uk/files/rcog- corp/uploaded-files/GT20aExternalCephalicVersion.pdf. 2010. p. 1–8.

5 The Royal Australian and New Zealand College of Obstetricians and Gynecologists. Management of Breech Presentation at Term. 2001. p. file:///C:/Users/PS13C/ Downloads/C–Obs%2011%20Mana.

6 Stacey D, Murray MA, Légaré F, Dunn S, Menard P, O’Connor AM. Decision Coaching to Support Shared Decision Making: A Framework, Evidence, and Implications for Nursing Practice, Education, and Policy. Worldviews on Evidence- Based Nursing, 2008 Mar 5 (1): 25-35

7 Nassar N, Roberts CL, Raynes-Greenow CH, Barratt A, Peat B. Evaluation of a decision aid for women with breech presentation at term: a randomised controlled trial [ISRCTN14570598]. BJOG [Internet]. 2007 Mar [cited 2014 Mar 20]; 114: 325– 333. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=24 08658&tool=pmcentrez&rendertype=abstract

8 Vlemmix F, Warendorf JK, Rosman A N, Kok M, Mol BWJ, Morris JM, et al. Decision aids to improve informed decision-making in pregnancy care: a systematic review. BJOG [Internet]. 2013; 120: 257–266. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/23145991

9 Kuntze J, van der Molen HT, Born MP. Increase in counselling communication skills after basic and advanced microskills training. Br J Educ Psychol [Internet]. 2009 Mar [cited 2014 May 30]; 79: 175–188. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/18559140

260 Chapter 13 Knowledge and decisional conflict scores

10 Rosman AN, Vlemmix F, Beuckens A, Rijnders ME, Opmeer BC, Mol BWJ, et al. Facilitators and barriers to external cephalic version for breech presentation at term among health care providers in the Netherlands: a quantitative analysis. Midwifery [Internet]. Elsevier; 2014 Mar [cited 2014 May 6]; 30: e145–50. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24530121

11 Rosman AN, Vlemmix F, Fleuren MAH, Rijnders ME, Beuckens A, Opmeer BC, et al. Patients’ and professionals’ barriers and facilitators to external cephalic version for breech presentation at term, a qualitative analysis in the Netherlands. Midwifery [Internet]. Elsevier; 2014 Mar [cited 2014 May 6]; 30: 324–330. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23664312

12 Verburgt, T; Offerhaus P. Uitwendige versie (External Cephalic Version). NTOG. 2007; 120: 4–7.

13 Otterlo Werkgroep dr. AJ van Loon en dr. MG van Pampus. NVOG richtlijn - Stuitligging- versie 2.0 (Guideline on breech presentation - version 2.0 of the Dutch Society of Obstetrics and Gynecology) [Internet]. 2008. Available from: file:///C:/ Users/PS13C/Downloads/Stuitligging (1).pdf

14 O’Connor AM. Validation of a decisional conflict scale. Med Decis Making [Internet]. 1995 [cited 2014 May 30]; 15: 25–30. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/7898294

261

14 Summary and general discussion

263 Chapter 14 Summary and general discussion

SUMMARY AND GENERAL DISCUSSION

The work presented in this thesis focuses on improvement of care for women with a term foetus in breech presentation in order to improve maternal and neonatal outcome. The results presented in part one emphasise the importance of ECV, as successful versions reduce the rate of planned vaginal breech delivery, and planned as well as emergency caesarean sections with the associated morbidity and mortality for both mother and child. In view of these health benefits associated with ECV, it is important that this intervention is adequately implemented in obstetric practice. The suboptimal implementation of ECV and strategies for improvement are the focus of research in part two and three. First, we have developed implementation strategies based on a determinant analysis to improve the three core elements of the Dutch ECV guidelines: counselling for ECV, advising ECV and arranging an ECV attempt (part two). We then evaluated the effect of these strategies on the implementation of ECV in a cluster randomized controlled trial (part three). Below, we first summarize the subject and main findings for each chapter. Subsequently we will discuss novel insights, clinical interpretation, and implications for daily practice. We debate on methodological and practical limitations of this work, and in terms of implications and recommendations for further research.

SUMMARY In chapter 1 we introduced the rationale of this thesis. ECV is the only treatment to prevent breech presentation at birth and thereby the adverse maternal and neonatal outcome during current or subsequent birth. According to the guidelines all women with a breech presentation at 36 weeks gestation and onwards should be offered an ECV attempt, the treatment should be advised and an attempt needs to be arranged after informed consent. However, guidelines are not automatically adhered to in daily practice and improvement of implementation of ECV is needed.

Part one: breech delivery

The first cohort study of this thesis, with data from the Dutch national perinatal registry (chapter 2), showed that publication of the Term Breech Trial resulted in a shift towards elective caesarean delivery up to 60%of

264 Chapter 14 Summary and general discussion breech birth, within three months of time and remained stable up to 2007. This increase in elective caesarean delivery led to a decrease in perinatal mortality and morbidity among women delivering a child in breech at term. Still, 40% of these women attempt vaginal birth. The relative safety of an elective caesarean should be weighed against the consequences of a scarred uterus in future pregnancies. Presumed risk selection of the women attempting vaginal breech birth since the Term Breech Trial, has not led to better outcome of the planned vaginal deliveries. We were not able to select a subgroup of women based on parity, type of breech, birth weight and onset of labour, with a low risk of adverse neonatal outcome during planned vaginal breech delivery, compared to elective caesarean delivery.

In chapter 3 we describe the consequences of increased elective caesarean delivery rates in nulliparous women with a term breech presentation for future pregnancies. Elective caesarean delivery for term breech presentation leads to a significant increase in maternal morbidity and adverse neonatal outcome in subsequent pregnancies. Combining outcomes of the first and the subsequent pregnancy only resulted in an overall reduction in neonatal morbidity during both pregnancies. Excluding those women with a trial of labour from the analyses, did not alter these findings, accept that the number needed to treat with an elective caesarean delivery to prevent one neonatal morbidity in both (first and subsequent) pregnancies decreased from 95 to 55.

Part two: Development of strategies to improve implementation of ECV

The systematic review of guidelines and literature on contraindications of ECV (chapter 4) showed a lack of consistency. Of the 22 contraindications mentioned in five international guidelines, only one (oligohydramnios) was mentioned in all. Literature does not provide much evidence (not higher than level III) for some of the mentioned contraindications in guidelines. Implementation of ECV is thus hampered by the fact that there is no general consensus on the eligibility for ECV. Therefore, we propose to limit contraindications for ECV to clear empirical evidence or to a clear pathophysiologic relevance.

In chapter 5 we report findings of a vignette study, evaluating women’s preference to opt for ECV regarding different levels of success rates,

265 Chapter 14 Summary and general discussion

experienced pain, use of uterine relaxants and risk of emergency caesarean delivery. Success rates and expected pain are the most important factors influencing the willingness to opt for ECV. These factors need to be addressed during counselling for ECV. The results of a qualitative analysis of clients’ and professionals’ barriers and facilitators towards ECV are presented in chapter 6. Among professionals the main barriers for ECV were a lack of knowledge and perceived skills to fully inform and counsel clients on ECV, and the experienced difficulty to counsel women who preferred a planned caesarean delivery. The main facilitator was an unambiguous policy on (counselling for) ECV within the region. More time and finance to counsel were also considered to be facilitators. Among pregnant women the main barriers for ECV were fear, incomplete and not well-timed information and incidents of severe birth complications (not necessarily breech related) of significant others. The main facilitator was confidence of the safety of ECV. The preferred mode of delivery could be either a barrier (wish for a planned caesarean delivery) or a facilitator (wish for a planned vaginal birth (at home)). In chapter 7 we describe the results of the quantitative analysis of the facilitators and barriers for ECV mentioned by the professionals. To do so, we performed an online survey. Professionals’ reported adherence to national guidelines varied from 84% who counselled all women, 73% who always advised ECV, and 82% who succeeded in arranging an ECV attempt for (almost) all their clients. Self-efficacy (degree to which the professional believes he or she is able to implement the activities involved in the guideline) was the most important determinant influencing adherence to national guidelines.

In chapter 8 we describe the outcome of a systematic review of decision aids in pregnancy care. This study was performed to evaluate the effectiveness of a decision aid on informed decision making. We identified ten decision aids, of which one was used to help decide women on ECV. There was a significant increase in knowledge, decrease in decisional conflict and decrease in anxiety among those women who were counselled using a decision aid. The uptake of decision aids in daily practice should be the focus of future studies.

266 Chapter 14 Summary and general discussion

Part three: Evaluating strategies to improve implementation of ECV

In chapter 9 we present the results of a retrospective cohort study on the implementation of ECV in the Netherlands. We evaluated 4,770 breech presentation at birth in 36 hospitals. External cephalic version was performed in 62.2% of the eligible women, with a range from 8% to 84%. Suboptimal implementation was mostly caused by care providers not offering the treatment and by women not opting for ECV. Implementation rates were higher in teaching hospitals, hospitals with special ECV office hours, larger obstetric units and hospitals located in larger cities. An interesting finding was the 7.4% of undiagnosed breech presentations until birth.

Chapter 10 describes the results of our cluster randomised controlled trial in which we evaluated the effect of two implementation strategies on the uptake of ECV. The two implementation strategies, a client strategy (information leaflets and decision aid) and a care provider strategy (one day counselling course focused on knowledge and counselling skills), were developed based on the results described in chapters 6, 7 and 8. The overall implementation rate of external cephalic version in the clusters participating in the cluster randomised controlled trial was 72% (1169 of 1613 eligible clients) with a range between clusters of 8 and 95%. Suboptimal implementation of ECV was mostly caused by the care provider not offering the treatment to eligible women (18.0%) and secondly due to women not opting for the offered attempt (9.5%). We could not demonstrate an effect from either the client nor care provider strategy or the combination of both strategies on the rate of counselled clients, clients opting for ECV and ECV attempts, compared to the care as usual. Adherence to implementation strategies was estimated at 20 to 23%. In per protocol analyses, the overall implementation rates of ECV were higher, but also this analysis did not demonstrate a beneficial effect of any implementation strategy. The cluster randomised controlled trial revealed important information regarding current practice, i.e. large variety in implementation of ECV among regions. We therefore also performed a cost-effectiveness analysis (chapter 11), in which we compared baseline measurements of the clusters (chapter 9) with their implementation rates after enrolment of the implementation strategies (chapter 10). The results of this analysis indicate that the combined client and care provider strategy can be successful in increasing

267 Chapter 14 Summary and general discussion

ECV implementation rates (mean increase of ECV attempts 7%, range among clusters 3-44%). The incremental cost effectiveness ratio for this strategy was €1,353 to gain one ECV attempt. In chapter 12 we present the delivery outcomes of all women in our cluster randomised controlled trial. The results confirm the benefits of ECV, with a significantly lower overall caesarean delivery rate in the group of women with an ECV attempt compared to women who did not receive an ECV. Women with a cephalic presentation after ECV had an decreased risk at instrumental vaginal delivery compared to those with a spontaneous cephalic presentation and an increased risk on an overall caesarean delivery compared to women with a spontaneous cephalic presentation. Chapter 13 describes the effect of our implementation strategies on knowledge and decisional conflict scores in women enrolled in our trial. Baseline knowledge was high and decisional conflict was low. The implementation strategies did not significantly affect these outcomes.

268 Chapter 14 Summary and general discussion

GENERAL DISCUSSION

In this thesis, the suboptimal implementation of ECV in Dutch clinical practice was systematically and thoroughly investigated. We provided a detailed insight in the implementation rates of ECV and quantified the reasons why ECV was suboptimal implemented in a cohort of clusters representative for the Netherlands. We identified barriers and facilitators for ECV on a client and care provider level and developed implementation strategies based on this knowledge. We systematically evaluated these strategies in a cluster randomised controlled trial with a factorial design, due to which we were able to investigate the effects of the single strategies and the combination of both strategies. We measured outcomes to evaluate effectiveness of the strategy as well as care provider adherence to the strategy and finally we also took cost-effectiveness into account. The evaluation of the strategies in the cluster randomised controlled trial did not show a significant effect which might be due to several factors which we will discuss below.

Baseline implementation rates of ECV What first stood out when the results of our retrospective cohort study (chapter 9) and the cluster randomised trial (chapter 10) became available, was the high overall implementation rate of ECV (72%). However, the variability among clusters was considerable , between 8 and 93,% and one of the most important additional findings of our studies. In advance, no data were available on this overall rate or variation. Based upon implementation rates reported in the literature we estimated the implementation rate to be around 50% in the Netherlands. This resulted in a sample size calculation with an underestimated baseline implementation rate and a relatively optimistic improvement from 50% to 80% .We anticipated to some potential imbalance in baseline implementation rate of ECV among the clusters by stratification on the presence of an ECV office hour as it is plausible that those care providers who have structured care for breech presentation, have better implementation rates than those clusters with a lack of this organized care. However, in randomized controlled trials with a small number of randomized units of observation (25 clusters in our study) even one or two allocated units could cause a considerable imbalance that could also influence the outcomes. The wide range of baseline ECV implementation might be an explanatory factor influencing the lack of significant differences between our implementation

269 Chapter 14 Summary and general discussion

strategies. In a situation with more variability, one needs larger differences in order to demonstrate a statistically significant effect. Re-evaluation of the data in the cost-effectiveness study, taking baseline implementation rates within the clusters into account, did show a potential effect of the combined strategy if clusters were compared to their own baseline measurement.

Shared decision making on ECV During the identification of barriers and facilitators for implementation of ECV, an important issue emerged: the apparent contradiction between pursuing a high uptake of a treatment (as recommended in national guidelines) and shared decision making , meaning counselling of clients by providing unbiased nondirective information on the advantages, disadvantages and expected risks and outcomes of all treatment options and assisting them in clarifying their personal values related to corresponding outcomes and adverse effects. The question is whether a 100% uptake among eligible women needs to be pursued. This might be the best aim regarding improving maternal and neonatal outcome among term breech presentations, and thus from a doctors or even social perspective. However, it does not leave any space for individual women to choose their desired care according to their personal values and believes. To improve the amount of women opting for ECV, we developed the client strategy based on the results of chapter 6, 7 and 8. The results of our client interviews indicated a lack of well-timed and objective information provision to women with a breech presentation (or at least to a part of them). Therefore we provided timely information presented in an objective information leaflet and decision aid. One could comment that tools like decision aids are developed to stimulate shared decision making, which does not necessarily affect the uptake of a treatment as intended by the guideline. However, it is a proven effective tool to inform clients uniformly and objectively and has a positive effect on informed decision making.

Effectiveness of implementation strategies Due to the set time frame of our study, limited financial resources and manpower, we were restricted in some ways with regard to both the development and a proper introduction of the implementation strategies. Regarding the issue of potential ineffective implementation strategies, some barriers and facilitators could not be incorporated within the implementation strategies. Determinants related to the context or social-political environment

270 Chapter 14 Summary and general discussion are difficult to change within a short amount of time and/or with limited financial resources.1 For instance, during the focus group meetings, the lack of a local transparent policy about informing, counselling, referral and performing ECV was frequently mentioned and presence of such a protocol in several clusters was considered to be a facilitator. We were not able to develop such a protocol, let alone influencing the local policy concerning the processes of care, for all participating regions. We have paid attention to this topic during the meetings in the care provider strategy, where we discussed the guideline and the barriers and facilitators for implementation of the guideline. We thereby aimed to improve insight into the care providers’ own clinical practice and uncover those areas where they could improve the provided care. We also presented two examples of well performing ECV office hours in other clusters, as an inspiration source for those clusters without a regional protocol on ECV. However, collaboration between care providers within a cluster is often embedded in a history of organization of care and the attitude of care providers to one another. A more intense way of guidance to improve collaboration within clusters might have been more effective.

Adherence to implementation strategies Our study and its intention to treat analysis reflects circumstances in daily practice. Just as there is extensive empirical evidence that the level of use of guidelines affects outcomes in clients, this is also true for adherence to implementation strategies.2 We need to take into account whether a strategy has actually been put into practice by professionals in order to determine its benefits. Otherwise, one may incorrectly conclude that an implementation strategy is not efficacious in itself when, in fact, it has not been properly implemented. There is a lack of publications on implementation strategies with non significant results, potentially indicating a publication bias in favour of those studies with significant results. This might hinder dissemination of knowledge on developed strategies which are effective itself, but which was not shown in the trial due to other influences. Therefore, other research groups cannot benefit from existing knowledge or learn from each other’s mistakes. Concerning the issue of inadequate adherence, we would like to reflect on two problems. First, inadequate adherence to the implementation strategy is partly caused by the fact that we were only able to organise an introduction session (client strategy) and single one-day course (care provider strategy)

271 Chapter 14 Summary and general discussion

instead of several consecutive ones. This means that some professionals were not able to attend the meeting and were thus not subjected to the implementation strategy, resulting in a potential lower effect of the strategy on implementation of ECV. Secondly, a systematic introduction of implementation strategies, including feedback on their uptake, takes some time. A similar implementation study has shown that it might take up to two years before optimal effect of the implementation strategy is reached.3 Ideally, one would first properly introduce the strategies to all care providers, provide for performance feedback, increase their self-confidence as a result of experience in practice and only then measure the results on adherence to the key recommendations in the guidelines. Within the cluster randomised controlled trial, the majority of eligible clients did not receive an ECV attempt because the care provider did not offer the treatment. Thus, the implementation is mainly limited by the care provider due to suboptimal adherence to existing guidelines, despite the effort to eliminate barriers and to promote facilitators within implementation strategies. Although it is interesting to discuss all practical and methodological issues of our implementation study, at the bottom line the question to all care providers remains why working according to state of the art guidelines is apparently not self-evident. Obviously, every care provider is allowed to deviate from advices in guidelines, if well argued. But this is clearly not the case for ECV as no notification of considering ECV was found in 40% of hospital charts. We have started this thesis with the statement that guidelines are not automatically adhered to by care providers. We would like to emphasize that all care providers are required to adhere to guidelines and to consider this as one of their core clinical tasks in daily practice to provide evidence based care to the clients. The significant difference in ECV implementation rates among clusters indicate that adhere to guidelines in daily practice is conceivable, and poor implementation rates in other regions should not be accepted within current evidence based practice.

Implications for daily practice Considering the broad practice variation in the implementation of ECV we recommend that counselling for ECV and ECV attempt outcomes (success rate and complication rates) will be included in national registries (LVR 1 and 2) to enable evaluation of care for term breech presentation in a systematic

272 Chapter 14 Summary and general discussion way on a yearly basis. This information can be used to provide tailored feedback and guidance to those regions with suboptimal adherence to national guidelines. The effect of these interventions will become clear in following years (and this will complete the circle of Deming (plan-do-check- act) for continuous improvement of health care). Evaluation of care for breech presentation from these national registries should focus on the proportion of women receiving an ECV attempt as well as the success rate of ECV in order to reach an optimal decrease in breech presentation at birth. We calculated the potential gain to make (optimal implementation and optimisation of ECV success rates) based on the data of our cluster randomised controlled trial and the Dutch perinatal registry 2012 (Figure 1 and 2). The data presented in this thesis show that around 7% of breech presentation from 36 weeks gestation onwards have a contraindication for ECV according to national guidelines. The mean rate of women not opting for ECV is 10%. Thus an implementation rate of 83% should be feasible in daily practice, compared to the current 72%. The increase in ECV rate could lead to a decrease of breech presentation at birth from 4,821 to 4,543 per year ( 5.8%) (Figure 1). (invoegen figuur 1) Some regions with optimal ECV implementation within our cluster randomised controlled trial achieved success rates up to 60%. These regions did have specialized ECV office hours and their success rates were not biased by case selection (meaning an optimal implementation of ECV rates among eligible women). Achieving such success rates among all ECV attempts can lead to a decrease in breech presentation at birth from 4,821 to 4,543 per year (30.9%) (Figure 2). (invoegen figuur 2) Two potential strategies to increase success rates could either be the implementation of similar ECV office hours in all regions or referral of clients to those regions with the most successful office hours. An third option would be the further evaluation of the use of uterine relaxants during ECV attempts, and the implementation of those uterine relaxants which are proven effective in order to increase ECV success rates.4

273 Chapter 14 Summary and general discussion

274 Chapter 14 Summary and general discussion

Implications for future research Although the implementation of ECV may implicate the final step in evidence based care for breech presentation, there are still some issues that remain unanswered.

1. Implementation With the current research, we gained valuable insight in the implementation of ECV in the Netherlands. Registration of ECV in national databases as a quality indicator will increase this insight. The developed combined client and care provider strategy showed some potential in the cost-effectiveness analysis and therefore could be used within a research setting, to evaluate its potentials in those clusters with poor ECV implementation rates.

2. Effectiveness of ECV Although the uptake of ECV is an important issue, it remains of importance that once an ECV is attempted, there is a high chance of success. We found that the highest ECV success rates where among regions with an ECV office. Future research should focus on why these practices are so successful. Are these high success rates caused by the skills of the care providers performing ECV, is it the use of uterine relaxants or a combination of these factors?

3. Diagnosing breech presentation We found that in 3.5% of the term breech deliveries, the breech presentation was only discovered during labour. This is in line with earlier research on detection of breech presentation, in which it is stated that clinical examination is not sensitive enough for detection and timely trimester ultrasound to detect these breech presentations should be performed.

275 Chapter 14 Summary and general discussion

REFERENCES

1 Huijg JM, Crone MR, Verheijden MW, van der Zouwe N, Middelkoop BJC, Gebhardt W a. Factors influencing the adoption, implementation, and continuation of physical activity interventions in primary health care: a Delphi study. BMC Fam Pract [Internet]. BMC Family Practice; 2013 Jan [cited 2014 May 27];14 : 142. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3848920&tool=p mcentrez&rendertype=abstract 2 Durlak J a, DuPre EP. Implementation matters: a review of research on the influence of implementation on program outcomes and the factors affecting implementation. Am J Community Psychol [Internet]. 2008 Jun [cited 2014 May 12]; 41: 327–350. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18322790 3 Fleuren M, Dusseldorp E, van den Bergh S, Vlek H, Wildschut J, van den Akker E, et al. Implementation of a shared care guideline for back pain: effect on unnecessary referrals. Int J Qual Health Care [Internet]. 2010 Oct; 22: 415–420. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2939752&tool=pmcent rez&rendertype=abstract 4 Cluver C, Gj H, Gml G, Sinclair M. Interventions for helping to turn term breech babies to head first presentation when using external cephalic version ( Review ). 2012;

276

15 Nederlandse samenvatting en algemene discussie

279 Chapter 15 Nederlandse samenvatting

SAMENVATTING EN ALGEMENE BESCHOUWING

Dit proefschrift gaat over het verbeteren van het beleid rondom stuitligging en de uitwendige versie. Het proefschrift bestaat uit drie delen. Het eerste deel beschrijft het belang van de uitwendige versie bij stuitligging: succesvolle versies leiden tot minder geplande vaginale stuitbevallingen, minder primaire keizersneden en tot minder spoedkeizersneden. Dit leidt weer tot minder maternale en neonatale morbiditeit en mortaliteit in de huidige, dan wel volgende zwangerschappen. De uitwendige versie is de enige behandeling om stuitligging bij de geboorte te voorkomen en dient derhalve aan elke zwangere met een stuitligging te worden aangeboden.

In deel twee en drie staat de implementatie van de uitwendige versie cen- traal. In deel twee wordt beschreven hoe nieuwe implementatiestrate- gieën zijn ontwikkeld op basis van een verrichte determinanten analyse. De ontwikkelde strategieën moeten er toe leiden dat de drie kernaanbevelingen uit de Nederlands richtlijnen die gaan over stuitligging en uitwendige versie (NVOG, stuitligging 2.0 en KNOV, standpunt uitwendige versie) beter worden nageleefd. Deze kernaanbevelingen zijn (1) het counselen voor een uitwendige versie, (2) het aanbevelen van deze behandeling en (3) het organiseren van deze behandeling voor de zwangere.

In deel drie worden de ontwikkelde implementatiestrategieën geëvalueerd in een cluster gerandomiseerd onderzoek. Binnen dit deel worden per hoofdstuk de belangrijkste bevindingen besproken.

Vervolgens bediscussiëren we nieuwe inzichten, methodologische en praktische belemmeringen van ons werk, de klinische betekenis en de impact op de dagelijkse praktijk, en tenslotte doen we aanbevelingen voor toekomstig onderzoek.

SAMENVATTING

In hoofdstuk 1 wordt de aanleiding tot dit proefschrift beschreven. Vanuit een historisch oogpunt wordt de uitwendige versie als behandeloptie bij stuitligging geïntroduceerd. De uitwendige versie is de enige behandeloptie die een moeder en kind kan behoeden voor de kans op slechtere perinatale uitkomsten door een geboorte in stuitligging te voorkomen. Tegenwoordig

280 Chapter 15 Nederlandse samenvatting wordt de uitwendige versie geadviseerd als eerste behandeloptie in na- tionale en internationale richtlijnen. Het probleem bij richtlijnen is echter dat de adviezen uit de richtlijnen niet automatisch worden opgevolgd in de dagelijkse praktijk, waardoor geadviseerde zorg niet optimaal wordt geïm- plementeerd. Voorgaande studies duiden er op dat dit ook het geval is voor de richtlijnen voor stuitligging en uitwendige versie zo te zijn. Om diverse re- denen krijgen vrouwen die in aanmerking komen voor een uitwendige versie dus niet de zorg die zij zouden moeten ontvangen. Rekening houdend met de vele voordelen van de uitwendige versie, is dat een ongewenst effect en is een betere implementatie van de richtlijnen en daarmee dus een betere implementatie van de uitwendige versie noodzakelijk.

Deel 1: stuitbevalling

De eerste studie die in dit proefschrift besproken wordt, is een cohortstudie met gegevens uit de Perinatale Registratie Nederland (PRN) (hoofdstuk 2). Deze studie laat zien dat in een periode van drie maanden na publicatie van de Term Breech Trial (2000), waarin de keizersnede als meest veilige manier van bevallen bij een stuitligging wordt geadviseerd, het percentage electieve keizersneden steeg van 30 naar 60%. Sindsdien is het percentage electieve keizersneden wegens stuitligging stabiel gebleven (tenminste tot 2007). Een keizersnede leidt tot een afname van de perinatale morbiditeit en mortaliteit, maar heeft als negatief gevolg dat er een litteken in de baarmoeder -ont staat. De relatieve veiligheid van een keizersnede moet worden afgewogen tegen de gevolgen van een litteken in de baarmoeder voor een volgende zwangerschap. Veronderstelde risicoselectie bij vrouwen die een vaginale stuitbevalling willen proberen (ongeveer 40%), heeft niet geleid tot betere uitkomsten. Op basis van patiëntkarakteristieken (pariteit, soort stuitligging, geboortegewicht en start bevalling) kon geen subgroep gedefinieerd worden met een laag risico op slechte perinatale uitkomsten tijdens en na een vaginale stuitbevalling in vergelijking met de electieve keizersnede.

In hoofdstuk 3 beschrijven wij de gevolgen van het toegenomen percentage keizersneden bij nullipara met een a terme stuitligging, op de uitkomst voor moeder en kind in een volgende zwangerschap. De electieve keizersnede leidt tot een significante toename van de maternale en neonatale morbiditeit in de volgende zwangerschap. Als we de uitkomsten van de eerste en

281 Chapter 15 Nederlandse samenvatting

de daaropvolgende zwangerschap combineren, dan leidt de electieve keizersnede vanwege de stuitligging alleen tot een afname van de neonatale morbiditeit in beide zwangerschappen. De kans op neonatale mortaliteit is niet significant verschillend. Wanneer we alle vrouwen die een proefbaring ondergaan buiten beschouwing laten, verandert deze uitkomst niet behalve dat het aantal vrouwen dat behandeld moet worden met een electieve keizersnede om één neonatale morbiditeit te voorkomen, afneemt van 95 naar 55.

Deel 2: ontwikkeling van implementatiestrategieën

In hoofdstuk 4 presenteren we een systematische review van richtlijnen en literatuur over contra-indicaties voor de uitwendige versie. Het overzicht van de richtlijnen laat zien dat zowel nationaal als internationaal weinig over- eenstemming bestaat over contra-indicaties voor een uitwendige versie. Vijf richtlijnen beschrijven 22 contra-indicaties; waarbij alleen een oligohydramnion in alle richtlijnen wordt genoemd. Daarnaast zijn er weinig artikelen gepubliceerd over de onderbouwing van de contra-indicaties te vinden en de bewijsniveaus zijn niet hoger dan niveau III. Een gevolg van het gebrek aan consensus en gebrekkige onderbouwing is een belemmering bij de implementatie van de uitwendige versie. Daarom stellen wij voor om de contra- indicaties voor de uitwendige versie te beperken tot die enkele waarvoor empirisch bewijs is dan wel die contra-indicaties waarvan het op pathofysi- ologisch gronden erg aannemelijk is om deze te handhaven.

In hoofdstuk 5 beschrijven we de resultaten van een zogenaamde vignetten- studie, waarin we onderzoeken wat de keuze van vrouwen om al dan niet een uitwendige versie te ondergaan, beïnvloed. De voorkeuren zijn getest aan de hand van verschillende succespercentages van de versie, pijnbeleving tijdens de versie, het gebruik van weeënremmers om de slagingskans te vergroten en de risico’s op een spoedkeizersnede ten gevolge van de versiepoging. De succespercentages en de te verwachten pijn blijken de belangrijkste factoren te zijn waarop vrouwen besluiten wel of geen versiepoging te ondergaan. Daarom moeten deze factoren benoemd en besproken worden tijdens het counselingsgesprek over de uitwendige versie.

282 Chapter 15 Nederlandse samenvatting

In hoofdstuk 6 presenteren we de resultaten van een kwalitatieve determinanten analyse onder patiënten en zorgverleners (gynaecologen en verloskundigen). De bevorderende en belemmerende factoren voor de implementatie van de uitwendige versie zijn onderzocht. Voor zorgverleners zijn de belangrijkste belemmerende factoren een gebrek aan kennis en kunde om vrouwen met een stuitligging goed te informeren en te counselen over de uitwendige versie. Dit speelt zeker een belangrijke rol voor de groep vrouwen die al voor de counseling een expliciete voorkeur voor een geplande keizersnede hebben. De belangrijkste bevorderende factor is een eenslu- idend beleid over (het counselen over) de uitwendige versie binnen de regio. Geld en tijd worden daarnaast ook genoemd als belangrijke bevorderende factoren.

Patiënten met een à terme stuitligging geven aan dat angst voor een versiepoging, incomplete en late informatie over de uitwendige versie de belangrijkste belemmerende factoren zijn. Het optreden van een ernstige zwangerschapscomplicatie bij kennissen en/ of familie is ook genoemd als barrière, ook als deze complicatie niet gerelateerd was aan stuitligging. De belangrijkste bevorderende factor was het vertrouwen dat de uitwendige versie een veilige behandeling is. De voorkeur voor de manier van bevallen is zowel een bevorderende (wens om thuis te bevallen) als een belemmerende (voorkeur voor een geplande keizersnede) factor.

In hoofdstuk 7 beschrijven we de uitkomsten van een kwantitatieve analyse van bevorderende en belemmerende factoren onder zorgverleners bij de implementatie van de uitwendige versie. De resultaten van een online vragenlijst geven aan dat 84% van de zorgverleners alle vrouwen counselt die in aanmerking komen voor een uitwendige versie, 73% adviseert een uitwendige versie en 82% regelt een uitwendige versiepoging. De determinant ‘self-efficacy’, de mate waarin zorgverleners zichzelf in staat achten om te werken conform de richtlijn en alle handelingen te verrichten die in de richtlijnen worden aanbevolen, is het sterkst verbonden met de opvolging (adherentie) van de kernaanbevelingen in de richtlijn.

283 Chapter 15 Nederlandse samenvatting

In hoofdstuk 8 worden de uitkomsten van een systematische review over het gebruik van decision aids in de zwangerschap beschreven. Een decision aid is een keuzehulpmiddel waarin het effect en de risico’s van een bepaal- de behandeling op een onbevooroordeelde, systematische manier worden weergegeven, met visuele ondersteuning. We identificeerden tien decision aids, waarvan één decision aid over de uitwendige versie. Na gebruik van een decision aid, is er een significante stijging van het kennisniveau, minder twijfel over de gemaakte keuze (decisional conflict) en minder angst voor de behandeling. Het is aan te bevelen routinematig gebruik van decision aids te bevorderen.

Deel 3: Het evalueren van de implementatie strategieën

In hoofdstuk 9 worden de resultaten van een retrospectieve cohort studie naar de implementatie van de uitwendige versie in Nederland beschreven. Wij evalueerden 4770 stuitbevallingen (vaginaal of keizersnede) in 36 ziekenhuizen. Bij 62.2% van de vrouwen die in aanmerking kwamen voor een uitwendige versie, werd deze behandeling ook toegepast (spreiding tussen ziekenhuizen: 8-84%). Suboptimale implementatie van de uitwendige versie werd in de meerderheid van de gevallen veroorzaakt door zorgverleners die geen versiepoging aanboden en ten tweede doordat zwangere vrouwen afzagen van een versiepoging. Het percentage versiepogingen lag hoger in opleidingsziekenhuizen, ziekenhuizen met speciale versiespreekuren, ziekenhuizen met grote verloskunde afdelingen en ziekenhuizen in de grotere steden. Een opvallende bevinding was dat in 7.4% van alle stuitbevallingen, de stuitligging pas tijdens de bevalling werd vastgesteld.

In hoofdstuk 10 presenteren we de resultaten van een cluster gerandomiseerd onderzoek waarin we de effecten van twee implementatiestrategieën op de implementatie van de uitwendige versie evalueren. De twee strat- egieën, een cliënt gerichte strategie (informatiebrochure en keuzehulpmiddel (decision aid)) en een zorgverlener gerichte strategie (eendaagse cursus gericht op kennis over en counseling voor en de uitwendige versie en stuitligging), zijn ontwikkeld op basis van de resultaten zoals beschreven in de hoofdstukken 6, 7 en 8. De implementatie van de uitwendige versie in alle

284 Chapter 15 Nederlandse samenvatting deelnemende clusters tezamen bedroeg 72% (1169 van 1613 in aanmerking komende vrouwen) met een range van 8 -95% tussen de clusters. Een suboptimale implementatie werd vooral veroorzaakt door het niet aanbieden van een uitwendige versie door zorgverleners (18.0%) en doordat zwangere vrouwen afzagen van een versiepoging (9.5%). Er werd geen effect aan- getoond van de implementatiestrategieën op het percentage gecounselde vrouwen, het aantal vrouwen dat voor een versiepoging opteerde en het percentage versiepogingen. Ook de combinatie van beide strategieën gaf geen verbetering met betrekking tot deze uitkomstmaten in vergelijking met de controle clusters. De adherentie aan de strategieën wisselde tussen de 20-23% per strategie. In een per protocol analyse was de implementatie van de uitwendige versie hoger dan in de intention-to-treat analyse, maar ook in deze analyse kwam geen van de strategieën als meest effectief naar voren.

De cluster gerandomiseerde onderzoek leverde veel informatie op over de huidige praktijkvoering met betrekking tot de uitwendige versie in Neder- land. Het liet een grote variatie zien van de implementatie van de uitwendige versie tussen verschillende regio’s. Daarom hebben we ook een kosten- effectiviteitsanalyse verrichthoofdstuk ( 11) waarin we de resultaten van de nulmeting uit de verschillende clustershoofdstuk ( 9) hebben vergeleken met de implementatie na toepassing van de strategieën (hoofdstuk 10). De resultaten wijzen in de richting van een potentieel effect van de gecombineerde cliënt en zorgverlener strategie (toename van 7% op de implementatie van de uitwendige versie, range tussen de clusters 3-44%). De extra kosten om één extra versiepoging te bewerkstelligen (ICER;) bedroeg in de gecombineerde strategie €1.353.

In hoofdstuk 12 is beschreven hoe alle deelnemers in het cluster gerandomiseerde onderzoek zijn bevallen. De resultaten laten zien dat in de groep vrouwen met een versiepoging significant minder keizersneden plaatsvinden ten opzichte van de groep vrouwen bij wie geen versiepoging werd verricht. Na een succesvolle versie hebben vrouwen minder kans op een vaginale kunstverlossing maar vinden er meer keizersneden plaats ten opzichte van vrouwen met een spontane hoofdligging tijdens de geboorte.

285 Chapter 15 Nederlandse samenvatting

In hoofdstuk 13 is gekeken naar het effect van de implementatiestrategieën op het kennisniveau en de decisional conflict scores bij de vrouwen die hebben deelgenomen aan het cluster gerandomiseerde onderzoek. Het kennisniveau bij aanvang van de studie bleek erg hoog en er was nauwelijks sprake van twijfel over de gemaakte keuze. De implementatiestrategieën hadden geen significante invloed op deze uitkomsten.

ALGEMENE BESCHOUWING

In dit proefschrift is de suboptimale implementatie van de uitwendige versie in Nederland systematisch en grondig onderzocht. Wij presenteerden een gedetailleerd overzicht van de implementatie van de uitwendige versie en kwantificeerden de oorzaken van suboptimale implementatie in een groot cohort van clusters, representatief voor Nederland. We identificeerden -bev orderende en belemmerende factoren voor de uitwendige versie op het niveau van de patiënt en op het niveau van de zorgverlener. Op basis van deze kennis ontwikkelden we implementatiestrategieën, welke we systematisch evalueerden in een cluster gerandomiseerd onderzoek met een facto- rieel design. Hierdoor waren we in staat om het effect van beide strategieën afzonderlijk en gecombineerd te onderzoeken. Naast de primaire uitkomstmaten voor de implementatie van de uitwendige versie, hebben we ook gekeken naar de adherentie van de zorgverleners aan de aangeboden strate- gieën. Tevens hebben we de kosteneffectiviteit in ogenschouw genomen. De evaluatie van de implementatiestrategieën in het cluster gerandomiseerde onderzoek liet geen significant effect zien. Dit kan worden verklaard door verschillende factoren, welke we hieronder zullen bediscussiëren.

Implementatie van de uitwendige versie bij aanvang van het onderzoek

De resultaten van de retrospectieve nulmeting (hoofdstuk 9) en het cluster gerandomiseerde onderzoek (hoofdstuk 10) toonden een gemiddeld hoge implementatiegraad van de uitwendige versie (72%), echter met grote verschillen tussen de clusters (8-93%). Dit percentage en de bijbehorende variatie is een van de belangrijkste nevenbevindingen van onze studie. Besta- ande publicaties ten tijde van het schrijven van het studieprotocol, gaven een schatting van de implementatiegraad van 50%. Hierdoor hebben we de

286 Chapter 15 Nederlandse samenvatting implementatiegraad bij aanvang van de studie onderschat en de mogelijke verbetering, van 50% naar 80% te optimistisch ingeschat. We hebben gean- ticipeerd op een mogelijke variabiliteit in implementatiegraad bij de start van de studie door te stratificeren op de aanwezigheid van een versiespreekuur binnen de clusters; indien een versiespreekuur is opgezet binnen een cluster, is het aannemelijk dat deze clusters een hogere implementatiegraad hebben, ten opzichte van die clusters die de zorg rondom stuitliggingen en de uitwendige versie niet hebben gestructureerd. Echter, in een gerandomiseerd onderzoek met slechts een klein aantal randomisaties (25 clusters in onze studie) kan zelfs een kleine scheve verdeling van een of twee clusters met een hogere implementatiegraad bij de nulmeting een aanzienlijke dis- balans veroorzaken, welke de resultaten beïnvloed.

Indien er meer variabiliteit is in nulmeting tussen clusters, is er een groter verschil in uitkomstmaat nodig om een statistisch significant effect aan te kunnen tonen. De grote variabiliteit in tussen de clusters kan dus een verklar- ende factor zijn voor het feit dat er geen verschil is gevonden tussen de verschillende implementatiestrategieën in het cluster gerandomiseerde onderzoek. Analyse van de data in een kosteneffectiviteit studie, waarbij rekening werd gehouden met de verschillen in nulmeting binnen de clusters door een vergelijking te maken van de verbetering binnen de clusters in plaats van de absolute implementatiegraad na implementatie van de strategieën, liet een potentieel effect zien van de gecombineerde cliënt en zorgverlener strategie.

Gezamenlijke besluitvorming ten aanzien van uitwendige versie

Tijdens de determinantenanalyse naar bevorderende en belemmerende factoren voor de uitwendige versie, kwam een belangrijk punt naar voren: de ogenschijnlijke tegenstrijdigheid tussen het optimaliseren van het versieper- centage door de behandeling aan te bevelen (zoals geadviseerd wordt in de richtlijn) en zogenaamde gezamenlijke besluitvorming. Dit laatste houdt in dat de zwangere vrouwen begeleid worden in het maken van een keuze door objectieve informatievoorziening over voor- en nadelen, mogelijke risico’s en te verwachten uitkomsten. De zwangere dient te worden begeleid in het verhelderen van haar eigen normen en waarden en geholpen te worden om op basis daarvan een keuze te maken. Daarom dient de vraag zich aan of een optimale implementatie (100%) van de uitwendige versie nagestreefd moet

287 Chapter 15 Nederlandse samenvatting

worden. Dit is de beste doelstelling met het oog op het verbeteren van de perinatale uitkomsten voor moeder en kind, en dus vanuit het oogpunt van de arts en misschien ook de maatschappij. Echter laat dit geen ruimte over voor de keuze van de individuele patiënt.

Om het percentage vrouwen dat opteert voor een versie te verbeteren, hebben we de cliënt strategie ontwikkelt op basis van de gegevens verworven met het onderzoek uit hoofdstuk 6, 7 en 8. De zwangere vrouwen gaven in de interviews te kennen dat er behoefte was aan eerdere informatievoorziening dan algemeen gebruikelijk. Daarom hebben we in de strategie de zorgverleners verzocht om de informatiefolder en decision aid vanaf 32 weken zwangerschapsduur te verstrekken indien er sprake was van een stuitligging. Het verstrekken van een decision aid, leidt niet noodzakelijkerwijs tot een hogere aantal versiepogingen. Het is wel bewezen dat een decision aid een nuttig middel is om patiënten uniforme informatie te verstrekken en een positief effect heeft op het maken van een geïnformeerde keuze door de patiënten.

Effectiviteit van de implementatie strategieën

Ten gevolgen van de beperkte tijd beschikbaar voor ons onderzoek en beperkte financiën, hebben we enkele concessies moeten doen bij de ontwikkeling van de implementatie strategieën en de introductie van de strategieën in de dagelijkse praktijk. Enkele bevorderende en belemmerende factoren die naar voren kwamen in de determinantenanalyse konden we hierdoor niet opne- men in de strategieën. Dit heeft het potentiele effect van deze strategieën mogelijk beïnvloed. Zo zijn factoren gerelateerd aan de sociaal politieke context moeilijk op korte termijn te beïnvloeden.1 Zo werd tijdens de focusgroep bijeenkomsten gesproken over het belang van een transparant beleid binnen een regio ten aanzien van de informatie voorziening, counseling, verwijzing en de eigenlijke versiepoging. Het bestaan van een regioprotocol werd als een bevorderende factor gezien. Wij waren niet in de mogelijkheid om zo een protocol te ontwikkelen voor alle deelnemende clusters, laat staan om het regionale zorgproces om te vormen. We hebben wel aandacht aan dit onderwerp besteed in de zorgverlener strategie, tijdens de bijscholingsdag. We beoogden hiermee het inzicht van de zorgverleners in hun eigen zorgproces te vergroten door die punten bloot te leggen waar verbetering mogelijk

288 Chapter 15 Nederlandse samenvatting was. We hebben twee voorbeelden van goedlopende versiespreekuren uit het land besproken ter inspiratie voor die clusters die nog geen regionaal versieprotocol hadden. Samenwerking tussen zorgverleners binnen een cluster gaat echter dieper dan een enkel onderwerp en ligt ingebed in er- varingen uit het verleden. Een meer intensieve begeleiding van clusters om samenwerking tussen de zorgverleners te stimuleren, zou effectiever kunnen zijn geweest.

Tot uitvoer brengen van de implementatie strategie

De resultaten van de intention-to-treat analyse van ons cluster gerandomiseerd onderzoek geeft een reëel beeld van de dagelijkse praktijk. Er is uitge- breid empirisch bewijs dat de mate waarin richtlijnen in de praktijk worden opgevolgd, de uitkomsten van de patiënten beïnvloed. En dit geldt ook voor implementatie strategieën.2 Om te beoordelen of een implementatiestrategie effectief is, moet er ook worden gekeken naar de mate waarin deze tot uitvoer is gebracht (adherentie) door de zorgverleners in de dagelijkse praktijk. Dit om te voorkomen dat te snel wordt geconcludeerd dat een strategie niet effectief is, terwijl het eigenlijk ontbreekt aan een goede implementatie van de strategie. Er is een hiaat in publicaties van implementatie strategieën met niet significante resultaten, wat potentieel duidt op een publicatie bias ten gunste van die studies met significante resultaten. Dit belemmerd ver- spreiding van kennis over ontwikkelde implementatie strategieën die van zichzelf potentieel effectief zijn, maar wat niet naar voren kwam in gerandomiseerd onderzoek door invloed van andere factoren. Daardoor kunnen andere onderzoeksgroepen niet hun voordeel doen van bestaande kennis of leren van andermans onvolkomenheden.

Met betrekking tot het probleem van suboptimale adherentie aan onze implementatie strategieën, willen we twee punten aandragen.

Allereerst is dit mogelijk deels veroorzaakt doordat we slechts één introductie sessie (cliëntstrategie) en één cursusdag (zorgverlener strategie) in plaats van meerdere opeenvolgende sessies konden organiseren. Dit had tot gevolg dat sommige zorgverleners niet in de gelegenheid waren om de bijeenkomst bij te wonen, met een mogelijk lager effect van de implementatie strategie op het implementatiegraad tot gevolg.

289 Chapter 15 Nederlandse samenvatting

Ten tweede kost het tijd om implementatie strategieën systematische te in- troduceren en zorgverleners terugkoppeling te geven over de adherentie. Een soortgelijk implementatie onderzoek heeft laten zien dat het tot twee jaar kan duren om een optimaal effect van de implementatie strategie te bereiken.3 Idealiter wordt dus eerst de implementatie strategie optimaal geïntroduceerd bij alle zorgverleners, inclusief terugkoppeling van hun adherentie aan de strategie, en om hun zelfvertrouwen door de ervaring met de strategie in de dagelijkse praktijk te vergroten, alvorens de uitkomstmaten (adherentie en implementatiegraad) worden gemeten.

In het cluster gerandomiseerde onderzoek kreeg de meerderheid van de zwangere vrouwen die geen versie ondergingen, deze niet aangeboden door de zorgverlener, ondanks de moeite die is gedaan om bestaande belemmerende factoren weg te nemen en bevorderende factoren te stimuleren. Het is natuurlijk belangrijk om alle praktische en methodologische discussie- punten van onze implementatie strategieën voor het voetlicht te brengen, maar uiteindelijk blijft ook de vraag over waarom het niet vanzelfsprekend is voor zorgverleners om volgens de meest recente richtlijnen te werken. Natuurlijk kan elke zorgverlener, indien weloverwogen onderbouwd, van de richtlijn afwijken. Echter met betrekking tot de uitwendige versie was er bij 40% van de vrouwen die geen versiepoging ondergingen niets vermeldt in de patiënten status over (de counseling voor) deze behandeling. In de introductie beschreven we dat richtlijn niet automatisch in gebruik worden genomen door alle zorgverleners. We willen de stelling poneren dat elke zorgverlener het als een kerntaak moet zien om volgens de richtlijn te werken om zo wetenschappelijk onderbouwde zorg aan de patiënten te leveren tijdens de dagelijkse praktijkvoering. De significante verschillen tussen clusters in implementatiegraad van de uitwendige versie laat zien dat goede adherentie in de dagelijkse praktijk mogelijk is en dat een slechte implementatiegraad onacceptabel is.

290 Chapter 15 Nederlandse samenvatting

IMPLICATIES VOOR DE DAGELIJKSE PRAKTIJK

Met het oog op de grote verschillen in implementatiegraad tussen clusters met betrekking tot de uitwendige versie, willen wij sterk aanbevelen dat de counseling voor deze behandeling en de versiepogingen zelf (inclusief slagingspercentages en complicaties) worden geregistreerd in een landelijk reg- istratiesysteem (LVR 1 en 2). Hierdoor wordt het mogelijk om de zorg voor a terme stuitligging systematisch en op jaarlijkse basis te evalueren. Deze informatie kan gebruikt worden voor terugkoppeling aan de clusters en begeleiding van die clusters met een lage implementatiegraad. Het effect van deze begeleiding zal duidelijk worden in de daarop volgende jaren, via de standaard registratie in de landelijke systemen. Hierdoor wordt de cirkel van Deming (plan-do-check-act) vervolmaakt wat leidt tot een continue verbetering van de zorg.

Evaluatie van zorg voor stuitligging uit deze landelijke registratie moet zich richten op zowel de implementatiegraad als de slagingspercentages van de uitwendige versie om een optimale daling van het aantal stuitliggingen bij de geboorte te bewerkstelligen. Op basis van de gegevens van ons cluster gerandomiseerde onderzoek en de perinatale registratie Nederland 2012 hebben we berekend wat de maximale winst is die nog te behalen is (met een optimale implementatie en optimalisatie van de slagingskansen). De data in dit proefschrift laten zien dat 7% van de stuitliggingen vanaf 36 weken zwangerschapsduur een absolute contra-indicatie hebben voor een uitwendige versie, conform de huidige richtlijnen. Het gemiddelde percentage vrouwen tussen de clusters dat af ziet van de behandeling is 10%. Dus een optimale implementatie ligt ongeveer rond 83% van alle vrouwen met een stuitligging vanaf 36 weken zwangerschapsduur. Momenteel ligt dit percentage echter op 72%. De optimalisatie van de implementatiegraad kan leiden tot een vermindering van het aantal stuitliggingen bij de geboorte van 4.821 naar 4.543 per jaar (5.8%).

Enkele clusters met een goede implementatiegraad van de uitwendige versie behaalden slagingspercentages van 60% of hoger. Dit waren clusters met versiespreekuren en het succespercentage was dus niet vertroebeld door een selectie van patiënten aangezien de implementatiegraad optimaal was. Indien deze slagingspercentages landelijk gehaald kunnen worden, daalt het aantal stuitliggingen bij geboorte van 4.821 naar 3.332 per jaar (30.9%). Mo- gelijk strategieën om deze slagingspercentages te bewerkstelligen zijn het

291 Chapter 15 Nederlandse samenvatting

organiseren van vergelijkbare versiespreekuren in die clusters die hier nog niet over beschikken, of het verwijzen van de zwangere vrouwen met een stuitligging naar die clusters met goed presterende versiespreekuren. Een derde optie is het verder evalueren van het gebruik van uterus relaxantia tijdens de uitwendige versie om de slagingskans te vergroten en het vervolgens optimaal implementeren van deze medicamenteuze behandeling in de dagelijkse praktijk.4

IMPLICATIE VOOR VERDER ONDERZOEK

Het onderwerp van dit proefschrift de implementatie van een wetenschappelijk bewezen effectieve behandeling, wat de laatste stap is in hetopti- maliseren van het zorgproces. Toch blijven er nog steeds enkele vragen over voor verder onderzoek:

1. Het optimaliseren van de implementatiegraad

Met het gepresenteerde onderzoek hebben we gedetailleerd inzicht gekregen in de implementatie van de uitwendige versie in Neder- land. Registratie in landelijke databases kan dit overzicht alleen maar verbeteren. De combinatie van de ontwikkelde cliënt- en zorgverlener strategie is potentieel effectief op basis van de uitkomsten van de kosteneffectiviteit analyse. Daarom kan gebruik van deze- strat egie geëvalueerd worden in die clusters met een lage implementatiegraad.

2. Effectiviteit van de uitwendige versie

Optimalisatie van de slagingskans van de uitwendige versie dient ook aandacht te krijgen in toekomstig onderzoek. De hoogste slagingspercentages werden bereikt binnen die clusters met een versiespreekuur. Toekomstig onderzoek moet nagaan waarom deze versiespreekuren zo succesvol zijn (bijvoorbeeld de expertise van zorgverleners, gebruik van uterus relaxantia of een combinatie van beide?).

292 Chapter 15 Nederlandse samenvatting

3. Diagnosticeren van stuitliggingen

Uit ons onderzoek bleek dat bij 3.5% van de a terme stuitpartussen, de stuitligging pas na aanvang van de bevalling werd gediagnos- ticeerd. Dit komt overeen met eerder onderzoek, waarin werd geconcludeerd dat uitwendig bimanueel onderzoek niet sensitief genoeg is voor de detectie en tijdige behandeling van liggingsafwijkingen.5 Een kosteneffectiviteit analyse van een liggingsecho om liggingsafwijkingen in het derde trimester op te sporen, moet derhalve worden verricht.

293 Chapter 15 Nederlandse samenvatting

REFERENTIELIJST

2 Durlak J a, DuPre EP. Implementation matters: a review of research on the influence of implementation on program outcomes and the factors affecting implementation. Am J Community Psychol [Inter net]. 2008 Jun [cited 2014 May 12]; 41: 327–350. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18322790

3 Fleuren M, Dusseldorp E, van den Bergh S, Vlek H, Wildschut J, van den Akker E, et al. Implementation of a shared care guideline for back pain: effect on unnecessary referrals. Int J Qual Health Care [Internet]. 2010 Oct; 22: 415–420. Available from: http://www. pubmedcentral.nih.gov/articlerender.fcgi?artid=2939752&tool=pmc entrez&rendertype=abstract

4 Cluver C, Gj H, Gml G, Sinclair M. Interventions for helping to turn term breech babies to head first presentation when using external cephalic version ( Review ). 2012;

5 Nassar N, Robberts CL, Cameron CA, Olive EC. Diagnostic accuracy of clinical examinations for detection of non-cephalic presentation in late pregnancy: cross sectional analytic study.BMJ, 2006 Seo; 16; 333(7568):578-80

294

16 Dankwoord Ageeth Rosman

297 Chapter 16 Dankwoord Ageeth Rosman

Dankwoord

Vijf jaar geleden begon ik aan een traject waarvan vandaag de afsluiting heeft plaatsgevonden. Een intensieve periode die ik niet zonder de steun van vele mensen heb kunnen volbrengen.

Ik licht er een aantal uit, wetende dat ik daarmee vele anderen passeer voor wie mijn dank niet minder groot is dan voor hen die ik er uit licht. Dus voor iedereen die niet met naam en toenaam wordt genoemd, mijn dank voor jullie is niet minder groot dan voor degene die wel genoemd worden.

Allereerst dank ik alle zwangere vrouwen die bereid waren deel te nemen aan de studie. Zonder hen was het resultaat dat er nu ligt nooit mogelijk geweest.

Ik dank alle verloskundigen, gynaecologen en arts-assistenten, werkzaam in de clusters die meegedaan hebben aan de studie voor hun inzet en doorzettingsvermogen. Meedoen aan een implementatiestudie is geen gemakkelijke opgave. Dank!

De leden van de promotie commissie: Prof. dr. C.J.M. de Groot, Prof. dr. M.L. Essink – Bot, Dr. E. de Miranda, en Dr. M. Zegers wil ik bedanken voor het beoordelen van het manuscript en het plaatsnemen in de commissie. I would like to thank Prof. dr. D. Bick for being here today. You are an expert in the field and it is a honour that you are one of my opponents.

Ik dank mijn promotoren, Prof. dr. J.A.M. van der Post en Prof. dr. B.W.J. Mol voor hun begeleiding.

Beste Joris. Mijn start in het AMC was niet gemakkelijk. Het kostte flink wat moeite om mijn plek en mijn weg in de organisatie te vinden. Ik moest wennen aan de gelaagdheid van de organisatie en het “Big Brother is watching you”- principe. Daar waar ik af en toe zeer twijfelde of ik wel een echte AMC- er zou worden, bleef jij vertrouwen houden dat het goed zou komen. Ik dank je voor de begeleiding in de afgelopen jaren en zeker bij de eindspurt van dit proefschrift. Ik weet niet wat de toekomst zal brengen, zeker nu er ook een andere professor op mijn pad is gekomen …. zal ik een AMC-er blijven of word ik een Erasmiaanse?

Beste Ben Willem. Tja, wat kan ik zeggen wat nog origineel is en niet door

298 Chapter 16 Dankwoord Ageeth Rosman vele andere promovendi, die mij voor zijn gegaan op jouw ongelooflijke lange lijst van promovendi, al is gezegd. Ik ben je dankbaar voor je begeleiding, ik bewonder je helikopterview en de snelheid waarmee je weet te antwoor- den, zelfs nu je aan de andere kant van de wereld woont. In het begin van ons gezamenlijke traject voelde ik me vaak een derde rad aan de wagen en had moeite om in te schatten wat jij nu werkelijk van mij dacht. Ik voelde me vaak overbodig. In de loop van de tijd veranderde dat en zeker nadat onze wegen ook op het hockeyveld gingen kruisen, zag ik ook een andere kant van je en werd je voor mij wat gemakkelijker te lezen. Ik hoop dat je heel gelukkig bent (en wordt) in Australië en ik wens je het allerbeste toe, samen met Yasmine en Nour. Wie weet waar we elkaar ooit weer tegen zullen komen.

Ik dank mijn copromotoren, dr. M. Kok en dr. M.A.H. Fleuren voor hun begeleiding en inzet gedurende het gehele promotietraject.

Beste Marjolein. Jouw eerste promovenda, samen met Floortje en dan ook nog eens een verloskundige. Op verloskundig gebied heel veel meer jaren ervaring en heel veel meer grijze haren, maar op het wetenschappelijk gebied veel minder. Het was wennen voor jou en voor mij en samen moesten wij zoeken naar de juiste weg. Lastig, vooral omdat onze achtergrond zoveel verschilde. Toch denk ik dat we uiteindelijk de juiste modus gevonden hebben en dank ik je voor je begeleiding in de afgelopen jaren. Dank!

Beste Margot. Een implementatiestrateeg van buiten de organisatie leverde in het begin wat onwennige situaties op. Gedurende het traject leerden we elkaar beter kennen en elkaars kennis en kunde beter begrijpen. Ik ben je dankbaar voor alle tijd die je voor en in mij hebt geïnvesteerd. Ik ben je dankbaar omdat jij in de eindfase vaak als eerste liet weten dat het goed ging en dat het allemaal goed ging komen. Ik was blij met je aanwezigheid op het ICM in Praag, waar ik voor de eerste keer een groot deel van het proefschrift voor een internationaal publiek presenteerde en kon laten zien wat ik allemaal van jouw kennis heb opgestoken. Dank!

Verder wil ik graag Marlies Rijnders en Antje Beuckens en Brent Opmeer (leden van de projectgroep) bedanken voor hun bijdrage aan het onderzoek. Simone Kuppens, Riet van der Donk, Olga Kok, Jan Molkenboer en Marijke Hoiting bedank ik voor hun bijdrage aan de ontwikkeling van de cursusdag in het kader van de cluster RCT. 299 Chapter 16 Dankwoord Ageeth Rosman

Antje en Marlies, jullie rol is meer op de achtergrond geweest maar daarom niet minder belangrijk. Dank voor het meelezen en meedenken tijdens het gehele traject.

Brent, ik ervaar jouw steun bij alle statistiek dat in dit proefschrift naar voren is gekomen, als ongekend groot. Ik hoop dat je geen punthoofd van mij hebt gekregen.

Debbie Bax en Marjan Du Prie, jullie zijn van ontelbare grote waarde voor de afdeling maar zeker voor ons als promovenda. Wat mag men blij zijn, met zulke parels als jullie!

Ik dank al mijn collega verloskundigen in het AMC voor hun luisterend oor, de ruimte die ik heb gekregen om dit promotietraject te kunnen doen, hun warme belangstelling en de medewerking bij het includeren van alle stuiten binnen het AMC. Ik dank alle verpleegkundigen, arts-assistenten en dokter- sassistenten voor hun steun en belangstelling en speciaal voor Joost, mijn “zoon”, ik zal nooooooooit meer zeggen dat je op iemand lijkt. Volgend jaar zit ik bij jou in de zaal, zet ‘m op.

Ik dank de Lieffies voor hun steun en luisterend oor en hun creativiteit in het vinden van een invaller in de laatste periode van mijn actieve hockeytijd als ik niet mee kon doen omdat het manuscript af moest. Ik weet niet hoe vaak ik jullie daarmee heb lastig gevallen, maar het zal zeker te vaak zijn geweest. Ik ben blij dat ik deel heb gemaakt van dit fantastische team en zoals op het tegeltje staat “Eens een Lieffie, altijd een Lieffie”. Ik ga verder als LIR maar zal zeker bij thuiswedstrijden regelmatig langs het veld staan en als er weer ruimte is in mijn hoofd, dan komt er ook wel weer tijd voor hockey en misschien mag ik dan ook weer toetreden als echte Lieffie?

En dan dank ik mijn internationale vrienden, Mercy Baafi, Laura Robbertson en Bill en Anna Folkerts.

Dear Mercy, I think we first met on the ICM in Durban in 2011 but I’m not sure anymore. However, we really met on the ACM in Hobart in 2013, where I was invited to present some parts of this thesis. I’m so glad I met you! I enjoy your enthusiasm, your support, your knowledge and your hospitality. We had a lot of pleasant conversations and I learnt a lot of you. We share the love for the Ghanaian people. I was very pleased to visit you at home with Gerrit and to meet your husband, Ernest. I will never forgot the blow

300 Chapter 16 Dankwoord Ageeth Rosman hole you showed us and the hurry we had to eat as the train will never wait… we will come back to Australia and to Wollongong to have a nice and relaxed diner together and in the meantime, you, Ernest, Wollongong and Australia are in our hearts. Thanks!

Dear Laura, my soul mate. Your sister wrote, we are looking like twins and I think she is right. I know what it is to be a twin and I can tell you, it feels okay! I really like you very much and I think we should keep on dreaming of our own breech clinic in Perth. You do a wonderful job and I’m looking forward to meet you or in the Netherlands once or in Perth, as we would come back to Australia once and visit your part of this wonderful continent. I wish I had a money tree in my backyard..

Dear Anna and Bill. Dear Anna, I met you on the ACM in Hobart and you told me about your Dutch husband. I was pleased that you invited me and my husband to your home. It was really close to Hobart and we were lucky as we did not organise our trip very well. If I didn’t met you, we should be sleep as tramps as we booked one night to short.. We were very pleased that we could also meet him and had a delicious diner together at your home. We are looking forward to meet you in our home in August and we enjoy that you visit Almere. We will take care of you and show you the most wonderful parts of our home town. We told our kids and family about you and we are so pleased that you come to the Netherlands.

Een woord van dank ook aan alle Westzeedijkers. Dank voor jullie belangstelling voor mijn promotietraject en jullie warme welkom als ‘Amsterdammer’ binnen Rotterdam. Ik voel me meer dan welkom in de groep en misschien komt dat ook wel omdat ik een kind ben uit een van de HP4all gebieden en daardoor precies begrijp en herken wat HP4all omvat en wil bereiken.

Als laatste ga ik traditiegetrouw over tot dankwoorden aan mijn familie en mijn maatje in deze strijd, Floortje.

Lieve Floortje. Ik kan je moeder zijn, zoveel verschillen wij in leeftijd. Geluk- kig heeft dit verschil nooit tussen ons in gestaan, want waar ik zoveel meer levensjaren inbracht, bracht jij zoveel meer wetenschap met je mee, waar ik dankbaar gebruik van mocht maken. Ik denk dat wij samen het voorbeeld zijn voor de samenwerking tussen verloskundigen en gynaecologen en ik denk dat wij, met de totstandkoming van dit proefschrift, laten zien, dat

301 Chapter 16 Dankwoord Ageeth Rosman

ook op het gebied van de wetenschap, verloskundigen en gynaecologen één moeten en kunnen zijn. Ik bewonder je zo enorm en heb je zien groeien als moeder van twee prachtige kinderen. Onze wegen zullen gaan scheiden na de dag van vandaag en morgen, we zullen beiden onze carrières weer een eigen invulling gaan geven, maar ik weet zeker dat jij een hele goede gynaecoloog gaat worden. Het ga je goed!

Papa en mama, dank voor jullie inzet, belangstelling, invalbeurten, oppassen en alles wat jullie nog meer voor mij hebben gedaan. In mooie en moeilijke momenten is het goed dat we elkaar hebben, dat we elkaar door en door kennen en dat we weten waar we aan toe zijn. We zijn en blijven “Grun- nigers” en doen dat zonder al te veel woorden en opsmuk, maar met kleine gebaren, weten we waar we aan toe zijn. Dank voor alles en ik ben blij dat jullie ook op deze dag naast mij staan.

Lieve broers en zussen. Ik voel me rijk met zoveel broers en zussen. We verschillen allen van dag tot dag, maar we hebben een gemeenschappelijk iets en dat is dat we broers en zussen van elkaar zijn. Ik dank ieder van jullie voor jullie belangstelling, medeleven, medewerking aan de totstandkoming van dit proefschrift. Ik ben blij dat ik nu meer tijd weer voor jullie krijg en spreek de hoop uit dat we elkaar wat vaker gaan zien en spreken, want dat is de afgelopen tijd een ondergeschoven kindje geweest.

Lieve “koude kant”. Ook jullie behoren tot onze familie en behoren tot de mensen die ik persoonlijk wil danken. Tot op de dag van vandaag doet het zeer dat we vroegtijdig afscheid moesten nemen van Pedro, maar ik weet zeker dat hij ook vandaag over onze schouders meekijkt. Ik dank jullie voor jullie begrip en ik hoop dat wij elkaar ook weer meer zullen zien.

Lieve neven en nichten, fijn dat jullie vandaag ook allemaal aanwezig zijn. Ik wilde jullie allen bij naam noemen, maar dan wordt het dankwoord onno- emelijk lang dus doe ik het maar zo. Jullie hebben allen een goed verstand meegekregen waarvan ik hoop dat jullie (liefst allemaal) in mijn voetsporen zullen treden en dat ik nog heel vaak bij een promotie aanwezig mag zijn. Zet ‘m op. Jullie kunnen het allemaal!

Een heel speciaal woord van dank is er nog voor Frank voor het kritisch lezen van onze Engelse teksten en het omzetten van mijn ‘Rudi Carell Engels’ in de juiste bewoordingen en Frank, er bestaat geen andere versie van de uitwendige versie. 302 Chapter 16 Dankwoord Ageeth Rosman

Lieve ma Schonewille, Tini en Bert. Dank voor jullie aanwezigheid op deze speciale dag. Ik weet zeker dat pa ook meekijkt en trots geweest zou zijn op zijn schoondochter. Ik ben blij dat jullie vanuit het zuiden naar Amster- dam zijn gekomen en hoop dat wij elkaar in de toekomst weer vaker zullen spreken en ontmoeten.

En dan mijn paranimfen, Marit en Lydia. Samen vormen wij een stevig blok tegenover de opponenten en ik voel me gesteund met twee aankomende dokters aan mijn zijde. Ik ben blij met de keuze die jullie gemaakt hebben en leef, lees en leer met jullie mee. Maak jullie dromen waar, zoals jullie ooit mijn grootste (wens)droom hebben waargemaakt………..moeder worden van een tweeling!

Rozemarijn, mijn jongste dochter. Op je eigen eigenwijze manier deel van ons gezin. Het liefst zou je ook als paranimf naast mij staan, maar dat gaat helaas niet en dus heb je jezelf benoemd tot ceremoniemeester. Een taak die je met je hele ziel en zaligheid hebt vervuld.

Samen gaan we er voor zorgen dat jij en ik het VWO examen (nog een keer) gaan halen en dat ook jij je droom waar kunt maken. Je kunt het en ik zal je ondersteunen daar waar ik kan en jij het toestaat.

Michel, onze eerste mannelijke aanhang binnen het gezin en dat maakt je positie uniek, maar houdt ook in dat je de kolen uit het vuur moet halen. Je kiest er één, je krijgt er vier! Je bent een uitstekend maatje voor Marit. We zijn blij met je en voel je welkom binnen ons gezin.

Als laatste, Gerrit, mijn maatje. Toen wij elkaar op een speciale manier leerden kennen, waren we beiden op zoek naar een maatje. In de loop der jaren heeft zich dat steeds meer ontwikkeld en ik denk dat wij nu, naar 24 jaar kunnen stellen dat we echt elkaars maatje zijn.

Dank voor al je steun in de afgelopen periode. Dat is niet gemakkelijk geweest want vaak kwam je op de laatste plaats te staan. Ik kijk uit naar de periode dat wij echt veel tijd voor elkaar hebben en we kunnen gaan reizen. Dank en ik hou van je.

303

17 Portfolio Ageeth Rosman

305 Chapter 17 Portfolio Ageeth Rosman

PORTFOLIO AGEETH ROSMAN

PhD period (June 2009-August 2013)

PhD supervisor: Prof. dr. J.A.M. van der Post

PhD training

Courses Academic Medical Center, Amsterdam

- Practical biostatistics (2010; 1.1)

- Clinical Epidemiology (2010; 1.1)

- Clinical data management (2011; 0.3)

- Advanced Biostatistics (2011; 1.1)

- Systematic Reviews (2010; 0.3)

- Scientific writing in English for publication (2010; 1,5)

- BROK (‘Basiscursus Regelgeving Klinisch Onderzoek’) (2010; 0.9)

Courses VU University Medical Center, Amsterdam

- Multilevel analyses (2012; 1.1)

Seminars, workshops and master classes

- Weekly department seminars dept. Obstetrics & Gynecology AMC

- Weekly department lunch meeting dept. Obstetrics & Gynecology ErasmusMC

306 Chapter 17 Portfolio Ageeth Rosman

Presentations

Oral presentations

- Contraindications of external cephalic version in breech presentation: a systematic review- ICM (International Congress of Midwives, Durban 2011)

- Improving the implementation of external cephalic version in breech position – WVGC (Women Visiting Gynecology Club England, Amsterdam 2011)

- Patients’ and professionals’ barriers and facilitators to external cephalic version for breech presentation at term, a qualitative analysis – (Kennispoortsymposium Utrecht, 2012)

- Implementatie van de uitwendige versie in Nederland – ( Gynaecongres Den Bosch 2013)

- Implementation of external cephalic version: a multicentre cluster randomised controlled trial –(Australian Colleges of Midwives, Hobart 2013, Australia)

- Impact of caesarean section rate due to term breech presentation on the outcome in subsequent pregnancies- (International Congress of Midwives Prague, 2014)

- Implementation of external cephalic version: a multicentre cluster randomised controlled trial – (International Congress of Midwives Prague, 2014)

- Improving care for breech presentation at term – (Erasmus MC, dept. Obstetrics & Gynae- cology, research meeting, 2014)

- Decisional conflict in women opting for External Cephalic Version –(Kennispoortsympo- sium Utrecht 2014)

Poster presentations

- Facilitators and barriers to external cephalic version for breech presentation at term among health care providers in the Netherlands: a quantitative study – (Kennispoortsym- posium Utrecht 2012)

- Contraindications of external cephalic version in breech position: a systematic review- (Australian Colleges of Midwives, Hobart 2013, Australia)

307 Chapter 17 Portfolio Ageeth Rosman

(Inter)national conferences

- ICM, Durban, 2011

- Kennispoortsymposium, Utrecht 2012

- 43ste Gynaecongres, Den Bosch, 2013

- ACM, Hobart Tasmania, 2013

- Kennispoortsymposium, Utrecht 2014

- ICM, Prague, 2014

Teaching, supervising

- L.M.L. de Baaij, M.I. Gerritsen, M.L.Vink (2011), Topic: Refer wise: A study on the difference in references within the Dutch obstetric health care between clients with Caucasian ethnicity and clients with Moroccan or Turkish ethnicity.

- Suzanne van den Berg (2012). Topic: Baseline implementation of ECV, a retrospective cohort study.

- Vincent Westendorp (2014). Topic: What skills use health care providers in communication with patients with low health?

- Lecture; ECV and scientific knowledge. ECV training course for midwifes, Andata coaching & training. November 2013 and March 2014.

308

18 Publications Ageeth Rosman

311 Chapter 18 Publications Ageeth Rosman

List of publications A.N. Rosman

1. Facilitators and barriers to external cephalic version for breech presentation at term among health care providers in the Netherlands: a quantitative analysis. Rosman AN, Vlemmix F, Beuckens A, Rijnders ME, Opmeer BC, Mol BW, Kok M, Fleuren MA. Midwifery. 2014 Mar;30(3):145-50.

2. Patients’ and professionals’ barriers and facilitators to external cephalic version for breech presentation at term, a qualitative analysis in the Netherlands. Rosman AN, Vlemmix F, Fleuren MA, Rijnders ME, Beuckens A, Opmeer BC, Mol BW, van Zwieten MC, Kok M. Midwifery. 2014 Mar;30(3):324-30.

3. Contraindications for external cephalic version in breech position at term: a systematic review. Rosman AN, Guijt A, Vlemmix F, Rijnders M, Mol BW, Kok M. Acta Obstet Gynecol Scand. 2013 Feb;92(2):137- 42.

4. Decision aids to improve informed decision-making in pregnancy care: a systematic review. Vlemmix F, Warendorf JK, Rosman AN, Kok M, Mol BW, Morris JM, Nassar N. BJOG. 2013 Feb;120(3):257-66.

5. Implementation of the external cephalic version in breech delivery. Dutch national implementation study of external cephalic version. Vlemmix F, Rosman AN, Fleuren MA, Rijnders ME, Beuckens A, Haak MC, Akerboom BM, Bais JM, Kuppens SM, Papatsonis DN, Opmeer BC, van der Post JA, Mol BW, Kok M. BMC Pregnancy Childbirth. 2010 May 10;10:20.

6. Implementation of external cephalic version in the Netherlands: a retrospective cohort study. Vlemmix, Floortje; Rosman, Ageeth; te Hoven, Susan; Van den Berg, Suzanne; Fleuren, Margot; Rijnders, Marlies; Beuckens, Antje; Opmeer, Brent; Mol, Ben; Kok, Marjolein. Birth 2014 March 13.

312 Chapter 18 Publications Ageeth Rosman

7. Term breech deliveries in the Netherlands: did the increased cesarean rate affect neonatal outcome? A population based cohort study. Floortje Vlemmix, Lester Bergenhenegouwen, Jelle M. Schaaf, Sabine Ensing, Ageeth N. Rosman, Anita C.J. Ravelli, Joris A.M. van der Post, Arno Verhoeven, Gerard H. Visser, Ben Willem J. Mol, Marjolein Kok. Acta Obstetricia et Gynecologica Scandinavica (AOGS,2014)

313

19 Curriculum Vitae Ageeth Rosman

315 Chapter 19 Curriculum vitae Ageeth Rosman

Curriculum vitae Ageeth Rosman

Ageeth Nicole Rosman was born on May 14th 1964 in Groningen as first part of a twin and grew up in a family of twins and one singleton. After graduation of high school in 1984 (Atheneum, Winschoter Scholengemeenschap, Win- schoten), she studied midwifery at the “Kweekschool voor vroedvrouwen” in Heerlen. In 1987 she obtained cum laude her bachelor degree.

In 1987 she started working as a midwife in primary care in Zwolle. She worked till 2000 in primary care. Thereafter she moved to the secondary care, till 2003 in the Isala Clinics, location De Weezenlanden at Zwolle, till 2007 in the Flevoziekenhuis at Almere and from 2007 and beyond in the Academic Medical Center of Amsterdam, where she still keeps on working on the delivery ward.

In 2005 she started with her Master of Science- midwifery at the University of Amsterdam. In 2007 she obtained her master’s degree in the sub specialism policy and management. In June 2009 she started as PhD student at the department of Obstetrics and Gynecology of the Academic Medical Center in Amsterdam on a project entitled ‘implementation of external cephalic version’. Her PhD thesis was supervised by Prof. dr. J.A.M. van der Post (promotor), Prof. dr. B.W.J. Mol (promotor), dr. M. Kok (co-promotor) and dr. M.A.H. Fleuren (co-promotor).

In 2013 she was part time set to work at the Erasmus MC where she started working on the project HP4all ( a Healthy Pregnancy for all- prof. dr. E.A.P. Steegers) and is still working on this project.

Ageeth is the mother of a twin (Marit and Lydia) and a singleton (Rozemarijn) and is married to Gerrit Schonewille. They live in Almere.

316

20 Dankwoord Floortje Vlemmix

319 Chapter 20 Dankwoord Floortje Vlemmix

Dankwoord

Onderzoek doe je gelukkig niet alleen, dit proefschrift was er niet gekomen zonder de hulp van velen. Een aantal mensen wil ik graag in het bijzonder bedanken.

Prof. dr. B.W.J. Mol, beste Ben Willem. Ik blijf je altijd dankbaar voor de kans die jij me hebt gegeven om te starten als promovenda binnen de afdeling, en de mogelijkheid om dit te combineren met het ANIOS-schap. Deze combinatie maakte beide trajecten dubbel zo leerzaam en het heeft me gebracht waar ik nu sta. Met jou werken en jou zien werken heeft me een hoop gel- eerd over onderzoek doen, alles wat bij onderzoek hoort, over mezelf, over doelen stellen, plichten vervullen, kansen creëren, dromen najagen en anderen daarin meenemen. Ik wens je een mooie toekomst toe in Australië en in alles wat mag volgen. Maar bovenal hoop ik dat je een gelukkig mens bent.

Prof. dr. J.A.M. van der Post, beste Joris. Je deur staat altijd open, de tele- foon wordt altijd opgenomen; nooit was je te druk om een luisterend oor te bieden en mij van advies te voorzien. Door de zaken van een andere kant te belichten gaf je mij weer nieuwe inzichten en ideeën om verder te werken aan onderzoek en bijbehorende artikelen. Daarnaast heb je altijd oog voor de persoon en ben je niet alleen in het werk, maar ook daar buiten een inspiratie: mijn nieuwe racefiets krijgt nog veel te verduren.

Dr. M. Kok, beste Marjolein. Jij zorgde voor een warm onthaal in het AMC en een gespreid bedje dankzij alles wat jij al in gang had gezet in opvolging van je eigen proefschrift. Je was als schakel tussen promotor en promovendus van onmisbare waarde voor mij! Het was heel fijn om met jou langer dan een ’15 minute pitch’ te kunnen sparren over organisatorische en inhoudelijke zaken van het onderzoek. Ik heb er bewondering voor hoeveel ballen jij in de lucht weet te houden binnen en naast je werk. Met altijd een vrolijke lach op je gezicht!

Dr. B.C. Opmeer, beste Brent. Wat was ik blij dat ik jou had voor de brood- nodige ondersteuning voor de methodologische kanten van het onderzoek. En wat heb jij een engelengeduld….Het was heel fijn om met jou te kunnen brainstormen over de opzet van databases, analyses en artikelen. Mijn dank is groot.

320 Chapter 20 Dankwoord Floortje Vlemmix

Dr. M.A.H. Fleuren, beste Margot. In het begin waren de verhoudingen wat afstandelijk maar door de loop van het onderzoek werd jou rol steeds groter. Je hebt ons de beginselen van implementatie onderzoek bijgebracht. Je was mijn derde co-promotor en je (persoonlijke) betrokkenheid in de laatste fase hield ons op de been.

De leden van de promotie commissie: Prof. dr. C.J.M. de Groot, Prof. dr. M.L. Essink – Bot,

Dr. A. de Jonge, en Dr. M. Zegers wil ik bedanken voor het beoordelen van het manuscript en het plaatsnemen in de commissie. I would like to thank Prof. dr. D. Bick for being here today. You are an expert in the field and it is a honour that you are one of my opponents.

Verder wil ik graag Marlies Rijnders en Antje Beukens, leden van de projectgroep, bedanken voor hun bijdrage aan het onderzoek. Simone Kuppens, Riet van der Donk, Olga Kok, Jan Molkenboer en Marijke Hoiting bedank ik voor hun bijdrage aan de ontwikkeling van de cursusdag in het kader van de cluster RCT.

Alle gynaecologen, verloskundigen en research medewerkers van de deelnemende ziekenhuizen en verloskundige praktijken bedank ik voor hun inzet voor ons onderzoek. Zonder data geen proefschrift!

Veel dank aan alle gynaecologen van het Kennemer Gasthuis, en in het bijzonder dr. Lips. Beste Jos, vanaf de eerste keer dat wij elkaar spraken gaf je aan dat het in het belang van mijn opleiding was om het proefschrift zo snel mogelijk af te ronden. Dankzij het begrip en de ruimte die jullie me hebben gegeven is dit gelukt. En tegelijkertijd heb ik met jullie hulp een hele fijne start met de opleiding tot gynaecoloog kunnen maken.

Beste Janna en Susan, jullie waren mijn wetenschappelijke stage studenten. Beide intelligent, ondernemend, zelfstandig en doorzetters. Janna, wat had ik nog toe te voegen nadat ik je wegwijs had gemaakt in het schrijven van een Systematic Review en je op weg was naar dr. Nassar? Bedankt voor de leuke tijd in Sydney waar jij mij weer wegwijs maakte. Susan, wat heb jij veel kilometers afgelegd en in veel stoffige ruimtes data voor ons boven tafel gekregen. Nooit was iets te veel gevraagd van je. Bedankt voor je enorme betrokkenheid bij het onderzoek. Ik wens jullie beide een mooie toekomst toe binnen de geneeskunde!

321 Chapter 20 Dankwoord Floortje Vlemmix

Lieve Tammo, mijn leermeester! Dankzij jou is het onderwerp van mijn proefschrift voor mij gaan leven. Wat was het leuk om samen de versiepogingen te doen en de euforie te delen als het weer gelukt was.

Dank aan alle collega onderzoekers en assistenten van de afdeling Ver- loskunde & Gynaecologie van het AMC (jullie zijn met te veel om bij naam te noemen). De gezellige sfeer en de bereidheid om elkaar te helpen, dat maakte het onderzoek doen leuk en leerzaam.

Lieve Marcella, heel wat artikelen hebben we samen beoordeeld voor de reviews. Vanaf de eerste dag ben je niet alleen een collega maar een maatje! Dat dit nog maar heel lang zo mag blijven.

Ha Joost, enorm gezellige ‘spring in het veld’. Ik ben heel blij dat jij aan een van mijn onderzoeken zo een goed vervolg hebt gegeven. Ik wens je heel veel succes met de afronding van deze RCT en natuurlijk van je gehele promotietraject!

Vissenkommers! Lieve Jolande, Margeet, Katrien en Emily. Dankzij jullie was het elke dag een feestje om naar het werk te gaan. Jullie hebben me in- houdelijk vaak weer even op weg geholpen als ik vast liep achter mijn pc. En wat was het heerlijk om al het lief en leed van het onderzoek, andere werkzaamheden en alles daarbuiten met jullie te kunnen delen. Ik verheug me op nog heel veel etentjes!

Zomer jaar 2000 begon het allemaal….Lieve Inge, Suzan, Marjolijn, Wobke, Kim en Eliane…. sindsdien zijn jullie er altijd voor mij geweest. Met een lach en een traan. Counter balance van dit hele promotietraject. Jullie weten niet half hoeveel jullie voor mij betekenen!

Lieve hockeyers die (bijna) niet meer hockeyen; Marjolein, Marinka, Annemarie, Jelle en Mark. Gelukkig hebben we de derde helft nog, ik verheug me op ons volgende weekendje weg!

Dames van het eerste uur. Lieve Pamela, ik kijk altijd uit naar de gezellige uitstapjes samen. Met en zonder kids. Lieve Kirstin, wat geniet ik er van dat wij elkaar een paar jaar geleden weer tegenkwamen.

322 Chapter 20 Dankwoord Floortje Vlemmix

Lieve Suzanne, we leerden elkaar kennen tijdens onze bachelorstage. Onder- weg naar onze Masterstages in Montréal in 2005 begon een mooie vriendschap. Ook al zien we elkaar niet vaak, even bijkletsen tijdens een kopje koffie of biertje is altijd fijn.

Mijn paranimfen, wat fijn dat jullie naast me komen staan. Lieve Margreet, vanaf de eerste maand in het AMC volgden we hetzelfde traject. Alleen een wat andere volgorde; jij wat eerder in opleiding, ik wat eerder moeder. Het is super fijn om zo samen op te trekken en alles te kunnen delen. Je relati- verend vermogen houdt het leven (als AIOS) luchtig en daarin ben je een voorbeeld voor mij. Lieve Inge, de eerste keer dat wij elkaar ontmoetten, had ik niet kunnen vermoeden dat er zo een mooie vriendschap zou ontstaan. Je enorme ambities in het leven prikkelen mij om ook verder te zoeken naar wat ik wil.

Lieve Ageeth, voortstuwende kracht binnen ons promotie traject. Wat heb ik het getroffen met jou; iemand met wie je te allen tijde harmonieus kan sa- menwerken. Dankzij jouw doorzettingsvermogen en landelijke netwerk zijn de inclusies van de RCT rond gekomen. Ik denk dat we elkaar mooi aanvulden en daarmee het klaar hebben weten te spelen dat dit boekwerk er nu ligt. En dan nog te bedenken dat je in dezelfde periode een huis hebt gebouwd! Je bent een onuitputbare bron van energie. Alles wat jij doet, en dat is veel tegelijk, doe je voor de volle 100%. Ik heb het je met twee zwangerschappen niet makkelijk gemaakt, in die periodes heb je het grootste deel van de promotie op je schouders genomen. Je hebt een groot hart!

Lieve Gerrit, Marit, Lydia en Rozemarijn, bedankt dat ik Ageeth zo lang en zo vaak van jullie heb mogen lenen. Heel veel plezier met alle tijd die jullie nu terugkrijgen als familie. Marit en Lydia, bedankt voor jullie hulp bij de data invoer.

Lieve schoonfamilie, Ineke & Frank, Stefanie & Barth, bedankt voor alle gezellige familie etentjes en uitjes! Ineke en Frank, enorm veel dank voor de goede zorgen voor mij en Krijn tijdens de afrondende fase van dit proefschrift.

Lieve mama, wat zou je hebben genoten van deze dag. In gedachten ben je altijd bij me.

323 Chapter 20 Dankwoord Floortje Vlemmix

Lieve papa, je bent er altijd voor me geweest. Ik koester de bijzondere band die wij hebben, die sinds 2007 alleen maar intenser is geworden. Naast de lieve, behulpzame vader die altijd voor me klaar stond, ben je ook diegene geworden die me 24/7 een luisterend oor biedt. Soms voorzien van een rela- tiverend en adviserend antwoord en soms alleen dat luisterend oor….waar ik even tegenaan kan kletsen over de dingen van alle dag. Ik geniet van alles wat we samen doen en nu dit promotietraject achter de rug is, komt er alleen maar meer tijd vrij voor familie etentjes en vakanties!

Lieve Tim en Stijn, wat bof ik met jullie. Altijd gezellig, altijd lachen, en op zijn tijd serieuze praat. Twee broers die me uitdagen om net als zij zelf grenzen te verleggen. Nooit zal ik vergeten hoe jullie stonden te juichen en mee renden tijdens de laatste kilometer van de Amsterdam Marathon. Jullie gaven mij vleugels.

Lieve Eline en Karin, lieve schoonzussen, bedankt dat jullie bij onze familie zijn gekomen!

En dan, lieve Tijs, jij maakt mijn leven compleet. Al twaalf jaar ben jij de nuchtere persoon naast me, die tijdig op de rem gaat staan zodra werk te veel tijd dreigt op te slokken. En tegelijk ben je de persoon die me de ruimte geeft om mijn werk te doen en in de tussen tijd heel goed voor mij zorgt. Je aanstekelijke enthousiasme voor van alles en nog wat is heerlijk. Ik geniet er van om samen lekker te koken, uren te tafelen, te borrelen aan de bar, te varen door de grachten, of gewoon lekker samen te hangen op de bank. Wat ben ik blij dat jij net zo veel van de bergen houdt als ik. En sinds twee jaar weet ik dat jij de allerliefste, relaxte en leukste vader bent die er is. Daar bof- fen Krijn en Cato maar mee! Nu dit promotietraject is afgerond komt er meer tijd voor onze bucket list. Op een heel mooi leven!

324

21 Portfolio Floortje Vlemmix

327 Chapter 21 Portfolio Floortje Vlemmix

PORTFOLIO FLOORTJE VLEMMIX

PhD training PhD supervisor: Prof.dr. B.W.J. Mol

Courses Academic Medical Center, Amsterdam - Practical biostatistics (2009; 1.1) - Clinical Epidemiology (2009; 1.1) - Clinical data management (2009; 0.3) - Advanced Biostatistics (2010; 1.1) - Systematic Reviews (2009; 0.3) - Scientific writing in English for publication (2010; 1,5) - BROK (‘Basiscursus Regelgeving Klinisch Onderzoek’) (2009; 0.9)

Courses VU University Medical Center, Amsterdam - Multilevel analyses (2012; 1.1)

Seminars, workshops and master classes - Weekly department seminars dept. Obstetrics & Gynecology AMC - Multivariable Model- building by P. Royston & W. Sauerbrei (2009; 1.1) - Castle course clinical epidemiology, France (2010; 0.8) - Consortium Training Days, ; Veldhoven (2012; 0.8) - Internship with dr. N. Nassar and Prof. J.M. Morris, Kolling Institute of Medical Research, University of Sydney, March 2011.

328 Chapter 21 Portfolio Floortje Vlemmix

Presentations - External validation of a prediction model for successful external cephalic version in breech presentation at term. SMFM,30th annual meeting, Chicago, 2010.Poster presentation - Facilitators and barriers of external cephalic version in women with a breech presentation at term. ISDM, th6 conference, Maastricht 2011. Oral presentation. - Why do women opt for or refuse external cephalic version in breech position. ISDM, 6th conference, - Maastricht, 2011. Poster presentation. - Breech deliveries in the Netherlands: effect of increased caesarean section rate on perinatal outcome. SMFM, - 32nd annual meeting, Dallas, 2012. Poster presentation - Impact of caesarean section rate due to term breech presentation on the outcome in subsequent - pregnancies. SMFM, 33th annual meeting, San Francisco, 2013. Post- er presentation

- Implementation of external cephalic version; a multicentre cluster randomised controlled trial. SMFM, 33th - annual meeting, San Francisco, 2013. Poster presentation - National implementation trial ‘external cephalic version’. 43ste Gyn- aecongres, Den Bosch, 2013. Oral presentation. - Term breech deliveries in the Netherlands: effect of increased cesarean rate on neonatal outcome. A prospective cohort study. 10th Posthumus Meyjessymposium, Haarlem, 2013. Oral presentation. - Subsequent pregnancy outcome after breech delivery, a population based cohort study, a national cohort study. 44ste Gynaecongres, Arnhem, 2013. Oral presentation.

329 Chapter 21 Portfolio Floortje Vlemmix

(Inter)national conferences - SMFM,30th annual meeting, Chicago, 2010 - ISDM, 6th conference, Maastricht 2011 - SMFM, 33th annual meeting, San Francisco, 2013 - 43ste Gynaecongres, Den Bosch, 2013 - 44ste Gynaecongres, Arnhem, 2013 Teaching, supervising - Marjon Kuitert (2010). Topic: Patient’s willingness to opt for external cephalic version. - Janna Warendorf (2011). Topic: Decision aids to improve informed decision-making in pregnancy care: a systematic review. - Susan te Hoven (2011). Topic: Baseline implementation of ECV, a retrospective cohort study. - Lecture; ECV and scientific knowledge. ECV trainingcourse for midwifes, Andata coaching & training. November 2013 and March 2014.

330

22 Publications Floortje Vlemmix

333 Chapter 22 Publications Floortje Vlemmix

List of publications F. Vlemmix 1. Facilitators and barriers to external cephalic version for breech presentation at term among health care providers in the Netherlands: a quantitative analysis. Rosman AN, Vlemmix F, Beuckens A, Rijnders ME, Opmeer BC, Mol BW, Kok M, Fleuren MA. Midwifery. 2014 Mar;30(3):e145-50.

3. Patient’s willingness to opt for external cephalic version. Vlemmix F, Kuitert M, Bais J, Opmeer B, van der Post J, Mol BW, Kok M. J Psychosom Obstet Gynaecol. 2013 Mar;34(1):15-21.

4. Decision aids to improve informed decision-making in pregnancy care: a systematic review. Vlemmix F, Warendorf JK, Rosman AN, Kok M, Mol BW, Morris JM, Nassar N. BJOG. 2013 Feb;120(3):257-66.

5. Risk of developmental dysplasia of the hip in breech presentation: the effect of successful external cephalic version. Lambeek AF, De Hundt M, Vlemmix F, Akerboom BM, Bais JM, Papatsonis DN, Mol BW, Kok M. BJOG. 2013 Apr;120(5):607-12.

6. Comment on: a meta-analysis of common risk factors associated with the diagnosis of developmental dysplasia of the hip in newborns. de Hundt M, Vlemmix F, Mol BW, Kok M. Eur J Radiol. 2013 Jan;82(1):199.

7. Contraindications for external cephalic version in breech position at term: a systematic review. Rosman AN, Guijt A, Vlemmix F, Rijnders M, Mol BW, Kok M. Acta Obstet Gynecol Scand. 2013 Feb;92(2):137-42.

8. Risk factors for developmental dysplasia of the hip: a meta-analysis. de Hundt M, Vlemmix F, Bais JM, Hutton EK, de Groot CJ, Mol BW, Kok M.Eur J Obstet Gynecol Reprod Biol. 2012 Nov;165(1):8-17.

9. External validation of a prediction model for successful external cephalic version. De Hundt M, Vlemmix F, Kok M, Van Der Steeg JW, Bais JM, Mol BW, Van Der Post JA. Am J Perinatol. 2012 Mar;29(3):231-6.

334 Chapter 22 Publications Floortje Vlemmix

10. Implementation of the external cephalic version in breech delivery. Dutch national implementation study of external cephalic version. Vlemmix F, Rosman AN, Fleuren MA, Rijnders ME, Beuckens A, Haak MC, Akerboom BM, Bais JM, Kuppens SM, Papatsonis DN, Opmeer BC, van der Post JA, Mol BW, Kok M. BMC Pregnancy Childbirth. 2010 May 10;10:20.

11. The cost-effectiveness of drug-eluting stents: a systematic review. Ligthart S, Vlemmix F, Dendukuri N, Brophy JM. CMAJ. 2007 Jan 16;176(2):199-205.

12. Hyperhomocysteinaemia and vitamin B12 deficiency: the long-term effects in cardiovascular disease. van Oijen MG, Vlemmix F, Laheij RJ, Paloheimo L, Jansen JB, Verheugt FW. Cardiology. 2007;107(1):57-62.

13. Health-related quality of life in patients with cardiovascular disease--the effect of upper gastrointestinal symptom treatment. van Rossum LG, Laheij RJ, Vlemmix F, Jansen JB, Verheugt FW. Aliment Pharmacol Ther. 2004 May 15;19(10):1099-104.

335

23 Curriculum vitae Floortje Vlemmix

337 Chapter 23 Curriculum vitae Floortje Vlemmix

CURRICULUM VITAE – FLOORTJE VLEMMIX

Floortje Vlemmix, daughter of Wim Vlemmix and Anita Vlemmix-Hendriks, was born on September the 16th 1982 in Gouda as second baby of a twin. Together with her two brothers, she grew up in Waddinxveen. After graduation of high school in 2000 (Atheneum, De Goudse Waarden, Gouda), she studied Biomedical Sciences at the Radboud University Nijmegen. In 2003, she started medical school at the same university. In 2006 she obtained her Master of Science degree in Health Technology Assessment and in 2008 her medical degree.

In September 2008 she started as PhD student at the department of Ob- stetrics and Gynecology of the Academic Medical Center in Amsterdam on a project entitled ‘implementation of external cephalic version’. Her PhD thesis was supervised by Prof. dr. B.W.J. Mol (promotor), Prof. dr. J.A.M. van der Post (promotor), dr. M. Kok (co-promotor) and dr. B.C. Opmeer (co-promotor). During the same period, she started working as a resident at the same department.

In September 2012 she started her residency in Obstetrics and Gynecology at the Kennemer Gasthuis in Haarlem (Head of the department dr J.P. Lips until 2013, current head of the department dr. J. Gianotten).

Floortje Vlemmix lives with Tijs van Wijk and their two children Krijn and Cato in Amsterdam.

338

24 List of co-authors and their affiliations

341 Chapter 24 List of co-authors and their affiliations

Authors’ affiliations

1. Dr. M. Kok, Academic Medical Centre, Amsterdam, the Netherlands 2. Dr. M.A.H. Fleuren, TNO Quality of life, Leiden, the Netherlands 3. Dr. B.C. Opmeer, Academic Medical Centre, Amsterdam, the Netherlands 4. Dr. M.E. Rijnders, TNO Quality of life, Leiden, the Netherlands 5. A. Beuckens, MSc, Royal Organisation of Midwives, Utrecht, the Netherlands 6. Prof. dr. J.A.M. van der Post, Academic Medical Centre, Amsterdam, the Netherlands 7. Prof. dr. B.W.J. Mol, The Robinson Institute, University of Adelaide, Australia 8. Dr. M.C. Haak, VU Medical Centre, Amsterdam, the Netherlands 9. Dr. B.M.C. Akerboom, Albert Schweitzer Hospital, Dordrecht, the Netherlands 10. Dr. J.M.J. Bais, Medical Centre Alkmaar, Alkmaar, the Netherlands 11. Dr. S.M.I. Kuppens, Catharina Hospital Eindhoven, the Netherlands 12. Dr. D.N. Papatsonis, Amphia Hospital, Breda, the Netherlands 13. Drs. L. Bergenhenegouwen, Ziekenhuis Groep Twente, Almelo, the Netherlands 14. Dr. J. Schaaf, Academic Medical Centre, Amsterdam, the Netherlands 15. Drs. S. Ensing, Academic Medical Centre, Amsterdam, the Netherlands 16. Dr. A.C.J. Ravelli, Academic Medical Centre, Amsterdam, the Netherlands 17. Drs. A.T.M. Verhoeven, retired, Arnhem, the Netherlands 18. Prof. dr. G.H. Visser, University Medical Centre Utrecht, Utrecht, the Netherlands 19. Drs. B.M. Kazemier, Kennemer gasthuis, Haarlem, the Netherlands 20. Drs. M. Hof, Academic Medical Centre, Amsterdam, the Netherlands 21. Dr. J.J. Duvekot, Erasmus Medical Centre, Rotterdam, the Netherlands 22. Drs. A. Guijt, Verloskundige praktijk Waterland Oost, Volendam, the Netherlands 23. Drs. M. Kuitert, Medical Centre Alkmaar, Alkmaar, the Netherlands 24. Dr. M.C. van Zwieten, Academic Medical Centre, Amsterdam, the Netherlands 25. Drs. J.K. Warendorf, Academic Medical Centre, Amsterdam, the Netherlands 26. Dr. N. Nassar, University of Sydney, Sydney, Australia 27. Prof. dr. J.M. Morris, University of Sydney, Sydney, Australia 28. Drs. S. te Hoven, Academic Medical Centre, Amsterdam, the Netherlands 29. Drs. S. van de Berg, Academic Medical Centre, Amsterdam, the Netherlands 30. Drs. J. Velzel, Academic Medical Centre, Amsterdam, the Netherlands 31. Mw. H. Rota, Rode Kruisziekenhuis, Beverwijk, the Netherlands 32. Drs. M.C. de Hundt, VU Medical Centre, Amsterdam, the Netherlands 33. Dr. E.A.F. Dancet, Academic Medical Centre, Amsterdam, the Netherlands

342

25 Addendum

345 Chapter 25 Addendum

ABSTRACT

Background Breech presentation occurs in 3 to 4 % of all term pregnancies. External cephalic version (ECV) is proven effective to prevent vaginal breech deliveries and therefore it is recommended by clinical guidelines of the Royal Dutch Organisation for Midwives (KNOV) and the Dutch Society for Obstetrics and Gynaecology (NVOG). Implementation of ECV does not exceed 50 to 60% and probably less. We aim to improve the implementation of ECV to decrease maternal and neonatal morbidity and mortality due to breech presentations. This will be done by defining barriers and facilitators of implementation of ECV in the Netherlands. An innovative implementation strategy will be developed based on improved patient counselling and thorough instructions of health care providers for counselling.

Method/ design The ultimate purpose of this implementation study is to improve counselling of pregnant women and information of clinicians to realize a better implementation of ECV. The first phase of the project is to detect the barriers and facilitators of ECV. The next step is to develop an implementation strategy to inform and counsel pregnant women with a breech presentation, and to inform and educate care providers. In the third phase, the effectiveness of the developed implementation strategy will be evaluated in a randomised trial. The study population is a random selection of midwives and gynaecologists from 60 to 100 hospitals and practices. Primary endpoints are number of counselled women. Secondary endpoints are process indicators, the amount of fetes in cephalic presentation at birth, complications due to ECV, the number of caesarean sections and perinatal condition of mother and child. Cost effectiveness of the implementation strategy will be measured.

Discussion This study will provide evidence for the cost effectiveness of a structural implementation of external cephalic versions to reduce the number of breech presentations at term.

Trial Registration Dutch Trial Register (NTR): 1878

346 Chapter 25 Addendum

BACKGROUND Breech presentation occurs in 3 to 4 % of all term pregnancies.1 In pregnancies complicated by breech presentation, perinatal mortality, neonatal mortality or serious neonatal morbidity are increased as compared to pregnancies where the fetes is in cephalic postion.2 Breech position at term occurs in 8.000 pregnancies in the Netherlands each year. Until 2000 the mode of breech delivery was controversial. Since publication of the results of the term breech trial, the caesarean delivery rate in women with a fetes in breech presentation in the Netherlands has increased from 45% to around 80%.3 Although an elective caesarean section is safer for the baby, it increases maternal morbidity .4 Moreover, the uterine scar carries a risk for future pregnancies. 5 External cephalic version (ECV) reduces the rate of non-cephalic presentations at term with 40-50%, and thus the number of caesarean deliveries performed for at term breech presentation, without any increased risk to the baby.6 The high caesarean delivery rate for breech presentation makes ECV an important obstetric intervention and it is therefore recommended by the Royal College of Obstetricians and Gynaecologists in the Clinical Green Top Guidelines. A recent inventory among Dutch gynaecologists showed that 5% of the practices never offered ECV, whereas one practice offered ECV onlyto multipara. In 28% of the practices an ECV was performed by all gynaecologists, whereas in the other 72% ECV was performed by a smaller team of specialists. The palette of relative contraindications differed widely, and only 19% of the responding clinics registered their success rates. The number of patients refusing ECV was estimated to be 20 to 30%. At present, 60 to 70% of the women with a foetus in breech position undergoes ECV in the Netherlands. This rate is probably an overestimation, as they are based on self-reported rates by midwives and gynaecologists, in the absence of registrations. The true number of women that undergoes ECV for term breech is probably below 50%.

With 6.000 term breech deliveries each year, of which 5.000 are delivered by caesarean section, there is clearly room for improvement. The number of caesarean sections for breech delivery can be reduced with approximately 2.000 per year. As the additional costs of a caesarean section as compared to vaginal delivery are estimated to be 1.500 Euros, the potential saving of a better implementation of ECV can reduce costs with 2 to 3 million Euros per year for direct medical costs only. As pregnant women with a previous

347 Chapter 25 Addendum

caesarean section are at increased risk of complications, and therefore always deliver under responsibility of a gynaecologist, the potential savings are even higher. In the Netherlands, there is professional consensus among the Dutch Society for Obstetrics and Gynaecology (NVOG) as well as the Royal Dutch Organisation for Midwives (KNOV) that ECV should be offered to all women with a fetes in breech presentation at term. The KNOV has produced a leaflet for patients in which ECV is recommended. Nevertheless, the number of women potential suitable for ECV who were not offered an attempt range from 4% to 33%.7-9 Moreover, a substantial number of women refuse ECV, and opt immediately for a caesarean section. In Australia, Raynes-Greenow and others investigated pregnant women’s preferences and knowledge of term breech management.10 39% would choose ECV, 39% would not choose ECV, and the remaining 22% were uncertain. Yogev and colleagues performed a similar study in Israel.11 They reported that in 1995, more than half the women (52.7%) had heard of ECV and 53.8% were willing to consider it, whereas in 2001, 73.2% had heard of it but only 23.9% were willing to consider it. Johanson reported that out of a group of 323 pregnant women with a foetus in breech presentation, 65% opted for external cephalic version after they were informed.12 They also demonstrated an association between the gynaecologists who provided information and the level of uptake by the women. The number of women with a breech presentation considering ECV and their knowledge about the procedure are unknown in the Netherlands. Until publication of the new NVOG and KNOV guidelines in 2001 and 2002, ECV used to be a controversial intervention. Since the introduction of these guidelines, there is no evidence for a raise in the number of ECVs performed. However, the number of birth in breech presentation at term is stable since 2002, indicating a low implementation grade of ECV. Implementation requires a clear and deliverable evidence based message.13 However, there are often major discrepancies between best evidence and practice, resulting in a large variation between professionals.14-17 Based on this national and international literature, two main barriers for ECV implementation are identified 1) lack of patients’ knowledge about risk for and consequences of ECV and breech delivery 2) the attitude and knowledge of midwives and gynaecologists towards ECV. This study will show us if an implementation strategy, tailored on main barriers combining both patient and health care provider interventions,

348 Chapter 25 Addendum is more cost-effective than usual care. We will also evaluate the process of implementation to ascertain which elements of the strategy can be particularly associated with successful implementation of ECV. Moreover, our proposal is a collaborative effort of midwives and gynaecologists, and is supported by both the KNOV and NVOG.

349 Chapter 25 Addendum

METHODS / DESIGN Aims The main objective of this study is to assess barriers and facilitators of implementation of ECV in the Netherlands. We would like to create a broad social basis and awareness for the need of cooperation in care for women with a fetes in breech position. An innovative implementation strategy will be developed based on improved patient counselling and thorough instructions of health care providers for counselling. Research questions are: What are barriers and facilitators of implementation of ECV in the Netherlands? What are the costs and effects of an innovative implementation strategy based on improved patient counselling and information of health care providers to implement the guidelines on ECV for breech presentation? What is the feasibility of this implementation strategy?

Methods The proposal will contain three phases. In the first phase, we will identify facilitators and barriers of implementation of ECV and create awareness of the need for improvement. Subsequently, we will develop an implementation strategy targeted on patient counselling and information of health care providers, and evaluate the cost-effectiveness of the developed strategy.

Identification of facilitators and barriers of implementation of ECV A topic list will be defined by an expert panel. The topic list will be constructed on the basis of existing knowledge and theories on implementation factors, primarily based on the clinical guidelines by the NVOG and KNOV. This list will be used to guide and direct semi-structured interviews and focus groups, patients, gynaecologists and midwives will be asked for their attitude towards ECV. We will assess fear for complications, skills and attitudes. A framework approach will be used to analyse the data and to derive potential factors that might influence the implementation of ECV.18 In the second step the exploratory findings will be used to develop a series of questionnaires. These will then be distributed among a much larger number of actors. Quantitative analyses will be conducted, accounting for the multilevel nature of the problem and the corresponding data-collection, to study the prevalence of factors that facilitate and/or inhibit the implementation of external cephalic version in the Netherlands. In this process, we will take into account regional and socio-economic differences. To identify and explore relevant factors, semi-structured interviews will be held with pregnant women

350 Chapter 25 Addendum with a fetes in breech presentation (N=20), general practitioners (N=20), midwives (N=20), gynaecologists (N=20), and representatives of health care insurers. Respondents will be ‘purposively’ selected.19 Gynaecologists and midwives will be selected on relevant characteristics of their practices (region, hospital). Moreover, we will interview partners of pregnant women separately, but also as a couple. The interviews will be guided by the topic list that is defined on the basis of existing knowledge and theories on implementation.20 The topic list will be used to direct the interviews, but will be informally and flexible applied in order to prevent us from imposing our preconceptions. The topic list will contain themes that include knowledge of the complications of ECV, about the risks of caesarean section in terms of short- and long term maternal and neonatal mortality and morbidity, and about available evidence. We will ask couples about their attitude towards the problem, about their expectations and previous experiences, and about their beliefs about health care in general. Midwives and gynaecologists will be interviewed on similar subjects, but they will be asked more extensively about the potential effectiveness of ECV (relative risks, numbers needed to treat, numbers needed to harm) as well as the influence of wider societal, organisational, financial, and contextual factors. We expect the following categories of factors to play a role: socio-political factors (opinion stated by opinion leaders, comments in journals, opinion of experts within the own department, attitude of the head of the department), organizational factors, and knowledge and beliefs (doctor and patient level, (the lack of) knowledge, about complications of vaginal breech delivery and caesarean section, anxiety for ECV and the demand for caesarean section (patient level only)). The semi-structured interviews will be analysed on the basis of a framework approach 18. The analyses will help us identify the important factors that are most likely to play a role in the implementation of ECV. We expect that we will be able to distil a list of approximately fifteen issues that are of potential importance. The exploratory findings will be used to develop a questionnaire, which is supposed to be extensive and complete. The questionnaire will be sent to all midwives practices and to the departments of obstetrics and gynaecology of all hospitals in the Netherlands. Moreover, all health insurers will be targeted. Ethical approval is not needed in accordance to the prescriptions of the medical ethical committee of the AMC, if individual patients will be approached for semi structured interviews, without provocation of emotional arousal and with anonymous processing of the data study.

351 Chapter 25 Addendum

Development of an implementation plan After analysis, the results will be discussed in a final meeting with the study group and representative participants (clients, midwives, gynaecologists). Based on the discussion of these results, an implementation plan, with different actions and interventions, will be developed. Although parts of the plan will highly depend on the obstacles of change to be identified (e.g. knowledge, attitude, behaviour, logistics), we anticipate the following ingredients: 1) an educational brochure/website with evidence, one designed for gynaecologists/midwives and one designed for patients. 2) group/unit meetings to discuss the protocol, the resistance to it, and how to overcome problems in adaptation. 3) if necessary, a training in counselling patients on ECV for midwifes and/or gynaecologists can be developed 4) reminders; regular self-monitoring; regular observation by heads of units. This plan will be developed according to phase 2 to 5 of Grols 5-step implementation model. Depending on the results of the proposed project, additional funding will be sought for these phases.

Evaluation of the developed implementation strategies design A cluster randomised controlled trial with an economic evaluation will be performed alongside. Ethical approval will be requested before the start of this part of the study. Information on counselling of patients, required for ethical approval, will be specified in the first phase of the trial. A set of quality indicators will be extracted from the NVOG and KNOV guidelines. The indicator development will be performed according to the RAND-modified Delphi method.21 First of all, key recommendations from the guidelines will be extracted by two or three experts, the project leaders. Subsequently, the clinical relevance of all key recommendations for patient’s health benefit and efficacy will be tested in two rounds among an independent panel of 12- 15 experts (KNOV and NVOG guideline writers, KNOV and NVOG members, quality of care experts and patients). The key recommendations with the highest scores will be selected and made ready for use, in measurable elements (process indicators). Moreover, measurement instruments will be developed and the participating hospitals will be informed about the study. A detailed protocol on the randomised controlled trial will be written after finishing the first phase. This is a requirement to receive further funding of the further project. This protocol is in development; it depends mainly in the outcome of the first phase. This will be followed by a pilot, feasibility study where two different

352 Chapter 25 Addendum implementation strategies will be tested in four participating hospitals, one primary care centre specialised in ECV and one independent midwifery practice. Before the pilot study starts, baseline characteristics and primary and secondary outcomes will be measured with questionnaires and data gathering on amount of version attempts, outcomes after ECV and caesarean section rate. The data gathering lists and questionnaires will also be developed during the first phase. After the pilot study a cluster randomised controlled trial will be performed in which two implementation strategies will be evaluated: a patient centred strategy and a health care provider (midwife and gynaecologist) centred strategy. In the randomised clinical trial, we will allocate centres and their regions at random to four groups: A. Patient strategy alone, B. Healthcare providers-strategy alone, C. Both the patient and healthcare provider strategies, D. No specific implementation intervention. Up to 30% of eligible patients are not offered an ECV attempt and upto 40% of counselled patients do not choose for ECV, which might be the result of insufficient information. With the decision aid we intend to increase the number of well-informed patients from 50 to 80%. To correct for the degree of similarity among responses within a cluster, an intra-cluster correlation coefficient of 0.10 was integrated in the power analyses. To be able to show the difference of 30% with a power of 80% and an alpha error of 5%, we need 20 clusters of a 30 patients each. The gynaecologists in the participating hospitals will give informed consent for this trial to confirm that they will treat their patients conform one of the four strategies were they will be allocated to, as long as the trial is open for inclusion.

Outcome measures and process indicators Primary endpoint is the number of patients that has an ECV performed. Secondary endpoints are guidelines’ adherence rates, complications of ECV, the number of fetes that are in cephalic position at delivery, the number of caesarean sections and the perinatal condition of mother and child. Moreover, we will assess patient knowledge (e.g. ECV, breech delivery, caesarean section), patient decisional conflict and patient satisfaction. We will also calculate costs of both implementation interventions and medical interventions. In case one or both implementation interventions are effective, their cost-effectiveness will be assessed.

353 Chapter 25 Addendum

Implementation study An effect and process evaluation will be performed. An effect evaluation of the two strategies will be carried out using the primary and secondary outcome measures and the set of process indicators derived from the NVOG and KNOV guidelines. This will be done among pregnant women with breech presentation at term and professionals involved in the care for these women. The measurements will be performed in the 20 participating hospitals before and after implementation of the different strategies by a medical record search, added with questionnaires among professionals and patients. The medical records will be searched using standardised registration forms. A process evaluation will be performed to study the feasibility of the two strategies. The extent by which clinicians, midwives and patients used these elements will be measured.

ABBREVIATIONS

External cephalic version (ECV), Dutch Society for obstetrics and gynaecology (NVOG), The royal Dutch organisation for midwives (KNOV).

COMPETING INTERESTS The study is sponsored by ZonMW; the Dutch organisation for health research and innovation. The author(s) declare that they have no competing interests.

AUTHORS’ CONTRIBUTIONS BWJM, MK, JAMvdP, MAHF, AB and MEBR were involved in conception and design of the study. BWMJ, MK, BCO, AR and FV drafted the first manuscript. All authors mentioned in the manuscript are members of the ‘ECV implementation study group’. They participated in the design ofthe study during several meetings and are local investigators in the participating centers. All authors read and approved the final manuscript.

AUTHORS’ INFORMATION BWJM (clinical epidemiologist and gynaecologist) has been involved in many projects in the field of Obstetrics and Gynaecology. His thesis, which focused on the evaluation of diagnostic and prognostic tests in sub fertility, was awarded with the Jan Swammerdam prize. In 2002, he has initiated the Dutch OFO-project, a study that aims to evaluate the effectiveness and cost

354 Chapter 25 Addendum effectiveness of the basic fertility work-up. The key paper of this project has been published in The Lancet (Lancet 2006;368:216-21). A study proposal entitled: Use of probabilistic decision rules in Obstetrics and Gynaecology was granted in the VIDI program of ZonMw, the Netherlands. He has (co-)authored over 150 peer reviewed publications, and he has supervised nine doctorates. He was instrumental in founding a Dutch consortium of obstetricians that cooperates in performing studies in obstetrics (http://www.studies-obsgyn. nl/index.asp). He is chair of the Dutch guideline committee of the NVOG. He is involved in implementation studies in obstetrics.

PMO practiced ten years in several independent midwifery practices. In 2001 she graduated in Health Science at the University of Maastricht. From 1995 until 2003 she worked as midwife researcher in several projects, at TNO-PG and later at WOK (UMCN), e.g. a study to assess the adherence and barriers for implementation of the KNOV guideline Anemia for primary care midwifes. Since 2003 she is staff member of the Royal Dutch Organisation for Midwifes (KNOV), and there she is responsible for the guideline development. In 2006 she was co-auteur of the KNOV guideline ‘External Cephalic Version’. Offerhaus P, Fleuren M, Wensing M. Guidelines on anaemia: effect on primary care midwives in the Netherlands. Midwifery 2005;(21): 204-211

MR works as a research-midwife at TNO quality of Life. She has extensive experience in conducting research projects in independent midwifery practices: for example the Serinam study, a RCT in 55 midwifery practices, the ECV prevalence study in 50 midwifery practices (both still ongoing) and a retrospective study on satisfaction and mode of delivery in 8 midwifery practices. Furthermore, she conducted 2 studies on ECV: a retrospective study of all ECV between 1996 and 2000 in the Slotervaart Hospital and a prospective study on ECV in the same hospital including expectations and experiences if women with ECV (publication in press). Other research projects she participated in were a cost effectiveness study on screenings strategies for EOGBS and a study on information and informed consent of women with neonatal screening. Finally she was co-author of the first midwifery guideline “Anaemia in first line obstetric practice”.

JAMvdP is clinical director of obstetrics in the AMC and his main research focus is high risk pregnancy. His thesis dealt with pathofysiology of pre-eclampsia. He is currently involved in management of the Perinatal Research Unit, a

355 Chapter 25 Addendum

collaborative initiative from the departments of obstetrics and neonatology for the monitoring of multicenter clinical studies, and a consortium of perinatal centres in the Netherlands. Work in progress concerns early diagnosis of pre-eclampsia, external version for breech presentation (RCT) and treatment of recurrent abortion (RCT), the value of the intrauterine pressure catheter IUPC trial (RCT).

BCO (psychologist, methodologist) has participated in the coordination of European collaborative studies (European Communities 4th Medical and Health Research Program COMAC; BIOMED1) concerning health services research and is since 1999 working in the department of Clinical Epidemiology and Biostatistics in the Academic Medical Center. He has been involved in several evaluation studies (RCTs) in the fields of Dermatology, Surgery and Infectious diseases respectively, especially focusing on the economic and methodological perspective. He has conducted a study on methods for determining patients’ treatment preferences and trade-offs in health care technology assessment research (ZONMW 2001-2003). He is currently involved in several studies in the field of obstetrics, gynaecology and fertility research.

ACKNOWLEDGEMENTS This study is funded by ZonMW.

356