• No results found

From Department of Medicine, Solna Karolinska Institutet, Stockholm, Sweden SAFE LABOR - REAPPRAISAL OF LABOR DURATION AND MANAGEMENT FOR SAFE MATERNAL AND INFANT OUTCOME

N/A
N/A
Protected

Academic year: 2022

Share "From Department of Medicine, Solna Karolinska Institutet, Stockholm, Sweden SAFE LABOR - REAPPRAISAL OF LABOR DURATION AND MANAGEMENT FOR SAFE MATERNAL AND INFANT OUTCOME"

Copied!
103
0
0

Loading.... (view fulltext now)

Full text

(1)

From Department of Medicine, Solna Karolinska Institutet, Stockholm, Sweden

SAFE LABOR - REAPPRAISAL OF LABOR DURATION AND MANAGEMENT FOR SAFE MATERNAL AND INFANT OUTCOME

Louise Lundborg

(2)

All previously published papers were reproduced with permission from the publisher.

Published by Karolinska Institutet.

Printed by Universitetsservice US-AB, 2022

© Louise Lundborg, 2022 ISBN 978-91-8016-546-4

Cover illustration: Ture Lundborg

(3)

Safe Labor- Reappraisal of labor duration and manag ement for safe maternal and infant outcome

THESIS FOR DOCTORAL DEGREE (Ph.D.)

By

Louise Lundborg

The thesis will be defended in public at CMB, Karolinska Institutet, Solna Friday May 13th, 2022 at 09.00 am.

Principal Supervisor:

Mia Ahlberg PhD Karolinska Institutet Department of Medicine Clinical Epidemiology Division

Co-supervisor(s):

Professor Olof Stephansson Karolinska Institutet

Department of Medicine Clinical Epidemiology Division

Xingrong Liu PhD Karolinska Institutet Department of Medicine Clinical Epidemiology Division

Associate Professor Anna Sandström Karolinska Institutet

Department of Medicine Clinical Epidemiology Division

Opponent:

Associate Professor Stine Bernitz Oslo Metropolitan University Department of Nursing and Health Promotion

Examination Board:

Professor Susanne Georgsson

Swedish Red Cross University/Karolinska Institutet

Department of Clinical Science, Intervention and Technology

Professor Helena Lindgren Karolinska Institutet

Department of Women’s and Children’s Health

Associate Professor Andreas Herbst Lund University

Department of Obstetrics and Gynaecology Institution of Clinical Sciences

(4)
(5)

ABSTRACT

Safe Labor - Reappraisal of labor duration and management for safe maternal and infant outcome

Background:

The partograph, today used worldwide for assessing normal labor progress and identifying prolonged labor is based on Friedman’s research findings from the 1950s, based on 500 deliveries. Friedman established a series of definitions of time

parameters, labor curves and phases to detect normal and abnormal labor progress and incorporated it into the Partograph. However, since the early 2000s, there is increasing evidence suggesting that the described relationship between cervical dilatation and duration of the first stage of labor may be inappropriate, which could be due to a misinterpretation of Friedman’s and others’ findings. Importantly,

contemporary changes in maternal demographics and obstetrical management raise serious concerns to whether Friedman´s definitions can be applied to contemporary childbearing women. Improved understanding of correct thresholds for what should be defined as normal labor progression and duration is important to improve every aspect of childbirth. Therefore, the overall aim of this thesis was to establish updated and evidence-based measures of normal labor duration, focusing on labors with spontaneous onset.

Methods: All four studies were population-based cohort studies with data from the Stockholm Gotland Obstetric cohort. The study population was term gestations with spontaneous labor onset. In Study I we investigated duration of the first stage of labor and trajectory of cervical change in both nulliparous and multiparous women.

Study II investigated the association between body mass index (BMI) and first stage labor duration, the outcome was first stage labor duration and the analysis were stratified by BMI class and tested for effect modification of age. The study population

(6)

delivery. In Study IV the relationship between increasing active first stage of labor duration and adverse neonatal outcomes was examined. The study population was term gestations with spontaneous labor onset, allocated to Robson group 1. Here, the investigated outcome was a composite of either moderate or severe outcomes.

Modified Poisson regression was used to test the association at different thresholds of labor duration and relative risks with 95% confidence intervals (CI) were

calculated.

Main findings: Labor progression and duration varied largely between women, indicating that using measures of central tendency is not useful to identify slow labor.

Labor does not seem to accelerate until beyond a cervical dilation of 5-6 cm. Obese women had longer active first stage and total active labor duration and this

association was modified by age. Increasing duration of active first stage of labor was also associated with a linear increase of duration of second stage of labor until a plateau at 12 hours of first stage duration. Risk for adverse neonatal outcomes was increasing along with a cumulatively increasing duration ranging from 5-10 hours, we tested the association in relation to two composite neonatal outcomes, based on clinical knowledge of severity and risks of long-term consequences. The risk for moderate neonatal outcome started to increase from 5.1 hours (aRR 1.40, CI 1.24, 1.58) and were twofold increased for women beyond the 90th percentile (10.1 hours).

Moreover, among severe neonatal outcomes the risk was found significant beyond the 90th percentile (10.1 hours, aRR 1.53, CI 1.26, 1.87). Effect decomposition showed that only approximately 1/5 of the found association was mediated by second stage of labor duration for moderate neonatal outcomes. For severe outcomes with a labor duration beyond the 90th percentile (10.1 hours),

approximately 13 % of the association was mediated through second stage of labor duration.

Conclusions: Variations in both the total duration and the trajectory of cervical change over time was large and the progressive pace of cervical dilation is found around cervical dilation of 5–6 centimeters. Labor duration is a continuous process, BMI and Age and labor duration during the active first stage will have impact on labor duration and interventions in second stage. The biological interpretation is that labor duration and progression is a function of observed and unobserved maternal and neonatal anthropometrics and management norms, which cannot be

constrained into a model where one duration or one pace fits all. Adverse neonatal

(7)

outcomes are rare, and any implications of a duration threshold for when the

absolute risk of an event occurs is not possible to establish, however the relative risk increased with increasing duration at different examined percentiles, i.e. 5, 7 and 10 hours of labor duration.

Implications: The findings from all studies within this thesis may safeguard against potential unintended fatal consequences of practice change based on average separate thresholds for slow progression during first and second stage of labor. Vice versa, potential unintended consequences of a “better safe than sorry regime” needs to be balanced against the risk of long-term consequences for both mother and neonate with terminating labor during due to slow labor per se. Future research should aim to identify those women and fetuses who are vulnerable to longer labor durations to provide clinicians and pregnant women with information about who can continue with labor and who´s labor that should be terminated earlier. Consequently, it would also be beneficial with more research emphasis on the etiologic causes for slow labor to customize any use of interventions during childbirth.

(8)

LIST OF SCIENTIFIC PAPERS

I. Lundborg L, Åberg K, Sandström A, Discacciati A, Tilden L E, Stephansson O, Ahlberg M

First stage progression in women with spontanous onset of labor: A large population-based cohort study

Plos One. 2020 Sept 25; 15(9)e0239724

II. Lundborg L, Åberg K, Liu X, Sandström A, Tilden L E, Stephansson O, Ahlberg M

Association of body mass index and maternal age with first stage duration of labour

Scientific Reports. 2021 July 5; 11(1):13843

III. Lundborg L, Åberg K, Sandström A, Liu X, Tilden L E, Stephansson O, Ahlberg M

Association between first and second stage of labour duration and mode of delivery: A population-based cohort study

Paediatric and perinatal epidemiology 2021 December 29.

doi:10.1111/ppe.12848

IV. Lundborg L, Åberg K, Liu X, Sandström A, Tilden L E, Bolk J, Ladfors L, Stephansson O, Ahlberg M

Labour duration and risk of adverse neonatal outcome In manuscript

(9)

CONTENTS

1 INTRODUCTION ... 1

1.1 Background and clinical context ... 1

2 LITERATURE REVIEW... 3

2.1 Normal labor and birth ... 3

2.2 Progress of labor ... 3

2.3 Spontaneous start, stages and phases of labor ... 4

2.4 Onset of labor ... 5

2.5 Latent phase ... 5

2.6 Active phase ... 6

2.7 The Partograph ... 8

2.8 Contemporary research on definition of active phase and labor progression ... 11

2.9 Interventions used during labors with spontaneous onset ... 16

2.9.1 Active management of labor ... 16

2.9.2 Amniotomy ... 17

2.9.3 Natural and synthetic Oxytocin ... 18

2.9.4 Epidural ... 19

2.9.5 Cesarean delivery ... 19

2.9.6 Robson classification ... 20

2.10 Maternal anthropometrics influence on the course of labor ... 21

3 RESEARCH AIMS ... 23

3.1 Specific aims:... 23

4 MATERIALS AND METHODS ... 25

4.1 Setting ... 25

4.2 Data sources ... 25

4.2.1 The Stockholm-Gotland Obstetric cohort ... 25

4.2.2 The Swedish Neonatal Quality Register... 26

4.3 Study population and study designs ... 26

4.3.1 Overview ... 26

4.3.2 Overview and detailed description for the definition of active first stage of labor, used as both an outcome and exposure of interest in this thesis ... 29

4.4 Statistical overview ... 31

(10)

6.1.2 Confounders ... 39

6.1.3 Directed Acyclic Graphs ... 39

6.1.4 Effect modification ... 41

6.1.5 Systematic and random errors overview ... 42

6.2 Methodological considerations for each study ... 43

6.2.1 Study I ... 43

6.2.2 Study II ... 45

6.2.3 Study III ... 47

6.2.4 Study IV... 49

7 RESULTS ... 55

7.1.1 Study I ... 55

7.1.2 Study II ... 57

7.1.3 Study III ... 58

7.1.4 Study IV... 61

8 DISCUSSION ... 63

8.1 Summary of results and interpretation ... 63

8.1.1 Study I ... 63

8.1.2 Study II ... 63

8.1.3 Study III ... 64

8.1.4 Study IV... 64

9 CONCLUSIONS ... 67

10 POINTS OF PERSPECTIVE ... 69

11 POPULÄRVETENSKAPLIG SAMMANFATTNING ... 71

12 ACKNOWLEDGEMENTS ... 75

13 REFERENCES ... 79

(11)

LIST OF ABBREVIATIONS

ACOG American College of Obstetricians and Gynecologists

aOR Adjusted odds ratio

BMI Body Mass Index

BD Base deficit

BE Base excess

CI Confidence interval

CD Caesarean delivery

CPAP Continuous positive airway pressure

CS Cesarean section

CTG Cardiotocography

DAG Directed acyclic graph

FHR Fetal heart rate

GEE Generalized estimation equation HIE Hypoxic ischemic encephalopathy

ICD-9 and ICD-10 International Classification of Diseases, 9:th, 10:th revisions

MAS Meconium aspiration syndrome

MVU Montevideo units

NICE National Institute for Health and Clinical Excellence NICU Neonatal intensive care unit

pH Power of hydrogen

RCT Randomized controlled trial

OR Odds ratio

RR Relative risk

SBF Swedish Midwifery Association

(12)
(13)

1 INTRODUCTION

1.1 BACKGROUND AND CLINICAL CONTEXT

“How long will it take before our baby is born” Over the years, working in the delivery clinic as a midwife, I have been asked this question thousands of times. My answer usually was something along the lines: “We do not know, but I can promise you a baby in the end of this journey”. Later in my career, I attempted to give a more scientific answer to the awaiting parents, using a prognostic equation based on the standard definition I had been taught in midwifery school: “it should take

approximately 1 hour for your cervix to dilate 1 cm, and if it’s going too slowly we can help speeding up the pace with different interventions, such as rupturing the

membranes or providing you with intravenous oxytocin “. This approximation for progress of labor is based on the partograph, a tool used to monitor labor progress.

It is known that the duration of labor, or the time between the onset of labor and the birth of the baby, is an important factor that impacts mother, baby and obstetric management. However, decades of research and universal use of the partograph, indicate that the factual impact of labor duration on delivery outcomes may to some extent have been misunderstood by obstetricians and midwives.

After some years of practice in the delivery ward, I started to reflect on the fact that many laboring women did not follow the predefined pattern described in the

partograph and in such cases we tried to “fit” women into the standardized

progression pace, exposing them to interventions since we were taught that waiting with interventions could lead to labor arrest. There were cases when I could sense that a woman was progressing in labor and then finding myself disappointed when the woman’s cervical dilation had not progressed in line with the partograph. It could be frustrating to describe to colleagues outside of the delivery room that “she is progressing but her cervical dilation is not” For example, it was not unusual to find women dilating “too slowly” around 3 cm but later “very fast” at 8 cm. Over the years, I have had endless discussions with colleagues in the delivery wards, reflecting on the timelines in the partograph and the effect it has in clinical practice. Consequently, engaging in a project aiming to examine labor progress among women today and how labor duration actually impacts obstetric and perinatal outcomes laid very close to my heart.

(14)

exist despite a lack of evidence that support maternal and neonatal benefits with rates above 19 % (15, 16).

In health care, the term “unwarranted variation” is defined as an unexplained

difference in use of health care that cannot be attributed to differences in population characteristics (17). Cesarean delivery frequency is a well-established estimator for quality of obstetric care. The increasing trend in caesarean delivery rates is

multifactorial and concerns have been raised regarding discrepancy not only between countries but also within countries.

Notably, the large variations in-between regions in Sweden remain after adjustment for confounding factors (i.e. parity, mode of birth) and for case mix. This indicates that there is an overuse of cesarean delivery in obstetric care also in Sweden, despite a relatively low proportion compared to other high-income countries (5, 18- 25). Since the single most common indication for cesarean delivery is failed

progression/slow labor, it became important to do a reevaluation of the data that had represented the standard for normal labor duration for more than five decades. Thus, if the thresholds for what is considered to be a prolonged or abnormal labor duration are inaccurate, it will by default increase the frequency of caesarean deliveries on a population level (5, 23, 26). Therefore, the intriguing question in this project became:

For the mother and infant, how long is too long during active labor?

(15)

2 LITERATURE REVIEW

2.1 NORMAL LABOR AND BIRTH

Low-risk singleton pregnancy with spontaneous onset of labor, low risk at the start of labor and remaining so throughout labor and birth. The infant is born spontaneously in the vertex position between 37+ and 41+6 (weeks+days) of pregnancy. After birth, mother and infant are in good condition.

This is the definition of normal labor and birth according to the WHO (27). However, in reality, there is no standard definition, what is considered normal in some settings is defined differently in others. Normal labor and childbirth include a variety of factors that needs to be in place to achieve the goal; a healthy mother and a healthy baby.

One important factor impacting the mother, the baby and the obstetric management is time. The increased medicalization of childbirth over the last three decades, where slow labor duration is a driving factor, has highlighted the distinction between what could be considered normal and abnormal labor duration.

In 2018, WHO revised their definition of a normal labor and birth to include both service delivery and the woman´s experience. The new guideline’s goal is to achieve the best possible physical, emotional and psychological outcomes for mother and infant. It also articulated a more woman-centered care, encouraging women to both choose their birth position, keep a companion with them throughout labor as well as a recommendation to avoid unnecessary medical interventions, provide respectful care and accurate information about options for pain relief. Good-quality and

evidence-based care should not undermine the woman´s own capability to give birth, it should rather consider the complexity and diverse nature of different models of care. Furthermore, a strong implication to reduce high rates of unnecessary interventions and cesarean deliveries is also stated in the recommendation (28).

2.2 PROGRESS OF LABOR

The time that elapses between the onset of labor and the birth of the baby is

commonly described as labor progression. To describe this time, different concepts are applied, duration and length are interchangeable and refers to a total time, e.g. 3 or 6 hours. Progression, on the other hand, is mainly used to describe trajectories and patterns of change in duration of labor (i.e. the time from 3 cm dilatation to 4 cm

(16)

2.3 SPONTANEOUS START, STAGES AND PHASES OF LABOR

Historically, and reported in literature since the 18th century, based on advancing knowledge of anatomy, the labor has been divided into three stages, the first stage of labor is described as the opening of the cervix, ending at full cervical dilation, the second stage is the descent phase for the baby into the birth canal and birth of the baby while the third stage includes the delivery of the placenta (28), as depicted in Figure 2.

Figure 2. Stages of labor (illustration downloaded from istockphoto.com)

In the 1950´s Friedman published the first scientific evidence on labor progression, describing the relationship between duration (hours) of labor and cervical dilation.

The statistical observations of 500 nulliparous women were compiled into a

spreadsheet where the duration of labor was plotted on the X-axis with the cervical dilation on the Y-axis (Figure 3). Along with dividing labor into specific stages and phases, the relationship between duration of labor and cervical dilation was

represented graphically as a sigmoid curve. Herby, it introduced an exact index for understanding of labor progression (30, 31). Using this curve, the first stage of labor was then further divided into latent and active phase of labor, with the active phase being characterized by a more rapid labor progression. In the Friedman days, it was considered normal with a labor duration of 24 hours for nulliparous women in

comparison to the 12-hour cutoff that is applied today (30-33).

(17)

Figure 3. The Friedman curve for normal labor(30) 2.4 ONSET OF LABOR

Labor onset is difficult to describe, since it could take on different expressions in different women. Historically, it has been defined as either the beginning of painful contractions, rupture of membranes or, in cases where the women is already admitted to hospital, the first clinical assessment. In recent times, women´s self- reported time of labor onset have been used in research. The definition of onset of labor is important since it is the first known timepoint for start of labor and the subsequent phases of labor (34-40). A large systematic review of defining onset of labor, including studies from 1978-2014 stated that there is a little consensus regarding definition of labor onset in research literature and that more research is needed in order to understand consequences of different definitions for maternal and neonatal morbidity (41).

2.5 LATENT PHASE

Latent phase, sometimes referred to as “early labor”, starting with the onset of labor is a stage of labor characterized by slow cervical change ranging from 0-3 cm (0-6 cm in more seminal findings which will be later described) and almost complete effacement (the thinning out) of the cervix, while active phase includes a significant cervical change from 3 cm dilation and onwards according to the research by

(18)

change are criteria for the transition from latent to active phase. This threshold to categorize women into being in latent or active phase is of importance not

necessarily for the woman in labor, but it has become a benchmark for determining the level of care (such as triage for access to hospital care) (14, 33, 34, 44-46). In 2018, the WHO published two large systematic reviews, evaluating the level of evidence of research on labor duration, including latent phase duration (33, 47). The findings from these reviews indicate very low-certainty evidence, reporting a median duration of the latent phase of 6.0-7.5 hours and a mean duration of 5.1-7.1 in nulliparous women with spontaneous onset (46, 48). These findings are in line with the duration of latent phase reported by Friedman who reported a mean of 8.6 hours and a median of 7.5 hours for nulliparous women and corresponding for parous women mean (5.3 hours), median (4.5 hours) and maximum (13.6 hours) (33, 47).

Changes in maternal characteristics in combination with raising cesarean rates and improved statistical methods have challenged the early findings of Friedman and others and the current consensus is that normal latent phase lasts longer than previously observed.

2.6 ACTIVE PHASE

Most women are admitted to the hospital for care during the active phase.

Convincing evidence over the last decade indicate that this previous sectioning of phases is inaccurate, proposing that the active phase, where the labor seems to rapidly progress, is more likely at 5-6 cm dilation and not at 3-4 cm as previously anticipated. In addition, absence of cervical dilation for 4 hours may be normal in latent phase, (<6cm) but is considered prolonged after a dilation of 6 cm for

nulliparous women, and perhaps even later for multiparous women (28, 33, 34, 49, 50). To compare research on labor progression is complex because of non-

consensus in defining onset of labor, the transition between phases including the start of active phase and inherited limitations due to different management routines in clinical settings based on the partograph (41).

Local and national guidelines use different definitions for onset of the active phase of labor and many have been updated over the last years transitioning from only describing centimeters of dilation to include other parameters such as effacement, regularity and intensity of contractions and descent of the fetal head. Table 1 shows an overview of the most commonly used definitions of the start of active phase of labor.

(19)

Sweden WHO ACOG NICE

Current

Regular painful contractions in combination with a cervical dilation of 5 cm (2021)

Progress is evaluated through an overall assessment of the women´s own experience of labor, well-being or the midwife/obstetrician’s observations of the course of labor.

Regular, painful uterine

contractions, a substantial degree of cervical

effacement and more rapid cervical

dilatation from 5 cm until full dilatation for first and subsequent labours (2018)

Cervical dilation of 6 cm and beyond (2014)

Progressive cervical dilation (2014)

Previous

2/3 criteria:

• Cervical dilation of 4 cm or complete

effacement and dilation > 1 cm

• ≥ 2 regular, painful contractions every 10 minutes

• Rupture of the amniotic membranes

Cervical dilation of 3 cm (1994) and 4 cm (2007) and beyond

Cervical dilation of 3-4 cm and beyond (2003)

(20)

Noteworthy, both median and mean durations have previously been reported which impacts duration measurements. The only study published, actually examining accuracy of different starting definitions is an article by Harper (51) showing that the definition of onset of active phase of labor did not significantly impact maternal and neonatal morbidity (51). Despite a variety of definitions used, identified by both clinicians and researchers, this is not necessarily in conjunction with women’s perceptions of progression during labor. Women have reported that they consider labor as a continuous process and being in different phases is more something they are informed about from their midwife (40).

Table 2. Overview of how different studies define the start and end point for active first stage of labor. From the Systematic review by Abalos, Oladapo (33) 2018.

2.7 THE PARTOGRAPH

In 1972 Philpott (52), in line with previous work by Friedman, introduced the cervicograph, where two lines were added to Friedman’s graphic presentation of labor progression (Figure 3) (52, 53). These two lines became the well-established labor progression lines, the “alert” and “action” lines. The “alert line” was initially created as an early indication tool to identify women in need of transfer to suitable locations at which obstetric care was available. The alert line represented a pace of 1 cm cervical dilation/hour, a landmark reference that is still in use today. Parallel to

(21)

the alert line, a second line was drawn 4-hours to the right, named the “action line”, used to indicate “slow labor progression”. The purpose of the action line was to identify “slow laboring women” i.e. labor dystocia, in need of interventions such as amniotomy, oxytocin or instrumental delivery (52, 53). Philpott found that almost 11% of the included women in his study crossed the action line (i.e. women who progressed too slowly. Aim with the use was reducing the incidence of labor dystocia and consequently the risk of caesarean delivery and perinatal mortality (53).

Figure 4. The WHO partograph

(22)

documenting the descent of the presenting part of the fetus, maternal and fetal vital parameters and both medication and tools for augmentation of labor, i.e. amniotomy or oxytocin during labor. Use of the partograph to monitor labor progression has been promoted globally by the WHO since the 1990s as part of the Safe Motherhood initiative to reduce maternal and fetal mortality (56). This endorsement was based on results found in a multicenter study including 35 484 low-risk, term women with spontaneous onset. The study concluded that using the partograph could reduce both prolonged labor (>18 hours; reduced from 6.4 % to 3.4 %) and the proportion of labors with augmentation (20.7 to 9.1%), emergency cesarean deliveries (9.9% to 8.3%) and stillbirth (0.5 to 0.3 %) (56, 57). Interestingly, the mean duration of labor was only marginally changed, (5.72 to 5.05). Based on this, a routine use of the partograph was recommended in all obstetric settings. The aim with the WHO- partograph initiative was to record all essential observations in labor, i.e.

contractions, fetal head descent, maternal parameters in a routine manner to

improve the timing of critical management decisions during labor. Hence, partograph use was introduced in tandem with recommendations to use local guidelines and protocols for individual interventions during labor (56, 57). The WHO changed their definition of active labor onset from 3 to 4 cm in 2007. Modified versions of the WHO-partograph have since been introduced worldwide with action lines at other intervals, for example two or three hours to the rights of the alert line. In Sweden, a partograph with a three-hour action line is used, accompanied with guidelines for interventions (12).

As mentioned, the intention of using the partograph is to identify women at risk and to act or intervene when needed but this requires adequately trained midwives and obstetricians with knowledge of how to use the tool, and a capacity to act in

accordance with intervention guidelines (58). Use of the partograph has been reported to be challenging in both low and high resource settings (59, 60).

Inadequate training in using the partograph, described by birth attendants in low- resource settings, has led to lack of confidence in using the tool (61) and challenges in initiating necessary interventions due to inaccessible resources (59). The opposite pattern was found in a study from Sweden where 42.6% of low-risk women who were treated with oxytocin for augmentation of labor had partograph patterns that did not indicate any need for interventions (62). A retrospective completion of

partographs due to fear of litigation has also been reported from both high and low resource settings (59). A Cochrane review from 2018, evaluating the effect of the partograph to monitor normal and abnormal labor progression, concluded that its effect is uncertain and that there is no clear evidence that one partograph’s design is superior to another. For future trials, it was emphasized that the clinical environment in which the partograph will be used, should be taken into consideration in the design to provide less heterogeneity within the results (63).

(23)

A diagnostic accuracy test of the 1 cm/h threshold based on a systematic review of more than 17,000 women, showed that the threshold/alert line is suboptimal for identifying women at risk and that this approach both led to false negative findings and false positive findings which could lead to unnecessary labor interventions that may even be potentially harmful. (64). Although use of the partograph shows no clear positive effect on outcome for mother and infant, it represents the only alternative for labor progression monitoring. In 2018, the WHO concluded in their guideline Intrapartum care for a positive childbirth that an updated monitoring tool for documenting and evaluating labor progression is needed (11).

2.8 CONTEMPORARY RESEARCH ON DEFINITION OF ACTIVE PHASE AND LABOR PROGRESSION

After the launch of the partograph, research was sparse in terms of evaluating labor progression and duration. However, in the late 1990s a few studies were published, questioning the pioneer work of Friedman and indicating that the first stage of labor may be longer than presented by Friedman (32, 65). Over the last two decades, with more advanced statistical methods, new research on labor duration and progression has emerged. Zhang et al (66-68) have published a series of articles on labor

progression challenging the historical definitions. These studies from the Consortium on Safe Labor focused on 62 000 women singleton deliveries at term in vertex presentation with vaginal deliveries and normal perinatal outcomes and were based on a contemporary cohort of women at 19 different US medical centers during 2002- 2008. In this selected cohort, 84 % of the women used epidural and 47 % received augmentation of labor which reflects current obstetrical management in the US (69, 70).

According to Zhang (69) the median duration of active stage of labor for nulliparous women, admitted at 4 cm, was 5.3 hours and that the 95th percentiles was at 16.4 hours, which is slower than the rates presented by Friedman (55). Moreover, Zhang reports that labor is slower between 3-4 cm than between 5-6 cm, suggesting that 6 cm dilation seems to be the transition point from latent to active phase where labor accelerates faster. Interval censoring regression for calculating traverse time for every centimeter’s dilation (3-4, 4-5 etc.) was used and presented for the 50th and 95th percentile distribution of duration. Zhang and colleges fitted a model with a log normal distribution and interval censored time-to-event data among those having a

(24)

The shape of the curve also appeared to have a less clear inflection point compared to the labor curve presented by Friedman. Figure 6 shows a graphical overview of Zhang and Friedmans curves, along with data from Study I in this thesis.

(25)

Figure 6. Graph comparing labor progression curves for both nulliparous and

multiparous women with data from Friedman, Zhang and Lundborg (31, 42, 72). This graph is constructed for graphical presentation in this thesis and is not presented in any of the studies.

Similar slower rates and a flatter labor in line with the curves by Zhang was found by Harper et al (51). A systematic review by Neal (73) found that labor duration does not follow a normal distributed curve but rather that a more hyperbolic labor curve may improve clinical decision making and concluded that today’s labor dystocia definitions may be too stringent for some women . Based on the evident variations in duration it was suggested in a document from the American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine that a more relaxed approach to prolonged labor during the first stage of labor should be

implemented in obstetric care in the US (74).These results led to changes in US guidelines for defining onset of active phase of labor, aiming to reduce the increasing numbers of primary cesarean deliveries, i.e. the first operative delivery for a

nulliparous woman (74).

(26)

The new ACOG guidelines were tested in a pre and post implementation study in France in 2014. The cutoff point for active phase was moved from 4 to 6 cm which had implications for the definition of labor dystocia. In this study the median duration of labor before cesarean delivery was 480 min (240-660) pre-implementation vs 600 min (300-840 p-value <0.01) in the post- implementation period and a significant decrease in the rates of both cesarean and operative vaginal delivery for nulliparous women was found with no simultaneous increase in short-term maternal and fetal morbidity (75). Zhang’s publications were the first of many taking advantage of more advanced and contemporary statistical modeling. To address speculations that arose, that differences in characteristics could account for the differences in the shape of the curve and the median time, Zhang used the same modeling techniques in the cohort of women used by Friedman (67). Despite anthropometric differences in populations, the new curves based on repeated-measures and polynomial modeling were similar in shape. However, duration of the active first stage of labor was different between the populations. Some of the issues debated with new research is that all retrospective analysis to some extent are prone to include some selection bias. One criticism leveled was that when a proportion of women with a

“slow labor” is excluded from the analysis, the duration for those left in the cohort will appear shorter than otherwise. This has been discussed in several studies using this statistical model (29, 76, 77).

The Zhang estimated labor curve was further evaluated in the Labor Progression Study (LAPS) pragmatic trial in Norway (3). The LAPS trial took place during 2014- 2017, randomly enrolling fourteen obstetric units to either a control group (WHO partograph) or an intervention group (Zhang’s guideline) with the primary endpoint being rates of intrapartum cesarean section (ICS). Zhang’s guideline was a modified version of the labor progression curves described in the findings by Zhang(69).

Results from the LAPS trial showed no significant difference in the frequency of ICS between the control and intervention groups, however, a decrease in the frequency of ICS in both groups during the study period was found, explained by a close focus on assessing labor progression more than the use of a particular guideline (3). The unadjusted median duration during active first stage of labor was longer for women in the intervention group, median 6.1 h (95th percentile 15.0) compared to 5.0 h (95th percentile 12.5) in the control group. The adjusted median duration was 7.0 and 6.2, respectively (78).

In a prospectively collected observational study of 5 606 women, with vaginal births in 13 maternity hospitals in Nigeria and Uganda between 2014 and 2015, with spontaneous labor onset and normal perinatal outcomes, the results similarly

showed that average labor curves were different from those published by Friedman.

It was also concluded that using an average measure of tendency does not capture the true variations of women’s labor progression during the first stage of labor which could unintentionally misclassify women with normal labor progress as slow laborers

(27)

which may lead to unnecessary labor interventions (79). It has since then been evident that that the estimates from Zhang’s study are not universally applicable (47, 69, 77, 80). Additional important research evaluating duration of labor contributing to this standpoint are the systematic reviews preparing the evidence base for the new WHO recommendation on intrapartum care (2018) (2, 28, 33, 47, 64, 81).

These systematic reviews question the accuracy of the alert and action lines and conclude that the existing body of evidence does not support the 1cm/hour threshold for identifying those at risk of adverse maternal and neonatal outcomes.

Furthermore, they conclude that assessment of cervical dilation over time is a poor predictor for severe adverse birth outcomes and question the expectation of linear labor progress in all women. However, there has not yet been universal agreements on these conclusions and some criticism have been proposed against both the newer statistical approaches and findings (11, 13, 14, 75, 82-84).

Following the update of its global recommendations on intrapartum care in 2018 the WHO has initiated a revision of the use of a partograph. The WHO most recently launched a new tool “The next generation partograph” for documenting and

providing essential care during labor in the Labour care guide (85). It is not yet in use and more detailed plans for the implementation worldwide will take place during 2022 (86). It has been described as a shift from a more static tool to a more flexible dynamic tool which take into consideration the evidence-based time limits presented over the last decade. The most obvious differences to the WHO partograph are the absence of the diagonal action and alert lines and the fact that the designation for use during labor is a minimum of 5 cm cervical dilation (86, 87).

(28)

Figure 7. Labour care guide WHO, (LCG), launched 2020, not yet in use (85).

2.9 INTERVENTIONS USED DURING LABORS WITH SPONTANEOUS ONSET 2.9.1 Active management of labor

The “Active management of labor” (AOL) approach was introduced by O'Driscoll, Jackson (88) in 1969 to prevent long labors for low risk women. The original study, set in Dublin, Ireland, included 1000 nulliparous women with singleton pregnancies and term gestation in a setting with a low cesarean delivery (4%) rate and where the recommendation was that nulliparous women should not have labors lasting more

(29)

than 24 hours (88). With the AOL this recommendation was later reduced to 12 hours for nulliparous women (89). This package included a strict definition of labor dystocia, early amniotomy, regular examinations to evaluate labor progression and augmentation of labor with high doses of oxytocin and continuous one-to-one

support during labor (89).The intention with the package was to keep the total active labor duration <12 hours, which was considered a safe way to maintain a low

cesarean delivery rate (90). The first randomized trial was performed in 1995, randomizing 1934 nulliparous women to either AOL or standard care and showed a reduced labor duration for women allocated to AOL but no difference in cesarean delivery rate (19.5 vs 19.4 %) (91). More recently, a systematic review including 14 trials concluded that women receiving AOL will have a reduced mean duration (1.3 hours measured from admission to birth) of labor and a clinically modest reduction in cesarean delivery rate. It has been suggested that the most important factor in the AOL package was continuous care during labor and that early rupture of membranes and high doses of oxytocin may not be the contributing factors since the reduction in cesarean delivery seen in the early trial have not been replicated in subsequent studies. The concept of AOL was later adopted worldwide, in different forms, with indefinite implications on cesarean delivery rates (92, 93).

2.9.2 Amniotomy

Amniotomy is an artificial rupture of the membranes performed during labor under the condition that the cervix is more than 1 cm dilated in order to reach the

membranes. Amniotomy is often used routinely in contemporary clinical practice, aiming to prevent prolonged labor. In the Cochrane review by Smyth, Markham (94), analyzing spontaneous labor and comparing routine amniotomy with intact

membranes, no significant difference was found in first stage duration between the groups, but when stratified by parity, a non-significant shorter duration of 58 minutes was found for nulliparous women with a routine amniotomy. A number of randomized trials have reported a decrease in duration of labor but no increase in cesarean delivery rate (95-99). Likewise, it has been shown that a routine use of amniotomy for women with spontaneous labor and normal progress may reduce the use of oxytocin for augmentation (100).

All these studies have used the measure of mean duration of labor and have not evaluated the distribution of duration or median duration, which are considered the

(30)

2.9.3 Natural and synthetic Oxytocin

The natural hormone oxytocin is released during labor into the systemic circulation.

Largely undiscovered we do know that it stimulates the oxytocin receptors in the myometrium of the uterus. The oxytocin molecule also crosses the placenta (101).

Infusion with synthetic oxytocin is a common routine for augmentation of slow labor, and an accelerated trend in the use of oxytocin has been observed worldwide (25).

Synthetic oxytocin does not fully replicate biological oxytocin and long-term exposure to synthetic oxytocin may lead to reduced contractility of the uterine muscle (102). In Friedman´s cohort only 9% of the women had augmentation with oxytocin. Today, augmentation with oxytocin is a very common intervention (>50% of labors)

worldwide, and is used mainly to prevent slow labor, to put it differently; it is the only medicative drug used during labor to increase frequency and intensity of

contractions (70, 103). However, it is not clear to what extent it impacts first and second stage labor duration, since it is not dose-response steady on an individual level. Bugg showed that oxytocin decreased labor duration, in some women (104).

Conversely, in a prospective study evaluating first stage progression, Oladapo et al (79) found that excluding women with oxytocin from the cohort, slightly faster labors were noted.

Inconsistent high use of oxytocin has been reported by Bernitz et al (105) , who found that 42 % of women who received oxytocin to augment labor had not fulfilled the criteria for labor dystocia before oxytocin was started.

Similarly, Nysted et al (106) also found an inverse relationship between misuse of oxytocin in labors without a diagnosis of dystocia, however women with prolonged labor were also not augmented according to guidelines.

This paradox was also found in a Swedish cohort with 617 nulliparous women and 646 multiparous women, where augmentation with oxytocin was done inadequately and in a unstructured manner and where labor progress many times was

accelerated with oxytocin despite there not being a diagnosis of labor dystocia (57 % for nulliparous and 83% for multiparous) (62). More recently, the Norwegian LAPS trial found that when including women with a spontaneous start, term gestations, women who had been allocated to the WHO partograph group received more

augmentation with oxytocin compared to women randomized to Zhang’s labor curve guidelines. The study further found a higher rate of augmentation with oxytocin when a two hours’ action line was used compared to a four-hour action line (107). A

Cochrane review from 2013, evaluating the effect oxytocin may have on natural childbirth, concluded that early administration shortens labor duration, but does not contribute to a reduction in operative delivery (104). Despite its wide use, oxytocin augmentation has been associated with negative side-effect such as uterine hyperstimulation with insufficient placental circulation for the neonate. Hence, suboptimal use of oxytocin could cause adverse neonatal outcomes (108-113).

(31)

2.9.4 Epidural

Epidural is a type of anesthesia used to provide pain relief during labor. The

anesthesia is injected near the bottom of the spine and inhibits nerve pulse effect in the lower back, resulting in a decreased feeling in the lower parts of the body. The goal is to provide women pain relief without total numbness. There are different provider types of epidurals in use, the “heavy motor-blockade” epidural have an effect on body motor skills and it may become impossible to move the lower parts of the body, conversely the “stand up” epidural has no or little effect on body motor skills, with a dose-response effect for low and high doses. The anesthetics administered in the epidural pass into the circulation and cross the placenta, but duration of epidural use have not been associated with adverse effect on neonatal outcomes (114-116) . The impact of epidural use on the course of labor remains unclear, in observational studies the difference in labor duration seems to be

overestimated, but in RCTs most of these differences in duration mostly vanishes. It was concluded in a Cochrane review from 2018, with low evidence grading that it seems that epidural is associated with increase in assisted vaginal birth, however this is not seen in recent studies, after 2005 (117) (118).

Early administration in latent phase of labor (<3 cm dilation) have in most trials been compared to late administration with the outcome being cesarean delivery rather than labor duration (119). The estimated negative side effects of epidural use have been allocated to second stage of labor duration and mode of birth (117, 120-122).

Methodologically, studies differ in how they define onset of labor, and different types of epidurals are administrated in different settings, which highly explain the overall contradictive results. The effect of epidural use on labor duration is highly debated in clinical settings, and there is not enough body of evidence for evaluating its true effect on duration of first stage of labor. To clarify, any found association with use of epidural and higher frequency of cesarean delivery has been observed, it is however not clarified if it the epidural per se or a prolonged labor that is the driver for this association.

2.9.5 Cesarean delivery

Slow labor, labor dystocia, is the primary indication for primary cesarean delivery, and will also have an impact on future pregnancies since it will increase the risk of a

(32)

and obesity are known risk factors for labor dystocia and risk factors associated with cesarean delivery (106, 127). Different from many high-income countries, the

frequency of emergency cesarean deliveries in Sweden for nulliparous women has remained stable during the study period (Figure 8).

Figure 8. Frequency of emergency cesarean deliveries in Sweden during 2008-2019 in nulliparous women with spontaneous start, downloaded from Socialstyrelsens statistikdatabas.

2.9.6 Robson classification

The Robson classification was proposed by the WHO in 2015 to be used as a global standard for comparing caesarean delivery rates (128). In short, the core of the classification includes five maternal/neonatal/obstetric characteristics routinely collected in prenatal and delivery care; parity, number of fetuses, previous cesarean delivery, onset of labor, gestational age and fetal presentation. Women allocated to Robson 1 are nulliparous women, term gestation (>37 weeks) with a singleton fetus in vertex presentation (128, 129). In this project, comprising mainly nulliparous women, the Robson classification has been used in three of the studies to facilitate external interpretation of the results.

(33)

Figure 9. Robson classification system. Reproduced with permission from World Health Organization (128)

2.10 MATERNAL ANTHROPOMETRICS INFLUENCE ON THE COURSE OF LABOR

Both total duration and progression of labor is believed to be influenced by several factors including birth interval, parity, spontaneous start, gestational week, fetal presentation, position of the fetus, fetal weight and a combination of frequency and intensity of uterine contractions and maternal anthropometrics such as body mass

(34)

birth, labor induction, labor dystocia, instrumental deliveries including caesarean delivery, postpartum hemorrhage, and surgical wound infections (152-157). Despite these findings, defining labor arrest due to dystocia is identical across different BMI groups, and management during labor is standardized with a possible negative impact on the individual women. In addition, it is still unclear whether the distribution of the association between BMI and labor duration is similar or varies across

different maternal age groups (133, 138-140, 143, 155, 158-166).

Laughon, Branch (49) showed in a large retrospective comparison of nulliparous women in spontaneous labor, that today are childbearing women in the US older, have higher BMI, more often use epidural and oxytocin and more often experience prolonged labor duration and intrapartum cesarean delivery compared with women in the 1950-60´s. In Sweden, the proportion of women giving birth at the age 35 or above has increased considerably since 1973 from approximately 5% to 20% in 2019 and the mean age for having the first child has increased from 24 years of age to 29 years old over this period of time (167, 168). Taken together, women giving birth today differ widely compared to Friedman´s cohort, where women were on average 20 years old and only 1.8% gave birth by caesarean section (31).

Age and its impact on labors with spontaneous onset have been more sparsely studied and the definition of advanced maternal age has mostly been set at 35 years or older with studies using small cohorts with nulliparous women (133, 158, 164, 169, 170). In many settings there are guidelines for inducing labors for advanced age mothers, and most of the research have focused on outcomes for induced labors or differences in outcomes between spontaneous and induced labors (171- 173). Advanced maternal age is a risk factor for developing pregnancy-related complications and there is evidence to support an age-related decline of uterine performance with the possible consequence of increased risk of labor dystocia and cesarean delivery (135, 174). A decreased number of oxytocin receptors,

physiological changes in myometrial contractility have been proposed as biological causes of this decline in function. Another hypothesis is that postponing childbirth to an advanced age will have impact on uterine contractility since a long use of

contraceptives may cause an artificial hormone stimulation of the myometrium (175).

It has also been proposed that the frequency of all age-related co-morbidity and absolute risks are underestimated in relation to all birth outcomes. This may in part be explained by lack of consistency and insufficient study design to capture the biological joint effects of aging maternal anthropometrics (169, 176).

(35)

3 RESEARCH AIMS

The overall aim was to establish updated and evidence-based measures of normal labor duration in labors with spontaneous onset.

3.1 SPECIFIC AIMS:

Study I To describe duration, progression and pattern of first stage of labor in different groups of women.

Study II To investigate the total effect of early-pregnancy BMI on labor duration, and examine if the association between BMI and duration is modified by age.

Study III To investigate the association between active first stage labor duration and second stage labor duration to address the continuous course of labor in relation to time.

Study IV To investigate if increasing active first stage labor duration is associated with adverse neonatal outcomes

(36)
(37)

4 MATERIALS AND METHODS

4.1 SETTING

Sweden has a long tradition of collecting rigorous information on its citizens,

including data from the school system, tax system and medical records, Health care and welfare are tax-financed, enabling equity in access to health care during the course of life. Pregnant women have had access to standardized and free basic antenatal and delivery care for almost 60 years, covering 98% of the pregnant population and 99 % of births occur in a hospital setting. Normal pregnancy and normal labors are cared for by certified midwifes. Pregnancies and labors deviating from normality are cared for by midwifes in close collaboration with obstetricians.

Routinely used ultrasound scans are offered all women before gestational week 20 to confirm gestational age and also include screening for malformations and multiple fetuses. The use of a unique personal identification number, introduced in 1947, is the key element for this documentation of citizens, a newborn is given this personal identification number directly after birth, provided by the Swedish citizen registration office by linkage to the medical charts of the mother. Citizens living in Sweden for longer than one year after migration are also equipped with a unique personal identification number (177).

4.2 DATA SOURCES

The Stockholm-Gotland Obstetric Cohort is used in all four studies. For

completeness of data on infant outcome, the Stockholm-Gotland Obstetric Cohort has been linked with the Swedish Neonatal Quality Register (SNQ) containing specified information on infant diagnoses from neonatal care units.

4.2.1 The Stockholm-Gotland Obstetric cohort

The first version of this population-based cohort includes all pregnancies and all births between January 1st 2008 and October 31st 2014 in the regions of Stockholm and Gotland. Annually, approximately 25 000 -30 000 deliveries take place in these two regions, accounting for about 25 % of all deliveries in Sweden. For the register, data is collected from seven different hospitals in these two regions and during this time period consisted of 175 522 singleton deliveries. All pregnancy- and labor- related data is collected and forwarded from the electronic medical records system

(38)

birthweight. The database was established at the Clinical Epidemiology Division at Karolinska Institutet and funded by the Swedish Research Council. During 2020 the second version of this cohort was created. The new version consists of data on births occurring between 1st November 2014 to 30st and June 2020, including a total of 335 153 unique pregnancies. Both the first and the second version of this cohort are universally superior owing to their unique size, completeness on clinically relevant variables, including date and time data on cervical dilatation for each woman. To date, several studies based on the Stockholm-Gotland Obstetric Database have been published, a formal validation has not yet been performed.

Both versions of the Stockholm Gotland Obstetric Cohort have permanent linkages to other quality registers.

4.2.2 The Swedish Neonatal Quality Register

The Stockholm Gotland Obstetric Cohort has been linked to the SNQ. This register includes information about the mother, pregnancy complications and delivery characteristics and information about admission to neonatal care, diagnoses and interventions performed in direct relation to the birth of the neonate. Any level of neonatal care within the first 28 days after birth is found in the register that is nationwide and, since 2012, covers all neonatal units in Sweden. The register was started in 2001 and have been validated with excellent agreement and completeness (178).

4.3 STUDY POPULATION AND STUDY DESIGNS 4.3.1 Overview

The different aims of the four studies along with updates of the database required different cohorts to suit the different objectives. In all the studies, the aim was to evaluate labors with spontaneous onset, herby the Target population for all studies are all women in the Stockholm Gotland Obstetric Cohort were a spontaneous onset can be identified. When studying labors and subsequent birth outcomes,

stratification on parity is necessary. Hence, Target population identification is described for all four studies and all exposures, outcomes, and covariates have all been collected from the databases previously described. An overview of cohort selection for each study is presented in Table 2.

(39)

Study and Aims

I

To describe duration, progression and pattern of first stage of labor

Study population Exposure Outcome

Data from Stockholm-Gotland Obstetric Cohort 2008–2014.

Included: 85 408 women, term gestations (<37+0-41+6 weeks), singleton, vertex, live fetuses

experiencing spontaneous labor onset and vaginal delivery with normal neonatal outcomes.

Excluded: Induction of labor or spontaneous onset with cesarean delivery, or previous cesarean delivery and women with less than two cervical examinations during first stage and/or had records in which timing of the complete dilation of cervix was missing.

Also excluded where when neonates had an Apgar score less than 7 at 5 minutes of age or any of the following:

hypoxic ischæmic encephalopathy, convulsion and meconium aspiration.

Descriptive cohort study.

Pattern of labor progression and duration was

calculated only for descriptive purposes.

Cumulative labor duration and traverse time, during first stage of labor.

II

To investigate the total effect of early- pregnancy BMI on labor duration, and examine if the association between BMI and duration is modified by age.

Data from Stockholm-Gotland Obstetric Cohort 2008–2014.

Included: 13 794 women. Women allocated to Robson group 1:

nulliparous, term gestations

(≤37+0 weeks) with vertex, singleton pregnancies who experienced

spontaneous labor onset and with a live fetus.

Excluded: Women with missing notation of the start point or endpoint of the outcome or of the exposure (BMI).

Women with caesarean delivery during first stage, since the endpoint of the

Early pregnancy weight status, BMI, categorized into 3 different weight groups, normal weight (<25), overweight (25-29.9) and obesity (30)

Primary outcome:

Duration of active first stage of labor at different quantile levels Secondary outcome: Total active labor duration until birth at different quantile levels

(40)

Table 2. Overview of aim, study population, exposure and outcome for each of the four studies

labor duration and second stage of labor duration

pregnancies who experienced

spontaneous labor onset and with a live fetus.

Excluded: Women with missing notation of start the exposure or end of the exposure. Women with caesarean delivery during first stage since the end of the exposure could not be identified.

Categorized exposure:

active first stage of labor duration, categorized into 4 different quartiles, Q1-Q4, where Q1 was used as the

reference.

Secondary outcome:

Mode of delivery

IV

To investigate if there is an association between first stage labor duration and adverse neonatal outcome

Data from Stockholm-Gotland Obstetric Cohort 2008–2020.

Included: The final study cohort consists of 46 040 women. Women allocated to Robson group 1: nulliparous, term gestations (≤37+0 weeks) with vertex, singleton pregnancies who experienced spontaneous labor onset and with a live fetus. To address missing data on cervical dilation of 5 cm (start of the exposure) single imputation models were performed on women with informed data on a minimum of two notations of cervical dilation during first stage of labor. Ten different imputation patterns were generated to define start of the exposure, i.e. when each woman with missing data in the partograph on 5 cm was estimated to have been dilated 5 cm respectively, this was done to reduce selection bias and improve precision

Excluded: Women with missing

identified notation of start the exposure or end of the exposure.

Continuous exposure:

Categorized into 4 levels of duration, C1-C4, where C1 was used as the

reference.

Adverse neonatal outcome, categorized into a

composite of severe or moderate outcome based on severity grading.

(41)

4.3.2 Overview and detailed description for the definition of active first stage of labor, used as both an outcome and exposure of interest in this thesis

Study I

In Study I the start of the active first stage of labor was defined by two strict criterions:

✓ time characterized by well effaced cervix (complete or nearly 100%

effacement), cervical dilation of 3 centimeter in presence of regular painful uterine contractions

✓ time characterized by well effaced cervix (complete or nearly 100%

effacement), a rupture of membrane in combination with cervical dilation of 3 centimeter.

Following this hierarchy of the two criterions, when there were two or three time points fulfilled in one criterion, the second time point was chosen as the starting point of first stage of labor. If a woman did not fulfill all the parameters of criterion one, she was evaluated according to criterion two. Women who did not fulfill criterion one or two were not included in the study cohort since the start of the active first stage of labor could not be clearly defined. The end of the active first stage was defined by the time-point that the cervix was fully dilated. The estimated median duration of the first stage of labor was extrapolated from the cervical examination on admission and subsequent examinations performed during labor documented in the partograph.

Labor duration prior to hospital admission was not considered in this study.

Study II

In Study II start of the active first stage of labor was defined by four strict criterions for customizing the start of the active first stage for each woman.

✓ cervical efacement, dilation of 3 to 4 cm and painful contractions at admission;

✓ cervical efacement, dilation of 3 to 4 cm and spontaneous rupture of

(42)

If a woman did not fulfill all the parameters of criterion one, she was evaluated according to criterion two, three or four. Women who did not fulfill any criterion were not included in the study cohort since the start of the active first stage of labor could not be clearly defined. The end of the active first stage was defined by the time-point that the cervix was fully dilated. The estimated median duration of the first stage of labor was extrapolated from the cervical examination on admission and subsequent examinations performed during labor documented in the partograph. Labor duration prior to hospital admission was not considered in this study.

Study III

To identify the correct timepoint for customizing the start of the exposure for each woman, the onset of active first stage was defined to start when the cervix was completely or nearly 100% effaced along with one of the following four criteria:

✓ dilation 3–4 cm with painful contractions

✓ dilation 3–4 cm with spontaneous rupture of membranes

✓ dilation 4 cm with painful and regular contractions

✓ dilation 5 cm with painful and regular contractions Study IV

Following new recommendations from WHO (27) we updated the previous

definitions used in this thesis. Here, to identify the correct timepoint for start of the active first stage of labor, we customized the start to each woman, identified by one data point:

✓ the first timepoint of a cervical dilation of 5 cm recorded in the partograph.

Timepoint of cervical dilation of 5 cm was not available in a large proportion of women. To minimize potential risk of selection bias and improve precision, 10 different imputation patterns were generated to define start of the active first stage, we estimated when each woman with missing information in the partograph on 5 cm was assumed to have been dilated 5 cm. Each individual where we had recorded data on cervical dilation and a timepoint before 5 cm and after 5 cm were subjected to imputation. For example, 42.5 % (N= 6482) of the imputed start of the exposure (5 cm dilation) was among women with known data on time for dilation 4 cm and 6 cm.

Estimations was calculated upon previous knowledge of the populations traverse time to progress from one centimeter to full dilation, with the time trajectory weighted by a population-based percentage, this is described more in detail under

Methodological considerations. The end of first stage was defined by the timepoint of the first notion of full cervical dilation.

(43)

4.4 STATISTICAL OVERVIEW

A brief description of methods for each study will follow. Detailed descriptions of methods and materials are provided for each study in the original articles at the end of the thesis and are discussed in detail under Methodological considerations.

4.4.1 Study I

A population-based cohort study including both nulliparous and multiparous women with spontaneous labor onset, Table 2 shows the study population in detail. The final study cohort comprised 85 408 women, stratified into parity groups. For sensitivity analysis, we also created a target population of 101 730 women, expanding the study cohort with women with a a caesarean delivery during second stage or

adverse neonatal outcomes and performed the same analysis to investigate if there were differences in progression patterns between the cohorts.

Interval censored regression was used to calculate traverse time (in hours) to progress from one centimeter to the next during the first stage of labor. Cervical dilation was the only covariate in the model. The outcome measure was estimated and presented for the 5th, 50th and 95th percentiles.

Labor dilation curves were created to display labor patterns for women. These curves were based on the traverse time calculated and modelled by using restricted cubic splines with three knots. When constructing cervical dilation curves, time zero was considered to be the timepoint for a fully retracted cervix. Cumulative labor duration stratified by cervical dilation at admission was done using analogous

interval-censored regression approach, same as used for the traverse time, in these analyses, time zero was considered to be the cervical dilation at admission. The cervical dilation at admission could be 3, 4, 5 or 6 cm. The method was done identical to the approach for traverse time estimations.

4.4.2 Study II

A population-based cohort study including nulliparous women with spontaneous labor onset, Table 2 shows the study population in detail. The final study cohort comprised 13 794 women. The Robson Ten-Group Classification system was used

References

Related documents

Figure 1 shows simulated responses to a birth in year 1 for the simple probit model, random effects MSL probit model, AR(1) probit model, and dynamic probit with first order state

This paper investigates the effect of right to full-time policies implemented to decrease involuntary part-time work for public care workers employed by Swedish municipalities..

Hence, given that individuals with substantial shares of their total income from capital have kept their share of realized capital gains, at the same time as the share

The VoC literature is thought to open up “substantial new perspectives on both economic and social policymaking,” with proponents hypothesizing that institutional change and

This paper reframes the labor question according to the normal juridical principle of imputation whose application to property appropriation is the modern treatment of the old

Without international capital mobility, overall gains from migration in the immigration region are only a small fraction of total losses to native workers, but with

The follow- ing variables were not associated with the duration of any breastfeeding in the linear regression analysis on univariate level: mother’s country of birth, partner’s

ABSTRACT: Introduction: Most women who give birth for the first time experience some form of perineal trauma. Second-degree tears contribute to long-term consequences for women and are