University of Ghana          http://ugspace.ug.edu.gh
RJ60. G4 G15 
bite C.l 
G379649
University of Ghana          http://ugspace.ug.edu.gh
THE RELATIONSHIP BETW EEN THE USE OF THE PARTOGRAPH  
AND BIRTH OUTCOM ES AT KORLE-BU TEACHING HOSPITAL
THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF 
GHANA, LEGON, IN PARTIAL FULFILM ENT OF THE  
REQUIREMENT FOR THE AWARD OF MASTER OF PHILOSOPHY  
DEGREE IN NURSING
FLORENCE GANS-LARTEY
Co-supervisors:
Mrs Faustina Oware-Gyekye 
Dr Beverley O ’Brien
July, 2007
University of Ghana          http://ugspace.ug.edu.gh
DECLARATION
I, Florence Gans-Lartey, author o f this thesis “The Relationship between the use o f  the 
Partograph and Birth Outcomes” do hereby declare that apart from references to past and 
current literature duly cited in this thesis, the entire research work presented in this thesis was 
done by me as a student o f  the School o f  Nursing, University o f  Ghana, Legon.
This work has never been presented in whole or part for any other degree in this university or 
elsewhere.
cz—W  ■
------------- , ---------------------------
Florence Gans-Lartey 
(STUDENT)
This thesis has been presented for examination with our approval as supervisors.
Mrs Faustina Oware-Gyekye 
(MAJOR SUPERVISOR)
(CO-SUPERVISOR)
University of Ghana          http://ugspace.ug.edu.gh
DEDICATION
This thesis is dedicated to all parents, husbands, children, siblings or friends who have lost 
their daughters, wives, mothers, sisters or friends through pregnancy and childbirth.
University of Ghana          http://ugspace.ug.edu.gh
ACKNOW LEDGEM ENT
Without the support and encouragement o f many people this thesis would not be a reality.
1 am grateful to my supervisors, Mrs. Faustina Owarc-Gyekye and Dr. Beverley O ’Brien for 
their direction, suggestions and patience that enabled this research project to reach its 
conclusion.
I wish to thank committee members for the time spent and suggestions made in reviewing my 
proposal and final thesis.
I would like to thank Mr Annan, the records officer at the Medical Statistics and 
Documentation Department o f the Korle-Bu Teaching Flospital M aternity Block, for the 
timely pulling o f charts. I am also grateful to all the medical records staff who accommodated 
me in coordinating the charts to be reviewed.
I am also indebted to my colleagues, librarians and friends for their support and 
encouragement to complete this project.
Finally. I wish to acknowledge my God, my creator and my source o f  strength.
I V
University of Ghana          http://ugspace.ug.edu.gh
TABLE OF CONTENTS
DECLARATION  ii
DEDICATION  iii
ACKNOW LEDGEM ENT  iv
TABLE OF CONTENTS v
LIST OF TABLES vii
LIST OF FIGURES viii
LIST OF ABBREVIATIONS ix
ABSTRACT x
CHAPTER ONE: 1 INTRODUCTION 1
1.1 Background to the Study 3
1.2 Statement o f the Problem 6
1.3 Purpose o f the Study 8
1.4 Significance o f the Study 9
1.5 Definition o f Terms 9
CHAPTER TWO: 2 REVIEW  OF LITERATURE 11
2.1 Maternal M ortality Worldwide 13
2.2 Maternal Mortality in Ghana 14
2.3 Programmes to Reduce Maternal Mortality 17
2.4 The WHO Partograph as a SMI Intervention 19
2.5 Theoretical Framework 25
University of Ghana          http://ugspace.ug.edu.gh
CHAPTER THREE: 3 RESEARCH PROCESS 27
3.1 Design 30
3.2 Setting 32
3.3 Sample 34
3.4 Procedure 35
3.5 Assumptions 36
3.6 Analysis 36
3.7 Variables 36
3.8 Ethical Considerations 38
CHAPTER FOUR: 4 RESULTS 40
4.1 Demographic Characteristics 41
4.2 Research Hypotheses 42
4.3 Partograph 43
4.4 Maternal Outcomes 46
4.5 Newborn Outcomes 49
CHAPTER FIVE: 5 DISCUSSION OF RESULTS 52
5.1 Summary of Main Findings 52
5.2 Discussion o f Results 52
5.3 Threats to Validity 55
5.4 Limitations 57
VI
University of Ghana          http://ugspace.ug.edu.gh
5.5 Implications to Nursing Practice
5.6 Recommendations
5.7 Conclusion
Appendix A 
Appendix B 
Appendix C 
References
University of Ghana          http://ugspace.ug.edu.gh
LIST OF TABLES
Table 1: Components o f Partograph and levels of omission 31
Table 2: Levels of omission resulting in adequate and inadequate partograph 32
Table 3: Parity 41
Table 4: Mode o f delivery 42
Table 5: Components o f partograph 43
Table 5: Various maternal intervention 46
LIST OF FIGURES
Figure 1: Obstetric admission in Korle-Bu maternity block in May and June 2003 40
v i i i
University of Ghana          http://ugspace.ug.edu.gh
LIST OF ABBREVIATIONS
APGAR - Appearance, Pulse, Grimace, Activity and Respiration o f the newborn
ARM - Artificial rupture o f membranes
C/S - Caesarean section
F/D - Forceps delivery
FHR - Foetal heart rate
LGA - Large for gestational age
MMR - Maternal mortality rate
N/A Non-applicable
NICU - Neonatal intensive care unit
NSVB - Normal spontaneous vaginal birth
PPH Postpartum haemorrhage
SGA - Small for gestational age
SMI - Safe motherhood initiative
SMP - Safe motherhood programme
UTI - Urinary tract infection
V/E Vacuum extraction
ix
University of Ghana          http://ugspace.ug.edu.gh
ABSTRACT
The purpose o f this study was to examine the relationship between the use o f the 
partograph and maternal interventions as well as maternal and neonatal outcomes. A 
retrospective chart review o f 1,845 primiparous and multiparous deliveries which met the 
inclusion criteria was conducted. Using pre-established criteria only 472 (25.6%) charts were 
adequately completed and on inspection some o f these appeared to be completed 
retrospectively. W hether or not timely action was taken if  the “action line” was crossed on the 
partograph was also assessed. Chi-square analysis showed that when the partograph was 
adequately used, there were more interventions to augment labour. There were differences in 
maternal outcomes with women being less likely to lose more than 500 millilitres o f blood if  
partographs were adequately used. Although there were 13 intra partum deaths only one chart 
could be retrieved. While there appeared to be less pyrexia with the group whose partograph 
was inadequate, it is possible that this group also had their temperature monitored less 
frequently. Decreases in newborn outcomes including Apgar scores, birth injuries, NICU 
admissions and mortality were not found.
Recommendations were made to facilitate accurate documentation and completion o f 
the partograph to help in taking prompt decisions or interventions. Periodic in-service training 
for record officers in health facilities is recommended to improve filing system to ensure easy 
retrieval o f patients charts. Recommendations to shorten the length o f time that action is taken 
when the action line was crossed.
X
University of Ghana          http://ugspace.ug.edu.gh
CHAPTER ONE 
IN TR O D U C TIO N
Pregnancy and childbirth involve many biological, psychological and social changes 
that require a great deal o f  adjustment and adaptation on the part o f the mother. Pregnancy is 
considered a time o f crisis or transition, although the changes are manifested more clearly in a 
w om an's life when the baby is born (Barclay, Everitt and Rogan, 1997; Mercer, 1986; Rubin, 
1984). It is a time when w om en’s health can be placed at risk by a host o f factors. For 
example, they may be anxious and apprehensive about the outcome o f the pregnancy itself or 
about the process of labour and delivery. They may also be apprehensive about being a 
mother. However, the outcome of pregnancy, for both mother and baby, is associated with the 
health of the mother and the quality of antenatal and intrapartum care. It is affirmed in several 
reports from higher income countries that women with uncomplicated pregnancies may be 
cared for by midwives, and taken through labour and delivery at home or in birth centres with 
no increased risk o f adverse maternal or perinatal outcomes when compared to similarly low- 
risk pregnant women who deliver in hospital under the care o f physicians (Ackermann- 
Liebrich, Voegeli and Gunter-Witt, 1996; David, Kraker von Schwarzenfeld, Dimer and 
Kentenich. 1999; Rooks, Weatherby, Ernst, Stapleton. Rosen and Rosenfield, 1989). The 
achievement o f these comparable outcomes, however, can be attributed to early recognition of 
complications and prompt intervention or referral and transfer to hospitals for further care. It 
may also be attributed to the health status o f the pregnant women (Nkyekyer, 2000).
University of Ghana          http://ugspace.ug.edu.gh
Labour is seldom a problem if  it goes at its proper pace. Most complications start when 
labour is delayed or obstructed. To manage such complications it is imperative for the 
provider to identify the occurrence o f the problem as early as possible through effective 
methods o f monitoring labour. One such method is the partograph. The introduction and use 
o f the partograph in labour wards was to help prevent obstructed labour and uterine rupture. 
These conditions cause 70 percent o f maternal deaths in lower income countries. The 
partograph can be used to monitor all labours in hospital in order to avoid unnecessary 
interventions so that maternal and neonatal morbidity are not needlessly increased. 
Furthermore, an appropriate use o f the partograph will assist the provider to intervene in a 
timely manner so that maternal and neonatal morbidity and mortality can be avoided 
(Maternal and Neonatal Health. 2003).
The normal pattern o f labour, including the latent and active phases o f cervical 
dilatation, was first documented on a graph in the 1950s. In the 1960s and 1970s. further 
research led to improved tools to chart the progression o f labour and to build a scientific basis 
for interventions to prevent prolonged labour. Friedman (1975) described a graphic analysis of 
labour, correlating the duration of labour with the rate o f cervical dilatation. Philpott and 
Castle (1972; 1972b) further developed this tool and added a line to alert providers of 
prolonged labours as well as to record other intrapartum details. The partograph provided 
health professionals with a continuous pictorial overview o f labour and allowed early 
identification and diagnosis o f pathological labour. These early partographs formed the 
foundation for the WHO model of the partograph, which was developed as an international 
standard in 1988 following the launch of the worldwide safe motherhood initiative.
2
University of Ghana          http://ugspace.ug.edu.gh
1.1 Background to the Study
Many people believed that the dawn o f the 21st century would be associated with the 
disappearance o f maternal mortality. Unfortunately, that is not the case, particularly in low- 
income countries including Ghana. On the Africa continent, one in 16 childbearing women 
will not survive because o f a pregnancy related complication (Gerlus, Moran and Segu, 1999). 
Sub-Saharan African women are 175 times more likely to die from complications in 
pregnancy and childbirth compared to those in higher income countries. Most maternal deaths 
and disabilities result from delays in recognising complications, reaching a medical centre or 
receiving quality care (CBC News, 2003).
Although, through the safe motherhood programme, simple interventions were adopted 
to improve wom en’s health and reduce maternal mortality and morbidity, maternal deaths still 
remain unacceptably high. Maternal mortality is defined as “death o f  women while pregnant 
or within six weeks o f the termination of a pregnancy, irrespective of the duration and site of 
the pregnancy, from any cause related to or aggravated by the pregnancy or its management 
but not from accident or incidental causes’1 (WHO, 1997). The World Health Organisation 
(WHO) and United Nations International Children’s Emergency Fund (UNICEF) estimate that 
each year 585,000 women die from causes related to pregnancy and childbirth.
Every day at least 1,600 women die from complications associated with pregnancy and 
childbirth. Almost 90 percent of these deaths occur in Sub-Sahara Africa and Asia. (Maine. 
Akalin, Ward and Kamara, 1997). In Ghana, for example, although the maternal mortality rate 
has been officially reported to be 214 per 100.000 live births (Ghana Demographic Health 
Survey (GDHS), 2003), figures ol'734.4 and 1140 per 100,000 deliveries have been reported
University of Ghana          http://ugspace.ug.edu.gh
from two teaching hospitals (Lassey and Wilson, 1998; Martey, Djan, Twum, Browne and 
Opoku, 1998). Among the national health objectives, health leaders in Ghana stated a goal of 
reducing the rate of maternal death to 100 per 100,000 live births by the year 2001 (Annual 
Report. Reproductive and Child Health (RCH). 1999). but the institutional maternal mortality 
ratio (MMR) increased to 260 per 100,000 live births in 2001 despite efforts to reduce it 
(Annual Report. RCH, 2001). The MMR target o f 180 per 100,000 live births for the year 
2003 was not achieved. Eight hundred and fifty four institutional maternal deaths were 
recorded in 2003. representing a MMR of 205 per 100,000 live births, higher than the 2002 
ratio o f 204 per 100,000 (Annual Report, RCH, 2003).
Causes o f maternal deaths are similar all around the world. About 80 percent o f these 
deaths have direct causes resulting from obstetric complications o f pregnancy, labour and 
puerperium. Deaths may also occur from interventions, omissions, incorrect treatment or a 
chain o f events resulting from any o f these complications (Bennett & Brown, 1999). In 
Ghana, the direct or major causes o f maternal death are infection/sepsis, 
haemorrhage/bleeding, hypertensive disorders of pregnancy, obstructed labour and abortion 
complications. These direct causes result in about 80 percent o f all maternal deaths. Other 
direct causes include ectopic and molar pregnancies as well as complications associated with 
anaesthesia. The indirect causes include anaemia, malaria and heart disease. These conditions 
affect the health status o f all women who are o f childbearing and are aggravated by 
pregnancy, (WHO, 1997). Indirect causes o f death may arise as a result of previous existing 
disease or disease that developed during pregnancy and was aggravated by pregnancy. Other 
indirect causes are malnutrition, sickle cell diseases, infectious diseases, diabetes mellitus and
i
University of Ghana          http://ugspace.ug.edu.gh
hypertension (Network Family Health International, 1999). Most o f the causes o f maternal 
mortality and morbidity can be prevented with improved health status and quality o f care 
given by health care workers, as well as by access to emergency interventions for most o f the 
pregnant women in hospital and in their communities (Starrs, 1998). The issue o f  maternal 
mortality has raised a lot o f concern among governments and health workers, particularly 
those in low-income countries. The concern stems from the fact that proper care during 
pregnancy, delivery and puerperium may avert many of these deaths. There is evidence that 
adequate care during delivery and the puerperium alone can decrease the number o f maternal 
deaths by 50 to 80 percent (Twum-Baah. Nyarko, Quarshie, Caiquo and Amuah, 1994). In 
adopting the safe motherhood programme, policy makers in many countries, including Ghana, 
hoped to improve w om en’s health in general and especially, to reduce maternal and newborn 
morbidity and mortality (RCH Annual Report 2001). The “safe motherhood initiative” 
includes a series o f interventions. One of these is the use o f the partograph.
The partograph is a tool that when used appropriately should result in timely 
management o f prolonged labour. Labour and delivery is considered a critical time for the 
mother and interventions at this time may be essential to reduce maternal morbidity and 
mortality. Even with adequate screening during the antenatal period, the intrapartum period 
can be complicated, requiring emergency intervention. Delivery must be safe, atraumatic and 
clean if maternal mortality is to be reduced. As much as possible delays, such as delay in 
deciding to seek care, or in reaching an appropriate care facility because o f difficulties with 
transportation and referral system, or in receiving care after arriving at a health facility, should 
be reduced (Safe Motherhood Newsletter, 1995).
5
University of Ghana          http://ugspace.ug.edu.gh
The WHO partograph is an instrument used to improve maternal health care through 
early detection o f complications during labour by personnel functioning with or without the 
support o f an obstetrician (WHO/MCH. 88.3, 1988). However, more emphasis is being placed 
on midwifery notes rather than partographic documentation to the extent that the partograph is 
often being completed in retrospectively to correspond with midwifery notes (Groeschel and 
Glover, 1998). It is therefore not clear whether the partograph is being used adequately or 
inadequately. The question then is “What is the relationship between the appropriate use and 
the outcomes o f birth?”
1.2 Statement of the Problem
A parameter to describe the level of care provided by obstetric services is the maternal 
mortality ratio (MMR) or the number of maternal deaths per 100,000 deliveries). The MMR 
may be the gold standard for obstetric epidemiology but less than 30 percent o f  countries 
provide reliable maternal mortality data and maternal deaths are frequently underreported in 
both higher and lower income countries (Pittrof, 1997). Maternal mortality continues to be a 
major problem in sub-Sahara Africa. It is estimated to be about 10 per 1000 live births or 
higher in some rural communities in sub-Sahara Africa and these rates are among the highest 
in the world. Complications during delivery directly cause about three-quarters o f all maternal 
deaths. The remaining deaths are the result o f medical conditions such as viral hepatitis, 
anaemia and cardiovascular disease that are aggravated by pregnancy (Herz and Measham. 
1987).
Ghanaians, mostly women and children, die each year from preventable causes, and 
the official maternal mortality rate was reported as 260 per 100,000 live births (MOH, Ghana.
6
University of Ghana          http://ugspace.ug.edu.gh
1990 and RCH Annual Report, 2002). This trend is unacceptable and calls for all concerned 
with issues related to women and children to develop strategies or interventions to prevent or 
reduce this horrific toll o f maternal deaths. One such intervention is the introduction o f the 
WHO partograph in the labour wards as a tool for early identification and diagnosis of 
abnormal progress o f labour. This should lead to timely interventions. The partograph is a 
simple chart that is used to guide the health care provider in recording the progress o f labour 
as well as monitor the health status o f both the foetus and mother. It is an inexpensive, low 
technology tool, which requires basic, though vital, midwifery skills to save lives (WHO, 
1994).
The partograph consists o f three components: the foetal condition, the progress of 
labour and the maternal condition. It can be used for all labours in hospitals and health centers 
and at the domiciliary level where trained nurse/midwives practice midwifery. The purpose o f 
the partograph is to reduce maternal and perinatal mortality and morbidity worldwide 
(Preventing Prolonged labour: A Practice Guide. The Partograph Part 1. 
WHO/FHE/MSM/93.8, 1993). As soon as a woman enters labour, the chart is used to record 
the progress of her labour, in particular the rate of cervical dilatation and the pattern o f uterine 
contractions. The partograph helps providers ensure that women are carefully monitored 
during labour, unnecessary interventions are avoided and complications recognised and 
responded to in a timely manner. It gives providers objective data on which to base their 
clinical decisions and enhances communication among members o f the team o f providers who 
are caring for the woman in labour.
7
University of Ghana          http://ugspace.ug.edu.gh
According to the WHO (1994), the partograph when used appropriately has the 
potential not only to reduce foetal and maternal mortality and morbidity but also to reduce the 
rising tide o f caesarean deliveries. In projects in Indonesia, Uganda, Nigeria and Ghana where 
the use o f the partograph in community obstetrics was examined, it was revealed that its use 
was successful in urban community settings. Yet prolonged and difficult labour contributes to 
at least a third o f all maternal deaths (Lennox and Kwast, 1995).
The most problematic time in pregnancy is the intrapartum period. The Safe 
Motherhood Initiative introduced strategies to address this problem. An important strategy is 
the use o f the WHO partograph, which is intended to be used to prevent prolonged labour. 
Important questions are: 1) ‘i s  the partograph being used properly?” and 2) “When properly 
used, does it reduce the M M R?”
The specific research questions were:
1. W'hat is the relationship between outcomes o f labour and appropriate use o f the 
partograph.
2. Do health care professionals take action in a timely manner when the action line 
is crossed.
1.3 Purpose of the Study
The partograph has been shown to be an efficacious tool for monitoring labour and 
identifying women in need o f obstetric intervention. Partographs have been introduced in 
many countries and the introduction is supported by guidelines and training workshops but 
often with little follow-up and supervision. Moreover, there are concerns that the use of the
8
University of Ghana          http://ugspace.ug.edu.gh
partograph may increase the rate o f unnecessary obstetric interventions. The purpose of this 
study is to explore the relationship between the use of the partograph and birth outcomes.
1.4 Significance of the Study
Health care providers/midwives will gain more understanding in the use o f partograph 
and this will increase the observation and interpretation skills related to the progress of labour. 
Health workers can only use the partograph effectively if  they have adequate training. The 
results o f the study will help identify the extent to which the partograph is used, determine 
whether it is used correctly, identify the relationship between correct use and outcomes, and 
identify areas where education or training needs to be strengthened. Utilisation o f the findings 
will inform policy makers who make decisions about training programmes o f personnel needs 
that will allow them to reach the millions o f women who are delivered with no attendant or 
with the assistance only o f a relative or other untrained person. Findings will guide further 
research in the area o f safe(r) motherhood in labour and delivery, with the view o f improving 
maternity care and reducing maternal morbidity and mortality.
1.5 Definition of Terms
Maternal mortality - is defined as the death o f a woman during childbirth and/or within 2 
weeks after delivery.
Maternal morbidity -  is any sign or symptom resulting from or made worse by pregnancy or 
childbirth. In this study, maternal morbidity is defined as assisted birth, pyrexia and 
haemorrhage of 5()0mls or more.
Neonatal morbidity - any complication to the newborn resulting from childbirth. In this 
study, neonatal morbidity will be reflected in 1 and 5 minutes APGAR scores of 6 or less.
9
University of Ghana          http://ugspace.ug.edu.gh
cephal haematoma or other birth injury and admission to the Neonatal Intensive Care Unit 
(NICU).
Neonatal mortality -  death o f a newborn during birthing and/or within 2 weeks after birth 
Appropriate or adequate use of partograph -  correctly filled out partograph with no 
omission or minor omissions. Adequacy is determined by guides established by the WHO.
10
University of Ghana          http://ugspace.ug.edu.gh
C H A P TE R  TW O
REVIEW OF LITERATURE
I'he purpose o f the review o f the literature is to summarise and critically analyse 
existing theoretical and research literature relating to maternal mortality and morbidity, the 
Safe Motherhood Initiative and the partograph as a specific intervention for reducing maternal 
mortality. Using electronic searches such as Databases (ERIC, EMBASE, CINAHL, Pubmed, 
Medline). Internet search engines such as Google and abstracts from dissertations, a wide 
range o f articles relevant to the proposed research question were found. References were cited 
from 1987 to 2003. However, studies published earlier than 1987 but found to be relevant to 
the proposed research study were included. Key words used for the search included safe 
motherhood, WHO partograph, maternal mortality and morbidity, quality maternal services, 
programme evaluation and pregnancy outcome.
The magnitude o f maternal mortality will be evaluated. Strategies to reduce rates, 
particularly in low-income countries, will be discussed and an emphasis will be placed on the 
WHO partograph as an intervention. Every minute a woman dies of causes related to 
pregnancy or childbirth (Gerlus, Moran and Segu, 1999). She is most likely to be young, 
already a mother, and living in a low-income country or within the disadvantaged minority 
groups of relatively high-income nations. For each woman who dies, an estimated 100 women 
survive childbearing but are afflicted by disease, disability or physical damage caused by 
pregnancy-related complications (KobIinsky, 1993). In the late 1980s maternal deaths were 
often described as "silent’, because so little had been heard o f the women who died, the 
circumstances in which they died, and what may have contributed to their deaths. Because it
11
University of Ghana          http://ugspace.ug.edu.gh
was 'silent' it was not known how many maternal deaths occurred in low-income countries. 
“Unheard, unseen, uncounted, unattended, they died" (Peters, 2000 pp 2-7). This is no longer 
so. The silence has been cracked, if not broken, by the gaining o f  a much better understanding 
of why these women die, how many, and in what circumstances.
All pregnant women are at risk of obstetric complications. Most life-threatening 
complications occur during labour and delivery, and these cannot all be predicted. Prenatal 
screening does not identify all o f the women who will develop complications (Rooks, 
W inikoff and Bruce, 1990). Women not identified as ‘high-risk' can and do develop obstetric 
complications. Most obstetric complications occur among women with no obvious risk 
factors. Similarly, evidence that obstetric complications may occur despite intensive prenatal 
care and screening was illustrated in a study of deliveries in out-of-hospital birth centres in the 
United States (Weatherby, 1990). One of every 13 women categorized as ‘low-risk’ and 
receiving intensive prenatal care developed serious complications. Additionally, maternal 
deaths, with all the attendant personal loss a family and nation experience, is only a tip o f the 
iceberg because the true incidence o f maternal mortality and morbidity is extremely difficult 
to measure, particularly in low-income countries, and has a wide margin o f error. Implications 
for economic productivity for the nation, health o f the family and loss o f personal fulfillment 
of individual women are incalculable. The difficulty o f measuring maternal mortality has long 
proved to be an impediment to progress in alerting health planners and others to the magnitude 
and causes of this problem and hence to effective interventions on an appropriate scale (WHO 
Maternal Mortality Rates, 1986).
12
University of Ghana          http://ugspace.ug.edu.gh
2.1 Maternal Mortality W orldwide
Profound differences still exist in maternal mortality ratios for women depending on 
where they live in the world. Maternal deaths range from less than 10 per 100.000 live births 
in Northern European countries to more than 1.000 per 100,000 live births in some low 
income countries (WHO, 1992). Women from Southern Asia and Africa are at highest risk. A 
community-based study in Bangladesh showed ratios o f approximately 600 maternal deaths 
per 100,000 live births. In a rural India a study in Andhra Pradesh, a ratio o f  874 per 100,000 
was reported (WHO. Maternal Mortality Rates. 1986).
Northern and Southern Africa report lower mortality rates than the rest of Africa. 
However, the very high fertility rate (an average o f eight live births per woman plus foetal 
deaths) yields a lifetime risk o f greater than five percent that a woman will die from a 
pregnancy related death (WHO. Maternal Mortality Rates, 1986). In rural Africa, ratios of
1,000 per 100.000 live births have been recorded in several studies and in urban studies ratios 
of over 500 per 100,000 live births with an overall ratio o f 640 per 100.000. Thus Africa, 
while contributing eighteen percent of the world’s live births, contributes thirty percent o f the 
world’s maternal mortality. On the other hand, the higher income nations contribute fourteen 
percent o f the world’s live births, but only one percent of the maternal mortality (WHO, 
Maternal Mortality Rates, 1986).
Latin American mortality rates vary markedly between temperate and tropical areas. It 
is estimated that in tropical South America mortality ratios are an average o f 270 per 100.000 
live births, with temperate climate ratios somewhat lower. Fifty percent o f the Latin American 
maternal deaths are secondary to complications from illegal abortions. Latin American women
13
University of Ghana          http://ugspace.ug.edu.gh
die in unacceptably high numbers due to lack of access to safe abortions. In contrast to 
Southern Asia, Eastern Asia enjoys quite low maternal mortality ratios. China reports rural 
ratios o f 59 maternal deaths per 100.000 live births and urban ratios o f 25. Widespread access 
to prenatal care and safe abortion are two important contributors to these relatively low ratios. 
However, underreporting in China may account for the reduced rates. USSR maternal 
mortality data were not available.
Maternal mortality ratios in Northern and Western Europe, Canada, the United States, 
Japan, Australia and New Zealand are all in the range o f 10 per 100.000 or lower. Canada in 
the year 2000, recorded that 20 women died giving birth, for a rate o f one in 8,700. In the 
United States, there were 660 deaths during the same year, for a rate of one in 2,500. Britain 
recorded 85 deaths, a risk o f one in 3,300. These deaths include incidental deaths which are 
frequently related to low income and this may account for the slightly higher rales in Britain. 
In Europe, North America and Australia, the average is one in 2,800. However, the World 
Health Organisation, the Children’s Fund and the Population Fund gave no reason for the 
higher number o f deaths in the United States or the differences among higher income 
countries (CBC News. 2003). The Nordic countries are the lowest with ratios o f  2 to 6 per 
100,000. Romania in Eastern Europe has a ratio out o f line with the rest o f the region, i.e., 
more than 30, largely due to abortion related deaths (WHO, Maternal Mortality Rates, 1986).
2.2 Maternal Mortality in Ghana
What is known about maternal mortality in Ghana is fragmented. Ampofo reviewed the 
causes of maternal death from 1963 to 67 at Korle-Bu Maternity Hospital in Accra. Korle-Bu 
is one of the two teaching hospitals in Ghana and a major referral point for patients within the
14
University of Ghana          http://ugspace.ug.edu.gh
Greater Accra Region. It is also a referral centre for other parts o f the country and 
neighbouring West African countries. During the five-year period reviewed by Ampofo, 238 
maternal deaths were recorded for 35,193 births, giving a ratio o f 676 per 100,000 births. Of 
the 238 deaths, 63 were attributed to haemorrhage, retained placenta and hypofibrinoginaemia. 
Sixty-one deaths were attributed to obstructed labour, including 33 who suffered from 
ruptured uteri. Twenty-four died o f puerperal sepsis. Ghana has the additional infection 
burden caused by unregistered patients self-treating with herbs and various medicinal potions 
applied prior to admission to the vaginal canal in the form o f pessary (Ampofo, 1969).
Seven deaths were attributed to severe anaemia o f pregnancy and 23 deaths were 
attributed to sickle cell anaemia. Due to the difficulty in obtaining laboratory confirmation, 
these diagnoses were made clinically or from autopsy findings. Twenty-seven o f  the deaths 
were due to ectopic pregnancy, four o f which were advanced extra-uterine pregnancies.
Within the study period 952 ectopic pregnancies were diagnosed with a mortality rate o f about 
3 percent. Abortion accounted for only 33 deaths in this study. Abortion, though a common 
practice, was illegal in Ghana during that period. Treatment o f complications o f abortion w'as 
the most common condition seen during the study period and this could be due to sepsis 
(Ampofo. 1969).
Nine women died from trophoblastic disease, two from hydatidiform mole and seven 
from choriocarcinoma. Though it was noted that these women ran a rapid course to death, it is 
clear that these deaths were within 42 days of delivery. Thirty-five deaths were due to 
eclampsia; 30 of these women died prior to delivery. O f the indirect obstetrical deaths, liver 
disease was the leading cause of death claiming 3 1 women (Ampofo, 1969).
15
University of Ghana          http://ugspace.ug.edu.gh
Korle-Bu teaching hospital was the site for a broader maternal and child health study 
conducted by Family Health International, North Carolina. This Maternity Care Monitoring 
Project was a multi-site study, which eventually included deliveries of more than half a 
million women in 64 countries. Information was collected on a uniform data collection sheet, 
which ran one page per mother-child unit. Information was collected at Korle-Bu teaching 
hospital on 4.990 deliveries occurring between August 1981 and August 1982. Eleven deaths 
were recorded from this study group o f women for a ratio o f 230 per 100,000 live births. Only 
in-hospital deaths were recorded. Any of these women who died after discharge but before 42 
days postpartum were lost to follow up and not included; therefore the number o f maternal 
deaths was underestimated. Although an incomplete ratio, this represents a tremendous change 
from the 1.080 deaths per 100,000 live births noted in the previous study. Within the study 
group, 5.2 percent of the women delivered by caesarean section and 12.7 percent o f the births 
were attended by a physician (Arruda, Covington, Janowitz, Morris and Rodriques, 1985).
The maternal mortality rate remains high in Ghana with estimates ranging from 214 
(Ghana Statistical Survey (GSS), 1994) to 740 deaths (WHO/UNICEF, 1996) per 100.000 live 
births. Institutional data estimated the Maternal Mortality Ratio (MMR) in 2001 as 
260/100,000 live births. Central and Western regions continued to record the highest rates 
with 450 deaths per 100,000 live births, fhe Greater Accra region registered the lowest MMR 
with 150/100,000 live births (RCH Annual Report, 2001). Other health issues o f specific 
concern to women involve a very low level o f knowledge among women about causes, 
prevention, risk factors and signs of gynaecological malignancies such as cancer o f the cervix.
16
University of Ghana          http://ugspace.ug.edu.gh
Cervical cancer deaths are high at 57.8 percent, followed by ovarian cancer at 25.3 percent of 
all female genital tract malignancies (W omen's Health Foundation’s Brochure, 2003).
Global experience suggests that the majority of pregnancy-related deaths are 
preventable, and a significant body o f research on strategies for reducing maternal mortality 
has been generated but there appears to be no decrease in the maternal mortality rate since 
1975. Specific attention to wom en’s health is thus warranted and in fact well overdue.
2.3 Programmes to Reduce Maternal Mortality
Since maternal mortality was known to be an important public health problem, there 
have been efforts to reduce it. though some strategies have proved to be more effective than 
others. To ensure safer motherhood, numerous organisations and groups have introduced 
several mechanisms or strategies to help prevent or reduce maternal mortality. Examples of 
such organisations include WHO, UNICEF, UNFPA, USAID, JHPIEGO, the World Bank, the 
Population Council, IPPF and the Carnegie Corporation of New York. However, efforts to 
increase safe motherhood vary among countries depending on the resources available and the 
social and cultural environment in which women live. The Safe M otherhood Initiative was 
launched in 1987 by international agencies and governments to raise global awareness about 
the impact of maternal mortality and morbidity, and to find solutions.
Authors o f safe motherhood programmes assume that maternal health services can 
reduce maternal mortality (Campbell, Heichelheim and Koblinsky. 1999; National Research 
Council, 1997; Tinker and Koblinsky, 1993). The assumption that services make a difference 
is supported by historical data on the timing o f reduction in maternal mortality, which suggests 
that such declines have not coincided with socioeconomic gains but with improvements in
17
University of Ghana          http://ugspace.ug.edu.gh
specific services (Loudon, 1991; Papiernik, 1995). Data supporting the effect o f services is 
coming from case-control and experimental studies. For instance, Midhet, Becker and 
Berender (1998) linked lower maternal mortality in rural Pakistan to health facility staffing 
and the availability o f  essential obstetric care. Frankerberg and Thomas (2001) linked the 
addition of village midwives in Indonesia, if not to decreases in maternal mortality, at least to 
an increase in body mass among women o f reproductive age.
An important question is which maternal health services reduce maternal mortality. 
Shiffman (2000) indicates that, in cross-national regressions, having trained attendants at 
delivery is strongly associated with lower rates o f maternal mortality. Sloan, W inikoff and 
Fikree (2001) counter that the case for any intervention is unprovable from such data, because 
of strong interdependence among possible determinants o f maternal mortality. Such ‘single 
solutions.’ they argue, do not take into account o f interdependence among services. The Inter­
Agency Group for Sale Motherhood (1997) reports that ensuring that a trained attendant is 
present at all deliveries has now become the cornerstone o f safe motherhood programmes. 
However, having trained attendants at birth or having supervised delivery may not be enough 
to reduce maternal deaths. All birth attendants, including midwives, must possess the requisite 
skills and competencies, and be knowledgeable in the use o f effective technologies that 
monitor mother and foetus during pregnancy and labour, so as to make pregnancy and birthing 
safer and less painful for the mother (Smeenk and Ten-Have, 2003).
In lurther research to improve maternal health services, the focus was on indicators 
such as socioeconomic development. Hertz, Hebert and Landon (1994), for instance, used the 
availability of medical resources as an indicator, and found that maternal mortality across
18
University of Ghana          http://ugspace.ug.edu.gh
countries was associated with total energy consumption and excess energy consumed as fat. 
Other authors explored a variety of socioeconomic variables relating to development and 
women’s status, and similarly supported the hypothesis that maternal mortality levels were 
dependent on some of them (Shen and Williamson, 1999).
2.4 The WHO partograph as a SMI Intervention
As part of its contribution to the Safe M otherhood Initiative (SMI) which was launched 
in 1987. the WHO introduced and promoted the use o f the partograph with a view to 
monitoring labour and introducing quick and effective interventions. The magnitude of 
maternal/inlant morbidity and mortality to the family, community and the nation cannot be 
over emphasized. A mother ’s death greatly influences the health and livelihood o f her 
surviving children. When a mother dies, her surviving children are three to ten times more 
likely to die within two years than those with both parents living (Tinker, 1997). In addition, 
surviving children often do not receive adequate health care and education as they grow up. 
The death of a mother has an impact beyond that o f her immediate family. A productive 
worker, one who rears and guides the next generation, cares for the elderly, and contributes 
stability to the community is lost. As such, the SMI had among its objectives to reduce 
maternal mortality by 50 percent by the year 2000, and included supervised delivery as one of 
its interventions to achieve this objective. Supervised delivery is defined as the percentage of 
deliveries attended by skilled health personnel irrespective o f the outcome. This is one of the 
important components of the safe motherhood programme and also an important aspect in 
midwifery practice.
19
University of Ghana          http://ugspace.ug.edu.gh
Pregnancy and childbirth have the potential for being intimate and emotionally charged 
experiences for all involved. Whether the experience is positive or negative, midwives are 
involved in providing support for women and their families, and in dealing with the feelings 
that arise. Midwives everywhere arc concerned at the loss o f a wom an’s life at or near the time 
o f giving birth and would therefore use every skill they possess and all their expertise to 
facilitate the safe delivery o f both mother and infant. All women seem to develop expectations 
of childbirth and the kinds o f expectations vary among women, as does how realistic they are. 
The pain women expect to feel can provoke anxiety and fear, because the quality and duration 
of the potential pain are unknown (Beaton and Gupton 1990). Women also express fears of 
loss o f control, abandonment, self-injury, or injury to the baby (Green, 1993; Lowe, 1996; 
Sjogren. 1997).
W omen’s satisfaction with maternity services primarily depends on the quality of 
communication between the women and their caregivers. This enables these women to 
experience an enhanced sense o f recognition and feel more involved in their care as they build 
a relationship with a "known midwife” (McCourt and Pierce, 2000). Tarkka and Paunonen 
(1996) also reiterated that the midwife was the main source o f emotional support for mothers 
during labour and that the mothers attached much importance to the presence o f the midwives, 
for the encouragement they offered, and for their individual treatment. The mothers felt safer 
if they had their own midwife with them throughout childbirth.
lhe  partograph is an instrument used to improve maternal health care through early 
detection by any trained birth attendant o f complications during labour. The progress of 
labour, especially the rate of cervical dilatation, is recorded in the partograph. It helps to detect
20
University of Ghana          http://ugspace.ug.edu.gh
abnormal progress in labour so that appropriate interventions can be taken which may save the 
lives o f mother(s) and infant(s). The partograph has been in use in many countries for over 
25years but adoption in lower income countries has been slower. Research on the use o f this 
partograph has taken place mainly in hospital settings. WHO undertook the largest and best- 
known study in South East Asia through the SMI, in which 35,000 partographs from eight 
hospitals were surveyed (WHO/FHE/MSM/94.4, 1994). Statistically significant differences 
were found in numbers o f prolonged labours, caesarean sections and labours requiring 
oxytocin augmentation when partographs were used.
There was also a marginal improvement in newborn outcomes as measured by 
intrapartum foetal deaths, Apgar scores, resuscitation measures and admission to neonatal 
special care facilities. However, the partograph played a minimal role in 47 cases o f maternal 
mortality and 55 cases o f uterine rupture, as most o f these occurred as a result o f delayed 
presentation of complications. In Ghana, specifically in Accra, the WHO partograph was 
introduced in the labour wards in 1990 and by the end of 1991 it was used in almost all the 
regional and district hospitals. The partograph was intended to be used routinely for ever)' 
delivery attended by midwives and other health care practitioners from all levels o f health 
care.
There is increasing awareness of the extent of maternal ill health, and more knowledge 
about interventions that can be effective in reducing this burden (Ronsmans et al. 2001). Yet 
there has not been enough progress. It is an international disgrace that mothers in poor 
countries, at the beginning o f the 21sl century, should be experiencing unimaginable suffering
21
University of Ghana          http://ugspace.ug.edu.gh
and death, due to a scandalous lack of effective care during pregnancy and childbirth, which is 
resulting in approximately 600,000 women dying annually (Walley, 2002).
Unlike higher income countries, birth for most women in Ghana is quite hazardous, 
especially in the rural areas where most women give birth in their homes or rural clinics, far 
from emergency maternity care. The majority o f the women are delivered by traditional birth 
attendants (TBAs) or village women who have minimal or no formal training (Matthews, 
1999). The introduction of the use o f the partograph in maternity care is to help reduce 
prolonged labour and the sequelae of maternal morbidity and mortality for both women and 
infants, to improve the quality o f  care for women in labour and to increase the observation and 
interpretation skills o f all birth attendants including midwives and other health professionals.
Nakkaki (1998) audited 1000 partographs for patients in Nsambya Hospital, Uganda, 
between June 1998 and June 1999. She looked at the use o f  the partograph as an indicator for 
the quality o f care provided in the hospital. She wanted to know whether partographs were 
being used in the labour wards and whether midwives and doctors were using them correctly. 
The results revealed that hospital staff used the partograph for the majority o f labours and that 
most of the partographs were completed correctly. However, a few partographs had errors, 
which might have contributed to poor labour outcomes. Trainees, new doctors and midwives 
who had never used a partograph outside the classroom committed these errors.
A prospective population-based study conducted among 20,326 pregnant women in 
seven primarily urban areas was used to ascertain whether clinical risk factors for late stillbirth 
in West Africa can be detected during antenatal care or only during labour. The investigators 
concluded that the principal risk factor for late stillbirth could be detected only in the late
22
University of Ghana          http://ugspace.ug.edu.gh
antenatal and intrapartum period. The potential benefits o f partograph use are highlighted in 
these findings (Bouvier-Colle, Breart and Chalumeau and the MOMA Group, 2001). Seffah 
(2002) conducted a study to find out whether the introduction o f the WHO partograph to the 
labour wards in Accra, Ghana in 1990 had reduced the incidence o f  ruptured uterus, maternal 
mortality and perinatal mortality. There were 136 cases o f ruptured uterus diagnosed for 
71.067 deliveries between January 1996 and December 2001 (incidence 1 in 523). In 1971 
when the partograph had not been introduced in Accra, the incidence was 1 in 257. Other 
variables such as improved technology and increased surveillance during labour may have 
influenced some of these maternal outcomes.
The quality o f monitoring during labour for routine maternity care after partograph 
introduction in three hospitals in southern Tanzania (two governments, one mission) was 
assessed. In each hospital, all (62; 65; 69; totalling 196) deliveries were observed and 
partographs for each woman were analysed during a 4-week period (Bosse, Massaw'e and 
Jahn. 2002). They concluded that the partograph helped to improve delivery' outcomes. 
However, its implementation needs continuous reinforcement and quality assurance. Concerns 
that the partograph leads to higher rates of caesarean section were not supported by the data. 
The sample size may be too small to make any definitive statement. The use o f the partograph 
in monitoring labour in a woman who had a caesarean section was studied at Korle-Bu 
Teaching Hospital in 1991 1992 (Amaniampong, Seffah, and Wilson, 1994). They analysed 
the case notes o f 15,347 deliveries and confirmed that prolonged labour can be prevented 
using the partograph as described in the WHO manuscripts.
23
University of Ghana          http://ugspace.ug.edu.gh
The impact of breech labour management on foetal and maternal outcomes o f labour 
was assessed using the WHO partograph (Lennox, Kwast and Parley, 1997). In all 1,740 
breech presentations in a larger multicenter hospital-based were studied. The researchers 
concluded that the use o f  the WHO partograph in the management o f breech labour reduced 
prolonged labour and (among multigravida) caesarean section and also improved foetal 
outcomes. However, in the study, caesarean section was reported to be a safer method of 
delivery for the baby, regardless o f whether or not the partograph was used (International 
Federation o f Gynaecology and Obstetrics, 1998). Other studies have shown that using the 
partograph can be highly effective in reducing complications for the mother from prolonged 
labour for the mother (postpartum haemorrhage, sepsis, uterine rupture) and in preventing 
newborn death from anoxia and infections (MSH and UNICEF, 1998).
Some researchers argued that the partograph can be used as a traditional format with 
separate pages for the deliver)' summary and midwifery notes, whereas others incorporated the 
summary, the partograph and notes into one comprehensive document. Thus, the use o f  the 
partograph reduces excessive documentation on the part o f the midwife as well as reducing 
prolonged labour as asserted by other researchers (Groeschel and Glover, 1998).
In another study, Lavender and Malcolmson (1999) studied midwivcs" views on the 
value o f action lines when using the partograph and reported that 83 percent o f  midwives 
surveyed felt that action lines helps to ‘manage labour’ and ‘diagnose prolonged labour at a 
glance’ However, others felt that it takes away autonomy and individual clinical decision­
making. I hough the study does not indicate the number o f midwives interviewed, it is worth
24
University of Ghana          http://ugspace.ug.edu.gh
noting that documenting the information correctly on the partograph can help in early 
identification o f the abnormal progress of labour and make timely intervention more likely.
Ghanaians adopted the use o f the WHO partograph in the labour wards in 1 9 9 0 - 1991, 
to monitor labour and help to identify women in need (abnormal progress of labour) of an 
obstetrical intervention, with a view to reducing the number o f prolonged labours, 
appropriately augmenting labour with oxytocin, reducing rates o f caesarean section and 
reducing the incidence o f infection, thereby improving maternal health and reducing maternal 
and perinatal mortality. Has the introduction o f the WHO partograph helped in making 
decisions early enough for women in need o f interventions? Are providers still using the 
partograph? Are they using it correctly? Is the correct use associated with changes in 
outcomes for mother and baby?
2.5 Theoretical Framework
O f all the health indicators monitored by WHO, maternal mortality rates (MMR) 
demonstrate the largest disparity between higher and lower income countries. The disparity of 
MMR between countries can largely be attributed to knowledge gap and lack o f human, 
financial and material resources available to women o f low socioeconomic status with less 
education and living in poorer countries.
To help bridge this gap, series of activities were designed by the WHO under its Safe 
Motherhood Programme to update the knowledge and skills o f providers o f maternity 
services. I he Life Saving Skill (LSS) training programme for midwives and doctors was 
developed to provide clinical training and practice in such areas as use o f the partograph.
25
University of Ghana          http://ugspace.ug.edu.gh
manual removal ol'lhe placenta, prevention and management o f  haemorrhage and post­
abortion care including manual vacuum aspiration. It is anticipated by those concerned with 
maternal and neonatal health that if  the partograph is adequately used and necessary 
interventions are promptly initiated, most complications occurring during childbirth will be 
prevented. This will reduce maternal and neonatal deaths to a level comparable to those 
observed in higher income countries.
This approach is consistent with the theoretical approach taken by the Neumcm Systems 
Model (George, 1985), where three levels o f prevention, that is, primary, secondary and 
tertiary preventions are descibed. Primary prevention relates to general knowledge applied to 
client assessment in an attempt to identify stressors before they occur. Secondary prevention 
relates to symptomatology which includes interventions that are initiated after an encounter 
with a stressor (George, 1985). Nurses and midwives uniquely use this model to assist 
individuals, families and other groups to attain and maintain maximum levels o f total wellness 
by purposeful interventions. The use o f the partograph is a secondary prevention strategy. It 
enables providers to monitor labour effectively so that timely interventions can be taken as 
needed to ensure the safety o f the mother and the newborn.
26
University of Ghana          http://ugspace.ug.edu.gh
C H A P TE R  TH R E E  
R ES EA R C H  P R O C E S S
Although there have been studies on evaluation of the use o f the WHO partograph in 
relation to maternal and neonatal outcomes in other parts of the world, there has not been 
adequate assessment o f this tool in monitoring labour and its outcome in Ghana. This is a 
retrospective study, which is an example of a non-experimental research design to evaluate the 
relationship between the use of the partograph and the outcomes o f birth. Retrospective 
studies are almost always cross-sectional and data are collected concurrently with regard to the 
independent (partograph) and dependent (maternal and neonatal outcomes and interventions) 
variables. The discussion o f the research method includes information about a) sample, b) 
instrument, and c) procedure.
Advantages of a Retrospective Study
• It is an efficient and effective means o f collecting a large amount of data about a 
problem area.
• It tends to be high in realism and it is seldom criticised for its artificiality.
• It helps in advancing our understanding of what the world around us is like. 
Disadvantages of a Retrospective Study
• It is weak in its ability to reveal causal relationship among variables. Manipulation o f 
an independent variable and control o f extraneous variables are not possible because 
the events occurred prior to data collection.
• It is susceptible to faulty interpretation (Polit and Hungler, 1997).
University of Ghana          http://ugspace.ug.edu.gh
Three hypotheses each with accompanying sub-hypothesis, were advanced to answer 
research questions posed for this study.
Hypothesis 1:
There will be more interventions documented in the partographs that are used adequately 
than in those that are used inadequately.
a) There will be more augmentation documented in the charts o f women where the 
partograph was used adequately than in those where the partograph was used inadequately.
b) There will be more artificial rupture of membranes (ARM) documented in the charts 
of women where the partograph was used adequately than in those where the partograph was 
used inadequately.
c) There will be more assisted births (i.e., vacuum extraction, forceps delivery, 
Caesarean section) documented in the charts o f women where the partograph was used 
adequateh than in those where the partograph was used inadequately.
Hypothesis 2:
There will be less intrapartum and postpartum (within 2 weeks of birth) morbidity 
documented in the charts o f women where the partograph was used adequately than in those 
where the partograph was used inadequately.
a) There will be less pyrexia at any time during the intrapartum and postpartum period 
documented in the charts o f women where the partograph was used adequately than in those 
where the partograph was used inadequately.
28
University of Ghana          http://ugspace.ug.edu.gh
b) There will be less haemorrhage (i.e., blood loss estimated to be more than 500 ml) at 
any time during the postpartum period documented in the charts o f women where the 
partograph was used adequately than in those where the partograph was used inadequately.
c) There will be fewer readmissions during the first 2 weeks of the postpartum period 
documented in the charts o f women where the partograph was used adequately than in those 
where the partograph was used inadequately.
d) There will be less mortality at any time during the intrapartum and postpartum period 
documented in the charts o f women where the partograph was used adequately than in those 
where the partograph was used inadequately.
Hypothesis 3:
There will be less newborn morbidity documented in the charts o f women where the 
partograph was used adequately than in those where the partograph was used inadequately.
a) There will be more newborns with Apgar scores o f 7 or more at 1 minute o f age 
and 5 minutes of age documented in the charts o f women where the partograph was 
used adequately than in those where the partograph was used inadequately.
b) There will be few birth injuries (i.e., cephal haematoma) to newborns documented 
in the charts o f women where the partograph was used adequately than in those 
where the partograph was used inadequately.
c) There will be fewer admissions to the neonatal intensive care unit (NICU) 
documented in the charts o f women where the partograph was used adequately 
than in those where the partograph was used inadequately.
29
University of Ghana          http://ugspace.ug.edu.gh
d) There will be fewer neonatal deaths (during the intrapartum and within 2 weeks of 
birth) documented in the charts o f women where the partograph was used adequately 
than in those where the partograph was used inadequately.
3.1 Design
To assess the relationship between the use o f  the partograph and birth 
outcomes, a retrospective comparative chart review was used and charts were assigned 
to one o f 2 groups depending on whether or not a partograph was used and adequately 
completed.
The partograph consists o f three components: foetal condition, progress of labour and 
maternal condition.
a  Foetal condition consists of foetal heart rate (FHR), status o f membranes and 
appearance o f liquor as well as presence o f moulding. During labour the foetal heart rate must 
be checked and recorded every hour, but if the foetal heart is irregular then it should be 
checked and recorded every fifteen minutes. The state o f the membranes and liquor and 
moulding should be checked and recorded every four hours.
b. Progress o f  labour comprises cervical dilatation, descent o f the presenting part and 
uterine contractions. In labour, dilatation o f the cervix and descent o f the presenting part 
should be assessed on admission and repeated at least every four hours. More frequent 
examination should be done as the need arises. Uterine contractions should be checked and 
recorded every half hour.
c. Maternal condition has the following components: maternal blood pressure, pulse, 
temperature and assessment of urine for presence of protein and keynotes. Pulse rale should be
30
University of Ghana          http://ugspace.ug.edu.gh
assessed and recorded hourly, blood pressure, temperature and urinalysis should be assessed 
and recorded every four hours.
A partograph was considered to be adequately or appropriately used if all the three 
components: progress o f labour, maternal condition and foetal condition were assessed and the 
partograph was filled out in the manner recommended in the WHO guidelines (Appendix B). 
Furthermore, if there were no omissions in any of the above areas or if the omissions were 
deemed to be minor (1-2 omissions in any of the components) the partograph was considered 
adequate or appropriate. A partograph was considered to be inadequate if  there were moderate 
(3-4 omissions in any o f the components) or major (5 or more) omissions. Definitions of 
levels of omissions are presented in the following table.
Table 1: Components of Partograph and levels of omission
Components of Partogragh No omission Omission
a. Foetal condition 
I. FHR
II. Moulding
III. Liquor
b. Progress o f labour 
IV. Descent of foetal head
V. Cervical dilatation
VI. Uterine contractions
c. Maternal condition 
VII. Pulse
31
University of Ghana          http://ugspace.ug.edu.gh
VIII. BP
IX. Temperature
X. Urine (Urinalysis)
A score of 10 was possible for all partographs , one point for each o f the ten variables under 
the components o f partograph chosen for this study and classified as adequate or inadequate.
Tabic 2: Levels of Omissions resulting in adequate and inadequate partograph
Adequate Partograph No Omission Minor Omissions, i.e. 
1 -  2 omissions
Inadequate Partograph Moderate Omissions, i.e. Major Omissions, i.e.
3 - 4  omissions 5 and above omissions
3.2 Setting
Korle-Bu Teaching Hospital, the site for the study, is situated on the west part o f Accra 
about 0.5 kilometers from the Korle lagoon. Korle-Bu Teaching Hospital is the oldest, largest 
and first teaching hospital in the country and serves as the teaching hospital for the College of 
Health Sciences, which is comprised of the Medical and Dental School, School of Nursing, 
School of Public Health, Noguchi Memorial Institute for Medical Research and Allied Health 
Sciences. Currently, it has about 1.560 beds, making it one of the biggest hospitals in the 
country. It consists o f the following departments:
• Medicine and Therapeutics
• Surgery, Theatres and Recovery
• Allied Surgery
32
University of Ghana          http://ugspace.ug.edu.gh
• National Cardio-Thoracic Centre
• O bstetrics and G ynaecology
• Child Health
• Reconstructive Plastic Surgery and Burns Centre
• Radiotherapy Centre
• Surgical and M edical Em ergency,
• Accident. Casualty U nit and Central O ut-Patients
• Physiotherapy and Pharm acy
• Polyclinic and A dm inistration
The D epartm ent o f  O bstetrics and G ynaecology has two m ain units, that is, the 
D epartm ent o f  O bstetrics (M aternity  Block) and the D epartm ent o f  G ynaecology. The setting 
for this study was the records departm ent at the M aternity B lock o f  the D epartm ent o f 
Obstetrics and G ynaecology, w hich handles about 12.000 deliveries annually, w ith an average 
o f  about 993 deliveries per m onth. The unit serves as referral point for various low er-level 
health facilities w ith in  m etropolitan  A ccra and surrounding areas. The unit is m anaged under 5 
team s o f  doctors w ith 14 physicians on each team  and 253 nursing and m idw ifery staff.
The departm ent has six floors or w ards w ith 54 beds each except the 6Ih floor, which 
has 24 beds. It has 2 labour wards and 2 obstetrics theatres, an out-patient unit including 8 
consulting room s and an adm ission bay where all patients adm itted to the facility are 
registered and exam ined prior to adm ission. The N eonatal Intensive Care U nit (NICU). which
33
University of Ghana          http://ugspace.ug.edu.gh
is pari o f  the D epartm ent o f  Child Health, is also housed on the 3rcl floor o f  the m aternity 
block. There is a laboratory unit, an X-ray, Pharm acy and M edical statistics and 
docum entation units.
The departm ent operates 24-hour em ergency services for both antenatal and postnatal 
w om en as well as m anagem ent o f  labour and delivery and ultrasonologic services. There is 
also the Public Health unit w here im m unisations for m others and babies are done, counselling 
for family
planning take place and cervical cancer screening m others antenatally  and postnatally.
The D epartm ent o f  G ynaecology is behind the m ain A dm inistra tion  Block o f  the 
Korle-Bu Teaching H ospital. It has an O ut-Patient unit w ith 7 consulting  room s and an 
em ergency room. It also has 3 m ain wards, nam ely, W ard E w ith 34 beds, C henard A w ith 50 
beds. Chenard A nnex w ith 30 beds and a G ynaecology theatre. The unit handles all 
reproductive health cases including early pregnancy care, obstetrics em ergency care, cervical 
cancer screening services and fam ily planning services.
3.3 Sample
To avoid selection bias, charts o f  w om en who delivered at K orle Bu and m et inclusion 
criteria were enrolled as sequentially as possible. The charts w ere from  both labour wards 
(i.e., 1 & 2). Charts o f  m ultiparous and prim iparous w om en at any gestational age greater 
than 24 weeks and who w ere deem ed to be in active labour (regular uterine contractions, i.e.. 3 
contractions in 10 m inutes) and had a cervical dilation o f  least 4 centim etres but not fully 
dilated were included.
34
University of Ghana          http://ugspace.ug.edu.gh
Charts o f  w om en treated for hypertension during the index pregnancy or who had been 
adm itted to hospital with a blood pressure greater than 140 systolic and/or 100 diastolic were
excluded. Charts o f  w om en who arrived at the hospital in second stage o f  labour or were 
booked for elective C aesarean section or had experienced intrauterine death prior to labour in 
the index pregnancy were also excluded.
The sam ple consisted o f  charts from the defined cohort o f  births occurring at Korle-Bu 
teaching hospital during the m onths o f  May 1 to June 30, 2003. These m onths had the highest 
recorded deliveries. 1.187 and 1,215 respectively. In all. 1,845 charts w ere review ed. About 
100 to 150 charts w ere review ed daily and about 24 percent o f  these charts did not m eet the 
inclusion criteria.
3.4 Procedure
The protocol for data collection was as follows:
a) Data w ere extracted m anually from the m edical records o f  the selected 
cohort and recorded on the attached data collection sheets (A ppendix B).
b) E ligible charts w ere selected serially as arranged by m edical records
c) For the first 10 charts, the researcher assessed inter-rater consistency in 
recording the adequacy o f  the partograph. the use o f  interventions, and all m aternal and 
newborn outcom es by having a colleague w ith about the sam e level o f  skill and experience 
also com plete data collection forms. An agreem ent o f  90%  w as accepted.
d) Data from the data collection sheets were transferred to the com puter database.
e) A colleague checked every 20lh entry for accuracy.
35
University of Ghana          http://ugspace.ug.edu.gh
3.5 Assumptions
It was assum ed Lhat inform ation recorded on the charts accurately reflected the 
characteristics o f  the w om en, the interventions that were perform ed, and level o f  m aternal and 
neonatal m orbidity and mortality.
3.6 Analysis
Data were analysed using a com parative design because the researcher is interested in 
assessing the relationship betw een 2 variables [i.e., the adequate use o f  the partograph 
(independent variable) and birth outcom es (dependent variable)]. A database com patible with 
the m ost recent version o f  the Statistical Package for the Social Sciences (SPSS) was 
developed.
The analyses consisted  o f  two sections. First, the dem ographic characteristics and 
outcom es recorded in the charts were described and relationships am ong selected variables 
were evaluated. Categorical variables w ere described using percentages and bar charts. 
Numerical variables w ere described using m easures o f  central tendency (m eans, m edians) 
and variability (ranges, standard deviations). R elationships am ong selected variables were 
calculated using sim ple correlations and Chi-square analysis. N ext relationships between 
selected outcom es and the use o f  the partograph were assessed using non- param etric (i.e., 
Chi-square) statistics.
3.7 Variables
Data were collected for the follow ing dem ographic, m aternal intervention, m aternal and 
new born outcom e, and partograph variables.
1. D em ographic
36
University of Ghana          http://ugspace.ug.edu.gh
a. Age (actual age)
b. Parity (0, 1, 2-5, more than 5)
c. M arital status (m arried, single, divorced, w idow ed)
d. O ccupation (unem ployed, inform al private sector, form al private sector, public 
sector)
e. M ode o f  last delivery prior to the index pregnancy (N/A, N SV B , C/S, V acuum  
extraction, forceps)
f. G estational age at onset o f  labour (37 w eeks or less, 38 w eeks or more)
2. M aternal Interventions
a. O xytocin augm entation (yes, no)
b. A rtificial rupture o f  m em branes during labour (yes, no)
c. A ssisted delivery (C/S. vacuum  extraction, forceps, episiotom y, breech
extraction)
3. M aternal O utcom es
a. Pyrexia above 37.5 C at any tim e (yes, no)
b. M aternal haem orrhage o f  500 ml and above (less than 500m l is no,
5 0 0 -1000ml is m oderate: m ore than 1000ml is severe)
c. R eadm ission to hospital w ithin 2 w eeks o f  birth (yes. no)
d. M ortality (yes, no)
4. N ew born O utcom es
A. A pgar Scores (6 or less) or (7 -  10) at one m inute o f  birth
b. A pgar Scores (6 or less) or (7 -  10) at five m inutes o f  birth
37
University of Ghana          http://ugspace.ug.edu.gh
c. B irth injury (cephal haem atom a, others)
d. A dm itted to NICU (yes, yes but after operative birth, no)
e. M ortality (yes, no)
5. Partograph adequacy (yes, no)
a. A dequate (1 if  no om issions, 2 if  m inor om issions)
b. Inadequate (3 if  m oderate om issions, 4 if  m ajor om issions)
c. Date and tim e action line w as crossed
d. A ction taken in tim ely m anner (yes, no)
A data collection sheet w as developed to reflect values for the above variables (A ppendix C).
3.8 Ethical Considerations
A letter requesting perm ission lo conduct the study was sent to the D eputy D irector o f  N ursing 
Services (DDNS) o f  the K orle-B u Teaching H ospital, the D eputy D irector o f  N ursing 
Services (DDNS) o f  the D epartm ent o f  O bstetrics and G ynaecology and the head o f  the 
M edical Statistics and D ocum entation  D epartm ent o f  the study area (A ppendix D). This 
proposal was subm itted to the Institutional R eview  Board o f  the N oguchi M em orial Institute 
for M edical Research (N M IM R ) and approval obtained before the com m encem ent o f  data 
collection. All data w ere extracted from  the m edical records by the researcher. C onfidentiality 
was m aintained since no nam es or identifying inform ation w ere collected w ith the data. A 
colleague who does not w ork at the m aternity unit and therefore was unlikely to know  the 
nam es o f the subjects whose charts were being review ed assisted in exam ining the charts. This 
also ensured confidentiality in the charts reviewed. N o consent from  subjects was required 
because charts rather than individuals were studied. D ata sheets w ere stored in a safe locked
38
University of Ghana          http://ugspace.ug.edu.gh
cabinet in the School o f  Nursing, University o f  Ghana and secondary analysis will be done 
only with approval o f  an appropriate ethics com m ittee.
39
University of Ghana          http://ugspace.ug.edu.gh
CHAPTER FOUR 
RESULTS
The results o f  the data analysis are presented in this chapter. First, the dem ographic 
characteristics (including age, parity, m arital status, occupational status, m ode o f  last delivery 
and gestational age o f  onset o f  labour) are presented. Then the relationship betw een the 
independent variable (i.e., adequacy o f  the partograph) and the frequency o f  m aternal 
interventions, m aternal outcom es and new born outcom es are reported. A total o f  1,845 charts 
that met inclusion criteria for this study w ere review ed. These included the charts o f  930 
w om en adm itted to the M aternity B lock o f  the K orle Bu T eaching H ospital in M ay 2003 and 
915 in June 2003. A  sum m ary o f  how  eligible charts w ere selected from  the 2,622 adm issions 
to the M aternity B lock during that tim e is presented in F igure 1.
Figure 1: O bstetric A dm ission in Korle-Bu M aternity B lock in M ay and June2003
Obstctric Adm ission in Korle-Bu Maternity Block
2622
M ay June
1310 1312
In labour Not in labciui In labour Not in labour 
1187 123 1214 98
Included Excluded Included Excluded 
Planned use of Partograph not use  (Eclampsia, Planned u se o l Partograph not u se  (Eclampsia, 
Partograph elective C/* 2 stage arrival. Partograph Elective C/S, 2 nri stage arrival,
QfU IUFD) 223 07? It |FD> 7 4 ?
i
C h a n s  found Charts not found Chaits found Charts nol found
930 (M issing charts Including deaths) 915 (M issing charts including deaths)
34 57
40
University of Ghana          http://ugspace.ug.edu.gh
4.1 Demographic characteristics
4.1.1 Age
The average age o f  the w om en whose charts were included was 27.7 years old (SD
6.09 years). Their ages ranged from  15 to 45 years old. The m ajority  o f  the w om en were 
w ithin the age category o f  21 to 30 years old.
4.1.2 Parity
Out o f  the 1,845 w om en w hose charts were review ed, 1,125(61.0% ) w ere m ultiparous 
while 717(38.9% ) w ere nulliparous. The modal category for parity  w as the 2 to 5 category. A 
more detailed description o f  parity is presented in Table 3.
Table 3. Parity'
Parity Frequency Percent
0 (Prim igravid) 717 38.9
1 505 27.4
2 - 5 569 30.8
More than 5 51 2.8
M issing data Jo 0.2
Total 1845 100.0
4.1.3 M arital status
O f the 1,845 charts review ed, marital status was reported for 1,843 wom en. O f these.
1,612 were m arried w om en living with their husbands. The rem ainder reported their m arital 
status as single.
4.1.4 Occupation
O f the 1,845 charts review ed, 1,831 w om en reported their occupations. O f these. 1.321 
(71.6% ) were em ployed either in the inform al private sector, form al private sector or public 
sector. Ih e  rem ainder w ere unem ployed. For 14 (1%), their em ploym ent status was not 
known.
41
University of Ghana          http://ugspace.ug.edu.gh
4.1.5 M ode o f  last delivery
O f 1.128 m ultiparous w om en whose charts w ere review ed, 989(88% ) had a previous 
norm al spontaneous vaginal birth, while the rest delivered either by caesarean section or 
vacuum  extraction. The previous delivery experience o f  m ultiparous w om en is provided in 
m ore detail in Table 4.
Table 4: Mode of Last Delivery
M ode o f  last delivery Frequency Percent
N orm al spontaneous vaginal birth (NSVB) 974 52.8
Caesarean section (C/S) 132 7.2
Vacuum  extraction (V/E) 7 0.4
N on-application (Prim igravid) 705 38.2
M issing data 27 1.5
Total 1845 100.0
4.1.6 G estational age o f  onset o f  labour
For the m ajority o f  the 1,824 w om en w hose gestational ages were included on their 
charts, 1,396 (76.5% ) experienced the onset o f  labour w hen their gestational age w'as 38 w eeks 
or more. For 428 (23.5% ), the onset o f  labour occurred prior to 37 com pleted w eeks o f  
gestation. Inform ation about gestational age was m issing for 21 charts.
4.2 Research Hypotheses
In the next sections the research hypotheses will be addressed. Each o f  the three 
hypotheses will be listed with the accom panying sub-hypotheses. The results o f  the analysis 
o f  each individual sub-hypothesis will be reported. Before reporting findings from research
42
University of Ghana          http://ugspace.ug.edu.gh
hypotheses, guidelines used to establish the adequacy ol the independent variables (i.e., 
partograph and whether or not tim ely action was taken) will be described.
4.3 Partograph
The independent variable is the adequacy o f  the partograph. U sing guidelines based 
on W orld Health O rganization (W HO) criteria established a priori to determ ine w hether or 
not the partograph w as judged  to be adequate, m aternal in terventions as well as m aternal and 
newborn outcom es w ere com pared w ith respect to w hether or not the partograph was 
adequately used.
The levels o f  om ission in plotting the partograph as required by W H O  guidelines are 
presented in Table 4. For the m ajority o f  charts that w ere review ed for this study, assessm ents 
o f  com ponents o f  the partograph (i.e., FHR, liquor and m em branes, and presence o f  caput or 
m oulding, as well as uterine activity, m aternal pulse and urine analysis) w ere not plotted to 
specification. There were one or m ore om issions during plotting or charting o f  the partograph.
Table 5. Components of Partograph and Levels of Omission
Com ponents o f  Partograph No om issions O m issions
Frequency Percent Frequency Percent
a. Foetal conditions 
i. FHR 121 6.6 1724 93.4
ii. M oulding 286 15.5 1559 84.5
iii. Liquor 839 45.5 1006 54.5
b. Progress o f  labour 
iv. Descent o f  head 1699 92.1 146 7.9
v. Cervical dilatation 1759 95.3 86 4.7
vi. Uterine contractions 211 11.4 1634 88.6
c. M aternal conditions 
vii. Pulse 194 10.5 1651 89.5
viii. BP 1568 85.0 277 15.0
xi. Tem perature 1221 66.2 624 33.8
x. Urine (Urinalysis) 304 16.5 1541 83.5
•13
University of Ghana          http://ugspace.ug.edu.gh
O f the 1,845 charts review ed, the partographs w ere correctly filled out in accordance with 
W HO guidelines on 472 (25.6% ) charts while 1,373 (74.4% ) partographs w ere incorrectly 
filled out.
4.3.1. Action Line on Partograph:
A lthough the use o f  the partograph by care providers w as filled on 464 occasions, it 
was also noted that w hen the action line was crossed, action w as only taken in a tim ely 
m anner on 226 (48.7% ) occasions and was not taken on 238 (51.3% ) occasions. Therefore, 
m aternal and new born outcom es w ere com pared on a) w hether or not the partograph was 
adequately filled out and b) the charts w ere further analyzed so that those for w hom  action 
was taken when the action line was crossed w ere com pared w ith those for w hom  no action 
was taken.
Out o f  the 1.845 charts review ed, for 1,381(74.9% ) w om en, labour ended w ithout the 
action line being cross. For 464 (25.1% ) the action line w as crossed. O f  these, action was not 
taken in a tim ely m anner to augm ent labour for 238(51.3% ). T im ely action w as therefore 
taken to augm ent labour for only 226(48.7% ) women.
Hypothesis 1:
There will be m ore in terventions docum ented in the partographs that are used adequately than 
in those that are used inadequately.
a) There will be m ore augm entation docum ented in the charts o f  w om en w here the 
partograph was used adequately than in those w here the partograph was used 
inadequately.
44
University of Ghana          http://ugspace.ug.edu.gh
O f the 1,845 records, data were available lor 1,844. O f these, the labour o f  623 
w om en was augm ented w ith oxytocin. Those whose partograph was judged to be adequate 
were m ore likely to have their labour augm ented than those w hose partographs w ere judged  to 
be inadequate (X2=175.8, d f  1, p<0.0009). The hypothesis was supported.
b) There will be m ore artificial rupture o f  m em branes (A R M ) docum ented in the charts 
o f  w om en w here the partograph was used adequately than in those w here the 
partograph w as used inadequately.
Data were available on 1,835 o f  the 1,845 records. O f these, 673 had their labour 
augm ented by having their m em branes artificially ruptured during labour. Those whose 
partographs were judged  to be adequate were more likely to have their m em branes ruptured 
(X2=87.3. d f  l.p < 0 .0 0 0 9 ). The hypothesis was supported.
c) There will be m ore assisted births (i.e., vacuum  extraction, forceps delivery,
Caesarean Section) docum ented in the charts o f  w om en w here the partograph was used 
adequately than in those where the partograph was used inadequately.
O f the 1,845 records, data with respect to assisted birth w ere available on 1.839. O f 
these. 778 w om en had assisted births. These included C aesarean Section (n=235); vacuum  
extraction (n=38); forceps extraction (n=5); episiotom y (n=455); and assisted breech birth 
(n=44). This hypothesis was also supported as there were m ore assisted births in group whose 
partographs were adequately filled out (X2=108.3, d f  1, p<0.0009).
4.4.2 Types o f  m aternal interventions
M aternal interventions included oxytocin augm entation, artificial rupture o f  
m em branes (ARM ) during labour and assisted deliveries such as C aesarean section, vacuum
University of Ghana          http://ugspace.ug.edu.gh
extraction, forceps, episiotom y and breech delivery extraction. Six hundred and tw enty-four 
(33.8% ) o f  the w om en had their labours augm ented w ith oxytocin and 673(36.5% ) had their 
m em branes artificially ruptured during labour. Out o f  the 778(42.2% ) w om en who were 
assisted during delivery, 456(58.6% ) o f  them  had norm al spontaneous vaginal births with 
episiotom y w hile 235(30.2% ) had caesarean sections. I f  only Caesarean Section, vacuum  
extractions or forceps deliveries w ere considered as assisted birth, there w ere differences 
between those w ho had a partograph that was adequately filled out and those w ho d id n ’t 
(X 2=5.16, d f  2, p<0.03).
Table 6: Various M aternal Interventions
M aternal Interventions Yes No Total
a. Oxytocin augm entation 624 (33.8) 1221 (66.2) 100.0
b. ARM  during labour 673 (36.5) 1172 (63.5) 100.0
c. Assisted delivery 778 (42.2) 1067 (57.8) 100.0
4.4 Maternal Outcomes 
Hypothesis 2:
There will be less in trapartum  and postpartum  (w ithin 2 w eeks o f  birth) m orbidity docum ented 
in the charts o f  w om en w here the partograph was used adequately than in those w here the 
partograph was used inadequately.
a) There will be less pyrexia at any tim e during the intrapartum  and postpartum  period 
docum ented in the charts o f  w om en where the partograph w as used adequately than in 
those w here the partograph was used inadequately.
M aternal tem perature was available for all 1,845 records. O f these, 427 (23.1% ) o f  the 
wom en had a body tem perature record at or above 37.5 degrees C entigrade (C). at som e time 
during their intra partum  or postpartum  period. Those w hose partographs w ere not adequately
46
University of Ghana          http://ugspace.ug.edu.gh
com pleted were less likely to have a m aternal tem perature above 37.5 recorded on their 
partograph (X2=13.2, d f  1, p<0.0009). The hypothesis was not supported.
However, m aternal tem perature was also com pared betw een those for w hom  tim ely 
action was taken if  the action line was crossed and those w hom  tim ely action was not taken. 
There were no betw een-group differences (X 2=0.64. d f  1, p=0.44).
b) There will be less haem orrhage (i.e., blood loss o f  m ore than 500 m l) at any tim e 
during the postpartum  period docum ented in the charts o f  w om en w here the partograph 
w as used adequately than in those where the partograph w as used inadequately.
D ata w ere available on 1,811 o f  the 1,845 records. The records o f  1,360(73.7% ) 
wom en indicated that they had experienced post delivery b lood loss o f  less than 500m ls while 
485(26.3% ) experienced either m oderate or severe blood loss. The m axim um  blood loss was 
assessed to be 2500m ls.
There w ere betw een-group differences w ith those w hose partographs were adequately 
com pleted being less likely to have a m oderate or severe haem orrhage (X z=81.5, d f  2, 
p<0.0009). The hypothesis w as supported .
W hen the presence o f  m inim al, m oderate or severe b lood loss w as com pared between 
those where action w as taken in a tim ely m anner w hen the action line w as crossed and those 
where action was not taken, differences were not found.
d) There will be few er readm issions during the first 2 w eeks o f  the postpartum  period 
docum ented in the charts o f  w om en where the partograph was used adequately than in 
those where the partograph was used inadequately.
47
University of Ghana          http://ugspace.ug.edu.gh
O f the 1,845 records, inform ation about w hether or not the m other w as re-adm itted to 
hospital during the postpartum  period was available in 1,836. It w as indicated in 92 charts 
that w om en were readm itted to the m aternity unit. O f these, 26 (27.2% ) o f  the w om en were 
readm itted w ith pre-eclam psia, 21 (19% ) with puerperal sepsis, 19 (21% ) w ith m alaria or UTI 
and 8 (9%) w ith severe anaem ia. The rem aining 18 (20% ) w ere adm itted w ith com plications 
such as wound infection, gaping episiotom y, bleeding per vagina (due to secondary PPH) and 
sickle cell crisis.
M ore w om en in the group w hose partographs w ere adequately com pleted were 
readmitted during the first 2 w eeks postnatally (X2=5.7, d f  1, p=0.01). The hypothesis was not 
supported.
W ith respect to w hether or not action was taken in a tim ely m anner, there wrere no 
differences betw een groups in rates o f  m aternal readm ission to hospital w ithin the first 2 
weeks follow ing birth (X2“0.03, d f  1, p=0.87).
There w ill be less m ortality  at any tim e during the in trapartum  and postpartum  period 
docum ented in the charts o f  w om en w here the partograph w as used adequately than in those 
where the partograph w as used inadequately.
Total adm issions and deliveries recorded during the study period M ay and June. 2003 
were 2,622 and 2,401. respectively. The unit recorded 29 m aternal deaths during the sam e 
period. O f these, it was reported in the log book that there w ere 13 intrapartum  m aternal 
deaths but only 1 chart was located. The charts o f  the rem aining 12 w om en who died during 
labour were am ong those that w ere m issing, l he researcher becam e aw are o f  the num ber o f 
deaths and their causes through scrutiny of records kept by nursing adm inistrators.
48
University of Ghana          http://ugspace.ug.edu.gh
The num ber o f  eligible charts o f  wom en who died was insufficient to assess between 
group differences w ith respect to the adequacy o f  the partograph. Therefore, the hypothesis 
was not tested.
4.5 Newborn Outcomes 
Hypothesis 3:
There will be less new born m orbidity docum ented in the charts o f  w om en w here the 
partograph was used adequately than in those w here the partograph w as used inadequately,
a) There will be m ore new borns w ith A pgar scores o f  7 or m ore at 1 m inute o f  age and 5 
m inutes o f  age docum ented in the charts o f  w om en w here the partograph was used 
adequately than in those w here the partograph was used inadequately.
Out o f  the 1,845 charts review ed, A pgar scores at 1-m inute o f  age w ere available for
1,843 births. A total o f  1.885 babies including 40 sets o f  tw ins w ere delivered during the 
study period. For this study only the Apgar scores of the first twin were recorded. O f 
these, 704 (38.1% ) babies had A pgar score o f  6 or less at 1 m inute o f  age and 1,139(61.9% ) 
babies had scores o f  7 to 10 w ithin the same period o f  time. W hile 1 80 A pgar scores o f  less 
than 6 were expected, 238 w ere observed in the groups w hose partograph w as adequately 
com pleted. Conversely, w hile 524 A pgar scores o f  6 or less w ere expected in the group where 
the partograph w as inadequately com pleted, only 466 w ere observed. The differences were 
significant, in that A pgar scores were higher in those w hose partographs w ere not filled out 
appropriated (X2=40.8. d f  1, p<0.0009). The hypothesis was not supported .
However, when the 226 cases where action was taken in a tim ely m anner was 
exam ined, Apgar scores o f  6 or less were expected for 114 o f  those w here action w as taken in
49
University of Ghana          http://ugspace.ug.edu.gh
a tim ely m anner; only 102 w ere observed. Conversely, the 238 w here action was not taken in 
a tim ely m anner, 120 cases w ere expected and 133 were observed. In this case, those who 
received tim ely action were less likely to have babies w ith low  A pgar scores (X2==5.9, d f  1, 
p=0.0009)
Apgar scores w ere recorded on all charts for 1,845 births at 5 m inutes o f  age. O f 
these, 207 babies (11.2% ) had scores o f  6 or less and 1,638 (88.8% ) babies scored 7 to 10 at 5 
m inutes after birth. T hose w hose partograph were adequately filled out w ere m ore likely to 
have scores o f  6 or less at 5 m inutes o f  age (X2= 5 .6 ,d f 1, p=0.01). The hypothesis w as not 
supported.
W hen those w here action was taken in a tim ely m anner were com pared w ith those 
where action was not taken in a tim ely m anner w ith respect to 5-m inutes A pgar scores, those 
where action was not taken w ere significantly higher.
b) There will be few  birth  injuries (i.e., cephal haem atom a) to new borns docum ented in 
the charts o f  w om en w here the partograph was used adequately than in those w here the 
partograph was used inadequately.
O f the 1,885 babies delivered from  the 1,845 charts, only 17 birth injuries were 
recorded. Again, only data for the first o f  each o f  the 40 sets o f  tw ins w ere available. O f these 
3(0.2% ) suffered cephal haem atom a w hile 14 w ere born w ith congenital abnorm alities such as 
anencephaly, extra digits, haem agiolym phangiom a, ectropion o f  left eye. hydrospadias, 
exom phalous and supra-um bilical cord. Birth injuries w ere recorded for 6 neonates in the 
group whose partographs were adequately com pleted, while 11 occurred in the group whose 
partographs were inadequately com pleted. The num ber was sm all but Chi Square analysis
50
University of Ghana          http://ugspace.ug.edu.gh
was done. No betw een-group differences were found. The hypothesis was not supported (X2 
-0 .9 , d f l ,p = 0 .2 5 ) .
l or the 464 charts w here it was recorded that the action line was crossed, there were no 
differences between those w here tim ely action was taken and those w here tim ely action was 
not taken (X2=0.004. d f  1 p=0.63).
c) There will be fewer adm issions to the neonatal intensive carc unit (NICU) docum ented 
in the charts o f  w om en w here the partograph was used adequately than in those where 
the partograph w as used inadequately.
Inform ation was available on 1,841 records. O f these, 243 new borns, excluding twins 
were adm itted to N IC U . O f these, 62 were routinely adm itted because o f  C/S and 181 were 
adm itted because o f  other problem s.
The tim eliness o f  intervention after the action line was crossed was evaluated. Again, 
there were more adm issions to NICU in the group w here action was not taken in a timely 
m anner (X 2=7.8 d f  2. p<0.02).
d) There will be few er neonatal deaths (during the intrapartum  and w ithin 2 w eeks o f  
birth) docum ented in the charts o f  w om en w here the partograph was used adequately 
than in those w here the partograph was used inadequately.
There were 29 neonatal deaths. The sam ple size was small but Chi Square analysis was done 
and there were no differences in the num ber o f  deaths betw een those w hose partograph was 
adequate and those w hose partograph was inadequate. The hypothesis was not supported 
(X2“0.03, d f -  1, p<0.52)
51
University of Ghana          http://ugspace.ug.edu.gh
CHAPTER FIVE 
DISCUSSION
5.1 Summary of Main Findings
In this study, 1,845 retrospective charts were review ed and the adequate and 
inadequate use o f  the partograph was analysed using three m ain hypotheses. The study 
findings revealed that there w ere m ore m aternal interventions, i.e., augm entation, A R M  and 
assisted births, recorded in the charts o f  w om en w hose partograph was adequately used. There 
were slightly m ore w om en w ith pyrexia and others readm itted to hospital docum ented in the 
adequately used charts. H ow ever, when the partograph w as adequately used, w om en w ere less 
likely to haem orrhage (i.e., experience a blood loss o f  m ore than 500 m illilitres). W ith respect 
to newborn outcom es, those w hose partographs were adequate w ere less likely to have A pgar 
scores o f  7 or m ore at 1 m inute o f  age. In those charts w here partographs w ere inadequate, 
newborns had few er adm issions to the NICU. N o betw een-group differences in the num ber o f  
birth injuries or neonatal deaths were recorded in the charts o f  w om en w hose partograph w ere 
used adequately. The study also revealed that 472 (25.6% ) out o f  the 1,845 charts review ed 
were adequately filled out in accordance w ith W HO guidelines.
D ata w ere further explored by com paring w hether or not tim ely action w as taken when 
the action line on the partograph was crossed. The action line w as crossed on the partographs 
o f  464 charts. In these, action was taken in a tim ely m anner on 238 (48.7% ) occasions. W hen 
action was taken in a tim ely m anner, 1-m inute A pgar scores w ere higher.
5.2 Discussion of Results
This study was designed to investigate the relationship betw een the use o f  the partograph 
and birth outcom es. A retrospective chart review  o f both prim iparous and m ultiparous
52
University of Ghana          http://ugspace.ug.edu.gh
deliveries from May 01 to June 30, 2003 was conductcd. C harts ol 1,845 w om en who m et the 
inclusion criteria were assigned to two groups; om issions or no om issions based on W HO 
criteria. D ata were analysed and discussed for each o f  the three m ain hypotheses that guided 
the data collection. Each hypothesis had three or four sub-hypothesis and the d iscussion 
covered all these hypotheses.
Hypothesis 1 was highly significant. It clearly revealed that adequate use o f  the 
partograph indicates m ore interventions, in term s o f  oxytocin augm entation, artificial rupture 
o f  m em branes and assisted deliveries. These findings are com pared w ith a previous m ulti­
centre trial conducted by W HO in South East A sia on the use o f  the partograph in clinical 
practice, w hich revealed that com petent use o f  the partograph can save m aternal lives by 
ensuring that labour is closely m onitored and that life-threatening com plications such as 
obstructed labour are identified and treated (M aternal and N eonatal H ealth, 2002). However, 
this was not evident in th is study since only 226(48.7% ) out o f  the 464 w om en w hose labour 
crossed the action line had interventions taken in a tim ely m anner. Even though action was 
taken for the rem aining 238(51.7% ) w om en, they could have been am ong those w om en whose 
partograph was used inadequately.
It was hypothesized (Hypothesis 2) that there will be less in trapartum  and postpartum  
(within 2 w eeks o f  birth) m orbidity  docum ented in the charts o f  w om en w here the partograph 
was used adequately than in those where the partograph w as used inadequately. There were 
slightly m ore w om en who had pyrexia as was recorded on the partograph w hich was 
adequately used. These w om en may have been m onitored m ore accurately; it is also possible
53
University of Ghana          http://ugspace.ug.edu.gh
that those whose charts were inadequately m onitored had their charts w ere filled out after 
delivery.
W om en w ere less likely to haem orrhage if  their partograph was adequately filled out. 
However, slightly m ore o f  these w om en were readm itted to hospital during the first 2 w eeks 
o f  the postnatal period. The health status o f  a w om an m ay determ ine her ability to survive 
pregnancy, labour and delivery as well as puerperium  (Royston and A rm strong, 1989). There 
were 13 m aternal deaths relating to com plications o f  labour and delivery recorded during the 
stud)' period but only 1 chart w as retrieved. This num ber w as insufficient to analyze and 
therefore the relationship betw een m aternal death and the adequacy o f  partograph use cannot 
be assessed. W HO (1994) literature suggested that if  the partograph w as appropriately used, it 
has the potential o f  reducing infant and m aternal m orbidity and m ortality.
It was further hypothesized (Hypothesis 3) that there w ill be less new born m orbidity 
docum ented in the charts o f  w om en where the partograph w as used adequately. There were 
more new borns w ith A pgar score o f  7 or m ore and reduced num ber o f  adm issions to the N IC U  
in those w om en w hose partographs w ere inadequately filled out. No betw een-group 
differences w ere observed in the num ber o f  birth injuries or neonatal deaths w ith in  2 w eeks o f  
birth. This finding differs from  that found in a W HO study in South East A sia in w hich 35.000 
partographs from  eight hospitals were surveyed. They reported a rem arkable im provem ent in 
newborn outcom es in the area o f  intrapartum  foetal deaths, A pgar scores, resuscitation 
m easures and adm ission to neonatal spccial care facilities. It w as o f  interest that there did not 
appear to be an im pact on neonatal health or that it may have even been com prom ised by 
adequate use ol the partograph. However, it m ust be noted that there w ere m ore m aternal
54
University of Ghana          http://ugspace.ug.edu.gh
interventions during labour for the group w hose partographs w ere adequately com pleted.
These assisted births could well have contributed to the higher m orbidity  that was observed.
There did not appear to be a “buy-in” for partograph use in G hana since only 26 
percent o f  the partographs in this study were adequately used. It is possible that partograph 
use in South East A sia w as m ore w idespread so that greater opportunities existed to assess its 
impact on the outcom es o f  m aternal and neonatal health.
5.3 Threats to Validity
In this section, potential threats to the validity o f  this study will be addressed and 
strategies that w ere used to  m inim ize the threats w ill be described. Threats to internal validity 
will be identified so that rival explanations for study findings can be considered. H istory is a 
m ajor threat to the internal validity o f  this study in that data w ere transcribed from  the records 
o f 1,845 available charts. Som e charts were m issing including all but one o f  the 13 w om en 
who died during the intrapartum  period. B ecause o f  this, the num ber o f  m aternal deaths 
(n=29) is likely underestim ated. It is not possible to know  w hether these m issing records 
would alter the relationship  between the adequacy o f  the partograph and m aternal and neonatal 
mortality. A nother threat to history and also selection is that data w ere gathered during the 
two busiest m onths o f  the year (i.e., M ay and June) in that there w ere m ore births and m ore 
m aternal m ortality than during other m onths. It is not possible to assess the im pact that 
increased w orkload may have had on m aternal and neonatal outcom es in a clim ate where 
serious nursing shortages already exist.
The historical threat that is o f  greatest concern is always a concern for epidem iologists. 
Inform ation on the charts w ould be affected by integrity o f  the many individuals who would
55
University of Ghana          http://ugspace.ug.edu.gh
have recorded the inform ation. Entries w ere m ade by providers in a very large national 
hospital as well as by those in sm aller hospitals and polyclinics who m ade a decision to 
transfer their patients to K orle Bu Teaching Hospital. On inspection o f  several o f  the 
partographs, it was clear that all inform ation was recorded at the sam e tim e; w e can speculate 
that this would have happened at the tim e o f  birth, transfer to the tertiary care facility or when 
a labouring w o m an 's  condition  had clearly deteriorated. Since few  providers adequately 
com pleted the partograph (25.6% ) the “buy-in” by providers has not occurred. Finally, w ith 
respect to history, the investigator is not aw are o f  m edia reports or other events during the 
study period that may have influenced attitudes tow ard the independent variable (i.e., 
partograph). B ecause it w as unclear w hether or not even the adequately  com pleted 
partographs w ere used to m onitor labour rather than being com pleted retrospectively, the 
tim eliness o f  in terventions once the action line was crossed w as also assessed.
W hile data entered into patient records was beyond the control o f  the researcher, the 
criteria for adequacy o f  the partograph was established a priori and this criteria w ere applied 
to each partograph that was assessed, thus reducing the possib ility  in random  error when 
determ ining the adequacy o f  the instrum ent. The researcher w as always present when data 
were recorded. W hen she had assistants, inter-rater reliability w as form ally assessed. Any 
atypical records w ere d iscussed and consensus was reached. The sam ple size w as large and the 
entire population o f  eligible charts that could be located w ere included.
It is also im portant to assess the external validity or generalizability  o f  study findings, 
fhe charts o f labouring wom en, including those with atypical pregnancies or m ultiple 
gestations that could have a partograph were included. Only those in second stage labour.
56
University of Ghana          http://ugspace.ug.edu.gh
adm itted for planned instrum ental delivery or where the foetus w as know n to have died in 
utero were excluded. W hile a non-probability sam ple was used, it included the entire 
population for a specific period o f  tim e and was sufficiently large to overcom e selection 
effects.
5.4 Limitations
In addition to threats to both internal and external validity w hich have already been 
addressed, im portant lim itations will be outlined. The design can be described as an After- 
only Non-equivalent Control Group Design (Cam pbell & Stanley. 1966). Ideally, a 
prospective study w ould have been conducted so that m ore control could  be m aintained. 
W om en would have been  follow ed from  the onset o f  labour until 2 w eeks follow ing their 
birth. H ow ever, a p rospective design is not feasible for this study because the use o f  a tool 
(i.e., the partograph) in a real and pre-existing social setting w as being assessed.
N on-param etric tests w ere used in the data analysis as they are based on few er 
assum ptions but are less pow erful (W ood & Ross-K err, 2006). The design for this study is 
quasi-experim ental in that it was not possible to exert further control by m anipulating  the 
independent variable or random izing charts to study groups since the condition already existed 
in nature. Since full experim ental control was not possible, given the nature o f  this study, one 
is less confident o f  “cause and effect” relationships.
Com puterized patient records w ere not used in the study setting so all data were 
recorded by hand from  available paper records. D ata from  data collection sheets were 
transferred to a com puter database. Obviously, errors in both collecting and entering data are 
likely. To m inim ize this type o f  random  error, for the first 10 charts the researcher assessed
57
University of Ghana          http://ugspace.ug.edu.gh
inter-rater consistency in recording the adequacy o f  the partograph, the use o f  interventions, 
and all m aternal and neonatal outcom es by having a colleague w ith about the sam e level o f  
skill and experience also com plete data collection forms. A n agreem ent o f  90 percent was 
accepted. Further, a colleague checked every tw entieth  entry for accuracy.
5.5 Implications for Nursing Practice
There appears to be an association betw een the appropriate use o f  the partograph and 
birth outcom es but m ore needs to be done in the tim ing o f  interventions to help save the lives 
o f  m others and their babies. A lthough the providers used the partographs for all w om en in 
labour most did not use them  adequately or perhaps even prospectively. Som e im portant 
com ponents, for instance, uterine contractions, foetal condition and m aternal pulse w ere not 
assessed and plo tted  correctly. In all, it appeared the partograph w as not correctly filled out as 
specified in the W H O  guidelines.
The partograph serves as an “early w arning system ” and assists in early referral, 
augm entation and term ination. It is therefore im perative that the nurses/m idw ives and 
obstetricians be well inform ed or rem inded o f  their responsibilities regarding labour 
m anagem ent and how they affect birthing outcom es.
5.5.1 Implications for Nursing Education
Periodic in-service education program m es and supervision for all nurses, m idw ives 
and other health care w orkers providing m aternity services to pregnant w om en are o f  
param ount im portance in order to secure correct use o f  the partograph and im prove labour 
managem ent.
58
University of Ghana          http://ugspace.ug.edu.gh
Em phasize the im portance o f  m onitoring and appropriate action  that is needed w hen 
the progress o f  labour, m aternal vital signs, and neonatal assessm ent fall outside acceptable 
variations.
A priority in providing a high standard o f  m aternity care needs to be put in place to 
ensure that students are adequately supervised by highly skilled preceptors or m entors who 
can serve as m odels w ho provide a high level m aternity care.
5.5.2 Implications for Research
The value o f  the partograph as a Safe M otherhood intervention needs to be assessed 
since neonates in this study did not appear to have im proved outcom es and in terest am ong 
providers appeared low.
Q ualitative study to further understand why health  professionals are not com m itted to 
using partograph.
Q uantitative studies to further understand why tim ely action w as not taken w hen the 
action line w as crossed. A vailability  o f  intervention? A vailability  o f  personnel?
5.6 Recommendations
In view  o f  the extensive literature review s on the use o f  the partograph and its effect on 
labour, m ost o f  the research to date has been on prospective studies. The social, econom ic and 
em otional burden that fam ilies face w hen they lose any relative through childbirth  cannot be 
over-em phasized. The researcher w ishes to m ake the follow ing recom m endations
• O rganize continuous educational program m e to update know ledge and skills o f  
nurses/m idw ives in labour m anagem ent and care o f  the new born.
59
University of Ghana          http://ugspace.ug.edu.gh
• D evelop educational m odels that can be delivered to m ore rem ote clinics as well as 
be used by nurses/m idw ives in larger tertiary care centres that will address both the 
com m itm ent needed and skill level necessary to appropriately docum ent m aternal 
progress and m aternal and foetal status using a partograph.
• Develop a sim ilar educational m odel that may be used to assist nurses/m idw ives in 
decision m aking once an action line is crossed.
•  A protocol for appropriate and tim ely action using recom m ended interventions 
needs to be developed and prom inently displayed on the w ards so that attending 
m idw ives can apply them prom ptly and effectively.
• D evelop periodic in-service training for personnel at the M edical S tatistics and 
D ocum entation  U nit (Records departm ent) in the health care facilities to im prove 
the filing system  and ensure easy retrieval o f  patien ts’ charts/folders. M issing 
records m ust be located and appropriately filed.
• Lobby the Reproductive and Child Health (RCH) Unit o f  the M inistry o f  Health to 
consider com puterizing m aternal/child  records as a priority.
• W ork w ith key representatives in the Maternal and Child Health Unit o f  the 
M inistry o f  Health to evaluate the efficacy o f  the use o f  the partograph in G hana 
and identify barriers to appropriate use.
• P resentation o f  findings locally to the tertiary institutions (K orle-B u and Komfo 
A nokye Teaching H ospitals), the RCH unit o f  the various health care facilities. 
N urses’ and M idw ives Council for Ghana and Ghana R egistered N urses/M idw ives
60
University of Ghana          http://ugspace.ug.edu.gh
A ssociations. Further, to w ider A frican audience and W H O  (Safe M otherhood 
Project).
5.7 Conclusion
In sum m ary, the study w hich was a retrospective chart review  sought to assess the 
relationship between the use o f  the partograph and birthing outcom es. The m ethod used has 
inform ation about sam ple, instrum ent and procedure. The data w ere analysed using sim ple 
statistical tools such as percentages, bar charts, m easures o f  central tendency (m eans, m edian), 
variability (ranges, standard deviations) and chi-square. The results o f  the study revealed that 
472 (25.6% ) partographs w ere appropriately filled and tim ely in terventions w ere taken for 
226(48.7% ) w om en w hose labour crossed the action line. Tw o hundred and fifty-nine (13.7% ) 
babies were adm itted to N IC U  w ith varying conditions. The partograph m ay be o f  great value 
in clinical practice today if  it is accurately used by skilled nurses, m idw ives and obstetricians 
to help reduce m aternal and infant m orbidity and m ortality.
61
UniversAityf 'off' fGchNan&a  l X        hftt\p://ugspace.ug.edu.gh
mioMomm P A R T O G R A P H
' Page 8
N a m e ........................................................................................................................................................
Date o f  adm iss ion
University of Ghana          http://ugspace.ug.edu.gh
PARTOGRAPH
DELIVERY AND C O M PL IC A T IO N S
M E T H O D  OF D E L IV E R Y ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DATE A N D  T I M E  O F  D EL IV ERY.
D U R A T I O N  O F  L A B O U R : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C O M P L I C A T I O N S : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
BABY: S EX : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C O N D I T I O N  OF B ABY: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A P G A R  S C O R E :
IMN________ 5INI INS
A PPEA R A N CE  ___________________
PULSE
GRIMACE
ACTIV ITY
RESPIRATION
TOTAL
Delivery By:
University of Ghana          http://ugspace.ug.edu.gh
APPENDIX B 
WHO GUIDELINES IN THE USE OF THE PARTOGRAPH  
A. FOETAL CONDITION
I. FHR record FHR every hour but if  foetal heart is irregular, check it every 15 
m inutes
II. Membranes and liquor - assess and record the state o f  m em branes and liquor on 
adm ission and every 4 hours
III. M oulding - record degree o f  m oulding 4 hourly
B. PROGRESS OF LABOUR
IV. Descent of foetal head -  assess level o f  presenting part on adm ission and every 4 
hours
V. Cervical d ilatation- assess cervical dilatation on adm ission and every 4 hours. 
V aginal exam ination  m ay done frequently if  the need arises
VI. Uterine contractions -  plot uterine contractions h a lf  hourly
C. MATERNAL CONDITION
VII. Blood pressure -  record BP 4 hourly
VIII. Pulse -  record pulse hourly
IX. Temperature -  record tem perature 4 hourly
X. Urine (urinalysis) -  test urine 4 hourly.
63
University of Ghana          http://ugspace.ug.edu.gh
APPENDIX C 
CHECKLIST (Data Collection Sheet)
Participant No.
DEMOGRAPHIC CHARACTERISTICS OUTCOMES
1 2 3 4
1. Dem ographic characteristics
a. Age
b. Parity
0 ’
1
2 - 5  
More than 5
c. Marital status 
married 
single
d Occupation 
Unemployed
Self-employed (inform al private sector) 
Formal private sector 
Public sector
e. Mode o f last delivery 
Non-applicable (N/A)
Normal spontaneous vaginal birth(NSV B) 
Caesarean section (C/S)
Vacuum extraction (V/E)
Forceps delivery (F/D)
f. Gestational age on onset o f  labour
37 weeks or less
38 weeks or m ore
2. Maternal Interventions
a. Oxytocin augm entation Yes
No
b. ARM during labour
Yes
No
c. A ssisted delivery
C/S
V/E
F/D
Fipisiotomy
Breech
64
University of Ghana          http://ugspace.ug.edu.gh
3 M aternal O utcom es
a. Pyrexia above 37.5C
Yes
No
b. H aem orrhage o f  500ml and
above
No haem orrhage -  less than 500ml 
M oderate - 5 0 0 -  1000ml 
Severe - m ore than 1000ml
c. Readm ission to hospital w ithin 2 
weeks o f  birth
d. M ortality Yes
No
4. N ew born O utcom es
a. Apgar scores at 1 m inute o f  age
6 or less 
7 - 1 0
b. Apgar scores at 5 m inutes o f  age
6 or less 
7 - 1 0
c. Birth injury
Cephal haem atom a 
O thers
d. Adm itted to N ICU  
Yes
Yes but after operative delivery' 
No
e. M ortality 
Yes
No
5. Partograph
a. Adequate
No om ission 
M inor om ission
b. Inadequate
M oderate om ission 
M ajor om ission
c. Date and tim e action line was crossed
d. Action taken in tim ely m anner
Yes
No
65
University of Ghana          http://ugspace.ug.edu.gh
REFERENCES
Ackerm an-1,iebrich, U., Voegeli, 1'., & G unter-W ili, K. (1996). H om e versus hospital 
deliveries: follow  up study o f  m atched pairs for procedures and outcom e. British Medical 
Journal 131. 1313-1318.
Addai. I. (2000). D eterm inants o f  use m aternal-child health in rural Ghana. Journal 
Biosocial Sciences 32, 1-15.
Am aniapong, K., Seffah, J.D ., & W ilson, J.B. (1994). The use o f  the partograph in 
m onitoring labour in a prior caesarean section. International Journal o f  Gynaecology and  
Obstetric 45. 281-288.
Ampofo. D. A. (1969). C auses o f  m aternal death and com m ents, m aternity hospital. Accra. 
1963-1967, West African Medical Journal, 75-81.
Annual report, reproductive and child health, (1999). M inistry  o f  Health.
Annual report, reproductive and child health. (2001). M inistry o f  Health.
Annual report, reproductive and child health, (2003). M inistry  o f  Health.
Arruda. J.M ., Covington. D.L.. Janow itz, B., M orris, L. & R odriques, (1985).Caesarean 
delivery in N orthern Region o f  Brazil, 1978-80. American Journal o f  Public Health 75 (5), 
560-562.
Barclay. L., Everitt. L., & Rogan, F. (1997). Becom ing a m other -  analysis o f  w om en 's 
experience o f  early m otherhood. Journal of Advanced Nursing 25. 719-728.
Beaton, J., & G upton, A. (1990). Childbirth  expectations: a qualita tive analysis. Midwifery
6, 133-139.
66
University of Ghana          http://ugspace.ug.edu.gh
Bennett, L. R., & Brown L.K. (1 9 9 9 / Myles textbook fo r  midwifery, (13th Ed) Edinburgh, 
Churchill Livingstone.
Best Practiccs: The partograph, an essential tool for decision-m aking during labour. 
Maternal Neonatal Health, 30 Septem ber, 2002.
Bosse, G., M assaw e, S., & Jahn, A. (2002). The partograph in daily practice: i t ’s quality 
that matters. International Journal o f  Gvnaecolosy and Obstetrics 77 (3), 243-244.
Bouvier-Colle, M .H .. Breart, G „ C halum eau, M., & the M O M A  G roup. (2002). Can risk 
factors for late stillbirth  in W est A frica be detected during antenatal care or during labour? 
International Journal o f  Epidemiology 31, 661-668.
Bulatao, R .A ., & Ross, J.A. (2003). W hich health services reduce m aternal m ortality? 
Evidence from ratings o f  m aternal health services. Troyical Medicine and International 
Health 8 (8), 710-721.
CBC N ew s (2003). Childbirth  deaths rates differ sharply betw een U .S., Canada: UN. Last 
Updated: 21 O ctober 2003.
David, M ., K raker von Schw arzenfeld, K., D im er, J.A .S., &  K entenich, H. (1999). 
Perinatal outcom e in hospital and birth centre obstetric care. International Journal o f  
Gynaecology and Obstetrics 65. 149-156.
Frankenberg, E., & Thom as D. (2001). W om en 's health and pregnancy outcom es: do 
services m ake a difference? Demography 38. 253-265.
Friedm an, E.A. (1975). D isordered labour, objective evaluation and m anagem ent. Journal 
Family Practice 2. 1 67.
67
University of Ghana          http://ugspace.ug.edu.gh
Friedman, E.A .,&  Sachtleben, M.R. (1963). D ysfunctional labour: V Therapeutic trial o f  
oxytocin in secondary arrest. Obstetrics and Gynaecolos\’ 21, 13.
Gerlus, R., M oran, E., & Segu, D. (1999). M aternal m ortality  in developing countries: 
causes and strategies. Fall, 1999.
Ghana D em ographic H ealth Survey, 2003. Population Council Prelim inary Report.
Ghana Statistical Services. 1994. M inistry o f  Health Annual Report. 2003.
Ghana M inistry o f  H ealth, 1990. Maternal and child health/family planning: N ational 
report. Accra: G hana Publishing Corporation.
G ibbins, J., & Thom as, A.M . (2001). W om en 's expectations and experiences o f  childbirth. 
Midwifery 17, 303-313.
Green. J.M . (1993). Expectations and experiences o f  pain in labour: findings from  a large 
prospective study. Birth 20 (2). 65-72.
Groeshel, N .. & Glover, P. (1998). The partograph used daily but rarely questioned. 
Australian College o f  Midwives Incorporated 14 (3), 22-27.
Hertz. E.. Hebert, J.R ., & Landon. J. (1994). Social and environm ental factors and life 
expectancy, infant m ortality  and m aternal m ortality rates: results o f  a cross-national 
com parison. Social Science and Medicine 39, 105-114.
Herz, B., & M easham  A.R. (1987). The safe m otherhood initiative: proposal for action. 
World Bank Discussion Paper, No. 9, W orld Bank. W ashington DC.
Huddart, J.A., Lyons, J.V., & B jerragaard, D. (1999). Integrating reproductive health into 
NGO program m es, volum e 2, safer m otherhood for com m unities. Boston. Initiative Inc.
68
University of Ghana          http://ugspace.ug.edu.gh
Inter-Agency group for safe m otherhood, the safe m otherhood action agenda: priorities for 
the next decade. Report on the safe m otherhood technical consultation , 18-23 O ctober, 1997, 
Colom bo, Sri Lanka.
Kasongo Project Team  (1984). “A ntenatal screening for foetopelvic dystocias. A cost 
effectiveness approach to the choice o f  sim ple indicators for use by auxiliary personnel” 
Journal o f  Tropical Medicine and Hygiene 87  |4 ). 173-183.
Koblinsky, M .A. (1993). M other and more: a broader perspective on w om en’s health. The 
Health o f  Women: A Global Perspective. Boulder, CO: W estview  Press, Inc. pp. 33-62.
Koblinsky, M .A ., C am pbell. O., & H eichelheim , J. (1999). O rganising delivery care: w hat 
works for safe m otherhood. Bulletin o f  the World Health Organisation 77. 399-406.
Kusiako, T., R onsm ans, C., & Van der Paal, L. (2000). Perinatal m ortality  attributable to 
com plications o f  ch ildb irth  in M atlab, Bangladesh. Bulletin o f  the World Health Organisation 
Z 8 (5), 621-627.
Lassey, A .T., & W ilson, J.B. (1998). Trends in m aternal m ortality  in K orle-B u Teaching 
Hospital, 1984-94. Ghana M edical Journal 32a. 910-916.
Lavender, T., & M alcolm son, L. (1999). Is the partogram  a help or a h indrance? An 
exploratory study o f  m idw ives’view. The Practising Midwife 2 (8), 23-27.
Lennox, C .E., & K w ast B.E. (1995). The partograph in com m unity obstetrics. Tropical 
Doctor 25. 56-63.
Lennox, C .E., K w ast, B .E., & Farley, T .M .M . (1998). Breech labour on the W HO 
partoRraph. International Journal o f  Gynaecology and Obstetrics 62. 1 17-127.
69
University of Ghana          http://ugspace.ug.edu.gh
Loudon, I. (1991). On m aternal and infant m ortality 1900-1960. Social History o f M edicine
4. 29-73.
Lowe. N.K. (1996). The pain and discom fort o f  labour and birth. .Journal of Obstetric 
Gvnaecolozic Neonatal Nursing 25 (1). 82-92.
M aine, D., A kalin , M .Z., W ard, V .M ., & K am ara, A. (1997). The design and evaluation o f  
maternal mortality programmes. Centre for Population and Fam ily H ealth, School o f  Public 
Health, C olum bia U niversity.
Maine, D.. & R osenfield, A. (1999). The safe m otherhood initiative: why has it stalled? 
American Journal o f  Public Health 89. 480-482.
Martey. J.O., D jan, J.O ., Tw um , S. Brown, E.N., & Opoku, S.A. (1998). R eferrals for 
obstetric com plications from  Ejisu district Ghana. West African Journal o f  Medicine 17. 58­
63.
M aternal and N eonatal H ealth (2002). The Partograph: An essential tool for decision­
m aking during labour.
M aternal and N eonatal H ealth (2003). Normal labour and child birth presentation graphic 
notes. Last U pdated: 02 July 2003.
M atthews, K. (1999). M atercare prevention program m e trains 34 1'BAs for safe deliveries. 
MCI News and C om m ent quarterly new sletter, Spring 1999, Edition.
M cCourt, M ., & P ierce, A. (2000). Does continuity o f  care m atter to w om en from  m inority 
ethnic group. Midwife 16. 145-152.
M ercer. R. (1986). first-tim e motherhood. New York. Springer.
70
University of Ghana          http://ugspace.ug.edu.gh
M idhet. F., Becker, S., & Berendes, FI.W. (1998). C ontextual determ inants o f  m aternal 
mortality in rural Pakistan. Social Science and Medicine 46, 1587-1598.
M inistry o f  Health (G hana), 2001, A nnual report. Maternal and Child Health and Family 
Planning, MOFI, Accra.
MSH and U N IC EF, (1998). The guide to m anaging for quality: partograph.
N akkazi, S. (1998). Beyond the call o f  duty: m idw ives w ork to im prove care in Uganda 
Hospital. M aternal and N eonatal Health, Solutions for Survival.
National R esearch Council (1997). Reproductive health in developing countries: expanding 
dim ensions, build ing solutions, A.O. Tsui, J.N. W asserheit, & J.G. H aaga, eds. W ashington 
DC. N ational A cadcm y Press.
N etw ork Fam ily H ealth International (1999). M easure evaluation. C arolina Population 
Center.
Nkyekyer, K. (2000). Peripartum  referrals to K orle-Bu T eaching Flospital, G hana -  a 
descriptive study. Tropical Medicine and International Health 5 (11), 811-817.
Papiernik, E. (1995). The role o f  em ergency obstetric care in preventing m aternal deaths: an 
historic perspective on European figures since 1751. International Journal o fG ynaecolosi’ 
and Obstetrics 50, 573-577.
Peters. M. (2000). Safe m otherhood beyond the year 2000: a global perspective. Midwife 
16, 2-7. H arcourt Publishers Ltd.
71
University of Ghana          http://ugspace.ug.edu.gh
Petterson, K.O., Svensson, M .L., & Christensson, K. (2000). Evaluation o f  an adapted 
model o f  the W orld H ealth O rganisation partograph used by A ngolan m idw ives in a peripheral 
delivery unit. Midwife 16, 82-88. H arcourt Publishers Ltd.
Philpott, R.H .. & C astle. W .M . (1972). C ervicographs in the m anagem ent o f  labour in 
primigravida. Journal o f  Obstetrics and Gynaecolog\> o f the British Commonwealth 79, 592­
593.
Philpott, R.H .. & C astle, W .M . (1972b). C ervicographs in the m anagem ent o f  labour in 
primigravida. Journal o f  Obstetrics and Gynaecology o f  the British Commonwealth 79. 599­
603.
Pittroff, R. (1997). O bserved versus expected obstetric com plications: an assessm ent o f  
coverage o f  obstetric care in developing countries. Tropical Doctor 2 7, 25-29.
Preventing M aternal Death: Safe motherhood Newsletter 19 (3), 1995. H arcourt Publishers 
Lid.
Preventing Prolonged Labour: A Practicce Guide. The Partograph Part 1. 
W H O /FH E/M SM /93.8, G eneva, W H 0.1993.
Polit, D .F., & H ungler, B.P. (1997). Essentials o f  nursing research, methods, appraisal and  
utilization. 4 th Ed, Philadelphia, L ippincott-R aven Publishers.
Ronsmans, C., Endang, A., G unaw , S., Zazri, A., M cD erm ott, J., K oblinsky, M „ &
M arshall, T. (2001). Evaluation o f  a com prehensive hom e-based m idw ifery program m e in 
South K alim antan, Indonesia. Tropical Medicine and International Health 6 (10). 799.
72
University of Ghana          http://ugspace.ug.edu.gh
Rooks, J.P., W inikoff, B., & Bruce, J. (1990). Technical sum m ary: sem inar on 
"reassessment o f  the concept o f  reproductive risk in maternity care and fam ily planning  
services ", N ew  York, The Population Council.
Rooks, J.P., W eatherby, N.L., Ernst, E .K .M ., Stapleton, S., R osen, D., & R osenfield, A. 
(1989). O utcom es o f  care in the birth centres. New England Journal o f  Medicine 327, 1804­
1811.
Rosenfield, A. (1992). M aternal mortality: com m unity based interventions. International 
Journal o f  Gynaecology and Obstetrics 38, 17-22.
Royston, E.. & A rm strong. S. (1989). Preventing maternal death. G eneva: W HO.
Rubin. R. (1984). M aternal identity and the maternal experience. N ew  York: Springer.
Seffah, J.D. (2002). Ruptured uterus and the partograph. International Journal of  
Gynaecology' and Obstetrics 8U, 169-170.
Shiffm an, J. (2000). Can poor countries surm ount high m aternal m ortality? S tudies in 
Family Planning 31. 274-288.
Sinclair. M. (2001). M idw ives’ attitudes to the use o f  the cardiotocograph m achine. Journal 
o f Advanced Nursing 35 (4), 599-606.
Slen, C., & W illiam son. J.B. (1999). M aternal m ortality, w o m en 's  status and econom ic 
dependency in less developed countries: a cross-national analysis. Social Science and 
Medicine 49. 197-214.
Sloan. N .L., W inkol'f, B., & Fikree, F.F (2001). An ecological analysis o f  m aternal 
m ortality ratios. Studies in Family Planning 32, 352-355.
73
University of Ghana          http://ugspace.ug.edu.gh
Smeenk. A .D .J., & ten Have, H .A .M .J. (2003). M edicalisation and obstetric care: an 
analysis o f  developm ents in Dutch m idwifery. Medicine, Health Care and Philosophy 6, 153­
165,
Sjogren, B. (1997). R eason for anxiety about childbirth  in 100 pregnant wom en. Journal o f  
Psychosomatic, Obstetrics and Gynaecology 18. 266-272.
Starrs, A. (1998). The safe motherhood action agenda: priorities for the next decade.
Tarkka. M .T., 7Paunonem , M. (1996). Social support and its im pact on m others’ 
experiences o f  childbirth. Journal o f  Advanced Nursing 23, 70-75.
Thornton. J.G ., & Lilford, R.J. (1994). A ctive m anagem ent o f  labour; current know ledge 
and research issues, BM J 309, 366-369.
Tinker, A. (1997). Safe m otherhood as an econom ic and social investm ent. P resentation at 
Safe m otherhood Technical C onsultation in Sri Lanka, 18-23 O ctober 1997.
Tinker, A., & K oblinsky, M .A. (1993). M aking m otherhood safe. World Bank Discussion 
Paper, No. 202, W ashington DC: W orld Bank.
Tw um -Baah, K.A., N yarko, P.E., Q uarshie, S.E., Caiquo, I.B., & A m uah, E. (1994). Infant, 
child and m aternal m ortality  study in Ghana, Accra, Ghana: GSS, U N IC EF, M OH.
W alley, R. (2002). Safe m otherhood and the “conspiracy o f  silence” Rom e, ZENIT 5-6 
December 2002.
W eatherby, N. (1990). "A ntenatal assessm ent o f  intrapartum  risk: national birth centre 
study". Paper presented at sem inar o f  reassessm ent o f  the concept o f  reproductive risk in 
m aternal care and fam ily planning services. New York: The Population  Centre.
W om en’s Health Foundation Brochure, 2003.
14
University of Ghana          http://ugspace.ug.edu.gh
W orld Bank (1993). World Development Report: Investing in health, W ashington.
W orld Health O rganisation. Maternal Mortality Rates, 1986.
W orld Health O rganisation (1988). The partograph, section I. The principle and strategy, 
W HO/M CH/88.3. D ivision o f  Fam ily Health. Geneva.
W orld H ealth O rganisation (1989). The partograph. section IV, guidelines for operations 
research on the application  o f  the partograph. W H O /M C H /89.1. D ivision o f  Fam ily Health. 
Geneva.
W'orld H ealth O rganisation. W HO partograph reduces com plications o f  labour and 
childbirth. Press Release W H O /45. 7 June 1994.
W orld H ealth  O rganisation. W HO partograph in m anagem ent o f  labour. Lancet 1994: 343: 
1399-1404.
W orld H ealth O rganisation (1994). The partograph: the application o f  the W H O  partograph 
in the m anagem ent o f  labour. Report o f  a m ulticentre study 1990-1991.
W I-IO/FHE/M SM /94.4. D ivision o f  Fam ily Health, Geneva.
W orld H ealth O rganisation, (1996). Revised 1990 estim ates o f  m aternal mortality: a new 
approach by W HO and U N IC EF W H O /FR H /M SM  96.11.
World Flealth O rganisation, (1997). M aternal health around the w orld. W all Chart, W HO, 
Geneva.
75