British Journal of Anaesthesia 104 (5): 613–18 (2010)
doi:10.1093/bja/aeq053
Advance Access publication March 25, 2010
OBSTETRICS
Randomized controlled trial of external cephalic version in term
multiparae with or without spinal analgesia†
C. F. Weiniger 1*, Y. Ginosar 1, U. Elchalal 2, H. Y. Sela 2, C. Weissman 1 and Y. Ezra 2
1
Department of Anesthesiology and Critical Care Medicine and 2Department of Obstetrics and Gynecology,
Hadassah-Hebrew University Medical Centre, Ein Kerem, Jerusalem 91120, Israel
*Corresponding author. E-mail: [email protected]
Background. Neuraxial analgesia significantly increases the success rate of external cephalic
version (ECV) among nulliparae. The study objective was to compare ECV success among multiparae with and without spinal analgesia.
Methods. Prospective randomized controlled trial performed over a pre-defined 6 yr period
in a tertiary referral delivery suite. Healthy multiparae at term requesting ECV for breech presentation, without fetal or uterine anomaly, were enrolled after written informed consent.
Women were randomized to receive either spinal analgesia (bupivacaine 7.5 mg) or no analgesia before the ECV. The primary outcome was successful conversion from breech to vertex
presentation, confirmed by ultrasound. Visual analogue pain score and adverse outcomes
(complications of anaesthesia or ECV) were recorded. Statistical analysis was performed
according to intention to treat using two-sided tests.
Results. Among 265 multiparae who underwent ECV, 65 consented to enrol, one subsequently refused ECV; therefore, data from 64 women were analysed. ECV was successful in
27 of 31 patients (87.1%) receiving spinal analgesia vs 19 of 33 (57.5%) with no analgesia
(P¼0.009; 95% CI of difference: 0.075 – 0.48). ECV with spinal analgesia reduced visual analogue
pain score, mean (SD) 1.7 (2.4) vs 5.5 (2.9) without (P,0.0001). Maternal hypotension was
seen after spinal analgesia in 10 of 31 (32%) (P¼0.0003) and easily treated without adverse
outcome. No complications were noted after the ECV.
Conclusions. Administration of spinal analgesia significantly increased the rate of successful
ECV among multiparae at term with increased patient comfort.
The trial was registered at the National Institute of Health Trials Registry, NCT00119184,
www.clinicaltrials.gov.
Br J Anaesth 2010; 104: 613–18
Keywords: anaesthesia, obstetric; anaesthetic techniques, subarachnoid; fetus
Accepted for publication: January 10, 2010
Almost 90% of women in developed countries with a
breech-presenting fetus at term are delivered by Caesarean
section because of concerns about fetal safety.1 This exposes
women to risks such as uterine rupture and placenta accreta
during subsequent deliveries.2 – 4 Furthermore, maternal and
fetal morbidity is lower with vertex vaginal delivery than
with Caesarean section.5 6 Since women with a breechpresenting fetus are more likely to have a breech-presenting
fetus during a subsequent pregnancy,7 it becomes important
to succeed with external cephalic version (ECV) in each
individual pregnancy in order to avoid unnecessary
Caesarean sections with their cumulative risk.8
Neuraxial analgesia significantly increases the chance of
successful ECV among nulliparae, thus an otherwise mandatory Caesarean delivery can potentially be avoided.9
The success rate of ECV among multiparae is known to be
higher (57 – 67%) than among nulliparae; therefore, neuraxial analgesia may not have the same positive impact on
the success rate of ECV among multiparae.10 11
†
An abstract of these data was presented as an oral presentation at
the Society of Maternal-Foetal Medicine 2009 Annual Meeting, San
Diego, CA, USA.
# The Author [2010]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved.
For Permissions, please email: [email protected]
Weiniger et al.
The primary aim of this randomized prospective controlled
trial was to examine the effect of spinal analgesia on the
success rate of ECV performed among multiparae at term.
Methods
This prospective randomized controlled trial was performed in a tertiary referral centre in the Post-Anaesthesia
Care Unit of a Labour and Delivery suite with about 5000
deliveries per year. After Institutional Review Board (IRB)
approval (January 20 – 25, 2002), all eligible multiparae
requesting ECV after 37 weeks gestation were approached
for recruitment before the ECV. Patients declining to participate were not offered spinal analgesia during their
ECV. This did not entail a deviation from routine practice
since neuraxial analgesia was not offered to women undergoing ECV during or before the study period.1
Women were randomized using numbered sealed envelopes containing concealed cards allocated at random by a
physician (H.Y.S.) not involved in study enrolment. Eligible
women were enrolled and written informed consent was
obtained before revealing the study assignment. The allocation cards designating patients to either receive or not
receive spinal analgesia were revealed either by the recruiting anaesthetist (C.F.W.) or by the obstetrician (Y.E.).
Inclusion criteria were: ASA status I–II, 37–40 complete
weeks gestation, no fetal abnormality (including intrauterine
growth restriction), no contraindication for vaginal delivery,
or no contraindication for regional analgesia. Exclusion criteria were: previous Caesarean section, previous myomectomy with uterine cavity penetration or uterine anomaly,
morbid obesity (BMI .40 kg m22), amniotic fluid index
,7 cm, neuropathy, severe back pain with radicular radiation, patient refusal of regional analgesia, poor communication, or request for elective Caesarean section (either after
failed ECV at another institution or not wishing to attempt
ECV). Department protocol required that women presenting
for ECV were fasted for 6 h before the procedure.
Uterine relaxation
Ritodrine 50 mg i.v. was used for uterine relaxation until
it became unavailable after April 2003 and was replaced
by nifedipine 20 mg orally. Patients were recumbent with
a 308 left lateral tilt facilitated by judicious placement of
pillows and underwent continuous non-invasive arterial
pressure (AP) and fetal heart rate monitoring once the
uterine relaxant was administered.
Intervention
After randomization and fetal heart rate monitor
(non-stress test), the spinal analgesia group received 1000
ml of Ringer’s lactate solution. In the sitting position,
single-shot spinal analgesia was performed using a pencilpoint needle. Plain bupivacaine, Kamacaine (Kamada Ltd,
Beit Kama, MP Negev, Israel), 7.5 mg was injected
intrathecally at L3– 4 or L4– 5. The patient then reclined
in a left lateral tilt. Once analgesia was attained (defined
as a sensory level of T6), the ECV was performed. The
senior obstetricians (Y.E. and U.E.) each had more than
5 yr experience with the ECV and were assisted by
another obstetrician. Conversion to vertex position was
attempted using a forward roll manoeuvre. Each attempt
was a separate manoeuvre and ultrasound surveillance was
used. The ECV was halted if any of the following
occurred: successful conversion, maternal request to stop
(indicative of pain), fetal bradycardia, or suspected placental abruption. If immediate emergency Caesarean section
was required, sodium citrate would be administered and
general anaesthesia could be performed if the block height
was insufficient. A maximum of three manipulations were
performed during the ECV attempt.
The control group received no analgesia but did receive
1000 ml of Ringer’s lactate solution before the ECV.
These women were monitored as for the spinal group, and
ECV was performed as described.
The two experienced ECV performing obstetricians
(Y.E. and U.E.) were not blinded. Data were obtained from
personal patient interviews before the ECV attempt, from
ultrasound examination, and from the senior obstetrician
(Y.E. or U.E.) performing the ECV. After delivery, data
were retrieved from computer files and telephone interviews.
Endpoints
The primary endpoint was successful ECV confirmed using
ultrasound. Secondary endpoints measured included pain
score during the procedure (visual analogue scale 0, no pain;
10, severe pain), complications associated with ECV or with
analgesia such as maternal hypotension (.20% reduction in
systolic AP), post-dural puncture headache, and failure of
analgesia, and successful vaginal delivery after successful
ECV. In addition, the inability to perform the ECV because
of maternal distress or discomfort was recorded, as was the
reason for halting the procedure (successful ECV, fetal
bradycardia, maternal discomfort).
Sample size
A one-sided hypothesis was used to estimate the sample
size in view of the direction of the outcome from previous
reports.12 13 Before the study, the ECV success rates in the
control (no analgesia) and spinal groups were estimated to
be 60% and 80%, respectively, among multiparae. A
sample size of 130 recruits was calculated for a power of
80% to detect a 20% difference in the ECV success rate
with an a priori one-sided a-level of 5%. Towards the end
of the predetermined 6 yr study period, an independent
bio-statistician (L.D.) was consulted with regard to the low
recruitment (64 patients). L.D. was blinded to treatment
allocation since study groups were masked during data
insertion, and performed a sample size recalculation.
Among the 64 study recruits, 46 of them (72%) were
614
Spinal analgesia for external cephalic version
found to be ECV success. Assuming that 50% of the subjects were randomized to each group, and that the ECV
success rate in the control group is 60%, the success rate
in the intervention group is 84%. On the basis of these
assumptions, a sample size of 64 subjects would provide
70% power to see an increase of 24% in the ECV success
rate with a one-sided a-level of 5%. Therefore, given the
slow recruitment and these calculations, it was decided to
terminate the study. The database was locked and the data
were analysed for 64 women.
Statistical analysis
Statistical analysis was performed using SPSS 14.0 (SPSS
Inc., Chicago, IL, USA). Results were analysed on an
intention-to-treat basis, comparing women randomized to
receive spinal analgesia with those receiving no analgesia.
All statistical tests were two-sided and a P-value of 5% or
less was considered statistically significant. Quantitative
variables were compared between the two study groups
using the independent samples t-test and are presented as
means and standard deviations. Categorical data were
compared between the study groups using the x2 test or
Fisher’s exact test and are presented as percentages.
Results
During the study period, 265 multiparae underwent ECV
at our institution and the rate of breech presentation
among all deliveries averaged 3.7% over the study period.
Sixty-five women requesting ECV consented to enrol in
the study, and 64 women were analysed (Fig. 1, Table 1).
The success rate of ECV among multiparae refusing to
participate in the study (therefore not receiving spinal
analgesia) was 65.2% (129/198).
All women received a uterine relaxant; the first 15
received ritodrine, then, when it was no longer available,
the remaining 49 received nifedipine. The success rate of
the ECV in patients with spinal analgesia was 11 of 15
(73.3%) when ritodrine was given, vs 35 of 49 (71.4%)
when nifedipine was given, P¼0.886 (Table 1). There
were no differences between these drugs with regard to
uterine tone by clinical assessment, ease of palpating the
fetal head, incidence of fetal bradycardia, or maternal
hypotension.
One patient randomized to receive spinal analgesia did
not receive the intended treatment as the anaesthetist was
unable to locate the dura (her ECV was unsuccessful but
she was analysed as intention to treat in the spinal
265 parturients eligible for
ECV procedure
200 refused to participate,
data on successful ECV outcome
available on 129/198 (65.2%)
65 randomly assigned
after consent, before
ECV performed
32 allocated to receive spinal
analgesia (study group)
Study allocation
33 allocated to receive no analgesia
(control group)
1 pt refused ECV after
randomization
31 analysed per study protocol for
success of ECV procedure
Analysis
Fig 1 CONSORT flowchart of trial profile.
615
33 analysed per study protocol for
success of ECV procedure
Weiniger et al.
analgesia group). ECV was significantly more successful
among women who were randomized to receive spinal, 27
of 31 (87.1%), vs no analgesia, 19 of 33 (57.5%)
(P¼0.009, 95% CI of the difference 0.075 – 0.48, Table 2).
There was no difference in the overall number of manipulations performed, mean (SD) 1.4 (0.8) in the spinal group
vs 1.3 (0.6) in the no analgesia group (P¼0.535).
Owing to the relatively small number of enrolled
women compared with those who declined to enrol, a sensitivity analysis was performed comparing the ECV
success rate in the non-analgesia control group of our
study with those not enrolled. Therefore, we compared 19
of 33 (57.5%) with 129 of 198 (65.2%) and found no statistically significant difference between these two rates,
P¼0.401 (x2 test). Two women with breech presentation
in consecutive pregnancies were enrolled twice in the
Table 1 Characteristics of treatment groups presented as mean (range), mean
(SD) or number (%). US, ultrasound measurement
Spinal analgesia
(n531)
Maternal age (yr)
28.5 (21– 40)
Gestational age (weeks)
38.1 (0.9)
Weight at the time of ECV (kg)
73.6 (10.3)
Height (cm)
163.3 (6.1)
Parity
1
13 (41.9%)
2
7 (22.6%)
3
4 (12.9%)
4
5 (16.1%)
5
1 (3.2%)
8
1 (3.2%)
Breech in past pregnancy
5 (16.1%)
Estimated fetal weight
3142.5 (422.4)
(ultrasound) (g)
Amniotic fluid index (cm)
13.4 (4.5)
Placental position (US)
Anterior
11/30 (37%)
Posterior
10/30 (33%)
Fundal
9/30 (30%)
Fetal presentation (US)
Frank breech
7/31 (22.6%)
Complete breech
24/31 (77.4%)
Position of fetal spine
Anterior
17/28 (60.7%)
Posterior
2/28 (7.1%)
Lateral
9/28 (32.1%)
Tocolytic before ECV attempt
Ritodrine
7 (22.6%)
Nifedipine
24 (77.4%)
No analgesia
(n533)
28.6
38.2
71.4
164.3
(20 –36)
(1.1)
(10.0)
(5.0)
21
5
3
4
0
0
3
3048.8
(63.6%)
(15.2%)
(9.1%)
(12.1%)
(0%)
(0%)
(9.1%)
(349.5)
12.9 (4.3)
17/32 (53%)
8/32 (25%)
7/32 (22%)
7/33 (21.2%)
26/33 (78.8%)
20/32 (62.5%)
4/32 (12.5%)
8/32 (25.0%)
8 (24.2%)
25 (75.8%)
current study. A further analysis without these women was
performed to exclude potential bias for the primary
outcome. The success of ECV with spinal analgesia
excluding the repeat patients was 23 of 27 (85.1%) vs 19
of 33 (57.5%) without analgesia (P¼0.02).
There was no difference according to parity for the rate
of successful ECV within the intention-to-treat groups.
There were no complications such as abruption or
unplanned delivery after ECV (Table 3). Maternal hypotension occurred in 10 cases (32%) after spinal analgesia,
and in all cases, it was successfully managed by left
lateral tilt and a bolus of up to 10 mg i.v. ephedrine. One
case of transient fetal heart rate deceleration occurred
immediately after the spinal analgesia (despite normal
maternal AP), and recovered after administration of ephedrine 10 mg. No case of post-dural puncture headache was
reported.
Vaginal delivery after successful ECV occurred in 25 of
27 (92.6%) spinal group women and 19 (100%) in the
control group (Table 2). Three women in the control group
had a breech vaginal delivery. In addition, 10 women had
vertex vaginal delivery, despite failed ECV in the study.
Seven women presented again for ECV, which included
the use of spinal analgesia, either to our or another institution, two women had spontaneous conversion of the
fetus to vertex position, and for one woman, no details
were available.
Discussion
In this study, we found that administration of spinal
analgesia significantly increased the success rate of ECV
among multiparae, in addition to reducing the pain experienced during the procedure.
Bupivacaine 7.5 mg spinal dose used in the current
study may be considered excessive, particularly in view of
the frequency of hypotension compared with our previous
study using the same spinal dose.9 However, the choice of
drug dose administered is relevant for determining the
success of ECV. Randomized administration of a lower
dose of bupivacaine 2.5 mg plus opioid demonstrated a
low ECV success rate, similar to our non-analgesia control
group.12 14 15 Conversely, after administration of surgical
epidural analgesia to a sensory level of T6– T10, ECV
success rates were significantly increased compared with
Table 2 Outcome and findings after ECV, presented as n (%) or mean (SD). *Statistically significant difference analysed by t-test or Fisher’s exact test as
appropriate
Successful version to vertex position
Vaginal delivery
Uterine tone relaxed
Easy palpation of fetal head
VAS pain score (0 –10)
Spinal analgesia
group (n531)
No analgesia
group (n533)
P-value (95% CI of the
difference)
27/31 (87.1%)
27/31 (87.1%)
24/31 (77.4%)
28/31 (90.3%)
1.7 (2.4)
19/33 (57.6%)
30/33 (90.9%)
26/29 (89.7%)
24/33 (72.7%)
5.5 (2.9)
0.012*
0.7039
0.110
0.172
,0.0001*
616
(0.075 – 0.48)
(20.22 to 0.14)
(20.31 to 0.07)
(20.019 to 0.36)
(25.16 to – 2.49)
Spinal analgesia for external cephalic version
Table 3 Adverse outcomes after ECV, presented as number (%).
*Significantly analysed using Fisher’s exact test
Spinal
analgesia
(n531)
Transient fetal
bradycardia before
ECV
Maternal
hypotension
(systolic AP ,20
mm Hg baseline)
Transient fetal
bradycardia after
ECV
Control
group
(n533)
1 (3.2%)
0 (0)
10 (32.3%)
0 (0)
2 (6.5%)
1 (3.0%)
P-value (95% CI of the
difference)
0.48 (20.075 to 0.16)
0.0003* (0.15 –0.49)
0.61 (20.09 to 0.18)
control groups with no anaesthesia.16 17 The rate of hypotension may not be a consequence of the spinal dose,
since similar incidence of hypotension was seen after
administration of bupivacaine 2.5 mg.14 In the current
study, this was easily treated with ephedrine with no
adverse sequelae.
The current study also demonstrated that spinal analgesia significantly improved patient comfort. Spinal analgesia may contribute to successful ECV. Sullivan and
colleagues14 randomized patients to receive ECV after
administration of either spinal bupivacaine 2.5 mg or i.v.
fentanyl and found that although women receiving spinal
analgesia experienced significantly less pain, the ECV
success rate was not higher than the i.v. analgesia control.
In non-randomized studies, the success rate of ECV was
increased by using neuraxial analgesia after failed ECV
without analgesia.13 18 However, increased number of
manipulations may be associated with greater overall force
applied to the fetus.19 Multiple manipulations may cause
subclinical fetal injury or release of fetal DNA into
maternal circulation.20
Previous studies demonstrating increased success of
ECV after neuraxial analgesia did not differentiate multiparae and nulliparae.14 – 17 A previous study among nulliparae
demonstrated that spinal analgesia significantly increases
their success rate of ECV;9 however, this result cannot be
extrapolated to multiparae due to their higher baseline
success rate for the ECV. Interestingly, multiparae in the
current study considered that their chance of success with
ECV was high without spinal; therefore, many multiparae
declined to enrol in the current study. Practising obstetricians
may also consider that there is no need to request spinal
analgesia for multiparae undergoing ECV. Importantly, this
study demonstrates that multiparae also benefit from spinal
analgesia for ECV.
Block height was designed for excellent ECV conditions16 – 18 and not for emergency Caesarean surgery
which may be needed in 0.5% or fewer.21 In the case of
severe fetal compromise after ECV with inadequate spinal
anaesthesia level, general anaesthesia could have been performed safely in these patients who were all fasted.
The current study was limited by the smaller than
planned study sample size. The blinded power and sample
size re-assessment supported the decision to terminate
recruitment at the predetermined date, and the difference
between the study groups attained statistical significance
even after sensitivity analysis and adjustment for potential
prognostic factors. The experienced obstetricians were not
blinded, and this is a potential bias seen also in other ECV
studies where analgesia was used.12 13
Future studies should consider the benefit of immediate
Caesarean delivery after failed ECV. In order to avoid the
beginning of labour before attempting to convert the fetal
presentation by ECV, the procedure was usually performed
by 38 weeks. However, among women who failed ECV,
even under spinal analgesia, delivery was delayed to at
least 39 weeks in order to minimize the risk for neonatal
respiratory disorders.22
In conclusion, this prospective randomized controlled
trial of multiparae with a breech-presenting fetus demonstrated that spinal analgesia increased the success rate of
ECV to allow vaginal delivery. In addition, spinal analgesia provided enhanced patient comfort.
Acknowledgement
Lisa Deutsch (L.D.), PhD, of BioStats-Statistical Consulting Services,
Israel, performed the blinded analyses of sample size re-assessment and
other statistical analyses.
Conflict of interest
None declared.
Funding
This work was supported by grants from the Chief
Scientist Office of the Ministry of Health, Israel (grant no.
6189), and the Hadassah-Hebrew University Medical
Centre Women’s Health Research Fund.
References
617
1 ACOG Committee on Obstetric Practice. ACOG Committee
Opinion No. 340. Mode of term singleton breech delivery. Obstet
Gynecol 2006; 108: 235 – 7
2 Silver RM, Landon MB, Rouse DJ, et al. National Institute of Child
Health and Human Development Maternal-Foetal Medicine Units
Network. Maternal morbidity associated with multiple repeat
cesarean deliveries. Obstet Gynecol 2006; 107: 1226 – 32
3 Kennare R, Tucker G, Heard A, Chan A. Risks of adverse outcomes in the next birth after a first cesarean delivery. Obstet
Gynecol 2007; 109: 270 – 6
4 ACOG Committee Opinion No. 394, December 2007. Cesarean
delivery on maternal request. American College of Obstetricians
and Gynecologists. Obstet Gynecol 2007; 110: 1501
5 Jackson N, Paterson-Brown S. Physical sequelae of caesarean
section. Best Pract Res Clin Obstet Gynecol 2001; 15: 49 – 61
6 Kolas T, Saugstad OD, Daltveit AK, Nilsen ST, Oian P. Planned
cesarean versus planned vaginal delivery at term: comparison of
Weiniger et al.
7
8
9
10
11
12
13
14
newborn infant outcomes. Am J Obstet Gynecol 2006; 195:
1538– 43
Albrechtsen S, Rasmussen S, Dalaker K, Irgens LM. Reproductive
career after breech presentation: subsequent pregnancy rates,
interpregnancy interval, and recurrence. Obstet Gynecol 1998; 92:
345 –50
Nisenblat V, Barak S, Griness OB, Degani S, Ohel G, Gonen R.
Maternal complications associated with multiple cesarean deliveries. Obstet Gynecol 2006; 108: 21– 6
Weiniger CF, Ginosar Y, Elchalal U, Sharon E, Nokrian M, Ezra Y.
External cephalic version for breech presentation with or
without spinal analgesia in nulliparous women at term: a randomized controlled trial. Obstet Gynecol 2007; 110: 1343 – 50
Salim R, Zafran N, Nachum Z, Edelstein S, Shalev E. Employing
nifedipine as a tocolytic agent before external cephalic version.
Acta Obstet Gynecol Scand 2008; 87: 434 –7
Ezra Y, Elram T, Plotkin V, Elchalal U. Significance of success rate
of external cephalic versions and vaginal breech deliveries in
counseling women with breech presentation at term. Eur J Obstet
Gynecol Reprod Biol 2000; 90: 63 –6
Macarthur AJ, Gagnon S, Tureanu LM, Downey KN. Anesthesia
facilitation of external cephalic version: a meta-analysis. Am J
Obstet Gynecol 2004; 191: 1219 – 24
Birnbach DJ, Matut J, Stein DJ, et al. The effect of intrathecal
analgesia on the success of external cephalic version. Anesth
Analg 2001; 93: 410 – 3
Sullivan JT, Grobman WA, Bauchat JR, et al. A randomized controlled trial of the effect of combined spinal– epidural analgesia
15
16
17
18
19
20
21
22
618
on the success of external cephalic version for breech presentation. Int J Obstet Anesth 2009; 18: 328 – 34
Dugoff L, Stamm CA, Jones OW, III, et al. The effect of spinal
anesthesia on the success rate of external cephalic version: a randomized trial. Obstet Gynecol 1999; 93: 345 – 9
Schorr SJ, Speights SE, Ross EL, et al. A randomized trial of epidural anesthesia to improve external cephalic version success. Am
J Obstet Gynecol 1997; 177: 1133 – 7
Mancuso KM, Yancey MK, Murphy JA, Markenson GR. Epidural
analgesia for cephalic version: a randomized trial. Obstet Gynecol
2000; 95: 648 – 51
Cherayil G, Feinberg B, Robinson J, Tsen LC. Central neuraxial
blockade promotes external cephalic version success after a
failed attempt. Anesth Analg 2002; 94: 1589 – 92
Leung TY, Sahota DS, Fok WY, Chan LW, Lau TK. External cephalic version induced foetal cerebral and umbilical blood flow
changes are related to the amount of pressure exerted. Br J
Obstet Gynaecol 2004; 111: 430 – 5
Lau TK, Lo KW, Chan LY, Leung TY, Lo YM. Cell-free foetal deoxyribonucleic acid in maternal circulation as a marker of foetal–
maternal hemorrhage in patients undergoing external cephalic
version near term. Am J Obstet Gynecol 2000; 183: 712– 6
Collins S, Ellaway P, Harrington D, Pandit M, Impey LW. The complications of external cephalic version: results from 805 consecutive attempts. Br J Obstet Gynaecol 2007; 114: 636– 8
Hansen AK, Wisborg K, Uldbjerg N, Henriksen TB. Risk of respiratory morbidity in term infants delivered by elective caesarean
section: cohort study. Br Med J 2008; 336: 85 – 7