BJOG: an International Journal of Obstetrics and Gynaecology DOI: 10.1111/j.1471-0528.2004.00518.x May 2005, Vol. 112, pp. 627–631

Tocolysis for repeat external cephalic version in breech presentation at term: a randomised, double-blinded, placebo-controlled trial

Lawrence Impey,a Meghana Panditb

Background External cephalic version (ECV) reduces the incidence of breech presentation at term and caesarean section for non-cephalic births. Tocolytics may improve success rates, but are time consuming, may cause side effects and have not been proven to alter caesarean section rates. The aim of this trial was to determine whether tocolysis should be used if ECV is being re-attempted after a failed attempt. Objective To determine whether tocolysis should be used if ECV is being re-attempted after a failed attempt. Design Randomised, double-blinded, placebo-controlled trial. Setting UK teaching hospital. Population One hundred and twenty-four women with a breech presentation at term who had undergone an unsuccessful attempt at ECV. Methods Relative risks with 95% confidence intervals for categorical variables and a t test for continuous variables. Analysis was by intention to treat. Main outcome measures Incidence of cephalic presentation at delivery. Secondary outcomes were caesarean section and measures of neonatal and maternal morbidity. Results The use of tocolysis for a repeat attempt at ECV significantly increases the incidence of cephalic presentation at delivery (RR 3.21; 95% CI 1.23–8.39) and reduces the incidence of caesarean section (RR 0.33; 95% CI 0.14–0.80). The effects were most marked in multiparous women (RR for cephalic presentation at delivery 9.38; 95% CI 1.64–53.62). Maternal and neonatal morbidity remain unchanged. Conclusions The use of tocolysis increases the success rate of repeat ECV and reduces the incidence of caesarean section. A policy of only using tocolysis where an initial attempt has failed leads to a relatively high success rate with minimum usage of tocolysis.

INTRODUCTION whether it is best administered before every attempt at ECV or whether it should be used only after a failed attempt. Breech presentation occurs at birth in 3–4% of all The aim of this trial was to test the null hypothesis that deliveries and is a major contributor to the caesarean tocolysis for a repeat ECV (i.e. after a failed ECV) does not section rate.1 This problem has increased2 following the increase the success rates of ECV. publication of the Term Breech Trial.3 Although external cephalic version (ECV) reduces both the incidence of breech presentation at delivery and the caesarean section METHODS rate,4 this effect is limited by low success rates, especially in white British women.5 Tocolysis may improve success, The trial was undertaken in the breech clinic, held in but the effect on non-cephalic births is limited.6 It is also the delivery ward of the John Radcliffe Hospital, Oxford. time consuming and may be unpleasant, and it is not known Ethical approval for the study was obtained from the Cen- tral Oxford Research Ethics Committee (C00.024) in April 2000. Women were referred to the clinic from hospital or community clinics, including those attached to other a local hospitals, if they had a singleton breech presenta- Obstetrics and Fetal Medicine, Oxford Fetal Medicine tion at 36 or more (nulliparous) or 37 or more (multiparous) Unit, Level 6, The Women’s Centre, John Radcliffe weeks. Health professionals were encouraged to refer all Hospital, Oxford, UK b breech presentations to the clinic for ECV before consid- Obstetrics and Gynaecology, Milton Keynes General ering contraindications or counselling regarding mode of Hospital, UK delivery. Correspondence: Mr L. Impey, Oxford Fetal Medicine Unit, Level 6, Women were eligible for the trial if they had undergone The Women’s Centre, John Radcliffe Hospital, Oxford, OX3 9DU, UK. an unsuccessful attempt at ECV (without tocolysis) for a

D RCOG 2004 BJOG: an International Journal of Obstetrics and Gynaecology www.blackwellpublishing.com/bjog 628 L. IMPEY & M. PANDIT

allocations in sealed numbered opaque envelopes opened in sequential order on the delivery ward. Both the patient and the ECV operator were blinded to the result of random- isation, including maternal observations, and the latter was not allowed to enquire about side effects. Tocolysis was administered as ritodrine hydrochloride (Yutopar) infusion of 50 mg (10 mg/mL) added to 12 mL dextrose saline (total 17 mL of ritodrine 3 mg/mL). This was administered by a syringe pump, starting at 1 mL/hour and increasing at 10-minute intervals to a maximum of 5 mL/hour, provided the maternal pulse did not exceed 120/minute. The placebo consisted of 17 mL dextrose sa- line solution by the same route and in similar increments, for a maximum of 50 minutes. Maternal pulse and blood pressure were measured every 10 minutes in both arms. ECV was re-attempted at 50 minutes or when the maternal pulse exceeded 120/minute, whichever was sooner. Wher- ever possible, the repeat ECV was performed by the same operator as for the first attempt. Data were collected prospectively by the research mid- wife. The primary outcome was the incidence of cephalic presentation at delivery. For the power calculation, we estimated a spontaneous version rate in the control group of 5%.7 Tocolysis for all women increases success rates from approximately 40% to 55%.8,9 Therefore, of the 60 women out of 100 who have had a failed attempt with- out tocolysis, a further 15 (25%) could reasonably be ex- pected to have a second successful attempt with tocolysis. To detect an increase in the incidence of cephalic presen- tation from 5% to 25% with 90% power (a ¼ 0.05), 124 patients were required. Secondary outcomes were the incidence of successful ECV, incidence of caesarean section, length of hospital inpatient stay, incidence of neonatal Apgar scores <7at Fig. 1. Trial profile. 5 minutes, neonatal intensive care unit admission or other rarer neonatal outcomes and mean cord arterial pH. A patient assessment of discomfort using an abbre- viated, modified McGill pain intensity score10 was also breech presentation. They were required to have had a compared. normal cardiotocograph following this, and no potential Analysis was by intention to treat. For categorical contraindications to the use of ritodrine (cardiac disease, variables, relative risks were calculated, with 95% confi- pre-existing or gestational diabetes, hypertension). Women dence intervals. Continuous variables, where normally ineligible for an initial attempt at ECV were those with a distributed, were compared by a t test. pre-existing indication for caesarean section, suspected unstable lie, pre-eclampsia, recent (<4 weeks) antepartum haemorrhage, suspected fetal compromise (abdominal cir- Table 1. Demographics. Values are presented as n (%) or mean [SD]. cumference below the third centile, with either an umbilical Tocolysis (n ¼ 62) Placebo (n ¼ 62) artery resistance index above the 97th centile or deepest amniotic fluid pocket <2 cm) or rhesus isoimmunisation. Maternal age 30.6 [4.5] 30.9 [5.5] Information sheets regarding the trial were given in ad- Nulliparous 44 (71.0) 45 (72.6) Booking weight 66.5 [13.1] 63.0 [12.0] vance of the initial ECV attempt. After a failed attempt on Height (cm) 164.6 [7.8] 165.3 [5.7] an eligible patient, written consent was obtained by the re- Caucasian race 61 (98.4) 61 (98.4) search midwife. Consenting patients were then randomly Extended breech 38 (61.3) 42 (67.7) allocated, in a ratio of 1:1, to receive either tocolysis or Liquor pool depth (mm) 43.3 [16.0] 47.7 [13.7] placebo for a repeat attempt at ECV. Randomisation was Duration first ECV (minutes) 9.1 [2.1] 9.5 [2.7] Gestation at ECV (weeks) 37.5 [0.81] 37.5 [0.85] achieved using random block sizes up to 20, with the D RCOG 2004 BJOG: an International Journal of Obstetrics and Gynaecology 112, pp. 627–631 RANDOMISED TRIAL OF TOCOLYSIS FOR A REPEAT EXTERNAL CEPHALIC VERSION 629

Table 2. Effect of tocolysis. The values are presented as n (%) or mean [SD].

Tocolysis (n ¼ 62) Placebo (n ¼ 62) RR (95% CI)/P

Primary outcome Cephalic presentation at delivery 19 (29.0) 7 (11.3) 3.21 (1.23–8.39)

Secondary outcomes Successful ECV 17 (27.4) 5 (8.1) 4.31 (1.48–12.57) Caesarean section 41 (66.1) 53 (85.5) 0.33 (0.14–0.80) Elective caesarean 40 (64.5) 48 (77.4) 0.53 (0.24–1.17) Breech delivery 5 (6.5) 5 (6.5) 1.00 (0.24–4.19) Birthweight (g) 3169 [484] 3310 [429] P ¼ 0.658 Gestation at delivery (weeks) 40.0 [1.2] 38.8 [1.0] P ¼ 0.076 SCBU admission 2 (3.2) 2 (3.2) 1.00 (0.14–7.33) Length SCBU stay (days) 0.11 [0.76] 0.08 [0.45] P ¼ 0.831 Apgar < 7at5 0 0 Mean arterial pH 7.25 [0.07] 7.26 [0.08] P ¼ 0.752 Neonatal seizures 0 0 McGill score 2.3 [0.7] 2.1 [0.7] P ¼ 0.303 Maternal stay (days) 3.9 [0.3] 4.4 [0.3] P ¼ 0.176 Maternal morbidity 0 0

Data were entered prospectively on a database SPSS 8.0 An increase in cephalic presentation at delivery was (Chicago, Illinois). Demographic and antenatal data were found (Table 2). This was due to an increase in the success obtained retrospectively at the time of recruitment. Out- of repeat ECV and was followed by a decrease in the inci- come details were collected retrospectively from hospital dence of caesarean section. This was largely due to a re- computerised records. ECV details were collected pro- duction in the incidence of elective caesarean section for spectively. The treatment allocation was recorded only as breech presentation. There were no significant differences a number: randomisation details were revealed only when in maternal postnatal stay or in neonatal outcomes. There recruitment was complete. was no serious neonatal or maternal morbidity in the entire The funding source had no role in the study design, data cohort. collection, analysis or interpretation or in the writing of this Among 505 women referred to the Breech clinic during report. the time period, 485 (96%) were eligible for ECV and underwent it, and in 178 (36.7%) it was successful. An additional attempt at ECV using tocolysis in all eligible RESULTS women with a failed ECV would therefore have increased the number of successful ECVs by 78, giving an overall Between 12 June 2000 and 11 November 2003, 505 success rate of 52.8%. women with a breech presentation were seen in the clinic The effect on the success of ECV was most marked in and 124 women were recruited to the trial (Fig. 1). Follow up multiparous women (Table 3). In nulliparous women, a was available on all women, and in all cases women statistically significant effect was not seen in the success underwent the treatment to which they had been allocated. rate or incidence of cephalic presentation at delivery or Demographic variables are shown in Table 1. The vast caesarean section. majority were of white British origin and over 70% were While there was no difference in pain scores between nulliparous. the groups, only one woman experienced no pain and

Table 3. Influence of parity on effect of tocolysis. Values are presented as n (%).

Parity ¼ 0 Tocolysis (n ¼ 44) Placebo (n ¼ 45) RR (95% CI)

Cephalic presentation at delivery 8 (18.2) 5 (11.1) 1.78 (0.53–5.93) Successful ECV 8 (18.2) 4 (8.9) 2.28 (0.63–8.20) Caesarean section 35 (79.5) 41 (91.1) 0.38 (0.11–1.34)

Parity ¼ 1þ Tocolysis (n ¼ 18) Placebo (n ¼ 17) RR (95% CI)

Cephalic presentation at delivery 10 (55.6) 2 (11.8) 9.38 (1.64–53.62) Successful ECV 9 (50.0) 1 (5.9) 16.00 (1.74–147.54) Caesarean section 6 (33.3) 12 (70.6) 0.21 (0.05–0.87)

D RCOG 2004 BJOG: an International Journal of Obstetrics and Gynaecology 112, pp. 627–631 630 L. IMPEY & M. PANDIT

93 (75.8%) described the procedure as ‘discomforting’. in spite of our points about why ECV could be more Eight women (6.5%) described it as ‘horrible’ or worse. difficult in our group of women. This probably under-estimates maternal discomfort at ECV The other alternative of using tocolysis only in nullipa- because those who found the initial attempt uncomfortable rous women, as has been advocated by some,9 could have would be expected to have been less likely to take part in led to fewer overall successes. This is because the effect of the trial. tocolysis seemed most marked in the small group of mul- tiparous women. This influence of parity may also mean that a larger effect of tocolysis would have been evident if DISCUSSION multiparous women had not been under-represented in the trial. This trial shows that where ECV has failed, tocolysis A further question concerns the choice of tocolytic. We should be used for a repeat attempt. It also confirms the used ritodrine because this has been clearly shown to benefits of ECV in reducing the caesarean section rate, improve overall ECV success rates6,13 while some others particularly now that vaginal breech birth is rare. have not. It may, however, be that atosiban (Tractocile) is There are potential limitations of our study. Regarding as effective with fewer side effects and a faster onset of the blinding, firstly, women receiving ritodrine developed action: its use should be investigated in a randomised, increased sympathetic tone. The operator, however, did not controlled trial. record or see the maternal observations, and as the use of ultrasound required the room to be darkened, any facial flushing could not be seen. Secondly, we cannot prove that CONCLUSIONS the operator could not deduce the allocation from the time between the initial attempt and the trial attempt because External cephalic version success rates can be markedly the pulse in women receiving tocolysis might reach the improved, and non-cephalic births and caesarean section maximum allowed before the 50 minutes for which women for breech presentation reduced, by the use of tocolysis if a allocated to the placebo arm had to wait. In practice, the second attempt is made. Restriction of its usage to this person performing ECV was also covering the delivery indication might lead to a higher overall success rate than ward and all timing was determined by the research mid- when it is used simply for all nulliparous women and lead wife. Another criticism concerns the initial success rate, to less use of tocolysis and therefore side effects, time and which was low when compared with some published series. resources than when it is simply used for all. This, however, is likely to be due to not using tocolysis, to the small number of contraindications to ECV and to the fact that ECV was always attempted in consenting women Acknowledgments however difficult it might be. In addition, there was a small percentage of women in whom ECV is less difficult: This trial was funded by a grant from the NHS Executive multiparous women11 and those of non-white British race.5 South East Research and Development Directorate (CE0. Given that tocolysis is effective at improving ECV 093). success rates, the choice is whether it should be used for The authors gratefully acknowledge the participating all, or where an initial attempt has failed, or just for women, the research midwives Kate Sharp and Pauline nulliparous women. This trial does not directly answer this Ellaway, Rebecca Black and Deborah Harrington for question, but deductions can be made. performing some of the ECVs, Peter Brocklehurst and Despite evidence as to the safety of ECV,7 concerns Simon Gates for statistical advice and Ian MacKenzie for among women and obstetricians remain. This can result co-application for funding. in precautions such as starvation of women, intravenous access and use of an operating theatre. We did not use any of these because we, and others,11 have not needed to resort to emergency delivery. Therefore, the risks of emergency References delivery are much lower than, say, for normal labour, when such precautions are not routine. Although the procedure 1. Thomas J, Paranjothy S. Royal College of Obstetricians and Gynae- cologists Clinical Effectiveness Support Unit. National Sentinel Cae- does not take long, counselling and monitoring afterward sarean Section Audit Report. London: RCOG Press, 2001. may do, and considerable manpower and resources are 2. Molkenboer JF, Bouckaert PX, Roumen FJ. Recent trends in breech 12 required. The time and monitoring of tocolysis and the delivery in the Netherlands. Br J Obstet Gynaecol 2003;110:948–951. side effects of tocolysis will increase this requirement 3. Hannah ME, Hannah WJ, Hewson SA, Hodnett ED, Saigal S, Willan considerably. With a policy of tocolysis only after a failed AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Term Breech initial attempt, we would have achieved a similar overall Trial Collaborative Group. Lancet 2000;356(9239):1375–1383. success rate (52.8%) to other series using tocolysis for 4. Hofmeyr GJ, Kulier R. External cephalic version for breech presen- 8,9 all, despite using it in only 63.3% of all women. This is tation at term. Cochrane Database Syst Rev 2000;(2):CD000083. D RCOG 2004 BJOG: an International Journal of Obstetrics and Gynaecology 112, pp. 627–631 RANDOMISED TRIAL OF TOCOLYSIS FOR A REPEAT EXTERNAL CEPHALIC VERSION 631

5. Hofmeyr GJ, Sadan O, Myer IG, Galal KC, Simko G. External cephalic 10. Melzack R. The McGill pain questionnaire: major properties and version and spontaneous version rates: ethnic and other determinants. scoring methods. Pain 1975;1:277–299. Br J Obstet Gynaecol 1986;93:13–16. 11. Boucher M, Bujold E, Marquette GP, Vezina Y. The relationship be- 6. Hofmeyr G. Interventions to help external cephalic version for tween amniotic fluid index and successful external cephalic version: a breech presentation at term. Cochrane Database Syst Rev 2004;1: 14-year experience. Am J Obstet Gynecol 2003;189:751–754. CD000184. 12. Mauldin JG, Mauldin PD, Feng TI, Adams EK, Durkalski VL. 7. Impey L, Lissoni D. Outcome of external cephalic version after 36 Determining the clinical efficacy and cost savings of successful weeks’ gestation without tocolysis. J Matern Fetal Med 1999;8:203–207. external cephalic version. Am J Obstet Gynecol 1996;175:1639–1644. 8. Chung T, Neale E, Lau TK, Rogers M. A randomized, double blind, 13. Bujold E, Marquette GP, Ferreira E, Gauthier RJ, Boucher M. controlled trial of tocolysis to assist external cephalic version in late Sublingual nitroglycerin versus intravenous ritodrine as tocolytic for pregnancy. Acta Obstet Gynecol Scand 1996;75:720–724. external cephalic version: a double-blinded randomized trial. Am J 9. Marquette GP, Boucher M, Theriault D, Rinfret D. Does the use of a Obstet Gynecol 2003;188:1454–1457. tocolytic agent affect the success rate of external cephalic version? Am J Obstet Gynecol 1996;175:859–861. Accepted 18 September 2004

D RCOG 2004 BJOG: an International Journal of Obstetrics and Gynaecology 112, pp. 627–631