Bu Linda 1

Cochrane
Library
Cochrane Database of Systematic Reviews

Effect of partograph use on outcomes for women in spontaneous
labour at term and their babies (Review)
Lavender T, Cuthbert A, Smyth RMD
Lavender T, Cuthbert A, Smyth RMD.

Effect of partograph use on outcomes for women in spontaneous labour at term and their babies.
Cochrane Database of Systematic Reviews 2018, Issue 8. Art. No.: CD005461.
DOI: 10.1002/14651858.CD005461.pub5.
www.cochranelibrary.com

Effect of partograph use on outcomes for women in spontaneous labour at term and their babies (Review)
Copyright © 2018 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
Cochrane Trusted evidence.
Informed decisions.

Library Better health. Cochrane Database of Systematic Reviews
TABLE OF CONTENTS
HEADER......................................................................................................................................................................................................... 1
ABSTRACT..................................................................................................................................................................................................... 1
PLAIN LANGUAGE SUMMARY....................................................................................................................................................................... 2
SUMMARY OF FINDINGS.............................................................................................................................................................................. 4
BACKGROUND.............................................................................................................................................................................................. 6
Figure 1.................................................................................................................................................................................................. 6
OBJECTIVES.................................................................................................................................................................................................. 7
METHODS..................................................................................................................................................................................................... 8
RESULTS........................................................................................................................................................................................................ 12
Figure 2.................................................................................................................................................................................................. 13
Figure 3.................................................................................................................................................................................................. 17
DISCUSSION.................................................................................................................................................................................................. 24
AUTHORS' CONCLUSIONS........................................................................................................................................................................... 26
ACKNOWLEDGEMENTS................................................................................................................................................................................ 26
REFERENCES................................................................................................................................................................................................ 28
CHARACTERISTICS OF STUDIES.................................................................................................................................................................. 31
DATA AND ANALYSES.................................................................................................................................................................................... 48
Analysis 1.1. Comparison 1 Partograph versus no partograph (studies carried out in high- and low-resource settings), Outcome 50
1 Caesarean section (overall)...............................................................................................................................................................
2 Oxytocin augmentation.....................................................................................................................................................................
3 Duration of first stage of labour........................................................................................................................................................
4 Low Apgar score (less than 7 at 5 minutes).....................................................................................................................................
5 Instrumental vaginal birth.................................................................................................................................................................
6 Regional analgesia - epidural............................................................................................................................................................
7 Performance of artificial rupture of membranes during labour.....................................................................................................
8 Antibiotic use.....................................................................................................................................................................................
9 Duration of second stage of labour (hours).....................................................................................................................................
10 Number of vaginal examinations....................................................................................................................................................
11 Admission to special care nursery..................................................................................................................................................
Analysis 2.1. Comparison 2 Partograph with 2-hour action line versus partograph with 4-hour action line (studies carried out in 56
a high- and low-resource settings), Outcome 1 Caesarean section (overall)....................................................................................
a high- and low-resource settings), Outcome 2 Oxytocin augmentation..........................................................................................
a high- and low-resource settings), Outcome 3 Duration of first stage of labour (length of labour greater than 18 hours, length
of labour greater than 12 hours).........................................................................................................................................................
a high- and low-resource settings), Outcome 4 Maternal experience of childbirth - negative childbirth experience.....................
a high- and low-resource settings), Outcome 5 Low Apgar score (less than 7 at 5 minutes)...........................................................
a high- and low-resource settings), Outcome 6 Serious maternal morbidity or death....................................................................
a high- and low-resource settings), Outcome 7 Caesarean section (distress)...................................................................................
Effect of partograph use on outcomes for women in spontaneous labour at term and their babies (Review) i
Informed decisions.

a high- and low-resource settings), Outcome 8 Caesarean section (delay)......................................................................................
a high- and low-resource settings), Outcome 9 Instrumental vaginal delivery.................................................................................
Analysis 2.10. Comparison 2 Partograph with 2-hour action line versus partograph with 4-hour action line (studies carried out 60
in a high- and low-resource settings), Outcome 10 Postpartum haemorrhage - blood loss > 500 mL.............................................
in a high- and low-resource settings), Outcome 11 Regional analgesia - epidural...........................................................................
in a high- and low-resource settings), Outcome 12 Performance of artificial rupture of the membranes during labour................
in a high- and low-resource settings), Outcome 13 Number of vaginal examinations in labour......................................................
in a high- and low-resource settings), Outcome 14 Serious neonatal morbidity or perinatal death...............................................
in a high- and low-resource settings), Outcome 15 Admission to special care nursery...................................................................
in a high- and low-resource settings), Outcome 16 Cord blood arterial pH less than 7.1................................................................
Analysis 3.1. Comparison 3 Partograph with 2-hour action line versus partograph with 3-hour action line (study carried out in 63
a high-resource setting), Outcome 1 Caesarean section (overall).....................................................................................................
a high-resource setting), Outcome 2 Oxytocin augmentation...........................................................................................................
a high-resource setting), Outcome 3 Maternal experience of childbirth - negative childbirth experience......................................
a high-resource setting), Outcome 4 Low Apgar score (less than 7 at 5 minutes)............................................................................
a high-resource setting), Outcome 5 Serious maternal morbidity or death.....................................................................................
a high-resource setting), Outcome 6 Caesarean section (distress)....................................................................................................
a high-resource setting), Outcome 7 Caesarean section (delay).......................................................................................................
a high-resource setting), Outcome 8 Instrumental vaginal delivery..................................................................................................
a high-resource setting), Outcome 9 Postpartum haemorrhage - blood loss > 500 mL...................................................................
a high-resource setting), Outcome 10 Regional analgesia - epidural................................................................................................
a high-resource setting), Outcome 11 Performance of artificial rupture of membranes during labour...........................................
a high-resource setting), Outcome 12 Vaginal examinations.............................................................................................................
a high-resource setting), Outcome 13 Serious neonatal morbidity or perinatal death....................................................................
a high-resource setting), Outcome 14 Admission to special care nursery........................................................................................
a high-resource setting), Outcome 15 Cord blood arterial pH less than 7.1.....................................................................................
a high-resource setting), Outcome 1 Caesarean section (overall).....................................................................................................
a high-resource setting), Outcome 2 Oxytocin augmentation...........................................................................................................
a high-resource setting), Outcome 3 Maternal experience of childbirth - negative childbirth experience......................................
a high-resource setting), Outcome 4 Low Apgar score (less than 7 at 5 minutes)............................................................................
Effect of partograph use on outcomes for women in spontaneous labour at term and their babies (Review) ii
Informed decisions.

a high-resource setting), Outcome 5 Serious maternal morbidity or death.....................................................................................
a high-resource setting), Outcome 6 Caesarean section (distress)....................................................................................................
a high-resource setting), Outcome 7 Caesarean section (delay).......................................................................................................
a high-resource setting), Outcome 8 Instrumental vaginal delivery..................................................................................................
a high-resource setting), Outcome 9 Postpartum haemorrhage - blood loss > 500 mL...................................................................
a high-resource setting), Outcome 10 Regional analgesia - epidural................................................................................................
a high-resource setting), Outcome 11 Performance of artificial rupture of membranes during labour...........................................
a high-resource setting), Outcome 12 Number of vaginal examinations in labour..........................................................................
a high-resource setting), Outcome 13 Serious neonatal morbidity or perinatal death....................................................................
a high-resource setting), Outcome 14 Admission to special care nursery........................................................................................
a high-resource setting), Outcome 15 Cord blood arterial pH less than 7.1.....................................................................................
Analysis 5.1. Comparison 5 Partograph with alert line only versus partograph with alert and action line (study carried out in a 74
low-resource setting), Outcome 1 Caesarean section (overall).........................................................................................................
low-resource setting), Outcome 2 Oxytocin augmentation...............................................................................................................
low-resource setting), Outcome 3 Low Apgar score (less than 7 at 5 minutes)................................................................................
low-resource setting), Outcome 4 Instrumental vaginal delivery......................................................................................................
low-resource setting), Outcome 5 Serious neonatal morbidity or perinatal death..........................................................................
Analysis 6.1. Comparison 6 Partograph with latent phase versus partograph without latent phase (study carried out in a low- 76
resource setting), Outcome 1 Caesarean section (overall).................................................................................................................
resource setting), Outcome 2 Oxytocin augmentation.......................................................................................................................
resource setting), Outcome 3 Low Apgar score (less than 7 at 5 minutes).......................................................................................
resource setting), Outcome 4 Caesarean section (distress)...............................................................................................................
resource setting), Outcome 5 Caesarean section (delay)...................................................................................................................
resource setting), Outcome 6 Instrumental vaginal delivery.............................................................................................................
resource setting), Outcome 7 Admission to special care nursery......................................................................................................
resource setting), Outcome 8 Usability: user-friendliness score........................................................................................................
Analysis 7.1. Comparison 7 Partograph with 2-hour action line versus partograph with stepped dystocia line, Outcome 1 80
Caesarean section.................................................................................................................................................................................
Oxytocin augmentation........................................................................................................................................................................
Duration of first stage of labour (labour longer than 12 hours).........................................................................................................
Maternal experience of childbirth (BSS-R score)................................................................................................................................
Effect of partograph use on outcomes for women in spontaneous labour at term and their babies (Review) iii
Informed decisions.

Analysis 7.5. Comparison 7 Partograph with 2-hour action line versus partograph with stepped dystocia line, Outcome 5 Low 81
Apgar score (less than 4 at 4 min).......................................................................................................................................................
Instrumental vaginal birth....................................................................................................................................................................
Postpartum haemorrhage (> 500 mL).................................................................................................................................................
Regional analgesia................................................................................................................................................................................
Analysis 7.9. Comparison 7 Partograph with 2-hour action line versus partograph with stepped dystocia line, Outcome 9 Opioid 82
use..........................................................................................................................................................................................................
Analysis 7.10. Comparison 7 Partograph with 2-hour action line versus partograph with stepped dystocia line, Outcome 10 Need 82
for intubation at birth...........................................................................................................................................................................
Analysis 8.1. Comparison 8 Partograph versus labour scale (study carried out in a low-resource setting), Outcome 1 Caesarean 83
section (overall).....................................................................................................................................................................................
Analysis 8.2. Comparison 8 Partograph versus labour scale (study carried out in a low-resource setting), Outcome 2 Oxytocin 84
augmentation........................................................................................................................................................................................
Analysis 8.3. Comparison 8 Partograph versus labour scale (study carried out in a low-resource setting), Outcome 3 Duration 84
of first stage of labour..........................................................................................................................................................................
Analysis 8.4. Comparison 8 Partograph versus labour scale (study carried out in a low-resource setting), Outcome 4 Low Apgar 84
score (less than 7 at 5 minutes)...........................................................................................................................................................
Analysis 8.5. Comparison 8 Partograph versus labour scale (study carried out in a low-resource setting), Outcome 5 Caesarean 84
section (delay).......................................................................................................................................................................................
Analysis 8.6. Comparison 8 Partograph versus labour scale (study carried out in a low-resource setting), Outcome 6 Stillbirth, 85
neonatal death or neonatal morbidity................................................................................................................................................
Analysis 8.7. Comparison 8 Partograph versus labour scale (study carried out in a low-resource setting), Outcome 7 Birth injuries 85
and PPH (non-prespecified outcome).................................................................................................................................................
WHAT'S NEW................................................................................................................................................................................................. 85
HISTORY........................................................................................................................................................................................................ 86
CONTRIBUTIONS OF AUTHORS................................................................................................................................................................... 86
DECLARATIONS OF INTEREST..................................................................................................................................................................... 86
SOURCES OF SUPPORT............................................................................................................................................................................... 87
DIFFERENCES BETWEEN PROTOCOL AND REVIEW.................................................................................................................................... 87
INDEX TERMS............................................................................................................................................................................................... 87
Effect of partograph use on outcomes for women in spontaneous labour at term and their babies (Review) iv
Informed decisions.

[Intervention Review]
Effect of partograph use on outcomes for women in spontaneous labour

at term and their babies
Tina Lavender1, Anna Cuthbert2, Rebecca MD Smyth1
1Division of Nursing Midwifery and Social Work, The University of Manchester, Manchester, UK. 2Cochrane Pregnancy and Childbirth Group,
Department of Women's and Children's Health, The University of Liverpool, Liverpool, UK
Contact address: Tina Lavender, Division of Nursing Midwifery and Social Work, The University of Manchester, Oxford Road, Manchester,
M13 9PL, UK. tina.lavender@manchester.ac.uk.
Editorial group: Cochrane Pregnancy and Childbirth Group

Publication status and date: New search for studies and content updated (conclusions changed), published in Issue 8, 2018.
Citation: Lavender T, Cuthbert A, Smyth RMD. Effect of partograph use on outcomes for women in spontaneous labour at term and their
babies. Cochrane Database of Systematic Reviews 2018, Issue 8. Art. No.: CD005461. DOI: 10.1002/14651858.CD005461.pub5.
ABSTRACT
Background
The partograph (sometimes known as partogram) is usually a pre-printed paper form on which labour observations are recorded. The aim
of the partograph is to provide a pictorial overview of labour, and to alert midwives and obstetricians to deviations in maternal or fetal
well-being and labour progress. Charts have traditionally contained pre-printed alert and action lines. An alert line, which is based on the
slowest 10% of primigravid women's labours, signifies slow progress. An action line is placed a number of hours after the alert line (usually
two or four hours) to prompt effective management of slow progress of labour.
This review is an update of a review last published in 2013.
Objectives
The primary objective was to determine the effectiveness and safety of partograph use on perinatal and maternal morbidity and mortality.
The secondary objective was to determine which partograph design is most effective for perinatal and maternal morbidity and mortality
outcomes.
Search methods
We searched Cochrane Pregnancy and Childbirth's Trials Register (31 August 2017), ClinicalTrials.gov, the World Health Organization
(WHO) International Clinical Trials Registry Platform (ICTRP) (31 August 2017) and reference lists of retrieved studies.
Selection criteria
Randomised, cluster-randomised, and quasi-randomised controlled trials involving a comparison of partograph use with no partograph,
or comparison between different partograph designs.
Data collection and analysis

Three review authors independently assessed eligibility, quality and extracted data. When one review author was also the trial author, the
two remaining review authors assessed the studies independently. We assessed the evidence using the GRADE approach.
Main results
We have included 11 studies, involving 9475 women in this review; three studies assessed partograph use versus no partograph, seven
assessed different partograph designs, and one assessed partograph use versus labour scale. Risk of bias varied in all studies. It was in-
feasible to blind staff or women to the intervention. Two studies did not adequately conceal allocation. Loss to follow-up was low in all
Effect of partograph use on outcomes for women in spontaneous labour at term and their babies (Review) 1
Informed decisions.

studies. We assessed the evidence for partograph use versus no partograph using the GRADE approach; downgrading decisions were due
to study design, inconsistency, indirectness, and imprecision of effect estimates.
Most trials reported caesarean section rates and Apgar scores less than 7 at five minutes; all other outcomes were not consistently reported
(e.g. duration of first stage of labour and maternal experience of childbirth).
Partograph versus no partograph (3 trials, 1813 women)
It is uncertain whether there is any clear difference between partograph use and no partograph in caesarean section rates (average risk
ratio (RR) 0.77, 95% confidence interval (CI) 0.40 to 1.46; n = 1813; 3 trials; I2 = 87%; very low-quality evidence); oxytocin augmentation (RR
1.02, 95% CI 0.95 to 1.10; n = 1156; 1 trial; moderate-quality evidence); duration of first stage of labour (mean difference (MD) 0.80 hours,
95% CI -0.06 to 1.66; n = 1156; 1 trial; low-quality evidence); or Apgar score less than 7 at five minutes (RR 0.76, 95% CI 0.29 to 2.03; n =
1596; 2 trials; I2 = 87%; very low-quality evidence).
Partograph with different placement of action lines (4 trials, 5051 women)
When compared to a four-hour action line, women in the two-hour action line group were more likely to receive oxytocin augmentation
(average RR 2.44, 95% CI 1.36 to 4.35; n = 4749; 4 trials; I2 = 96%). There was no clear difference in caesarean section rates (RR 1.06, 95%
CI 0.88 to 1.28; n = 4749; 4 trials); duration of first stage of labour (RR 0.81 hours, 95% CI 0.32 to 2.04; n = 948; 1 trial); maternal experience
of childbirth (average RR 0.61, 95% CI 0.28 to 1.35; n = 2269; 2 trials; I2 = 83%); or Apgar score less than 7 at five minutes (RR 0.93, 95% CI
0.61 to 1.42; n = 4749; 4 trials) between the two- and four-hour action line.
The following comparisons only include data from single studies. Fewer women reported negative childbirth experiences in the two-hour
action line group compared to the three-hour action line group (RR 0.49, 95% CI 0.27 to 0.90; n = 348; 1 trial). When we compared the three-
and four-hour action line groups, the caesarean section rate was higher in the three-hour action line group (RR 1.70, 95% CI 1.07 to 2.70; n
= 613; 1 trial). We did not observe any clear differences in any of the other outcomes in these comparisons.
Partograph with alert line only versus partograph with alert and action line (1 trial, 694 women)
The caesarean section rate was lower in the alert line only group (RR 0.68, 95% CI 0.50 to 0.93). There were no clear differences between
groups for oxytocin augmentation, low Apgar score, instrumental vaginal birth and perinatal death.
Partograph with latent phase (composite) versus partograph without latent phase (modified) (1 trial, 743 women)
The caesarean section and oxytocin augmentation rates were higher in the partograph with a latent phase (RR 2.45, 95% CI 1.72 to 3.50;
and RR 2.18, 95% CI 1.67 to 2.83, respectively). There were no clear differences between groups for oxytocin augmentation, and Apgar
score less than 7 at five minutes.
Partograph with two-hour action line versus partograph with stepped dystocia line (1 trial, 99 women)
Fewer women received oxytocin augmentation in the dystocia line group (RR 0.62, 95% CI 0.39 to 0.98). We did not observe any clear
differences in any of the other primary outcomes in this comparison.
Partograph versus labour scale (1 trial, 122 women)
The use of the partograph compared with the labour scale resulted in fewer women receiving oxytocin augmentation (RR 0.32, 95% CI 0.18
to 0.54), but did not produce any clear differences for any of the other primary outcomes.
Authors' conclusions
On the basis of the findings of this review, we cannot be certain of the effects of routine use of the partograph as part of standard labour
management and care, or which design, if any, are most effective. Further trial evidence is required to establish the efficacy of partograph
use per se and its optimum design.
PLAIN LANGUAGE SUMMARY
Effect of partograph use on outcomes for women in spontaneous labour at term
What is the issue?
Does the use of the partograph during spontaneous labour at term improve the health outcomes for women and babies?
Do different partograph designs make any difference to the health outcomes for women and babies?
Why is this important?
Informed decisions.

A partograph is usually a pre-printed form, the aim of which is to provide a pictorial overview of labour progress and to alert health pro-
fessionals to any problems with the mother or baby. It has been unclear whether a partograph should be used and, if so, which design of
partograph is better for women and babies.
What evidence did we find?
We searched for evidence in August 2017 and have now included 11 studies involving 9475 women. Three studies looked at using a par-
tograph versus no partograph, seven studies looked at different partograph designs, and one study looked at using a partograph versus
a new labour scale.
Partograph versus no partograph (3 studies, 1703 women)
It is uncertain whether using a partograph has any effect on the number of women having a caesarean section or babies born with low
Apgar scores (a score which measures the physical condition of the newborn, with a low score indicating poor condition) because the
quality of evidence is very low. Using a partograph may make little or no difference to length of labour (low-quality evidence), or the
number of women who receive oxytocin to speed up their labour (moderate-quality evidence).
Partograph with different placement of action lines (4 trials, 5051 women)
When compared to a four-hour action line, women in the two-hour action line group were more likely to have their labour speeded up
using oxytocin. There was no clear difference between women in the two- and four-hour action line groups in having caesarean sections,
the lengths of first stage of labour, maternal experiences of childbirth, or low Apgar scores.
When we compared a two-hour action line with a three-hour action line, fewer women reported negative childbirth experiences in the
two-hour action line group. When we compared the three- and four-hour action line groups, the caesarean section rate was higher in the
three-hour action line group. There were no clear differences between the two-, three-, or four-hour action line groups in any of the other
outcomes measured.
Partograph with alert line only versus partograph with alert and action line (1 trial, 694 women)
The caesarean section rate was lower in the alert line only group. There were no clear differences between groups for oxytocin augmen-
tation, low Apgar score, instrumental vaginal birth, and perinatal death.
Partograph with latent phase versus partograph without latent phase (1 trial, 743 women)
When we compared a partograph with the latent phase (including early stages of labour) and one without the latent phase, the caesarean
section and oxytocin augmentation rates were higher in the partograph with a latent phase. There were no clear differences between
groups for oxytocin augmentation, and Apgar score less than 7 at 5 minutes.
Partograph with two-hour action line versus partograph with stepped dystocia line (1 trial, 99 women)
When we compared a partograph with a two-hour action line and a stepped dystocia line, fewer women received oxytocin augmentation
in the dystocia line group. We did not observe any clear differences in any of the other primary outcomes in this comparison.
Partograph versus labour scale (1 trial, 122 women)
The labour scale compared with the partograph resulted in fewer women receiving oxytocin augmentation, but did not produce any clear
difference for any of the other primary outcomes.
What does this mean?
On the basis of the findings of this review, we cannot be certain of the effects of routine use of the partograph as part of standard labour
care, or of the different partograph designs. Further trial evidence is required to establish the efficacy of partograph use per se and its
optimum design.
SUMMARY OF FINDINGS

Summary of findings for the main comparison. Partograph compared to no partograph (studies carried out in high- and low-resource settings) for
Library
Cochrane
women in spontaneous labour at term
Partograph compared to no partograph (studies carried out in high- and low-resource settings) for women in spontaneous labour at term
Patient or population: women in spontaneous labour at term

Setting: hospital settings in Canada, India, and Mexico
Better health.
Informed decisions.
Trusted evidence.
Intervention: partograph
Comparison: no partograph (studies carried out in high- and low-resource settings)
Outcomes Anticipated absolute effects* Relative effect № of partici- Quality of the Comments
(95% CI) (95% CI) pants evidence
(studies) (GRADE)
Risk with no Risk with
partograph partograph
(studies car-
ried out in
high- and
low-resource
settings)
Caesarean section (overall) Study population Average RR 0.77 1813 ⊕⊝⊝⊝

(0.40 to 1.46) (3 RCTs) VERY LOW a b
214 per 1000 164 per 1000 c
(85 to 312)
Oxytocin augmentation Study population RR 1.02 1156 ⊕⊕⊕⊝

(0.95 to 1.10) (1 RCT) MODERATE d
715 per 1000 730 per 1000
(680 to 787)

Duration of first stage of labour The mean The mean In the partograph 1156 ⊕⊕⊝⊝
duration of duration of group, mean dura- (1 RCT) LOW c d
first stage of first stage of tion of first stage
labour was 16 labour was was 0.80 hours
hours (SD 7.6) 16.8 hours (SD longer (0.06 hours
7.3) shorter to 1.66
hours longer)
Maternal experience of childbirth (as defined by trial au- Study population - (0 study) - No trials re-
thors) ported this out-
See comment See comment come.
4

Apgar score less than 7 at 5 minutes Study population RR 0.76 1596 ⊕⊝⊝⊝
(0.29 to 2.03) (2 RCTs) VERY LOW e f g
11 per 1000 9 per 1000
Library
Cochrane
(3 to 24)
Serious maternal morbidity or death (e.g. ruptured Study population - (0 study) - No trials re-
uterus, admission to intensive care unit, septicaemia, ported this out-
organ failure) See comment See comment come.
Better health.
Informed decisions.
Trusted evidence.
Stillbirth or neonatal death or neonatal morbidity, ex- Study population - (0 study) - No trials re-
cluding fatal malformations (e.g. seizures, birth asphyx- ported this out-
ia, neonatal encephalopathy) See comment See comment come.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and
its 95% CI).
CI: confidence interval; RCT: randomised controlled trial; RR: risk ratio; SD: standard deviation.
GRADE Working Group grades of evidence

High quality: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is sub-
stantially different
Low quality: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect
Very low quality: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect
aTwo out of three studies have design limitations, serious design limitations less than 40% of pooled effect (-1).
bSevere heterogeneity > 60% (-1).
cWide confidence intervals crossing the line of no effect (-1).
dOne trial included which only included low-risk women (-1).
eOne study with serious design limitations but only contributes 17% to pooled analysis (not downgraded).
fStudies only included term, low risk women (-1).
gWide confidence intervals crossing the line of no effect and few events (-2).

5

Informed decisions.

BACKGROUND services (Neilson 2003). Although, for some women, acceleration of

labour is possible with amniotomy and oxytocin, an increase in as-
Description of the condition sociated pains is more likely to result in an epidural, vaginal instru-
mental birth or caesarean section (Nystedt 2014). Thus accurate di-
The optimum length of labour has been a source of debate that
agnosis of prolonged labour which prompts appropriate action is
stems back to the early days of Friedman (Friedman 1954), and
pivotal.
is still being re-evaluated today (Oladapo 2017). The lack of a de-
finition of normal labour has resulted in challenges in defining Description of the intervention
and detecting prolonged labour. Nevertheless, prolonged labour is
known to be a complication that affects around 8% of all labour- The partograph (or partogram) is a simple, inexpensive tool to pro-
ing women and negatively affects obstetric outcome and women's vide a continuous pictorial overview of labour. The partograph is a
experiences (Nystedt 2014). Detection of prolonged labour is impre-printed form, usually in paper version, on which midwives and
portant, as longer labours are associated with short- and long-term obstetricians record labour observations. Most partographs have
morbidity. In the shorter term, postpartum haemorrhage and in- three distinct sections where observations are entered on maternal
fection are more common in women with long labours and in the condition, fetal condition and labour progress; this last section as-
longer term, fistulas are more prevalent (Neilson 2003). These risks sists in the detection of prolonged labour (Figure 1).
are greater in low-income countries with poorly-resourced health

Figure 1. Modified partograph
Informed decisions.

The first obstetrician to describe the progress of labour graphically labours requiring augmentation (20.7% to 9.1%), emergency cae-
was Friedman (Friedman 1954), following his study of the cervical sarean section (from 9.9% to 8.3%), and stillbirths (from 0.5% to
dilatation of 100 African primigravidae at term. The women were 0.3%).
given frequent rectal examinations and their progress was record-
ed in centimetres of dilatation per hour, producing a slope resem- A belief that partograph use is not affected by racial, cultural and
bling a sigmoid curve ('S' shaped). This became know as the cer- socioeconomic differences, led to the approach finding favour in
vicograph. In an attempt to utilise midwives efficiently in a hos- both high-income and low- to middle-income countries. However,
pital and clinic service in Zimbabwe (then Rhodesia), where doc- in practice, it is conceivable that such variations in care between
tors were in short supply, Philpott 1972a developed a partograph countries, and even units, may alter the use of the partograph
from this original cervicograph. This provided a practical tool for and subsequent effectiveness, in terms of maternal and neonatal
recording all intrapartum details, not just cervical dilatation. An outcomes. As a consequence, some practitioners have questioned
'alert line' was added following the results of a prospective study of its effectiveness, particularly when used in high-income countries
624 women (Philpott 1972b). The alert line was straight, not curved, (Groeschel 2001; Walsh 1994).
and was a modification of the mean rate of cervical dilatation of
There is some evidence to suggest that the partograph has practical
the slowest 10% of primigravid women who were in the active
benefits in terms of aiding referral and transfer, ease of use, time re-
phase of labour. This line represented a progress rate of 1 cm per
sourcefulness, continuity of care, educational assistance and pro-
hour. Should a woman's cervical dilatation progress more slowly, it
fessional accountability (Lavender 1999; Orhue 2012; Rakotonirina
would cross this alert line and arrangements were made to transfer
2013; Rotich 2011; Yisma 2013). These positive aspects may con-
her from a peripheral unit to a central unit where prolonged labour
tribute to improving maternal and fetal outcomes. Conversely, it
could be managed. The next stage of partograph development was
has been reported that the partograph's status within some ob-
the introduction of an 'action line', four hours to the right of the alert
stetric units is such that they may restrict clinical practice, create
line (Philpott 1972c). This line was developed to identify primary in-
unnecessary interference, reduce midwife autonomy and limit the
efficient uterine activity to prompt appropriate management. Cor-
flexibility to treat each woman as an individual (Lavender 1999;
rection of primary inefficient uterine activity would usually be with
Lavender 2011; Walraven 1994; Walsh 1994), factors which could al-
an intervention such as amniotomy or oxytocin infusion, or both.
so impact on clinical and psychological outcomes.
There have been a number of challenges associated with parto-
Barriers and facilitators of partograph use were explored in a re-
graph completion, including shortages of human resources, low
cent realist review (Bedwell 2017), which included 92 papers de-
status within labour wards and inadequate training (Bedwell 2017;
scribing studies from low- and high-income settings. A realist re-
Fatusi 2007; Lavender 2011). These challenges have resulted in a
view looks to explore what it is about a particular intervention
number of adaptations to the original partograph, one of which is
that works or does not work, who it works for, and in which set-
the simplified partograph, which does not include monitoring of
tings (Pawson 2004). Importantly, this realist review highlighted
the latent phase of labour (WHO 2003). In a small cross-over tri-
the fact that health provider support for using the partograph of-
al, this simplified partograph was shown to be more 'user-friend-
ten did not translate into practice. The review highlighted the es-
ly' (Mathews 2007). More recently, a randomised controlled trial, in
sential core components necessary within the healthcare environ-
India (Kenchaveeriah 2011), comparing the simplified (not includ-
ment which are required to support partograph use. These compo-
ing latent phase) with the traditional partograph (including latent
nents can be broadly categorised into health worker acceptability,
phase), confirmed a preference amongst medical staff for using the
health system support, effective referral systems, human resources
simplified version.
and health worker competence.
How the intervention might work
Why it is important to do this review
The partograph has been heralded as one of the most important
The partograph has become an integral part of routine labour care
advances in modern obstetric care (Agarwal 2013); however, this
in most parts of the world; assessment of its efficacy is therefore
was prior to any rigorous evaluation. Furthermore, the majority of
imperative.
early studies took place in hospital settings where most maternal
deaths occur among women admitted with severe complications Different designs of partograph exist, and Cartmill 1992 hypothe-
and often neglected labour (Lennox 1995). More than 20 years af- sised that the way a partograph is presented may affect an obste-
ter its introduction, and using a partograph adapted from that for- trician's perception of the labour progress and thus influence deci-
mulated by Philpott and Castle (Philpott 1972b; Philpott 1972c), the sion-making. This hypothesis has received some support from oth-
World Health Organization (WHO) conducted a prospective non- ers (Lavender 1998b; Tay 1996), who have suggested that the slope
randomised study of 35,484 women in South-East Asia (WHO 1994), and position of the action line have an impact on caesarean section
and concluded that the partograph was a necessary tool in the and intervention rates and maternal satisfaction.
management of labour, and recommended its universal applica-
tion. In this study, four pairs of hospitals participated (two pairs in The aim of this review is to assess the benefits and harms of parto-
Indonesia, one each in Thailand and Malaysia). A staged approach graph use on women in labour and to enable women and clinicians
was adopted, whereby for the first five months of the study, all eight to make informed evidence-based decisions. This review is the lat-
centres collected baseline data; after five months, the partograph est update of a review that was first published in 2008.
was randomly introduced into one of each pair; in the remaining
five months, the partograph was introduced into all hospital sites. OBJECTIVES
Introduction of the partograph, and agreed management protocol,
reduced prolonged labour (from 6.4% to 3.4%), the proportion of The primary objective was to determine the effectiveness and safe-
ty of partograph use on perinatal and maternal morbidity and mor-
Informed decisions.

tality. The secondary objective was to determine which partograph Secondary outcomes
design is most effective for perinatal and maternal morbidity and
Outcomes for mother
mortality outcomes.
Short-term maternal outcomes
METHODS
• Serious maternal morbidity or death (e.g. ruptured uterus, ad-
Criteria for considering studies for this review mission to intensive care unit, septicaemia, organ failure)
• Caesarean section for fetal distress
Types of studies • Caesarean section for delay in labour
We included in this review all published, unpublished and ongo- • Instrumental vaginal birth
ing randomised, quasi-randomised, and cluster-controlled trials • Vaginal birth not achieved within 24 hours, from onset of labour
that compared outcomes, as listed below, between partograph use (as defined by trial authors)
and non-use. We included randomised controlled trials of differ-
• Postpartum haemorrhage (as defined by the trial authors)
ent designs of partograph for secondary analysis. We included trials
that used quasi-random allocations (e.g. alternation). We excluded • Blood transfusion
studies reported in abstract form, without sufficient information on • Regional analgesia
study methods or where results were not clear, only after an unsuc- • Opioid use
cessful attempt to contact the study author for further information. • Duration of rupture of the membranes at the time of birth
We also excluded cross-over trials. • Performance of artificial rupture of the membranes during
labour
Types of participants
• Deep venous thrombosis
All women with singleton pregnancies and cephalic presentations,
• Pulmonary embolism
in spontaneous labour at term.
• Antibiotic use
Types of interventions • Duration of second stage of labour
We compared labour management using a partograph with labour • Number of vaginal examinations in labour
management where no partograph was used. The two groups had • Episiotomy
to differ only in the partograph usage and not in other labour ward • Third- and fourth-degree tears
interventions, such as psychological support, early amniotomy or • Shoulder dystocia
use of analgesia.
Long-term maternal outcomes
To meet the second objective, we included studies reporting com-
parisons between different designs of partograph. • Postnatal depression (as defined by trial authors)
• Breastfeeding failure (as defined by trial authors)
These are complex interventions. The partograph will be used in a • Fistulae
way dictated by the accompanying guidelines and this may influ- • Perineal pain
ence outcomes. Therefore, wherever possible, we have contextu-
alised trial findings by describing the associated clinical guidelines. • Dyspareunia
• Abdominal pain
Types of outcome measures • Backache reported six weeks postnatal
Primary outcomes • Prolapse or urinary incontinence
• Faecal incontinence
Outcomes for mother
• Relationship with baby (as defined by trial authors)
Short-term maternal outcomes
• Subsequent pregnancy complications
• Caesarean section • Postpartum rehospitalisation
• Oxytocin augmentation
Outcomes for baby
• Duration of first stage of labour (length of labour greater than 18
hours, length of labour greater than 12 hours) • Stillbirth or neonatal death or neonatal morbidity, excluding
• Maternal experience of childbirth (as defined by trial authors) fatal malformations (e.g. seizures, birth asphyxia, neonatal en-
cephalopathy)
Outcome for baby • Admission to special care nursery
Short-term neonatal outcomes • Need for intubation at birth
• Neonatal septicaemia
• Low Apgar score (less than 7 at 5 minutes)
• Intrapartum fetal death
• Jaundice (as defined by trial authors)
• Cord blood arterial pH less than 7.1
• Birth trauma (e.g. Erb's palsy, fractured skull, cephal-
haematoma, fractured clavicle)
• Childhood disability (as defined by trial author)
Informed decisions.

Staff For this update, we used the following methods for assessing the 13
relevant reports that we identified as a result of the updated search.
• Usability
• Ability to audit The following methods section of this review is based on a standard
template used by Cochrane Pregnancy and Childbirth.
Search methods for identification of studies
Selection of studies
The following methods section of this review is based on a standard
template used by Cochrane Pregnancy and Childbirth. Two review authors independently assessed for inclusion all the
potential studies identified as a result of the search strategy. We re-
Electronic searches solved any disagreement through discussion or, if required, we con-
We searched Cochrane Pregnancy and Childbirth's Trials Register sulted the third review author. When one review author was also
by contacting their Information Specialist (31 August 2017). the trial author (Lavender 1998a; Lavender 2006), the two remain-
ing authors assessed the studies independently.
The Register is a database containing over 24,000 reports of con-
trolled trials in the field of pregnancy and childbirth. For full search Data extraction and management
methods used to populate Pregnancy and Childbirth's Trials Regis- We designed a form to extract data. For eligible studies, two re-
ter, including the detailed search strategies for CENTRAL, MEDLINE, view authors extracted the data using the agreed form. We resolved
Embase and CINAHL; the list of handsearched journals and confer- discrepancies through discussion or, if required, we consulted the
ence proceedings, and the list of journals reviewed via the current third review author. We entered data into Review Manager 5 soft-
awareness service, please follow this link to the editorial informa- ware (Review Manager 2014), and checked for accuracy.
tion about Cochrane Pregnancy and Childbirth in the Cochrane Li-
brary and select the 'Specialized Register' section from the options When information regarding any of the above was unclear, we
on the left side of the screen. planned to contact authors of the original reports to provide further
details.
Briefly, Cochrane Pregnancy and Childbirth's Trials Register is
maintained by their Information Specialist and contains trials iden- Assessment of risk of bias in included studies
tified from:
Two review authors (AC, RS) independently assessed risk of bias for
1. monthly searches of the Cochrane Central Register of Controlled each study using the criteria outlined in the Cochrane Handbook for
Trials (CENTRAL); Systematic Reviews of Interventions (Higgins 2011). Any disagree-
2. weekly searches of MEDLINE (Ovid); ment was resolved by discussion or by involving the third review
author (TL).
3. weekly searches of Embase (Ovid);
4. monthly searches of CINAHL (EBSCO); (1) Random sequence generation (checking for possible
5. handsearches of 30 journals and the proceedings of major con- selection bias)
ferences; We described the method used to generate the allocation sequence
6. weekly current awareness alerts for a further 44 journals plus in sufficient detail to allow an assessment of whether it should pro-
monthly BioMed Central email alerts. duce comparable groups.
Two people screen the search results and review the full text of all For each included study we assessed the method as being at:
relevant trial reports identified through the searching activities de-
scribed above. Based on the intervention described, each trial re- • low risk of bias (any truly random process, e.g. random number
port is assigned a number that corresponds to a specific Pregnan- table; computer random number generator);
cy and Childbirth review topic (or topics), and is then added to the • high risk of bias (any non-random process, e.g. odd or even date
Register. The Information Specialist searches the Register for each of birth; hospital or clinic record number);
review using this topic number rather than keywords. This results • unclear risk of bias.
in a more specific search set that has been fully accounted for in the
relevant review sections (Included studies; Excluded studies; Stud- (2) Allocation concealment (checking for possible selection bias)
ies awaiting classification; Ongoing studies).
For each included study we described the method used to con-
In addition, we searched ClinicalTrials.gov and the World Health Or- ceal allocation to interventions prior to assignment and assessed
ganization (WHO) International Clinical Trials Registry Platform (IC- whether intervention allocation could have been foreseen in ad-
TRP) (31 August 2017), for unpublished, planned and ongoing trial vance of, or during recruitment, or changed after assignment.
reports using the search string: partogram OR partograph.
We assessed the methods as being at:
Searching other resources
• low risk of bias (e.g. telephone or central randomisation; con-
We searched the reference lists of retrieved studies. We did not ap- secutively numbered sealed opaque envelopes);
ply any language or date restrictions. • high risk of bias (open random allocation; unsealed or non-
opaque envelopes, alternation; date of birth);
Data collection and analysis
• unclear risk of bias.
For methods used in the previous version of this review, see Laven-
der 2013.
Informed decisions.

(3.1) Blinding of participants and personnel (checking for not prespecified; outcomes of interest were reported incom-
possible performance bias) pletely and so could not be used; study failed to include results
of a key outcome that would have been expected to have been
For each included study we described the methods used, if any, to
reported);
blind study participants and personnel from knowledge of which
intervention a participant received. We considered that studies • unclear risk of bias.
were at low risk of bias if they were blinded, or if we judged that the
(6) Other bias (checking for bias due to problems not covered by
lack of blinding was unlikely to affect results. We assessed blinding
(1) to (5) above)
separately for different outcomes or classes of outcomes.
For each included study we described any important concerns we
We assessed the methods as being at: had about other possible sources of bias.
• low, high or unclear risk of bias for participants; (7) Overall risk of bias
• low, high or unclear risk of bias for personnel.
We made explicit judgements about whether studies were at high
(3.2) Blinding of outcome assessment (checking for possible risk of bias, according to the criteria given in the Cochrane Hand-
detection bias) book for Systematic Reviews of Interventions (Higgins 2011). With
reference to (1) to (6) above, we planned to assess the likely mag-
For each included study we described the methods used, if any, to nitude and direction of the bias and whether we considered it was
blind outcome assessors from knowledge of which intervention a likely to have an impact on the findings. In future updates, we will
participant received. We assessed blinding separately for different explore the impact of the level of bias through undertaking sensi-
outcomes or classes of outcomes. tivity analyses (Sensitivity analysis).
We assessed methods used to blind outcome assessment as being Measures of treatment effect
at:
Dichotomous data
• low, high or unclear risk of bias.
For dichotomous data, we presented results as summary risk ratio
(4) Incomplete outcome data (checking for possible attrition (RR) with 95% confidence intervals (CIs).
bias due to the amount, nature and handling of incomplete
Continuous data
outcome data)
We used the mean difference (MD) if outcomes were measured in
For each included study, and for each outcome or class of out-
the same way between trials. If necessary, in future updates we will
comes, we described the completeness of data, including attri-
use the standardised mean difference (SMD) to combine trials that
tion and exclusions from the analysis. We stated whether attrition
measured the same outcome, but used different methods.
and exclusions were reported and the numbers included in the
analysis at each stage (compared with the total randomised par- Unit of analysis issues
ticipants), reasons for attrition or exclusion where reported, and
whether missing data were balanced across groups or were related Cluster-randomised trials
to outcomes. Where sufficient information was reported, or could We did not identify any cluster-randomised trials. In future updates,
be supplied by the trial authors, we planned to re-include missing if we identify any cluster-randomised trials, we will include them in
data in the analyses that we undertook. the analyses along with individually-randomised trials. We will ad-
just their sample sizes using the methods described in the Cochrane
We assessed methods as being at:
Handbook for Systematic Reviews of Interventions (Higgins 2011),
• low risk of bias (e.g. no missing outcome data; missing outcome Section 16.3.4, using an estimate of the intracluster correlation co-
data balanced across groups); efficient (ICC) derived from the trial (if possible), from a similar trial
or from a study of a similar population. If we use ICCs from other
• high risk of bias (e.g. numbers or reasons for missing data im-
sources, we will report this and conduct sensitivity analyses to in-
balanced across groups; ‘as treated’ analysis done with substan-
vestigate the effect of variation in the ICC. If we identify both clus-
tial departure of intervention received from that assigned at ran-
ter-randomised trials and individually-randomised trials, we plan
domisation);
to synthesise the relevant information. We will consider it reason-
• unclear risk of bias. able to combine the results from both if there is little heterogeneity
between the study designs and the interaction between the effect
(5) Selective reporting (checking for reporting bias)
of intervention and the choice of randomisation unit is considered
For each included study we described how we investigated the pos- to be unlikely.
sibility of selective outcome reporting bias and what we found.
We will also acknowledge heterogeneity in the randomisation unit
We assessed the methods as being at: and perform a sensitivity analysis to investigate the effects of the
randomisation unit.
• low risk of bias (where it was clear that all of the study's prespec-
ified outcomes and all expected outcomes of interest to the re- Cross-over trials
view have been reported);
Cross-over trials are not a valid design for this review and we ex-
• high risk of bias (where not all the study's prespecified outcomes cluded them.
were reported; one or more reported primary outcomes were
Informed decisions.

Other unit of analysis issues Subgroup analysis and investigation of heterogeneity

Multiple pregnancies If we identified substantial heterogeneity, we planned to investi-
Women with a multiple pregnancy were not eligible for inclusion in gate it using subgroup analyses and sensitivity analyses and to con-
this review. sider whether an overall summary was meaningful, and if it was, we
used random-effects analysis.
Trials with multiple treatment arms
We planned to carry out the following subgroup analyses.
Lavender 1998a was a three-arm trial. For each analysis, we used
the data of one intervention pair and excluded the others, following • Resource setting: low-resource setting versus high-resource set-
methods outlined in section 16.5.4 of the Cochrane Handbook for ting.
Systematic Reviews of Interventions (Higgins 2011).
We planned to use the following outcomes in subgroup analyses:
Dealing with missing data caesarean section, oxytocin augmentation, duration of first stage of
labour, maternal experience of childbirth, Apgar score and admis-
For included studies, we noted levels of attrition. In future updates,
sion to special care nursery.
if we include more eligible studies, we will explore the impact of
including studies with high levels of missing data in the overall as- We planned to assess subgroup differences by interaction tests
sessment of treatment effect by using sensitivity analysis. available within Review Manager (Review Manager 2014), and to re-
port the results of subgroup analyses, quoting the Chi2 statistic and
For all outcomes, we carried out analyses, as far as possible, on
P value, and the interaction test I2 value.
an intention-to-treat basis, i.e. we attempted to include all partici-
pants randomised to each group in the analyses. The denominator Given the small number of studies in each analysis, we did not
for each outcome in each trial was the number randomised minus perform our planned subgroup analysis as it would not produce a
any participants whose outcomes were known to be missing. meaningful analysis. We did, however, present the data under high-
and low-resource setting subheadings. If more studies are identi-
Assessment of heterogeneity
fied in future updates, we will perform a subgroup analysis.
We assessed statistical heterogeneity in each meta-analysis using
the Tau2, I2 and Chi2 statistics. We regarded heterogeneity as sub- Sensitivity analysis
stantial if I2 was greater than 30% and either Tau2 was greater than We planned to carry out sensitivity analyses to explore the effect
zero, or there was a low P value (less than 0.10) in the Chi2 test of allocation concealment and attrition bias on overall analyses, so
for heterogeneity. If we identified substantial heterogeneity (above that studies at high risk of bias for these domains were excluded
30%), we planned to explore it by prespecified subgroup analysis. from the analyses to see if this made any difference to the overall
results. Two studies did not conceal allocation adequately (Sinha
Assessment of reporting biases
2017; Walss Rodriguez 1987), and we removed them as part of this
In future updates, if there are 10 or more studies in the meta-analy- sensitivity analysis.
sis we will investigate reporting biases (such as publication bias)
using funnel plots. We will assess funnel plot asymmetry visually. Assessment of the quality of the evidence using the GRADE
If asymmetry is suggested by a visual assessment, we will perform approach
exploratory analyses to investigate it. For this update, we assessed the quality of the evidence using the
GRADE approach, as outlined in the GRADE handbook in order to
Data synthesis
assess the quality of the body of evidence relating to the follow-
We carried out statistical analysis using the Review Manager 5 soft- ing outcomes for the main comparison: partograph compared to no
ware (Review Manager 2014). We used fixed-effect meta-analysis partograph (studies carried out in high- and low-resource settings).
for combining data where it was reasonable to assume that studies
were estimating the same underlying treatment effect, i.e. where • Caesarean section (overall)
trials were examining the same intervention, we judged the trials' • Oxytocin augmentation
populations and methods sufficiently similar. • Duration of first stage of labour (length of labour greater than 18
hours, length of labour greater than 12 hours)
If there was clinical heterogeneity sufficient to expect that the un-
• Maternal experience of childbirth (as defined by trial authors)
derlying treatment effects differed between trials, or if substan-
tial statistical heterogeneity was detected, we used random-effects • Low Apgar score (less than 7 at 5 minutes)
meta-analysis to produce an overall summary if we considered • Serious maternal morbidity or death (e.g. ruptured uterus, ad-
an average treatment effect across trials clinically meaningful. We mission to intensive care unit, septicaemia, organ failure)
treated the random-effects summary as the average range of possi- • Stillbirth or neonatal death or neonatal morbidity, excluding
ble treatment effects and we planned to discuss the clinical impli- fatal malformations (e.g. seizures, birth asphyxia, neonatal en-
cations of treatment effects differing between trials. If the average cephalopathy)
treatment effect was not clinically meaningful, we planned not to
combine trials. If we used random-effects analyses, we presented We used GRADEpro Guideline Development Tool to import data
the results as the average treatment effect with 95% CIs, and the from Review Manager 5 (Review Manager 2014), in order to create
estimates of Tau2 and I2. 'Summary of findings' tables. We produced a summary of the in-
tervention effect and a measure of quality for each of the above
outcomes using the GRADE approach. The GRADE approach uses
Informed decisions.

five considerations (study limitations, consistency of effect, impre- RESULTS

cision, indirectness and publication bias) to assess the quality of
the body of evidence for each outcome. The evidence can be down- Description of studies
graded from 'high quality' by one level for serious (or by two levels
Results of the search
for very serious) limitations, depending on assessments for risk of
bias, indirectness of evidence, serious inconsistency, imprecision See: Figure 2.
of effect estimates or potential publication bias.

Informed decisions.

Figure 2. Study flow diagram.

Informed decisions.

Figure 2. (Continued)
Informed decisions.

We retrieved 13 relevant reports of 10 trials from an updated search unit in the Walss Rodriguez 1987 or Windrim 2006 studies; it is un-
in August 2017. Five of the trials met our inclusion criteria and have clear if the partograph was used routinely in the Rani 2015 study.
contributed data to this update (Lee 2015; Orhue 2013; Rani 2015;
Shazly 2017; Sinha 2017). Four studies compared partographs with different placement of
action lines (Lavender 1998a; Lavender 2006; Orhue 2013; Sinha
This review now includes 11 studies involving 9475 women 2017). The Lavender 2006 study was a two-arm trial and the Laven-
(Kenchaveeriah 2011; Lavender 1998a; Lavender 2006; Lee 2015; der 1998a study was a three-arm trial. Other than the placement
Orhue 2013; Pattinson 2003; Rani 2015; Shazly 2017; Sinha 2017; of the action line, labour management remained consistent. If
Walss Rodriguez 1987; Windrim 2006). Three studies are ongoing progress crossed the action line, a diagnosis of prolonged labour
(NCT02714270; NCT02741141; NTR5543), and we excluded eight was made and managed according to standard protocol; this in-
(Ajoodani 2011; Cartmill 1992; Fahdhy 2005; Hamilton 2001; Hamil- volved clinical assessment and augmentation, as appropriate.
ton 2004; Kogovsek 2000; Mathews 2007; WHO 1994), two in this
update (Ajoodani 2011; WHO 1994). Following the updated search, One study in South Africa compared a partograph with an alert
one study is awaiting assessment as we are awaiting further infor- and action line with one which contained an alert line only (Pat-
mation from authors (NCT02911272). tinson 2003). In this study, the group that received a partograph
with only an alert line received more aggressive intrapartum man-
Included studies agement; a vaginal examination was carried out every two hours
and oxytocin infusion advocated when progress crossed the line.
Methods
Those with an alert and action line had more expectant manage-
All trials were randomised controlled trials with individual ran- ment, vaginal examinations every four hours, and commencement
domisation. The Walss Rodriguez 1987 trial was quasi-randomised. of oxytocin if progress crossed the four-hour action line. The Pat-
The studies by Kenchaveeriah 2011 and Walss Rodriguez 1987 gen- tinson 2003 study was financially supported by the South African
erally lacked detail, making assessment of quality and contextual- Medical Research Council but did not report whether the authors
isation of the results difficult. disclosed any conflicts of interest. The Kenchaveeriah 2011 study,
conducted in India, compared two partographs - a composite par-
Participants and settings tograph including the latent phase with a modified one without the
Included studies took place in hospital settings in Australia (Lee latent phase. This trial was carried out in India were the use of the
2015), Canada (Windrim 2006), Egypt (Shazly 2017), India (Ken- partograph has not been incorporated and practiced widely, even
chaveeriah 2011; Rani 2015; Sinha 2017), Mexico (Walss Rodriguez at the tertiary level. The plotting of the composite partograph was
1987), Nigeria (Orhue 2013), South Africa (Pattinson 2003), and the started as soon as the woman was in labour. In the modified par-
UK (Lavender 1998a; Lavender 2006). tograph, the plotting of the partograph was started with at least 4
cm of cervical dilatation. Prolonged labour was defined when the
Three studies took place between 1995 and 2005 (Lavender 1998a; woman was in labour for more than 12 hours in the active phase.
Lavender 2006; Windrim 2006), two between 2005 and 2015 (Ken-
chaveeriah 2011; Rani 2015), two between 2015 and 2016 (Lee 2015; One study (Shazly 2017), conducted in a hospital in Egypt, com-
Shazly 2017), and one took place in 1985 (Walss Rodriguez 1987). pared a labour scale with a traditional WHO partograph. The labour
The Orhue 2013, Pattinson 2003, and Sinha 2017 trials did not re- scale was developed from the WHO partograph, and was designed
port the dates of their studies. to help clinicians recognise determinants and manage potential
labour dystocia (Shazly 2014). The labour scale has particular trig-
Most studies only included primiparous women with uncomplicat- ger points throughout first and second stage where management
ed, low-risk pregnancies in spontaneous labour (Lavender 1998a; is reviewed instead of having a fixed action line. Vaginal examina-
Lavender 2006; Lee 2015; Orhue 2013; Pattinson 2003; Shazly 2017; tions took place two-hourly in this trial. The labour scale monitors
Sinha 2017; Windrim 2006). The Kenchaveeriah 2011 and Walss labour by filling in boxes, corresponding with cervical dilatation
Rodriguez 1987 studies did not specify parity, but only included with dots, lines, or shading, depending on if there is long progress
women with uncomplicated pregnancies. The Rani 2015 study only (> 2 cm in two hours), short progress (1 cm in two hours), or no
included high-risk primiparous women. progress, respectively. Progress can be 'reset' if long progress is
identified regardless of previous progress. No progress after two
Interventions and comparisons hours was followed either by artificial rupture of membranes, or
oxytocin augmentation. For women undergoing artificial rupture of
Three studies compared partograph use with no partograph (Rani
membranes, oxytocin augmentation was not started unless there
2015; Walss Rodriguez 1987; Windrim 2006). The Rani 2015 study
was no change in dilatation at the next examination. Dashed lines
took place in India, the Windrim 2006 study in Canada, and the
at points throughout the scale recommend clinical re-evaluation:
Walss Rodriguez 1987 study in Mexico; therefore, they were from
"after 8 hours of the latent phase, at the membrane line, at the aug-
very different settings. The Windrim 2006 and Walss Rodriguez 1987
mentation line and after 1 hour of the second stage" (Shazly 2014).
studies both compared their usual descriptive, sequential, record-
We took most of the data for this trial from an unpublished trial re-
ing of intrapartum details, with an experimental arm, i.e. the par-
port provided by the authors.
tograph. The Rani 2015 study compared noting progress in the
women's case notes with utilising the World Health Organization One pilot study took place in Australia and compared a traditional
(WHO) modified partograph. In the Windrim 2006 study, the par- partograph with a two-hour action line (Lee 2015), with a dystocia
tograph used incorporated a two-hour alert line, but no action line. The dystocia line partograph starts when the woman's cervix is
line. The Walss Rodriguez 1987 study used a 'Friedman' partograph a minimum of 4 cm dilated. The stepped line steepens at 6 cm to ac-
(Friedman 1954). The partograph was not currently in use in either count for the "progressive acceleration" of labour. If the action line
or vertical step line were crossed, a vaginal examination was per-
Informed decisions.

formed in two hours to assess whether the line had been crossed, port from the South African Medical Research Council. None of the
and oxytocin augmentation was then discussed. other studies reported their funding sources.
Outcomes None of the studies reported if any authors had conflicts of interest.
Only two outcomes were reported by all trials; caesarean section Excluded studies
rates and Apgar score. Other outcomes were not consistently re-
ported; no trials reported vaginal birth not achieved within 24 We excluded eight studies from this review: two trials lacked suf-
hours, from onset of labour (as defined by trial authors), blood ficient detail to assess them adequately (Hamilton 2001; Kogov-
transfusion, duration of rupture of the membranes at the time of sek 2000), and we were unable to contact the authors; two studies
birth, deep venous thrombosis, pulmonary embolism, episiotomy, did not have partographs as an intervention (Fahdhy 2005; Hamil-
third- and fourth-degree tears, shoulder dystocia, neonatal septi- ton 2004); one used partographs in both trial arms (Ajoodani 2011);
caemia, jaundice, ability to audit, or any of the review's long-term one was a cross-over trial (Mathews 2007); and two were not ran-
maternal or baby outcomes. domised controlled trials (Cartmill 1992; WHO 1994).
Funding and conflicts of interest Risk of bias in included studies

The Windrim 2006 study was supported by a grant from the Physi- We assessed included studies for methodological quality on the ba-
cians' Services Incorporated Foundation, Canada, The Lavender sis of sequence generation, allocation concealment, blinding, at-
1998a and Lavender 2006 studies were funded by the Liverpool trition and other concerns about bias (see Methods section above,
Women's Hospital, and the Pattinson 2003 study had financial sup- and Figure 3).

Informed decisions.

Figure 3. Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Informed decisions.

Figure 3. (Continued)
Informed decisions.

Allocation were at unclear risk: the Lavender 1998a and Lavender 2006 trials
did not randomise all eligible women; the Orhue 2013, Sinha 2017,
We graded sequence generation as adequate (and therefore at low
and Walss Rodriguez 1987 trials did not provide sufficient informa-
risk of bias) in seven studies, with studies reporting the use of either
tion to adequately asses this domain; and the Windrim 2006 trial
a random number table or a computer random number generator
gave no information on the number of women approached or the
(Kenchaveeriah 2011; Lavender 1998a; Lavender 2006; Pattinson
numbers of eligible women declining participation. We considered
2003; Rani 2015; Shazly 2017; Windrim 2006), high risk in two due to
the Pattinson 2003 study to be at high risk of bias as recruitment
quasi-randomisation and lack of information provided (Sinha 2017;
was stopped before the required sample size was reached due to
Walss Rodriguez 1987), and unclear in two (Lee 2015; Orhue 2013).
funding constraints.
The Lee 2015 study used block randomisation, and the Orhue 2013
study gave no information. Effects of interventions
Allocation concealment was unclear in four trials (Kenchaveeriah See: Summary of findings for the main comparison Partograph
2011; Orhue 2013; Pattinson 2003; Rani 2015); low risk in five trials compared to no partograph (studies carried out in high- and low-
(Lavender 1998a; Lavender 2006; Lee 2015; Shazly 2017; Windrim resource settings) for women in spontaneous labour at term
2006); and high risk in the Sinha 2017 and Walss Rodriguez 1987
trials. The Walss Rodriguez 1987 study was quasi-randomised but No trials reported: vaginal birth not achieved within 24 hours,
the method of randomisation was not reported. The method of ran- from onset of labour (as defined by trial authors); blood transfu-
domisation was not clear in the Sinha 2017 study, however they sion; duration of rupture of the membranes at the time of birth;
reported the participants were “equally divided into two groups” deep venous thrombosis; pulmonary embolism; episiotomy, third-
which suggests quasi-randomisation. There was also no informa- and fourth-degree tears; shoulder dystocia; neonatal septicaemia;
tion in either study on how women were allocated to groups. jaundice; ability to audit; or any of the review's long-term maternal
or baby outcomes.
Blinding
1. Partograph versus no partograph
Due to the nature of the intervention, it was not feasible to blind the
women or clinical staff in any of the studies. Blinding of outcome as- We included three randomised trials in this comparison, with 1813
sessors was not attempted in three studies (Orhue 2013; Rani 2015; women participating (Rani 2015; Walss Rodriguez 1987; Windrim
Sinha 2017), and was not mentioned in five studies (Kenchaveeri- 2006). The Walss Rodriguez 1987 study reported only three out-
ah 2011; Pattinson 2003; Shazly 2017; Walss Rodriguez 1987; Win- comes, and the Rani 2015 study only two outcomes relevant to this
drim 2006). Three studies were at low risk of detection bias due to review, therefore we only pooled results for these outcomes. We
the blinding of the statistician or researcher involved in outcome assessed evidence using the GRADE approach for this comparison
assessment (Lavender 1998a; Lavender 2006; Lee 2015). (see Summary of findings for the main comparison).
Incomplete outcome data Primary outcomes

We judged attrition to be low in seven studies (Kenchaveeriah 2011; Caesarean section
Lavender 1998a; Lavender 2006; Lee 2015; Pattinson 2003; Walss There was no clear difference between groups in caesarean section
Rodriguez 1987; Windrim 2006), with less than 1% of participants rates (average risk ratio (RR) 0.77, 95% confidence interval (CI) 0.40
excluded or lost to follow-up. In one trial (Lavender 1998a), there to 1.46; n = 1813; 3 trials; I2 = 87%; very low-quality evidence; Analy-
were higher levels of missing data (13.5%) for the maternal expe- sis 1.1). There were high levels of heterogeneity for this result (het-
rience outcome. In this study, maternal experience was only as- erogeneity: Tau2 = 0.28; Chi2 = 15.07 (P = 0.0005); I2 = 87%) so this
sessed in a subset of women (n = 615); this comprised all women re- result should be interpreted with caution.
cruited over a prespecified 12-month period, of whom 519 respond-
ed. Four studies were at unclear risk of attrition bias either due to Oxytocin augmentation
lack of information in the trials reports (Orhue 2013; Rani 2015; Sin-
The Windrim 2006 study reported no clear difference between
ha 2017), or a small number of women withdrawing prior to the in-
groups in oxytocin augmentation (RR 1.02, 95% CI 0.95 to 1.10; n =
tervention and analysis (Shazly 2017).
1156; 1 trial; moderate-quality evidence; Analysis 1.2).
Selective reporting
Duration of first stage of labour
Seven studies appeared to report all planned outcomes and we The Windrim 2006 study reported no clear difference between
judged them to be at low risk of bias (Kenchaveeriah 2011; Lavender groups in duration of first stage of labour (mean difference (MD)
1998a; Lavender 2006; Lee 2015; Pattinson 2003; Rani 2015; Win- 0.80 hours, 95% CI -0.06 to 1.66; n = 1156; 1 trial; low-quality evi-
drim 2006). Reporting bias was unclear in the Walss Rodriguez 1987 dence; Analysis 1.3).
trial due to lack of prespecified outcomes reported. There was too
little information in the Orhue 2013 and Sinha 2017 trials to assess Maternal experience of childbirth
this domain adequately. We judged the Shazly 2017 study at high
risk of reporting bias as some of the trial's primary outcomes were Not reported in this comparison.
not reported. Low Apgar score (less than 7 at 5 minutes)
Other potential sources of bias There was no clear difference between groups in Apgar score less
than 7 at five minutes (RR 0.76, 95% CI 0.29 to 2.03; n = 1596; 2 trials;
We judged the Lee 2015, Kenchaveeriah 2011, Rani 2015, and Shaz-
very low-quality evidence; Analysis 1.4).
ly 2017 studies as being at low risk of other sources of bias. Six trials
Informed decisions.

Secondary outcomes n = 4749; 4 trials; I2 = 96%; Analysis 2.2). There is high heterogeneity
in this outcome and the results should be interpreted with caution
In addition to the general list of outcomes not reported in any of the
(Tau2 = 0.29; I2 = 96%). The effect was much greater for the women
trials (see under heading ‘Effects of Interventions’), no trials report-
in the two-hour action line group in the low-resource setting (av-
ed the following secondary outcomes for this comparison: serious
erage RR 7.22, 95% CI 2.49 to 20.91; n = 1148; 2 trials; I2 = 78%),
maternal morbidity or death; caesarean section for fetal distress;
than in the high-resource setting (average RR 1.14, 95% CI 1.05 to
caesarean section for delay in labour; instrumental vaginal birth;
1.22; n = 3601; 2 trials). However there was high heterogeneity with-
postpartum haemorrhage; opioid use; antibiotic use; stillbirth or
in the low-resource setting subgroup (Tau2 = 0.46; I2 = 78%) and this
neonatal death or neonatal morbidity, excluding fatal malforma-
should be interpreted with caution.
tions; need for intubation at birth; intrapartum fetal death; cord
blood arterial pH less than 7.1; birth trauma; childhood disability; Duration of first stage of labour
usability.
One trial reported "prolonged labour" (Orhue 2013). No clear differ-
Instrumental vaginal birth ence was observed between groups (RR 0.81, 95% CI 0.32 to 2.04; n
= 948; 1 trial; Analysis 2.3).
There was no clear difference between groups in instrumental vagi-
nal birth in high- or low-resource settings (RR 0.99, 95% CI 0.84 to Maternal experience of childbirth (reported as negative childbirth
1.15; n = 1813; 3 trials; Analysis 1.5). experience)
There was insufficient evidence of benefit or harm in any of the oth- There was no clear difference in number of women reporting neg-
er secondary maternal or neonatal outcomes, reported by the Win- ative childbirth experience between groups (average RR 0.61, 95%
drim 2006 study. CI 0.28 to 1.35; n = 2269; 2 trials; I2 = 83%; Analysis 2.4). Heterogene-
ity was high for the outcome (Tau2 = 0.27; I2 = 83%) and the results
• Regional analgesia - epidural (RR 1.01, 95% CI 0.98 to 1.05; n = should be interpreted with caution.
1156; 1 trial; Analysis 1.6).
Low Apgar score (less than 7 at 5 minutes)
• Performance of artificial rupture of membranes (RR 0.99, 95% CI
0.88 to 1.11; n = 1156; 1 trial; Analysis 1.7). There was no clear difference between groups in babies with low
• Antibiotic use (RR 1.23, 95% CI 0.88 to 1.73; n = 1156; 1 trial; Apgar scores (RR 0.93, 95% CI 0.61 to 1.42; n = 4749; 4 trials; Analysis
Analysis 1.8). 2.5).
• Duration of second stage of labour (MD 0.00 hours, 95% CI -0.21 Secondary outcomes
to 0.21; n = 1156; 1 trial; Analysis 1.9).
• Number of vaginal examinations in labour (mean of 4 examina- In addition to the general list of outcomes not reported in any of the
tions in labour for both groups; Analysis 1.10). trials (see under heading ‘Effects of Interventions’), no trials report-
ed the following secondary outcomes for this comparison: opioid
• Admission to special care nursery (RR 0.94, 95% CI 0.51 to 1.75;
use; antibiotic use; duration of second stage of labour; need for in-
n = 1156; 1 trial; Analysis 1.11).
tubation at birth; intrapartum fetal death; birth trauma; childhood
Sensitivity analysis disability; usability.
The Walss Rodriguez 1987 study had poor allocation concealment There were no clear differences in any of the secondary maternal or
and provided very little information on study methods. In view of neonatal outcomes reported in these trials.
the high risk of bias associated with this study, we carried out a
sensitivity analysis excluding it from Analysis 1.1, Analysis 1.4, and • Serious maternal morbidity or death (no events in either group;
Analysis 1.5. There were no clear differences between groups when Analysis 2.6).
we removed this study. • Caesarean section for fetal distress (RR 1.30, 95% CI 0.86 to 1.96;
n = 3601; 2 trials; Analysis 2.7).
2. Partograph with two-hour action line versus partograph
• Caesarean section for delay in labour (RR 0.98, 95% CI 0.77 to
with four-hour action line
1.25; n = 3601; 2 trials; Analysis 2.8).
Four randomised trials were included in this comparison with 4749 • Instrumental vaginal birth (RR 0.92, 95% CI 0.81 to 1.04; n = 3801;
women participating (Lavender 1998a; Lavender 2006; Orhue 2013; 3 trials; Analysis 2.9).
Sinha 2017). Two studies were carried out in the same high-re- • Postpartum haemorrhage - blood loss > 500 mL (RR 1.06, 95% CI
source setting (Lavender 1998a; Lavender 2006); two studies were 0.90 to 1.25; n = 4549; 3 trials; Analysis 2.10).
carried out in a low-resource setting (Orhue 2013; Sinha 2017).
• Regional analgesia - epidural (average RR 1.06, 95% CI 0.92 to
Primary outcomes 1.21; n = 3601; 2 trials; I2 = 35%; Analysis 2.11). Heterogeneity is
over 30% for this outcome (Tau2 = 0.00; I2 = 35%) and the results
Caesarean section should be interpreted cautiously.
There was no clear difference in caesarean section rates between • Performance of artificial rupture of membranes (average RR
the groups (average RR 1.06, 95% CI 0.88 to 1.28; n = 4749; 4 trials; 1.00, 95% CI 0.84 to 1.18; n = 3801; 3 trials; I2 = 55%; Analysis 2.12).
Analysis 2.1). Heterogeneity is over 30% for this outcome (Tau2 = 0.01; I2 = 55%)
and the results should be interpreted cautiously.
Oxytocin augmentation
• Number of vaginal examinations in labour (MD -0.08, 95% CI
Women in the two-hour action line group were more likely to re- -0.37 to 0.21; random-effects; n = 3601; 2 trials; I2 = 70%; Tau2 =
ceive oxytocin augmentation (average RR 2.44, 95% CI 1.36 to 4.35;
Informed decisions.

0.03; Analysis 2.13). High heterogeneity has been noted for this There were no clear differences in any of the secondary maternal or
outcome and the results should be interpreted cautiously. neonatal outcomes reported in this trial.
• Serious neonatal morbidity or perinatal death (no events in ei-
ther group; Analysis 2.14). • Serious maternal morbidity or death (no events in either group;
Analysis 3.5).
• Admission to special care nursery (RR 0.83, 95% CI 0.51 to 1.34;
n = 3801; 3 trials; Analysis 2.15). • Caesarean section for fetal distress (RR 0.96, 95% CI 0.44 to 2.10;
• Cord blood arterial pH less than 7.1 (RR 0.73, 95% CI 0.44 to 1.22;
n = 3601; 2 trials; Analysis 2.16). • Caesarean section for delay in labour (RR 0.71, 95% CI 0.42 to
1.19; n = 617; 1 trial; Analysis 3.7).
Sensitivity analysis • Instrumental vaginal birth (RR 0.93, 95% CI 0.69 to 1.26; n = 617;
1 trial; Analysis 3.8).
The Sinha 2017 trial did not give clear information on allocation
concealment but states that women were "equally divided into two • Postpartum haemorrhage - blood loss > 500 mL (RR 0.96, 95% CI
groups" which suggests quasi-randomisation was used. In view of 0.63 to 1.45; n = 617; 1 trial; Analysis 3.9).
the high risk of bias associated with this study, we carried out a sen- • Regional analgesia - epidural (RR 1.16, 95% CI 0.94 to 1.44; n =
sitivity analysis, excluding it from Analysis 2.1, Analysis 2.2, Analy- 617; 1 trial; Analysis 3.10).
sis 2.5, Analysis 2.9, Analysis 2.12, and Analysis 2.15. There were • Performance of artificial rupture of membranes during labour
no clear differences between groups when we removed this study. (RR 0.94, 95% CI 0.77 to 1.15; n = 617; 1 trial; Analysis 3.11).
Performance of artificial rupture of membranes (Analysis 2.12), ap- • Number of vaginal examinations in labour (MD 0.00, 95% CI -0.29
peared to favour the four-hour partograph when the Sinha 2017 tri- to 0.29; n = 617; 1 trial; Analysis 3.12).
al was removed, however, the lower CI still touched the line of no • Serious neonatal morbidity or perinatal death (no events in ei-
effect. ther group; Analysis 3.13).
3. Partograph with two-hour action line versus partograph • Admission to special care nursery (RR 3.83, 95% CI 0.43 to 34.12;
with three-hour action line n = 617; 1 trial; Analysis 3.14).
• Cord blood arterial pH less than 7.1 (RR 0.38, 95% CI 0.07 to 1.96;
Only one randomised trial (carried out in a high-resource setting)
compared a two-hour versus a three-hour action line, with 617
women participating (Lavender 1998a). 4. Partograph with three-hour action line versus partograph
with four-hour action line
Primary outcomes
Only one randomised trial, again carried out in a high-resource set-
Caesarean section
ting, compared a three-hour versus a four-hour action line, with 613
There was no clear difference in overall caesarean section rate (RR women participating (Lavender 1998a).
0.78, 95% CI 0.51 to 1.18; n = 617; 1 trial; Analysis 3.1).
Primary outcomes
Caesarean section
There was no clear difference between groups for oxytocin aug-
Caesarean section rate was higher in the three-hour action line
mentation (RR 1.02, 95% CI 0.85 to 1.21; n = 617; 1 trial; Analysis 3.2).
group (RR 1.70, 95% CI 1.07 to 2.70; n = 613; 1 trial; Analysis 4.1).
Not reported in this comparison.
There was no clear difference in oxytocin augmentation between
Maternal experience of childbirth (reported as negative childbirth the groups (RR 1.09, 95% CI 0.91 to 1.30; n = 613; 1 trial; Analysis 4.2).
experience)
Women in the two-hour action line group were less likely to report a
negative childbirth experience than those in the three-hour action Not reported in this comparison.
line group (RR 0.49, 95% CI 0.27 to 0.90; n = 348; 1 trial; Analysis 3.3). Maternal experience of childbirth (reported as negative childbirth
experience)
There was no clear difference in negative childbirth experiences be-
There was no clear difference between the groups in the number of
tween the groups (RR 0.80, 95% CI 0.51 to 1.27; n = 340; 1 trial; Analy-
babies with Apgar scores of less than 7 at five minutes (RR 1.44, 95%
sis 4.3). Loss to follow-up was noticeably high for this outcome.
CI 0.41 to 5.05; n = 617; 1 trial; Analysis 3.4).
Secondary outcomes
There was no clear difference between the groups in the number of
In addition to the general list of outcomes not reported in any of
babies with Apgar scores of less than 7 at five minutes (RR 0.82, 95%
the trials (see under heading ‘Effects of Interventions’), this trial did
CI 0.22 to 3.04; n = 613; 1 trial; Analysis 4.4).
not report the following secondary outcomes for this comparison:
opioid use; antibiotic use; duration of second stage of labour; need Secondary outcomes
for intubation at birth; intrapartum fetal death; birth trauma; child-
hood disability; usability. In addition to the general list of outcomes not reported in any of
the trials (see under heading ‘Effects of Interventions’), this trial did
Informed decisions.

not report the following secondary outcomes for this comparison: Low Apgar score (less than 7 at five minutes)
opioid use; antibiotic use; duration of second stage of labour; need There was no clear difference in number of low Apgar scores at five
for intubation at birth; intrapartum fetal death; birth trauma; child- minutes between the groups (RR 7.12, 95% CI 0.37 to 137.36; n =
hood disability; usability. 694; 1 trial; Analysis 5.3).
There were no clear differences between the groups for any of the Secondary outcomes
reported secondary outcomes.
• Serious maternal morbidity or death (no events in either group; the trials (see under heading ‘Effects of Interventions’), no trials re-
Analysis 4.5). ported the following secondary outcomes for this comparison: se-
• Caesarean section for fetal distress (RR 1.77, 95% CI 0.70 to 4.42; rious maternal morbidity or death; caesarean section for fetal dis-
n = 613; 1 trial; Analysis 4.6). tress; caesarean section for delay in labour; postpartum haemor-
• Caesarean section for delay in labour (RR 1.68, 95% CI 0.97 to rhage; regional analgesia; opioid use; performance of artificial rup-
2.91; n = 613; 1 trial; Analysis 4.7). ture of the membranes during labour; antibiotic use; duration of
• Instrumental vaginal birth (RR 0.96, 95% CI 0.72 to 1.28; n = 613; second stage of labour; number of vaginal examinations in labour;
one trial; Analysis 4.8). stillbirth or neonatal death or neonatal morbidity, excluding fatal
malformations; admission to special care nursery; need for intuba-
• Postpartum haemorrhage - blood loss > 500 mL (RR 1.03, 95% CI tion at birth; intrapartum fetal death; cord blood arterial pH less
0.68 to 1.56; n = 613; 1 trial; Analysis 4.9). than 7.1; birth trauma; childhood disability; usability
• Regional analgesia - epidural (RR 1.01, 95% CI 0.80 to 1.27; n =
613; 1 trial; Analysis 4.10). There were no clear differences in any of the secondary maternal or
• Performance of artificial rupture of membranes during labour neonatal outcomes reported in this trial.
(RR 1.04, 95% CI 0.85 to 1.26; n = 613; 1 trial; Analysis 4.11).
• Instrumental vaginal birth (RR 0.87, 95% CI 0.66 to 1.15; n = 694;
• Number of vaginal examinations in labour (MD 0.10, 95% CI -0.19
1 trial; Analysis 5.4).
to 0.39; n = 613; 1 trial; Analysis 4.12).
• Perinatal death (RR 7.12, 95% CI 0.37 to 137.36; n = 694; 1 trial;
• Serious neonatal morbidity or perinatal death (no events in ei-
Analysis 5.5).
ther group; Analysis 4.13).
• Admission to special care nursery (RR 0.51, 95% CI 0.05 to 5.65; 6. Partograph with latent phase versus partograph without
n = 613; 1 trial; Analysis 4.14). latent phase
• Cord blood arterial pH less than 7.1 (RR 2.57, 95% CI 0.50 to Only one study examined the comparison between partograph
13.17; n = 613; 1 trial; Analysis 4.15). with latent phase (composite) versus partograph without latent
5. Partograph with alert line only versus partograph with alert phase (modified) (Kenchaveeriah 2011), and 743 women participat-
and action line ed.
Only one randomised trial compared a partograph with an alert Primary outcomes
line only versus a partograph with an alert and action line, with 694 Caesarean section
women participating (Pattinson 2003). This trial was carried out in
a low-resource setting. The caesarean section rate was higher in the partograph with latent
phase (composite) group (95% RR 2.45, 95% CI 1.72 to 3.50; n = 743;
Primary outcomes 1 trial; Analysis 6.1).
Caesarean section Oxytocin augmentation
The caesarean section rate was lower in the alert line only group The partograph with latent phase condition produced more inci-
(RR 0.68, 95% CI 0.50 to 0.93; n = 694; 1 trial; Analysis 5.1). dences of augmentation of labour (RR 2.18, 95% CI 1.67 to 2.83; n =
743; 1 trial; Analysis 6.2).
There was no clear difference when oxytocin augmentation was Duration of first stage of labour
managed aggressively, with the use of a single alert and action line, Not reported in this comparison.
or conservatively with separate alert and action lines (RR 0.81, 95%
CI 0.62 to 1.05; n = 694; 1 trial; Analysis 5.2). Maternal experience of childbirth
Duration of first stage of labour Not reported in this comparison.
Not reported in this comparison. Low Apgar score (less than 7 at 5 minutes)
Maternal experience of childbirth There was no clear advantage for one condition with respect to Ap-
gar score less than 7 at five minutes (RR 0.75, 95% CI 0.21 to 2.63; n
Not reported in this comparison. = 743; 1 trial; Analysis 6.3).
Secondary outcomes
the trials (see under heading ‘Effects of Interventions’), no trials re-
Informed decisions.

ported the following secondary outcomes for this comparison: se- Duration of first stage of labour (labour longer than 12 hours)
rious maternal morbidity or death; postpartum haemorrhage; re- There was no clear difference between the groups in duration of the
gional analgesia; opioid use; performance of artificial rupture of the first stage of labour (labour longer than 12 hours) (RR 0.76, 95% CI
membranes during labour; antibiotic use; duration of second stage 0.31 to 1.89; n = 99; 1 trial; Analysis 7.3).
of labour; number of vaginal examinations in labour; stillbirth or
neonatal death or neonatal morbidity, excluding fatal malforma- Maternal experience of childbirth (Birth Satisfaction Score (BSS-R))
tions; need for intubation at birth; intrapartum fetal death; cord
blood arterial pH less than 7.1; birth trauma; childhood disability. Women in both groups gave similar birth satisfaction scores in a
survey completed in the immediate postpartum period (MD 0.00,
Caesarean section for fetal distress 95% CI -3.58 to 3.58; n = 90; 1 trial; Analysis 7.4).
The level of caesarean section fetal distress was higher in the par- Low Apgar score (less than 4 at 4 minutes) (non-prespecified outcome)
tograph with latent phase (composite) (RR 4.87, 95% CI 2.83 to 8.37;
n = 743; 1 trial; Analysis 6.4). One baby in each group had an Apgar score less than 4 at four min-
utes (RR 0.98, 95% CI 0.06 to 15.23; n = 99; 1 trial; Analysis 7.5). An
Caesarean section for delay in labour Apgar score below 7 at five minutes was not reported.
There was no clear difference between the groups in the caesarean Secondary outcomes
section rate for delay in labour (RR 1.35, 95% CI 0.59 to 3.08; n = 743;
1 trial; Analysis 6.5). In addition to the general list of outcomes not reported in any of the
trials (see under heading ‘Effects of Interventions’), no trials report-
Instrumental vaginal birth ed the following secondary outcomes for this comparison: serious
maternal morbidity or death; caesarean section for fetal distress;
There was no clear advantage for one condition with respect to in-
caesarean section for delay in labour; performance of artificial rup-
strumental vaginal birth (RR 1.04, 95% CI 0.61 to 1.77; n = 743; 1 tri-
ture of the membranes during labour; antibiotic use; duration of
al; Analysis 6.6).
second stage of labour; number of vaginal examinations in labour;
Admission to special care nursery stillbirth or neonatal death or neonatal morbidity, excluding fatal
malformations; admission to special care nursery; intrapartum fe-
There were fewer admissions to special care nursery in the parto- tal death; cord blood arterial pH less than 7.1; birth trauma; child-
graph without latent phase condition (RR 1.84, 95% CI 1.29 to 2.63; hood disability; usability.
There were no clear differences between groups for any of the sec-
Usability: user-friendliness score
ondary outcomes reported.
In the modified partograph group there was a higher user-friendli-
ness score (mean difference (MD) -7.89, 95% CI -8.14 to -7.64; n = • Instrumental vaginal birth (RR 0.75, 95% CI 0.37 to 1.56; n = 99;
743; 1 trial; Analysis 6.8); 93% of staff felt the partograph with latent 1 trial; Analysis 7.6).
phase (composite) was more difficult to use. The partographs were • Postpartum haemorrhage - blood loss > 500 mL (RR 1.57, 95% CI
scored on ease of teaching, usefulness, interpretation and overall 0.55 to 4.46; n = 99; 1 trial; Analysis 7.7).
rating. • Regional analgesia (RR 0.86, 95% CI 0.56 to 1.32; n = 99; 1 trial;
Analysis 7.8).
7. Partograph with two-hour action line versus partograph
• Opioid use (RR 0.98, 95% CI 0.45 to 2.14; n = 99; 1 trial; Analysis
with stepped dystocia line
7.9).
One pilot study involving 99 women took place in Australia (Lee • Need for intubation at birth. Described as "active resuscita-
2015), and compared a traditional partograph with a two-hour action" (RR 0.44, 95% CI 0.14 to 1.32; n = 99; 1 trial; Analysis 7.10).
tion line, with a dystocia line. The dystocia line partograph starts
when the woman's cervix is a minimum of 4 cm dilated. The stepped 8. Partograph versus labour scale
line steepens at 6 cm to account for the "progressive acceleration"
Only one small study compared a labour scale versus a tradition-
of labour. If the action line or vertical step line were crossed, a vagi-
al WHO partograph (Shazly 2017). One-hundred and twenty-two
nal examination was performed in two hours to assess whether the
women were randomised, though six women in each group with-
line had been crossed, and oxytocin augmentation was then dis-
drew prior to analysis. We have used the number randomised for all
cussed.
outcomes except for Apgar score less than 7 at five minutes, neona-
Primary outcomes tal death, and duration of first stage of labour.
Caesarean section Primary outcomes

There was no clear difference between the groups in rates of cae- Caesarean section
sarean section (RR 1.10, 95% CI 0.46 to 2.62; n = 99; 1 trial; Analysis
There were no clear differences between the groups for overall cae-
7.1).
sarean section (RR 0.42, 95% CI 0.16 to 1.11; n = 122; 1 trial; Analysis
Oxytocin augmentation 8.1).
Fewer women in the dystocia line group received oxytocin augmen- Oxytocin augmentation
tation than those in the two-hour action line group (RR 0.62, 95% CI
Fewer women in the labour scale group received oxytocin augmen-
0.39 to 0.98; n = 99; 1 trial; Analysis 7.2).
tation (RR 0.32, 95% CI 0.18 to 0.54; n = 122; 1 trial; Analysis 8.2).
Informed decisions.

Duration of first stage of labour tions (e.g. seizures, birth asphyxia, neonatal encephalopathy) were
The Shazly 2017 trial reported duration of active stage of labour. not reported under this comparison.
There was no clear difference between the labour scale and parto- Partograph with different placement of action lines
graph groups (MD 0.44 hours, 95% CI -0.40 to 1.28; n = 110; 1 trial; (Comparisons 2 - 4: 4 trials, 5051 women)
Analysis 8.3).
When compared to a four-hour action line, women in the two-hour
Maternal experience of childbirth action line group were more likely to receive oxytocin augmenta-
Not reported in this comparison. tion. There was no clear difference in caesarean section rates; du-
ration of first stage of labour; maternal experience of childbirth; or
Low Apgar score (less than 7 at 5 minutes) Apgar score less than 7 at five minutes between the two- and four-
hour action line.
No babies in either group had Apgar scores of less than 7 at five min-
utes (110 babies; Analysis 8.4). The following comparisons only include data from single studies.
Fewer women reported negative childbirth experiences in the two-
Secondary outcomes
hour action line group than the three-hour action line group but
In addition to the general list of outcomes not reported in any of there were no clear differences between the groups in caesarean
the trials (see under heading ‘Effects of Interventions’), no trials section, oxytocin augmentation, or low Apgar score. Duration of
reported the following secondary outcomes for this comparison: first stage of labour was not reported.
serious maternal morbidity or death; caesarean section for fetal
distress; instrumental vaginal birth; postpartum haemorrhage; re- When we compared the three- and four-hour action line groups,
gional analgesia; opioid use; performance of artificial rupture of the the caesarean section rate was higher in the three-hour action line
membranes during labour; antibiotic use; duration of second stage group. We did not observe any clear differences in oxytocin aug-
of labour; number of vaginal examinations in labour; admission to mentation, maternal experience of childbirth, or low Apgar score.
special care nursery; need for intubation at birth; intrapartum fetal Duration of first stage of labour was not reported.
death; cord blood arterial pH less than 7.1; birth trauma; childhood
Partograph with alert line only versus partograph with alert
disability; usability.
and action line (Comparison 5: one trial, 694 women)
Caesarean section for delay in labour The caesarean section rate was lower in the alert line only group.
Fewer women in the labour scale group had a caesarean section for There were no clear differences between groups for oxytocin aug-
delay in labour (RR 0.20, 95% CI 0.05 to 0.88; n = 122; 1 trial; Analysis mentation, low Apgar score, instrumental vaginal birth, or perina-
8.5). tal death.
Stillbirth or neonatal death or neonatal morbidity Duration of first stage of labour, and maternal experience of child-
birth were not reported in this trial.
The trial reported that there were no neonatal complications (110
babies; Analysis 8.6). Partograph with latent phase versus partograph without
latent phase (Comparison 6: one trial, 743 women)
Birth injuries and postpartum haemorrhage (PPH) (non-prespecified
outcome) When we compared a partograph with a latent phase (composite)
There was no clear difference between the groups for birth injuries and one without (modified), the caesarean section and oxytocin
and PPH (RR 3.00, 95% CI 0.32 to 28.04; n = 122; 1 trial; Analysis 8.7). augmentation rates were higher in the partograph with a latent
phase. There were no clear differences between groups for oxytocin
DISCUSSION augmentation, or Apgar score less than 7 at five minutes.
Summary of main results Duration of the first stage of labour, and maternal experience of
childbirth were not reported.
A total of 9475 women were recruited from 11 trials comparing par-
tograph use; three trials comparing partograph use with no par- Partograph with two-hour action line versus partograph with
tograph, seven trials comparing different partograph designs, and stepped dystocia line (Comparison 7: 1 trial, 99 women)
one trial comparing a labour scale versus partograph.
When we compared a two-hour action line and a stepped dystocia
Partograph use versus no partograph (Comparison 1: 3 trials, line, fewer women received oxytocin augmentation in the dystocia
1703 women) line group. We did not observe any clear differences in caesarean
section rates, duration of the first stage of labour, maternal experi-
There was no evidence of any clear difference between partograph ence of childbirth, or low Apgar scores.
use and no partograph in caesarean section rates (very low-quali-
ty evidence); oxytocin augmentation (moderate-quality evidence); Partograph use versus labour scale (Comparison 8: 1 trial, 122
duration of first stage of labour (low-quality evidence); or Apgar women)
score less than 7 at five minutes (very low-quality evidence).
The labour scale compared with the partograph resulted in fewer
Serious maternal morbidity or death (e.g. ruptured uterus, admis- caesarean sections for delay in labour, and fewer women receiving
sion to intensive care unit, septicaemia, organ failure) and stillbirth oxytocin augmentation, but did not produce any clear difference
or neonatal death or neonatal morbidity, excluding fatal malforma- for overall caesarean section rate, duration of first stage of labour,
or low Apgar score.
Informed decisions.

Overall completeness and applicability of evidence Only two of the trials included multiparous women (Kenchaveeriah
2011; Walss Rodriguez 1987), one trial only included women with
Based on the limited evidence from the 11 trials included in this re- high-risk pregnancies (Rani 2015), all other trials included women
view, there remains uncertainty regarding the effectiveness of par- with uncomplicated pregnancies in spontaneous labour. However,
tograph use. in reality, the partograph is used for a wide spectrum of women in
the intrapartum period. Consideration needs to be given to the ap-
Only three trials compared partograph use with no partograph use.
plicability of these review findings to pregnant women who fall out-
These trials were conducted in different settings; one in a high-
side the inclusion criteria of the included trials. Further research on
resourced setting (Windrim 2006), and the other two in a low-re-
different populations would be preferable.
sourced setting (Rani 2015: Walss Rodriguez 1987). In two studies
(Walss Rodriguez 1987; Windrim 2006), the partograph was the ex- Important clinical outcomes were absent from the included trials,
perimental arm. It is not clear in the Rani 2015 study whether this particularly in low-resourced settings, e.g. length of first and sec-
was the experimental arm or not. These findings cannot be extrap- ond stage of labour. None of the trials examined the impact of the
olated to units where the partograph is currently in use; removing partograph on resource utilisation; a factor particularly important
the partograph as opposed to introducing it may produce different in low-resourced settings. Only the Lavender studies reported mea-
findings. sures of maternal childbirth experience (Lavender 1998a; Lavender
2006). None of the studies reported quality of care as an outcome.
Seven trials compared the use of different designs of partograph for
women in spontaneous labour. Combined evidence from four tri-
Quality of the evidence
als comparing the different placement of action lines showed little
difference in caesarean section rates and few differences in other Evidence from this review is inconclusive. Evidence from trials com-
maternal outcomes (Lavender 1998a; Lavender 2006; Orhue 2013; paring partograph use with no partograph was limited to only three
Sinha 2017). When we compared the two-hour action line with the trials with 1813 women of varying risk of bias (Rani 2015; Walss Ro-
four-hour action line, the only difference we found was an increase driguez 1987; Windrim 2006). We assessed prespecified outcomes
in oxytocin augmentation in the two-hour arm. This is unsurpris- using the GRADE approach for this comparison; we graded evi-
ing given that the associated guidelines advocated earlier use of dence for caesarean section, and low Apgar score (less than 7 at
oxytocin. When we compared the two-hour action line and three- five minutes) very low-quality, evidence for oxytocin augmentation
hour action line, we found differences in the self-reported maternal was moderate-quality, and duration of first stage of labour was low-
experience, with less women in the two-hour arm reporting a neg- quality. Maternal experience of childbirth, serious maternal and
ative experience. The relevance of these findings is uncertain, es- neonatal morbidity or death were not reported under this compari-
pecially as the comparison between the two-hour versus four-hour son. We downgraded evidence for limitations in study designs, high
arm and three-hour versus four-hour arm revealed no differences. heterogeneity, indirectness, and imprecision in effect estimates.
It may be that women in the two-hour arm perceived their labours
to be shorter, as the three-hour action line was current local policy. Overall, risk of bias varied in the included trials and no trial blind-
Alternatively, it may be that because those women whose labours ed women or staff. Two studies did not adequately conceal alloca-
were managed with the two-hour action line received more inter- tion. Loss to follow-up was low in all studies. The strongest study,
vention, they also received more labour support. There were no dif- in terms of quality, was that conducted by the Windrim 2006 study,
ferences in any neonatal outcomes. Although the findings of these which showed no differences in any clinical outcomes measured
studies were fairly consistent, both studies were from the same set- (caesarean section rate, duration of labour, oxytocin augmenta-
ting, and therefore their generalisability needs consideration. tion, amniotomy, epidural use, use of antibiotics in labour, Apgar
scores, or admissions to neonatal intensive care unit) following in-
We did not combine the South African trial with the previous trials troduction of the partograph. However, as acknowledged by the
(Pattinson 2003), as this was a trial which compared a partograph study authors, the findings may have been influenced by the rela-
with an alert line and aggressive management versus one with an tively high percentage of non-compliance in completing the parto-
alert and action line, with more conservative management. This tri- graph (20%) or the cross-contamination of care by staff, or both.
al described a package of care for labour management alongside
the partograph use, which advocated more frequent vaginal ex- Potential biases in the review process
aminations (two-hourly) for women in the aggressive management
We accept the possibility of introducing bias in the review process,
group, thereby suggesting a more complex intervention. This study
however we have taken steps to minimise this. Two review authors
compared different partograph designs that clearly demonstrated
independently assessed for inclusion all the potential studies iden-
a difference in caesarean section rates; the more aggressive arm
tified as a result of the search strategy. We resolved any disagree-
having the lower rate. Given that the partograph is a complex inter-
ment through discussion or, if required, we consulted the third re-
vention, used in conjunction with labour guidelines, the approach
view author. Two review authors independently also assessed risk
used in this study may be more appropriate. Utilising a reductionist
of bias for each study. Any disagreement was resolved by discussion
approach, to what is in essence a complex intervention, may pro-
or by involving a third assessor.
duce less meaningful findings.
Tina Lavender was investigator of two trials included in this review
Kenchaveeriah 2011 compared two partographs - a composite par-
(Lavender 1998a; Lavender 2006); therefore, two other reviewers
tograph including the latent phase with a modified one without the
(Smyth and Cuthbert) evaluated these studies.
latent phase. The trial confirmed that a partograph without a latent
phase was associated with a lower rate of caesarean section, indi-
cating labour can be managed without a latent phase being plotted
on the partograph.
Informed decisions.

Agreements and disagreements with other studies or ment in which it will be used. Standard care should be clearly de-
reviews scribed to allow for judgements on transferability of findings, to be
made. A cluster-randomised trial would be the most appropriate
We did not find any reviews that directly addressed clinical out- design, as this would enable consideration of key organisational is-
comes related to partograph use. However, two recent reviews ex- sues (e.g. training of individuals, hospital practices and clinical pro-
ploring barriers and facilitators to partograph use may assist any tocols). Using a cluster design would also reduce any contamina-
ongoing work (Bedwell 2017; Ollerhead 2014). Both reviews action between facility-based health professionals who would be sup-
knowledge the challenges of using current partographs but also porting many women in labour at the same time; this was an issue
highlight the multifaceted factors which may be contributing to its raised in the included trials.
success or failure as a labour monitoring tool.
Given the limitations of existing trials, future studies could consid-
A review of vaginal examinations has some relevance to this cur- er the inclusion of both primigravid and multigravid women, as, in
rent review (Downe 2013), particularly related to the cervicograph most units which use the partograph, the same chart is used irre-
part of the partograph, however the evidence currently presented spective of parity. Any future trials should stratify participants ac-
is limited to two small studies. cording to parity, services with low (20 or less per 1000) and high
perinatal mortality (more than 20) and low versus high intervention
There is a plethora of local, national and international guidelines,
rates in the first stage of labour. This clarity would also allow for
policies and training manuals which advocate partograph use;
more accurate comparability, both clinically, and also between tri-
most of these would recommend the World Health Organization
als for the purposes of systematic review by meta-analysis, allow-
(WHO) modified partograph. Until further evidence is published,
ing for more robust conclusions. The trials in this review included
these remain relevant. Emerging evidence which challenges be-
women at urban hospitals only. Whether the partograph is benefi-
liefs about normal patterns of labour progress are likely to result in
cial for women across all facilities is unclear and needs further in-
changes to partograph designs and subsequent recommendations
vestigation.
(Oladapo 2017).
It is essential to involve consumers in all stages of future trials, and
AUTHORS' CONCLUSIONS most significantly during the planning stages, in order to identify
those outcomes which are deemed most relevant. Important out-
Implications for practice
comes are absent from the existing trials and should be considered
There is no clear evidence that partographs improve outcomes and in future protocols. Important clinical outcomes, for example, re-
no clear evidence that one type of partograph is greatly superior to late to recognition of prolonged labour and include the length of
another. However, we acknowledge that many units, in high- and the first and second stage of labour. Moreover, maternal experience
low-income settings, currently use a partograph and have report- is of crucial importance and should be investigated using recog-
ed quality of care benefits in terms of ease of recording, provision nised validated tools in order to allow women to make informed
of pictorial overview of progress, auditing of care, training of clin- choices about their care. There was no information in any of the in-
icians and transferring of care (Bedwell 2017). Furthermore, there cluded trials regarding long-term outcomes for women and babies.
has been evidence from non-randomised trials of potential benefits We propose that future trialists should consider instituting some
of partograph use (Bedwell 2017; Bosse 2002; Fawole 2008). Given form of long-term follow-up which is feasible and appropriate for
the fact that the partograph is currently in widespread use, it ap- the study population in question. Any future trials should be of ad-
pears reasonable, until stronger evidence is available, for decisions equate size and data on economic outcomes should be obtained,
regarding whether or not to use a partograph and which one to use, to allow for allocation of resources and service planning.
to be locally determined following consultation with women and
clinicians. As detailed earlier, in one study (Windrim 2006), there was a rela-
tively high percentage of non-compliance in completing the parto-
Implications for research graph; we were unable to draw any conclusions about why this may
have happened. In future trials, consideration should be given to
Whether or not there is a need for a trial of partograph use versus the inclusion of a nested qualitative study, which would enable the
no partograph is open for debate. One could argue that a trial of capture of comprehensive information on the barriers and facilita-
partograph use versus no partograph could be conducted in set- tors to partograph use.
tings where a partograph is not used or by removing the partograph
in some settings and comparing it to those where it remains. How- ACKNOWLEDGEMENTS
ever, a consensus of international experts proposed that this is no
longer an important question (Fistula Care 2012), suggesting that Eckhart Buchmann and Cheryl Nikodem for an earlier draft of the
a more important question is which partograph should be used, as protocol. Patrick O'Brien for the existing protocol which guided this
current evidence, from the included trials, fails to provide robust review.
guidance. Trials comparing different partograph designs, are there-
fore needed. Therese Dowswell for help with assessment of risk of bias, data ex-
traction, and technical assistance.
Although the partograph is a low cost, low-invasive intervention, it
forms part of the overall management of labour care, making it part Sherif Shazly for kindly providing further information regarding the
of a complex intervention. This complexity may, in part, account for SLiP trial (Shazly 2017).
the heterogeneity within the results, particularly related to the dif-
Anna Hart for her contributions to earlier versions of the review.
ferences between high- and low-resourced settings. As such, any
future trials should be designed to consider the clinical environ-
Informed decisions.

As part of the pre-publication editorial process, this review has This project was supported by the National Institute for Health
been commented on by three peers (an editor and two referees who Research (NIHR), via Cochrane Infrastructure funding to Cochrane
are external to the editorial team), a member of Cochrane Preg- Pregnancy and Childbirth. The views and opinions expressed there-
nancy and Childbirth's international panel of consumers and the in are those of the authors and do not necessarily reflect those of
Group's Statistical Adviser. the Systematic Reviews Programme, NIHR, National Health Service
(NHS) or the Department of Health.
Informed decisions.

REFERENCES

References to studies included in this review Pattinson 2003 {published data only}
Kenchaveeriah 2011 {published data only} Pattinson RC, Howarth GR, Mdluli W, Macdonald AP, Makin JD,
Funk M. Aggressive or expectant management of labour: a
Kenchaveeriah SM, Patil KP, Singh TG. Comparison of two WHO
randomised trial. BJOG: an international journal of obstetrics
partographs: a one year randomized controlled trial [Iki DSO
and gynaecology 2003;110:457-61.
partografinin karsilastirilmasi: Bir yillik randomize kontrollu
calisma]. Journal of the Turkish German Gynecology Association Rani 2015 {published data only}
Artemis 2011;12(1):31-4.
Rani J, Sharma D, Sehgal A. Role of partogram in high risk
Lavender 1998a {published and unpublished data} pregnancies: an experience at a tertiary centre. Archives of
Gynecology & Obstetrics 2015;291(1):73-8.
* Lavender T, Walkinshaw S, Alfirevic Z. Partogram action line
study: a randomised trial. British Journal of Obstetrics and Shazly 2017 {published data only}
Gynaecology 1998;105:976-80.
NCT02486822. The management of spontaneous labour in
Lavender T, Walkinshaw SA, Walton I. A prospective study primigravida (SLIP): labor scale versus WHO partograph [Labor
of women's views of factors contributing to a positive birth scale versus WHO partograph in the management of labor].
experience. Midwifery 1999;15:40-6. clinicaltrials.gov/ct2/show/record/NCT02486822 (first received
24 June 2015).
Lavender T, Wallymahmed AH, Walkinshaw SA. Managing labor
using partograms with different action lines: a prospective * Shazly SA, Tolba S, Abbas AM, Nassr AA. Management of
study of womens views. Birth 1999;26:89-96. spontaneous labor in primigravidae: labor scale versus who
partograph (SLiP trial). American Journal of Obstetrics and
Lavender 2006 {published and unpublished data} Gynecology 2017;216(1):S510-1, Abstract no: 894.
ISRCTN78346801. Partogram Action Line Study (PALS2).
Sinha 2017 {published data only}
isrctn.com/ISRCTN78346801 (first received 6 July 2005).
Sinha D, Shrivastava S, Shrivastava S. A comparative study
* Lavender T, Alfirevic Z, Walkinshaw S. Effect of different of 4-hour versus 2-hour action line on WHO modified
partogram action lines on birth outcomes. Obstetrics & partograph. International Journal of Research in Medical
Gynecology 2006;108:295-302. Sciences 2017;5(3):876-9.
Lee 2015 {published data only} Walss Rodriguez 1987 {published data only}
* ACTRN12614000786695. A pilot randomised controlled trial to Walss-Rodriguez RJ, Gudino-Ruiz F, Tapia-Rodriguez S.
determine the effect of a physiological (step) versus a standard Comparative study between Friedman's partogram and
action (slope) labour progress lines on the rate of spontaneous conventional descriptive partogram. Ginecologia y Obstetricia
vaginal birth amongst low risk women in labour for the first de Mexico 1987;55:318-22.
time [For women in labour for the first time does a partograph
with a graduated dystocia line compared to a standard sloping Windrim 2006 {published and unpublished data}
action line increase the likelihood of a spontaneous vaginal Windrim R, Seaward G, Hodnett E, Akoury H, Kingdom J,
birth]. anzctr.org.au/Trial/Registration/TrialReview.aspx? Salenieks ME, et al. A randomized controlled trial of a
ACTRN=12614000786695 (first received 7 July 2014). bedside partogram in the active management of primiparous
labour. Journal of Obstetrics and Gynaecology Canada: JOGC
ACTRN12615000445572. A pilot randomised controlled 2006;29(1):27-34.
trial to determine the effect of two designs of partographs,
including labour progress lines, on the rate of spontaneous
vaginal birth amongst low risk women in labour for the first References to studies excluded from this review
time: The Partograph Trial [For women in labour for the
Ajoodani 2011 {published data only}
first time does a partograph with a graduated dystocia line
compared to a standard sloping action line increase the * Ajoodani Z, Heidari P, Nasrollahi S. Assessment of the results
likelihood of a spontaneous vaginal birth: Pilot Partograph of using partogram in labor management. Iranian Journal of
study 2]. anzctr.org.au/Trial/Registration/TrialReview.aspx? Obstetrics, Gynecology and Infertility 2011;14(2):20-7.
ACTRN=12615000445572 (first received 13 April 2015).
IRCT2012122811911N1. Labor management based on
Orhue 2013 {published data only} partogram. en.search.irct.ir/view/11959 (first received 24
September 2003).
Orhue A, Iribhogbe I. Outcome of a protocol for the use of the
partograph as a decision-making algorithm for the prevention Cartmill 1992 {published data only}
of prolonged labor. 1st FIGO African Regional Conference of
Gynecology and Obstetrics; 2013 Oct 2-5; Addis Ababa, Ethiopia. Cartmill R, Thornton J. Effect of presentation of partogram
2013. information on obstetric decision-making. Lancet
1992;339:1520-2.
Informed decisions.

Fahdhy 2005 {published data only} percentile normogram of the consortium on safe labor to
Fahdhy M, Chongsuvivatwong V. Evaluation of World Health evaluate early versus delayed cesarean section. trialregister.nl/
Organization partograph implementation by midwives trialreg/admin/rctview.asp?TC=5543 (first received 4 October
for maternity home birth in Medan, Indonesia. Midwifery 2015).
2005;21:301-10.

Hamilton 2001 {published data only} Additional references
Hamilton E, Platt RW, Gauthier RJ, McNamara H, Miner L, Agarwal 2013
Rothenberg S, et al. A multicenter trial of individualized labor Agarwal K, Agarwal L, Agarwal VK, Agarwal A, Sharma M.
curves and cesarean rates. American Journal of Obstetrics and Evaluation of Paperless Partogram as a Bedside Tool in the
Gynecology Vol. 185, issue 6 Suppl:S103. Management of Labor.. Journal of Family Medical and Primary
Care 2013;2:47-49.
Hamilton 2004 {published data only}
Hamilton E, Platt R, Gauthier R, McNamara H, Miner L, Bedwell 2017
Rothenberg S, et al. The effect of computer-assisted evaluation Bedwell C, Levin K, Pett C, Lavender T. A realist review of the
of labor on cesarean rates. Journal for Healthcare Quality partograph: when and how does it work for labour monitoring?.
2004;26(1):37-44. BMC Pregnancy and Childbirth 2017;17:31. [DOI: 10.1186/
s12884-016-1213-4]
Kogovsek 2000 {published data only}
Kogovsek U, Assejev V, Novak-Antolic Z. Partogram presentation Bosse 2002
and the outcome of labour and delivery. Effectiveness of Bosse G, Massawe S, Jahn A. The partograph in daily practice:
prenatal care in Slovenia; 2000 June 1-3; Slovenia. 2000:125-32. It's quality that matters. International Journal of Gynecology and
Obstetrics 2002;77(3):243-4.
Mathews 2007 {published data only}
Mathews JE, Rajaratnam A, George A, Mathai M. Comparison Downe 2013
of two World Health Organization partographs. International Downe S, Gyte GM, Dahlen HG, Singata M. Routine vaginal
Journal of Gynecology & Obstetrics 2007;96(2):147-50. examinations for assessing progress of labour to improve
outcomes for women and babies at term. Cochrane
WHO 1994 {published data only}
Database of Systematic Reviews 2013, Issue 7. [DOI:
Anonymous. World Health Organization partograph 10.1002/14651858.CD010088.pub2]
in management of labour. World Health Organization
Maternal Health and Safe Motherhood Programme. Lancet Fatusi 2007
1994;343(8910):1399-404. Fatusi AO, Makinde ON, Adetemi AB. Evaluation of health
workers' training in use of the partogram. International Journal

of Gynecology and Obstetrics 2007;100(1):41-4.
References to studies awaiting assessment
NCT02911272 {published data only} Fawole 2008
NCT02911272. Best partograph action line to prevent prolonged Fawole AO, Hunyinbo KI, Adekanle DA. Knowledge and
labour (partograph). clinicaltrials.gov/show/NCT02911272 (first utilization of the partograph among obstetric care givers in
received 14 September 2016). South West Nigeria. African Journal of Reproductive Health
2008;12(1):22-9.

References to ongoing studies Fistula Care 2012
Fistula Care and Maternal Health Task Force. Revitalizing
NCT02714270 {published data only}
the partograph: does the evidence support a global call
NCT02714270. A randomized clinical trial of paperless to action?. Report of an Expert Meeting; 2011 November
versus modified World Health Organization partograph in 15-16; New York (NY) (accessed 12 June 2012). New York
management of first stage of labour [Paperless partograph for (NY) www.fistulacare.org/pages/pdf/program-reports/
management of low risk labour]. clinicaltrials.gov/ct2/show/ EngenderHealth-Fistula-Care-Partograph-Meeting-Report-9-
NCT02714270 (first received 16 March 2016). April-12.pdf: EngenderHealth/Fistula Care, 2012.
NCT02741141 {published data only} Friedman 1954
NCT02741141. A comparative study of the effect of two Friedman E. Graphic analysis of labour. American Journal of
partographs on the cesarean section rate in women in Obstetrics and Gynecology 1954;68:1568-75.
spontaneous labour (PARTODYS). clinicaltrials.gov/show/
NCT02741141 (first received 12 January 2016). Groeschel 2001
Groeschel N, Glover P. The partograph. Used daily but rarely
NTR5543 {published data only}
questioned. Australian Journal of Midwifery 2001;14:22-7.
NTR5543. Randomized controlled trial comparing the currently
used Friedman partogram with a four hour action line to
the newly developed simple partogram, based on the 95th
Informed decisions.

Higgins 2011 Philpott 1972a

Higgins JP, Green S, editor(s). Cochrane Handbook for Philpott RH. Graphic records in labour. British Medical Journal
Systematic Reviews of Interventions Version 5.1.0 (updated 1972;4:163.
March 2011). The Cochrane Collaboration, 2011. Available from
www.handbook.cochrane.org. Philpott 1972b
Philpott RH, Castle WM. Cervicographs in the management of
Lavender 1998b labour in primigravidae. Journal of Obstetrics and Gynaecology
Lavender T, Alfirevic Z, Walkinshaw S. Partogram action line of the British Commonwealth 1972;79:592-8.
study: a randomised trial. British Journal of Obstetrics and
Gynaecology 1998;105(9):976-80. Philpott 1972c
Philpott RH, Castle WM. Cervicographs in the management of
Lavender 1999 labour in primigravidae. Journal of Obstetrics and Gynaecology
Lavender T, Malcolmson L. Is the partogram a help or a of the British Commonwealth 1972;79:599-602.
hindrance?. Practising Midwife 1999;2:23-7.
Rakotonirina 2013
Lavender 2011 Rakotonirina JE, Randrianantenainjatovo CH, Elyan Edwige BB,
Lavender T, Omoni G, Lee K, Wakasiaka S, Watiti J, Mathai M. Dorasse R, Rakotomanga J, Rakotovao JH. Assessment of the
Student nurses experiences of using the partograph in labour use of partographs in the region of Analamanga. International
wards in Kenya: A qualitative study. African Journal of Midwifery Journal of Reproduction, Contraception, Obstetrics and
and Women’s Health 2011;5(3):117-22. Gynecology 2013;2:257-62.
Lennox 1995 Review Manager 2014 [Computer program]

Lennox CE, Kwast BE. The partograph in community obstetrics. Nordic Cochrane Centre, The Cochrane Collaboration. Review
Tropical Doctor 1995;25:56-63. Manager 5 (RevMan 5). Version 5.3. Copenhagen: Nordic
Cochrane Centre, The Cochrane Collaboration, 2014.
Neilson 2003
Neilson J, Lavender T, Quenby S, Wray S. Obstructed labour. Rotich 2011
British Medical Bulletin 2003;67:191-204. Rotich E, Maina L, Njihia A, Christensson K. Evaluating
partograph use at two main referral hospitals in Kenya. African
Nystedt 2014 Journal of Midwifery and Women's Health 2011;5:21-4.
Nystedt A, Hildingsson I. Diverse definitions of prolonged
labour and its consequences with sometimes subsequent Shazly 2014
inappropriate treatment. BMC Pregnancy and Chldbirth Shazly SAM, Embaby LHO, Ali SS. The labour scale – assessment
2014;14:233. of the validity of a novel labour chart: A pilot study. Australian
and New Zealand Journal of Obstetrics and Gynaecology
Oladapo 2017 2014;54:322-6.
Oladapo OT, Diaz V, Bonet M, Abalos E, Thwin SS, Souza H,
et al. Cervical dilatation patterns of 'low-risk' women with Tay 1996
spontaneous labour and normal perinatal outcomes: a Tay SK, Yong TT. Visual effect of partogram designs on the
systematic review. BJOG: an international journal of obstetrics management and outcome of labour. Australian and New
and gynaecology 2017 Sep 11 [Epub ahead of print]. [DOI: Zealand Journal of Obstetrics and Gynaecology 1996;36:395-400.
10.1111/1471-0528.14930.]
Walraven 1994
Ollerhead 2014 Walraven GE. WHO partograph. Lancet 1994;344:617.
Ollerhead E, Osrin D. Barriers to and incentives for achieving
partograph use in obstetric practice in low- and middle-income Walsh 1994
countries: a systematic review. BMC Pregnancy and Childbirth Walsh D. Management of progress in the first stage of labour.
2014;14:281. Midwives Chronicle and Nursing Notes 1994;3:84-8.
Orhue 2012 WHO 2003

Orhue AAE, Aziken ME, Osemwenkha AP. Partograph as a tool for World Health Organization. Pregnancy, Childbirth, Postpartum
team work management of spontaneous labor. Nigerian Journal and Newborn care: A Guide for Essential Practice. Geneva:
of Clinical Practice 2012;15:1-8. World Health Organization, 2003.
Pawson 2004 Yisma 2013

Pawson R, Greenhalgh T, Harvey G, Walshe K. Realist synthesis: Yisma E, Dessalegn B, Astatkie A, Fesseha N. Knowledge and
an introduction. ESRC Research Methods Programme - utilisation of partograph among obstetric care givers in public
University of Manchester 2004. health institutions of Addis Ababa, Ethiopia. BMC Pregnancy
Childbirth 2013;13:1-9.
Informed decisions.

References to other published versions of this review Lavender 2012

Lavender 2005 Lavender T, Hart A, Smyth RMD. Effect of partogram use
on outcomes for women in spontaneous labour at term.
Lavender T, O'Brien P, Hart A. Effect of partogram use
Cochrane Database of Systematic Reviews 2012, Issue 8. [DOI:
10.1002/14651858.CD005461.pub3]
10.1002/14651858.CD005461] Lavender 2013
Lavender 2008 Lavender T, Hart A, Smyth RMD. Effect of partogram use
Lavender T, Hart A, Smyth RMD. Effect of partogram use
10.1002/14651858.CD005461.pub4]
10.1002/14651858.CD005461.pub2]
* Indicates the major publication for the study

CHARACTERISTICS OF STUDIES
Characteristics of included studies [ordered by study ID]

Kenchaveeriah 2011
Methods 1-year randomised controlled trial. Conducted between November 2008 and October 2009.
Participants 743 women with uncomplicated pregnancy in spontaneous labour with term, singleton, vertex gesta-
tion.
Interventions Composite partograph including the latent phase versus a modified partograph without the latent
phase.
Outcomes • Rate of caesarean section

• Augmentation of labour
• Labour crossing the alert and action line
• Perinatal outcome
• User-friendliness
• Maternal complications
Notes Declarations of interest: none declared
Funding source: not reported
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence genera- Low risk Computer-generated random number table.

tion (selection bias)
Allocation concealment Unclear risk No information on how women were allocated to groups following random se-
(selection bias) quence generation.
Blinding of participants High risk Not feasible to blind women or clinical staff.
and personnel (perfor-
mance bias)
All outcomes
Blinding of outcome as- Unclear risk Blinding not reported.

sessment (detection bias)
All outcomes
Informed decisions.

Kenchaveeriah 2011 (Continued)
Incomplete outcome data Low risk No apparent loss to follow-up.
(attrition bias)
All outcomes
Selective reporting (re- Low risk All outcomes prespecified in methods reported.
porting bias)
Other bias Low risk Additional analysis performed, not prespecified in methods section of publica-
tion regarding indication for caesarean section.

Lavender 1998a
Methods Prospective randomised clinical trial. Random allocation by sealed, opaque envelopes. Conducted be-
tween January 1996 and August 1997.
Participants 928 primigravid women from the North-West of England, with uncomplicated pregnancies who pre-
sented in spontaneous labour at term.
Interventions Women were randomised to have their progress of labour recorded on a partogram with an action line
2, 3 or 4 hours to the right of the alert line.
Outcomes • Caesarean section rate

• Maternal satisfaction
• Instrumental delivery rate
• Need for augmentation
• Randomisation to delivery interval
• Use of epidural
• Cord blood gas analysis
• Blood loss > 500 mL
• Number of vaginal examinations
• Apgar score
• Admission to special care baby unit
Notes Maternal satisfaction was only assessed in a subset of women, i.e. all women recruited over a prespeci-
fied 12 month period (n = 615).
Declarations of interest: not reported
Funding source: study took place at the Liverpool Women's Hospital
Risk of bias
Random sequence genera- Low risk Table of random numbers.

Allocation concealment Low risk Consecutively numbered, sealed, opaque envelopes.

(selection bias)
mance bias)
All outcomes
Informed decisions.

Lavender 1998a (Continued)
Blinding of outcome as- Low risk Statistician blind to treatment allocation.
All outcomes
Incomplete outcome data Low risk Small loss to follow-up after randomisation (less than 1% attrition) for out-
(attrition bias) comes measured in labour. Higher attrition rates for the maternal satisfaction
All outcomes outcomes measured in the postnatal period.
porting bias)
Other bias Unclear risk 10% (who were otherwise eligible) were not approached (overall, 57% of eligi-
ble women were randomised).

Lavender 2006
Methods Prospective randomised clinical trial. Random allocation by sealed, opaque envelopes. Conducted be-
tween August 1998 and March 2005.
Participants 2975 primigravid women from the North-West of England, with uncomplicated pregnancies, in sponta-
neous labour at term.
Interventions Women were randomised to have their progress of labour recorded on a partogram with an action line
2 or 4 hours to the right of the alert line.
Outcomes Outcomes were stratified according to intended place of birth (midwife-led unit or obstetric unit).
• Caesarean section rate

• Maternal satisfaction
• Instrumental delivery rate
• Randomisation to delivery interval
• Use of epidural
• Cord blood gas analysis
• Blood loss > 500 mL
• Apgar score
• Admission to special care baby unit
Notes Declarations of interest: not reported
Funding source: Liverpool Women's Hospital
Risk of bias
Random sequence genera- Low risk Table of random numbers. Randomisation stratified by intended place of birth
tion (selection bias) (2 participating units).
Allocation concealment Low risk Consecutively numbered, sealed, opaque envelopes.

(selection bias)
Informed decisions.

Lavender 2006 (Continued)
mance bias)
All outcomes
Blinding of outcome as- Low risk Statistician blind to treatment allocation.

All outcomes
Incomplete outcome data Low risk Less than 1% attrition after randomisation.
(attrition bias)
All outcomes
porting bias)
Other bias Unclear risk Large numbers of women who were otherwise eligible were not approached to
participate. The numbers not approached varied depending on the recruiting
unit, 26% not approached in the midwifery and 61% in the delivery unit.

Lee 2015
Methods Pilot parallel (1:1), randomised, single-blinded study. Conducted January until May 2015 at a metropol-
itan hospital in Brisbane, Australia.
Participants 99 nulliparous women with spontaneous labour onset at term gestation (37 to 41 + 6 weeks) with a sin-
gle cephalic presentation.
Exclusion criteria
• History of 3 or more consecutive miscarriages

• Previous fetal death in utero
• Previous mid-trimester loss/cervical incompetence/cone biopsy/known uterine anomaly
• Rhesus isoimmunisation
• Complications during the current pregnancy (such as multiple pregnancy or fetal abnormality)
• Precluding medical conditions such as cardiac disease, essential hypertension, renal disease, pre-
existing diabetes, previous gestational diabetes, epilepsy, severe asthma, substance use, significant
psychiatric disorders, age 40 years, or a body mass index at beginning of pregnancy 35 (WHO Obesity
Class II or higher)
Interventions Women were randomised to have their progress of labour recorded on a partograph with a stepped
dystocia line partograph with steps beginning at 4 cm cervical dilatation or a 2-hour action line rising at
1 cm per hour.
Outcomes • Compliance
• Rate of artificial rupture of membranes
• Operative birth
• Postpartum haemorrhage
• Adverse effects
• Apgar score < 7 at 5 minutes
• NICU admission
• Need for neonatal resuscitation
• Neonatal systemic infection
Informed decisions.

Lee 2015 (Continued)
Notes Participants' labours were assessed as per the unit's standard care, i.e. vaginal exams every 4 hours, or
more frequently if deemed necessary. Amniotomies were not routinely performed among women with
epidurals. Providing the plot of the progress of labour stayed to the left of the Action or Dystocia line,
augmentation was not indicated. If the plot of progress reached the Action line, or the vertical 'step'
of the dystocia line, a vaginal assessment was performed in 2 hours to determine if the line had been
crossed; this triggered discussion regarding the need for oxytocic augmentation (following amniotomy
if membranes still intact). If individual maternal/fetal status warranted, augmentation could be initiat-
ed earlier according to clinician judgement.
Funding: this study was supported by a grant provided by the Mater Research Institute, Mater Health
Services, Brisbane, Australia
Declarations of interest: the authors report that there is no conflict of interest
Risk of bias
Random sequence genera- Unclear risk Block randomisation (block size 4) was prepared by an independent statisti-
tion (selection bias) cian.
Allocation concealment Low risk Opaque envelopes containing either instruction to use a standard care par-
(selection bias) tograph, upon which an Action line was drawn by the midwife or a preprint-
ed dystocia line partograph were prepared by an independent person in the
Mater Research Support Unit. Envelopes were stored in a locked cupboard on
the Birth Suite, with the key held by the midwifery team leader. Regardless of
allocation, each envelope contained the same number of pages to prevent dif-
ferentiation by weight.
Blinding of participants High risk Although the attending clinicians were aware of the allocation, the partic-
and personnel (perfor- ipants were blinded to the treatment group. Staff monitoring labour and
mance bias) prompting interventions were unblinded.
All outcomes
Blinding of outcome as- Low risk Researchers conducting the analysis were blinded to the treatment group,
sessment (detection bias) however the staff recording the outcomes were unblinded.
All outcomes
Incomplete outcome data Low risk Data for all women are reported. 46/50 and 44/49 maternal surveys complet-
(attrition bias) ed.
All outcomes
Selective reporting (re- Low risk All outcomes reported for protocol.
porting bias)
Other bias Low risk Baseline characteristics are similar.

Orhue 2013
Methods Prospective controlled study. Dates of study not reported.
Participants 948 term nulliparous women were allocated to a partograph with 2-hour and 4-hour action lines and
the same protocol was used for labour care.
Interventions Partograph with 2-hour action line versus partograph with a 4-hour action line.

Informed decisions.

Orhue 2013 (Continued)
• Primary postpartum haemorrhage
• Five minutes neonatal asphyxia
• Prolonged labour
Notes Awaiting response from authors for more information - contacted 21/04/2017.
By the protocol, admission was in active labour by the midwife at 4 cm dilatation and amniotomy per-
formed. Vaginal examinations were repeated 2 hourly on the 2-hour or 4 hourly on the 4-hour action
line partograph and plotted.
Declarations of interest: not reported
Funding source: not reported
Risk of bias
Random sequence genera- Unclear risk Not reported.

Allocation concealment Unclear risk It was stated in the protocol that allocation was random and masked but no
(selection bias) details given.
Blinding of participants High risk The protocol stated that the study was masked, but blinding not usually feasi-
and personnel (perfor- ble for this type of intervention.
mance bias)
All outcomes
Blinding of outcome as- High risk Blinding not feasible for this type of intervention and outcomes would be
sessment (detection bias) recorded by those providing clinical care who would be aware of the interven-
All outcomes tion.
Incomplete outcome data Unclear risk No information on dropouts or missing data in the brief study report.
(attrition bias)
All outcomes
Selective reporting (re- Unclear risk A protocol was available (trial registration) but there was too little information
porting bias) in the study report to assess whether the study was carried out according to
protocol.
Other bias Unclear risk Too little information to assess.

Pattinson 2003
Methods Prospective randomised clinical trial. Random allocation by sealed, opaque envelopes. Details of dates
of study not provided.
Participants 694 healthy nulliparous women from South Africa, who were in active spontaneous labour, at term,
with a healthy singleton pregnancy and cephalic presentation.
Interventions Women were randomised to either aggressive or expectant management protocols.
Aggressive management entailed: using a single line partograph, a vaginal examination every 2 hours
and use of oxytocin if the line was crossed.
Informed decisions.

Pattinson 2003 (Continued)
Expectant management entailed: using a 2-line partograph, with the alert line and a parallel action line
4 hours to the right, with a vaginal examination every 4 hours. If the action line was reached, oxytocin
was started.

• Operative deliveries
• Oxytocin use
• Received analgesia
• Apgar score
• Perinatal death
Notes Declarations of interest: not reported
Funding source: financial support came from the South African Medical Research Council
Risk of bias
Random sequence genera- Low risk Computer-generated list of random numbers.

Allocation concealment Unclear risk Sealed, opaque envelopes.

(selection bias)
mance bias)
All outcomes
Blinding of outcome as- Unclear risk Not reported.

All outcomes
Incomplete outcome data Low risk Low attrition after randomisation (less than 1%). Where women did not re-
(attrition bias) ceive the allocated intervention, there was intention-to-treat analyses.
All outcomes
porting bias)
Other bias High risk Recruitment stopped before required sample size reached due to funding con-
straints.

Rani 2015
Methods Prospective randomised controlled study.
Trial conducted at Patna Medical College and Hospital (tertiary hospital), Bihar, from October 2006 to
August 2008.
Participants 400 mothers with high-risk pregnancy, enrolled in active labour - 223 women with spontaneous onset.
Inclusion criteria
• Primigravidas with high-risk pregnancies like pregnancy-induced hypertension
Informed decisions.

Rani 2015 (Continued)
• Gestational diabetes mellitus
• Oligohydramnios
• Post-dates
• Cholestasis
• Hypothyroidism
• Patients with medical illness like seizure disorder, viral infections, at 37 weeks' gestation with spon-
taneous or induced labour.
• Patients with 1 previous LSCS willing for trial of labour after taking informed consent
Exclusion criteria
• Multiple pregnancy
• Malpresentation
• Heart disease
• Severe anaemia (Hb <6 gm/dl), antepartum eclampsia or haemorrhage
• Major degree of cephalopelvic disproportion
• Epidural analgesia
• elective LSCS or emergency LSCS in latent phase in high-risk pregnancy group
Interventions Partograph use versus no partograph.
Partograph group: progress of labour along with maternal vitals were recorded as standard notes on
case record file as well as on modified WHO partograph. n = 200 (only women with spontaneous onset
of labour are included in our data and analyses n = 110).
No partograph group: the course of labour and maternal condition were documented only as notes
in case record file. n = 200 (only women with spontaneous onset of labour are included in our data and
analyses n = 113).
Outcomes • Duration of first and second stage of labour

• Mode of delivery
• Neonatal Apgar scores
• Meconium staining of liquor
• Fetal heart rate abnormalities
• Incidence of NICU admission
• Maternal complications, i.e. prolonged labour, postpartum haemorrhage and puerperal sepsis
Notes Active phase of labour was diagnosed as regular contractions every 10 min or less, lasting more than 40
secs, with cervical effacement more than 80 % and cervical dilatation of 4 cm.
Vaginal examinations were performed 2-hourly or when indicated. Standard care included early am-
niotomy.
Funding: no funding reported
Declarations of interest: none declared
Risk of bias
Random sequence genera- Low risk It was reported that women were randomised to either group by comput-
tion (selection bias) er-generated random number. Cases were stratified into 2 groups as with
spontaneous onset of labour or following induced labour.
Informed decisions.

Rani 2015 (Continued)
Allocation concealment Unclear risk Not reported.
(selection bias)
Blinding of participants High risk Not mentioned, assumed not possible to blind.
mance bias)
All outcomes
Blinding of outcome as- High risk Not reported, assumed not due to nature of intervention. Outcomes likely to
sessment (detection bias) be reported by unblinded clinicians.
All outcomes
Incomplete outcome data Unclear risk Loss to follow-up and missing data were not mentioned; there was no study
(attrition bias) flow diagram.
All outcomes
Selective reporting (re- Low risk All outcomes prespecified in methods reported. Protocol not seen.
porting bias)
Other bias Low risk Baseline characteristics are similar between groups.

Shazly 2017
Methods A single centre, double-blinded, randomised trial conducted between July 2015 and June 2016 at Assi-
ut University Hosptal, Egypt.
Participants 122 nulliparous women randomised (data analysed for 110 women).
Inclusion criteria
• Nulliparous women
• Women who experienced spontaneous onset of labour
• Pregnant women 38 to 41 weeks with a singleton viable fetus
• Women with vertex presented
• Estimated fetal weights between 2500 g and 3800 g
Exclusion criteria
• Women with medical and obstetric comorbidities or fetal abnormalities

• Women with pre-labour rupture of membranes
Interventions Women either had labour managed by labour scale or by the traditional WHO partograph.
The labour scale was developed from the WHO partograph with the NICE Intrapartum Guidelines. It is
designed to help clinicians recognise determinants and management of potential labour dystocia. In-
stead of having a fixed 4-hour action line, the labour scale has particular trigger points throughout first
and second stage where management is reviewed. The labour scale aims to collect more detailed re-
porting of labour than the traditional partograph.
Outcomes • Successful vaginal delivery

• Proportion who delivered vaginal versus those indicated for caesarean section for labour dystocia
• Intrapartum maternal distress
• Maternal birth injuries
• Primary postpartum haemorrhage
• Maternal fever/postpartum infections
• Intrapartum fetal distress
Informed decisions.

Shazly 2017 (Continued)
• Birth injuries of the newborn
• Neonatal distress "asphyxia" (as reported 1 and 5 minutes Apgar score, resuscitation event, umbilical
artery pH, admission to NICU, length of stay and any further medical complications)
Notes Data from unpublished draft kindly provided by Sherif Shazly.
Funding: no funding sources
Risk of bias
Random sequence genera- Low risk A computer-generated randomisation list was created by one of the investiga-
tion (selection bias) tors (AMA).
Allocation concealment Low risk Opaque sequentially numbered envelopes were sealed and delivered to the
(selection bias) nurse in duty who assigned every new patient to the next envelope.
Blinding of participants High risk Blinding of women and clinicians not feasible
mance bias)
All outcomes
Blinding of outcome as- Unclear risk Blinded for the principal investigator who received the final data, a comput-
sessment (detection bias) er-generated randomisation list was created by one of the investigators (AMA).
All outcomes Assignment was single-blinded; double-blindness was not logistically applica-
ble. Randomisation number was used to collect and analyse data afterwards.
Incomplete outcome data Unclear risk 55/61 in each group were analysed: 7 women decided to withdraw from the
(attrition bias) study before management was initiated (3 from labour scale, 4 from WHO par-
All outcomes tograph) and 5 women decided to proceed directly to caesarean section (3 in
labour scale, 2 in WHO partograph). Eventually, 55 women remained in each
arm and data were collected for analysis. No further loss to follow-up.
Selective reporting (re- High risk Vaginal/assisted vaginal births not reported. Maternal complications were re-
porting bias) ported as a composite and most neonatal outcomes, such as admission to
special care unit were not mentioned.
Other bias Low risk Similar baseline characteristics between groups.

Sinha 2017
Methods Prospective hospital based randomised study. Dates of trial not given
Participants 200 women randomised
Inclusion criteria
• Primigravida
• Aged 19 to 29 at term (had crossed 37 weeks of gestation) with singleton live, term, cephalic presen-
tation
• Uncomplicated pregnancy in spontaneous labour
Exclusion criteria
Informed decisions.

Sinha 2017 (Continued)
• Multipara
• Short stature < 140 cm
• Teenage pregnancy
• Elderly primigravida
• Malpresentation
• Post-caesarean pregnancy
• Post-term pregnancy
• Preterm labour
• Oligohydramnios
• Intrauterine death
• Associated medical complications, i.e. diabetes, essential hypertension, heart disease, anaemia
• Associated obstetrical complications, i.e. antepartum haemorrhage, placenta previa, pre-eclamp-
sia/eclampsia
Interventions 100 womens' labour was monitored with 4-hour action line on WHO modified partograph (group A), the
other group (B) had the whole duration of labour monitored with 2-hour action line on the partograph.
Outcomes • Duration of labour

• Fetal outcome in terms of Apgar score
• Admission to NICU
Notes The study was carried out in the department of Obstetrics and Gynecology, Gauhati Medical College
and Hospital, Guwahati, Assam, India. Both groups recording of labour on WHO modified partograph
were started only in active phase of labour. According to guidelines of WHO modified partograph, ac-
tive phase has been considered at 4 cm of cervical dilatation.
The management of labouring women in both groups was unaffected if labour followed the expect-
ed rate of progress. However, if cervical dilatation crossed the allocated action line, a clinical assess-
ment was made and guidelines for the management of prolonged labour were followed. Where aug-
mentation was required, this involved oxytocin alone when membranes were ruptured or amniotomy
followed by oxytocin in the presence of intact membranes.
Funding: no funding sources
Conflict of interest: none declared
Risk of bias
Random sequence genera- High risk Not described - “randomised study” although it states groups were “divided
tion (selection bias) equally” into 2 groups which describes a quasi-randomisation process.
Allocation concealment High risk No clear information given but “equally divided into two groups” suggests
(selection bias) quasi-randomisation.
Blinding of participants High risk Not mentioned, but blinding of staff is not feasible with this type of interven-
and personnel (perfor- tion.
mance bias)
All outcomes
Blinding of outcome as- High risk Not reported, assumed not due to nature of intervention. Outcomes likely to
sessment (detection bias) be reported by unblinded clinicians.
Informed decisions.

Sinha 2017 (Continued)
All outcomes
Incomplete outcome data Unclear risk Loss to follow-up was not mentioned. No study flow diagram.
(attrition bias)
All outcomes
Selective reporting (re- Unclear risk Duration of labour not reported in this paper. Protocol not available and little
porting bias) information on methods.
Other bias Unclear risk Baseline characteristics of groups not mentioned.

Walss Rodriguez 1987
Methods Prospective study in which women "at random" were distributed in 1 of 2 groups. Conducted between
15 September to 25 October 1985.
Participants 434 women in Mexico, with term pregnancies who presented in labour (cervix 2 cm or more dilated)
with live, singleton, cephalic presentation.
Interventions One group had their labour managed according to the Friedman partogram and the other had labour
managed using a non-graphic, descriptive record.
Outcomes • Caesarean section

• Forceps delivery
• Normal delivery
• Apgar score
Notes This study was translated into English.
Declarations of interest: not mentioned in translation
Funding source: not mentioned in translation
Risk of bias
Random sequence genera- High risk Quasi-randomised study. No information on how randomisation was achieved.
Allocation concealment High risk No information on how women were allocated to groups, not clear that group
(selection bias) allocation was truly random.
mance bias)
All outcomes

All outcomes
Incomplete outcome data Low risk No apparent loss to follow-up.

(attrition bias)
All outcomes
Informed decisions.

Walss Rodriguez 1987 (Continued)

Selective reporting (re- Unclear risk Unable to assess from translation.
porting bias)
Other bias Unclear risk Very little information on study methods was provided.

Windrim 2006
Methods Prospective randomised clinical trial. Computerised allocation, by telephone. Conducted between July
1997 and December 1999.
Participants 1932 primiparous women, in Toronto, Canada, with uncomplicated pregnancies at term, with con-
tractions every 3 to 5 minutes and cervix at least 3 cm dilated. Outcomes were stratified according to
whether labour was spontaneous or induced. Only data from women not induced were included (n =
1156).
Interventions Women were randomised to 1 of 2 groups: the standard group, who had the progress of labour chart-
ed in written notes, or the partograph group, whose progress in labour was recorded using a bedside
graphical partograph as well as written notes.
Outcomes • Rate of caesarean section

• Operative vaginal delivery
• Spontaneous vaginal delivery
• Duration of first stage of labour
• Duration of second stage of labour
• Epidural analgesia use
• Artificial rupture of membranes
• Evaluation for non-reassuring fetal heart tracing
• Maternal and neonatal morbidity
Notes Only data from those in spontaneous labour are included in the review.
Funding source: trial was supported by Grant # 96–33 from the Physicians’ Services Incorporated
Foundation, Canada
Risk of bias
Random sequence genera- Low risk Stratified randomisation by off-site computerised randomisation service.
Allocation concealment Low risk By telephone to off-site service.

(selection bias)
Blinding of participants High risk Not feasible to blind women or clinical staff. Used bedside charts.
mance bias)
All outcomes

Informed decisions.

Windrim 2006 (Continued)
All outcomes
Incomplete outcome data Low risk No missing data apparent.

(attrition bias)
All outcomes
porting bias)
Other bias Unclear risk No information on the number of women approached or the numbers of eligi-
ble women declining participation.
LSCS: lower segment caesarean section

NICE: National Institute for Health and Care Excellence
NICU: neonatal intensive care unit
WHO: World Health Organization

Characteristics of excluded studies [ordered by study ID]

Study Reason for exclusion
Ajoodani 2011 Both groups had labour monitored by a partograph.
Cartmill 1992 A report of a hypothetical study. No research conducted and no data presented.
Fahdhy 2005 This was a cluster-randomised trial in which midwives were randomised to receive training, along-
side using the partograph. The intervention was therefore the training and not the partograph.
There is no description of what midwives in the control group received.
Hamilton 2001 This study was presented in abstract form only and lacked detail. It was particularly unclear
whether participants were in spontaneous labour and whether they were at term. We attempted to
contact the trial author, without success.
Hamilton 2004 The study intervention was a computerised reference range, not a partograph.
Kogovsek 2000 It was unclear from the presentation of data which outcome data were from women in sponta-
neous labour. We were unable to contact any of the authors.
Mathews 2007 This was a cross-over trial comparing 2 partographs; 1 which included a latent phase and 1 which
did not. In this study all physicians posted to the labour ward used the first partograph (composite
or simplified depending on the random allocation) for 10 days. After 1 week's break, all physicians
used the second partograph. Study participants were therefore physicians and not women.
WHO 1994 This was an observational study, not a randomised controlled trial.

Characteristics of studies awaiting assessment [ordered by study ID]

NCT02911272
Methods Randomised controlled trial in a university teaching hospital in Nigeria.
Participants Nulliparous women in labour at term
Inclusion criteria
Informed decisions.

NCT02911272 (Continued)
• Healthy nulliparous women at term in spontaneous active phase labour with singleton pregnan-
cies and intact fetal membranes at first vaginal examination in labour ward
Exclusion criteria
• Multiparous patients
• Multiple pregnancies
• Pre-existing medical conditions complicating pregnancy
• Non-cephalic presenting fetus at term
• Prelabour rupture of membrane at term
• Preterm labour
• Poor fetal growth in pregnancy
Interventions Partograph with 2-hour action line versus partograph with a 4-hour action line.
Outcomes Prolonged labour, mode of birth, neonatal outcome.
Notes Awaiting more information from authors - contacted 21/04/2017 'draziken@ubth.org' as could not
locate Dr Orhue's personal email.

Characteristics of ongoing studies [ordered by study ID]

NCT02714270
Trial name or title A randomised clinical trial of paperless versus modified World Health Organization partograph in
management of first stage of labour
Methods Single-blind, parallel randomised controlled trial
Participants Inclusion criteria
• Age: 18 to 40 years
• Gestational age 38 to 42 weeks
• Singleton pregnancy
• Vertex presentation
• Women who agree to participate in the study
Exclusion criteria
• Malpresentation
• Induced labour
• Medical diseases with pregnancy
Interventions Modified partograph versus paperless partograph
Outcomes Vaginal births
Starting date April 2016
Contact information Ahmed Mohamed Abbas, Assiut University
No contact details given
Notes Estimated completion date: August 2017
Informed decisions.

Trial conducted in Egypt

NCT02741141
Trial name or title A comparative study of the effect of two partographs on the cesarean section rate in women in
spontaneous labour (PARTODYS)
Methods Randomised controlled trial
• Age ≥ 18 years
• Affiliation to a social security insurance
• Written consent given
• Cephalic presentation
• ≥ 37 gestational weeks
• Spontaneous onset of labour
Exclusion criteria
• Previous caesarean section

• Induction of labour
• Intrauterine growth restriction
• In utero fetal death
• Congenital malformation
• Chorioamnionitis
• Placenta praevia
• Need for an emergency delivery (fetal heart rate abnormalities at admission)
• Contraindication for vaginal delivery
• Patient under temporary guardianship, guardianship or judicial protection
• Patient included in another study which could interfere with the results of this study
Interventions Classic partograph (labour dystocia is diagnosed when cervical dilation is less than 1 cm per hour
or after 3 hours at complete cervical dilation without engagement of the presentation. In this case,
active management of labour is started with introduction of oxytocin, artificial rupture of mem-
branes and supportive therapy) versus new partograph (the second strategy is based on the par-
tograph developed by Neal and Lowe. An active management of labour is started when crossing
the dystocia line or when there are no cervical modifications after 4 hours beyond 5 cm of cervical
dilation. In this case, active management of labour is started with introduction of oxytocin, artifi-
cial rupture of membranes and supportive therapy.)
Outcomes Primary outcome
Secondary outcomes
• Total amount of oxytocin used

• Uterine hyperstimulation
• Uterine rupture
• Retained placenta
• Need for artificial rupture of membranes
• Duration of labour
Informed decisions.

• Rate of vaginal birth
• Need for epidural
• Transfusion rate
• Maternal fever
• Apgar score < 7 at 5 minutes
• Neonatal intensive care unit admission
• Neonatal death
• Neonatal infection rate
Starting date January 2016
Contact information Contact: Adrien GAUDINEAU, MD 03 88 12 74 61
adrien.gaudineau@chru-strasbourg.fr
Notes Estimated end date: January 2018
Trial conducted at University Hospital, Strasbourg, France

NTR5543
Trial name or title Randomised controlled trial comparing the currently used Friedman partogram with a 4-hour ac-
tion line to the newly developed SIMPLE partogram, based on the 95th percentile normogram of
the Consortium on Safe Labor to evaluate early versus delayed cesarean section (SIMPLE III)
Methods Randomised controlled trial
• Nulliparous
• Cephalic presentation
• ≥ 37 weeks of pregnancy
• ≥ 4 cm dilatation
• Non-progressing labour (< 1 cm/hour)
Exclusion criteria
• < 18 years of age

• Unable to read or understand informed consent
• Fetus with relevant congenital malformation
• Planned caesarean section
• Non-reassuring fetal condition
• Inadequate pain medication (as judged by the women herself, no absolute need for epidural, but
pain needs to be considered manageable for continuation of delivery)
Interventions When after all regular interventions for non-progressing labour (amniotomy, oxytocin augmenta-
tion, empty bladder, pain medication) the Friedman partogram action line is crossed, randomisa-
tion occurs in which performing a caesarean section is the regular conduct. Women in the interven-
tion group wait until the Simple partogram action line is crossed (based on the 95th percentile of
the nomogram of the Consortium on Safe Labor)
Outcomes Primary outcomes
Informed decisions.

NTR5543 (Continued)
• Mortality and composite morbidity
* maternal intensive care unit admittance
* Apgar score < 7 after 5 minutes
* pH < 7.10
* neonatal intensive care unit admittance
Secondary outcomes
• Delivery mode
• Shoulder dystocia
• Anal sphincter lesions
• Blood loss
• Need for blood transfusion
• Maternal infection
• Neonatal infection
• Duration of admission to the hospital
• Total number of caesarean sections in the target population (including non-participating women)
• Cost evaluation
• Budget impact
• Patient preference
• Patient satisfaction
Starting date December 2015
Contact information hcj.scheepers@mumc.nl
Notes Estimated end date: December 2018
Trial website: http://www.studies-obsgyn.nl/simple3

DATA AND ANALYSES

Comparison 1. Partograph versus no partograph (studies carried out in high- and low-resource settings)
Outcome or subgroup title No. of No. of Statistical method Effect size

studies partici-
pants
1 Caesarean section (overall) 3 1813 Risk Ratio (M-H, Random, 95% 0.77 [0.40, 1.46]
CI)
1.1 Low-resource setting 2 657 Risk Ratio (M-H, Random, 95% 0.65 [0.22, 1.91]
CI)
1.2 High-resource setting 1 1156 Risk Ratio (M-H, Random, 95% 1.03 [0.82, 1.28]
CI)
2 Oxytocin augmentation 1 1156 Risk Ratio (M-H, Fixed, 95% CI) 1.02 [0.95, 1.10]
2.1 High-resource setting 1 1156 Risk Ratio (M-H, Fixed, 95% CI) 1.02 [0.95, 1.10]
3 Duration of first stage of labour 1 1156 Mean Difference (IV, Fixed, 95% 0.80 [-0.06, 1.66]
CI)
Informed decisions.


studies partici-
pants
3.1 High-resource setting 1 1156 Mean Difference (IV, Fixed, 95% 0.80 [-0.06, 1.66]
CI)
4 Low Apgar score (less than 7 at 5 minutes) 2 1596 Risk Ratio (M-H, Fixed, 95% CI) 0.76 [0.29, 2.03]
4.1 Low-resource setting 1 440 Risk Ratio (M-H, Fixed, 95% CI) 0.46 [0.04, 5.00]
5 Instrumental vaginal birth 3 1813 Risk Ratio (M-H, Fixed, 95% CI) 0.99 [0.84, 1.15]
6 Regional analgesia - epidural 1 1156 Risk Ratio (M-H, Fixed, 95% CI) 1.01 [0.98, 1.05]
7 Performance of artificial rupture of mem- 1 1156 Risk Ratio (M-H, Fixed, 95% CI) 0.99 [0.88, 1.11]
branes during labour
8 Antibiotic use 1 1156 Risk Ratio (M-H, Fixed, 95% CI) 1.23 [0.88, 1.73]
9 Duration of second stage of labour (hours) 1 1156 Mean Difference (IV, Fixed, 95% 0.0 [-0.21, 0.21]
CI)
9.1 High-resource setting 1 1156 Mean Difference (IV, Fixed, 95% 0.0 [-0.21, 0.21]
CI)
10 Number of vaginal examinations 1 1156 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
10.1 High-resource setting 1 1156 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
11 Admission to special care nursery 1 1156 Risk Ratio (M-H, Fixed, 95% CI) 0.94 [0.51, 1.75]

Informed decisions.

Analysis 1.1. Comparison 1 Partograph versus no partograph (studies carried

out in high- and low-resource settings), Outcome 1 Caesarean section (overall).
Study or subgroup Partograph No partograph Risk Ratio Weight Risk Ratio
n/N n/N M-H, Random, 95% CI M-H, Random, 95% CI
1.1.1 Low-resource setting
Rani 2015 21/110 19/113 30.2% 1.14[0.65,1.99]
Walss Rodriguez 1987 21/224 52/210 32.41% 0.38[0.24,0.61]
Subtotal (95% CI) 334 323 62.62% 0.65[0.22,1.91]
Total events: 42 (Partograph), 71 (No partograph)
Heterogeneity: Tau2=0.54; Chi2=8.66, df=1(P=0); I2=88.45%
Test for overall effect: Z=0.79(P=0.43)

1.1.2 High-resource setting
Windrim 2006 125/580 121/576 37.38% 1.03[0.82,1.28]
Subtotal (95% CI) 580 576 37.38% 1.03[0.82,1.28]
Heterogeneity: Not applicable

Total (95% CI) 914 899 100% 0.77[0.4,1.46]
Heterogeneity: Tau2=0.28; Chi2=15.07, df=2(P=0); I2=86.73%
Test for subgroup differences: Chi2=0.67, df=1 (P=0.41), I2=0%
Favours partograph 0.1 0.2 0.5 1 2 5 10 Favours no partograph

out in high- and low-resource settings), Outcome 2 Oxytocin augmentation.
n/N n/N M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Windrim 2006 423/580 412/576 100% 1.02[0.95,1.1]
Subtotal (95% CI) 580 576 100% 1.02[0.95,1.1]

Total (95% CI) 580 576 100% 1.02[0.95,1.1]

Informed decisions.

Analysis 1.3. Comparison 1 Partograph versus no partograph (studies carried out

in high- and low-resource settings), Outcome 3 Duration of first stage of labour.
Study or subgroup Partograph No partograph Mean Difference Weight Mean Difference
N Mean(SD) N Mean(SD) Fixed, 95% CI Fixed, 95% CI
Windrim 2006 580 16.8 (7.3) 576 16 (7.6) 100% 0.8[-0.06,1.66]
Subtotal *** 580 576 100% 0.8[-0.06,1.66]
Heterogeneity: Tau2=0; Chi2=0, df=0(P<0.0001); I2=100%

Total *** 580 576 100% 0.8[-0.06,1.66]
Favours partograph -10 -5 0 5 10 Favours no partograph

Analysis 1.4. Comparison 1 Partograph versus no partograph (studies carried out in
high- and low-resource settings), Outcome 4 Low Apgar score (less than 7 at 5 minutes).
Walss Rodriguez 1987 1/230 2/210 22.94% 0.46[0.04,5]
Subtotal (95% CI) 230 210 22.94% 0.46[0.04,5]

Windrim 2006 6/580 7/576 77.06% 0.85[0.29,2.52]
Subtotal (95% CI) 580 576 77.06% 0.85[0.29,2.52]

Total (95% CI) 810 786 100% 0.76[0.29,2.03]
Heterogeneity: Tau2=0; Chi2=0.22, df=1(P=0.64); I2=0%
Favours partograph 0.05 0.2 1 5 20 Favours no partograph

out in high- and low-resource settings), Outcome 5 Instrumental vaginal birth.
Rani 2015 12/110 16/113 6.82% 0.77[0.38,1.55]
Informed decisions.


Walss Rodriguez 1987 45/224 36/210 16.05% 1.17[0.79,1.74]
Subtotal (95% CI) 334 323 22.86% 1.05[0.75,1.48]
Heterogeneity: Tau2=0; Chi2=1.04, df=1(P=0.31); I2=4.22%

Windrim 2006 173/580 178/576 77.14% 0.97[0.81,1.15]
Subtotal (95% CI) 580 576 77.14% 0.97[0.81,1.15]

Total (95% CI) 914 899 100% 0.99[0.84,1.15]

out in high- and low-resource settings), Outcome 6 Regional analgesia - epidural.
Windrim 2006 532/580 521/576 100% 1.01[0.98,1.05]
Subtotal (95% CI) 580 576 100% 1.01[0.98,1.05]

Total (95% CI) 580 576 100% 1.01[0.98,1.05]

Analysis 1.7. Comparison 1 Partograph versus no partograph (studies carried out in high- and
low-resource settings), Outcome 7 Performance of artificial rupture of membranes during labour.
Windrim 2006 283/580 284/576 100% 0.99[0.88,1.11]
Subtotal (95% CI) 580 576 100% 0.99[0.88,1.11]
Informed decisions.



Total (95% CI) 580 576 100% 0.99[0.88,1.11]

Analysis 1.8. Comparison 1 Partograph versus no partograph (studies
carried out in high- and low-resource settings), Outcome 8 Antibiotic use.
Windrim 2006 67/580 54/576 100% 1.23[0.88,1.73]
Subtotal (95% CI) 580 576 100% 1.23[0.88,1.73]

Total (95% CI) 580 576 100% 1.23[0.88,1.73]

Analysis 1.9. Comparison 1 Partograph versus no partograph (studies carried out in
high- and low-resource settings), Outcome 9 Duration of second stage of labour (hours).
Windrim 2006 580 2.4 (1.8) 576 2.4 (1.9) 100% 0[-0.21,0.21]
Subtotal *** 580 576 100% 0[-0.21,0.21]
Test for overall effect: Not applicable

Total *** 580 576 100% 0[-0.21,0.21]

Informed decisions.


in high- and low-resource settings), Outcome 10 Number of vaginal examinations.
Windrim 2006 580 4 (0) 576 4 (0) Not estimable
Subtotal *** 580 576 Not estimable

Total *** 580 576 Not estimable

in high- and low-resource settings), Outcome 11 Admission to special care nursery.
Windrim 2006 19/580 20/576 100% 0.94[0.51,1.75]
Subtotal (95% CI) 580 576 100% 0.94[0.51,1.75]

Total (95% CI) 580 576 100% 0.94[0.51,1.75]

Comparison 2. Partograph with 2-hour action line versus partograph with 4-hour action line (studies carried out in
a high- and low-resource settings)

studies partici-
pants
1 Caesarean section (overall) 4 4749 Risk Ratio (M-H, Fixed, 95% CI) 1.06 [0.88, 1.28]
2 Oxytocin augmentation 4 4749 Risk Ratio (M-H, Random, 95% CI) 2.44 [1.36, 4.35]
2.1 Low-resource setting 2 1148 Risk Ratio (M-H, Random, 95% CI) 7.22 [2.49, 20.91]
Informed decisions.


studies partici-
pants
2.2 High-resource setting 2 3601 Risk Ratio (M-H, Random, 95% CI) 1.14 [1.05, 1.22]
3 Duration of first stage of labour 1 948 Risk Ratio (M-H, Fixed, 95% CI) 0.81 [0.32, 2.04]
(length of labour greater than 18
hours, length of labour greater
than 12 hours)
4 Maternal experience of child- 2 2269 Risk Ratio (M-H, Random, 95% CI) 0.61 [0.28, 1.35]
birth - negative childbirth experi-
ence
5 Low Apgar score (less than 7 at 4 4749 Risk Ratio (M-H, Fixed, 95% CI) 0.93 [0.61, 1.42]
5 minutes)
6 Serious maternal morbidity or 2 3601 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
death
7 Caesarean section (distress) 2 3601 Risk Ratio (M-H, Fixed, 95% CI) 1.30 [0.86, 1.96]
8 Caesarean section (delay) 2 3601 Risk Ratio (M-H, Fixed, 95% CI) 0.98 [0.77, 1.25]
9 Instrumental vaginal delivery 3 3801 Risk Ratio (M-H, Fixed, 95% CI) 0.92 [0.81, 1.04]
10 Postpartum haemorrhage - 3 4549 Risk Ratio (M-H, Fixed, 95% CI) 1.06 [0.90, 1.25]
blood loss > 500 mL
11 Regional analgesia - epidural 2 3601 Risk Ratio (M-H, Random, 95% CI) 1.06 [0.92, 1.21]
12 Performance of artificial rup- 3 3801 Risk Ratio (M-H, Random, 95% CI) 1.00 [0.84, 1.18]
ture of the membranes during
labour
13 Number of vaginal examina- 2 3601 Mean Difference (IV, Random, 95% CI) -0.08 [-0.37, 0.21]
tions in labour
14 Serious neonatal morbidity or 2 3601 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
perinatal death
15 Admission to special care 3 3801 Risk Ratio (M-H, Fixed, 95% CI) 0.83 [0.51, 1.34]
nursery
Informed decisions.


studies partici-
pants
16 Cord blood arterial pH less 2 3601 Risk Ratio (M-H, Fixed, 95% CI) 0.73 [0.44, 1.22]
than 7.1

Analysis 2.1. Comparison 2 Partograph with 2-hour action line versus partograph with 4-hour action
line (studies carried out in a high- and low-resource settings), Outcome 1 Caesarean section (overall).
Study or subgroup 2-hour ac- 4-hour ac- Risk Ratio Weight Risk Ratio
tion line tion line
Orhue 2013 29/470 28/478 14.01% 1.05[0.64,1.74]
Sinha 2017 11/100 9/100 4.54% 1.22[0.53,2.82]
Subtotal (95% CI) 570 578 18.55% 1.09[0.71,1.68]
Total events: 40 (2-hour action line), 37 (4-hour action line)

Lavender 1998a 35/315 26/311 13.2% 1.33[0.82,2.15]
Lavender 2006 136/1490 135/1485 68.24% 1[0.8,1.26]
Subtotal (95% CI) 1805 1796 81.45% 1.06[0.86,1.3]

Total (95% CI) 2375 2374 100% 1.06[0.88,1.28]
Favours 2-hour 0.1 0.2 0.5 1 2 5 10 Favours 4-hour

line (studies carried out in a high- and low-resource settings), Outcome 2 Oxytocin augmentation.
tion line tion line
Orhue 2013 118/470 10/478 21.99% 12[6.37,22.6]
Sinha 2017 29/100 7/100 19.41% 4.14[1.9,9.01]
Subtotal (95% CI) 570 578 41.4% 7.22[2.49,20.91]
Heterogeneity: Tau2=0.46; Chi2=4.51, df=1(P=0.03); I2=77.83%
Test for overall effect: Z=3.64(P=0)

Favours 2-hour 0.02 0.1 1 10 50 Favours 4-hour
Informed decisions.

tion line tion line
Lavender 1998a 144/315 129/311 28.98% 1.1[0.92,1.32]
Lavender 2006 696/1490 607/1485 29.62% 1.14[1.05,1.24]
Subtotal (95% CI) 1805 1796 58.6% 1.14[1.05,1.22]

Total (95% CI) 2375 2374 100% 2.44[1.36,4.35]
Heterogeneity: Tau2=0.29; Chi2=69.4, df=3(P<0.0001); I2=95.68%
Test for subgroup differences: Chi2=11.57, df=1 (P=0), I2=91.36%

Analysis 2.3. Comparison 2 Partograph with 2-hour action line versus partograph with 4-hour
action line (studies carried out in a high- and low-resource settings), Outcome 3 Duration of first
stage of labour (length of labour greater than 18 hours, length of labour greater than 12 hours).
tion line tion line
Orhue 2013 8/470 10/478 100% 0.81[0.32,2.04]

Total (95% CI) 470 478 100% 0.81[0.32,2.04]

Analysis 2.4. Comparison 2 Partograph with 2-hour action line versus partograph
with 4-hour action line (studies carried out in a high- and low-resource settings),
Outcome 4 Maternal experience of childbirth - negative childbirth experience.
tion line tion line
Lavender 1998a 14/179 34/171 45.03% 0.39[0.22,0.71]
Lavender 2006 72/962 81/957 54.97% 0.88[0.65,1.2]

Total (95% CI) 1141 1128 100% 0.61[0.28,1.35]

Informed decisions.

Analysis 2.5. Comparison 2 Partograph with 2-hour action line versus partograph with 4-hour action line
(studies carried out in a high- and low-resource settings), Outcome 5 Low Apgar score (less than 7 at 5 minutes).
tion line tion line
Orhue 2013 7/470 6/478 13.51% 1.19[0.4,3.5]
Sinha 2017 6/100 4/100 9.08% 1.5[0.44,5.15]
Subtotal (95% CI) 570 578 22.6% 1.31[0.58,2.96]

Lavender 1998a 6/315 5/311 11.43% 1.18[0.37,3.84]
Lavender 2006 22/1490 29/1485 65.97% 0.76[0.44,1.31]
Subtotal (95% CI) 1805 1796 77.4% 0.82[0.5,1.35]

Total (95% CI) 2375 2374 100% 0.93[0.61,1.42]

(studies carried out in a high- and low-resource settings), Outcome 6 Serious maternal morbidity or death.
tion line tion line
Lavender 1998a 0/315 0/311 Not estimable
Lavender 2006 0/1490 0/1485 Not estimable

Total (95% CI) 1805 1796 Not estimable

Informed decisions.

line (studies carried out in a high- and low-resource settings), Outcome 7 Caesarean section (distress).
tion line tion line
Lavender 1998a 12/315 7/311 18.02% 1.69[0.68,4.24]
Lavender 2006 39/1490 32/1485 81.98% 1.21[0.77,1.93]

Total (95% CI) 1805 1796 100% 1.3[0.86,1.96]

line (studies carried out in a high- and low-resource settings), Outcome 8 Caesarean section (delay).
tion line tion line
Lavender 1998a 23/315 19/311 15.64% 1.2[0.66,2.15]
Lavender 2006 97/1490 103/1485 84.36% 0.94[0.72,1.23]

Total (95% CI) 1805 1796 100% 0.98[0.77,1.25]

line (studies carried out in a high- and low-resource settings), Outcome 9 Instrumental vaginal delivery.
tion line tion line
Lavender 1998a 66/315 73/311 18.28% 0.89[0.67,1.2]
Lavender 2006 294/1490 320/1485 79.73% 0.92[0.8,1.05]
Sinha 2017 9/100 8/100 1.99% 1.13[0.45,2.8]

Total (95% CI) 1905 1896 100% 0.92[0.81,1.04]

Informed decisions.

Analysis 2.10. Comparison 2 Partograph with 2-hour action line versus partograph with 4-hour action line (studies
carried out in a high- and low-resource settings), Outcome 10 Postpartum haemorrhage - blood loss > 500 mL.
tion line tion line
Orhue 2013 5/470 6/478 2.58% 0.85[0.26,2.76]
Subtotal (95% CI) 470 478 2.58% 0.85[0.26,2.76]

Lavender 1998a 39/315 39/311 17.03% 0.99[0.65,1.5]
Lavender 2006 201/1490 185/1485 80.39% 1.08[0.9,1.3]
Subtotal (95% CI) 1805 1796 97.42% 1.07[0.9,1.26]

Total (95% CI) 2275 2274 100% 1.06[0.9,1.25]

line (studies carried out in a high- and low-resource settings), Outcome 11 Regional analgesia - epidural.
tion line tion line
Lavender 1998a 120/315 101/311 30.13% 1.17[0.95,1.45]
Lavender 2006 479/1490 473/1485 69.87% 1.01[0.91,1.12]

Total (95% CI) 1805 1796 100% 1.06[0.92,1.21]
Favours 2-hour 1 Favours 4-hour

Informed decisions.

with 4-hour action line (studies carried out in a high- and low-resource settings),
Outcome 12 Performance of artificial rupture of the membranes during labour.
tion line tion line
Lavender 1998a 120/315 121/311 34.46% 0.98[0.8,1.19]
Lavender 2006 715/1490 657/1485 55.74% 1.08[1,1.17]
Sinha 2017 19/100 29/100 9.8% 0.66[0.39,1.09]

Total (95% CI) 1905 1896 100% 1[0.84,1.18]
Favours 2-hour 0.5 0.7 1 1.5 2 Favours 4-hour

(studies carried out in a high- and low-resource settings), Outcome 13 Number of vaginal examinations in labour.
Study or subgroup 2-hour action line 4-hour action line Mean Difference Weight Mean Difference
N Mean(SD) N Mean(SD) Random, 95% CI Random, 95% CI
Lavender 1998a 315 4 (1.9) 311 3.9 (1.8) 40.69% 0.1[-0.19,0.39]
Lavender 2006 1490 3.2 (1.9) 1485 3.4 (2) 59.31% -0.2[-0.34,-0.06]

Total *** 1805 1796 100% -0.08[-0.37,0.21]
Favours 2-hour -0.5 -0.25 0 0.25 0.5 Favours 4-hour

Analysis 2.14. Comparison 2 Partograph with 2-hour action line versus partograph with 4-hour action line (studies
carried out in a high- and low-resource settings), Outcome 14 Serious neonatal morbidity or perinatal death.
tion line tion line
Lavender 2006 0/1490 0/1485 Not estimable


Informed decisions.

(studies carried out in a high- and low-resource settings), Outcome 15 Admission to special care nursery.
tion line tion line
Sinha 2017 4/100 3/100 8.56% 1.33[0.31,5.81]
Subtotal (95% CI) 100 100 8.56% 1.33[0.31,5.81]

Lavender 1998a 4/315 2/311 5.74% 1.97[0.36,10.7]
Lavender 2006 21/1490 30/1485 85.7% 0.7[0.4,1.21]
Subtotal (95% CI) 1805 1796 91.44% 0.78[0.46,1.31]

Total (95% CI) 1905 1896 100% 0.83[0.51,1.34]

(studies carried out in a high- and low-resource settings), Outcome 16 Cord blood arterial pH less than 7.1.
tion line tion line
Lavender 1998a 2/315 2/311 5.91% 0.99[0.14,6.97]
Lavender 2006 23/1490 32/1485 94.09% 0.72[0.42,1.22]

Total (95% CI) 1805 1796 100% 0.73[0.44,1.22]

Comparison 3. Partograph with 2-hour action line versus partograph with 3-hour action line (study carried out in a
high-resource setting)

studies partici-
pants
Informed decisions.


studies partici-
pants
3 Maternal experience of childbirth 1 348 Risk Ratio (M-H, Fixed, 95% CI) 0.49 [0.27, 0.90]
- negative childbirth experience
4 Low Apgar score (less than 7 at 5 1 617 Risk Ratio (M-H, Fixed, 95% CI) 1.44 [0.41, 5.05]
minutes)
death
9 Postpartum haemorrhage - 1 617 Risk Ratio (M-H, Fixed, 95% CI) 0.96 [0.63, 1.45]
blood loss > 500 mL
11 Performance of artificial rup- 1 617 Risk Ratio (M-H, Fixed, 95% CI) 0.94 [0.77, 1.15]
ture of membranes during labour
12 Vaginal examinations 1 617 Mean Difference (IV, Fixed, 95% CI) 0.0 [-0.29, 0.29]
perinatal death
14 Admission to special care nurs- 1 617 Risk Ratio (M-H, Fixed, 95% CI) 3.83 [0.43, 34.12]
ery
15 Cord blood arterial pH less than 1 617 Risk Ratio (M-H, Fixed, 95% CI) 0.38 [0.07, 1.96]
7.1

action line (study carried out in a high-resource setting), Outcome 1 Caesarean section (overall).
tion line tion line
Lavender 1998a 35/315 43/302 100% 0.78[0.51,1.18]

Total (95% CI) 315 302 100% 0.78[0.51,1.18]
Informed decisions.


action line (study carried out in a high-resource setting), Outcome 2 Oxytocin augmentation.
tion line tion line
Lavender 1998a 144/315 136/302 100% 1.02[0.85,1.21]

Total (95% CI) 315 302 100% 1.02[0.85,1.21]

Analysis 3.3. Comparison 3 Partograph with 2-hour action line versus
partograph with 3-hour action line (study carried out in a high-resource setting),
tion line tion line
Lavender 1998a 14/179 27/169 100% 0.49[0.27,0.9]

Total (95% CI) 179 169 100% 0.49[0.27,0.9]

line (study carried out in a high-resource setting), Outcome 4 Low Apgar score (less than 7 at 5 minutes).
tion line tion line
Lavender 1998a 6/315 4/302 100% 1.44[0.41,5.05]

Total (95% CI) 315 302 100% 1.44[0.41,5.05]

Informed decisions.

line (study carried out in a high-resource setting), Outcome 5 Serious maternal morbidity or death.
tion line tion line


action line (study carried out in a high-resource setting), Outcome 6 Caesarean section (distress).
tion line tion line
Lavender 1998a 12/315 12/302 100% 0.96[0.44,2.1]

Total (95% CI) 315 302 100% 0.96[0.44,2.1]

action line (study carried out in a high-resource setting), Outcome 7 Caesarean section (delay).
tion line tion line
Lavender 1998a 23/315 31/302 100% 0.71[0.42,1.19]

Total (95% CI) 315 302 100% 0.71[0.42,1.19]

Informed decisions.

action line (study carried out in a high-resource setting), Outcome 8 Instrumental vaginal delivery.
tion line tion line
Lavender 1998a 66/315 68/302 100% 0.93[0.69,1.26]

Total (95% CI) 315 302 100% 0.93[0.69,1.26]

(study carried out in a high-resource setting), Outcome 9 Postpartum haemorrhage - blood loss > 500 mL.
tion line tion line
Lavender 1998a 39/315 39/302 100% 0.96[0.63,1.45]

Total (95% CI) 315 302 100% 0.96[0.63,1.45]

action line (study carried out in a high-resource setting), Outcome 10 Regional analgesia - epidural.
tion line tion line
Lavender 1998a 120/315 99/302 100% 1.16[0.94,1.44]

Total (95% CI) 315 302 100% 1.16[0.94,1.44]

Informed decisions.

Analysis 3.11. Comparison 3 Partograph with 2-hour action line versus partograph with 3-hour action line (study
carried out in a high-resource setting), Outcome 11 Performance of artificial rupture of membranes during labour.
tion line tion line
Lavender 1998a 120/315 122/302 100% 0.94[0.77,1.15]

Total (95% CI) 315 302 100% 0.94[0.77,1.15]

Analysis 3.12. Comparison 3 Partograph with 2-hour action line versus partograph with 3-
hour action line (study carried out in a high-resource setting), Outcome 12 Vaginal examinations.
Lavender 1998a 315 4 (1.9) 302 4 (1.8) 100% 0[-0.29,0.29]

Total *** 315 302 100% 0[-0.29,0.29]
Favours 2-hour -10 -5 0 5 10 Favours 3-hour

(study carried out in a high-resource setting), Outcome 13 Serious neonatal morbidity or perinatal death.
tion line tion line


line (study carried out in a high-resource setting), Outcome 14 Admission to special care nursery.
tion line tion line
Lavender 1998a 4/315 1/302 100% 3.83[0.43,34.12]
Informed decisions.

tion line tion line

Total (95% CI) 315 302 100% 3.83[0.43,34.12]

line (study carried out in a high-resource setting), Outcome 15 Cord blood arterial pH less than 7.1.
tion line tion line
Lavender 1998a 2/315 5/302 100% 0.38[0.07,1.96]

Total (95% CI) 315 302 100% 0.38[0.07,1.96]

Comparison 4. Partograph with 3-hour action line versus partograph with 4-hour action line (study carried out in a
high-resource setting)

studies partici-
pants
3 Maternal experience of childbirth - 1 340 Risk Ratio (M-H, Fixed, 95% CI) 0.80 [0.51, 1.27]
negative childbirth experience
minutes)
death
9 Postpartum haemorrhage - blood 1 613 Risk Ratio (M-H, Fixed, 95% CI) 1.03 [0.68, 1.56]
loss > 500 mL
Informed decisions.


studies partici-
pants
11 Performance of artificial rupture 1 613 Risk Ratio (M-H, Fixed, 95% CI) 1.04 [0.85, 1.26]
of membranes during labour
12 Number of vaginal examinations 1 613 Mean Difference (IV, Fixed, 95% CI) 0.10 [-0.19, 0.39]
in labour
perinatal death
14 Admission to special care nursery 1 613 Risk Ratio (M-H, Fixed, 95% CI) 0.51 [0.05, 5.65]
15 Cord blood arterial pH less than 1 613 Risk Ratio (M-H, Fixed, 95% CI) 2.57 [0.50, 13.17]
7.1

action line (study carried out in a high-resource setting), Outcome 1 Caesarean section (overall).
tion line tion line
Lavender 1998a 43/302 26/311 100% 1.7[1.07,2.7]

Total (95% CI) 302 311 100% 1.7[1.07,2.7]

action line (study carried out in a high-resource setting), Outcome 2 Oxytocin augmentation.
tion line tion line
Lavender 1998a 136/302 129/311 100% 1.09[0.91,1.3]

Total (95% CI) 302 311 100% 1.09[0.91,1.3]

Informed decisions.


partograph with 4-hour action line (study carried out in a high-resource setting),
tion line tion line
Lavender 1998a 27/169 34/171 100% 0.8[0.51,1.27]

Total (95% CI) 169 171 100% 0.8[0.51,1.27]

line (study carried out in a high-resource setting), Outcome 4 Low Apgar score (less than 7 at 5 minutes).
tion line tion line
Lavender 1998a 4/302 5/311 100% 0.82[0.22,3.04]

Total (95% CI) 302 311 100% 0.82[0.22,3.04]

line (study carried out in a high-resource setting), Outcome 5 Serious maternal morbidity or death.
tion line tion line


Informed decisions.

action line (study carried out in a high-resource setting), Outcome 6 Caesarean section (distress).
tion line tion line
Lavender 1998a 12/302 7/311 100% 1.77[0.7,4.42]

Total (95% CI) 302 311 100% 1.77[0.7,4.42]

action line (study carried out in a high-resource setting), Outcome 7 Caesarean section (delay).
tion line tion line
Lavender 1998a 31/302 19/311 100% 1.68[0.97,2.91]

Total (95% CI) 302 311 100% 1.68[0.97,2.91]

action line (study carried out in a high-resource setting), Outcome 8 Instrumental vaginal delivery.
tion line tion line
Lavender 1998a 68/302 73/311 100% 0.96[0.72,1.28]

Total (95% CI) 302 311 100% 0.96[0.72,1.28]

Informed decisions.

(study carried out in a high-resource setting), Outcome 9 Postpartum haemorrhage - blood loss > 500 mL.
tion line tion line
Lavender 1998a 39/302 39/311 100% 1.03[0.68,1.56]

Total (95% CI) 302 311 100% 1.03[0.68,1.56]

action line (study carried out in a high-resource setting), Outcome 10 Regional analgesia - epidural.
tion line tion line
Lavender 1998a 99/302 101/311 100% 1.01[0.8,1.27]

Total (95% CI) 302 311 100% 1.01[0.8,1.27]

Analysis 4.11. Comparison 4 Partograph with 3-hour action line versus partograph with 4-hour action line (study
carried out in a high-resource setting), Outcome 11 Performance of artificial rupture of membranes during labour.
tion line tion line
Lavender 1998a 122/302 121/311 100% 1.04[0.85,1.26]

Total (95% CI) 302 311 100% 1.04[0.85,1.26]

Informed decisions.

line (study carried out in a high-resource setting), Outcome 12 Number of vaginal examinations in labour.
Lavender 1998a 302 4 (1.8) 311 3.9 (1.8) 100% 0.1[-0.19,0.39]

Total *** 302 311 100% 0.1[-0.19,0.39]
Favours 3-hour -10 -5 0 5 10 Favours 4-hour

(study carried out in a high-resource setting), Outcome 13 Serious neonatal morbidity or perinatal death.
tion line tion line


line (study carried out in a high-resource setting), Outcome 14 Admission to special care nursery.
tion line tion line
Lavender 1998a 1/302 2/311 100% 0.51[0.05,5.65]

Total (95% CI) 302 311 100% 0.51[0.05,5.65]

line (study carried out in a high-resource setting), Outcome 15 Cord blood arterial pH less than 7.1.
tion line tion line
Lavender 1998a 5/302 2/311 100% 2.57[0.5,13.17]
Informed decisions.

tion line tion line

Total (95% CI) 302 311 100% 2.57[0.5,13.17]

Comparison 5. Partograph with alert line only versus partograph with alert and action line (study carried out in a
low-resource setting)

studies partici-
pants
minutes)
perinatal death

Analysis 5.1. Comparison 5 Partograph with alert line only versus partograph with alert and
action line (study carried out in a low-resource setting), Outcome 1 Caesarean section (overall).
Study or subgroup alert line only alert and action Risk Ratio Weight Risk Ratio
Pattinson 2003 55/344 82/350 100% 0.68[0.5,0.93]

Total (95% CI) 344 350 100% 0.68[0.5,0.93]
Total events: 55 (alert line only), 82 (alert and action)
Favours alert line only 0.1 0.2 0.5 1 2 5 10 Favours alert and action

Informed decisions.

action line (study carried out in a low-resource setting), Outcome 2 Oxytocin augmentation.
Study or subgroup Alert line only Alert and action Risk Ratio Weight Risk Ratio
Pattinson 2003 77/344 97/350 100% 0.81[0.62,1.05]

Total (95% CI) 344 350 100% 0.81[0.62,1.05]
Total events: 77 (Alert line only), 97 (Alert and action)

Analysis 5.3. Comparison 5 Partograph with alert line only versus partograph with alert and action
line (study carried out in a low-resource setting), Outcome 3 Low Apgar score (less than 7 at 5 minutes).
Pattinson 2003 3/344 0/350 100% 7.12[0.37,137.36]

Total (95% CI) 344 350 100% 7.12[0.37,137.36]
Favours alert line only 0.001 0.1 1 10 1000 Favours alert and action

action line (study carried out in a low-resource setting), Outcome 4 Instrumental vaginal delivery.
Pattinson 2003 70/344 82/350 100% 0.87[0.66,1.15]

Total (95% CI) 344 350 100% 0.87[0.66,1.15]

Analysis 5.5. Comparison 5 Partograph with alert line only versus partograph with alert and action line
(study carried out in a low-resource setting), Outcome 5 Serious neonatal morbidity or perinatal death.
Pattinson 2003 3/344 0/350 100% 7.12[0.37,137.36]

Total (95% CI) 344 350 100% 7.12[0.37,137.36]
Informed decisions.


Comparison 6. Partograph with latent phase versus partograph without latent phase (study carried out in a low-
resource setting)

studies partici-
pants
3 Low Apgar score (less than 7 1 743 Risk Ratio (M-H, Fixed, 95% CI) 0.75 [0.21, 2.63]
at 5 minutes)
7 Admission to special care 1 743 Risk Ratio (M-H, Fixed, 95% CI) 1.84 [1.29, 2.63]
nursery
8 Usability: user-friendliness 1 743 Mean Difference (IV, Fixed, 95% CI) -7.89 [-8.14, -7.64]
score

Analysis 6.1. Comparison 6 Partograph with latent phase versus partograph without latent
phase (study carried out in a low-resource setting), Outcome 1 Caesarean section (overall).
Study or subgroup Partograph Partograph Risk Ratio Weight Risk Ratio
with la- without la-
tent phase tent phase
Kenchaveeriah 2011 83/350 38/393 100% 2.45[1.72,3.5]

Total (95% CI) 350 393 100% 2.45[1.72,3.5]
Total events: 83 (Partograph with latent phase), 38 (Partograph without
latent phase)
Test for overall effect: Z=4.94(P<0.0001)
Partograph with latent phase 0.1 0.2 0.5 1 2 5 10 Partograph without latent phase

Informed decisions.

phase (study carried out in a low-resource setting), Outcome 2 Oxytocin augmentation.
Kenchaveeriah 2011 126/350 65/393 100% 2.18[1.67,2.83]

Total (95% CI) 350 393 100% 2.18[1.67,2.83]
latent phase)

Analysis 6.3. Comparison 6 Partograph with latent phase versus partograph without latent phase
(study carried out in a low-resource setting), Outcome 3 Low Apgar score (less than 7 at 5 minutes).
Kenchaveeriah 2011 4/350 6/393 100% 0.75[0.21,2.63]

Total (95% CI) 350 393 100% 0.75[0.21,2.63]
Total events: 4 (Partograph with latent phase), 6 (Partograph without la-
tent phase)
Partograph with latent phase 0.02 0.1 1 10 50 Partograph without latent phase

phase (study carried out in a low-resource setting), Outcome 4 Caesarean section (distress).
Kenchaveeriah 2011 65/350 15/393 100% 4.87[2.83,8.37]

Total (95% CI) 350 393 100% 4.87[2.83,8.37]
latent phase)

Informed decisions.

phase (study carried out in a low-resource setting), Outcome 5 Caesarean section (delay).
Kenchaveeriah 2011 12/350 10/393 100% 1.35[0.59,3.08]

Total (95% CI) 350 393 100% 1.35[0.59,3.08]
latent phase)

phase (study carried out in a low-resource setting), Outcome 6 Instrumental vaginal delivery.
Kenchaveeriah 2011 24/350 26/393 100% 1.04[0.61,1.77]

Total (95% CI) 350 393 100% 1.04[0.61,1.77]
latent phase)

phase (study carried out in a low-resource setting), Outcome 7 Admission to special care nursery.
Kenchaveeriah 2011 69/350 42/393 100% 1.84[1.29,2.63]

Total (95% CI) 350 393 100% 1.84[1.29,2.63]
latent phase)

Informed decisions.

phase (study carried out in a low-resource setting), Outcome 8 Usability: user-friendliness score.
Study or subgroup Partograph with Partograph with- Mean Difference Weight Mean Difference
latent phase out latent phase
Kenchaveeriah 2011 350 2.9 (1.9) 393 10.8 (1.6) 100% -7.89[-8.14,-7.64]

Total *** 350 393 100% -7.89[-8.14,-7.64]
Partograph without latent -10 -5 0 5 10 Partograph with latent

Comparison 7. Partograph with 2-hour action line versus partograph with stepped dystocia line

studies partici-
pants
1 Caesarean section 1 99 Risk Ratio (M-H, Fixed, 95% CI) 1.10 [0.46, 2.62]
3 Duration of first stage of 1 99 Risk Ratio (M-H, Fixed, 95% CI) 0.76 [0.31, 1.89]
labour (labour longer than 12
hours)
4 Maternal experience of 1 90 Mean Difference (IV, Fixed, 95% CI) 0.0 [-3.58, 3.58]
childbirth (BSS-R score)
5 Low Apgar score (less than 1 99 Risk Ratio (M-H, Fixed, 95% CI) 0.98 [0.06, 15.23]
4 at 4 min)
6 Instrumental vaginal birth 1 99 Risk Ratio (M-H, Fixed, 95% CI) 0.75 [0.37, 1.56]
7 Postpartum haemorrhage 1 99 Risk Ratio (M-H, Fixed, 95% CI) 1.57 [0.55, 4.46]
(> 500 mL)
8 Regional analgesia 1 99 Risk Ratio (M-H, Fixed, 95% CI) 0.86 [0.56, 1.32]
9 Opioid use 1 99 Risk Ratio (M-H, Fixed, 95% CI) 0.98 [0.45, 2.14]
10 Need for intubation at 1 99 Risk Ratio (M-H, Fixed, 95% CI) 0.44 [0.14, 1.32]
birth

Informed decisions.


partograph with stepped dystocia line, Outcome 1 Caesarean section.
Study or subgroup Dystocia line 2-hour ac- Risk Ratio Weight Risk Ratio
tion line
Lee 2015 9/50 8/49 100% 1.1[0.46,2.62]

Total (95% CI) 50 49 100% 1.1[0.46,2.62]
Total events: 9 (Dystocia line), 8 (2-hour action line)
Favours dystocia line 0.01 0.1 1 10 100 Favours 2-hour

partograph with stepped dystocia line, Outcome 2 Oxytocin augmentation.
tion line
Lee 2015 17/50 27/49 100% 0.62[0.39,0.98]

Total (95% CI) 50 49 100% 0.62[0.39,0.98]
Favours dystocia line 0.5 0.7 1 1.5 2 Favours 2-hour

Analysis 7.3. Comparison 7 Partograph with 2-hour action line versus partograph with stepped
dystocia line, Outcome 3 Duration of first stage of labour (labour longer than 12 hours).
tion line
Lee 2015 7/50 9/49 100% 0.76[0.31,1.89]

Total (95% CI) 50 49 100% 0.76[0.31,1.89]

Informed decisions.

with stepped dystocia line, Outcome 4 Maternal experience of childbirth (BSS-R score).
Study or subgroup Dystocia line 2-hour action line Mean Difference Weight Mean Difference
Lee 2015 46 24.3 (8.7) 44 24.3 (8.7) 100% 0[-3.58,3.58]

Total *** 46 44 100% 0[-3.58,3.58]
Favours 2-hour -10 -5 0 5 10 Favours dystocia line

with stepped dystocia line, Outcome 5 Low Apgar score (less than 4 at 4 min).
tion line
Lee 2015 1/50 1/49 100% 0.98[0.06,15.23]

Total (95% CI) 50 49 100% 0.98[0.06,15.23]

partograph with stepped dystocia line, Outcome 6 Instrumental vaginal birth.
tion line
Lee 2015 10/50 13/49 100% 0.75[0.37,1.56]

Total (95% CI) 50 49 100% 0.75[0.37,1.56]

with stepped dystocia line, Outcome 7 Postpartum haemorrhage (> 500 mL).
tion line
Lee 2015 8/50 5/49 100% 1.57[0.55,4.46]
Informed decisions.

tion line

Total (95% CI) 50 49 100% 1.57[0.55,4.46]

partograph with stepped dystocia line, Outcome 8 Regional analgesia.
tion line
Lee 2015 21/50 24/49 100% 0.86[0.56,1.32]

Total (95% CI) 50 49 100% 0.86[0.56,1.32]

Analysis 7.9. Comparison 7 Partograph with 2-hour action line
versus partograph with stepped dystocia line, Outcome 9 Opioid use.
tion line
Lee 2015 10/50 10/49 100% 0.98[0.45,2.14]

Total (95% CI) 50 49 100% 0.98[0.45,2.14]
Favours dystocia use 0.01 0.1 1 10 100 Favours 2-hour

partograph with stepped dystocia line, Outcome 10 Need for intubation at birth.
tion line
Lee 2015 4/50 9/49 100% 0.44[0.14,1.32]

Total (95% CI) 50 49 100% 0.44[0.14,1.32]
Informed decisions.

tion line

Comparison 8. Partograph versus labour scale (study carried out in a low-resource setting)

studies partici-
pants
3 Duration of first stage of labour 1 110 Mean Difference (IV, Fixed, 95% CI) 0.44 [-0.40, 1.28]
minutes)
6 Stillbirth, neonatal death or 1 110 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
neonatal morbidity
7 Birth injuries and PPH (non-pre- 1 122 Risk Ratio (M-H, Fixed, 95% CI) 3.0 [0.32, 28.04]
specified outcome)

Analysis 8.1. Comparison 8 Partograph versus labour scale (study carried
out in a low-resource setting), Outcome 1 Caesarean section (overall).
Study or subgroup Labour scale WHO par- Risk Ratio Weight Risk Ratio
tograph
Shazly 2017 5/61 12/61 100% 0.42[0.16,1.11]

Total (95% CI) 61 61 100% 0.42[0.16,1.11]
Total events: 5 (Labour scale), 12 (WHO partograph)
Favours labour scale 0.01 0.1 1 10 100 Favours partograph

Informed decisions.

Analysis 8.2. Comparison 8 Partograph versus labour scale (study

carried out in a low-resource setting), Outcome 2 Oxytocin augmentation.
tograph
Shazly 2017 12/61 38/61 100% 0.32[0.18,0.54]

Total (95% CI) 61 61 100% 0.32[0.18,0.54]

out in a low-resource setting), Outcome 3 Duration of first stage of labour.
Study or subgroup Labour scale WHO partograph Mean Difference Weight Mean Difference
Shazly 2017 55 4.8 (2.5) 55 4.4 (2) 100% 0.44[-0.4,1.28]

Total *** 55 55 100% 0.44[-0.4,1.28]
Favours labour scale -2 -1 0 1 2 Favours partograph

Analysis 8.4. Comparison 8 Partograph versus labour scale (study carried out in
a low-resource setting), Outcome 4 Low Apgar score (less than 7 at 5 minutes).
tograph
Shazly 2017 0/55 0/55 Not estimable


out in a low-resource setting), Outcome 5 Caesarean section (delay).
tograph
Shazly 2017 2/61 10/61 100% 0.2[0.05,0.88]
Informed decisions.

tograph

Total (95% CI) 61 61 100% 0.2[0.05,0.88]

Analysis 8.6. Comparison 8 Partograph versus labour scale (study carried out in a
low-resource setting), Outcome 6 Stillbirth, neonatal death or neonatal morbidity.
tograph
Shazly 2017 0/55 0/55 Not estimable


Analysis 8.7. Comparison 8 Partograph versus labour scale (study carried out in a
low-resource setting), Outcome 7 Birth injuries and PPH (non-prespecified outcome).
tograph
Shazly 2017 3/61 1/61 100% 3[0.32,28.04]

Total (95% CI) 61 61 100% 3[0.32,28.04]

WHAT'S NEW

Date Event Description
31 August 2017 New search has been performed We updated the search and added five new trials (Lee 2015;
Orhue 2013; Rani 2015; Shazly 2017; Sinha 2017). We updated
the Background and Discussion. We incorporated a 'Summary of
findings' table for this update.
Informed decisions.

31 August 2017 New citation required and conclusions We slightly amended the Conclusion, to reflect the fact that other
have changed 'partograph' designs are being explored.

HISTORY
Protocol first published: Issue 3, 2005
Review first published: Issue 4, 2008

17 June 2013 New citation required and conclusions Review updated and corrections made to errors in data for out-
have changed comes 1.5 and 5.4 in previous publication (Lavender 2012). The
corrections do not change the conclusions.
17 June 2013 New search has been performed Search updated in May 2013. No new trials identified. Ajoodani
2011a remains in Studies awaiting classification, awaiting trans-
lation.
14 June 2012 New citation required but conclusions Review updated.

have not changed
14 June 2012 New search has been performed Search updated in May 2012. Three new trials identified. One
has been included (Kenchaveeriah 2011), one has been exclud-
ed (Hamilton 2004), and one is awaiting classification (Ajoodani
2011a).
This review is now comprised of six included studies (involving

7706 women).
12 May 2009 Amended Corrected typographical error in results section.
10 November 2008 Amended Contact details edited.
29 April 2008 Amended Converted to new review format.

CONTRIBUTIONS OF AUTHORS
Tina Lavender, Anna Cuthbert and Rebecca Smyth assessed studies for inclusion independently and extracted all the data. All review
authors interpreted the results individually. Tina Lavender drafted and finalised the text of the review. Anna Cuthbert and Rebecca Smyth
contributed to the content.
All authors contributed to the 2017 update.
DECLARATIONS OF INTEREST
Tina Lavender: was investigator of two trials included in this review (Lavender 1998a; Lavender 2006); therefore, she was not involved with
evaluating these studies. No other relevant conflicts of interest.
Anna Cuthbert: is employed by the University of Liverpool as a Research Associate with Cochrane Pregnancy and Childbirth. Her employ-
ment is supported by the National Institute for Health Research (NIHR), via Cochrane Infrastructure funding to Cochrane Pregnancy and
Childbirth. She had no involvement with the editorial processes for this review update.
Rebecca Smyth: none known.
Informed decisions.

SOURCES OF SUPPORT
Internal sources
• University of Liverpool, UK.
External sources
• (2012 update) National Institute for Health Research, UK.
NIHR Programme of centrally-managed pregnancy and childbirth systematic reviews of priority to the National Health Service (NHS)
and users of the NHS: 10/4001/02
• (2017 update) UNDP-UNFPA-UNICEF-WHO-World Bank Special Programme of Research, Development and Research Training in Human
Reproduction (HRP), Department of Reproductive Health and Research (RHR), World Health Organization, Switzerland.
DIFFERENCES BETWEEN PROTOCOL AND REVIEW
For this update, we assessed trial quality for seven selected outcomes using the GRADE approach (see Summary of findings for the main
comparison).
In a previous version of this review (Lavender 2013), we recorded caesarean section for fetal distress and caesarean section for delay in
labour. We have therefore added these to the outcomes in this update.
We changed the title from 'Effect of partogram use on outcomes for women in spontaneous labour at term' to 'Effect of partograph use on
outcomes for women in spontaneous labour at term'. We felt that this terminology is more widely used.
INDEX TERMS
Medical Subject Headings (MeSH)

*Medical Records; *Pregnancy Outcome; Cesarean Section [statistics & numerical data]; Delivery, Obstetric [methods]; Labor, Obstetric
[*physiology]; Oxytocics [administration & dosage]; Oxytocin [administration & dosage]; Randomized Controlled Trials as Topic; Term
Birth [*physiology]; Time Factors; Uterine Inertia [diagnosis]; Uterine Monitoring [*methods]
MeSH check words

Female; Humans; Pregnancy

Bu Linda 1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bu Linda 1

Uploaded by

Copyright:

Available Formats

Lavender T, Cuthbert A, Smyth RMD

Lavender T, Cuthbert A, Smyth RMD.

Effect of partograph use on outcomes for women in spontaneous labour

Tina Lavender1, Anna Cuthbert2, Rebecca MD Smyth1

Editorial group: Cochrane Pregnancy and Childbirth Group

This review is an update of a review last published in 2013.

Data collection and analysis

Partograph versus no partograph (3 trials, 1813 women)

Partograph with different placement of action lines (4 trials, 5051 women)

Partograph versus labour scale (1 trial, 122 women)

Effect of partograph use on outcomes for women in spontaneous labour at term

What is the issue?

Why is this important?

What evidence did we find?

Partograph versus no partograph (3 studies, 1703 women)

Partograph with different placement of action lines (4 trials, 5051 women)

Partograph versus labour scale (1 trial, 122 women)

What does this mean?

Patient or population: women in spontaneous labour at term

Caesarean section (overall) Study population Average RR 0.77 1813 ⊕⊝⊝⊝

Oxytocin augmentation Study population RR 1.02 1156 ⊕⊕⊕⊝

Cochrane Database of Systematic Reviews

GRADE Working Group grades of evidence

Cochrane Database of Systematic Reviews

BACKGROUND services (Neilson 2003). Although, for some women, acceleration of

Other unit of analysis issues Subgroup analysis and investigation of heterogeneity

five considerations (study limitations, consistency of effect, impre- RESULTS

Figure 2. Study flow diagram.

Funding and conflicts of interest Risk of bias in included studies

Incomplete outcome data Primary outcomes

Duration of first stage of labour Not reported in this comparison.

Caesarean section Primary outcomes

Higgins 2011 Philpott 1972a

Lennox 1995 Review Manager 2014 [Computer program]

Orhue 2012 WHO 2003

Pawson 2004 Yisma 2013

References to other published versions of this review Lavender 2012

Characteristics of included studies [ordered by study ID]

Outcomes • Rate of caesarean section

Notes Declarations of interest: none declared

Funding source: not reported

Bias Authors' judgement Support for judgement

Random sequence genera- Low risk Computer-generated random number table.

Blinding of outcome as- Unclear risk Blinding not reported.

Outcomes • Caesarean section rate

Declarations of interest: not reported

Funding source: study took place at the Liverpool Women's Hospital

Bias Authors' judgement Support for judgement

Random sequence genera- Low risk Table of random numbers.

Allocation concealment Low risk Consecutively numbered, sealed, opaque envelopes.

• Caesarean section rate

Notes Declarations of interest: not reported

Funding source: Liverpool Women's Hospital

Bias Authors' judgement Support for judgement

Allocation concealment Low risk Consecutively numbered, sealed, opaque envelopes.

Blinding of outcome as- Low risk Statistician blind to treatment allocation.

• History of 3 or more consecutive miscarriages

Declarations of interest: the authors report that there is no conflict of interest

Bias Authors' judgement Support for judgement

Other bias Low risk Baseline characteristics are similar.

Outcomes • Caesarean section rate

Declarations of interest: not reported

Funding source: not reported