Content uploaded by Doreth A M Teunissen
Author content
All content in this area was uploaded by Doreth A M Teunissen on Nov 26, 2014
Content may be subject to copyright.
Birthing positions De Jonge et al.
Journal of Psychosomatic Obstetrics & Gynecology 35
March 2004J Psychosom Obstet Gynecol 2004;25:35–45
Supine position compared to other
positions during the second stage
of labor: A meta-analytic review
A. De Jonge, T. A. M. Teunissen and
A. L. M. Lagro-Janssen
A. De Jonge, T. A. M.
Teunissen and A. L. M.
Lagro-Janssen, Department
of General Practice and
Social Medicine, Nijmegen
University, The Netherlands
*Correspondence to: A. Lagro-Janssen, Department of General Practice and Social Medicine, Nijmegen University, UMC St.
Radboud, HSV-229, Postbus 9101, 6500 HB, Nijmegen, The Netherlands. Email: A.Lagro-Janssen@hag.umcn.nl
The routine use of the supine position during the second stage of labor can be
considered to be an intervention in the natural course of labor. This study aimed
to establish whether the continuation of this intervention is justified. Nine
randomized controlled trials and one cohort study were included. A meta-analysis
indicated a higher rate of instrumental deliveries and episiotomies in the supine
position. A lower estimated blood loss and lower rate of postpartum hemorrhage
were found in the supine position, however it is not clear whether this is a real or
only an observed difference. Heterogenous, non-pooled data showed that women
experienced more severe pain in the supine position and had a preference for
other birthing positions.
Many methodological problems were identified in the studies and the
appropriateness of a randomized controlled trial to study this subject is called into
question. A cohort study is recommended as a more appropriate methodology,
supplemented by a qualitative method to study women’s experiences. Objective
laboratory measurements are advised to examine the difference in blood loss.
In conclusion, the results do not justify the continuation of the routine use of
the supine position during the second stage of labor.
Key words: supine, upright, lateral birthing position, birth experience
INTRODUCTION
Before the 17th century the upright birthing
position was common in western countries
1,2
.
Women only started to adopt the supine
position on a large scale when obstetric
instruments were introduced such as the
delivery forceps
3
. The supine position became
popular because of the convenience for health
professionals rather than the benefits for
women
4
. In countries where western health-
care has not had much influence, the upright
position is still very common
3,5,6
. The wide-
spread use of the supine position during the
second stage of labor, even for women who
do not need an instrumental delivery, can be
considered as an intervention in the natural
course of labor. The delivery bed can there-
fore be regarded as a midwifery instru-
ment
7
. Based on experiences in non-western
countries and in western countries before the
17th century, it can be assumed that women
will not only lie down during the second stage
of labor, if they feel free to use other positions.
Studies have confirmed that women use
various positions, supine and non-supine, if
they are left to choose
6,8,9
. The supine position,
however, has become so common that neither
healthworkers nor women now regard this as
an intervention
3
.
In the last few decades of the 20th century,
alternatives to the supine position have
gained some popularity. The invention of
Doppler ultrasound transducers has made it
easier to listen to the fetal heart when the
woman is in non-supine positions
10,11
. Physical
© 2004 Parthenon Publishing. A member of the Taylor & Francis Group
DOI: 10.1080/01674820410001737423
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.
Birthing positions
De Jonge et al.
36 Journal of Psychosomatic Obstetrics & Gynecology
benefits supposedly associated with non-
supine positions are increased uterine
pressure, more effective bearing down efforts,
improved fetal positioning, reduced risk of
aorto-caval compression and increased
diameters of the pelvis
3,5,12–15
. Psychological
benefits that have been ascribed to upright
positions include reduced experience of
pain, increased feeling of being in control,
communication with the delivery attendant
on a more equal basis and more active
involvement of the woman’s partner
7,16,17
.
These aspects are important in view of the
increasing emphasis on the autonomy of
women and on a positive birth experience
18
.
Expecting women to adopt one particular
position, whether supine or non-supine,
during the second stage of labor, can only be
justified if there is good evidence that this
has important advantages for the health of
either the mother or the baby
19
.
A meta-analysis has been conducted into
the benefits and risks of different positions
during the second stage of labor
20
. The
authors conclude that the use of any upright
or lateral position, compared with a supine
or lithotomy position, was associated with a
reduced duration of the second stage of labor,
a reduced reporting of severe pain and a
reduction in assisted deliveries, abnormal
fetal heart rate patterns and episiotomies. On
the other hand, they found an increase in
second degree tears and an increased risk of
blood loss of more than 500 ml. Randomized
controlled trials (RCTs) were included in
which lateral or lateral tilt and supine posi-
tions were combined as recumbent positions.
In this article it is not assumed that there
is one superior position for the second stage
of labor. However, since the supine position
is often used routinely in western countries
3
,
this meta-analysis aimed to establish the
benefits of this intervention in the light of
evidence-based medicine. Studies included
should therefore compare the supine position
to the use of one other or several positions.
The key question was: what are the benefits
for women of the routine use of the supine
position for the second stage of labor com-
pared to other positions, in terms of maternal
morbidity and comfort and the morbidity of
the baby?
METHODS
Formulation of the problem
This meta-analysis focused on women in the
second stage of labor who were expected to
have a vaginal birth. The onset of the second
stage was defined as full dilatation of the
cervix or from the time of expulsive effort if
full dilatation was not established. The supine
position was defined as the woman lying on
her back, supported with pillows or a bed rest
to a maximum of 45° from the horizontal. If
authors did not specify when exactly the
position had been used, it was assumed that
it had been adopted during most of the
second stage. Randomized controlled trials
were included as well as case-control and
cohort studies. The following outcomes for
the mother were included: medical inter-
ventions for failure to progress, trauma to the
birth canal, estimated or measured blood loss,
postpartum hemorrhage (more than 500 ml),
hemoglobin levels after delivery, incon-
tinence of urine or faeces, pelvic pain or
instability and the mother’s satisfaction with
the birth experience including perception of
pain. Trauma to the birth canal was defined
as: intact perineum, first, second or third
degree tear. For the child the following out-
comes were included: abnormal fetal heart
rate patterns, Apgar scores, mean umbilical
cord artery pH and the need for neonatal
resuscitation.
Although in many trials the duration of
the second stage has been compared between
groups, this criteria was not included. The
onset of the second stage is very arbitrary.
Some take full dilatation as the onset, others
the start of active pushing. In addition, it is
questionable whether the duration of the
second stage is a clinically important variable.
It is more important whether intervention
was needed because of a delay in the progress
of the second stage; therefore this criteria was
chosen instead.
Findings of studies in which health profes-
sionals appear to be unconfident in assisting
births in non-supine positions could easily
have been biased and such studies were
therefore excluded. In some articles the
inexperience of professionals was described
by the authors, in others it was clear that the
use of a new birthing position had been
introduced at the start of the study with
which the professionals were unfamiliar.
While developing the protocol it was
decided to perform subanalyses on supine vs.
upright positions and on supine vs. lateral
positions, on primigravidas and multigravi-
das, and on inclusion or exclusion of women
who had used oxytocin infusion or epidural
anaesthesia.
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.
Birthing positions De Jonge et al.
Journal of Psychosomatic Obstetrics & Gynecology 37
Search strategy
Literature was searched between 1 February
2001 and 31 March 2002 from 1966 (or from
the earliest available date) onwards via
Medline, Embase, The Cochrane Library
(including the CENTRAL/CCTR database), Web
of Science, Cinahl, Midirs (Midwifery Data-
base), and Picarta (keywords and related
articles). The following trial registers have
been contacted: National Institutes of Health
Inventory of Clinical Trials and Studies,
Colombia Registry of Clinical Trials and Inter-
national Registry of Perinatal Trials. Some
articles were found via reference lists from
other studies. From the articles that were
initially found, a cited reference search was
done. The search was limited to the English,
German, French and Dutch languages. Key-
words used were: delivery, birth, birthing,
bearing down, pushing, upright and position,
birth(ing) stool/chair/cushion, second and
stage and labor.
Inclusion criteria
Initially, 46 studies were found, two of which
were unpublished trials. All studies that were
excluded and the reason for exclusion are
given in Table 1. In 12 studies the position or
outcomes did not meet the criteria of the
protocol. For example, in some studies the
supine position also included a lateral tilt. For
four other trials only an abstract was found.
Several attempts were made to obtain more
details about these studies from the authors,
their places of work and publishers but these
efforts were unsuccessful. Two of the
researchers (A. De Jonge and D. Teunissen)
assessed whether the professionals in the
remaining studies were competent in the
management of labor in all the positions that
Table 1 Studies excluded from the review
Study (first autor) Reason for exclusion
Aarnoudse (1984)
39
No outcome measures in line with protocol
Aikins Murphy (1998)
40
Position not clearly defined
Allahbadia (1992)
41
Professionals unfamiliar with squatting position
Bastian (1994)
42
Cohort with quality mark 1
Bhardwaj (1994)
43
Only abstract found
Bomfim-Hyppolito (1998)
44
RCT with quality mark 3
Chan (1963)
45
RCT with quality mark 3.5
Crowley (1991)
46
More senior midwives in birth chair group, medical students only
involved in recumbent deliveries
Drähne (1982)
47
Only abstract found
Gardosi (1989a)
38
Professionals unfamiliar with upright position
Gardosi (1989b)
48
Semi-recumbent and lateral position combined in one group
Gåreberg (1994)
49
Comparison of two upright positons
Golay (1993)
11
Cohort with quality mark 1.75
Gupta (1989a)
50
Professionals unfamiliar with squatting position
Hagymasy (1998)
36
Cohort with quality mark 1.5
Hemminki (1986)
51
Professionals unfamiliar with birth chair
Kafka (1994)
52
Cohort with quality mark 1.25
Kleine-Tebbe (1996)
53
Control group includes other than supine position
Liddell (1985)
54
RCT with quality mark 3.5
Liu (1974)
55
Semi-upright position defined as 30° from horizontal
Liu (1989)
56
Semi-upright position defined as 30° from horizontal
McManus (1978)
57
Control group adopted lateral recumbent position
Moll (1985)
58
Cohort with quality mark 1
Nodine (1987)
59
Cohort with quality mark 1
Olson (1990)
60
Cohort with quality mark 0.75
Racinet (1999)
61
Professionals unfamiliar with squatting position
Radkey (1991)
62
Only abstract found
Roberts (1984)
63
Position not well defined
Rohrbacher (1998)
64
Position not well defined
Romney (1984)
65
Professionals not familiar with birth chair
Schneider-Affeld (1982)
66
Only abstract found
Shannahan (1985)
67
Cohort with quality mark 0.75
Shannahan (1989)
68
Cohort with quality mark 0.75
Shorten (2002)
69
Cohort with quality mark 1.5
Stewart (1989)
70
Supine position includes lateral tilt
Van Diem (2002)
37
Supine position includes lateral position
RCT, randomized controlled trial
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.
Birthing positions
De Jonge et al.
38 Journal of Psychosomatic Obstetrics & Gynecology
occurred in the study and unanimously
decided to exclude seven studies. For one
researcher (D. Teunissen) the studies were
blinded for evaluation. (She was only given
the information that was essential for
assessing the trial; names of authors, journals
and institutions, places where studies were
undertaken and years of publication were
removed.) The remaining 23 studies, 13 RCTs
and 10 cohort studies, were assessed using a
quality criteria list based on the Delphi-list
21
(Table 2).
The two researchers scored the studies
independently. The maximum score given to
RCTs was 9 and for cohort studies was 3. In
seven studies D. Teunissen scored 0.5 point
higher than A. De Jonge and in one study a
point higher. In another study, A. De Jonge
scored 0.5 point higher than D. Teunissen.
Table 3 lists all the studies included in this
meta-analysis with the average quality score
for each. The cut-off point for inclusion for
RCTs was a score of ≥ 4 and for cohort studies
≥ 2. This low cut-off point was taken because
almost all studies scored 0 on all the items
concerning blinding. Only one study scored
0.5 on the item of blinding of the outcome
assessor because an independent person
recorded hemoglobin levels and women’s
experiences but the attending midwife
assessed other outcomes. Nevertheless, the
items about blinding remained in the list to
indicate their impor-tance and to show the
loss of quality because these criteria could not
be met. Only in one cohort study
11
the
difference in scoring led to disagreement
about whether to include the trial or not. This
was resolved by asking the opinion of a third
researcher (A. Lagro-Janssen) after which the
study was excluded. After the quality
assessment, nine RCTs and one cohort study
were finally included in the review. The
maximum score for method-ological quality
of the RCTs was 6.75 out of 9 and the median
was 4.5. The only cohort study scored 2.
Statistical analysis
A meta-analysis was performed on the nine
RCT’s for all physical outcomes, using a ran-
Table 2 Quality criteria list (modified Delphi list)
21
A. Was randomization in a randomized controlled trial conducted in a concealed manner?
Randomization and concealed (computer, table with random numbers) = 1, randomized but not
concealed (hospital number, date of birth, length, alternation) = 0.5, not randomized or not clear how
randomization was performed = 0
B. Were groups similar at baseline regarding the most important prognostic indicators?
• Parity
• Maternal age
• Induction or augmentation with prostin and/or oxytocin infusion
• Epidural or pethidine for pain relief
• Birthweight
If groups are similar for all but one indicators = 1, if more than one indicator is not mentioned where it
should have been = 0.5, if more than 2 are not mentioned = 0, if for 1 or more indicators there is a
significant difference between the groups (p < 0.05) = 0.
C. Were the eligibility criteria specified (for inclusion in the trial as well as for the inclusion in either study
or the control group)?
Clearly specified = 1, partly described = 0.5, not specified = 0
If the supine group includes women in lateral position, and it has not been specified how many women
adopted this position, the study should be excluded.
D. Was the outcome assessor blinded?
E. Was the care provider blinded?
F. Was the patient blinded?
G. Were point estimates and measures of variability presented for the primary outcome measures?
Yes = 1, partly so = 0.5, no = 0
H. Did the analysis in an RCT include an intention to treat analysis?
Yes = 1, doubtful = 0.5, no = 0
I. Is the compliance rate (in each group) in an RCT unlikely to cause bias?
If > 80% have adopted the allocated position (in each group) = 1, 70-80% = 0.5, <70% or not known
= 0
Per item the score is 1 if the criteria has been satisfied, 0 if not and 0.5 if partly satisfied.
Randomized controlled trial: maximum score 9: included if score is ≥ 4
Cohort study: only B, C and G are scored: maximum score 3, included if score is ≥ 2.
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.
Birthing positions De Jonge et al.
Journal of Psychosomatic Obstetrics & Gynecology 39
dom effects model. The RevMan software was
used which was developed by the Cochrane
Collaboration
22
. Analyses were performed
according to the random effects model.
Heterogenous outcomes, dealing with the
experiences of the mother, and those from
the cohort study were not pooled together
and were described separately.
Odds ratios were given for categorical data
and weighted mean differences for con-
tinuous data. The p-values were based on the
normal Z-test. Statistical significance was
defined as p < 0.05.
RESULTS
The compliance rate varied greatly between
the studies (from 49.3% to 100%). Some studies
did not include standard deviations for
continuous variables and could therefore not
be included in the meta-analyses for these
outcomes. Only one birthing position, at the
time of the actual birth, was mentioned. It
was not always clear whether this position
had been adopted throughout the entire
second stage. Table 4 shows the outcomes of
the meta-analysis.
The supine position compared with other
positions was associated with an increased
rate of instrumental deliveries. In the supine
position there was a decreased estimated
blood loss and the incidence of postpartum
hemorrhage was also decreased. Both these
differences were only significant for multi-
gravidas and when supine and upright
positions were compared. Only two studies
compared women in supine position to those
with a lateral tilt
23,24
. Only one of these looked
at the difference in estimated blood loss
(without analysing separately for primi-
gravidas and multigravidas) and found no
significant difference between the two
groups
24
. There was no significant difference
in the requirement for a blood transfusion.
In only one study in which postpartum
hemorrhage was an outcome measure,
neither oxytocin nor epidural infusion were
used during the first stage of labor
25
. Women
in this study were randomized in a supine and
a squatting group. In this study, which was
also the best quality study in the review, the
lowest incidence of postpartum hemorrhage
was reported and there was no difference
between the two groups.
An increase in episiotomies was found in
the supine position. There was a tendency
towards a decrease in second degree tears but
this was not significant. When episiotomies
and second degree tears were combined, to
give an impression of perineal damage in
need of suturing, the rate was higher in the
supine position (almost statistically signifi-
cant) (p = 0.05). The incidence of third degree
tears was not reported in any of the studies.
In the cohort study, a higher rate of
episiotomies was found in lithotomy and
semi-sitting (25–45°) position compared to
alternative positions (34% and 11% vs. 9%)
26
.
The relative risk of an episiotomy in an
alternative position was 0.59 (confidence
interval, CI 0.37, 0.93, p < 0.02). There was no
significant difference in perineal tears
between the three groups.
There were no significant differences in
Apgar scores, abnormal fetal heart rate
patterns or requirement of neonatal resuscita-
tion. The difference in mean artery pH of –0.02
in supine position was almost statistically
significant (p = 0.05). Four studies measured
aspects of women’s birth experience (see Table
Table 3 Studies included in the review
Study Quality
Study (first author) design mark n
a
Position (n
b
)
Chen (1987)
29
RCT 5.25 116 supine (43) vs. birthing chair (73)
De Jong (1997)
25
RCT 6.75 517 supine (260) vs. squatting on step stool (257)
Hillan (1984)
31
RCT 4.5 500 supine (250) vs. chair (250)
Humphrey (1973)
23
RCT 4.0 40 supine (20) vs. lateral tilt (20)
Johnstone (1987)
24
RCT 4.25 58 supine (30) vs. lateral tilt (28)
Lydon-Rochelle (1995)
26
Cohort 2.0 393 supine (197) vs. other positions (196)
Marttila (1983)
28
RCT 4.5 100 supine (50) vs. chair made from bed (50)
Stewart (1983)
30
RCT 4.5 189 supine (90) vs. chair (94)
Turner (1986)
32
RCT 4.0 636 supine (370) vs. chair (266);
for analysis 313 vs. 226 (no intention to treat
analysis)
Waldenström (1991)
27
RCT 5.0 294 supine (146) vs. birthing stool (148)
a
total number of women in the study;
b
number of women in the sub-group; RCT, randomized controlled
trial
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.
Birthing positions
De Jonge et al.
40 Journal of Psychosomatic Obstetrics & Gynecology
5)
25,27–9
. They all used very different methods.
Three studies asked women about the expe-
rience of pain and about the satisfaction with
the birth
25,27,28
. De Jong et al. found a significant
trend towards women reporting more signifi-
cant pain in the supine position
25
. Marttila
et al. stated that more women reported
intolerable pain in supine position
28
. They did
not describe how and by whom women were
asked about this. Waldenstrøm and Gottvall
asked women to indicate their pain on a scale
from 1 to 10 and found an average of 7.6 for
women in supine position as opposed to 6.9
for women on a birthing stool
27
.
De Jong et al. also asked women about their
satisfaction by letting them choose between
five options
25
. There was no significant differ-
ence in satisfaction between the two groups.
Marttila et al. found that more women re-
ported the delivery to be an unpleasant
experience in the supine group compared to
the half-sitting group
28
. Waldenstrøm and
Gottvall
27
found a significant trend for
women on the birthing stool to have a better
experience than those in supine position. No
women in either group reported to have a very
bad experience.
Chen et al. asked women how easy it was
to bear down
29
. More nulliparas reported
difficulty in bearing down in the supine
position compared to those in the birthing
chair. Fewer multiparas in the supine position
said they found it easier to bear down than
during the previous delivery, which would
probably have been in supine position.
Two studies asked women in which
position they would like to give birth next
time
27,28
. In Marttila et al.’s study the majority
of women in both groups would like to give
birth in a half-sitting rather than supine
position next time
28
. In Waldenstrøm and
Gottvall’s study fewer women in the supine
group would like to use the same position
next time compared to the birthing stool
group and when women who actually gave
birth in the allocated position were compared
the difference was much larger (49% vs. 90%)
27
.
The other outcomes mentioned in the
protocol were not measured in the studies in
this review.
DISCUSSION
The approach taken in this meta-analyisis was
different from the one conducted by Gupta
and Nikodem
20
. The outcomes are not very
different from their comparison between
supine or lithotomy position and upright or
lateral position. We too, found an increase in
instrumental deliveries and epsiotomies in
the supine position compared to other
positions and a decrease in blood loss and
postpartum hemorrhage. The decrease in
second degree tears in supine position found
in their study did not reach significance in
ours (p = 0.09). They observed more abnormal
fetal heart rate patterns in the supine
position, whereas we did not. In both meta-
analyses, no differences in low Apgar scores
or neonatal resuscitation were found. The
Table 4 Outcomes of meta-analysis. Supine vs. non-supine positions: maternal and perinatal morbidity
Supine Not supine OR or
Outcome Studies n
a
or o/n
b
n
a
or o/n
b
WMD 95% CI p-value
Instrumental delivery
24, 25, 27, 28, 30–32
130/1139 88/1053 1.37 [1.03, 1.84] 0.03
EBL
24, 30, 31
370 372 –58.98 ml [–88.55, –29.41] 0.00009
EBL supine vs. upright
30,31
340 344 –71.63 [–107.70, –35.57] 0.0001
EBL supine vs. lateral tilt
24
30 28 –33.00 [–84.67, 18.67] n.s
EBL primigravida’s
30, 31
161 165 –43.07 [–101.95, 15.81] 0.15
EBL multigravida’s
30, 31
179 179 –92.04 [–134.58, –49.51] 0.00002
PPH (> 500 ml)
25, 27, 30–32
53/1017 90/943 0.52 [0.36, 0.75] 0.0004
Bloodtransfusion
25, 27
1/406 4/405 0.35 [0.05, 2.30] 0.3
Episiotomy
24, 25, 27, 30–32
394/1089 252/1003 1.73 [1.20, 2.50] 0.003
Second degree tear
25, 27, 30,
31
111/746 139/749 0.74 [0.52, 1.04] 0.09
Episiotomy and second
degree tear
25, 27, 30,
31
371/746 297/749 1.56 [0.99, 2.45] 0.05
Abnormal fetal heart pattern
25, 28
36/310 32/307 1.52 [0.30, 7.59] 0.6
Apgar ≤ 7 at 1 min
24, 25, 28
17/340 11/335 1.38 [0.59, 3.23] 0.5
Apgar ≤ 7 at 5 mins
24, 28, 32
3/393 1/304 1.85 [0.27, 12.79] 0.5
Mean artery pH
23, 24,29
93 121 –0.02 [–0.05, 0.00] 0.05
Neonatal resuscitation
24, 25
16/290 12/285 1.32 [0.61, 2.86] 0.5
a
total number of subjects in the subgroup (i..e. supine or not supine);
b
number with outcome of categorical variable/total number of subjects in the subgroup; OR, odds ratio;WMD, weighted mean difference;
CI, confidence interval; EBL, estimated blood loss; PPH, postpartum hemorrhage; n.s, non-significant
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.
Birthing positions De Jonge et al.
Journal of Psychosomatic Obstetrics & Gynecology 41
inclusion by Gupta and Nikodem
20
of studies
in which professionals appeared not to be
confident in assisting women in all the
positions which occurred, did not produce
different outcomes to our meta-analysis
which excluded these studies. Inexperience
with births in different positions should
therefore not be an argument against
allowing women to use them. We found no
studies which investigated the use of various
positions by women during the second stage.
Many methodological problems were
observed in the studies. The exclusion rate of
‘unsuitable participants’ was not always given
but appeared to be considerable in some
studies. One reason for exclusion was a
preference of the woman for a particular
birthing position. It is possible that the
women who had this preference had different
characteristics to the women who were
included in the studies. None of the studies
examined this possibility of bias.
Many problems are encountered when
setting up an RCT into birthing positions. The
fact that blinding is not possible meant that
these studies received only moderate quality
scores and may have caused several forms
of bias. The results should therefore be
interpreted with caution. The variation in
compliance rate may partly be explained by
the fact that some positions are easier to
adopt than others. The lowest compliance was
in a study where women gave birth on a
birthing stool
27
. A 100% compliance rate
appeared to be met in studies where women
used a chair or a lateral tilt position
23,24,28–32
.
Women may also have been encouraged to
adopt the allocated position even if they
Table 5 Supine vs. non-supine position: women’s birth experiences
Study n
a
Method Results p value
Experience of pain
De Jong
25
517 Independent midwife recorded pain day Fewer women in squatting group reported trend
after delivery; mild, moderate, severe, significant pain. 0.0034
extreme
Marttila
28
100 Not described Fewer women in half-sitting position < 0.05
reported intolerable pain (0 vs. 4).
Waldenström
27
287 Questionnaire given by assisting midwife Birthing stool lower average level of pain 0.02
2 hours after delivery; indicate level of compared to supine (6.9 vs. 7.6).
pain on scale from 1 to 10
Bearing down effort
Chen
29
116 Questionnaire filled in on 6th day More nullipara’s on birthing chair found it
postpartum easy to bear down. < 0.05
More multipara’s on birthing chair found it
easier to bear down than at the previous
delivery. < 0.05
Experience
De Jong
25
517 Independent midwife recorded satisfaction No significant difference in trend
day after delivery: very unhappy, slightly maternal satisfaction. 0.09
unhappy, satisfied, very satisfied, very
happy
Marttila
28
100 Not described In half-sitting position 5 women had a very 1
unpleasant experience, in supine position it
was 9 women.
Waldenström
27
287 Questionnaire given by assisting midwife Women on birthing stool more positive trend
2 hours after delivery; excellent, fairly experience. 0.011
good, neither good nor bad, not good
or very bad experience
Preferred position for next delivery
Marttila
28
100 Not described 96% women in half-sitting position and
86% in supine position prefer half-sitting
position next time.
Waldenström
27
287 Questionnaire given by assisting 53% women on birthing stool and 41% in
Midwife 2 hrs after delivery supine position prefer same position for
next delivery.
Of women who actually delivered in
allocated position these percentages were
90% and 49%, respectively.
a
number of women who were assessed for these outcomes
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.
Birthing positions
De Jonge et al.
42 Journal of Psychosomatic Obstetrics & Gynecology
would rather move to another position
during the second stage, or they may have felt
obliged to comply. It is debatable whether it
is ethical to ask women to participate in a
study whereby they agree to adopt a certain
birthing position even though they do not
know how they will feel during labor. The
main advantage of an RCT is that through
randomization and blinding several sources
of bias are reduced. Since blinding is hardly
possible and randomization results in
problems regarding ethical and compliance
issues, a cohort study is more appropriate to
research birthing positions. In our included
cohort study the supine position was asso-
ciated with more episiotomies
26
. However,
many more occurred in the lithotomy rather
than in the semi-sitting (25–45° ) position.
The authors suggest that some healthcare
providers may have changed the woman’s
position to lithotomy in order to perform an
episiotomy. In a cohort study information
on possible confounders, such as medical
reasons for a change in position, should be
carefully recorded.
In some studies it was not clear how the
second stage was defined and therefore, it was
not clear how long the position at the time
of birth had been adopted for. In addition, if
a second stage in supine position lasts only a
few minutes an upright position during the
hour before may still have influenced the
birth outcome. Much could be learnt from
knowing all the positions women adopt
during the entire second stage and during the
last hour of the first stage. This may also bring
to light benefits of using several positions
rather than just one during the second stage.
Authors have suggested that the movement
from one position to another may be benefi-
cial, but there is a lack of evidence to support
this
33,34
.
Many of the studies were carried out in
settings with a high rate of obstetric inter-
ventions such as the use of oxytocin or
epidural infusions. The rate of postpartum
hemorrhage was much lower in one study
without these interventions
25
. Although the
numbers involved are low, this suggests that
outcomes may be different in low inter-
vention settings. Well designed cohort studies
in low intervention settings may produce
useful results for professionals who use very
few methods of obstetric intervention.
Most outcomes were subjectively assessed
by the assisting health professional. Especially
when it comes to estimating blood loss,
professionals are known to underestimate the
amount, in particular when the loss is con-
siderable
35
. In an upright position the blood
loss may appear more than in supine position
because it can be collected in a recepta-
cle
27,31,36
. Even if more blood may be lost at the
time of the birth due to the force of gravity,
the subsequent lochia may be reduced
27,31
. It
has also been suggested that the increased
pressure on the perineum in an upright
position may cause an increased blood loss
from perineal damage rather than an atonic
uterus
31,37,38
. Even if there were a real de-
creased blood loss in the supine position, the
question remains whether this difference is
clinically significant. Although the difference
in blood loss found in this review was statis-
tically significant, it was only a difference
of almost 60 mls and a difference in the
requirement of a blood transfusion was not
found. The risk of severe blood loss may not
be the same for every woman. In this meta-
analysis, the difference was only significant
for multigravidas. It has been suggested that
multigravidas who give birth very quickly
tend to have a greater blood loss
31
. In daily
clinical practice the assisting health profes-
sional may suggest to these women that they
lie down and use the supine position to slow
the progress of labor. An RCT in which these
women adopt the allocated position will not
take these clinical differences into account
and may therefore overestimate the risk of
blood loss for all women.
The only difference in neonatal outcome
which almost reached significance was the
difference in umbilical artery pH (p = 0.05) but
the difference of –0.02 in supine position is
unlikely to be clinically significant.
Most women preferred positions other
than the supine position and more women
had a good experience in other positions.
More women reported severe pain in the
supine position and more women found it
difficult to bear down. These results should
be interpreted cautiously because of method-
ological problems. One study does not explain
how they asked women about their experi-
ence of pain
28
. In another one the assisting
midwife handed out the questionnaire and
this may have influenced the results
27
. The
fact that no women in either group in this
study reported to have had a very bad
experience may be because they did not want
to offend her. The wording used for the
questions varied and was sometimes open to
various interpretations. For example, women
may have had difficulty choosing between
very satisfied and very happy
25
. Finally,
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.
Birthing positions De Jonge et al.
Journal of Psychosomatic Obstetrics & Gynecology 43
although the non-pooled data give some
impression of women’s experiences, they do
not explain how different birthing positions
contribute to women’s experiences. For
example; women reported severe pain more
often in the supine position. They may have
felt more physical pain or the partner may
have been less able to give support, or women
may have felt less in control and were
therefore less able to cope with the pain. To
investigate these underlying mechanisms, a
qualitative research method will be useful.
In conclusion, due to methodological
problems in the studies included in this meta-
analysis the results need to be considered
cautiously. Nevertheless, it appears that the
routine use of the supine position may have
some disadvantages in terms of more instru-
mental deliveries and episiotomies. Also,
more women appear to experience significant
pain in this position and to prefer other
positions. The observed reduced blood loss in
the supine position may not be an actual,
physical difference and may not be clinically
significant. In summary, these results do not
justify the routine use of the supine position
for all women during the second stage of
labor.
A cohort study in a low intervention set-
ting complemented by a qualitative method
is suggested as the most appropriate form of
research into this subject. Not only should the
position at the moment of birth be registered,
but also the positions during the entire
second stage and during the last hour of the
first stage in order to measure their influence
on the birth outcome. Additionally, this
would allow the investigation of benefits of
using various positions during the final stages
of labor. Information on possible confounders
should be carefully registered. Objective
outcome measures such as blood indices
postpartum will be needed to test the
difference in blood loss.
ACKNOWLEDGEMENTS
We would like to thank Dr. A. Verhagen for
her advice on our quality criteria list and Dr.
P. Lucassen for his ideas on developing our
protocol. Many thanks too to A. Jackson for
correcting the use of the English language
and to C. Agyemang for his editing assistance.
REFERENCES
1. Banks E. Laboring in comfort. Nurs Times 1992;
88:40–1
2. Gélis J. Hastening the hour of deliverance. In
History of Childbirth: Fertility, Pregnancy and Birth
in Early Modern Europe. Translated by Rosemary
Morris. [L’arbre et le fruit. Arthème Fayard,
1984] Cambridge: Polity Press, 1991:121–33
3. Atwood RJ. Parturitional posture and related
birth behaviour. Acta Obstet Gynecol Scand
1976;(Suppl 57):1–25
4. Boyle M. Childbirth in bed. The historical
perspective. Pract Midwife 2000;3:21–4
5. Russell AGB. The rationale of primitive delivery
positions. Br J Obstet Gynaecol 1982;89:712–15
6. Hunt LM, Glantz NM, Halperin DC. Childbirth
care-seeking behavior in Chiapas. Health Care
Women Int 2002;23:98–118
7. Henty D. Brought to bed: a critical look at
birthing positions. RCM Midwives J 1998;1:
310–13
8. Carlson J, Sachtleben-Murray M, Fenwick L.
Maternal position during parturition in
normal labor. Obstet Gynaecol 1986;68:443–7
9. Hanson L. Second-stage positioning in nurse-
midwifery practices. Part 1: Position use and
preferences. J Nurse Midwifery 1998;43:320–5
10. Gibb D, Arulkumaran S. Introduction. In Fetal
monitoring in practice. Oxford: Butterworth-
Heinemann, 1992:3
11. Golay J, Vedam S, Sorger L. The squatting
position for the second stage of labor: effects
on labor and on maternal and fetal well-being.
Birth 1993;20:73–8
12. Caldeyro-Barcia R, Noriega-Guerra L, Cibilis
LA, et al. Effect of position changes on the
intensity and frequency of uterine contractions
during labor. Am J Obstet Gynecol 1960;80:284–90
13. Kurz CS. The influence of the maternal
position on the fetal transcutaneous oxygen
pressure. J Perinat Med 1982;10 (Suppl.2):74–5
14. Carbonne B, Benachi A, Lévèque M, et al.
Maternal position during labor: effects on
fetal oxygen saturation measured by pulse
oximetry. Obstet Gynecol 1996;88:797–800
15. Hofmeyr GJ, Kulier R. Hands/knees posture in
late pregnancy or labor for fetal malposition
(lateral or posterior). Cochrane Database Syst Rev
2000;(2): CD001063
16. Zuidenga A. Een literatuuronderzoek. De
verticale baringshouding. [A literature review.
The vertical birthing position]. Tijdschrift voor
Verloskundigen 1996;21:11–15
17. Van Diem M Th. De invloed van kenmerken
van de vrouw en de vroedvrouw op de
baringshouding. [The influence of women’s
and midwives’ characteristics on the birthing
position]. Tijdschrift voor Verloskunidigen 1997;
22:18–27
18. Moore S. Psychosocial Support during Labor.
In Henderson, C, Jones, K, eds. Essential
Midwifery. London: Mosby, 1997:219–27
19. Enkin M, Keirse MJNC, Neilson J, et al. In A
guide to effective care in pregnancy and childbirth.
Oxford: Oxford University Press, 2000:291–2
20.Gupta JK, Nikodem VC. Woman’s position
during the second stage of labor (Cochrane
Review). Oxford: Cochrane Library, Update
Software, 2000:issue 4
21. Verhagen AP, De Vet HCW, De Bie RA, et al. The
Delphi list: A criteria list for quality assessment
of randomized clinical trials for conducting
systematic reviews developed by Delphi
Consensus. J Clin Epidemiol 1998;51:1235–41
22.Mulrow C, Oxman A, eds. Cochrane
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.
Birthing positions
De Jonge et al.
44 Journal of Psychosomatic Obstetrics & Gynecology
Collaboration Handbook. Oxford: Cochrane
Library. Update Software, 1997:March
23.Humphrey M, Hounslow D, Morgan S, Wood C.
The influence of maternal posture at birth on
the fetus. J Obstet Gynaecol Br Commonwealth
1973;80:1075–80
24.Johnstone FD, Aboelmagd MS, Harouny AK.
Maternal posture in second stage and fetal
acid base status. Br J Obstet Gynaecol 1987;94:
753–7
25.De Jong PR, Johanson RB, Baxen P, et al.
Randomised trial comparing the upright and
supine positions for the second stage of labor.
Br J Obstet Gynaecol 1997;104:567-71
26.Lydon-Rochelle MT, Albers L, Teaf D. Perineal
outcomes and nurse-midwifery management.
J Nurse Midwifery 1995;40:13–18
27. Waldenstrøm U, Gottvall K. A randomized
trial of birthing stool or conventional
semirecumbent position for second-stage
labor. Birth 1991;18:5–10
28.Marttila M, Kajanoja P, Ylikorkala O. Maternal
half-sitting position in the second stage of
labor. J Perinat Med 1983;11:286–9
29.Chen SZ, Aisaka K, Mori H, Kigawa T. Effects of
sitting position on uterine activity during
labor. Obstet Gynecol 1987;69:67–73
30.Stewart P, Hillan E, Calder AA. A randomised
trial to evaluate the use of a birth chair for
delivery. Lancet 1983;1:1296–8
31. Hillan EM. The birthing chair trial. In Research
and the Midwife. Conference Proceedings 1983.
Manchester: Research and the Midwife, 1984:
22–37
32.Turner MJ, Romney ML, Webb JB, Gordon H.
The birthing chair: an obstetric hazard?
J Obstet Gynaecol Br Commonwealth 1986;6:232–5
33.Smulders B, Croon M. Veilig bevallen. Het complete
handboek. [Giving birth safely. The complete
guide]. Utrecht: Kosmos-Z & K Uitgevers B.V.,
1996:73
34.Rosser J. Women’s position in second stage.
The Practising Midwife 2000;3:10–11
35.Brant HA. Precise estimation of postpartum
hemorrhage: difficulties and importance. Br
Med J 1967;1:398–400
36.Hagymásy L, Gaál J. A comparative study of
vertical and horizontal deliveries in the
presence and with the assistance of the
woman’s partner. J Psychosom Obstet Gynecol
1998;19:98–103
37. Van Diem M, Herschderfer K, Aitink M,
Buitendijk S. Measured blood loss instead of
estimated blood loss and delivery in the upright
position: an observational study. Leiden: Leiden
Universitair Medisch Centrum, Unpublished:
2002
38.Gardosi J, Hutson N, Lynch CB. Randomised,
controlled trial of squatting in the second
stage of labor. The Lancet 1989a;2:74–7
39.Aarnoudse JG, Romney ML, Gordon H. Does
fetal oxygen supply improve in the birthing
chair? J Obstet Gynecol 1984;4:141–2
40.Aikins Murphy P, Baker Feinland J. Perineal
outcomes in a home birth setting. Birth 1998;
25:226–34
41. Allahbadia GN, Vaidya PR. Why deliver in the
supine position? Australia N Z J Obstet Gynaecol
1992;32:104–6
42.Bastian H. Birth positions and the perineum:
experiences and outcomes at home births in
Australia. Homebirth Australia Newsletter 1994;
36:4–8
43.Bhardwaj N, Kukade JA. Randomized controlled
trial on modified squatting position of
birthing [abstract]. Int J Obstet Gynecol 1994;46:
118
44.Bomfim-Hyppolito S. Influence of the position
of the mother at delivery over some maternal
and neonatal outcomes. Int J Gynecol Obstet
1998;63(Suppl.1):S67–73
45.Chan DPC. Positions during labor. Br Med J
1963;1:100–2
46.Crowley P, Elbourne D, Ashurst H, et al. Delivery
in an obstetric birth chair: a randomised
controlled trial. Br J Obstet Gynaecol 1991;98:
667–74
47. Drähne A, Prang E, Werner Ch. The various
positions for delivery [abstract]. J Perinatal Med
1982;10(Suppl.2):72–3
48.Gardosi J, Sylvester S, Lynch CB. Alternative
positions in the second stage of labor: a
randomised controlled trial. Br J Obstet Gynaecol
1989b;96:1290–6
49.Gåreberg B, Magnusson B, Sultan B, et al. Birth
in standing position: a high frequency of third
degree tears. Acta Obstet Gynaecol Scand 1994;
73:630–3
50.Gupta JK, Brayshaw EM, Lilford RJ. An
experiment of squatting birth. Eur J Obstet
Gynaecol Reprod Biol 1989;30:217–20
51. Hemminki E, Virkkunen A, Mäkelä A, et al. A
trial of delivery in a birth chair. J Obstet Gynecol
1986;6:162–5
52.Kafka M, Riss P, Von Trotsenburg M, Maly Z.
Gebarhocker – ein geburtshilfliches Risiko?
[The birthing stool – an obstetrical risk?].
Geburtshilfe und Frauenheilkunde 1994;54:529–31
53.Kleine-Tebbe A, David M, Farkic M. Aufrechte
Gebärpositionen – mehr Geburtswegsverlet-
zungen? Ergebnisse einer retrospektiven
vergleichenden Untersuchung. [Upright
birthing position – more birth canal injuries?
Results of a retrospective comparative study].
Zentralblatt Gynakologie 1996;118:448–52
54.Liddell HS, Fisher PR. The birthing chair in the
second stage of labor. Australia N Z J Obstet
Gynaecol 1985;25:65–8
55.Liu YC. Effects of an upright position during
labor. Am J Nursing 1974;74:2202–5
56.Liu YC. The effects of the upright position
during childbirth. J Nursing Scholarship 1989;
21:14–18
57. McManus TJ, Calder AA. Upright posture and
the efficiency of labor. The Lancet 1978;1:72–4
58.Moll L, Van Dam J, De Haan S. Een vergelijkend
onderzoek naar horizontaal en verticaal baren. [A
comparative study into supine and upright
delivery]. Amsterdam: School of Midwifery,
1985 (unpublished)
59.Nodine PM, Roberts J. Factors associated with
perineal outcome during childbirth. J Nurse
Midwifery 1987;32:123–30
60.Olson R, Olson C, Cox NS. Maternal birthing
positions and perineal injury. J Family Practice
1990;30:553–7
61. Racinet C, Eymery P, Philibert L, Lucas C.
L’Accouchement en position accroupie. Essai
randomisé comparant la position accroupie à
la position classique en phase d’expulsion.
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.
Birthing positions De Jonge et al.
Journal of Psychosomatic Obstetrics & Gynecology 45
[Labor in the squatting position. A randomized
trial comparing the squatting position with
the classical position for the expulsion phase].
J Gynecol Obstet Biol Reprod 1999;28:263–70
62.Radkey AL, Liston RM, Scott KE, Young C.
Squatting: Preventive medicine in childbirth?
[abstract]. In Proceedings of the annual meeting of
the Society of Obstetricians and Gynaecologists of
Canada. Toronto: Canada, 1991:76
63.Roberts JE, Kriz DM. Delivery positions and
perineal outcome. J Nurse Midwifery 1984;29:
186–9
64.Rohrbacher A, Salzer H. Roma-Geburtsrad:
1 Jahr klinische Erfahrung an einem
Schwerpunktkrankenhaus. [The Roma
birthing wheel: 1 year clinical experience in a
specialty hospital] Gynäkologisch-Geburtshilfliche
Rundschau 1998;38:158–63
65.Romney M. Chair Project. In Research and the
Midwife. Conference Proceedings 1983. Manchester:
Research and the Midwife, 1984:69–80
66.Schneider-Affeld F, Martin K. Delivery from a
sitting position [abstract]. J Perinatal Med 1982;
2:70–1
67. Shannahan MD, Cottrell BH. Effect of the
birth chair on duration of second stage labor,
fetal outcome and maternal blood loss.
Nursing Res 1985;34:89–92
68.Shannahan MD, Cottrell BH. The effects of
birth chair delivery on maternal perceptions.
J Obstet Gynecol Neonatal Nurs 1989;18:323–6
69.Shorten A, Donsante J, Shorten B. Birth
position, accoucheur, and perineal outcomes:
informing women about choices for vaginal
birth. Birth 2002;29:18–27
70.Stewart P, Spiby H. A randomised study of the
sitting position for delivery using a newly
designed obstetric chair. Br J Obstet Gynaecol
1989;96:327–33
Current knowledge on this subject
• Previous studies have identified benefits and risks of different
positions during the second stage of labor
• The use of any upright or lateral position has been associated
with a reduced duration of the second stage of labor, a reduced
reporting of severe pain and a reduction in assisted deliveries,
abnormal fetal heart rate patterns and episiotomies
• However, an increase in second degree tears and an increased
risk of blood loss of more than 500 ml has been found
What this study adds
• Supine position is often used routinely in western countries, this
meta-analysis aimed to establish the benefits of this intervention
in the light of evidence-based medicine
• The use of the supine position compared to other positions was
not associated with clear benefits for the mother or the baby
• Routine use of supine position for all women is not justified
• There are methodological and ethical problems in conducting a
randomized controlled trial into birthing positions
• A cohort study complemented by a qualitative method is sug-
gested as the most appropriate form of research into this subject
J Psychosom Obstet Gynaecol Downloaded from informahealthcare.com by University of Groningen on 11/26/14
For personal use only.